E. Albert Reece

How many of us are thoroughly, truly prepared to manage shoulder dystocia, to use forceps, or to perform a vaginal breech delivery?

It is not a silly question to ask ourselves, since these are critical, high-risk situations and skills that most of us do not encounter as frequently as routine vaginal delivery. When we are not practicing critical skills, we tend to lose them.

The question is increasingly important, moreover, because technological advances are making obstetric simulation more feasible and affordable for a variety of different settings. Realistic, low-fidelity mannequins that take up little space in an office or an old exam room are now relatively inexpensive.

Obstetric simulation training has become a tool that we simply must take advantage of. It is not only making its way into academia, with a small but growing body of literature showing that it improves competence and performance when a real event occurs, but is also gaining acceptance among practicing physicians as a valuable means of practicing skills and preparing for obstetric emergencies, such as shoulder dystocia, breech vaginal delivery, postpartum hemorrhage, and eclampsia.

I see and hear about attending physicians who join residents in the growing number of simulation programs that exist in academic institutions because they realize that they, too, can benefit from the practice. The American College of Obstetricians and Gynecologists is watching this trend; its Task Force on Simulation is examining the role of simulation in obstetrics and ways in which practicing obstetricians can take advantage of simulation technology.

Some professional liability organizations, meanwhile, are considering giving physicians discounts on their malpractice insurance premiums if they practice simulation; Harvard-affiliated obstetricians have been offered such discounts, and Kaiser Permanente is implementing simulation programs (including birth simulation training) as part of its initiatives for quality and patient safety.

It seems only a matter of time before more health care institutions draw on obstetric simulation to help practicing physicians update and reinforce their skills, and before certifying bodies also embrace the notion. (General surgery is on the cusp of establishing simulation centers for certification and recertification.)

In the meantime, obstetricians can take it upon themselves to use available technology and prepare for the high-acuity, low-frequency emergencies that are encountered by every obstetrician at some time.

Safety, Liability

That obstetric simulation is on the radar screen—and probably on its way to becoming mainstream—makes perfect sense.

Professionals in the airline industry, the military, and the nuclear power industry are already using simulation for teaching and for maintaining and evaluating skills. Simulation is a safety-first tool in these industries, and it often utilizes evidence-based protocols.

In medicine, we make evidence-based decisions all the time, and patient safety is a huge issue. The Joint Commission on Accreditation of Healthcare Organizations recently looked at all perinatal sentinel events across the country in all types of institutions, and found that 47% were linked with staff competence issues. Among the other identified root causes were communication issues (72%), the orientation and training process (40%), and organization culture as a barrier to effective communication and teamwork (55%). Simulation could play a significant role in addressing each of these issues.

Shoulder dystocia complicates up to 2% of all vaginal deliveries, and potentially causes permanent brachial plexus injury, clavicular fracture, hypoxic brain injury, and other significant long-term complications. Although we encounter shoulder dystocia infrequently, the risk for serious and permanent injury to the infant is so high that we ought to be prepared.

Similarly, approximately 3%–4% of singleton babies are in the breech position, yet only a minority of obstetricians are able to perform vaginal breech deliveries. In one recent study, only 33% of surveyed attending physicians performed vaginal breech deliveries. The rest do not do them anymore.

Although vaginal breech deliveries are discouraged, vaginal delivery is sometimes unavoidable or even preferable. (When the breech is on the perineum, for instance, it's riskier to go to cesarean delivery). We are putting our patients and ourselves at risk by not practicing and knowing how to do this with proper technique,

We do not like to talk about the litigation aspect, but we cannot hide it: Fetal injury that is related to emergencies like shoulder dystocia is a potential source of medical malpractice lawsuits and one that we can minimize by reinforcing and maintaining our skills through simulation.

Today's Simulators

Obstetricians worry about how they can do a simulation. Many think of simulators as too big, too expensive, and not lifelike. Some worry about doing a simulation in front of others and are too intimidated to try.

Some of the simulators available today are expensive. A full-body, high-fidelity obstetric simulator with all the bells and whistles—touch-screen computer technology that enables manipulation of the labor course, for instance, and varying vital signs and fetal heart rhythms—can cost up to $40,000.

These expensive models are often purchased by academic institutions that are interested in simulation for a multitude of purposes, including team training, but such models are not necessary to simulate at least several obstetric emergencies, including vaginal breech delivery, shoulder dystocia, and the use of forceps.

For these situations, low-fidelity simulators—which may be just a model of the pelvis through which a model baby can be manually pushed—are perfectly fine. They can be purchased for $2,000-$3,000, stored in a closet, and placed in an extra exam room where physicians can practice, either with a mentor or expert or by themselves.

Nothing is as real as a true patient or a real-life situation, of course, but many of these mannequins are surprisingly lifelike, with features like an anatomically correct bony pelvis, a stretchable perineum, and a silicone pelvic-floor musculature. A mannequin's cervix, for instance, really feels like a cervix.

When I was in resident training, I practiced using the forceps on a high-fidelity mannequin. This gave me an opportunity to practice all the necessary maneuvers and to know whether I performed all critical tasks, from inserting the posterior blade first, for instance, to holding the left blade of the forceps with my left hand while using my right hand as a guide.

Later, when I was in a real and urgent situation requiring forceps, I knew just what to do. It worked like clockwork. Simulation on a low-fidelity mannequin, if that was what my institution had had, would have been just as beneficial.

Simulation also provides opportunities to create protocols. In the middle of a forceps delivery simulation, for instance, you may realize that “this needs to be done all the time just like this.” Alternatively, you may think, “Let's not do it this way next time.”

Similarly, simulation affords us opportunities to practice and fine-tune communication and teamwork.

Improved Competence

I recently oversaw a resident who had previously done simulation training with high-fidelity mannequins as part of her curriculum at the Washington Hospital Center, and was now in a real and difficult delivery involving shoulder dystocia.

She performed the recommended initial maneuvers—like placing the patient in the McRobert's position and applying suprapubic pressure—but without success. She then immediately proceeded, without any prompting, to deliver the posterior arm, which relieved the shoulder dystocia. Afterward, the resident told me that “if I hadn't done the shoulder dystocia simulation lab, I would not have known to do that.” I hear such stories often.

Studies are beginning to document the effects of obstetric simulation training on competence and performance.

In a study published several years ago, for instance, residents at Georgetown University in Washington and the Uniformed Services University of Health Sciences in Bethesda, Md., were randomized to receive training on shoulder dystocia management using a high-fidelity obstetric simulator or to receive no special training. Each resident was subsequently tested without prior notice in another simulation scenario.

Those who had practiced shoulder dystocia management on mannequins completed more critical tasks and had significantly higher scores on timeliness of their interactions, proper performance of maneuvers, and overall performance (Obstet. Gynecol. 2004:103;1224–8).

Although not randomized, another more recent study at Georgetown University showed that high-fidelity simulation training improved resident performance of vaginal breech delivery. Residents were more likely after simulation training to perform critical maneuvers correctly and to deliver in a safe manner than they were before the training (Obstet. Gynecol. 2006:107;86–9).

Research from the University of Bristol (England) is also yielding interesting results. Investigators there have reported, for instance, that obstetric emergency training courses using simulation were associated with a significant reduction in low 5-minute APGAR scores and lower rates of hypoxic-ischemic encephalopathy (BJOG 2006;113:177–82).

Another study of shoulder dystocia has shown that, whereas training with high-fidelity mannequins provides additional benefits, training with low-fidelity mannequins is also effective in improving management of the obstetric situation by obstetricians and midwives (Obstet. Gynecol. 2006;108:1477–85).

A study from the Bristol investigators in which participants were tested on a standardized simulation before a simulation workshop, and then at 3 weeks, 6 months, and 12 months afterward, shows that improved performance appears to be sustained. Those who were proficient 3 weeks after the training retained their skills at the later dates. The researchers concluded that annual training may be adequate for some physicians, whereas others may need more frequent practice (Obstet. Gynecol. 2007;110:1069–74).

Soon-to-be-published research that we have recently completed at Georgetown University and the Washington Hospital Center similarly indicates that obstetricians generally should strive for continuing simulation training at least once a year. Residents in our study who were initially taught on the simulator scored higher when tested a year later than did residents who received no simulation training. Overall, however, everyone's scores declined.

Obstetric simulation is part of our future. New physicians of the future will enter practice having done simulation training in a variety of high-acuity, low-frequency scenarios—rather than learning solely through lectures and impromptu teaching after events have occurred—and those of us already in practice will likely find that working occasionally with low-fidelity mannequins enables us to provide better, safer patient care while reducing our liability risk.

Dr. Marsha Solomon, chief resident at the Washington Hospital Center, is shown performing a simulated forceps delivery in the photo at left. In the photo at right, Dr. Solomon performs a simulated breech vaginal delivery. Photos courtesy Dr. Tamika C. Auguste

Obstetric Simulation

Do you think that you would like to have your next airplane flight piloted by someone who has not flown a plane in several years or who has little experience in landing?

Pilots are among the professionals who gain their greatest experience and expertise through the development of skills using various simulation technologies.

Simulation training is common in the aviation industry, as it is in aeronautics and in some branches of engineering, which makes this question significantly less worrisome and less relevant than if simulation were not common.

Medicine in general—and obstetrics in particular—has been practiced worldwide using the apprenticeship model, in which residents and interns work with attending physicians to learn the art of medicine.

While caring for patients with various disorders and in various scenarios, physicians-in-training work alongside the more senior practitioners, taking on progressive amounts of responsibility. Experience is gained accordingly.

This approach has been very successful over the years, and will remain so. It may be enhanced, however, as the simulation approach is slowly integrated into medicine and into obstetrics training.

The use of simulation training in medicine makes intuitive sense. The acquisition of the greatest possible skill or expertise—or the enhancement of skills if there is a hiatus in practice—makes sense from quality-of-care and patient-safety perspectives, and also because of the litigious environment in which we live and practice.

The question, “How would you like to have your baby delivered by an obstetrician who has not used forceps or managed shoulder dystocia in over a year?” is a valid one for patients who realize that less-common delivery scenarios are unpredictable.

This month's Master Class will focus on the utility, practicability, and application of simulation technology in obstetrics as a means of maximizing not only the skills of the resident, but also the skills of the practicing clinician.

Our guest author, Dr. Tamika C. Auguste, is the director of obstetric simulation at Washington Hospital Center and assistant professor of obstetrics/gynecology at Georgetown University in Washington. She speaks in various forums on the issue of simulation for both residents and practicing physicians, and is fast becoming a young leader and expert from whom we can expect to hear more in the future.

Key Points on Simulation

1. Simulation can be used to practice classic obstetric skills and high-risk, low-frequency obstetric emergencies.

2. Simulation is not only for those in academic medicine but also for those in private practice.

3. Low-fidelity simulators can be just as useful as high-fidelity simulators.

4. Simulation is becoming the norm in residency training programs.

How many of us are thoroughly, truly prepared to manage shoulder dystocia, to use forceps, or to perform a vaginal breech delivery?

It is not a silly question to ask ourselves, since these are critical, high-risk situations and skills that most of us do not encounter as frequently as routine vaginal delivery. When we are not practicing critical skills, we tend to lose them.

The question is increasingly important, moreover, because technological advances are making obstetric simulation more feasible and affordable for a variety of different settings. Realistic, low-fidelity mannequins that take up little space in an office or an old exam room are now relatively inexpensive.

Obstetric simulation training has become a tool that we simply must take advantage of. It is not only making its way into academia, with a small but growing body of literature showing that it improves competence and performance when a real event occurs, but is also gaining acceptance among practicing physicians as a valuable means of practicing skills and preparing for obstetric emergencies, such as shoulder dystocia, breech vaginal delivery, postpartum hemorrhage, and eclampsia.

I see and hear about attending physicians who join residents in the growing number of simulation programs that exist in academic institutions because they realize that they, too, can benefit from the practice. The American College of Obstetricians and Gynecologists is watching this trend; its Task Force on Simulation is examining the role of simulation in obstetrics and ways in which practicing obstetricians can take advantage of simulation technology.

Some professional liability organizations, meanwhile, are considering giving physicians discounts on their malpractice insurance premiums if they practice simulation; Harvard-affiliated obstetricians have been offered such discounts, and Kaiser Permanente is implementing simulation programs (including birth simulation training) as part of its initiatives for quality and patient safety.

It seems only a matter of time before more health care institutions draw on obstetric simulation to help practicing physicians update and reinforce their skills, and before certifying bodies also embrace the notion. (General surgery is on the cusp of establishing simulation centers for certification and recertification.)

In the meantime, obstetricians can take it upon themselves to use available technology and prepare for the high-acuity, low-frequency emergencies that are encountered by every obstetrician at some time.

Safety, Liability

That obstetric simulation is on the radar screen—and probably on its way to becoming mainstream—makes perfect sense.

Professionals in the airline industry, the military, and the nuclear power industry are already using simulation for teaching and for maintaining and evaluating skills. Simulation is a safety-first tool in these industries, and it often utilizes evidence-based protocols.

In medicine, we make evidence-based decisions all the time, and patient safety is a huge issue. The Joint Commission on Accreditation of Healthcare Organizations recently looked at all perinatal sentinel events across the country in all types of institutions, and found that 47% were linked with staff competence issues. Among the other identified root causes were communication issues (72%), the orientation and training process (40%), and organization culture as a barrier to effective communication and teamwork (55%). Simulation could play a significant role in addressing each of these issues.

Shoulder dystocia complicates up to 2% of all vaginal deliveries, and potentially causes permanent brachial plexus injury, clavicular fracture, hypoxic brain injury, and other significant long-term complications. Although we encounter shoulder dystocia infrequently, the risk for serious and permanent injury to the infant is so high that we ought to be prepared.

Similarly, approximately 3%–4% of singleton babies are in the breech position, yet only a minority of obstetricians are able to perform vaginal breech deliveries. In one recent study, only 33% of surveyed attending physicians performed vaginal breech deliveries. The rest do not do them anymore.

Although vaginal breech deliveries are discouraged, vaginal delivery is sometimes unavoidable or even preferable. (When the breech is on the perineum, for instance, it's riskier to go to cesarean delivery). We are putting our patients and ourselves at risk by not practicing and knowing how to do this with proper technique,

We do not like to talk about the litigation aspect, but we cannot hide it: Fetal injury that is related to emergencies like shoulder dystocia is a potential source of medical malpractice lawsuits and one that we can minimize by reinforcing and maintaining our skills through simulation.

Today's Simulators

Obstetricians worry about how they can do a simulation. Many think of simulators as too big, too expensive, and not lifelike. Some worry about doing a simulation in front of others and are too intimidated to try.

Some of the simulators available today are expensive. A full-body, high-fidelity obstetric simulator with all the bells and whistles—touch-screen computer technology that enables manipulation of the labor course, for instance, and varying vital signs and fetal heart rhythms—can cost up to $40,000.

These expensive models are often purchased by academic institutions that are interested in simulation for a multitude of purposes, including team training, but such models are not necessary to simulate at least several obstetric emergencies, including vaginal breech delivery, shoulder dystocia, and the use of forceps.

For these situations, low-fidelity simulators—which may be just a model of the pelvis through which a model baby can be manually pushed—are perfectly fine. They can be purchased for $2,000-$3,000, stored in a closet, and placed in an extra exam room where physicians can practice, either with a mentor or expert or by themselves.

Nothing is as real as a true patient or a real-life situation, of course, but many of these mannequins are surprisingly lifelike, with features like an anatomically correct bony pelvis, a stretchable perineum, and a silicone pelvic-floor musculature. A mannequin's cervix, for instance, really feels like a cervix.

When I was in resident training, I practiced using the forceps on a high-fidelity mannequin. This gave me an opportunity to practice all the necessary maneuvers and to know whether I performed all critical tasks, from inserting the posterior blade first, for instance, to holding the left blade of the forceps with my left hand while using my right hand as a guide.

Later, when I was in a real and urgent situation requiring forceps, I knew just what to do. It worked like clockwork. Simulation on a low-fidelity mannequin, if that was what my institution had had, would have been just as beneficial.

Simulation also provides opportunities to create protocols. In the middle of a forceps delivery simulation, for instance, you may realize that “this needs to be done all the time just like this.” Alternatively, you may think, “Let's not do it this way next time.”

Similarly, simulation affords us opportunities to practice and fine-tune communication and teamwork.

Improved Competence

I recently oversaw a resident who had previously done simulation training with high-fidelity mannequins as part of her curriculum at the Washington Hospital Center, and was now in a real and difficult delivery involving shoulder dystocia.

She performed the recommended initial maneuvers—like placing the patient in the McRobert's position and applying suprapubic pressure—but without success. She then immediately proceeded, without any prompting, to deliver the posterior arm, which relieved the shoulder dystocia. Afterward, the resident told me that “if I hadn't done the shoulder dystocia simulation lab, I would not have known to do that.” I hear such stories often.

Studies are beginning to document the effects of obstetric simulation training on competence and performance.

In a study published several years ago, for instance, residents at Georgetown University in Washington and the Uniformed Services University of Health Sciences in Bethesda, Md., were randomized to receive training on shoulder dystocia management using a high-fidelity obstetric simulator or to receive no special training. Each resident was subsequently tested without prior notice in another simulation scenario.

Those who had practiced shoulder dystocia management on mannequins completed more critical tasks and had significantly higher scores on timeliness of their interactions, proper performance of maneuvers, and overall performance (Obstet. Gynecol. 2004:103;1224–8).

Although not randomized, another more recent study at Georgetown University showed that high-fidelity simulation training improved resident performance of vaginal breech delivery. Residents were more likely after simulation training to perform critical maneuvers correctly and to deliver in a safe manner than they were before the training (Obstet. Gynecol. 2006:107;86–9).

Research from the University of Bristol (England) is also yielding interesting results. Investigators there have reported, for instance, that obstetric emergency training courses using simulation were associated with a significant reduction in low 5-minute APGAR scores and lower rates of hypoxic-ischemic encephalopathy (BJOG 2006;113:177–82).

Another study of shoulder dystocia has shown that, whereas training with high-fidelity mannequins provides additional benefits, training with low-fidelity mannequins is also effective in improving management of the obstetric situation by obstetricians and midwives (Obstet. Gynecol. 2006;108:1477–85).

A study from the Bristol investigators in which participants were tested on a standardized simulation before a simulation workshop, and then at 3 weeks, 6 months, and 12 months afterward, shows that improved performance appears to be sustained. Those who were proficient 3 weeks after the training retained their skills at the later dates. The researchers concluded that annual training may be adequate for some physicians, whereas others may need more frequent practice (Obstet. Gynecol. 2007;110:1069–74).

Soon-to-be-published research that we have recently completed at Georgetown University and the Washington Hospital Center similarly indicates that obstetricians generally should strive for continuing simulation training at least once a year. Residents in our study who were initially taught on the simulator scored higher when tested a year later than did residents who received no simulation training. Overall, however, everyone's scores declined.

Obstetric simulation is part of our future. New physicians of the future will enter practice having done simulation training in a variety of high-acuity, low-frequency scenarios—rather than learning solely through lectures and impromptu teaching after events have occurred—and those of us already in practice will likely find that working occasionally with low-fidelity mannequins enables us to provide better, safer patient care while reducing our liability risk.

Dr. Marsha Solomon, chief resident at the Washington Hospital Center, is shown performing a simulated forceps delivery in the photo at left. In the photo at right, Dr. Solomon performs a simulated breech vaginal delivery. Photos courtesy Dr. Tamika C. Auguste

Obstetric Simulation

Do you think that you would like to have your next airplane flight piloted by someone who has not flown a plane in several years or who has little experience in landing?

Pilots are among the professionals who gain their greatest experience and expertise through the development of skills using various simulation technologies.

Simulation training is common in the aviation industry, as it is in aeronautics and in some branches of engineering, which makes this question significantly less worrisome and less relevant than if simulation were not common.

Medicine in general—and obstetrics in particular—has been practiced worldwide using the apprenticeship model, in which residents and interns work with attending physicians to learn the art of medicine.

While caring for patients with various disorders and in various scenarios, physicians-in-training work alongside the more senior practitioners, taking on progressive amounts of responsibility. Experience is gained accordingly.

This approach has been very successful over the years, and will remain so. It may be enhanced, however, as the simulation approach is slowly integrated into medicine and into obstetrics training.

The use of simulation training in medicine makes intuitive sense. The acquisition of the greatest possible skill or expertise—or the enhancement of skills if there is a hiatus in practice—makes sense from quality-of-care and patient-safety perspectives, and also because of the litigious environment in which we live and practice.

The question, “How would you like to have your baby delivered by an obstetrician who has not used forceps or managed shoulder dystocia in over a year?” is a valid one for patients who realize that less-common delivery scenarios are unpredictable.

This month's Master Class will focus on the utility, practicability, and application of simulation technology in obstetrics as a means of maximizing not only the skills of the resident, but also the skills of the practicing clinician.

Our guest author, Dr. Tamika C. Auguste, is the director of obstetric simulation at Washington Hospital Center and assistant professor of obstetrics/gynecology at Georgetown University in Washington. She speaks in various forums on the issue of simulation for both residents and practicing physicians, and is fast becoming a young leader and expert from whom we can expect to hear more in the future.

Key Points on Simulation

1. Simulation can be used to practice classic obstetric skills and high-risk, low-frequency obstetric emergencies.

2. Simulation is not only for those in academic medicine but also for those in private practice.

3. Low-fidelity simulators can be just as useful as high-fidelity simulators.

4. Simulation is becoming the norm in residency training programs.

How many of us are thoroughly, truly prepared to manage shoulder dystocia, to use forceps, or to perform a vaginal breech delivery?

It is not a silly question to ask ourselves, since these are critical, high-risk situations and skills that most of us do not encounter as frequently as routine vaginal delivery. When we are not practicing critical skills, we tend to lose them.

The question is increasingly important, moreover, because technological advances are making obstetric simulation more feasible and affordable for a variety of different settings. Realistic, low-fidelity mannequins that take up little space in an office or an old exam room are now relatively inexpensive.

Obstetric simulation training has become a tool that we simply must take advantage of. It is not only making its way into academia, with a small but growing body of literature showing that it improves competence and performance when a real event occurs, but is also gaining acceptance among practicing physicians as a valuable means of practicing skills and preparing for obstetric emergencies, such as shoulder dystocia, breech vaginal delivery, postpartum hemorrhage, and eclampsia.

I see and hear about attending physicians who join residents in the growing number of simulation programs that exist in academic institutions because they realize that they, too, can benefit from the practice. The American College of Obstetricians and Gynecologists is watching this trend; its Task Force on Simulation is examining the role of simulation in obstetrics and ways in which practicing obstetricians can take advantage of simulation technology.

Some professional liability organizations, meanwhile, are considering giving physicians discounts on their malpractice insurance premiums if they practice simulation; Harvard-affiliated obstetricians have been offered such discounts, and Kaiser Permanente is implementing simulation programs (including birth simulation training) as part of its initiatives for quality and patient safety.

It seems only a matter of time before more health care institutions draw on obstetric simulation to help practicing physicians update and reinforce their skills, and before certifying bodies also embrace the notion. (General surgery is on the cusp of establishing simulation centers for certification and recertification.)

In the meantime, obstetricians can take it upon themselves to use available technology and prepare for the high-acuity, low-frequency emergencies that are encountered by every obstetrician at some time.

Safety, Liability

That obstetric simulation is on the radar screen—and probably on its way to becoming mainstream—makes perfect sense.

Professionals in the airline industry, the military, and the nuclear power industry are already using simulation for teaching and for maintaining and evaluating skills. Simulation is a safety-first tool in these industries, and it often utilizes evidence-based protocols.

In medicine, we make evidence-based decisions all the time, and patient safety is a huge issue. The Joint Commission on Accreditation of Healthcare Organizations recently looked at all perinatal sentinel events across the country in all types of institutions, and found that 47% were linked with staff competence issues. Among the other identified root causes were communication issues (72%), the orientation and training process (40%), and organization culture as a barrier to effective communication and teamwork (55%). Simulation could play a significant role in addressing each of these issues.

Shoulder dystocia complicates up to 2% of all vaginal deliveries, and potentially causes permanent brachial plexus injury, clavicular fracture, hypoxic brain injury, and other significant long-term complications. Although we encounter shoulder dystocia infrequently, the risk for serious and permanent injury to the infant is so high that we ought to be prepared.

Similarly, approximately 3%–4% of singleton babies are in the breech position, yet only a minority of obstetricians are able to perform vaginal breech deliveries. In one recent study, only 33% of surveyed attending physicians performed vaginal breech deliveries. The rest do not do them anymore.

Although vaginal breech deliveries are discouraged, vaginal delivery is sometimes unavoidable or even preferable. (When the breech is on the perineum, for instance, it's riskier to go to cesarean delivery). We are putting our patients and ourselves at risk by not practicing and knowing how to do this with proper technique,

We do not like to talk about the litigation aspect, but we cannot hide it: Fetal injury that is related to emergencies like shoulder dystocia is a potential source of medical malpractice lawsuits and one that we can minimize by reinforcing and maintaining our skills through simulation.

Today's Simulators

Obstetricians worry about how they can do a simulation. Many think of simulators as too big, too expensive, and not lifelike. Some worry about doing a simulation in front of others and are too intimidated to try.

Some of the simulators available today are expensive. A full-body, high-fidelity obstetric simulator with all the bells and whistles—touch-screen computer technology that enables manipulation of the labor course, for instance, and varying vital signs and fetal heart rhythms—can cost up to $40,000.

These expensive models are often purchased by academic institutions that are interested in simulation for a multitude of purposes, including team training, but such models are not necessary to simulate at least several obstetric emergencies, including vaginal breech delivery, shoulder dystocia, and the use of forceps.

For these situations, low-fidelity simulators—which may be just a model of the pelvis through which a model baby can be manually pushed—are perfectly fine. They can be purchased for $2,000-$3,000, stored in a closet, and placed in an extra exam room where physicians can practice, either with a mentor or expert or by themselves.

Nothing is as real as a true patient or a real-life situation, of course, but many of these mannequins are surprisingly lifelike, with features like an anatomically correct bony pelvis, a stretchable perineum, and a silicone pelvic-floor musculature. A mannequin's cervix, for instance, really feels like a cervix.

When I was in resident training, I practiced using the forceps on a high-fidelity mannequin. This gave me an opportunity to practice all the necessary maneuvers and to know whether I performed all critical tasks, from inserting the posterior blade first, for instance, to holding the left blade of the forceps with my left hand while using my right hand as a guide.

Later, when I was in a real and urgent situation requiring forceps, I knew just what to do. It worked like clockwork. Simulation on a low-fidelity mannequin, if that was what my institution had had, would have been just as beneficial.

Simulation also provides opportunities to create protocols. In the middle of a forceps delivery simulation, for instance, you may realize that “this needs to be done all the time just like this.” Alternatively, you may think, “Let's not do it this way next time.”

Similarly, simulation affords us opportunities to practice and fine-tune communication and teamwork.

Improved Competence

I recently oversaw a resident who had previously done simulation training with high-fidelity mannequins as part of her curriculum at the Washington Hospital Center, and was now in a real and difficult delivery involving shoulder dystocia.

She performed the recommended initial maneuvers—like placing the patient in the McRobert's position and applying suprapubic pressure—but without success. She then immediately proceeded, without any prompting, to deliver the posterior arm, which relieved the shoulder dystocia. Afterward, the resident told me that “if I hadn't done the shoulder dystocia simulation lab, I would not have known to do that.” I hear such stories often.

Studies are beginning to document the effects of obstetric simulation training on competence and performance.

In a study published several years ago, for instance, residents at Georgetown University in Washington and the Uniformed Services University of Health Sciences in Bethesda, Md., were randomized to receive training on shoulder dystocia management using a high-fidelity obstetric simulator or to receive no special training. Each resident was subsequently tested without prior notice in another simulation scenario.

Those who had practiced shoulder dystocia management on mannequins completed more critical tasks and had significantly higher scores on timeliness of their interactions, proper performance of maneuvers, and overall performance (Obstet. Gynecol. 2004:103;1224–8).

Although not randomized, another more recent study at Georgetown University showed that high-fidelity simulation training improved resident performance of vaginal breech delivery. Residents were more likely after simulation training to perform critical maneuvers correctly and to deliver in a safe manner than they were before the training (Obstet. Gynecol. 2006:107;86–9).

Research from the University of Bristol (England) is also yielding interesting results. Investigators there have reported, for instance, that obstetric emergency training courses using simulation were associated with a significant reduction in low 5-minute APGAR scores and lower rates of hypoxic-ischemic encephalopathy (BJOG 2006;113:177–82).

Another study of shoulder dystocia has shown that, whereas training with high-fidelity mannequins provides additional benefits, training with low-fidelity mannequins is also effective in improving management of the obstetric situation by obstetricians and midwives (Obstet. Gynecol. 2006;108:1477–85).

A study from the Bristol investigators in which participants were tested on a standardized simulation before a simulation workshop, and then at 3 weeks, 6 months, and 12 months afterward, shows that improved performance appears to be sustained. Those who were proficient 3 weeks after the training retained their skills at the later dates. The researchers concluded that annual training may be adequate for some physicians, whereas others may need more frequent practice (Obstet. Gynecol. 2007;110:1069–74).

Soon-to-be-published research that we have recently completed at Georgetown University and the Washington Hospital Center similarly indicates that obstetricians generally should strive for continuing simulation training at least once a year. Residents in our study who were initially taught on the simulator scored higher when tested a year later than did residents who received no simulation training. Overall, however, everyone's scores declined.

Obstetric simulation is part of our future. New physicians of the future will enter practice having done simulation training in a variety of high-acuity, low-frequency scenarios—rather than learning solely through lectures and impromptu teaching after events have occurred—and those of us already in practice will likely find that working occasionally with low-fidelity mannequins enables us to provide better, safer patient care while reducing our liability risk.

Dr. Marsha Solomon, chief resident at the Washington Hospital Center, is shown performing a simulated forceps delivery in the photo at left. In the photo at right, Dr. Solomon performs a simulated breech vaginal delivery. Photos courtesy Dr. Tamika C. Auguste

Obstetric Simulation

Do you think that you would like to have your next airplane flight piloted by someone who has not flown a plane in several years or who has little experience in landing?

Pilots are among the professionals who gain their greatest experience and expertise through the development of skills using various simulation technologies.

Simulation training is common in the aviation industry, as it is in aeronautics and in some branches of engineering, which makes this question significantly less worrisome and less relevant than if simulation were not common.

Medicine in general—and obstetrics in particular—has been practiced worldwide using the apprenticeship model, in which residents and interns work with attending physicians to learn the art of medicine.

While caring for patients with various disorders and in various scenarios, physicians-in-training work alongside the more senior practitioners, taking on progressive amounts of responsibility. Experience is gained accordingly.

This approach has been very successful over the years, and will remain so. It may be enhanced, however, as the simulation approach is slowly integrated into medicine and into obstetrics training.

The use of simulation training in medicine makes intuitive sense. The acquisition of the greatest possible skill or expertise—or the enhancement of skills if there is a hiatus in practice—makes sense from quality-of-care and patient-safety perspectives, and also because of the litigious environment in which we live and practice.

The question, “How would you like to have your baby delivered by an obstetrician who has not used forceps or managed shoulder dystocia in over a year?” is a valid one for patients who realize that less-common delivery scenarios are unpredictable.

This month's Master Class will focus on the utility, practicability, and application of simulation technology in obstetrics as a means of maximizing not only the skills of the resident, but also the skills of the practicing clinician.

Our guest author, Dr. Tamika C. Auguste, is the director of obstetric simulation at Washington Hospital Center and assistant professor of obstetrics/gynecology at Georgetown University in Washington. She speaks in various forums on the issue of simulation for both residents and practicing physicians, and is fast becoming a young leader and expert from whom we can expect to hear more in the future.

Key Points on Simulation

1. Simulation can be used to practice classic obstetric skills and high-risk, low-frequency obstetric emergencies.

2. Simulation is not only for those in academic medicine but also for those in private practice.

3. Low-fidelity simulators can be just as useful as high-fidelity simulators.

4. Simulation is becoming the norm in residency training programs.

Nausea and vomiting are common in pregnancy and can have a significant negative effect on women's health. Approximately half of all pregnant women in the United States have nausea and vomiting in early pregnancy, and about 25% have nausea alone. Only about 25% of pregnant women are free of any such problem.

The problem presents across a broad spectrum of severity, with the most severe form being hyperemesis gravidarum, a condition characterized by persistent vomiting, weight loss greater than 5%, ketonuria, electrolyte abnormalities, hypokalemia, and dehydration; this condition usually results in the need for hospitalization, treatment with intravenous fluids, and even intravenous feeding. Approximately 1% of pregnant women have vomiting severe enough to require hospitalization.

Persistent mild nausea, however, can also be a significant problem worthy of attentive management. It is not just “morning sickness” for many of these women. Approximately 35% of women with nausea during pregnancy lose time from work, and 25% cannot function well at home throughout the day.

Nausea and vomiting can significantly impair their routines, can negatively affect their relationships with their husbands and children, and are sometimes cited as reasons for an otherwise undesired pregnancy termination.

Women who are suffering from nausea and vomiting in pregnancy frequently do not seek or receive specific therapy out of concern over safety, yet such fear is often based on misinformation and misperceptions regarding teratogenesis. Women have numerous safe and effective options, including therapy with vitamin B₆ and doxylamine, as well as ginger and other nonpharmacologic approaches, and treatment with various antiemetic drugs.

Etiology, Differential Diagnosis

Some patients can identify the triggers of their nausea and thus can avoid aggravating odors or foods. Dietary modifications include eating frequent and small meals; taking fluids between meals; eating primarily bland, dry, and high-protein foods; and avoiding fatty or spicy foods.

Discontinuing prenatal vitamin tablets containing iron also can help, as the iron can contribute to nausea. Women who are having trouble can switch to a multivitamin with no or low iron for the first trimester and can resume prenatal vitamins after 3 months, or they can switch to folic acid alone, which is all that is needed to prevent birth defects.

We must also consider other diagnoses that can cause nausea and vomiting in pregnancy, from gastroenteritis, pancreatitis, appendicitis, and other gastrointestinal disorders, to gastrourologic problems such as pyelonephritis and various metabolic disorders.

There are sometimes clues that the nausea and vomiting cannot be attributed to the pregnancy alone: Fever, abdominal pain, and headache, for instance, result from something other than the pregnancy, as do serious changes in liver enzymes, bilirubin, and amylase or lipase.

Nausea and vomiting that begin later in the pregnancy also cannot be attributed to the pregnancy itself. The problem has an early onset, usually starting at the time of the missed menstrual period. It is fully manifested by 10 weeks of gestation, and—although it usually improves as the pregnancy progresses further—the problem may persist until the placenta is delivered.

In any case, a patient who has not had any nausea in the first 3 months of her pregnancy and begins experiencing nausea and headache at 16 weeks of gestation is probably having a migraine headache.

Many believe that nausea and vomiting are related to the presence of human chorionic gonadotropin (HCG), because HCG can stimulate the ovaries to produce estrogen, and estrogen can contribute to nausea. Indeed, the start, peak, and resolution of nausea and vomiting in pregnancy correlate closely with the curve of HCG concentration. Nausea and vomiting are also more common in patients with multiple gestations and hydatidiform moles, obstetric situations in which HCG is high.

Hormonal influences do not explain, however, why some women have problems with nausea and others do not.

Over the years, some have believed that the problem is psychological, but I and many others strongly discount this belief. Any psychological problems these women have are not a cause of their nausea and vomiting, but rather are an effect.

Gastrointestinal dysmotility and Helicobacter pylori infection have been cited as other possible associations. H. pylori seropositivity has been associated with hyperemesis or serious nausea and vomiting, but data are conflicting and investigators have not studied whether the problem resolves after treatment for the infection. Ulcer disease should register as a possibility in any differential diagnosis, particularly if the woman has pain, but whether it is more broadly causative of the nausea and vomiting of pregnancy is uncertain.

There is also some evidence that the nausea and vomiting of pregnancy may be related to vitamin B₆ deficiency, and indeed, a significant number of women respond to vitamin B₆ supplementation. Overall, however, the etiology of nausea and vomiting in pregnancy are unknown.

Vitamin-Based Therapy and Doxylamine

Vitamin B₆ is a good initial therapy for women whose nausea and vomiting cannot be managed with dietary change. It has been more widely tested for safety and efficacy than has any other vitamin-based therapy for the problem, and it is inexpensive and widely available.

In a study from our group conducted many years ago, pregnant patients with nausea and vomiting were randomized to 3 days of vitamin B₆ or a placebo at a dosage of 25 mg three times a day. Half of the patients stopped vomiting, and most patients with severe nausea reported a diminution to mild or moderate nausea. Several years later, another group of investigators used 10 mg of vitamin B₆ three times a day in a larger, randomized, placebo-controlled study. After 5 days of therapy, they also documented a significant decrease in nausea.

A once-a-day, extended-release formulation of vitamin B₆ (PremesisRx) is a good first-line option. It delivers 75 mg of vitamin B₆ over 24 hours—which is easier than taking 25 mg three times a day—and contains some vitamin B12, calcium carbonate, and 1 mg of folic acid. (The level of folic acid makes the formulation a prescription therapy.)

If vitamin B₆ alone is not successful, the antihistamine doxylamine can be added in a combination similar to the formulation that was available in Bendectin from 1956 to 1983. It is estimated that Bendectin was used by more than 30 million women during this time period and, at one point, by approximately 40% of pregnant women.

Although no other agent given in pregnancy has more conclusive safety data with regard to the incidence of birth defects (more than 6,000 patients exposed to the combination have been compared with more than 6,000 controls), Bendectin was removed from the U.S. market in 1983 by the manufacturer because of lawsuits that alleged a teratogenic effect.

The combination has been continually available in Canada as Diclectin (a sustained-release formulation), and interestingly, there are significantly fewer hospitalizations for nausea and vomiting in pregnancy in Canada than in the United States.

A version of the combination can be created by combining vitamin B₆ with the over-the-counter sleep aid Unisom SleepTabs, which contains 25 mg of doxylamine per tablet. The dose of doxylamine in Bendectin was 10 mg, and two tablets were recommended at night, so one full tablet of Unisom can be taken at night, along with a half tablet in the morning and a half-tablet in the afternoon if some nausea persists and, of course, 25 mg of vitamin B₆ at each of these times of the day.

The combination of vitamin B₆ and doxylamine can bring fast and dramatic relief for many patients, leading to significant improvements in the quality of their lives. There is always concern for obstetricians that a mother will claim that a child's birth defect was caused by a drug prescribed during the first trimester, but this is unlikely to happen with the combination of vitamin B₆ and doxylamine because legal precedents already hold that the drug does not cause birth defects.

Interestingly, some studies have suggested that women who have taken multivitamins containing vitamin B₆ before pregnancy have less nausea and vomiting.

Nonpharmacologic Approaches

Ginger ale has been a traditional remedy for nausea in various populations, and among pregnant women with nausea and vomiting, ginger is the alternative therapy with the strongest evidence base. The data on ginger have accumulated to the point at which concerns about its possible adverse effects have largely dissipated, which makes it worthy of consideration as a second-line agent.

Two small, randomized, double-blind trials used 250-mg ground ginger capsules or placebo four times a day, one in 70 outpatients with nausea and vomiting and one in women who were hospitalized with hyperemesis gravidarum. Investigators of both trials reported significantly reduced nausea and reductions in vomiting among the women in the ginger groups (Obstet. Gynecol. 2001;97:577–82; Eur. J. Obstet. Gynecol. Reprod. Biol. 1990;38:19–24).

Among more recent randomized trials was one of approximately 300 women that compared ginger with vitamin B₆. Women who received identical-looking capsules three times a day of 25 mg vitamin B₆ or 350 mg ginger had similar levels of improvement in nausea and vomiting at 1 week, 2 weeks, and 3 weeks.

There were no differences in fetal outcome or congenital anomalies; the only difference was that the women taking ginger reported more heartburn and belching (Obstet. Gynecol. 2004;103:639–45).

In a literature review, a group of Italian investigators identified six double-blind, randomized, controlled trials with a total of 675 participants that met criteria for methodological quality for the evaluation of efficacy. Of these six trials, four demonstrated the superiority of ginger over placebo, and two demonstrated the equivalence of ginger with vitamin B₆.

To review safety, the investigators looked at an observational cohort study involving 187 women as well as at the randomized trials. The studies showed no significant side effects and no adverse effects on pregnancy outcome (Obstet. Gynecol. 2005;105:849).

Acupuncture is another therapy worthy of consideration and one that can be added to the treatment regimen at any time. It has now been studied in two randomized trials in pregnant women who had nausea and vomiting, and although the results do not demonstrate broad efficacy, the findings together suggest that the therapy can be worth a try (Obstet. Gynecol. 2001:97;184–8; J. Pain Symptom Manage. 2000:20;273–9).

Nerve stimulation of the P6 acupuncture point also appears to decrease the nausea and vomiting of pregnancy for some women, whereas acupressure with devices like the Sea-Band or the Bioband appears to be less effective.

Antiemetic Drugs

Ginger and vitamin B₆—alone or in combination with doxylamine—do not work for everyone. In unsuccessful cases, we can move on to try other antihistamines and, if necessary, to consider the four other categories of antiemetic drugs: phenothiazines, prokinetic agents, serotonin (5-HT3) antagonists, and corticosteroids.

With the exception of doxylamine, which is a Food and Drug Administration category A drug, none are FDA approved for use in pregnancy. The drugs are underutilized, however, largely because of misperceptions of teratogenic risk.

In a supplement to the American Journal of Obstetrics and Gynecology on nausea and vomiting in pregnancy, Dr. L.A. Magee and associates reported on an evidence-based review of the safety and effectiveness of available antiemetics. They concluded that many medications, particularly the antihistamines and phenothiazines, are safe and effective for the treatment of varying degrees of nausea and vomiting (Am. J. Obstet. Gynecol. 2002:186;S256–61).

In the same supplement, Dr. Gideon Koren addressed the issue of perceived versus true risk of medications for nausea and vomiting, and presents an algorithm for management that includes a hierarchical use of antiemetic drugs based on the strength of evidence of fetal safety (Am. J. Obstet. Gynecol. 2002:186;S248–52).

Although few studies have compared the antihistamines for nausea and vomiting in pregnancy, sedation seems to be a main difference among the various drugs, with some—such as diphenhydramine (Benadryl)—sedating more than others. In addition to doxylamine and diphenhydramine, we can consider using dimenhydrinate (Dramamine), meclizine (Antivert), hydroxyzine (Vistaril, Atarax), and cetirizine (Zyrtec).

If the antihistamines as a class are not effective, the phenothiazines are a good choice. Promethazine (Phenergan) is widely used for nausea and vomiting in pregnancy, and prochlorperazine (Compazine) and chlorpromazine (Thorazine) are other options.

Possible adverse side effects of the phenothiazines include sedation, hypotension, dry mouth, and extrapyramidal symptoms. Compazine tablets are placed inside the cheek—a formulation that is helpful for women with moderate and severe nausea—and are generally well tolerated, with less drowsiness and sedation than the antihistamines.

The phenothiazine droperidol (Inapsine) was popular for some time, but there were reports of cardiac deaths and, in 2001, the FDA issued a black box warning stating that all patients need a 12-lead ECG before, during, and after administration. This drug has, consequently, fallen out of favor.

Metoclopramide (Reglan) can help some women when other drugs have failed. It is a prokinetic agent, increasing upper gastrointestinal motility and lower esophageal sphincter tone. A review of Medicaid data showed no increased risk of birth defects in 303 newborns in Michigan born to mothers who had ingested this drug.

The serotonin (5-HT3) antagonist ondansetron (Zofran) has been one of the most heavily marketed drugs for postoperative nausea and vomiting, and from the start many women and their obstetricians used the drug as a first-line or near-first-line antiemetic choice for nausea and vomiting in pregnancy, despite its high cost and the relative paucity of information on its use in pregnancy.

Several years of use and studies of several hundred patients have increased the comfort level related to ondansetron use. In general, this drug and the serotonin antagonists dolasetron (Anzemet) and granisetron (Kytril) are now felt to be safe. All are FDA category B drugs.

Zofran comes in an oral disintegrating tablet that, like Compazine, is useful in patients who have difficulty swallowing or who do not feel they are able to drink. In a randomized trial, Zofran was compared with Phenergan and was found to have similar efficacy, but with less sedation.

Corticosteroids may not be as beneficial as many first thought—there are now conflicting data about their effectiveness—and some studies have suggested an increased risk of cleft lip and palate when these agents are used before 10 weeks' gestation. The drugs are recommended, therefore, only after 10 weeks' gestation and in cases in which other medications have failed.

Neither I nor any member of my family has any financial connections with the pharmaceutical industry.

A Fairly Common Condition

Nausea and vomiting in pregnancy are fairly common conditions. By themselves, they are neither life threatening nor threatening to the pregnancy, but they are nevertheless terribly uncomfortable, displeasing, and sometimes disabling.

Nausea and vomiting can often interfere with a patient's ability to perform household or workplace duties and can be destabilizing to a patient's life at a time when there otherwise is joy and happiness. Sometimes, nausea and vomiting can be so severe that they result in hospitalization and intravenous feeding.

For all these reasons, they cannot be overlooked.

Unfortunately, they are also mysterious. Nausea and vomiting in pregnancy are relatively uncommon in the populations of certain continents, such as Africa and Asia, but relatively common among North American women. They occur for varying lengths of time in some North American women and not at all in others. The causation, in short, is unclear and likely complicated. A number of hypotheses have been advanced, but none has been proved.

Despite the uncertainties, the frequency and effects of nausea and vomiting in pregnancy demand our attention. It therefore seems important to present a Master Class on the subject and to review the various treatments that have been tried and are available for patients who struggle with the condition.

Some of these approaches will work for our patients, and some will not. With these options before us, however, we can offer our patients the very best care.

Here to address the issue is Dr. Jennifer R. Niebyl, professor and chair of the department of obstetrics and gynecology at the University of Iowa Hospitals and Clinics, Iowa City.

Dr. Niebyl is an expert in the area of drugs and pregnancy and an expert in maternal-fetal medicine, with a special interest in nausea and vomiting in pregnancy.

Key Points

▸ Vitamin B6 and doxylamine should be first-line therapy for nausea and vomiting in pregnancy.

▸ No antiemetic has been found to have any teratogenic risk.

▸ Some alternative remedies, such as ginger and acupuncture, have been shown to be effective.

Treatments for Nausea/Vomiting

Vitamin B6

PremesisRx

Antihistamines

Doxylamine (Unisom)

Dimenhydrinate (Dramamine)

Diphenhydramine (Benadryl)

Meclizine (Antivert)

Hydroxyzine (Vistaril, Atarax)

Cetirizine (Zyrtec)

Phenothiazines

Promethazine (Phenergan)

Prochlorperazine (Compazine)

Chlorpromazine (Thorazine)

Prokinetic Agent

Metoclopramide (Reglan)

5-HT3 Receptor Antagonists

Ondansetron (Zofran)

Dolasetron (Anzemet)

Granisetron (Kytril)

Corticosteroids

Acupuncture

Ginger

Nausea and vomiting are common in pregnancy and can have a significant negative effect on women's health. Approximately half of all pregnant women in the United States have nausea and vomiting in early pregnancy, and about 25% have nausea alone. Only about 25% of pregnant women are free of any such problem.

The problem presents across a broad spectrum of severity, with the most severe form being hyperemesis gravidarum, a condition characterized by persistent vomiting, weight loss greater than 5%, ketonuria, electrolyte abnormalities, hypokalemia, and dehydration; this condition usually results in the need for hospitalization, treatment with intravenous fluids, and even intravenous feeding. Approximately 1% of pregnant women have vomiting severe enough to require hospitalization.

Persistent mild nausea, however, can also be a significant problem worthy of attentive management. It is not just “morning sickness” for many of these women. Approximately 35% of women with nausea during pregnancy lose time from work, and 25% cannot function well at home throughout the day.

Nausea and vomiting can significantly impair their routines, can negatively affect their relationships with their husbands and children, and are sometimes cited as reasons for an otherwise undesired pregnancy termination.

Women who are suffering from nausea and vomiting in pregnancy frequently do not seek or receive specific therapy out of concern over safety, yet such fear is often based on misinformation and misperceptions regarding teratogenesis. Women have numerous safe and effective options, including therapy with vitamin B₆ and doxylamine, as well as ginger and other nonpharmacologic approaches, and treatment with various antiemetic drugs.

Etiology, Differential Diagnosis

Some patients can identify the triggers of their nausea and thus can avoid aggravating odors or foods. Dietary modifications include eating frequent and small meals; taking fluids between meals; eating primarily bland, dry, and high-protein foods; and avoiding fatty or spicy foods.

Discontinuing prenatal vitamin tablets containing iron also can help, as the iron can contribute to nausea. Women who are having trouble can switch to a multivitamin with no or low iron for the first trimester and can resume prenatal vitamins after 3 months, or they can switch to folic acid alone, which is all that is needed to prevent birth defects.

We must also consider other diagnoses that can cause nausea and vomiting in pregnancy, from gastroenteritis, pancreatitis, appendicitis, and other gastrointestinal disorders, to gastrourologic problems such as pyelonephritis and various metabolic disorders.

There are sometimes clues that the nausea and vomiting cannot be attributed to the pregnancy alone: Fever, abdominal pain, and headache, for instance, result from something other than the pregnancy, as do serious changes in liver enzymes, bilirubin, and amylase or lipase.

Nausea and vomiting that begin later in the pregnancy also cannot be attributed to the pregnancy itself. The problem has an early onset, usually starting at the time of the missed menstrual period. It is fully manifested by 10 weeks of gestation, and—although it usually improves as the pregnancy progresses further—the problem may persist until the placenta is delivered.

In any case, a patient who has not had any nausea in the first 3 months of her pregnancy and begins experiencing nausea and headache at 16 weeks of gestation is probably having a migraine headache.

Many believe that nausea and vomiting are related to the presence of human chorionic gonadotropin (HCG), because HCG can stimulate the ovaries to produce estrogen, and estrogen can contribute to nausea. Indeed, the start, peak, and resolution of nausea and vomiting in pregnancy correlate closely with the curve of HCG concentration. Nausea and vomiting are also more common in patients with multiple gestations and hydatidiform moles, obstetric situations in which HCG is high.

Hormonal influences do not explain, however, why some women have problems with nausea and others do not.

Over the years, some have believed that the problem is psychological, but I and many others strongly discount this belief. Any psychological problems these women have are not a cause of their nausea and vomiting, but rather are an effect.

Gastrointestinal dysmotility and Helicobacter pylori infection have been cited as other possible associations. H. pylori seropositivity has been associated with hyperemesis or serious nausea and vomiting, but data are conflicting and investigators have not studied whether the problem resolves after treatment for the infection. Ulcer disease should register as a possibility in any differential diagnosis, particularly if the woman has pain, but whether it is more broadly causative of the nausea and vomiting of pregnancy is uncertain.

There is also some evidence that the nausea and vomiting of pregnancy may be related to vitamin B₆ deficiency, and indeed, a significant number of women respond to vitamin B₆ supplementation. Overall, however, the etiology of nausea and vomiting in pregnancy are unknown.

Vitamin-Based Therapy and Doxylamine

Vitamin B₆ is a good initial therapy for women whose nausea and vomiting cannot be managed with dietary change. It has been more widely tested for safety and efficacy than has any other vitamin-based therapy for the problem, and it is inexpensive and widely available.

In a study from our group conducted many years ago, pregnant patients with nausea and vomiting were randomized to 3 days of vitamin B₆ or a placebo at a dosage of 25 mg three times a day. Half of the patients stopped vomiting, and most patients with severe nausea reported a diminution to mild or moderate nausea. Several years later, another group of investigators used 10 mg of vitamin B₆ three times a day in a larger, randomized, placebo-controlled study. After 5 days of therapy, they also documented a significant decrease in nausea.

A once-a-day, extended-release formulation of vitamin B₆ (PremesisRx) is a good first-line option. It delivers 75 mg of vitamin B₆ over 24 hours—which is easier than taking 25 mg three times a day—and contains some vitamin B12, calcium carbonate, and 1 mg of folic acid. (The level of folic acid makes the formulation a prescription therapy.)

If vitamin B₆ alone is not successful, the antihistamine doxylamine can be added in a combination similar to the formulation that was available in Bendectin from 1956 to 1983. It is estimated that Bendectin was used by more than 30 million women during this time period and, at one point, by approximately 40% of pregnant women.

Although no other agent given in pregnancy has more conclusive safety data with regard to the incidence of birth defects (more than 6,000 patients exposed to the combination have been compared with more than 6,000 controls), Bendectin was removed from the U.S. market in 1983 by the manufacturer because of lawsuits that alleged a teratogenic effect.

The combination has been continually available in Canada as Diclectin (a sustained-release formulation), and interestingly, there are significantly fewer hospitalizations for nausea and vomiting in pregnancy in Canada than in the United States.

A version of the combination can be created by combining vitamin B₆ with the over-the-counter sleep aid Unisom SleepTabs, which contains 25 mg of doxylamine per tablet. The dose of doxylamine in Bendectin was 10 mg, and two tablets were recommended at night, so one full tablet of Unisom can be taken at night, along with a half tablet in the morning and a half-tablet in the afternoon if some nausea persists and, of course, 25 mg of vitamin B₆ at each of these times of the day.

The combination of vitamin B₆ and doxylamine can bring fast and dramatic relief for many patients, leading to significant improvements in the quality of their lives. There is always concern for obstetricians that a mother will claim that a child's birth defect was caused by a drug prescribed during the first trimester, but this is unlikely to happen with the combination of vitamin B₆ and doxylamine because legal precedents already hold that the drug does not cause birth defects.

Interestingly, some studies have suggested that women who have taken multivitamins containing vitamin B₆ before pregnancy have less nausea and vomiting.

Nonpharmacologic Approaches

Ginger ale has been a traditional remedy for nausea in various populations, and among pregnant women with nausea and vomiting, ginger is the alternative therapy with the strongest evidence base. The data on ginger have accumulated to the point at which concerns about its possible adverse effects have largely dissipated, which makes it worthy of consideration as a second-line agent.

Two small, randomized, double-blind trials used 250-mg ground ginger capsules or placebo four times a day, one in 70 outpatients with nausea and vomiting and one in women who were hospitalized with hyperemesis gravidarum. Investigators of both trials reported significantly reduced nausea and reductions in vomiting among the women in the ginger groups (Obstet. Gynecol. 2001;97:577–82; Eur. J. Obstet. Gynecol. Reprod. Biol. 1990;38:19–24).

Among more recent randomized trials was one of approximately 300 women that compared ginger with vitamin B₆. Women who received identical-looking capsules three times a day of 25 mg vitamin B₆ or 350 mg ginger had similar levels of improvement in nausea and vomiting at 1 week, 2 weeks, and 3 weeks.

There were no differences in fetal outcome or congenital anomalies; the only difference was that the women taking ginger reported more heartburn and belching (Obstet. Gynecol. 2004;103:639–45).

In a literature review, a group of Italian investigators identified six double-blind, randomized, controlled trials with a total of 675 participants that met criteria for methodological quality for the evaluation of efficacy. Of these six trials, four demonstrated the superiority of ginger over placebo, and two demonstrated the equivalence of ginger with vitamin B₆.

To review safety, the investigators looked at an observational cohort study involving 187 women as well as at the randomized trials. The studies showed no significant side effects and no adverse effects on pregnancy outcome (Obstet. Gynecol. 2005;105:849).

Acupuncture is another therapy worthy of consideration and one that can be added to the treatment regimen at any time. It has now been studied in two randomized trials in pregnant women who had nausea and vomiting, and although the results do not demonstrate broad efficacy, the findings together suggest that the therapy can be worth a try (Obstet. Gynecol. 2001:97;184–8; J. Pain Symptom Manage. 2000:20;273–9).

Nerve stimulation of the P6 acupuncture point also appears to decrease the nausea and vomiting of pregnancy for some women, whereas acupressure with devices like the Sea-Band or the Bioband appears to be less effective.

Antiemetic Drugs

Ginger and vitamin B₆—alone or in combination with doxylamine—do not work for everyone. In unsuccessful cases, we can move on to try other antihistamines and, if necessary, to consider the four other categories of antiemetic drugs: phenothiazines, prokinetic agents, serotonin (5-HT3) antagonists, and corticosteroids.

With the exception of doxylamine, which is a Food and Drug Administration category A drug, none are FDA approved for use in pregnancy. The drugs are underutilized, however, largely because of misperceptions of teratogenic risk.

In a supplement to the American Journal of Obstetrics and Gynecology on nausea and vomiting in pregnancy, Dr. L.A. Magee and associates reported on an evidence-based review of the safety and effectiveness of available antiemetics. They concluded that many medications, particularly the antihistamines and phenothiazines, are safe and effective for the treatment of varying degrees of nausea and vomiting (Am. J. Obstet. Gynecol. 2002:186;S256–61).

In the same supplement, Dr. Gideon Koren addressed the issue of perceived versus true risk of medications for nausea and vomiting, and presents an algorithm for management that includes a hierarchical use of antiemetic drugs based on the strength of evidence of fetal safety (Am. J. Obstet. Gynecol. 2002:186;S248–52).

Although few studies have compared the antihistamines for nausea and vomiting in pregnancy, sedation seems to be a main difference among the various drugs, with some—such as diphenhydramine (Benadryl)—sedating more than others. In addition to doxylamine and diphenhydramine, we can consider using dimenhydrinate (Dramamine), meclizine (Antivert), hydroxyzine (Vistaril, Atarax), and cetirizine (Zyrtec).

If the antihistamines as a class are not effective, the phenothiazines are a good choice. Promethazine (Phenergan) is widely used for nausea and vomiting in pregnancy, and prochlorperazine (Compazine) and chlorpromazine (Thorazine) are other options.

Possible adverse side effects of the phenothiazines include sedation, hypotension, dry mouth, and extrapyramidal symptoms. Compazine tablets are placed inside the cheek—a formulation that is helpful for women with moderate and severe nausea—and are generally well tolerated, with less drowsiness and sedation than the antihistamines.

The phenothiazine droperidol (Inapsine) was popular for some time, but there were reports of cardiac deaths and, in 2001, the FDA issued a black box warning stating that all patients need a 12-lead ECG before, during, and after administration. This drug has, consequently, fallen out of favor.

Metoclopramide (Reglan) can help some women when other drugs have failed. It is a prokinetic agent, increasing upper gastrointestinal motility and lower esophageal sphincter tone. A review of Medicaid data showed no increased risk of birth defects in 303 newborns in Michigan born to mothers who had ingested this drug.

The serotonin (5-HT3) antagonist ondansetron (Zofran) has been one of the most heavily marketed drugs for postoperative nausea and vomiting, and from the start many women and their obstetricians used the drug as a first-line or near-first-line antiemetic choice for nausea and vomiting in pregnancy, despite its high cost and the relative paucity of information on its use in pregnancy.

Several years of use and studies of several hundred patients have increased the comfort level related to ondansetron use. In general, this drug and the serotonin antagonists dolasetron (Anzemet) and granisetron (Kytril) are now felt to be safe. All are FDA category B drugs.

Zofran comes in an oral disintegrating tablet that, like Compazine, is useful in patients who have difficulty swallowing or who do not feel they are able to drink. In a randomized trial, Zofran was compared with Phenergan and was found to have similar efficacy, but with less sedation.

Corticosteroids may not be as beneficial as many first thought—there are now conflicting data about their effectiveness—and some studies have suggested an increased risk of cleft lip and palate when these agents are used before 10 weeks' gestation. The drugs are recommended, therefore, only after 10 weeks' gestation and in cases in which other medications have failed.

Neither I nor any member of my family has any financial connections with the pharmaceutical industry.

A Fairly Common Condition

Nausea and vomiting in pregnancy are fairly common conditions. By themselves, they are neither life threatening nor threatening to the pregnancy, but they are nevertheless terribly uncomfortable, displeasing, and sometimes disabling.

Nausea and vomiting can often interfere with a patient's ability to perform household or workplace duties and can be destabilizing to a patient's life at a time when there otherwise is joy and happiness. Sometimes, nausea and vomiting can be so severe that they result in hospitalization and intravenous feeding.

For all these reasons, they cannot be overlooked.

Unfortunately, they are also mysterious. Nausea and vomiting in pregnancy are relatively uncommon in the populations of certain continents, such as Africa and Asia, but relatively common among North American women. They occur for varying lengths of time in some North American women and not at all in others. The causation, in short, is unclear and likely complicated. A number of hypotheses have been advanced, but none has been proved.

Despite the uncertainties, the frequency and effects of nausea and vomiting in pregnancy demand our attention. It therefore seems important to present a Master Class on the subject and to review the various treatments that have been tried and are available for patients who struggle with the condition.

Some of these approaches will work for our patients, and some will not. With these options before us, however, we can offer our patients the very best care.

Here to address the issue is Dr. Jennifer R. Niebyl, professor and chair of the department of obstetrics and gynecology at the University of Iowa Hospitals and Clinics, Iowa City.

Dr. Niebyl is an expert in the area of drugs and pregnancy and an expert in maternal-fetal medicine, with a special interest in nausea and vomiting in pregnancy.

Key Points

▸ Vitamin B6 and doxylamine should be first-line therapy for nausea and vomiting in pregnancy.

▸ No antiemetic has been found to have any teratogenic risk.

▸ Some alternative remedies, such as ginger and acupuncture, have been shown to be effective.

Treatments for Nausea/Vomiting

Vitamin B6

PremesisRx

Antihistamines

Doxylamine (Unisom)

Dimenhydrinate (Dramamine)

Diphenhydramine (Benadryl)

Meclizine (Antivert)

Hydroxyzine (Vistaril, Atarax)

Cetirizine (Zyrtec)

Phenothiazines

Promethazine (Phenergan)

Prochlorperazine (Compazine)

Chlorpromazine (Thorazine)

Prokinetic Agent

Metoclopramide (Reglan)

5-HT3 Receptor Antagonists

Ondansetron (Zofran)

Dolasetron (Anzemet)

Granisetron (Kytril)

Corticosteroids

Acupuncture

Ginger

Nausea and vomiting are common in pregnancy and can have a significant negative effect on women's health. Approximately half of all pregnant women in the United States have nausea and vomiting in early pregnancy, and about 25% have nausea alone. Only about 25% of pregnant women are free of any such problem.

The problem presents across a broad spectrum of severity, with the most severe form being hyperemesis gravidarum, a condition characterized by persistent vomiting, weight loss greater than 5%, ketonuria, electrolyte abnormalities, hypokalemia, and dehydration; this condition usually results in the need for hospitalization, treatment with intravenous fluids, and even intravenous feeding. Approximately 1% of pregnant women have vomiting severe enough to require hospitalization.

Persistent mild nausea, however, can also be a significant problem worthy of attentive management. It is not just “morning sickness” for many of these women. Approximately 35% of women with nausea during pregnancy lose time from work, and 25% cannot function well at home throughout the day.

Nausea and vomiting can significantly impair their routines, can negatively affect their relationships with their husbands and children, and are sometimes cited as reasons for an otherwise undesired pregnancy termination.

Women who are suffering from nausea and vomiting in pregnancy frequently do not seek or receive specific therapy out of concern over safety, yet such fear is often based on misinformation and misperceptions regarding teratogenesis. Women have numerous safe and effective options, including therapy with vitamin B₆ and doxylamine, as well as ginger and other nonpharmacologic approaches, and treatment with various antiemetic drugs.

Etiology, Differential Diagnosis

Some patients can identify the triggers of their nausea and thus can avoid aggravating odors or foods. Dietary modifications include eating frequent and small meals; taking fluids between meals; eating primarily bland, dry, and high-protein foods; and avoiding fatty or spicy foods.

Discontinuing prenatal vitamin tablets containing iron also can help, as the iron can contribute to nausea. Women who are having trouble can switch to a multivitamin with no or low iron for the first trimester and can resume prenatal vitamins after 3 months, or they can switch to folic acid alone, which is all that is needed to prevent birth defects.

We must also consider other diagnoses that can cause nausea and vomiting in pregnancy, from gastroenteritis, pancreatitis, appendicitis, and other gastrointestinal disorders, to gastrourologic problems such as pyelonephritis and various metabolic disorders.

There are sometimes clues that the nausea and vomiting cannot be attributed to the pregnancy alone: Fever, abdominal pain, and headache, for instance, result from something other than the pregnancy, as do serious changes in liver enzymes, bilirubin, and amylase or lipase.

Nausea and vomiting that begin later in the pregnancy also cannot be attributed to the pregnancy itself. The problem has an early onset, usually starting at the time of the missed menstrual period. It is fully manifested by 10 weeks of gestation, and—although it usually improves as the pregnancy progresses further—the problem may persist until the placenta is delivered.

In any case, a patient who has not had any nausea in the first 3 months of her pregnancy and begins experiencing nausea and headache at 16 weeks of gestation is probably having a migraine headache.

Many believe that nausea and vomiting are related to the presence of human chorionic gonadotropin (HCG), because HCG can stimulate the ovaries to produce estrogen, and estrogen can contribute to nausea. Indeed, the start, peak, and resolution of nausea and vomiting in pregnancy correlate closely with the curve of HCG concentration. Nausea and vomiting are also more common in patients with multiple gestations and hydatidiform moles, obstetric situations in which HCG is high.

Hormonal influences do not explain, however, why some women have problems with nausea and others do not.

Over the years, some have believed that the problem is psychological, but I and many others strongly discount this belief. Any psychological problems these women have are not a cause of their nausea and vomiting, but rather are an effect.

Gastrointestinal dysmotility and Helicobacter pylori infection have been cited as other possible associations. H. pylori seropositivity has been associated with hyperemesis or serious nausea and vomiting, but data are conflicting and investigators have not studied whether the problem resolves after treatment for the infection. Ulcer disease should register as a possibility in any differential diagnosis, particularly if the woman has pain, but whether it is more broadly causative of the nausea and vomiting of pregnancy is uncertain.

There is also some evidence that the nausea and vomiting of pregnancy may be related to vitamin B₆ deficiency, and indeed, a significant number of women respond to vitamin B₆ supplementation. Overall, however, the etiology of nausea and vomiting in pregnancy are unknown.

Vitamin-Based Therapy and Doxylamine

Vitamin B₆ is a good initial therapy for women whose nausea and vomiting cannot be managed with dietary change. It has been more widely tested for safety and efficacy than has any other vitamin-based therapy for the problem, and it is inexpensive and widely available.

In a study from our group conducted many years ago, pregnant patients with nausea and vomiting were randomized to 3 days of vitamin B₆ or a placebo at a dosage of 25 mg three times a day. Half of the patients stopped vomiting, and most patients with severe nausea reported a diminution to mild or moderate nausea. Several years later, another group of investigators used 10 mg of vitamin B₆ three times a day in a larger, randomized, placebo-controlled study. After 5 days of therapy, they also documented a significant decrease in nausea.

A once-a-day, extended-release formulation of vitamin B₆ (PremesisRx) is a good first-line option. It delivers 75 mg of vitamin B₆ over 24 hours—which is easier than taking 25 mg three times a day—and contains some vitamin B12, calcium carbonate, and 1 mg of folic acid. (The level of folic acid makes the formulation a prescription therapy.)

If vitamin B₆ alone is not successful, the antihistamine doxylamine can be added in a combination similar to the formulation that was available in Bendectin from 1956 to 1983. It is estimated that Bendectin was used by more than 30 million women during this time period and, at one point, by approximately 40% of pregnant women.

Although no other agent given in pregnancy has more conclusive safety data with regard to the incidence of birth defects (more than 6,000 patients exposed to the combination have been compared with more than 6,000 controls), Bendectin was removed from the U.S. market in 1983 by the manufacturer because of lawsuits that alleged a teratogenic effect.

The combination has been continually available in Canada as Diclectin (a sustained-release formulation), and interestingly, there are significantly fewer hospitalizations for nausea and vomiting in pregnancy in Canada than in the United States.

A version of the combination can be created by combining vitamin B₆ with the over-the-counter sleep aid Unisom SleepTabs, which contains 25 mg of doxylamine per tablet. The dose of doxylamine in Bendectin was 10 mg, and two tablets were recommended at night, so one full tablet of Unisom can be taken at night, along with a half tablet in the morning and a half-tablet in the afternoon if some nausea persists and, of course, 25 mg of vitamin B₆ at each of these times of the day.

The combination of vitamin B₆ and doxylamine can bring fast and dramatic relief for many patients, leading to significant improvements in the quality of their lives. There is always concern for obstetricians that a mother will claim that a child's birth defect was caused by a drug prescribed during the first trimester, but this is unlikely to happen with the combination of vitamin B₆ and doxylamine because legal precedents already hold that the drug does not cause birth defects.

Interestingly, some studies have suggested that women who have taken multivitamins containing vitamin B₆ before pregnancy have less nausea and vomiting.

Nonpharmacologic Approaches

Ginger ale has been a traditional remedy for nausea in various populations, and among pregnant women with nausea and vomiting, ginger is the alternative therapy with the strongest evidence base. The data on ginger have accumulated to the point at which concerns about its possible adverse effects have largely dissipated, which makes it worthy of consideration as a second-line agent.

Two small, randomized, double-blind trials used 250-mg ground ginger capsules or placebo four times a day, one in 70 outpatients with nausea and vomiting and one in women who were hospitalized with hyperemesis gravidarum. Investigators of both trials reported significantly reduced nausea and reductions in vomiting among the women in the ginger groups (Obstet. Gynecol. 2001;97:577–82; Eur. J. Obstet. Gynecol. Reprod. Biol. 1990;38:19–24).

Among more recent randomized trials was one of approximately 300 women that compared ginger with vitamin B₆. Women who received identical-looking capsules three times a day of 25 mg vitamin B₆ or 350 mg ginger had similar levels of improvement in nausea and vomiting at 1 week, 2 weeks, and 3 weeks.

There were no differences in fetal outcome or congenital anomalies; the only difference was that the women taking ginger reported more heartburn and belching (Obstet. Gynecol. 2004;103:639–45).

In a literature review, a group of Italian investigators identified six double-blind, randomized, controlled trials with a total of 675 participants that met criteria for methodological quality for the evaluation of efficacy. Of these six trials, four demonstrated the superiority of ginger over placebo, and two demonstrated the equivalence of ginger with vitamin B₆.

To review safety, the investigators looked at an observational cohort study involving 187 women as well as at the randomized trials. The studies showed no significant side effects and no adverse effects on pregnancy outcome (Obstet. Gynecol. 2005;105:849).

Acupuncture is another therapy worthy of consideration and one that can be added to the treatment regimen at any time. It has now been studied in two randomized trials in pregnant women who had nausea and vomiting, and although the results do not demonstrate broad efficacy, the findings together suggest that the therapy can be worth a try (Obstet. Gynecol. 2001:97;184–8; J. Pain Symptom Manage. 2000:20;273–9).

Nerve stimulation of the P6 acupuncture point also appears to decrease the nausea and vomiting of pregnancy for some women, whereas acupressure with devices like the Sea-Band or the Bioband appears to be less effective.

Antiemetic Drugs

Ginger and vitamin B₆—alone or in combination with doxylamine—do not work for everyone. In unsuccessful cases, we can move on to try other antihistamines and, if necessary, to consider the four other categories of antiemetic drugs: phenothiazines, prokinetic agents, serotonin (5-HT3) antagonists, and corticosteroids.

With the exception of doxylamine, which is a Food and Drug Administration category A drug, none are FDA approved for use in pregnancy. The drugs are underutilized, however, largely because of misperceptions of teratogenic risk.

In a supplement to the American Journal of Obstetrics and Gynecology on nausea and vomiting in pregnancy, Dr. L.A. Magee and associates reported on an evidence-based review of the safety and effectiveness of available antiemetics. They concluded that many medications, particularly the antihistamines and phenothiazines, are safe and effective for the treatment of varying degrees of nausea and vomiting (Am. J. Obstet. Gynecol. 2002:186;S256–61).

In the same supplement, Dr. Gideon Koren addressed the issue of perceived versus true risk of medications for nausea and vomiting, and presents an algorithm for management that includes a hierarchical use of antiemetic drugs based on the strength of evidence of fetal safety (Am. J. Obstet. Gynecol. 2002:186;S248–52).

Although few studies have compared the antihistamines for nausea and vomiting in pregnancy, sedation seems to be a main difference among the various drugs, with some—such as diphenhydramine (Benadryl)—sedating more than others. In addition to doxylamine and diphenhydramine, we can consider using dimenhydrinate (Dramamine), meclizine (Antivert), hydroxyzine (Vistaril, Atarax), and cetirizine (Zyrtec).

If the antihistamines as a class are not effective, the phenothiazines are a good choice. Promethazine (Phenergan) is widely used for nausea and vomiting in pregnancy, and prochlorperazine (Compazine) and chlorpromazine (Thorazine) are other options.

Possible adverse side effects of the phenothiazines include sedation, hypotension, dry mouth, and extrapyramidal symptoms. Compazine tablets are placed inside the cheek—a formulation that is helpful for women with moderate and severe nausea—and are generally well tolerated, with less drowsiness and sedation than the antihistamines.

The phenothiazine droperidol (Inapsine) was popular for some time, but there were reports of cardiac deaths and, in 2001, the FDA issued a black box warning stating that all patients need a 12-lead ECG before, during, and after administration. This drug has, consequently, fallen out of favor.

Metoclopramide (Reglan) can help some women when other drugs have failed. It is a prokinetic agent, increasing upper gastrointestinal motility and lower esophageal sphincter tone. A review of Medicaid data showed no increased risk of birth defects in 303 newborns in Michigan born to mothers who had ingested this drug.

The serotonin (5-HT3) antagonist ondansetron (Zofran) has been one of the most heavily marketed drugs for postoperative nausea and vomiting, and from the start many women and their obstetricians used the drug as a first-line or near-first-line antiemetic choice for nausea and vomiting in pregnancy, despite its high cost and the relative paucity of information on its use in pregnancy.

Several years of use and studies of several hundred patients have increased the comfort level related to ondansetron use. In general, this drug and the serotonin antagonists dolasetron (Anzemet) and granisetron (Kytril) are now felt to be safe. All are FDA category B drugs.

Zofran comes in an oral disintegrating tablet that, like Compazine, is useful in patients who have difficulty swallowing or who do not feel they are able to drink. In a randomized trial, Zofran was compared with Phenergan and was found to have similar efficacy, but with less sedation.

Corticosteroids may not be as beneficial as many first thought—there are now conflicting data about their effectiveness—and some studies have suggested an increased risk of cleft lip and palate when these agents are used before 10 weeks' gestation. The drugs are recommended, therefore, only after 10 weeks' gestation and in cases in which other medications have failed.

Neither I nor any member of my family has any financial connections with the pharmaceutical industry.

A Fairly Common Condition

Nausea and vomiting in pregnancy are fairly common conditions. By themselves, they are neither life threatening nor threatening to the pregnancy, but they are nevertheless terribly uncomfortable, displeasing, and sometimes disabling.

Nausea and vomiting can often interfere with a patient's ability to perform household or workplace duties and can be destabilizing to a patient's life at a time when there otherwise is joy and happiness. Sometimes, nausea and vomiting can be so severe that they result in hospitalization and intravenous feeding.

For all these reasons, they cannot be overlooked.

Unfortunately, they are also mysterious. Nausea and vomiting in pregnancy are relatively uncommon in the populations of certain continents, such as Africa and Asia, but relatively common among North American women. They occur for varying lengths of time in some North American women and not at all in others. The causation, in short, is unclear and likely complicated. A number of hypotheses have been advanced, but none has been proved.

Despite the uncertainties, the frequency and effects of nausea and vomiting in pregnancy demand our attention. It therefore seems important to present a Master Class on the subject and to review the various treatments that have been tried and are available for patients who struggle with the condition.

Some of these approaches will work for our patients, and some will not. With these options before us, however, we can offer our patients the very best care.

Here to address the issue is Dr. Jennifer R. Niebyl, professor and chair of the department of obstetrics and gynecology at the University of Iowa Hospitals and Clinics, Iowa City.

Dr. Niebyl is an expert in the area of drugs and pregnancy and an expert in maternal-fetal medicine, with a special interest in nausea and vomiting in pregnancy.

Key Points

▸ Vitamin B6 and doxylamine should be first-line therapy for nausea and vomiting in pregnancy.

▸ No antiemetic has been found to have any teratogenic risk.

▸ Some alternative remedies, such as ginger and acupuncture, have been shown to be effective.

Treatments for Nausea/Vomiting

Vitamin B6

PremesisRx

Antihistamines

Doxylamine (Unisom)

Dimenhydrinate (Dramamine)

Diphenhydramine (Benadryl)

Meclizine (Antivert)

Hydroxyzine (Vistaril, Atarax)

Cetirizine (Zyrtec)

Phenothiazines

Promethazine (Phenergan)

Prochlorperazine (Compazine)

Chlorpromazine (Thorazine)

Prokinetic Agent

Metoclopramide (Reglan)

5-HT3 Receptor Antagonists

Ondansetron (Zofran)

Dolasetron (Anzemet)

Granisetron (Kytril)

Corticosteroids

Acupuncture

Ginger

The array-CGH test, which is already being used postnatally, will give obstetricians, geneticists, and their patients the opportunity in the prenatal setting to detect significantly more and smaller changes in the amount of chromosomal material present in individuals—and in significantly less time than a standard chromosome karyotype would take.

It may someday take the place of our standard techniques for cytogenetic analysis, but for now, it is a valuable addition to the available diagnostic tests.

Advances Over FISH

The technology, which has also been called chromosomal microarray, was first used to analyze gains and losses in chromosomal material in tumors and tumor cell lines. It is now a valuable tool in the postnatal testing of individuals with birth defects.

Between one-half and two-thirds of children with serious developmental abnormalities go undiagnosed and have a normal karyotype, so from a postnatal perspective, this new test has been welcomed at Johns Hopkins University and the Kennedy Krieger Institute, both in Baltimore, as well as at other institutions. Having a diagnosis facilitates the most appropriate therapy and allows parents to plan for future pregnancies and possible prenatal testing.

Yet it is the prenatal period for which array-CGH may have an even greater impact. Phenotypic features are not as apparent in the womb as at birth, making it more difficult to target testing with technology like rapid fluorescent in situ hybridization (FISH).

Along with standard karyotype analysis, the FISH technique has been the mainstay of cytogenetic analysis. It provides a targeted look at areas of the karyotype that are known to be associated with disease as a result of either the duplication or deletion of genetic material. In other words, it detects gains and losses in chromosomal material for just one or a few chromosome regions at a time.

Performing array-CGH is like doing FISH hundreds of times at once. Array-CGH testing may target the same chromosomal regions (and thus similar disorders) as a series of FISH tests, but array-CGH will target these regions at a much higher resolution, enabling the detection of much smaller deletions and duplications; it can also assess many regions associated with genetic disorders in a single test.

If we see on a prenatal ultrasound that a fetus has cardiac problems, for example, we might suspect the DiGeorge syndrome. The obstetrician today would probably perform an amniocentesis and order both a karyotype and FISH with a specific probe for the DiGeorge syndrome, which we know is caused by a deletion on chromosome 22, just as he or she would do in the postnatal period for a child with the syndrome's more obvious phenotypic features.

In the near future, the obstetrician facing this prenatal situation will likely proceed differently than he or she would in the postnatal period. The obstetrician will use array-CGH instead of FISH in order to cast a wider net—one that can catch a deletion on chromosome 22, as well as other possible deletions which may cause the heart defect.

Right now, the available array-CGH platforms can detect more than 40 syndromic chromosomal disorders. Just as with FISH, a normal result rules out only those conditions that correspond to the deletions or duplications that are covered on the array.

How Array-CGH Works

The technique involves labeling the patient's DNA in one fluorescent dye, labeling DNA from a normal control with a different fluorescent dye, allowing the DNA from both to mix, and then applying the mixture to a slide that contains small segments of DNA from known chromosomal regions.

The slide serves as the platform or the array. The mixture of the patient's DNA and the normal control DNA is allowed to match up, or hybridize, with the complementary DNA segments on the slide.

A scanner then reads the intensities of the two different dyes, determining their relative strength at each of the DNA spots on the array. If a patient has less DNA in a specified region of the genome—a deletion of chromosomal material—then the color of the control sample will be stronger at that point on the array. If a patient has more DNA in this specific region—a duplication of chromosomal material—then the color of the patient's sample will be stronger at that location.

Analysis can be performed on direct chorionic villi or amniotic fluid, or alternatively on cultured cells. For direct analysis, it might be necessary to amplify the amount of DNA obtained before running it on an array. In this case, it is essential that the amplification is uniform and does not introduce any bias.

Although many laboratories are using cultured cells at this point, some studies are demonstrating the feasibility of relying on uncultured samples, and ultimately, this is the direction in which we're heading. Direct testing of fetal DNA will save time and give us rapid results.

The Limitations of Array-CGH

Unlike standard karyotyping, array-CGH cannot detect defects in which the total amount of chromosomal material is unchanged. The test cannot, for instance, detect balance rearrangements, such as balanced reciprocal translocations, balanced Robertsonian translocations, and inversions.

In a couple with multiple miscarriages, a karyotype is still the appropriate test to perform on the parents' blood because a balanced rearrangement is what you would be looking for. You would not request array-CGH because balanced rearrangements are not detectable with this technique. On the other hand, array-CGH could be very useful on the products of conception from a miscarriage because very small deletions and duplications could be found.

Array-CGH also cannot detect point mutations, or small changes in the genes, like those that cause hemophilia or sickle cell disease. It is designed to detect the syndromes caused by duplications or deletions of larger amounts of chromosomal material. And it will not detect abnormalities that are not covered by the array.

Chromosomal mosaicism, in which only some cells show a particular abnormality, may or may not be more readily detected by array-CGH than by standard techniques.

On one hand, mosaicism may be more readily detected with array-CGH than with standard karyotype analysis because abnormal cells often do not divide as well and may be lost during the culture process that is part of the standard karyotyping methodology. On the other hand, experts believe that array-CGH may not detect mosaicism below a certain level—below the level, some say, at which the abnormality affects fewer than 15%–30% of cells.

Array-CGH will also inevitably detect normal variants (benign duplications and deletions that are not associated with any abnormal phenotype). Some variants will be difficult to explain. This has been true for karyotyping as well, and just as we have in the past, we will want to minimize parents' anxiety over the unknowns.

When we find variants of uncertain significance, we will turn to the parents, checking their blood samples for the same losses or gains of chromosomal material.

The Near Future

The clinicians and cytogeneticists who are using and offering array-CGH are on a learning curve. Experts seem to have been successful in ensuring that the test works for the disorders that are covered; there is an enormous amount of information and data being shared by centers and labs on what variants are associated with the normal phenotype, and on other issues as well.

At Johns Hopkins University and the Kennedy Krieger Institute, we have postnatal experience to draw upon as we bring array-CGH into the prenatal arena. Of the children with developmental delay and dysmorphic features who have had array-CGH, we have been able to give a specific syndromic diagnosis to approximately 5%–8%, depending on the array platform we utilize. In about 12%, we have detected variants that we know—through parental testing and the use of databases—are normal. In a much smaller percentage (3.4%) of these children, we have found variants that we cannot yet explain.

Until we learn more, we plan to limit prenatal array-CGH to cases in which there is a known abnormality on ultrasound, rather than offer the test more broadly as a screening tool for chromosomal abnormalities in high-risk pregnancies. And although we are moving in the postnatal setting toward more of a whole-genome screening, we will use targeted arrays in the prenatal setting.

Within this context—that of ultrasound-detected anomalies and targeted arrays—we can expect that 5%–10% of tests will provide a clear diagnosis.

The question of whether array-CGH could replace a karyotype in prenatal testing is an interesting one. For now, there are too many questions and issues (mosaicism and normal variants, for instance) to do away with karyotyping. We believe the role of array-CGH is to enhance our current approaches to prenatal testing, and in this sense, it is an exciting development.

Figure A shows a hybridized array of >4,200 BAC clones; B, one area enlarged; C, plot for chromosome 1 based on fluorescence ratios (patient vs. control DNA) showing normal copy number. Courtesy Dr. Denise Batista

Prenatal Diagnosis

In our contemporary society, where women and their physicians continue to seek as much information as possible early in their pregnancies, the field of prenatal diagnosis has rapidly become a well-established and central part of obstetrics. Prenatal diagnosis performed in the first trimester has become common practice—a far cry from the days in the not-so-distant past when the ultimate outcome of the fetus was not learned until the day of delivery.

As obstetricians and perinatologists, we benefit from being aware of and fully informed about the evolving technology that continues to move the field of prenatal diagnosis forward. The array of current prenatal diagnostic tools includes both invasive and noninvasive techniques that enable parents to assess the genetic, chromosomal, and biochemical aspects of their fetus considerably before the time of viability.

Parents and their physicians are using this information to guide them in pursuing potential therapeutic applications and interventions or, in some cases, interruption of the pregnancy.

Now there is a new technique called array-based comparative genomic hybridization, or array-CGH, which is entering the prenatal arena with promises of more comprehensive and faster detection capabilities than we now are afforded with the two current “gold standard” techniques: microscopic karyotype analysis and rapid fluorescent in situ hybridization.

Array-CGH is far from perfect in evaluating chromosomal material. It can only detect instances where there is a significant addition or deletion of genetic material. And, of course, it can only evaluate those genes encoded on the array.

As with every other prenatal diagnostic tool developed to date, the future use of this new technique involves many questions, including which variants are normal as opposed to abnormal, the technique's potential role as a screening tool, and other often vexing ambiguities and issues. However, its use in prenatal diagnosis will build upon a body of national experience in the postnatal setting.

To familiarize us with the new technology and discuss its role in prenatal diagnosis, I have invited Dr. Karin J. Blakemore to serve as the guest professor of this month's Master Class.

Dr. Blakemore is the director of maternal-fetal medicine and the Prenatal Genetics Service at Johns Hopkins University School of Medicine in Baltimore—an institution that is gearing up to use array-CGH as part of its armamentarium for prenatal diagnosis.

She is joined by her colleague Denise Batista, Ph.D., who is an assistant professor in the Johns Hopkins department of pathology and codirector of the university's prenatal cytogenetics laboratory. Dr. Batista also serves as the director of the cytogenetics laboratory at the Kennedy Krieger Institute in Baltimore.

Key Points for Array-CGH

▸ Detects: Unbalanced rearrangements, aneuploidy, gains and losses of regions represented in the array.

▸ Won't detect: Balanced rearrangements, point mutations, (possibly) low-level mosaicism.

▸ Pick-up rate: Estimated as 5%–10% from postnatal studies of developmentally delayed/dysmorphic children.

▸ Confirmation: By FISH probes.

▸ Parental studies: Might be necessary to sort out normal variants versus clinically significant changes.

▸ Copy number variants: Might find copy number variants of unknown significance.

▸ Platforms: Several commercial and home-brew arrays available with different genomic coverage.

The array-CGH test, which is already being used postnatally, will give obstetricians, geneticists, and their patients the opportunity in the prenatal setting to detect significantly more and smaller changes in the amount of chromosomal material present in individuals—and in significantly less time than a standard chromosome karyotype would take.

It may someday take the place of our standard techniques for cytogenetic analysis, but for now, it is a valuable addition to the available diagnostic tests.

Advances Over FISH

The technology, which has also been called chromosomal microarray, was first used to analyze gains and losses in chromosomal material in tumors and tumor cell lines. It is now a valuable tool in the postnatal testing of individuals with birth defects.

Between one-half and two-thirds of children with serious developmental abnormalities go undiagnosed and have a normal karyotype, so from a postnatal perspective, this new test has been welcomed at Johns Hopkins University and the Kennedy Krieger Institute, both in Baltimore, as well as at other institutions. Having a diagnosis facilitates the most appropriate therapy and allows parents to plan for future pregnancies and possible prenatal testing.

Yet it is the prenatal period for which array-CGH may have an even greater impact. Phenotypic features are not as apparent in the womb as at birth, making it more difficult to target testing with technology like rapid fluorescent in situ hybridization (FISH).

Along with standard karyotype analysis, the FISH technique has been the mainstay of cytogenetic analysis. It provides a targeted look at areas of the karyotype that are known to be associated with disease as a result of either the duplication or deletion of genetic material. In other words, it detects gains and losses in chromosomal material for just one or a few chromosome regions at a time.

Performing array-CGH is like doing FISH hundreds of times at once. Array-CGH testing may target the same chromosomal regions (and thus similar disorders) as a series of FISH tests, but array-CGH will target these regions at a much higher resolution, enabling the detection of much smaller deletions and duplications; it can also assess many regions associated with genetic disorders in a single test.

If we see on a prenatal ultrasound that a fetus has cardiac problems, for example, we might suspect the DiGeorge syndrome. The obstetrician today would probably perform an amniocentesis and order both a karyotype and FISH with a specific probe for the DiGeorge syndrome, which we know is caused by a deletion on chromosome 22, just as he or she would do in the postnatal period for a child with the syndrome's more obvious phenotypic features.

In the near future, the obstetrician facing this prenatal situation will likely proceed differently than he or she would in the postnatal period. The obstetrician will use array-CGH instead of FISH in order to cast a wider net—one that can catch a deletion on chromosome 22, as well as other possible deletions which may cause the heart defect.

Right now, the available array-CGH platforms can detect more than 40 syndromic chromosomal disorders. Just as with FISH, a normal result rules out only those conditions that correspond to the deletions or duplications that are covered on the array.

How Array-CGH Works

The technique involves labeling the patient's DNA in one fluorescent dye, labeling DNA from a normal control with a different fluorescent dye, allowing the DNA from both to mix, and then applying the mixture to a slide that contains small segments of DNA from known chromosomal regions.

The slide serves as the platform or the array. The mixture of the patient's DNA and the normal control DNA is allowed to match up, or hybridize, with the complementary DNA segments on the slide.

A scanner then reads the intensities of the two different dyes, determining their relative strength at each of the DNA spots on the array. If a patient has less DNA in a specified region of the genome—a deletion of chromosomal material—then the color of the control sample will be stronger at that point on the array. If a patient has more DNA in this specific region—a duplication of chromosomal material—then the color of the patient's sample will be stronger at that location.

Analysis can be performed on direct chorionic villi or amniotic fluid, or alternatively on cultured cells. For direct analysis, it might be necessary to amplify the amount of DNA obtained before running it on an array. In this case, it is essential that the amplification is uniform and does not introduce any bias.

Although many laboratories are using cultured cells at this point, some studies are demonstrating the feasibility of relying on uncultured samples, and ultimately, this is the direction in which we're heading. Direct testing of fetal DNA will save time and give us rapid results.

The Limitations of Array-CGH

Unlike standard karyotyping, array-CGH cannot detect defects in which the total amount of chromosomal material is unchanged. The test cannot, for instance, detect balance rearrangements, such as balanced reciprocal translocations, balanced Robertsonian translocations, and inversions.

In a couple with multiple miscarriages, a karyotype is still the appropriate test to perform on the parents' blood because a balanced rearrangement is what you would be looking for. You would not request array-CGH because balanced rearrangements are not detectable with this technique. On the other hand, array-CGH could be very useful on the products of conception from a miscarriage because very small deletions and duplications could be found.

Array-CGH also cannot detect point mutations, or small changes in the genes, like those that cause hemophilia or sickle cell disease. It is designed to detect the syndromes caused by duplications or deletions of larger amounts of chromosomal material. And it will not detect abnormalities that are not covered by the array.

Chromosomal mosaicism, in which only some cells show a particular abnormality, may or may not be more readily detected by array-CGH than by standard techniques.

On one hand, mosaicism may be more readily detected with array-CGH than with standard karyotype analysis because abnormal cells often do not divide as well and may be lost during the culture process that is part of the standard karyotyping methodology. On the other hand, experts believe that array-CGH may not detect mosaicism below a certain level—below the level, some say, at which the abnormality affects fewer than 15%–30% of cells.

Array-CGH will also inevitably detect normal variants (benign duplications and deletions that are not associated with any abnormal phenotype). Some variants will be difficult to explain. This has been true for karyotyping as well, and just as we have in the past, we will want to minimize parents' anxiety over the unknowns.

When we find variants of uncertain significance, we will turn to the parents, checking their blood samples for the same losses or gains of chromosomal material.

The Near Future

The clinicians and cytogeneticists who are using and offering array-CGH are on a learning curve. Experts seem to have been successful in ensuring that the test works for the disorders that are covered; there is an enormous amount of information and data being shared by centers and labs on what variants are associated with the normal phenotype, and on other issues as well.

At Johns Hopkins University and the Kennedy Krieger Institute, we have postnatal experience to draw upon as we bring array-CGH into the prenatal arena. Of the children with developmental delay and dysmorphic features who have had array-CGH, we have been able to give a specific syndromic diagnosis to approximately 5%–8%, depending on the array platform we utilize. In about 12%, we have detected variants that we know—through parental testing and the use of databases—are normal. In a much smaller percentage (3.4%) of these children, we have found variants that we cannot yet explain.

Until we learn more, we plan to limit prenatal array-CGH to cases in which there is a known abnormality on ultrasound, rather than offer the test more broadly as a screening tool for chromosomal abnormalities in high-risk pregnancies. And although we are moving in the postnatal setting toward more of a whole-genome screening, we will use targeted arrays in the prenatal setting.

Within this context—that of ultrasound-detected anomalies and targeted arrays—we can expect that 5%–10% of tests will provide a clear diagnosis.

The question of whether array-CGH could replace a karyotype in prenatal testing is an interesting one. For now, there are too many questions and issues (mosaicism and normal variants, for instance) to do away with karyotyping. We believe the role of array-CGH is to enhance our current approaches to prenatal testing, and in this sense, it is an exciting development.

Figure A shows a hybridized array of >4,200 BAC clones; B, one area enlarged; C, plot for chromosome 1 based on fluorescence ratios (patient vs. control DNA) showing normal copy number. Courtesy Dr. Denise Batista

Prenatal Diagnosis

In our contemporary society, where women and their physicians continue to seek as much information as possible early in their pregnancies, the field of prenatal diagnosis has rapidly become a well-established and central part of obstetrics. Prenatal diagnosis performed in the first trimester has become common practice—a far cry from the days in the not-so-distant past when the ultimate outcome of the fetus was not learned until the day of delivery.

As obstetricians and perinatologists, we benefit from being aware of and fully informed about the evolving technology that continues to move the field of prenatal diagnosis forward. The array of current prenatal diagnostic tools includes both invasive and noninvasive techniques that enable parents to assess the genetic, chromosomal, and biochemical aspects of their fetus considerably before the time of viability.

Parents and their physicians are using this information to guide them in pursuing potential therapeutic applications and interventions or, in some cases, interruption of the pregnancy.

Now there is a new technique called array-based comparative genomic hybridization, or array-CGH, which is entering the prenatal arena with promises of more comprehensive and faster detection capabilities than we now are afforded with the two current “gold standard” techniques: microscopic karyotype analysis and rapid fluorescent in situ hybridization.

Array-CGH is far from perfect in evaluating chromosomal material. It can only detect instances where there is a significant addition or deletion of genetic material. And, of course, it can only evaluate those genes encoded on the array.

As with every other prenatal diagnostic tool developed to date, the future use of this new technique involves many questions, including which variants are normal as opposed to abnormal, the technique's potential role as a screening tool, and other often vexing ambiguities and issues. However, its use in prenatal diagnosis will build upon a body of national experience in the postnatal setting.

To familiarize us with the new technology and discuss its role in prenatal diagnosis, I have invited Dr. Karin J. Blakemore to serve as the guest professor of this month's Master Class.

Dr. Blakemore is the director of maternal-fetal medicine and the Prenatal Genetics Service at Johns Hopkins University School of Medicine in Baltimore—an institution that is gearing up to use array-CGH as part of its armamentarium for prenatal diagnosis.

She is joined by her colleague Denise Batista, Ph.D., who is an assistant professor in the Johns Hopkins department of pathology and codirector of the university's prenatal cytogenetics laboratory. Dr. Batista also serves as the director of the cytogenetics laboratory at the Kennedy Krieger Institute in Baltimore.

Key Points for Array-CGH

▸ Detects: Unbalanced rearrangements, aneuploidy, gains and losses of regions represented in the array.

▸ Won't detect: Balanced rearrangements, point mutations, (possibly) low-level mosaicism.

▸ Pick-up rate: Estimated as 5%–10% from postnatal studies of developmentally delayed/dysmorphic children.

▸ Confirmation: By FISH probes.

▸ Parental studies: Might be necessary to sort out normal variants versus clinically significant changes.

▸ Copy number variants: Might find copy number variants of unknown significance.

▸ Platforms: Several commercial and home-brew arrays available with different genomic coverage.

The array-CGH test, which is already being used postnatally, will give obstetricians, geneticists, and their patients the opportunity in the prenatal setting to detect significantly more and smaller changes in the amount of chromosomal material present in individuals—and in significantly less time than a standard chromosome karyotype would take.

It may someday take the place of our standard techniques for cytogenetic analysis, but for now, it is a valuable addition to the available diagnostic tests.

Advances Over FISH

The technology, which has also been called chromosomal microarray, was first used to analyze gains and losses in chromosomal material in tumors and tumor cell lines. It is now a valuable tool in the postnatal testing of individuals with birth defects.

Between one-half and two-thirds of children with serious developmental abnormalities go undiagnosed and have a normal karyotype, so from a postnatal perspective, this new test has been welcomed at Johns Hopkins University and the Kennedy Krieger Institute, both in Baltimore, as well as at other institutions. Having a diagnosis facilitates the most appropriate therapy and allows parents to plan for future pregnancies and possible prenatal testing.

Yet it is the prenatal period for which array-CGH may have an even greater impact. Phenotypic features are not as apparent in the womb as at birth, making it more difficult to target testing with technology like rapid fluorescent in situ hybridization (FISH).

Along with standard karyotype analysis, the FISH technique has been the mainstay of cytogenetic analysis. It provides a targeted look at areas of the karyotype that are known to be associated with disease as a result of either the duplication or deletion of genetic material. In other words, it detects gains and losses in chromosomal material for just one or a few chromosome regions at a time.

Performing array-CGH is like doing FISH hundreds of times at once. Array-CGH testing may target the same chromosomal regions (and thus similar disorders) as a series of FISH tests, but array-CGH will target these regions at a much higher resolution, enabling the detection of much smaller deletions and duplications; it can also assess many regions associated with genetic disorders in a single test.

If we see on a prenatal ultrasound that a fetus has cardiac problems, for example, we might suspect the DiGeorge syndrome. The obstetrician today would probably perform an amniocentesis and order both a karyotype and FISH with a specific probe for the DiGeorge syndrome, which we know is caused by a deletion on chromosome 22, just as he or she would do in the postnatal period for a child with the syndrome's more obvious phenotypic features.

In the near future, the obstetrician facing this prenatal situation will likely proceed differently than he or she would in the postnatal period. The obstetrician will use array-CGH instead of FISH in order to cast a wider net—one that can catch a deletion on chromosome 22, as well as other possible deletions which may cause the heart defect.

Right now, the available array-CGH platforms can detect more than 40 syndromic chromosomal disorders. Just as with FISH, a normal result rules out only those conditions that correspond to the deletions or duplications that are covered on the array.

How Array-CGH Works

The technique involves labeling the patient's DNA in one fluorescent dye, labeling DNA from a normal control with a different fluorescent dye, allowing the DNA from both to mix, and then applying the mixture to a slide that contains small segments of DNA from known chromosomal regions.

The slide serves as the platform or the array. The mixture of the patient's DNA and the normal control DNA is allowed to match up, or hybridize, with the complementary DNA segments on the slide.

A scanner then reads the intensities of the two different dyes, determining their relative strength at each of the DNA spots on the array. If a patient has less DNA in a specified region of the genome—a deletion of chromosomal material—then the color of the control sample will be stronger at that point on the array. If a patient has more DNA in this specific region—a duplication of chromosomal material—then the color of the patient's sample will be stronger at that location.

Analysis can be performed on direct chorionic villi or amniotic fluid, or alternatively on cultured cells. For direct analysis, it might be necessary to amplify the amount of DNA obtained before running it on an array. In this case, it is essential that the amplification is uniform and does not introduce any bias.

Although many laboratories are using cultured cells at this point, some studies are demonstrating the feasibility of relying on uncultured samples, and ultimately, this is the direction in which we're heading. Direct testing of fetal DNA will save time and give us rapid results.

The Limitations of Array-CGH

Unlike standard karyotyping, array-CGH cannot detect defects in which the total amount of chromosomal material is unchanged. The test cannot, for instance, detect balance rearrangements, such as balanced reciprocal translocations, balanced Robertsonian translocations, and inversions.

In a couple with multiple miscarriages, a karyotype is still the appropriate test to perform on the parents' blood because a balanced rearrangement is what you would be looking for. You would not request array-CGH because balanced rearrangements are not detectable with this technique. On the other hand, array-CGH could be very useful on the products of conception from a miscarriage because very small deletions and duplications could be found.

Array-CGH also cannot detect point mutations, or small changes in the genes, like those that cause hemophilia or sickle cell disease. It is designed to detect the syndromes caused by duplications or deletions of larger amounts of chromosomal material. And it will not detect abnormalities that are not covered by the array.

Chromosomal mosaicism, in which only some cells show a particular abnormality, may or may not be more readily detected by array-CGH than by standard techniques.

On one hand, mosaicism may be more readily detected with array-CGH than with standard karyotype analysis because abnormal cells often do not divide as well and may be lost during the culture process that is part of the standard karyotyping methodology. On the other hand, experts believe that array-CGH may not detect mosaicism below a certain level—below the level, some say, at which the abnormality affects fewer than 15%–30% of cells.

Array-CGH will also inevitably detect normal variants (benign duplications and deletions that are not associated with any abnormal phenotype). Some variants will be difficult to explain. This has been true for karyotyping as well, and just as we have in the past, we will want to minimize parents' anxiety over the unknowns.

When we find variants of uncertain significance, we will turn to the parents, checking their blood samples for the same losses or gains of chromosomal material.

The Near Future

The clinicians and cytogeneticists who are using and offering array-CGH are on a learning curve. Experts seem to have been successful in ensuring that the test works for the disorders that are covered; there is an enormous amount of information and data being shared by centers and labs on what variants are associated with the normal phenotype, and on other issues as well.

At Johns Hopkins University and the Kennedy Krieger Institute, we have postnatal experience to draw upon as we bring array-CGH into the prenatal arena. Of the children with developmental delay and dysmorphic features who have had array-CGH, we have been able to give a specific syndromic diagnosis to approximately 5%–8%, depending on the array platform we utilize. In about 12%, we have detected variants that we know—through parental testing and the use of databases—are normal. In a much smaller percentage (3.4%) of these children, we have found variants that we cannot yet explain.

Until we learn more, we plan to limit prenatal array-CGH to cases in which there is a known abnormality on ultrasound, rather than offer the test more broadly as a screening tool for chromosomal abnormalities in high-risk pregnancies. And although we are moving in the postnatal setting toward more of a whole-genome screening, we will use targeted arrays in the prenatal setting.

Within this context—that of ultrasound-detected anomalies and targeted arrays—we can expect that 5%–10% of tests will provide a clear diagnosis.

The question of whether array-CGH could replace a karyotype in prenatal testing is an interesting one. For now, there are too many questions and issues (mosaicism and normal variants, for instance) to do away with karyotyping. We believe the role of array-CGH is to enhance our current approaches to prenatal testing, and in this sense, it is an exciting development.

Figure A shows a hybridized array of >4,200 BAC clones; B, one area enlarged; C, plot for chromosome 1 based on fluorescence ratios (patient vs. control DNA) showing normal copy number. Courtesy Dr. Denise Batista

Prenatal Diagnosis

In our contemporary society, where women and their physicians continue to seek as much information as possible early in their pregnancies, the field of prenatal diagnosis has rapidly become a well-established and central part of obstetrics. Prenatal diagnosis performed in the first trimester has become common practice—a far cry from the days in the not-so-distant past when the ultimate outcome of the fetus was not learned until the day of delivery.

As obstetricians and perinatologists, we benefit from being aware of and fully informed about the evolving technology that continues to move the field of prenatal diagnosis forward. The array of current prenatal diagnostic tools includes both invasive and noninvasive techniques that enable parents to assess the genetic, chromosomal, and biochemical aspects of their fetus considerably before the time of viability.

Parents and their physicians are using this information to guide them in pursuing potential therapeutic applications and interventions or, in some cases, interruption of the pregnancy.

Now there is a new technique called array-based comparative genomic hybridization, or array-CGH, which is entering the prenatal arena with promises of more comprehensive and faster detection capabilities than we now are afforded with the two current “gold standard” techniques: microscopic karyotype analysis and rapid fluorescent in situ hybridization.

Array-CGH is far from perfect in evaluating chromosomal material. It can only detect instances where there is a significant addition or deletion of genetic material. And, of course, it can only evaluate those genes encoded on the array.

As with every other prenatal diagnostic tool developed to date, the future use of this new technique involves many questions, including which variants are normal as opposed to abnormal, the technique's potential role as a screening tool, and other often vexing ambiguities and issues. However, its use in prenatal diagnosis will build upon a body of national experience in the postnatal setting.

To familiarize us with the new technology and discuss its role in prenatal diagnosis, I have invited Dr. Karin J. Blakemore to serve as the guest professor of this month's Master Class.

Dr. Blakemore is the director of maternal-fetal medicine and the Prenatal Genetics Service at Johns Hopkins University School of Medicine in Baltimore—an institution that is gearing up to use array-CGH as part of its armamentarium for prenatal diagnosis.

She is joined by her colleague Denise Batista, Ph.D., who is an assistant professor in the Johns Hopkins department of pathology and codirector of the university's prenatal cytogenetics laboratory. Dr. Batista also serves as the director of the cytogenetics laboratory at the Kennedy Krieger Institute in Baltimore.

Key Points for Array-CGH

▸ Detects: Unbalanced rearrangements, aneuploidy, gains and losses of regions represented in the array.

▸ Won't detect: Balanced rearrangements, point mutations, (possibly) low-level mosaicism.

▸ Pick-up rate: Estimated as 5%–10% from postnatal studies of developmentally delayed/dysmorphic children.

▸ Confirmation: By FISH probes.

▸ Parental studies: Might be necessary to sort out normal variants versus clinically significant changes.

▸ Copy number variants: Might find copy number variants of unknown significance.

▸ Platforms: Several commercial and home-brew arrays available with different genomic coverage.

Our understanding of its causes and effects has expanded rapidly. We now know that a spectrum of disease can result when unequal placental sharing and/or unequal blood-volume sharing occurs in monochorionic pregnancies, for instance, and that a significant imbalance in blood-flow exchange between twins' circulations is the primary contributor to the development of TTTS. On the other hand, our knowledge is still quite simplistic: We have much to learn about the pathophysiology and the natural history and progression of the syndrome.

The observations we have made, however, are significant enough to justify the treatment of severe TTTS—especially given the advances in ultrasound assessment, which allow us to detect the syndrome early, as well as the dramatic improvements in technology for minimally invasive intrauterine therapy that have come about in recent years.

Endoscopic laser ablation (or laser coagulation) of placental anastomoses has been shown in numerous studies—including a multicenter, randomized trial comparing it with serial amnioreduction—to be an effective treatment for TTTS, and a preferable first-line approach for severe TTTS that is diagnosed before 26 weeks' gestation.

Because intrauterine procedures require a high level of expertise and infrastructure, it is likely that the management of these conditions will remain regionalized. Improved referral patterns and support for families, however, will promote the development of a nationwide network of designated centers, making such therapy more accessible.

Pathophysiology and Consequences

Identical twins are monochorionic, and these pregnancies present several potential risks: the risk that one baby will not get its fair share of the placenta, the risk that blood volume will be shared unequally, and an overall risk of vascular instability in each twin.

When the predominant issue in an identical twin pregnancy is unequal placental sharing, the growth of one baby becomes restricted and the other baby grows normally, resulting in a condition called selective intrauterine growth restriction (selective IUGR).

The other main issue—that of unequal blood-volume sharing—is what fuels TTTS. In uncomplicated pregnancies, blood is exchanged equally through the vascular anastomoses that characterize all monochorionic pregnancies. In complicated pregnancies, however, the exchange is unbalanced, and blood is shared in one direction without adequate return.

Arteries emanating from the placental cord insertion of one twin, for instance, can drain into a vein returning to the other twin. Such arteriovenous anastomoses are in the substance of the placenta and act as one-way valves for blood flow. If the amount of blood flow in one direction is not balanced by enough flow in the opposite direction—that is, if the magnitude of blood flow through unidirectional arteriovenous anastomoses is not compensated by vascular channels that permit flow in the opposite direction—then an imbalance develops that is potentially harmful to both babies.

In TTTS, which develops in about 15% of monochorionic pregnancies, the imbalance progresses to the extent that one twin becomes a “donor” of blood volume and the other becomes the “recipient” twin.

The donor twin moves blood across the anastomoses to the placenta and to the recipient twin, and does not receive an equal amount in return. A decline in blood volume leads to decreased urine output to the extent that, eventually, bladder filling in the donor twin virtually ceases. Under these circumstances, oligohydramnios may progress to anhydramnios, and the twin may become “stuck” in an essentially empty amniotic sac.

The recipient twin, in the meantime, receives an excess amount of venous blood volume. The increase in intravascular blood volume drives an increase of filtration in the kidneys, which results in excess urination. The increased urinary frequency, which may even result in constant bladder filling, leads to polyhydramnios.

When the sac of the recipient twin becomes distended by amniotic fluid, and the donor twin is no longer producing urine, the membrane between the twins may become wrapped so tightly around the donor twin that it is barely visible on ultrasound (See image below.) When the donor twin is “stuck” to the uterine wall in such a way, the ultrasound appearance resembles that of identical twins with one amniotic cavity (monoamniotic twins).

Untreated TTTS has serious consequences for each twin and for the whole pregnancy. First, the resultant polyhydramnios can stimulate preterm labor because of uterine distention. Second, abnormalities in blood volume can lead to cardiac problems and cardiovascular compromise for the babies, most often for the recipient twin. The excess blood cells and volume overload that this twin faces can lead to cardiac failure and hydrops.

The donor twin, meanwhile, is at risk for abnormalities and long-term effects resulting from compression, failing placental function, malnutrition, and hypovolemia.

If one baby dies in utero, the placental anastomoses that cause TTTS in the first place—that is, the open vessel connections that exist between the twins—carry an additional danger. In artery-to-artery and vein-to-vein anastomoses, the direction of blood flow is determined by the difference in blood pressure on either side. If one twin dies, the resultant drop in blood pressure causes the surviving twin to lose a large amount of blood volume across the connecting vessels and into the dying twin. This puts the surviving twin at risk of hemorrhagic shock and a heart attack or stroke.

It is estimated that the risk for white-matter injury in the surviving twin at the time of birth may be as high as 50% following such an intrauterine event. The fates of both twins are thus essentially linked to each other through their placental anastomoses.

Although exact contributors still need to be determined, it is well established that, compared with nonidentical twins, identical twins have a higher incidence of cerebral palsy and other anomalies, and a higher rate of developmental delay at 2 years. Because the development of TTTS is one well-recognized contributor to these statistics, perinatal interventions in monochorionic pregnancies have primarily focused on its treatment.

Evolution of Management

It's most interesting to look at the evolution of management from a historical perspective. When TTTS was clinically recognized, before the days of multivessel Doppler assessment, patients would most often present with a massively distended uterus and preterm labor.

The natural management approach was amnioreduction, which involved the removal of large volumes of amniotic fluid in an effort to relieve uterine distention and prevent preterm delivery. Physicians recognized the need for serial amnioreduction, as the procedure leaves anastomoses open and does nothing to address the underlying problem.

This approach was often satisfactory when it was started at 26–27 weeks' gestation because chances to prolong pregnancy to 32–34 weeks with repeated drainage were reasonable. The patients who presented with massive polyhydramnios and severe TTTS at 20 weeks, however, were another story. Their outcomes with serial amnioreduction were poor; in fact, many physicians would offer pregnancy termination under these circumstances.

In the late 1990s several groups began to address the underlying problem by closing the problem vessels. Dr. Julian De Lia, at that time practicing in Utah, was the first to describe fetoscopic laser ablation of placental anastomoses. He and the team of Prof. Kypros Nicolaides in Europe used a nonselective technique that involved ablating blood vessels and the placental mass along a dividing line between the twins—essentially making the placenta functionally dichorionic—and then draining the amniotic fluid.

Developmental research on the equipment and modification of the technique proceeded. In 1999, Dr. Rubén A. Quintero in Florida published a five-stage classification system for the progression of TTTS, with stages I and II characterized primarily by imbalances in blood volume, stages III and IV signified by cardiovascular compromise, and stage V signified by the death of one or both twins.

This staging system marked a significant step in the management of TTTS because it established a unified diagnostic approach that was based on prenatal criteria. Until this point, the definitions of TTTS were based on an extrapolation of pediatric diagnostic criteria that were used at birth. The application of Dr. Quintero's staging system allowed a more objective comparison of treatment strategies, but required familiarity with arterial and venous Doppler techniques.

Dr. Quintero also argued that a nonselective approach with the laser—one that coagulates vessels that do not contribute to TTTS, as well as those that do—can rob one or both twins of placental territory that is vital for their survival. He developed a selective laser technique that involves identification and coagulation of the vessels that pass from one twin to the other, leaving normal placental territory and noncontributing vessels untouched.

In the meantime, the Eurofetus research consortium had formed in Europe, and had begun designing a trial to compare laser therapy with amnioreduction, with one of their premises being that laser therapy would most benefit twin pregnancies that are complicated by TTTS before 26 weeks' gestation. Perinatal mortality for untreated severe TTTS, they knew, was as high as 90%, with significant handicap in the survivors.

Results of the multicenter randomized study were published in 2004 (N. Engl. J. Med. 2004;351:136-44). Complication rates were basically comparable (approximately 9% in each arm), but the rates of survival of at least one twin at 28 days and at 6 months of age were significantly better in the group that underwent selective laser coagulation than in the amnioreduction group (76% vs. 56% at 28 days, and 76% vs. 51% at 6 months).

The differences existed in both the early and later stages of TTTS, although fetuses in the Quintero stages I or II had better outcomes than did those with higher stages in both treatment groups. (The study had been concluded early, after 72 women had been assigned to the laser group and 70 to the amnioreduction group, when an interim analysis demonstrated significant benefits.)

Gestational ages at the time of delivery were also significantly different: Patients in the laser group delivered, on average, at 33 weeks, whereas those in the amnioreduction group delivered at 29 weeks.

An intermediate-term look at neurologic outcomes favored laser surgery as well: At 6 months of age, infants in the laser group were more likely than those in the amnioreduction group to be free of neurologic deficits (52% vs. 31%, respectively).

At the center for advanced fetal care at the University of Maryland, Baltimore, which has served for almost a decade as a referral resource for minimally invasive fetal therapy, I have applied the identical technique utilized in the Eurofetus trial using a selective approach. Our treatment results have consistently mirrored the published statistics.

Our research, which we presented at the annual meeting of the Society for Maternal-Fetal Medicine, confirms that successful laser ablation corrects the abnormal blood volume distribution. This effect is first apparent for the donor twin and clinically presents with the reappearance of bladder filling, often on the day after the procedure.

Urination gradually normalizes in the recipient twin, typically over 1–2 weeks after the procedure. The mother feels better immediately after the procedure and continues to improve as fetal status normalizes.

Longer-term follow-up of neurologic abnormalities in the Eurofetus trial is underway. For now, however, an analysis of a series of patients who received intrauterine laser treatment for TTTS has shown that 78% of 89 surviving children had a normal neurodevelopmental status at about 2 years of age, whereas 11% had minor neurologic deficiencies and 11% had major neurologic deficits (Am. J. Obstet. Gynecol. 2003;188:876-80).

Although comparisons of patients managed in the randomized trial are pending, these rates of neurologic handicap compare favorably with those seen after amnioreduction.

Two large series indicate that severe TTTS is associated with poor neurodevelopment, and that up to 27% of survivors may have abnormal brain ultrasounds at the time of delivery. It is therefore widely accepted that the neuroprotective benefit of laser therapy is most marked in early onset TTTS (prior to 26 weeks), and that the difference in outcomes is attributable to lower rates of preterm delivery and prematurity-associated complications as well as to the elimination of the risks of ongoing TTTS.

Moving Into the Future

In Europe, the randomized trial basically brought the controversy over optimal treatment for TTTS to a close. In the United States, there are some who still lean toward performing an initial amnioreduction and moving on to laser surgery if necessary.

There are disadvantages to such an approach. An initial amnioreduction removes the amniotic fluid pocket that is necessary to successfully maneuver the fetoscope. Decompression of the placenta not only unpredictably affects shunt dynamics but also can create placental “valleys” that can impair visualization of anastomoses. Potential bleeding from the procedure, as well as advancing gestational age until a suitable fluid pocket has reestablished, can also make the fluid cloudier.

Investigators who have looked at the factors that influence outcomes of selective laser coagulation of placental anastomoses have reported that those who do poorly have more advanced TTTS; have shorter cervical length, and thus a higher incidence of preterm labor; have a history of prior amnioreduction; and have technically difficult laser procedures with poor visualization of anastomoses as contributing factors.

Amnioreduction still has a role, however, particularly for patients who present with TTTS beyond 26 weeks' gestation. These patients are not candidates for laser therapy because the efficacy and safety of the procedure at this gestational age has not been studied.

Even with the improved outcomes, the therapies are still not optimal, and our knowledge of TTTS is still full of gaps and differences in opinion. Some experts believe, for instance, that with selective laser therapy there is a risk of recurring TTTS—that is, as visible anastomoses are closed, intravascular pressures are diverted to very small vessels that are barely visible at the time of the laser procedure. Over time, it is believed, these vessels may expand and therefore become hemodynamically relevant contributors to recurring TTTS.

At this time, I believe it's important to keep an open mind after presumably successful laser therapy, and to follow the fetuses closely after surgery for TTTS. Continued evaluation of bladder filling, amniotic fluid volumes, and placental and venous Doppler studies may be necessary over extended periods of time.

It is also important to inform neonatologists and referring obstetricians of the special circumstances of these babies, who behave very differently—both in the NICU and beyond—than do other babies of similar size or with other underlying conditions.

Babies who matured in utero as TTTS “recipients” are chronically hypervolemic and will not respond well, for instance, to dopamine given in the NICU as the primary agent to boost blood pressure. Careful attention to fluid balance is essential to prevent neonatal complications under these circumstances.

Fortunately, technologic advances in equipment are making intrauterine therapy much more minimally invasive. The development of fetoscopes with a 2-mm lens offers superior visual resolution and facilitates a minimally invasive approach. Digital camera technology also enhances the visualization of the smallest blood vessels. Steerable and angulated optical devices tackle the problems of anterior placenta. The smaller caliber of the entry site also decreases the risk for complications.

Today, laser surgery is typically performed under local anesthesia that requires minimal hospitalization with only perioperative tocolysis. The average length of patient stay is 1 day at the University of Maryland Medical Center.

The initiation of the North American Fetal Therapy Network (NAFTNet), a research consortium, is a significant development in the United States.

The membrane is seen folding around the donor who becomes a “stuck twin.” Courtesy Dr. Ahmet A. Baschat

When bladder filling of the donor can no longer be demonstrated, progression to stage 2 TTTS is diagnosed. (Left arrow, small bladder; right arrow, empty bladder.)

Critically abnormal waveforms in the umbilical artery (left) and ductus venosus (right) indicate stage 3 TTTS.

Ultrasound findings of hydrops (arrows point to fluid in fetal abdomen) indicate stage 4 TTTS. Photos courtesy Dr. Ahmet A. Baschat

Twin-to-Twin Transfusion Syndrome

Despite the advances that have occurred in obstetrics over the years, who would have imagined that fetal surgery would be a viable therapeutic approach today? Well, indeed, this is where we are in the history of obstetrics.

Fetal evaluation has been conducted over the years using a variety of noninvasive techniques, most notably electronic fetal monitoring. More invasive techniques, such as amniocentesis, have also been used with very good success and with relatively low risk to mother and fetus. Certain conditions, however, cannot be addressed with noninvasive or slightly invasive approaches, but rather require either open surgery or more involved surgery of a minimally invasive nature.

One of these conditions is twin-to-twin transfusion syndrome, in which one of the fetuses may succumb during intrauterine life. Over the years, amniocentesis has been used with limited success. However, newer techniques involving endoscopic laser therapy are being introduced with improved outcomes. In this Master Class, we review both modalities in the management of these patients, with careful attention to advances in fetal laser therapy.

We are pleased to introduce Dr. Ahmet A. Baschat, of the department of obstetrics, gynecology, and reproductive sciences at the University of Maryland School of Medicine, Baltimore, as our guest professor this month. Dr. Baschat is considered a national expert in fetal therapy, including laser and other intrauterine surgical procedures.

EMILY BRANNAN, ILLUSTRATION

Our understanding of its causes and effects has expanded rapidly. We now know that a spectrum of disease can result when unequal placental sharing and/or unequal blood-volume sharing occurs in monochorionic pregnancies, for instance, and that a significant imbalance in blood-flow exchange between twins' circulations is the primary contributor to the development of TTTS. On the other hand, our knowledge is still quite simplistic: We have much to learn about the pathophysiology and the natural history and progression of the syndrome.

The observations we have made, however, are significant enough to justify the treatment of severe TTTS—especially given the advances in ultrasound assessment, which allow us to detect the syndrome early, as well as the dramatic improvements in technology for minimally invasive intrauterine therapy that have come about in recent years.

Endoscopic laser ablation (or laser coagulation) of placental anastomoses has been shown in numerous studies—including a multicenter, randomized trial comparing it with serial amnioreduction—to be an effective treatment for TTTS, and a preferable first-line approach for severe TTTS that is diagnosed before 26 weeks' gestation.

Because intrauterine procedures require a high level of expertise and infrastructure, it is likely that the management of these conditions will remain regionalized. Improved referral patterns and support for families, however, will promote the development of a nationwide network of designated centers, making such therapy more accessible.

Pathophysiology and Consequences

Identical twins are monochorionic, and these pregnancies present several potential risks: the risk that one baby will not get its fair share of the placenta, the risk that blood volume will be shared unequally, and an overall risk of vascular instability in each twin.

When the predominant issue in an identical twin pregnancy is unequal placental sharing, the growth of one baby becomes restricted and the other baby grows normally, resulting in a condition called selective intrauterine growth restriction (selective IUGR).

The other main issue—that of unequal blood-volume sharing—is what fuels TTTS. In uncomplicated pregnancies, blood is exchanged equally through the vascular anastomoses that characterize all monochorionic pregnancies. In complicated pregnancies, however, the exchange is unbalanced, and blood is shared in one direction without adequate return.

Arteries emanating from the placental cord insertion of one twin, for instance, can drain into a vein returning to the other twin. Such arteriovenous anastomoses are in the substance of the placenta and act as one-way valves for blood flow. If the amount of blood flow in one direction is not balanced by enough flow in the opposite direction—that is, if the magnitude of blood flow through unidirectional arteriovenous anastomoses is not compensated by vascular channels that permit flow in the opposite direction—then an imbalance develops that is potentially harmful to both babies.

In TTTS, which develops in about 15% of monochorionic pregnancies, the imbalance progresses to the extent that one twin becomes a “donor” of blood volume and the other becomes the “recipient” twin.

The donor twin moves blood across the anastomoses to the placenta and to the recipient twin, and does not receive an equal amount in return. A decline in blood volume leads to decreased urine output to the extent that, eventually, bladder filling in the donor twin virtually ceases. Under these circumstances, oligohydramnios may progress to anhydramnios, and the twin may become “stuck” in an essentially empty amniotic sac.

The recipient twin, in the meantime, receives an excess amount of venous blood volume. The increase in intravascular blood volume drives an increase of filtration in the kidneys, which results in excess urination. The increased urinary frequency, which may even result in constant bladder filling, leads to polyhydramnios.

When the sac of the recipient twin becomes distended by amniotic fluid, and the donor twin is no longer producing urine, the membrane between the twins may become wrapped so tightly around the donor twin that it is barely visible on ultrasound (See image below.) When the donor twin is “stuck” to the uterine wall in such a way, the ultrasound appearance resembles that of identical twins with one amniotic cavity (monoamniotic twins).

Untreated TTTS has serious consequences for each twin and for the whole pregnancy. First, the resultant polyhydramnios can stimulate preterm labor because of uterine distention. Second, abnormalities in blood volume can lead to cardiac problems and cardiovascular compromise for the babies, most often for the recipient twin. The excess blood cells and volume overload that this twin faces can lead to cardiac failure and hydrops.

The donor twin, meanwhile, is at risk for abnormalities and long-term effects resulting from compression, failing placental function, malnutrition, and hypovolemia.

If one baby dies in utero, the placental anastomoses that cause TTTS in the first place—that is, the open vessel connections that exist between the twins—carry an additional danger. In artery-to-artery and vein-to-vein anastomoses, the direction of blood flow is determined by the difference in blood pressure on either side. If one twin dies, the resultant drop in blood pressure causes the surviving twin to lose a large amount of blood volume across the connecting vessels and into the dying twin. This puts the surviving twin at risk of hemorrhagic shock and a heart attack or stroke.

It is estimated that the risk for white-matter injury in the surviving twin at the time of birth may be as high as 50% following such an intrauterine event. The fates of both twins are thus essentially linked to each other through their placental anastomoses.

Although exact contributors still need to be determined, it is well established that, compared with nonidentical twins, identical twins have a higher incidence of cerebral palsy and other anomalies, and a higher rate of developmental delay at 2 years. Because the development of TTTS is one well-recognized contributor to these statistics, perinatal interventions in monochorionic pregnancies have primarily focused on its treatment.

Evolution of Management

It's most interesting to look at the evolution of management from a historical perspective. When TTTS was clinically recognized, before the days of multivessel Doppler assessment, patients would most often present with a massively distended uterus and preterm labor.

The natural management approach was amnioreduction, which involved the removal of large volumes of amniotic fluid in an effort to relieve uterine distention and prevent preterm delivery. Physicians recognized the need for serial amnioreduction, as the procedure leaves anastomoses open and does nothing to address the underlying problem.

This approach was often satisfactory when it was started at 26–27 weeks' gestation because chances to prolong pregnancy to 32–34 weeks with repeated drainage were reasonable. The patients who presented with massive polyhydramnios and severe TTTS at 20 weeks, however, were another story. Their outcomes with serial amnioreduction were poor; in fact, many physicians would offer pregnancy termination under these circumstances.

In the late 1990s several groups began to address the underlying problem by closing the problem vessels. Dr. Julian De Lia, at that time practicing in Utah, was the first to describe fetoscopic laser ablation of placental anastomoses. He and the team of Prof. Kypros Nicolaides in Europe used a nonselective technique that involved ablating blood vessels and the placental mass along a dividing line between the twins—essentially making the placenta functionally dichorionic—and then draining the amniotic fluid.

Developmental research on the equipment and modification of the technique proceeded. In 1999, Dr. Rubén A. Quintero in Florida published a five-stage classification system for the progression of TTTS, with stages I and II characterized primarily by imbalances in blood volume, stages III and IV signified by cardiovascular compromise, and stage V signified by the death of one or both twins.

This staging system marked a significant step in the management of TTTS because it established a unified diagnostic approach that was based on prenatal criteria. Until this point, the definitions of TTTS were based on an extrapolation of pediatric diagnostic criteria that were used at birth. The application of Dr. Quintero's staging system allowed a more objective comparison of treatment strategies, but required familiarity with arterial and venous Doppler techniques.

Dr. Quintero also argued that a nonselective approach with the laser—one that coagulates vessels that do not contribute to TTTS, as well as those that do—can rob one or both twins of placental territory that is vital for their survival. He developed a selective laser technique that involves identification and coagulation of the vessels that pass from one twin to the other, leaving normal placental territory and noncontributing vessels untouched.

In the meantime, the Eurofetus research consortium had formed in Europe, and had begun designing a trial to compare laser therapy with amnioreduction, with one of their premises being that laser therapy would most benefit twin pregnancies that are complicated by TTTS before 26 weeks' gestation. Perinatal mortality for untreated severe TTTS, they knew, was as high as 90%, with significant handicap in the survivors.

Results of the multicenter randomized study were published in 2004 (N. Engl. J. Med. 2004;351:136-44). Complication rates were basically comparable (approximately 9% in each arm), but the rates of survival of at least one twin at 28 days and at 6 months of age were significantly better in the group that underwent selective laser coagulation than in the amnioreduction group (76% vs. 56% at 28 days, and 76% vs. 51% at 6 months).

The differences existed in both the early and later stages of TTTS, although fetuses in the Quintero stages I or II had better outcomes than did those with higher stages in both treatment groups. (The study had been concluded early, after 72 women had been assigned to the laser group and 70 to the amnioreduction group, when an interim analysis demonstrated significant benefits.)

Gestational ages at the time of delivery were also significantly different: Patients in the laser group delivered, on average, at 33 weeks, whereas those in the amnioreduction group delivered at 29 weeks.

An intermediate-term look at neurologic outcomes favored laser surgery as well: At 6 months of age, infants in the laser group were more likely than those in the amnioreduction group to be free of neurologic deficits (52% vs. 31%, respectively).

At the center for advanced fetal care at the University of Maryland, Baltimore, which has served for almost a decade as a referral resource for minimally invasive fetal therapy, I have applied the identical technique utilized in the Eurofetus trial using a selective approach. Our treatment results have consistently mirrored the published statistics.

Our research, which we presented at the annual meeting of the Society for Maternal-Fetal Medicine, confirms that successful laser ablation corrects the abnormal blood volume distribution. This effect is first apparent for the donor twin and clinically presents with the reappearance of bladder filling, often on the day after the procedure.

Urination gradually normalizes in the recipient twin, typically over 1–2 weeks after the procedure. The mother feels better immediately after the procedure and continues to improve as fetal status normalizes.

Longer-term follow-up of neurologic abnormalities in the Eurofetus trial is underway. For now, however, an analysis of a series of patients who received intrauterine laser treatment for TTTS has shown that 78% of 89 surviving children had a normal neurodevelopmental status at about 2 years of age, whereas 11% had minor neurologic deficiencies and 11% had major neurologic deficits (Am. J. Obstet. Gynecol. 2003;188:876-80).

Although comparisons of patients managed in the randomized trial are pending, these rates of neurologic handicap compare favorably with those seen after amnioreduction.

Two large series indicate that severe TTTS is associated with poor neurodevelopment, and that up to 27% of survivors may have abnormal brain ultrasounds at the time of delivery. It is therefore widely accepted that the neuroprotective benefit of laser therapy is most marked in early onset TTTS (prior to 26 weeks), and that the difference in outcomes is attributable to lower rates of preterm delivery and prematurity-associated complications as well as to the elimination of the risks of ongoing TTTS.

Moving Into the Future

In Europe, the randomized trial basically brought the controversy over optimal treatment for TTTS to a close. In the United States, there are some who still lean toward performing an initial amnioreduction and moving on to laser surgery if necessary.

There are disadvantages to such an approach. An initial amnioreduction removes the amniotic fluid pocket that is necessary to successfully maneuver the fetoscope. Decompression of the placenta not only unpredictably affects shunt dynamics but also can create placental “valleys” that can impair visualization of anastomoses. Potential bleeding from the procedure, as well as advancing gestational age until a suitable fluid pocket has reestablished, can also make the fluid cloudier.

Investigators who have looked at the factors that influence outcomes of selective laser coagulation of placental anastomoses have reported that those who do poorly have more advanced TTTS; have shorter cervical length, and thus a higher incidence of preterm labor; have a history of prior amnioreduction; and have technically difficult laser procedures with poor visualization of anastomoses as contributing factors.

Amnioreduction still has a role, however, particularly for patients who present with TTTS beyond 26 weeks' gestation. These patients are not candidates for laser therapy because the efficacy and safety of the procedure at this gestational age has not been studied.

Even with the improved outcomes, the therapies are still not optimal, and our knowledge of TTTS is still full of gaps and differences in opinion. Some experts believe, for instance, that with selective laser therapy there is a risk of recurring TTTS—that is, as visible anastomoses are closed, intravascular pressures are diverted to very small vessels that are barely visible at the time of the laser procedure. Over time, it is believed, these vessels may expand and therefore become hemodynamically relevant contributors to recurring TTTS.

At this time, I believe it's important to keep an open mind after presumably successful laser therapy, and to follow the fetuses closely after surgery for TTTS. Continued evaluation of bladder filling, amniotic fluid volumes, and placental and venous Doppler studies may be necessary over extended periods of time.

It is also important to inform neonatologists and referring obstetricians of the special circumstances of these babies, who behave very differently—both in the NICU and beyond—than do other babies of similar size or with other underlying conditions.

Babies who matured in utero as TTTS “recipients” are chronically hypervolemic and will not respond well, for instance, to dopamine given in the NICU as the primary agent to boost blood pressure. Careful attention to fluid balance is essential to prevent neonatal complications under these circumstances.

Fortunately, technologic advances in equipment are making intrauterine therapy much more minimally invasive. The development of fetoscopes with a 2-mm lens offers superior visual resolution and facilitates a minimally invasive approach. Digital camera technology also enhances the visualization of the smallest blood vessels. Steerable and angulated optical devices tackle the problems of anterior placenta. The smaller caliber of the entry site also decreases the risk for complications.

Today, laser surgery is typically performed under local anesthesia that requires minimal hospitalization with only perioperative tocolysis. The average length of patient stay is 1 day at the University of Maryland Medical Center.

The initiation of the North American Fetal Therapy Network (NAFTNet), a research consortium, is a significant development in the United States.

The membrane is seen folding around the donor who becomes a “stuck twin.” Courtesy Dr. Ahmet A. Baschat

When bladder filling of the donor can no longer be demonstrated, progression to stage 2 TTTS is diagnosed. (Left arrow, small bladder; right arrow, empty bladder.)

Critically abnormal waveforms in the umbilical artery (left) and ductus venosus (right) indicate stage 3 TTTS.

Ultrasound findings of hydrops (arrows point to fluid in fetal abdomen) indicate stage 4 TTTS. Photos courtesy Dr. Ahmet A. Baschat

Twin-to-Twin Transfusion Syndrome

Despite the advances that have occurred in obstetrics over the years, who would have imagined that fetal surgery would be a viable therapeutic approach today? Well, indeed, this is where we are in the history of obstetrics.

Fetal evaluation has been conducted over the years using a variety of noninvasive techniques, most notably electronic fetal monitoring. More invasive techniques, such as amniocentesis, have also been used with very good success and with relatively low risk to mother and fetus. Certain conditions, however, cannot be addressed with noninvasive or slightly invasive approaches, but rather require either open surgery or more involved surgery of a minimally invasive nature.

One of these conditions is twin-to-twin transfusion syndrome, in which one of the fetuses may succumb during intrauterine life. Over the years, amniocentesis has been used with limited success. However, newer techniques involving endoscopic laser therapy are being introduced with improved outcomes. In this Master Class, we review both modalities in the management of these patients, with careful attention to advances in fetal laser therapy.

We are pleased to introduce Dr. Ahmet A. Baschat, of the department of obstetrics, gynecology, and reproductive sciences at the University of Maryland School of Medicine, Baltimore, as our guest professor this month. Dr. Baschat is considered a national expert in fetal therapy, including laser and other intrauterine surgical procedures.

EMILY BRANNAN, ILLUSTRATION

Our understanding of its causes and effects has expanded rapidly. We now know that a spectrum of disease can result when unequal placental sharing and/or unequal blood-volume sharing occurs in monochorionic pregnancies, for instance, and that a significant imbalance in blood-flow exchange between twins' circulations is the primary contributor to the development of TTTS. On the other hand, our knowledge is still quite simplistic: We have much to learn about the pathophysiology and the natural history and progression of the syndrome.

The observations we have made, however, are significant enough to justify the treatment of severe TTTS—especially given the advances in ultrasound assessment, which allow us to detect the syndrome early, as well as the dramatic improvements in technology for minimally invasive intrauterine therapy that have come about in recent years.

Endoscopic laser ablation (or laser coagulation) of placental anastomoses has been shown in numerous studies—including a multicenter, randomized trial comparing it with serial amnioreduction—to be an effective treatment for TTTS, and a preferable first-line approach for severe TTTS that is diagnosed before 26 weeks' gestation.

Because intrauterine procedures require a high level of expertise and infrastructure, it is likely that the management of these conditions will remain regionalized. Improved referral patterns and support for families, however, will promote the development of a nationwide network of designated centers, making such therapy more accessible.

Pathophysiology and Consequences

Identical twins are monochorionic, and these pregnancies present several potential risks: the risk that one baby will not get its fair share of the placenta, the risk that blood volume will be shared unequally, and an overall risk of vascular instability in each twin.

When the predominant issue in an identical twin pregnancy is unequal placental sharing, the growth of one baby becomes restricted and the other baby grows normally, resulting in a condition called selective intrauterine growth restriction (selective IUGR).

The other main issue—that of unequal blood-volume sharing—is what fuels TTTS. In uncomplicated pregnancies, blood is exchanged equally through the vascular anastomoses that characterize all monochorionic pregnancies. In complicated pregnancies, however, the exchange is unbalanced, and blood is shared in one direction without adequate return.

Arteries emanating from the placental cord insertion of one twin, for instance, can drain into a vein returning to the other twin. Such arteriovenous anastomoses are in the substance of the placenta and act as one-way valves for blood flow. If the amount of blood flow in one direction is not balanced by enough flow in the opposite direction—that is, if the magnitude of blood flow through unidirectional arteriovenous anastomoses is not compensated by vascular channels that permit flow in the opposite direction—then an imbalance develops that is potentially harmful to both babies.

In TTTS, which develops in about 15% of monochorionic pregnancies, the imbalance progresses to the extent that one twin becomes a “donor” of blood volume and the other becomes the “recipient” twin.

The donor twin moves blood across the anastomoses to the placenta and to the recipient twin, and does not receive an equal amount in return. A decline in blood volume leads to decreased urine output to the extent that, eventually, bladder filling in the donor twin virtually ceases. Under these circumstances, oligohydramnios may progress to anhydramnios, and the twin may become “stuck” in an essentially empty amniotic sac.

The recipient twin, in the meantime, receives an excess amount of venous blood volume. The increase in intravascular blood volume drives an increase of filtration in the kidneys, which results in excess urination. The increased urinary frequency, which may even result in constant bladder filling, leads to polyhydramnios.

When the sac of the recipient twin becomes distended by amniotic fluid, and the donor twin is no longer producing urine, the membrane between the twins may become wrapped so tightly around the donor twin that it is barely visible on ultrasound (See image below.) When the donor twin is “stuck” to the uterine wall in such a way, the ultrasound appearance resembles that of identical twins with one amniotic cavity (monoamniotic twins).

Untreated TTTS has serious consequences for each twin and for the whole pregnancy. First, the resultant polyhydramnios can stimulate preterm labor because of uterine distention. Second, abnormalities in blood volume can lead to cardiac problems and cardiovascular compromise for the babies, most often for the recipient twin. The excess blood cells and volume overload that this twin faces can lead to cardiac failure and hydrops.

The donor twin, meanwhile, is at risk for abnormalities and long-term effects resulting from compression, failing placental function, malnutrition, and hypovolemia.

If one baby dies in utero, the placental anastomoses that cause TTTS in the first place—that is, the open vessel connections that exist between the twins—carry an additional danger. In artery-to-artery and vein-to-vein anastomoses, the direction of blood flow is determined by the difference in blood pressure on either side. If one twin dies, the resultant drop in blood pressure causes the surviving twin to lose a large amount of blood volume across the connecting vessels and into the dying twin. This puts the surviving twin at risk of hemorrhagic shock and a heart attack or stroke.

It is estimated that the risk for white-matter injury in the surviving twin at the time of birth may be as high as 50% following such an intrauterine event. The fates of both twins are thus essentially linked to each other through their placental anastomoses.

Although exact contributors still need to be determined, it is well established that, compared with nonidentical twins, identical twins have a higher incidence of cerebral palsy and other anomalies, and a higher rate of developmental delay at 2 years. Because the development of TTTS is one well-recognized contributor to these statistics, perinatal interventions in monochorionic pregnancies have primarily focused on its treatment.

Evolution of Management

It's most interesting to look at the evolution of management from a historical perspective. When TTTS was clinically recognized, before the days of multivessel Doppler assessment, patients would most often present with a massively distended uterus and preterm labor.

The natural management approach was amnioreduction, which involved the removal of large volumes of amniotic fluid in an effort to relieve uterine distention and prevent preterm delivery. Physicians recognized the need for serial amnioreduction, as the procedure leaves anastomoses open and does nothing to address the underlying problem.

This approach was often satisfactory when it was started at 26–27 weeks' gestation because chances to prolong pregnancy to 32–34 weeks with repeated drainage were reasonable. The patients who presented with massive polyhydramnios and severe TTTS at 20 weeks, however, were another story. Their outcomes with serial amnioreduction were poor; in fact, many physicians would offer pregnancy termination under these circumstances.

In the late 1990s several groups began to address the underlying problem by closing the problem vessels. Dr. Julian De Lia, at that time practicing in Utah, was the first to describe fetoscopic laser ablation of placental anastomoses. He and the team of Prof. Kypros Nicolaides in Europe used a nonselective technique that involved ablating blood vessels and the placental mass along a dividing line between the twins—essentially making the placenta functionally dichorionic—and then draining the amniotic fluid.

Developmental research on the equipment and modification of the technique proceeded. In 1999, Dr. Rubén A. Quintero in Florida published a five-stage classification system for the progression of TTTS, with stages I and II characterized primarily by imbalances in blood volume, stages III and IV signified by cardiovascular compromise, and stage V signified by the death of one or both twins.

This staging system marked a significant step in the management of TTTS because it established a unified diagnostic approach that was based on prenatal criteria. Until this point, the definitions of TTTS were based on an extrapolation of pediatric diagnostic criteria that were used at birth. The application of Dr. Quintero's staging system allowed a more objective comparison of treatment strategies, but required familiarity with arterial and venous Doppler techniques.

Dr. Quintero also argued that a nonselective approach with the laser—one that coagulates vessels that do not contribute to TTTS, as well as those that do—can rob one or both twins of placental territory that is vital for their survival. He developed a selective laser technique that involves identification and coagulation of the vessels that pass from one twin to the other, leaving normal placental territory and noncontributing vessels untouched.

In the meantime, the Eurofetus research consortium had formed in Europe, and had begun designing a trial to compare laser therapy with amnioreduction, with one of their premises being that laser therapy would most benefit twin pregnancies that are complicated by TTTS before 26 weeks' gestation. Perinatal mortality for untreated severe TTTS, they knew, was as high as 90%, with significant handicap in the survivors.

Results of the multicenter randomized study were published in 2004 (N. Engl. J. Med. 2004;351:136-44). Complication rates were basically comparable (approximately 9% in each arm), but the rates of survival of at least one twin at 28 days and at 6 months of age were significantly better in the group that underwent selective laser coagulation than in the amnioreduction group (76% vs. 56% at 28 days, and 76% vs. 51% at 6 months).

The differences existed in both the early and later stages of TTTS, although fetuses in the Quintero stages I or II had better outcomes than did those with higher stages in both treatment groups. (The study had been concluded early, after 72 women had been assigned to the laser group and 70 to the amnioreduction group, when an interim analysis demonstrated significant benefits.)

Gestational ages at the time of delivery were also significantly different: Patients in the laser group delivered, on average, at 33 weeks, whereas those in the amnioreduction group delivered at 29 weeks.

An intermediate-term look at neurologic outcomes favored laser surgery as well: At 6 months of age, infants in the laser group were more likely than those in the amnioreduction group to be free of neurologic deficits (52% vs. 31%, respectively).

At the center for advanced fetal care at the University of Maryland, Baltimore, which has served for almost a decade as a referral resource for minimally invasive fetal therapy, I have applied the identical technique utilized in the Eurofetus trial using a selective approach. Our treatment results have consistently mirrored the published statistics.

Our research, which we presented at the annual meeting of the Society for Maternal-Fetal Medicine, confirms that successful laser ablation corrects the abnormal blood volume distribution. This effect is first apparent for the donor twin and clinically presents with the reappearance of bladder filling, often on the day after the procedure.

Urination gradually normalizes in the recipient twin, typically over 1–2 weeks after the procedure. The mother feels better immediately after the procedure and continues to improve as fetal status normalizes.

Longer-term follow-up of neurologic abnormalities in the Eurofetus trial is underway. For now, however, an analysis of a series of patients who received intrauterine laser treatment for TTTS has shown that 78% of 89 surviving children had a normal neurodevelopmental status at about 2 years of age, whereas 11% had minor neurologic deficiencies and 11% had major neurologic deficits (Am. J. Obstet. Gynecol. 2003;188:876-80).

Although comparisons of patients managed in the randomized trial are pending, these rates of neurologic handicap compare favorably with those seen after amnioreduction.

Two large series indicate that severe TTTS is associated with poor neurodevelopment, and that up to 27% of survivors may have abnormal brain ultrasounds at the time of delivery. It is therefore widely accepted that the neuroprotective benefit of laser therapy is most marked in early onset TTTS (prior to 26 weeks), and that the difference in outcomes is attributable to lower rates of preterm delivery and prematurity-associated complications as well as to the elimination of the risks of ongoing TTTS.

Moving Into the Future

In Europe, the randomized trial basically brought the controversy over optimal treatment for TTTS to a close. In the United States, there are some who still lean toward performing an initial amnioreduction and moving on to laser surgery if necessary.

There are disadvantages to such an approach. An initial amnioreduction removes the amniotic fluid pocket that is necessary to successfully maneuver the fetoscope. Decompression of the placenta not only unpredictably affects shunt dynamics but also can create placental “valleys” that can impair visualization of anastomoses. Potential bleeding from the procedure, as well as advancing gestational age until a suitable fluid pocket has reestablished, can also make the fluid cloudier.

Investigators who have looked at the factors that influence outcomes of selective laser coagulation of placental anastomoses have reported that those who do poorly have more advanced TTTS; have shorter cervical length, and thus a higher incidence of preterm labor; have a history of prior amnioreduction; and have technically difficult laser procedures with poor visualization of anastomoses as contributing factors.

Amnioreduction still has a role, however, particularly for patients who present with TTTS beyond 26 weeks' gestation. These patients are not candidates for laser therapy because the efficacy and safety of the procedure at this gestational age has not been studied.

Even with the improved outcomes, the therapies are still not optimal, and our knowledge of TTTS is still full of gaps and differences in opinion. Some experts believe, for instance, that with selective laser therapy there is a risk of recurring TTTS—that is, as visible anastomoses are closed, intravascular pressures are diverted to very small vessels that are barely visible at the time of the laser procedure. Over time, it is believed, these vessels may expand and therefore become hemodynamically relevant contributors to recurring TTTS.

At this time, I believe it's important to keep an open mind after presumably successful laser therapy, and to follow the fetuses closely after surgery for TTTS. Continued evaluation of bladder filling, amniotic fluid volumes, and placental and venous Doppler studies may be necessary over extended periods of time.

It is also important to inform neonatologists and referring obstetricians of the special circumstances of these babies, who behave very differently—both in the NICU and beyond—than do other babies of similar size or with other underlying conditions.

Babies who matured in utero as TTTS “recipients” are chronically hypervolemic and will not respond well, for instance, to dopamine given in the NICU as the primary agent to boost blood pressure. Careful attention to fluid balance is essential to prevent neonatal complications under these circumstances.

Fortunately, technologic advances in equipment are making intrauterine therapy much more minimally invasive. The development of fetoscopes with a 2-mm lens offers superior visual resolution and facilitates a minimally invasive approach. Digital camera technology also enhances the visualization of the smallest blood vessels. Steerable and angulated optical devices tackle the problems of anterior placenta. The smaller caliber of the entry site also decreases the risk for complications.

Today, laser surgery is typically performed under local anesthesia that requires minimal hospitalization with only perioperative tocolysis. The average length of patient stay is 1 day at the University of Maryland Medical Center.

The initiation of the North American Fetal Therapy Network (NAFTNet), a research consortium, is a significant development in the United States.

The membrane is seen folding around the donor who becomes a “stuck twin.” Courtesy Dr. Ahmet A. Baschat

When bladder filling of the donor can no longer be demonstrated, progression to stage 2 TTTS is diagnosed. (Left arrow, small bladder; right arrow, empty bladder.)

Critically abnormal waveforms in the umbilical artery (left) and ductus venosus (right) indicate stage 3 TTTS.

Ultrasound findings of hydrops (arrows point to fluid in fetal abdomen) indicate stage 4 TTTS. Photos courtesy Dr. Ahmet A. Baschat

Twin-to-Twin Transfusion Syndrome

Despite the advances that have occurred in obstetrics over the years, who would have imagined that fetal surgery would be a viable therapeutic approach today? Well, indeed, this is where we are in the history of obstetrics.

Fetal evaluation has been conducted over the years using a variety of noninvasive techniques, most notably electronic fetal monitoring. More invasive techniques, such as amniocentesis, have also been used with very good success and with relatively low risk to mother and fetus. Certain conditions, however, cannot be addressed with noninvasive or slightly invasive approaches, but rather require either open surgery or more involved surgery of a minimally invasive nature.

One of these conditions is twin-to-twin transfusion syndrome, in which one of the fetuses may succumb during intrauterine life. Over the years, amniocentesis has been used with limited success. However, newer techniques involving endoscopic laser therapy are being introduced with improved outcomes. In this Master Class, we review both modalities in the management of these patients, with careful attention to advances in fetal laser therapy.

We are pleased to introduce Dr. Ahmet A. Baschat, of the department of obstetrics, gynecology, and reproductive sciences at the University of Maryland School of Medicine, Baltimore, as our guest professor this month. Dr. Baschat is considered a national expert in fetal therapy, including laser and other intrauterine surgical procedures.

EMILY BRANNAN, ILLUSTRATION

Patients are bearing children at older ages and are more active than the obstetric populations of generations ago. They are in the workplace, out on the roads, and exposed to common causes of trauma, such as motor vehicle accidents and falls.

It is helpful to know the numbers and be aware of the significance of the problem. Trauma in pregnancy is significantly more frequent, for instance, than the genetic problems we encounter and screen for in our practices.

As ob.gyns., we are bound to be called to treat trauma in pregnancy at some point, and it is important that we be prepared to optimally manage the seriously injured pregnant woman as either a primary manager or as an advisor or consultant, whatever the situation and phase of evaluation and treatment demand.

The Causes

The vast majority of women who experience trauma (95%–96%) suffer injuries from blunt, rather than penetrating, types of trauma. Most of these injuries, in turn, are a result of motor vehicle accidents.

Motor vehicle accidents are the leading cause of death in women aged 12-51 years and the leading cause of trauma in women of childbearing age, accounting for well over half of the major injuries experienced by pregnant women.

Falls are the second most common cause of blunt trauma.

Although relatively infrequent, penetrating trauma—often caused by gunshot wounds—is still a reality. Penetrating trauma leads to significant injury to the fetus more often than to the mother, because the mother's abdominal organs are shielded by the gravid uterus.

As ob.gyns., we can help our patients decrease injury in motor vehicle accidents by urging them to use seat belts. When worn correctly—with the lap belt fitting under the belly and close to the hips, and the shoulder belt resting between the breasts and over the shoulder—seat belts with shoulder restraints can definitively reduce the risk of death and injury for both the mother and fetus.

It is also important to remember that domestic or interpersonal violence may be more frequent during pregnancy than at other times. Such abuse is a less evident cause of blunt trauma, but a very real one. The incidence of interpersonal violence can be as high as 14%–20% in pregnant teenagers, and probably averages about 10%–11% in pregnant women overall.

When taking care of women who report injuries from falls and other events that do not correlate with the overall history or physical exam, we must—in a private and safe environment—address the possibility of partner abuse. We should also remind ourselves that women who are physically abused have a higher incidence of infection, low maternal weight gain, maternal alcohol and drug abuse, and low-birth-weight babies.

Physiologic Changes

Ob.gyns. will sometimes serve as consultants or advisors in managing trauma during pregnancy, and at other times will serve as primary managers. In any case, optimal evaluation and management require both teamwork (an integrated effort of multiple specialties) and a central role for the ob.gyn., whose understanding of the physiologic changes of pregnancy is vital to management decisions.

Key changes that occur secondary to pregnancy most often alter the patient's cardiovascular, hematologic, respiratory, urinary, gastrointestinal, and endocrine systems.

Cardiac output increases during pregnancy by 1-1.5 L/min, with a dramatic increase in the percentage of cardiac output that goes to the uterus. By week 36, the uterus receives up to 600 mL/min of blood—which represents about a tenth of the mother's cardiac output—compared with 60 mL/min in the nonpregnant state. Consequently, trauma to the uterus can result in significant hemorrhage and shock.

Blood volume increases by 45%–50% during pregnancy, and can be instrumental in concealing signs of shock from hypovolemia. Along with this, however, the amount of all clotting factors also increases, which predisposes the pregnant patient to embolic risks as well as coagulopathy from disseminated intravascular coagulation (DIC).

The respiratory system adapts for its role of oxygen delivery to the fetus. Tidal volume increases, with an overall increase in minute ventilation. Changes that result from this adaptation are a lowering of the maternal CO2 level and a decrease in alveolar residual volume. This can further result in a respiratory alkalotic state, which is corrected with a renal decrease in bicarbonate (a compensated respiratory alkalosis). The decreased residual volumes render the pregnant patient more susceptible to alveolar collapse and respiratory compromise.

Under these circumstances, it may be prudent to consider early intubation of the pregnant patient with respiratory compromise in order to preserve the exchange of gases across the fetoplacental unit.

Renal blood flow is increased in pregnancy, with a concomitant increase in creatinine clearance and a tendency to more rapidly clear drugs that are renal dependent.

Standard lab values must be seen in the context of the changes in maternal physiology. For example, blood gas values in pregnancy may reveal lower CO2 values, and hematocrits may be lower secondary to the hemodilution seen in blood volume expansion.

During evaluation or resuscitation, it is important for ob.gyns. to remind the team of the mass of the gravid uterus and its ability—when the woman is in the supine position—to compromise the return of blood flow to the heart by compressing the vena cava.

Beyond the second trimester, the patient should be tilted to the left by approximately 35 degrees, and when a full-body tilt is not possible, left lateral displacement of the uterus must be maintained. When spinal injury is suspected, care must be taken to keep the spine and neck aligned during tilting.

Questions About Imaging

The obstetrician often will be asked whether the imaging needed to diagnose various injuries is safe for the fetus. Doses of radiation used during trauma care and evaluation—for example, CT scans of the pelvis or chest, and chest x-rays—are usually in the range of less than 250 mGy, which is considered to be the intermediate range of exposure and reasonably safe for the fetus.

Whenever possible, the pelvis of a pregnant woman should be shielded, especially in the first trimester. It is important to know, however, that much of today's imaging equipment is faster than previous technology and therefore delivers much lower radiation exposure with more information in a single pass. Additionally, MRI has been shown to be a safe modality in pregnancy.

The bottom line is that imaging studies that are needed for the care of the critically injured patient should not be withheld because she may be pregnant.

Primary Assessment

On initial presentation, all efforts for the pregnant trauma patient must first be directed toward stabilizing the mother and maintaining oxygen delivery, with the ABCs (airway, breathing, and circulation) of trauma care being the first priority.

Women who are pregnant have a reduced ability to compensate for respiratory compromise. Maintaining a patent airway is critical for both maternal and fetal oxygenation, and this very well may require early intubation. We must make sure that the patient is moving oxygen in, with oxygen saturation levels better than 90%.

The fetus's oxygen uptake depends directly on oxygen delivery via uterine blood flow, so circulation—both to the mother's vital organs and to the uterus—is also key. Because of shunting and vasospasm, significant uterine blood flow compromise may exist even with normal-appearing blood pressure. Thus, it is important to control any significant bleeding and pursue vigorous volume replacement.

Only after the ABCs are addressed—and readdressed for effectiveness—can we turn our attention elsewhere. If we were to add a “D” to the trauma protocol, it would stand for “disability” and would involve a rapid neurologic evaluation to assess for any neurologic injury. It is worth considering that neurologic impairment in late pregnancy may be secondary to an eclamptic seizure that may have led to the trauma event.

Other often critical components of trauma injuries, such as fractures and intraperitoneal hemorrhage, are usually evaluated almost simultaneously by the trauma team.

Patients with pelvic fractures (common in motor vehicle accidents) are at risk of having retroperitoneal hemorrhage, which is not always obvious and requires careful diagnosis. A pelvic exam can reveal signs of lower pelvic fracture and possible vaginal lacerations from protruding bone fragments.

The Secondary Survey

Once we have stabilized the mother and evaluated her for other critical signs of trauma, we can turn our attention to fetal assessment. First, we should assess gestational age, either by taking a history if someone close to the mother is present, or through ultrasonography.

Ultrasonography is an important tool at this point for assessing several factors in short order. In addition to assessing the viability of the fetus, we can evaluate the intrauterine fluid volume and the placental location. (The question of viability, of course, depends on the level of neonatal intensive care services available).

A low amount of amniotic fluid should lead us to suspect rupture of the amniotic membranes or, in some cases, uterine rupture secondary to trauma. Although with expert hands it is possible to detect relatively small placental abruptions, abruptions are usually apparent only with larger separations.

Ultrasonography can also be used in determining intra-abdominal free fluid which is consistent with intraperitoneal hemorrhage.

When the woman is severely injured and needs surgery and if delivery of the fetus is unnecessary, we should focus on monitoring the fetus in the operating room. We can do so with an ultrasound probe or a fetal Doppler encased in a sterile sleeve.

After surgery—or when the woman does not need surgery and is considered stable enough to undergo observation—continuous monitoring of the viable fetus, with a longer-term view, should be done using external monitoring of fetal heart rate and uterine monitoring for signs of preterm labor. Frequent uterine contractions should be followed closely, and cervical dilatation should be evaluated.

The ob.gyn. becomes the primary provider when a viable fetus shows signs of fetal compromise that necessitate delivery, or when uterine rupture occurs, in which case urgent intervention is necessary for the mother.

Many studies have attempted to address the question of how long trauma patients should be monitored, and many guidelines have been proposed. In general, we can conclude that 12 hours of observation is adequate for stable patients who are not contracting and have reassuring fetal tracings and no signs of bleeding.

On the other hand, patients who are contracting, who have a nonreassuring fetal tracing, or who have had any form of vaginal bleeding should be observed for a minimum of 24 hours. The more severe the injury to the mother, the more likely there is to be an injury to the fetus, and the higher the risk that a placental abruption or other serious complication may surface.

If a patient is discharged after observation, regardless of the severity of trauma, she must be given precautions regarding any changes in fetal movement or the development of abdominal pain, vaginal bleeding, or fluid loss. Any one of these changes should prompt her to return for evaluation immediately.

The Kleihauer-Betke test may be useful in evaluating the degree of fetal-maternal hemorrhage and the amount of Rh immune globulin that may be needed in the Rh-negative mother. In general, Rh-negative mothers who are involved in trauma should be given a single dose of Rh immune globulin unless it is known that the fetus is Rh negative.

Cardiac arrest is sometimes the tragic outcome for a posttrauma victim. Maternal resuscitation should be undertaken immediately. If maternal resuscitation is not thought likely to be effective, and the fetus is considered viable, the best outcome for fetal survival occurs if delivery can be accomplished within approximately 5 minutes from arrest. Beyond this time, there is diminishing return for fetal survival.

Organized, rapid assessment and intervention hold the key to the best outcomes for the pregnant patient involved in a trauma. Following the rules of trauma resuscitation (those ABCs) provides the best chance of successful treatment of the mother, which in turn provides the best chance of a favorable fetal outcome.

Motor vehicle accidents are the leading cause of trauma and death in women of childbearing age. Courtesy Dr. Hugh Mighty

One situation in which the ob.gyn. becomes the primary provider is when delivery is necessary. Courtesy Dr. Hugh Mighty

Trauma in Pregnancy

High-risk obstetrics by its very nature involves a wide spectrum of diseases and events that complicate pregnancies and preclude or prevent their normal progression. Sometimes, high-risk obstetrics involves physical trauma that is inflicted externally upon an otherwise normal pregnancy.

Physical trauma is, in fact, one of the leading causes of morbidity and mortality during pregnancy. It has been estimated that physical trauma complicates approximately 1 in every 12 pregnancies—a staggering figure and one that we may not fully appreciate or think about often enough.

According to the American College of Obstetricians and Gynecologists, approximately two-thirds of all trauma during pregnancy in industrialized nations results from motor vehicle crashes.

Women not only are more likely to be involved in automobile accidents than are male drivers; they also are increasingly more likely to be victims of violence. In 1994, women were about six times more likely than men to be victims of violence—a significant increase from more than a decade before, when women were half as likely as men to be victims of violence.

Trauma in pregnant women has to be dealt with by a variety of specialists. Very often, these patients will present to emergency departments or urgent care centers, and will have to be seen by emergency physicians, surgeons, or general practitioners. This is a challenging situation and one that presents unusual challenges for obstetric staff.

Especially as the number of patients with traumatic injuries and complications increases, it is important that we review some of the key types of presentations and complications of trauma in pregnancy, and discuss how we may best develop therapeutic algorithms for dealing with them.

It is in this light that we have invited Dr. Hugh E. Mighty, chairman of the department of obstetrics, gynecology, and reproductive sciences at the University of Maryland, Baltimore, to discuss the management of seriously injured pregnant women.

Dr. Mighty is not only a maternal-fetal medicine specialist, but is certified in critical care medicine as well. We are pleased to welcome him as an expert on trauma in pregnancy and as this month's guest professor on the subject.

Patients are bearing children at older ages and are more active than the obstetric populations of generations ago. They are in the workplace, out on the roads, and exposed to common causes of trauma, such as motor vehicle accidents and falls.

It is helpful to know the numbers and be aware of the significance of the problem. Trauma in pregnancy is significantly more frequent, for instance, than the genetic problems we encounter and screen for in our practices.

As ob.gyns., we are bound to be called to treat trauma in pregnancy at some point, and it is important that we be prepared to optimally manage the seriously injured pregnant woman as either a primary manager or as an advisor or consultant, whatever the situation and phase of evaluation and treatment demand.

The Causes

The vast majority of women who experience trauma (95%–96%) suffer injuries from blunt, rather than penetrating, types of trauma. Most of these injuries, in turn, are a result of motor vehicle accidents.

Motor vehicle accidents are the leading cause of death in women aged 12-51 years and the leading cause of trauma in women of childbearing age, accounting for well over half of the major injuries experienced by pregnant women.

Falls are the second most common cause of blunt trauma.

Although relatively infrequent, penetrating trauma—often caused by gunshot wounds—is still a reality. Penetrating trauma leads to significant injury to the fetus more often than to the mother, because the mother's abdominal organs are shielded by the gravid uterus.

As ob.gyns., we can help our patients decrease injury in motor vehicle accidents by urging them to use seat belts. When worn correctly—with the lap belt fitting under the belly and close to the hips, and the shoulder belt resting between the breasts and over the shoulder—seat belts with shoulder restraints can definitively reduce the risk of death and injury for both the mother and fetus.

It is also important to remember that domestic or interpersonal violence may be more frequent during pregnancy than at other times. Such abuse is a less evident cause of blunt trauma, but a very real one. The incidence of interpersonal violence can be as high as 14%–20% in pregnant teenagers, and probably averages about 10%–11% in pregnant women overall.

When taking care of women who report injuries from falls and other events that do not correlate with the overall history or physical exam, we must—in a private and safe environment—address the possibility of partner abuse. We should also remind ourselves that women who are physically abused have a higher incidence of infection, low maternal weight gain, maternal alcohol and drug abuse, and low-birth-weight babies.

Physiologic Changes

Ob.gyns. will sometimes serve as consultants or advisors in managing trauma during pregnancy, and at other times will serve as primary managers. In any case, optimal evaluation and management require both teamwork (an integrated effort of multiple specialties) and a central role for the ob.gyn., whose understanding of the physiologic changes of pregnancy is vital to management decisions.

Key changes that occur secondary to pregnancy most often alter the patient's cardiovascular, hematologic, respiratory, urinary, gastrointestinal, and endocrine systems.

Cardiac output increases during pregnancy by 1-1.5 L/min, with a dramatic increase in the percentage of cardiac output that goes to the uterus. By week 36, the uterus receives up to 600 mL/min of blood—which represents about a tenth of the mother's cardiac output—compared with 60 mL/min in the nonpregnant state. Consequently, trauma to the uterus can result in significant hemorrhage and shock.

Blood volume increases by 45%–50% during pregnancy, and can be instrumental in concealing signs of shock from hypovolemia. Along with this, however, the amount of all clotting factors also increases, which predisposes the pregnant patient to embolic risks as well as coagulopathy from disseminated intravascular coagulation (DIC).

The respiratory system adapts for its role of oxygen delivery to the fetus. Tidal volume increases, with an overall increase in minute ventilation. Changes that result from this adaptation are a lowering of the maternal CO2 level and a decrease in alveolar residual volume. This can further result in a respiratory alkalotic state, which is corrected with a renal decrease in bicarbonate (a compensated respiratory alkalosis). The decreased residual volumes render the pregnant patient more susceptible to alveolar collapse and respiratory compromise.

Under these circumstances, it may be prudent to consider early intubation of the pregnant patient with respiratory compromise in order to preserve the exchange of gases across the fetoplacental unit.

Renal blood flow is increased in pregnancy, with a concomitant increase in creatinine clearance and a tendency to more rapidly clear drugs that are renal dependent.

Standard lab values must be seen in the context of the changes in maternal physiology. For example, blood gas values in pregnancy may reveal lower CO2 values, and hematocrits may be lower secondary to the hemodilution seen in blood volume expansion.

During evaluation or resuscitation, it is important for ob.gyns. to remind the team of the mass of the gravid uterus and its ability—when the woman is in the supine position—to compromise the return of blood flow to the heart by compressing the vena cava.

Beyond the second trimester, the patient should be tilted to the left by approximately 35 degrees, and when a full-body tilt is not possible, left lateral displacement of the uterus must be maintained. When spinal injury is suspected, care must be taken to keep the spine and neck aligned during tilting.

Questions About Imaging

The obstetrician often will be asked whether the imaging needed to diagnose various injuries is safe for the fetus. Doses of radiation used during trauma care and evaluation—for example, CT scans of the pelvis or chest, and chest x-rays—are usually in the range of less than 250 mGy, which is considered to be the intermediate range of exposure and reasonably safe for the fetus.

Whenever possible, the pelvis of a pregnant woman should be shielded, especially in the first trimester. It is important to know, however, that much of today's imaging equipment is faster than previous technology and therefore delivers much lower radiation exposure with more information in a single pass. Additionally, MRI has been shown to be a safe modality in pregnancy.

The bottom line is that imaging studies that are needed for the care of the critically injured patient should not be withheld because she may be pregnant.

Primary Assessment

On initial presentation, all efforts for the pregnant trauma patient must first be directed toward stabilizing the mother and maintaining oxygen delivery, with the ABCs (airway, breathing, and circulation) of trauma care being the first priority.

Women who are pregnant have a reduced ability to compensate for respiratory compromise. Maintaining a patent airway is critical for both maternal and fetal oxygenation, and this very well may require early intubation. We must make sure that the patient is moving oxygen in, with oxygen saturation levels better than 90%.

The fetus's oxygen uptake depends directly on oxygen delivery via uterine blood flow, so circulation—both to the mother's vital organs and to the uterus—is also key. Because of shunting and vasospasm, significant uterine blood flow compromise may exist even with normal-appearing blood pressure. Thus, it is important to control any significant bleeding and pursue vigorous volume replacement.

Only after the ABCs are addressed—and readdressed for effectiveness—can we turn our attention elsewhere. If we were to add a “D” to the trauma protocol, it would stand for “disability” and would involve a rapid neurologic evaluation to assess for any neurologic injury. It is worth considering that neurologic impairment in late pregnancy may be secondary to an eclamptic seizure that may have led to the trauma event.

Other often critical components of trauma injuries, such as fractures and intraperitoneal hemorrhage, are usually evaluated almost simultaneously by the trauma team.

Patients with pelvic fractures (common in motor vehicle accidents) are at risk of having retroperitoneal hemorrhage, which is not always obvious and requires careful diagnosis. A pelvic exam can reveal signs of lower pelvic fracture and possible vaginal lacerations from protruding bone fragments.

The Secondary Survey

Once we have stabilized the mother and evaluated her for other critical signs of trauma, we can turn our attention to fetal assessment. First, we should assess gestational age, either by taking a history if someone close to the mother is present, or through ultrasonography.

Ultrasonography is an important tool at this point for assessing several factors in short order. In addition to assessing the viability of the fetus, we can evaluate the intrauterine fluid volume and the placental location. (The question of viability, of course, depends on the level of neonatal intensive care services available).

A low amount of amniotic fluid should lead us to suspect rupture of the amniotic membranes or, in some cases, uterine rupture secondary to trauma. Although with expert hands it is possible to detect relatively small placental abruptions, abruptions are usually apparent only with larger separations.

Ultrasonography can also be used in determining intra-abdominal free fluid which is consistent with intraperitoneal hemorrhage.

When the woman is severely injured and needs surgery and if delivery of the fetus is unnecessary, we should focus on monitoring the fetus in the operating room. We can do so with an ultrasound probe or a fetal Doppler encased in a sterile sleeve.

After surgery—or when the woman does not need surgery and is considered stable enough to undergo observation—continuous monitoring of the viable fetus, with a longer-term view, should be done using external monitoring of fetal heart rate and uterine monitoring for signs of preterm labor. Frequent uterine contractions should be followed closely, and cervical dilatation should be evaluated.

The ob.gyn. becomes the primary provider when a viable fetus shows signs of fetal compromise that necessitate delivery, or when uterine rupture occurs, in which case urgent intervention is necessary for the mother.

Many studies have attempted to address the question of how long trauma patients should be monitored, and many guidelines have been proposed. In general, we can conclude that 12 hours of observation is adequate for stable patients who are not contracting and have reassuring fetal tracings and no signs of bleeding.

On the other hand, patients who are contracting, who have a nonreassuring fetal tracing, or who have had any form of vaginal bleeding should be observed for a minimum of 24 hours. The more severe the injury to the mother, the more likely there is to be an injury to the fetus, and the higher the risk that a placental abruption or other serious complication may surface.

If a patient is discharged after observation, regardless of the severity of trauma, she must be given precautions regarding any changes in fetal movement or the development of abdominal pain, vaginal bleeding, or fluid loss. Any one of these changes should prompt her to return for evaluation immediately.

The Kleihauer-Betke test may be useful in evaluating the degree of fetal-maternal hemorrhage and the amount of Rh immune globulin that may be needed in the Rh-negative mother. In general, Rh-negative mothers who are involved in trauma should be given a single dose of Rh immune globulin unless it is known that the fetus is Rh negative.

Cardiac arrest is sometimes the tragic outcome for a posttrauma victim. Maternal resuscitation should be undertaken immediately. If maternal resuscitation is not thought likely to be effective, and the fetus is considered viable, the best outcome for fetal survival occurs if delivery can be accomplished within approximately 5 minutes from arrest. Beyond this time, there is diminishing return for fetal survival.

Organized, rapid assessment and intervention hold the key to the best outcomes for the pregnant patient involved in a trauma. Following the rules of trauma resuscitation (those ABCs) provides the best chance of successful treatment of the mother, which in turn provides the best chance of a favorable fetal outcome.

Motor vehicle accidents are the leading cause of trauma and death in women of childbearing age. Courtesy Dr. Hugh Mighty

One situation in which the ob.gyn. becomes the primary provider is when delivery is necessary. Courtesy Dr. Hugh Mighty

Trauma in Pregnancy

High-risk obstetrics by its very nature involves a wide spectrum of diseases and events that complicate pregnancies and preclude or prevent their normal progression. Sometimes, high-risk obstetrics involves physical trauma that is inflicted externally upon an otherwise normal pregnancy.

Physical trauma is, in fact, one of the leading causes of morbidity and mortality during pregnancy. It has been estimated that physical trauma complicates approximately 1 in every 12 pregnancies—a staggering figure and one that we may not fully appreciate or think about often enough.

According to the American College of Obstetricians and Gynecologists, approximately two-thirds of all trauma during pregnancy in industrialized nations results from motor vehicle crashes.

Women not only are more likely to be involved in automobile accidents than are male drivers; they also are increasingly more likely to be victims of violence. In 1994, women were about six times more likely than men to be victims of violence—a significant increase from more than a decade before, when women were half as likely as men to be victims of violence.

Trauma in pregnant women has to be dealt with by a variety of specialists. Very often, these patients will present to emergency departments or urgent care centers, and will have to be seen by emergency physicians, surgeons, or general practitioners. This is a challenging situation and one that presents unusual challenges for obstetric staff.

Especially as the number of patients with traumatic injuries and complications increases, it is important that we review some of the key types of presentations and complications of trauma in pregnancy, and discuss how we may best develop therapeutic algorithms for dealing with them.

It is in this light that we have invited Dr. Hugh E. Mighty, chairman of the department of obstetrics, gynecology, and reproductive sciences at the University of Maryland, Baltimore, to discuss the management of seriously injured pregnant women.

Dr. Mighty is not only a maternal-fetal medicine specialist, but is certified in critical care medicine as well. We are pleased to welcome him as an expert on trauma in pregnancy and as this month's guest professor on the subject.

Patients are bearing children at older ages and are more active than the obstetric populations of generations ago. They are in the workplace, out on the roads, and exposed to common causes of trauma, such as motor vehicle accidents and falls.

It is helpful to know the numbers and be aware of the significance of the problem. Trauma in pregnancy is significantly more frequent, for instance, than the genetic problems we encounter and screen for in our practices.

As ob.gyns., we are bound to be called to treat trauma in pregnancy at some point, and it is important that we be prepared to optimally manage the seriously injured pregnant woman as either a primary manager or as an advisor or consultant, whatever the situation and phase of evaluation and treatment demand.

The Causes

The vast majority of women who experience trauma (95%–96%) suffer injuries from blunt, rather than penetrating, types of trauma. Most of these injuries, in turn, are a result of motor vehicle accidents.

Motor vehicle accidents are the leading cause of death in women aged 12-51 years and the leading cause of trauma in women of childbearing age, accounting for well over half of the major injuries experienced by pregnant women.

Falls are the second most common cause of blunt trauma.

Although relatively infrequent, penetrating trauma—often caused by gunshot wounds—is still a reality. Penetrating trauma leads to significant injury to the fetus more often than to the mother, because the mother's abdominal organs are shielded by the gravid uterus.

As ob.gyns., we can help our patients decrease injury in motor vehicle accidents by urging them to use seat belts. When worn correctly—with the lap belt fitting under the belly and close to the hips, and the shoulder belt resting between the breasts and over the shoulder—seat belts with shoulder restraints can definitively reduce the risk of death and injury for both the mother and fetus.

It is also important to remember that domestic or interpersonal violence may be more frequent during pregnancy than at other times. Such abuse is a less evident cause of blunt trauma, but a very real one. The incidence of interpersonal violence can be as high as 14%–20% in pregnant teenagers, and probably averages about 10%–11% in pregnant women overall.

When taking care of women who report injuries from falls and other events that do not correlate with the overall history or physical exam, we must—in a private and safe environment—address the possibility of partner abuse. We should also remind ourselves that women who are physically abused have a higher incidence of infection, low maternal weight gain, maternal alcohol and drug abuse, and low-birth-weight babies.

Physiologic Changes

Ob.gyns. will sometimes serve as consultants or advisors in managing trauma during pregnancy, and at other times will serve as primary managers. In any case, optimal evaluation and management require both teamwork (an integrated effort of multiple specialties) and a central role for the ob.gyn., whose understanding of the physiologic changes of pregnancy is vital to management decisions.

Key changes that occur secondary to pregnancy most often alter the patient's cardiovascular, hematologic, respiratory, urinary, gastrointestinal, and endocrine systems.

Cardiac output increases during pregnancy by 1-1.5 L/min, with a dramatic increase in the percentage of cardiac output that goes to the uterus. By week 36, the uterus receives up to 600 mL/min of blood—which represents about a tenth of the mother's cardiac output—compared with 60 mL/min in the nonpregnant state. Consequently, trauma to the uterus can result in significant hemorrhage and shock.

Blood volume increases by 45%–50% during pregnancy, and can be instrumental in concealing signs of shock from hypovolemia. Along with this, however, the amount of all clotting factors also increases, which predisposes the pregnant patient to embolic risks as well as coagulopathy from disseminated intravascular coagulation (DIC).

The respiratory system adapts for its role of oxygen delivery to the fetus. Tidal volume increases, with an overall increase in minute ventilation. Changes that result from this adaptation are a lowering of the maternal CO2 level and a decrease in alveolar residual volume. This can further result in a respiratory alkalotic state, which is corrected with a renal decrease in bicarbonate (a compensated respiratory alkalosis). The decreased residual volumes render the pregnant patient more susceptible to alveolar collapse and respiratory compromise.

Under these circumstances, it may be prudent to consider early intubation of the pregnant patient with respiratory compromise in order to preserve the exchange of gases across the fetoplacental unit.

Renal blood flow is increased in pregnancy, with a concomitant increase in creatinine clearance and a tendency to more rapidly clear drugs that are renal dependent.

Standard lab values must be seen in the context of the changes in maternal physiology. For example, blood gas values in pregnancy may reveal lower CO2 values, and hematocrits may be lower secondary to the hemodilution seen in blood volume expansion.

During evaluation or resuscitation, it is important for ob.gyns. to remind the team of the mass of the gravid uterus and its ability—when the woman is in the supine position—to compromise the return of blood flow to the heart by compressing the vena cava.

Beyond the second trimester, the patient should be tilted to the left by approximately 35 degrees, and when a full-body tilt is not possible, left lateral displacement of the uterus must be maintained. When spinal injury is suspected, care must be taken to keep the spine and neck aligned during tilting.

Questions About Imaging

The obstetrician often will be asked whether the imaging needed to diagnose various injuries is safe for the fetus. Doses of radiation used during trauma care and evaluation—for example, CT scans of the pelvis or chest, and chest x-rays—are usually in the range of less than 250 mGy, which is considered to be the intermediate range of exposure and reasonably safe for the fetus.

Whenever possible, the pelvis of a pregnant woman should be shielded, especially in the first trimester. It is important to know, however, that much of today's imaging equipment is faster than previous technology and therefore delivers much lower radiation exposure with more information in a single pass. Additionally, MRI has been shown to be a safe modality in pregnancy.

The bottom line is that imaging studies that are needed for the care of the critically injured patient should not be withheld because she may be pregnant.

Primary Assessment

On initial presentation, all efforts for the pregnant trauma patient must first be directed toward stabilizing the mother and maintaining oxygen delivery, with the ABCs (airway, breathing, and circulation) of trauma care being the first priority.

Women who are pregnant have a reduced ability to compensate for respiratory compromise. Maintaining a patent airway is critical for both maternal and fetal oxygenation, and this very well may require early intubation. We must make sure that the patient is moving oxygen in, with oxygen saturation levels better than 90%.

The fetus's oxygen uptake depends directly on oxygen delivery via uterine blood flow, so circulation—both to the mother's vital organs and to the uterus—is also key. Because of shunting and vasospasm, significant uterine blood flow compromise may exist even with normal-appearing blood pressure. Thus, it is important to control any significant bleeding and pursue vigorous volume replacement.

Only after the ABCs are addressed—and readdressed for effectiveness—can we turn our attention elsewhere. If we were to add a “D” to the trauma protocol, it would stand for “disability” and would involve a rapid neurologic evaluation to assess for any neurologic injury. It is worth considering that neurologic impairment in late pregnancy may be secondary to an eclamptic seizure that may have led to the trauma event.

Other often critical components of trauma injuries, such as fractures and intraperitoneal hemorrhage, are usually evaluated almost simultaneously by the trauma team.

Patients with pelvic fractures (common in motor vehicle accidents) are at risk of having retroperitoneal hemorrhage, which is not always obvious and requires careful diagnosis. A pelvic exam can reveal signs of lower pelvic fracture and possible vaginal lacerations from protruding bone fragments.

The Secondary Survey

Once we have stabilized the mother and evaluated her for other critical signs of trauma, we can turn our attention to fetal assessment. First, we should assess gestational age, either by taking a history if someone close to the mother is present, or through ultrasonography.

Ultrasonography is an important tool at this point for assessing several factors in short order. In addition to assessing the viability of the fetus, we can evaluate the intrauterine fluid volume and the placental location. (The question of viability, of course, depends on the level of neonatal intensive care services available).

A low amount of amniotic fluid should lead us to suspect rupture of the amniotic membranes or, in some cases, uterine rupture secondary to trauma. Although with expert hands it is possible to detect relatively small placental abruptions, abruptions are usually apparent only with larger separations.

Ultrasonography can also be used in determining intra-abdominal free fluid which is consistent with intraperitoneal hemorrhage.

When the woman is severely injured and needs surgery and if delivery of the fetus is unnecessary, we should focus on monitoring the fetus in the operating room. We can do so with an ultrasound probe or a fetal Doppler encased in a sterile sleeve.

After surgery—or when the woman does not need surgery and is considered stable enough to undergo observation—continuous monitoring of the viable fetus, with a longer-term view, should be done using external monitoring of fetal heart rate and uterine monitoring for signs of preterm labor. Frequent uterine contractions should be followed closely, and cervical dilatation should be evaluated.

The ob.gyn. becomes the primary provider when a viable fetus shows signs of fetal compromise that necessitate delivery, or when uterine rupture occurs, in which case urgent intervention is necessary for the mother.

Many studies have attempted to address the question of how long trauma patients should be monitored, and many guidelines have been proposed. In general, we can conclude that 12 hours of observation is adequate for stable patients who are not contracting and have reassuring fetal tracings and no signs of bleeding.

On the other hand, patients who are contracting, who have a nonreassuring fetal tracing, or who have had any form of vaginal bleeding should be observed for a minimum of 24 hours. The more severe the injury to the mother, the more likely there is to be an injury to the fetus, and the higher the risk that a placental abruption or other serious complication may surface.

If a patient is discharged after observation, regardless of the severity of trauma, she must be given precautions regarding any changes in fetal movement or the development of abdominal pain, vaginal bleeding, or fluid loss. Any one of these changes should prompt her to return for evaluation immediately.

The Kleihauer-Betke test may be useful in evaluating the degree of fetal-maternal hemorrhage and the amount of Rh immune globulin that may be needed in the Rh-negative mother. In general, Rh-negative mothers who are involved in trauma should be given a single dose of Rh immune globulin unless it is known that the fetus is Rh negative.

Cardiac arrest is sometimes the tragic outcome for a posttrauma victim. Maternal resuscitation should be undertaken immediately. If maternal resuscitation is not thought likely to be effective, and the fetus is considered viable, the best outcome for fetal survival occurs if delivery can be accomplished within approximately 5 minutes from arrest. Beyond this time, there is diminishing return for fetal survival.

Organized, rapid assessment and intervention hold the key to the best outcomes for the pregnant patient involved in a trauma. Following the rules of trauma resuscitation (those ABCs) provides the best chance of successful treatment of the mother, which in turn provides the best chance of a favorable fetal outcome.

Motor vehicle accidents are the leading cause of trauma and death in women of childbearing age. Courtesy Dr. Hugh Mighty

One situation in which the ob.gyn. becomes the primary provider is when delivery is necessary. Courtesy Dr. Hugh Mighty

Trauma in Pregnancy

High-risk obstetrics by its very nature involves a wide spectrum of diseases and events that complicate pregnancies and preclude or prevent their normal progression. Sometimes, high-risk obstetrics involves physical trauma that is inflicted externally upon an otherwise normal pregnancy.

Physical trauma is, in fact, one of the leading causes of morbidity and mortality during pregnancy. It has been estimated that physical trauma complicates approximately 1 in every 12 pregnancies—a staggering figure and one that we may not fully appreciate or think about often enough.

According to the American College of Obstetricians and Gynecologists, approximately two-thirds of all trauma during pregnancy in industrialized nations results from motor vehicle crashes.

Women not only are more likely to be involved in automobile accidents than are male drivers; they also are increasingly more likely to be victims of violence. In 1994, women were about six times more likely than men to be victims of violence—a significant increase from more than a decade before, when women were half as likely as men to be victims of violence.

Trauma in pregnant women has to be dealt with by a variety of specialists. Very often, these patients will present to emergency departments or urgent care centers, and will have to be seen by emergency physicians, surgeons, or general practitioners. This is a challenging situation and one that presents unusual challenges for obstetric staff.

Especially as the number of patients with traumatic injuries and complications increases, it is important that we review some of the key types of presentations and complications of trauma in pregnancy, and discuss how we may best develop therapeutic algorithms for dealing with them.

It is in this light that we have invited Dr. Hugh E. Mighty, chairman of the department of obstetrics, gynecology, and reproductive sciences at the University of Maryland, Baltimore, to discuss the management of seriously injured pregnant women.

Dr. Mighty is not only a maternal-fetal medicine specialist, but is certified in critical care medicine as well. We are pleased to welcome him as an expert on trauma in pregnancy and as this month's guest professor on the subject.

It was my job to explain to the baby's parents what the disorder was and how it happened. Many of the babies I saw had an autosomal recessive disorder, and it was necessary to explain to the baby's mother and father how this awful disease suddenly appeared in their child, and tell them that they faced a 25% risk of its happening again should they decide to have more children. Their options for the future, I would tell them, would be to adopt, to elect not have any more children, to use an anonymous donor for eggs or sperm, or—as many couples do—to carefully roll the genetic dice again with hopes of a better outcome.

In the last scenario, I would explain, the parents would have the opportunity to have a chorionic villus sampling or amniocentesis. But it goes without saying that the 12–15 weeks that unfold before such testing is done are often filled with anxiety: first, about having the test, and second, about the decision to be made if the results are not favorable.

Today, obstetricians have good reason to present such parents with another option—preimplantation genetic diagnosis (PGD)–and to be aware of its capabilities and limitations. Developments in PGD mean that we, also, in all of our preconceptual obstetrical work, have good reason to be cognizant of ethnicity-based risks for genetic disorders and to advise patients, when indicated, to have genetic screening themselves.

We can use PGD today, in conjunction with in vitro fertilization, to test for over 250 serious diseases and conditions caused by mutations or chromosomal abnormalities. Parents who choose the technology—usually couples who know they carry mutations or who have had another baby or a family member with a serious inherited illness—can learn that an embryo is free of the disease that their family is prone to, and can thus start a pregnancy with a commitment to continuing it.

The Genetics, the Process

As obstetricians, we are not trained geneticists. Therefore, it is helpful for us and for our patients if we understand the basic genetics behind PGD, and appreciate how PGD pushes diagnostic technology to its absolute limits—both its theoretical limits and its practical limits.

Every cell taken from an embryo (or any cell in our body) contains all the genes needed to make a new, complete individual. Each cell's DNA contains just four letters of the genetic alphabet: A, T, G, and C. The way in which these letters are strung together, just as letters are put together to make a book, will tell that cell what to do.

We can think of chromosomes as books of the gigantic encyclopedia of life, and genes as paragraphs within these chromosome books. Each chromosome has thousands of genes, most of which contribute something unique to the story of who we are, just as most paragraphs in a novel or encyclopedia contribute a unique element to the story or knowledge base. Some genes do not appear to be as important to our health as others, just as some paragraphs seem like “filler” in a story.

All genes have a defined beginning and a defined end, and just as a paragraph has an indentation, a gene has a promoter. Genes are made up of little bits called exons, just as paragraphs are made up of sentences. Some genes are gigantic; the gene whose mutation causes muscular dystrophy, for instance, would be equivalent to a paragraph about 158 pages long. Other genes are tiny, similar to a short phrase.

We all carry hundreds and hundreds of typographical errors in our personal encyclopedia, some of them inactivating the gene paragraph that contains them. Fortunately, the errors we inherit from our mothers are generally not matched by the errors we inherit from our fathers. Every once in a while, though, we choose a mate with a gene mutation in the same paragraph. When this occurs, the baby does not have a “backup” copy of the intact gene paragraph, and a recessive disease can occur.

Such unfortunate pairings happen more often in couples of similar ethnicity because many gene mutations are ancestral in a given ethnic population. Thus, one of our roles in preconceptual counseling is to think about the possibility that a patient who wants to discontinue birth control and start a family might carry a gene mutation for an inherited disorder common to his or her ethnic background.

In couples of Northern European ancestry, we think first of cystic fibrosis (with a carrier frequency of about 1 in 29 in the United States) and spinal muscular atrophy (1 in 5). In African Americans, we worry about sickle cell anemia (1 in 22).

If our patients have Southeast Asian ancestry, we should consider α-thalassemia (1 in 30), and for patients with Mediterranean ancestry, β-thalassemia. For patients of Ashkenazi Jewish ancestry, genetic screening is performed routinely for the ancestral mutations causing Tay-Sachs disease, familial dysautonomia, Gaucher's disease type 1, Niemann-Pick disease, Bloom syndrome, Canavan disease, and cystic fibrosis.

When we perform PGD, the testing is done overnight. Couples follow the same process that any infertile couple undergoing in vitro fertilization (IVF) would follow, but before implantation, a single cell is taken for analysis from each embryo in its day-3, eight-cell stage. The single-cell samples are sent by courier to a reference laboratory for overnight testing, and a report is electronically sent to the reproductive endocrinologist. Couples are notified of the results in time for embryo transfer on day 5.

In the future, we may be able to relax the timeline and allow more time for embryo transfer by performing the biopsies when the embryo is 5 days old. In this procedure, cells would be taken from the trophectoderm—the outer layer of the embryo that ultimately develops into the placenta—and the embryos would be frozen via a rapid freezing process called vitrification.

Ice crystals do not form in this method, so concerns about damage to the cells is alleviated. Women could then undergo embryo transfer the next month. For now, however, we follow a 5-day deadline for embryo transfer.

Regardless of what advances are made, we must appreciate the fact that this technology pushes medical diagnostics to its limits. PGD involves the testing of one single cell (the smallest unit of life) and one gene (the smallest unit of inheritance), for one typographical error in 3.3 billion DNA letters, and all of this occurs overnight.

Its Value and Accuracy

The nomenclature of what we have simply and rather loosely called PGD is actually changing a bit. Following Europe's lead, U.S. experts are beginning to use the term PGD to refer specifically to the actual diagnosis of a particular disease. PGS (preimplantation genetic screening) is exactly what its name implies as well—screening, largely for abnormal numbers of chromosomes—and not the actual disease diagnosis. Together, the terms fall under the general rubric of preimplantation genetics.

The differentiation is being made because everything about the two procedures—the technology; the people and issues involved; the risks and benefits; and importantly, the accuracy of the procedure—is different. All told, the error rate for preimplantation genetics is in the range of about 2%–4%. Analyzing chromosomes, however, is quite different from analyzing genes, just as counting books of the encyclopedia is quite different from opening a book and finding a letter error.

In analyzing chromosomes, we have to worry about the possibility of complex chromosomal mosaicism having occurred. This is a process by which chromosomes segregate unevenly to cells as the cells are dividing, and if it has occurred, some of the cells we biopsy may appear normal even though the rest of the cells are not.

Experts are increasingly concerned that the chromosome analysis component of preimplantation screening may not really be improving parents' chances of having a healthy baby. However, although the prognostic value of what we now should call PGS is unclear and confusing, there are no such doubts associated with PGD. Telling parents that their baby has a clear 25% chance of having a serious disease is quite different from telling parents that their baby may—or may not—have a chromosomal abnormality.

This is not to say, however, that PGS is without value. I would advise it in cases in which the woman already needs IVF and if she has had recurrent miscarriages.

Ethical Issues and the Future

The real issue with preimplantation genetics, I believe, is whether there are limits to when the technology can and should be used. We must continue, of course, to consider and address the questions associated with PGS and its value. But beyond this, we face numerous questions emanating not as much from a scientific or technologic perspective as from an ethical perspective.

For instance, couples who already have a child with a genetic disease and do not want it to happen again can test their embryos not only to learn which ones carry the genetic defect, but also to learn whether any of their embryos are an identical stem-cell match with their child who is ill.

At the time of delivery, then, they will have not only a healthy baby, but also a baby who can donate identically human leukocyte antigen (HLA)-matched cord blood for stem-cell transplantation to the sibling. Such testing for HLA matching happens daily in the United States.

It is also possible to screen embryos for genes that raise the risk of cancer—primarily breast, ovarian, and colon—in adulthood. We know this is being done in England today, where regulators ruled last year to allow it. The questions are a bit different with this issue, as I see it, because genotype in this case does not accurately predict phenotype. Having the BRCA1 or BRCA2 gene mutation does not mean, for instance, that a person will develop breast or ovarian cancer. So the question really is whether we should be testing embryos for a disease that may never occur.

As in other ethical debates, we must listen to all points of view. Many couples have watched multiple family members die from colon cancer or breast cancer and have decided that enough is enough, whereas other couples who are testing for HLA matching have a child with an incurable, often fatal disease. These couples know there is no such thing as a perfect baby. All they want is to have the A and the G and the T and the C in the right places, or to save their child while having the chance to have another healthy baby to love as well.

Most clinics have ethics teams to develop policies that address these issues and to describe which indications for preimplantation genetics are acceptable and which are not. Most clinics allow the use of technology for finding cancer genes and for HLA matching, for instance, but not for selecting gender.

Studies following children after IVF and preimplantation genetics that have been done in Europe—where the type of medical system allows investigators to effectively track patients for longer-term outcomes—are better than those done in the United States. Clearly, safety and good outcomes have been demonstrated. Thousands and thousands of babies have been born after having undergone IVF and PGS or PGS, with no evidence of birth defects.

Still, experts in the United States have been designing a database—a prospective registry of sorts—that, when implemented, will collect data on the use of preimplantation genetics, primarily regarding how much is being done and for what ends the technology is being used. Such data will help us to further understand and guide this fast-growing facet of reproductive medicine.

A human 8-cell embryo produced routinely in an IVF laboratory is undergoing the biopsy of one blastomere (cell) for testing. Courtesy Dr. Mark R. Hughes

The Quest for Prenatal Evaluation

The quest for fetal and embryonic evaluation has been of great interest to many scientists, physicians, parents, and members of the lay public. It has eventuated into the well-established field of prenatal diagnosis.

The focus, for the most part, has been on gleaning information during the early and midtrimester periods of pregnancy. Such evaluation has been found most useful in providing reassurance to parents when anomalies are excluded, or—under those uncommon circumstances when a diagnosis is made prenatally—in allowing parents and physicians to create appropriate medical and social strategies to deal with these diagnoses.

Despite the benefits afforded by prenatal diagnosis, there remains a subset of the potentially reproductive population for whom conceiving and delivering an abnormal child is not a rare event, but may in fact have a high degree of predictability. These aspiring parents have had to choose among high-risk pregnancy, nonconception, adoption, or egg or sperm donors.

Most recently, however, these patients and their physicians have been able to exploit the new technology of preimplantation genetic diagnosis. This is a rapidly expanding field that has the potential to make a significant difference in the lives of patients who could not otherwise be anywhere near certain that they would deliver a healthy, normal child.

The technology affords us a wide range of possibilities, both now and in the future, but it also presents a number of challenges, including ethical issues, financial coverage issues, and issues concerning the actual availability of the services to select populations and individuals.

This all makes preimplantation genetics quite a complex topic, with great positive potential and a great many implications and potential hurdles and obstacles, all of which must be discussed and deliberated.

It is for this reason that we have invited Dr. Mark R. Hughes, a leading international scholar in the area of preimplantation diagnosis and one of the pioneers of this technology, to serve as the guest author of this month's Master Class.

Currently, Dr. Hughes is the director of the Genesis Genetics Institute in Detroit. He previously served on the faculty of the schools of medicine at Baylor College of Medicine, Georgetown University, and Wayne State University, and was a member of the founding group of the Human Genome Institute at the National Institutes of Health.

It was my job to explain to the baby's parents what the disorder was and how it happened. Many of the babies I saw had an autosomal recessive disorder, and it was necessary to explain to the baby's mother and father how this awful disease suddenly appeared in their child, and tell them that they faced a 25% risk of its happening again should they decide to have more children. Their options for the future, I would tell them, would be to adopt, to elect not have any more children, to use an anonymous donor for eggs or sperm, or—as many couples do—to carefully roll the genetic dice again with hopes of a better outcome.

In the last scenario, I would explain, the parents would have the opportunity to have a chorionic villus sampling or amniocentesis. But it goes without saying that the 12–15 weeks that unfold before such testing is done are often filled with anxiety: first, about having the test, and second, about the decision to be made if the results are not favorable.

Today, obstetricians have good reason to present such parents with another option—preimplantation genetic diagnosis (PGD)–and to be aware of its capabilities and limitations. Developments in PGD mean that we, also, in all of our preconceptual obstetrical work, have good reason to be cognizant of ethnicity-based risks for genetic disorders and to advise patients, when indicated, to have genetic screening themselves.

We can use PGD today, in conjunction with in vitro fertilization, to test for over 250 serious diseases and conditions caused by mutations or chromosomal abnormalities. Parents who choose the technology—usually couples who know they carry mutations or who have had another baby or a family member with a serious inherited illness—can learn that an embryo is free of the disease that their family is prone to, and can thus start a pregnancy with a commitment to continuing it.

The Genetics, the Process

As obstetricians, we are not trained geneticists. Therefore, it is helpful for us and for our patients if we understand the basic genetics behind PGD, and appreciate how PGD pushes diagnostic technology to its absolute limits—both its theoretical limits and its practical limits.

Every cell taken from an embryo (or any cell in our body) contains all the genes needed to make a new, complete individual. Each cell's DNA contains just four letters of the genetic alphabet: A, T, G, and C. The way in which these letters are strung together, just as letters are put together to make a book, will tell that cell what to do.

We can think of chromosomes as books of the gigantic encyclopedia of life, and genes as paragraphs within these chromosome books. Each chromosome has thousands of genes, most of which contribute something unique to the story of who we are, just as most paragraphs in a novel or encyclopedia contribute a unique element to the story or knowledge base. Some genes do not appear to be as important to our health as others, just as some paragraphs seem like “filler” in a story.

All genes have a defined beginning and a defined end, and just as a paragraph has an indentation, a gene has a promoter. Genes are made up of little bits called exons, just as paragraphs are made up of sentences. Some genes are gigantic; the gene whose mutation causes muscular dystrophy, for instance, would be equivalent to a paragraph about 158 pages long. Other genes are tiny, similar to a short phrase.

We all carry hundreds and hundreds of typographical errors in our personal encyclopedia, some of them inactivating the gene paragraph that contains them. Fortunately, the errors we inherit from our mothers are generally not matched by the errors we inherit from our fathers. Every once in a while, though, we choose a mate with a gene mutation in the same paragraph. When this occurs, the baby does not have a “backup” copy of the intact gene paragraph, and a recessive disease can occur.

Such unfortunate pairings happen more often in couples of similar ethnicity because many gene mutations are ancestral in a given ethnic population. Thus, one of our roles in preconceptual counseling is to think about the possibility that a patient who wants to discontinue birth control and start a family might carry a gene mutation for an inherited disorder common to his or her ethnic background.

In couples of Northern European ancestry, we think first of cystic fibrosis (with a carrier frequency of about 1 in 29 in the United States) and spinal muscular atrophy (1 in 5). In African Americans, we worry about sickle cell anemia (1 in 22).

If our patients have Southeast Asian ancestry, we should consider α-thalassemia (1 in 30), and for patients with Mediterranean ancestry, β-thalassemia. For patients of Ashkenazi Jewish ancestry, genetic screening is performed routinely for the ancestral mutations causing Tay-Sachs disease, familial dysautonomia, Gaucher's disease type 1, Niemann-Pick disease, Bloom syndrome, Canavan disease, and cystic fibrosis.

When we perform PGD, the testing is done overnight. Couples follow the same process that any infertile couple undergoing in vitro fertilization (IVF) would follow, but before implantation, a single cell is taken for analysis from each embryo in its day-3, eight-cell stage. The single-cell samples are sent by courier to a reference laboratory for overnight testing, and a report is electronically sent to the reproductive endocrinologist. Couples are notified of the results in time for embryo transfer on day 5.

In the future, we may be able to relax the timeline and allow more time for embryo transfer by performing the biopsies when the embryo is 5 days old. In this procedure, cells would be taken from the trophectoderm—the outer layer of the embryo that ultimately develops into the placenta—and the embryos would be frozen via a rapid freezing process called vitrification.

Ice crystals do not form in this method, so concerns about damage to the cells is alleviated. Women could then undergo embryo transfer the next month. For now, however, we follow a 5-day deadline for embryo transfer.

Regardless of what advances are made, we must appreciate the fact that this technology pushes medical diagnostics to its limits. PGD involves the testing of one single cell (the smallest unit of life) and one gene (the smallest unit of inheritance), for one typographical error in 3.3 billion DNA letters, and all of this occurs overnight.

Its Value and Accuracy

The nomenclature of what we have simply and rather loosely called PGD is actually changing a bit. Following Europe's lead, U.S. experts are beginning to use the term PGD to refer specifically to the actual diagnosis of a particular disease. PGS (preimplantation genetic screening) is exactly what its name implies as well—screening, largely for abnormal numbers of chromosomes—and not the actual disease diagnosis. Together, the terms fall under the general rubric of preimplantation genetics.

The differentiation is being made because everything about the two procedures—the technology; the people and issues involved; the risks and benefits; and importantly, the accuracy of the procedure—is different. All told, the error rate for preimplantation genetics is in the range of about 2%–4%. Analyzing chromosomes, however, is quite different from analyzing genes, just as counting books of the encyclopedia is quite different from opening a book and finding a letter error.

In analyzing chromosomes, we have to worry about the possibility of complex chromosomal mosaicism having occurred. This is a process by which chromosomes segregate unevenly to cells as the cells are dividing, and if it has occurred, some of the cells we biopsy may appear normal even though the rest of the cells are not.

Experts are increasingly concerned that the chromosome analysis component of preimplantation screening may not really be improving parents' chances of having a healthy baby. However, although the prognostic value of what we now should call PGS is unclear and confusing, there are no such doubts associated with PGD. Telling parents that their baby has a clear 25% chance of having a serious disease is quite different from telling parents that their baby may—or may not—have a chromosomal abnormality.

This is not to say, however, that PGS is without value. I would advise it in cases in which the woman already needs IVF and if she has had recurrent miscarriages.

Ethical Issues and the Future

The real issue with preimplantation genetics, I believe, is whether there are limits to when the technology can and should be used. We must continue, of course, to consider and address the questions associated with PGS and its value. But beyond this, we face numerous questions emanating not as much from a scientific or technologic perspective as from an ethical perspective.

For instance, couples who already have a child with a genetic disease and do not want it to happen again can test their embryos not only to learn which ones carry the genetic defect, but also to learn whether any of their embryos are an identical stem-cell match with their child who is ill.

At the time of delivery, then, they will have not only a healthy baby, but also a baby who can donate identically human leukocyte antigen (HLA)-matched cord blood for stem-cell transplantation to the sibling. Such testing for HLA matching happens daily in the United States.

It is also possible to screen embryos for genes that raise the risk of cancer—primarily breast, ovarian, and colon—in adulthood. We know this is being done in England today, where regulators ruled last year to allow it. The questions are a bit different with this issue, as I see it, because genotype in this case does not accurately predict phenotype. Having the BRCA1 or BRCA2 gene mutation does not mean, for instance, that a person will develop breast or ovarian cancer. So the question really is whether we should be testing embryos for a disease that may never occur.

As in other ethical debates, we must listen to all points of view. Many couples have watched multiple family members die from colon cancer or breast cancer and have decided that enough is enough, whereas other couples who are testing for HLA matching have a child with an incurable, often fatal disease. These couples know there is no such thing as a perfect baby. All they want is to have the A and the G and the T and the C in the right places, or to save their child while having the chance to have another healthy baby to love as well.

Most clinics have ethics teams to develop policies that address these issues and to describe which indications for preimplantation genetics are acceptable and which are not. Most clinics allow the use of technology for finding cancer genes and for HLA matching, for instance, but not for selecting gender.

Studies following children after IVF and preimplantation genetics that have been done in Europe—where the type of medical system allows investigators to effectively track patients for longer-term outcomes—are better than those done in the United States. Clearly, safety and good outcomes have been demonstrated. Thousands and thousands of babies have been born after having undergone IVF and PGS or PGS, with no evidence of birth defects.

Still, experts in the United States have been designing a database—a prospective registry of sorts—that, when implemented, will collect data on the use of preimplantation genetics, primarily regarding how much is being done and for what ends the technology is being used. Such data will help us to further understand and guide this fast-growing facet of reproductive medicine.

A human 8-cell embryo produced routinely in an IVF laboratory is undergoing the biopsy of one blastomere (cell) for testing. Courtesy Dr. Mark R. Hughes

The Quest for Prenatal Evaluation

The quest for fetal and embryonic evaluation has been of great interest to many scientists, physicians, parents, and members of the lay public. It has eventuated into the well-established field of prenatal diagnosis.

The focus, for the most part, has been on gleaning information during the early and midtrimester periods of pregnancy. Such evaluation has been found most useful in providing reassurance to parents when anomalies are excluded, or—under those uncommon circumstances when a diagnosis is made prenatally—in allowing parents and physicians to create appropriate medical and social strategies to deal with these diagnoses.

Despite the benefits afforded by prenatal diagnosis, there remains a subset of the potentially reproductive population for whom conceiving and delivering an abnormal child is not a rare event, but may in fact have a high degree of predictability. These aspiring parents have had to choose among high-risk pregnancy, nonconception, adoption, or egg or sperm donors.

Most recently, however, these patients and their physicians have been able to exploit the new technology of preimplantation genetic diagnosis. This is a rapidly expanding field that has the potential to make a significant difference in the lives of patients who could not otherwise be anywhere near certain that they would deliver a healthy, normal child.

The technology affords us a wide range of possibilities, both now and in the future, but it also presents a number of challenges, including ethical issues, financial coverage issues, and issues concerning the actual availability of the services to select populations and individuals.

This all makes preimplantation genetics quite a complex topic, with great positive potential and a great many implications and potential hurdles and obstacles, all of which must be discussed and deliberated.

It is for this reason that we have invited Dr. Mark R. Hughes, a leading international scholar in the area of preimplantation diagnosis and one of the pioneers of this technology, to serve as the guest author of this month's Master Class.

Currently, Dr. Hughes is the director of the Genesis Genetics Institute in Detroit. He previously served on the faculty of the schools of medicine at Baylor College of Medicine, Georgetown University, and Wayne State University, and was a member of the founding group of the Human Genome Institute at the National Institutes of Health.

It was my job to explain to the baby's parents what the disorder was and how it happened. Many of the babies I saw had an autosomal recessive disorder, and it was necessary to explain to the baby's mother and father how this awful disease suddenly appeared in their child, and tell them that they faced a 25% risk of its happening again should they decide to have more children. Their options for the future, I would tell them, would be to adopt, to elect not have any more children, to use an anonymous donor for eggs or sperm, or—as many couples do—to carefully roll the genetic dice again with hopes of a better outcome.

In the last scenario, I would explain, the parents would have the opportunity to have a chorionic villus sampling or amniocentesis. But it goes without saying that the 12–15 weeks that unfold before such testing is done are often filled with anxiety: first, about having the test, and second, about the decision to be made if the results are not favorable.

Today, obstetricians have good reason to present such parents with another option—preimplantation genetic diagnosis (PGD)–and to be aware of its capabilities and limitations. Developments in PGD mean that we, also, in all of our preconceptual obstetrical work, have good reason to be cognizant of ethnicity-based risks for genetic disorders and to advise patients, when indicated, to have genetic screening themselves.

We can use PGD today, in conjunction with in vitro fertilization, to test for over 250 serious diseases and conditions caused by mutations or chromosomal abnormalities. Parents who choose the technology—usually couples who know they carry mutations or who have had another baby or a family member with a serious inherited illness—can learn that an embryo is free of the disease that their family is prone to, and can thus start a pregnancy with a commitment to continuing it.

The Genetics, the Process

As obstetricians, we are not trained geneticists. Therefore, it is helpful for us and for our patients if we understand the basic genetics behind PGD, and appreciate how PGD pushes diagnostic technology to its absolute limits—both its theoretical limits and its practical limits.

Every cell taken from an embryo (or any cell in our body) contains all the genes needed to make a new, complete individual. Each cell's DNA contains just four letters of the genetic alphabet: A, T, G, and C. The way in which these letters are strung together, just as letters are put together to make a book, will tell that cell what to do.

We can think of chromosomes as books of the gigantic encyclopedia of life, and genes as paragraphs within these chromosome books. Each chromosome has thousands of genes, most of which contribute something unique to the story of who we are, just as most paragraphs in a novel or encyclopedia contribute a unique element to the story or knowledge base. Some genes do not appear to be as important to our health as others, just as some paragraphs seem like “filler” in a story.

All genes have a defined beginning and a defined end, and just as a paragraph has an indentation, a gene has a promoter. Genes are made up of little bits called exons, just as paragraphs are made up of sentences. Some genes are gigantic; the gene whose mutation causes muscular dystrophy, for instance, would be equivalent to a paragraph about 158 pages long. Other genes are tiny, similar to a short phrase.

We all carry hundreds and hundreds of typographical errors in our personal encyclopedia, some of them inactivating the gene paragraph that contains them. Fortunately, the errors we inherit from our mothers are generally not matched by the errors we inherit from our fathers. Every once in a while, though, we choose a mate with a gene mutation in the same paragraph. When this occurs, the baby does not have a “backup” copy of the intact gene paragraph, and a recessive disease can occur.

Such unfortunate pairings happen more often in couples of similar ethnicity because many gene mutations are ancestral in a given ethnic population. Thus, one of our roles in preconceptual counseling is to think about the possibility that a patient who wants to discontinue birth control and start a family might carry a gene mutation for an inherited disorder common to his or her ethnic background.

In couples of Northern European ancestry, we think first of cystic fibrosis (with a carrier frequency of about 1 in 29 in the United States) and spinal muscular atrophy (1 in 5). In African Americans, we worry about sickle cell anemia (1 in 22).

If our patients have Southeast Asian ancestry, we should consider α-thalassemia (1 in 30), and for patients with Mediterranean ancestry, β-thalassemia. For patients of Ashkenazi Jewish ancestry, genetic screening is performed routinely for the ancestral mutations causing Tay-Sachs disease, familial dysautonomia, Gaucher's disease type 1, Niemann-Pick disease, Bloom syndrome, Canavan disease, and cystic fibrosis.

When we perform PGD, the testing is done overnight. Couples follow the same process that any infertile couple undergoing in vitro fertilization (IVF) would follow, but before implantation, a single cell is taken for analysis from each embryo in its day-3, eight-cell stage. The single-cell samples are sent by courier to a reference laboratory for overnight testing, and a report is electronically sent to the reproductive endocrinologist. Couples are notified of the results in time for embryo transfer on day 5.

In the future, we may be able to relax the timeline and allow more time for embryo transfer by performing the biopsies when the embryo is 5 days old. In this procedure, cells would be taken from the trophectoderm—the outer layer of the embryo that ultimately develops into the placenta—and the embryos would be frozen via a rapid freezing process called vitrification.

Ice crystals do not form in this method, so concerns about damage to the cells is alleviated. Women could then undergo embryo transfer the next month. For now, however, we follow a 5-day deadline for embryo transfer.

Regardless of what advances are made, we must appreciate the fact that this technology pushes medical diagnostics to its limits. PGD involves the testing of one single cell (the smallest unit of life) and one gene (the smallest unit of inheritance), for one typographical error in 3.3 billion DNA letters, and all of this occurs overnight.

Its Value and Accuracy

The nomenclature of what we have simply and rather loosely called PGD is actually changing a bit. Following Europe's lead, U.S. experts are beginning to use the term PGD to refer specifically to the actual diagnosis of a particular disease. PGS (preimplantation genetic screening) is exactly what its name implies as well—screening, largely for abnormal numbers of chromosomes—and not the actual disease diagnosis. Together, the terms fall under the general rubric of preimplantation genetics.

The differentiation is being made because everything about the two procedures—the technology; the people and issues involved; the risks and benefits; and importantly, the accuracy of the procedure—is different. All told, the error rate for preimplantation genetics is in the range of about 2%–4%. Analyzing chromosomes, however, is quite different from analyzing genes, just as counting books of the encyclopedia is quite different from opening a book and finding a letter error.

In analyzing chromosomes, we have to worry about the possibility of complex chromosomal mosaicism having occurred. This is a process by which chromosomes segregate unevenly to cells as the cells are dividing, and if it has occurred, some of the cells we biopsy may appear normal even though the rest of the cells are not.

Experts are increasingly concerned that the chromosome analysis component of preimplantation screening may not really be improving parents' chances of having a healthy baby. However, although the prognostic value of what we now should call PGS is unclear and confusing, there are no such doubts associated with PGD. Telling parents that their baby has a clear 25% chance of having a serious disease is quite different from telling parents that their baby may—or may not—have a chromosomal abnormality.

This is not to say, however, that PGS is without value. I would advise it in cases in which the woman already needs IVF and if she has had recurrent miscarriages.

Ethical Issues and the Future

The real issue with preimplantation genetics, I believe, is whether there are limits to when the technology can and should be used. We must continue, of course, to consider and address the questions associated with PGS and its value. But beyond this, we face numerous questions emanating not as much from a scientific or technologic perspective as from an ethical perspective.

For instance, couples who already have a child with a genetic disease and do not want it to happen again can test their embryos not only to learn which ones carry the genetic defect, but also to learn whether any of their embryos are an identical stem-cell match with their child who is ill.

At the time of delivery, then, they will have not only a healthy baby, but also a baby who can donate identically human leukocyte antigen (HLA)-matched cord blood for stem-cell transplantation to the sibling. Such testing for HLA matching happens daily in the United States.

It is also possible to screen embryos for genes that raise the risk of cancer—primarily breast, ovarian, and colon—in adulthood. We know this is being done in England today, where regulators ruled last year to allow it. The questions are a bit different with this issue, as I see it, because genotype in this case does not accurately predict phenotype. Having the BRCA1 or BRCA2 gene mutation does not mean, for instance, that a person will develop breast or ovarian cancer. So the question really is whether we should be testing embryos for a disease that may never occur.

As in other ethical debates, we must listen to all points of view. Many couples have watched multiple family members die from colon cancer or breast cancer and have decided that enough is enough, whereas other couples who are testing for HLA matching have a child with an incurable, often fatal disease. These couples know there is no such thing as a perfect baby. All they want is to have the A and the G and the T and the C in the right places, or to save their child while having the chance to have another healthy baby to love as well.

Most clinics have ethics teams to develop policies that address these issues and to describe which indications for preimplantation genetics are acceptable and which are not. Most clinics allow the use of technology for finding cancer genes and for HLA matching, for instance, but not for selecting gender.

Studies following children after IVF and preimplantation genetics that have been done in Europe—where the type of medical system allows investigators to effectively track patients for longer-term outcomes—are better than those done in the United States. Clearly, safety and good outcomes have been demonstrated. Thousands and thousands of babies have been born after having undergone IVF and PGS or PGS, with no evidence of birth defects.

Still, experts in the United States have been designing a database—a prospective registry of sorts—that, when implemented, will collect data on the use of preimplantation genetics, primarily regarding how much is being done and for what ends the technology is being used. Such data will help us to further understand and guide this fast-growing facet of reproductive medicine.

A human 8-cell embryo produced routinely in an IVF laboratory is undergoing the biopsy of one blastomere (cell) for testing. Courtesy Dr. Mark R. Hughes

The Quest for Prenatal Evaluation

The quest for fetal and embryonic evaluation has been of great interest to many scientists, physicians, parents, and members of the lay public. It has eventuated into the well-established field of prenatal diagnosis.

The focus, for the most part, has been on gleaning information during the early and midtrimester periods of pregnancy. Such evaluation has been found most useful in providing reassurance to parents when anomalies are excluded, or—under those uncommon circumstances when a diagnosis is made prenatally—in allowing parents and physicians to create appropriate medical and social strategies to deal with these diagnoses.

Despite the benefits afforded by prenatal diagnosis, there remains a subset of the potentially reproductive population for whom conceiving and delivering an abnormal child is not a rare event, but may in fact have a high degree of predictability. These aspiring parents have had to choose among high-risk pregnancy, nonconception, adoption, or egg or sperm donors.

Most recently, however, these patients and their physicians have been able to exploit the new technology of preimplantation genetic diagnosis. This is a rapidly expanding field that has the potential to make a significant difference in the lives of patients who could not otherwise be anywhere near certain that they would deliver a healthy, normal child.

The technology affords us a wide range of possibilities, both now and in the future, but it also presents a number of challenges, including ethical issues, financial coverage issues, and issues concerning the actual availability of the services to select populations and individuals.

This all makes preimplantation genetics quite a complex topic, with great positive potential and a great many implications and potential hurdles and obstacles, all of which must be discussed and deliberated.

It is for this reason that we have invited Dr. Mark R. Hughes, a leading international scholar in the area of preimplantation diagnosis and one of the pioneers of this technology, to serve as the guest author of this month's Master Class.

Currently, Dr. Hughes is the director of the Genesis Genetics Institute in Detroit. He previously served on the faculty of the schools of medicine at Baylor College of Medicine, Georgetown University, and Wayne State University, and was a member of the founding group of the Human Genome Institute at the National Institutes of Health.

The American College of Obstetricians and Gynecologists (ACOG) has described examples of commonly accepted indications, contraindications, and clinical conditions requiring special attention for an induction of labor. (See box p. 37.) We must remember that indications for labor induction are often not absolute and need to take maternal and fetal conditions, gestational age, and cervical status into account. Many contraindications are the same as those for either spontaneous labor or vaginal delivery; several obstetric conditions are not contraindications, but do necessitate special attention.

In 1988, the National Center for Health Statistics began requiring hospitals to indicate on birth certificates whether labor was induced or not. This requirement has provided us with remarkable insight into labor induction rates—insight that should cause us to pause, to reflect on available data and our own practices, and to demand that the issue receive more widespread attention.

Over a 10-year period beginning in 1989, the rate of labor induction doubled from about 9% to almost 19% of live births. (See chart.) The trend steadily continued into the new millennium, to the point where, in 2003, nearly 23% of all births involved induction of labor. Clearly, labor induction is one of the most common procedures in obstetrics.

Examining the Increase

The reasons for this significant increase over just 15 years relate to the availability of FDA-approved cervical ripening agents; to both the patient's desire and the physician's convenience; to the acceptance of added risks of cesarean delivery; and to increases in marginal or elective inductions for term pregnancies, especially those past 40 weeks. Inductions in which the reason is not evidence based now account for at least half of all term inductions, or up to 10% of all deliveries. The increase in medically indicated inductions was slower than the overall increase, suggesting that inductions for marginal or elective reasons have risen more rapidly.

Also contributing to the rising rate in inductions is our increasing success with cervical ripening and the fact that, in the current era of ultrasound availability and a more accurate dating of gestational age, we have had to worry less about iatrogenic prematurity.

When considering labor induction, we can view “elective” and “marginal” indications as being very similar, or we can differentiate the two, with “elective” meaning there is no plausible medical or obstetric reason for the induction, and “marginal” referring to cases in which obstetricians face or suspect problems but have no data to suggest that the benefits of labor induction outweigh the risks. I believe it is valuable to consider the terms separately as we attempt to understand the changes in induction rates.

Marginal indications include gestational hypertension; unexplained and mild fetal-growth restriction; idiopathic decreased amniotic fluid (which does not pose substantial danger unless it is accompanied by a recognized complication, such as hypertension or a small-for-gestational-age baby); and a pregnancy beyond 40 weeks. Prospective studies to recommend induction for these and other marginal indications are limited in size or design, or are nonexistent.

There is some rationale behind induction for suspected fetal macrosomia in nondiabetic pregnancies. Theoretically, eliminating further fetal growth should reduce the risks of shoulder dystocia and perhaps of cesarean delivery. However, there is no evidence-based justification for labor induction in these patients. Studies have shown, in fact, that the procedure approximately doubles the cesarean delivery risk, does not reduce neonatal morbidity, and does not appear to reduce the risk of shoulder dystocia.

There is also no published evidence to support the induction of labor for preterm mild preeclampsia, prior shoulder dystocia, and prior cephalopelvic disproportion.

ACOG weighed into the issue by approving “logistic reasons” for labor induction, such as a risk of rapid labor, a patient's unacceptable distance from the hospital, and psychosocial indications. This has left ob.gyns. with a substantial amount of latitude. For instance, one could argue that “psychosocial” reasons could include alleviating the concerns of a mother who previously had a stillborn infant, or alleviating the anxiety of a woman whose spouse is scheduled for deployment to Iraq before the delivery date.

In analyzing the increased rate of labor inductions, we can simply and easily make our own justifications for elective and marginal inductions—we are making our patients happy, for one thing—and put on the back burner the lack of evidence favoring non-medically indicated induction. No matter how appealing our justifications might be, however, we cannot ignore the paucity of published data on benefits, nor can we ignore the data that do exist on the risks of labor induction.

Appreciating the Risks

Studies have shown that induced labor is associated with an increase in epidurals, with the greatest risk of uterine rupture in patients with a scarred uterus, with perhaps an increase in instrumental vaginal deliveries, and with an increase in cesarean deliveries, particularly among nulliparas undergoing an induction with an unfavorable cervix.

Investigators of a large study published in 2005 found a 1.5-fold greater risk of diagnosing a nonreassuring fetal heart rate pattern, a twofold increase in the need for epidural anesthesia, and a 1.5-fold increased risk of having a cesarean delivery among women who had elective inductions of labor compared with women who had spontaneous labor.

The risks of oxytocin use are principally dose related. Excess or undesired uterine hyperstimulation and subsequent fetal heart rate decelerations (“hyperstimulation syndrome”) are the most common side effects. In addition, hyperstimulation is associated with a greater risk of abruptio placentae or uterine rupture. There does not appear to be a significant increase in adverse fetal outcomes from uterine tachysystole.

Uterine hyperstimulation is an adverse effect that is also dose-dependent for prostaglandins (misoprostol, dinoprostone) used as cervical-ripening agents. The potential risks associated with amniotomy include prolapse of the umbilical cord, chorioamnionitis, significant umbilical cord compression, and rupture of vasa previa. With close monitoring and proper precaution, these hazards are fortunately uncommon.

Even if no additional risks are found with elective and marginal indications, it is important to consider issues related to personnel and cost. In addition to increasing the primary cesarean rate—and even a small additional risk of cesarean delivery for nulliparous women who have their labor induced translates into a significantly larger number of cesarean deliveries nationally—labor that is induced requires more one-on-one care and thus more nurses or nursing time.

It also independently leads to significantly longer time in labor and delivery, as well as a prolonged maternal length of hospital stay. Investigators have demonstrated significant differences in the admission-to-delivery times and in-hospital costs between patients who have vaginal deliveries after induced labor as compared with those who have spontaneous labor, as well as with patients who have cesarean deliveries in both scenarios.

Other studies have shown that labor inductions can overload the labor and delivery departments of some hospitals during “popular” midweek times. Downstream, labor induction also leads to an excess number of vaginal births after cesarean (VBAC) or repeat cesarean procedures. I am convinced, moreover, that litigation will be a concern in the future, especially with our armamentarium of cervical ripening agents. When there is a negative outcome after induction, I believe we can anticipate an allegation of unnecessary induction due to the lack of a medical indication.

The frequency of elective inductions and inductions for marginal indications appears to be higher in community hospitals than at university hospitals. A study that my colleagues and I published in 2000 found that 5% of all labor inductions at a university hospital were elective or not medically indicated using the ACOG criteria. At two community hospitals, on the other hand, 44% and 57% of inductions were for elective reasons.

Physicians in academic settings—particularly those involved in clinical trials to assess the effectiveness of therapies for labor induction—are more likely to use the Bishop scoring system. The Bishop score, first described in 1964, is based on cervical dilation, effacement, consistency, and position, as well as on fetal station. Although the scale isn't used much outside of academia, the principles should be consistently and universally applied, particularly the assessment of dilation and cervical consistency.

Planning the Future

Investigators have looked and will continue to look for predictors of success and ideal conditions for labor induction, but at this point in time the only known conditions are a favorable cervix and a patient who has had a previous vaginal delivery. Multiparous women at term generally present with a more favorable cervix.

Right now, roughly half of women who have their labor induced—or roughly 10% of the overall pregnancy population—have an unfavorable cervix. Cesarean rates are high for nulliparas who undergo an induction with an unfavorable cervix. This is a picture that needs widespread attention and an awareness of the desirability of evidence-based decisions.

Obstetricians must construct consistent and evidence-based protocols for cervical ripening; formally evaluate physician and patient satisfaction with induction; and design and lead clinical trials to provide answers on the value of marginal indications.

In the meantime, labor induction rates for hospitals and physicians should be monitored, and patients should be educated about the risks of induction so that they can participate in decision making and be better able to balance concerns and benefits. It is quite possible that written consent may become a standard of care before any induction is undertaken.

Until we do so, we should be aware that we may be complicating the uncomplicated.

ELSEVIER GLOBAL MEDICAL NEWS

Induction of Labor

The timing of parturition remains a conundrum in obstetric medicine in that the majority of pregnancies will go to term and enter labor spontaneously, whereas another portion will go post term and often require induction, and still others will enter labor prematurely.

The concept of labor induction, therefore, has become very important in obstetric management, especially in addressing pregnancies that either go post term or pregnancies that require induction because of medical complications in the mother.

Increasingly, however, patients are apt to have labor induced for their own convenience, for personal reasons, for the convenience of the physician, and sometimes for all of these reasons.

This increasingly utilized social option ushers in a whole new perspective on the issue of induction, and the question is raised about whether or not the elective induction of labor brings with it added risk and more complications.

It is for this reason that we decided to develop a Master Class feature on this topic. It gives us the important opportunity to examine and consider the pros and cons of labor induction, the timing of labor induction, and the advisability of the various conditions under which induction can and does occur.

This month's guest professor is Dr. William F. Rayburn, professor and chairman of the department of ob.gyn. at the University of New Mexico, Albuquerque. Dr. Rayburn is a maternal and fetal medicine specialist with a national reputation in this area.

Indications and Contraindications

Indications

Abruptio placentae

Chorioamnionitis

Fetal demise

Pregnancy-induced hypertension

Premature rupture of membranes

Postterm pregnancy

Maternal medical conditions (such as diabetes mellitus, renal disease, chronic pulmonary disease, chronic hypertension)

Fetal compromise (such as severe fetal growth restriction, isoimmunization)

Preeclampsia, eclampsia

Contraindications

Vasa previa or complete placenta previa

Transverse fetal lie

Umbilical cord prolapse

Previous transfundal uterine surgery

Special Attention

One or more previous low-transverse cesarean deliveries

Breech presentation

Maternal heart disease

Multifetal pregnancy

Polyhydramnios

Presenting part above the pelvic inlet

Severe hypertension

Abnormal fetal heart rate patterns not necessitating emergent delivery

Source: Adapted from ACOG Practice Bulletin No. 10, “Induction of Labor” (Nov. 1999).

The American College of Obstetricians and Gynecologists (ACOG) has described examples of commonly accepted indications, contraindications, and clinical conditions requiring special attention for an induction of labor. (See box p. 37.) We must remember that indications for labor induction are often not absolute and need to take maternal and fetal conditions, gestational age, and cervical status into account. Many contraindications are the same as those for either spontaneous labor or vaginal delivery; several obstetric conditions are not contraindications, but do necessitate special attention.

In 1988, the National Center for Health Statistics began requiring hospitals to indicate on birth certificates whether labor was induced or not. This requirement has provided us with remarkable insight into labor induction rates—insight that should cause us to pause, to reflect on available data and our own practices, and to demand that the issue receive more widespread attention.

Over a 10-year period beginning in 1989, the rate of labor induction doubled from about 9% to almost 19% of live births. (See chart.) The trend steadily continued into the new millennium, to the point where, in 2003, nearly 23% of all births involved induction of labor. Clearly, labor induction is one of the most common procedures in obstetrics.

Examining the Increase

The reasons for this significant increase over just 15 years relate to the availability of FDA-approved cervical ripening agents; to both the patient's desire and the physician's convenience; to the acceptance of added risks of cesarean delivery; and to increases in marginal or elective inductions for term pregnancies, especially those past 40 weeks. Inductions in which the reason is not evidence based now account for at least half of all term inductions, or up to 10% of all deliveries. The increase in medically indicated inductions was slower than the overall increase, suggesting that inductions for marginal or elective reasons have risen more rapidly.

Also contributing to the rising rate in inductions is our increasing success with cervical ripening and the fact that, in the current era of ultrasound availability and a more accurate dating of gestational age, we have had to worry less about iatrogenic prematurity.

When considering labor induction, we can view “elective” and “marginal” indications as being very similar, or we can differentiate the two, with “elective” meaning there is no plausible medical or obstetric reason for the induction, and “marginal” referring to cases in which obstetricians face or suspect problems but have no data to suggest that the benefits of labor induction outweigh the risks. I believe it is valuable to consider the terms separately as we attempt to understand the changes in induction rates.

Marginal indications include gestational hypertension; unexplained and mild fetal-growth restriction; idiopathic decreased amniotic fluid (which does not pose substantial danger unless it is accompanied by a recognized complication, such as hypertension or a small-for-gestational-age baby); and a pregnancy beyond 40 weeks. Prospective studies to recommend induction for these and other marginal indications are limited in size or design, or are nonexistent.

There is some rationale behind induction for suspected fetal macrosomia in nondiabetic pregnancies. Theoretically, eliminating further fetal growth should reduce the risks of shoulder dystocia and perhaps of cesarean delivery. However, there is no evidence-based justification for labor induction in these patients. Studies have shown, in fact, that the procedure approximately doubles the cesarean delivery risk, does not reduce neonatal morbidity, and does not appear to reduce the risk of shoulder dystocia.

There is also no published evidence to support the induction of labor for preterm mild preeclampsia, prior shoulder dystocia, and prior cephalopelvic disproportion.

ACOG weighed into the issue by approving “logistic reasons” for labor induction, such as a risk of rapid labor, a patient's unacceptable distance from the hospital, and psychosocial indications. This has left ob.gyns. with a substantial amount of latitude. For instance, one could argue that “psychosocial” reasons could include alleviating the concerns of a mother who previously had a stillborn infant, or alleviating the anxiety of a woman whose spouse is scheduled for deployment to Iraq before the delivery date.

In analyzing the increased rate of labor inductions, we can simply and easily make our own justifications for elective and marginal inductions—we are making our patients happy, for one thing—and put on the back burner the lack of evidence favoring non-medically indicated induction. No matter how appealing our justifications might be, however, we cannot ignore the paucity of published data on benefits, nor can we ignore the data that do exist on the risks of labor induction.

Appreciating the Risks

Studies have shown that induced labor is associated with an increase in epidurals, with the greatest risk of uterine rupture in patients with a scarred uterus, with perhaps an increase in instrumental vaginal deliveries, and with an increase in cesarean deliveries, particularly among nulliparas undergoing an induction with an unfavorable cervix.

Investigators of a large study published in 2005 found a 1.5-fold greater risk of diagnosing a nonreassuring fetal heart rate pattern, a twofold increase in the need for epidural anesthesia, and a 1.5-fold increased risk of having a cesarean delivery among women who had elective inductions of labor compared with women who had spontaneous labor.

The risks of oxytocin use are principally dose related. Excess or undesired uterine hyperstimulation and subsequent fetal heart rate decelerations (“hyperstimulation syndrome”) are the most common side effects. In addition, hyperstimulation is associated with a greater risk of abruptio placentae or uterine rupture. There does not appear to be a significant increase in adverse fetal outcomes from uterine tachysystole.

Uterine hyperstimulation is an adverse effect that is also dose-dependent for prostaglandins (misoprostol, dinoprostone) used as cervical-ripening agents. The potential risks associated with amniotomy include prolapse of the umbilical cord, chorioamnionitis, significant umbilical cord compression, and rupture of vasa previa. With close monitoring and proper precaution, these hazards are fortunately uncommon.

Even if no additional risks are found with elective and marginal indications, it is important to consider issues related to personnel and cost. In addition to increasing the primary cesarean rate—and even a small additional risk of cesarean delivery for nulliparous women who have their labor induced translates into a significantly larger number of cesarean deliveries nationally—labor that is induced requires more one-on-one care and thus more nurses or nursing time.

It also independently leads to significantly longer time in labor and delivery, as well as a prolonged maternal length of hospital stay. Investigators have demonstrated significant differences in the admission-to-delivery times and in-hospital costs between patients who have vaginal deliveries after induced labor as compared with those who have spontaneous labor, as well as with patients who have cesarean deliveries in both scenarios.

Other studies have shown that labor inductions can overload the labor and delivery departments of some hospitals during “popular” midweek times. Downstream, labor induction also leads to an excess number of vaginal births after cesarean (VBAC) or repeat cesarean procedures. I am convinced, moreover, that litigation will be a concern in the future, especially with our armamentarium of cervical ripening agents. When there is a negative outcome after induction, I believe we can anticipate an allegation of unnecessary induction due to the lack of a medical indication.

The frequency of elective inductions and inductions for marginal indications appears to be higher in community hospitals than at university hospitals. A study that my colleagues and I published in 2000 found that 5% of all labor inductions at a university hospital were elective or not medically indicated using the ACOG criteria. At two community hospitals, on the other hand, 44% and 57% of inductions were for elective reasons.

Physicians in academic settings—particularly those involved in clinical trials to assess the effectiveness of therapies for labor induction—are more likely to use the Bishop scoring system. The Bishop score, first described in 1964, is based on cervical dilation, effacement, consistency, and position, as well as on fetal station. Although the scale isn't used much outside of academia, the principles should be consistently and universally applied, particularly the assessment of dilation and cervical consistency.

Planning the Future

Investigators have looked and will continue to look for predictors of success and ideal conditions for labor induction, but at this point in time the only known conditions are a favorable cervix and a patient who has had a previous vaginal delivery. Multiparous women at term generally present with a more favorable cervix.

Right now, roughly half of women who have their labor induced—or roughly 10% of the overall pregnancy population—have an unfavorable cervix. Cesarean rates are high for nulliparas who undergo an induction with an unfavorable cervix. This is a picture that needs widespread attention and an awareness of the desirability of evidence-based decisions.

Obstetricians must construct consistent and evidence-based protocols for cervical ripening; formally evaluate physician and patient satisfaction with induction; and design and lead clinical trials to provide answers on the value of marginal indications.

In the meantime, labor induction rates for hospitals and physicians should be monitored, and patients should be educated about the risks of induction so that they can participate in decision making and be better able to balance concerns and benefits. It is quite possible that written consent may become a standard of care before any induction is undertaken.

Until we do so, we should be aware that we may be complicating the uncomplicated.

ELSEVIER GLOBAL MEDICAL NEWS

Induction of Labor

The timing of parturition remains a conundrum in obstetric medicine in that the majority of pregnancies will go to term and enter labor spontaneously, whereas another portion will go post term and often require induction, and still others will enter labor prematurely.

The concept of labor induction, therefore, has become very important in obstetric management, especially in addressing pregnancies that either go post term or pregnancies that require induction because of medical complications in the mother.

Increasingly, however, patients are apt to have labor induced for their own convenience, for personal reasons, for the convenience of the physician, and sometimes for all of these reasons.

This increasingly utilized social option ushers in a whole new perspective on the issue of induction, and the question is raised about whether or not the elective induction of labor brings with it added risk and more complications.

It is for this reason that we decided to develop a Master Class feature on this topic. It gives us the important opportunity to examine and consider the pros and cons of labor induction, the timing of labor induction, and the advisability of the various conditions under which induction can and does occur.

This month's guest professor is Dr. William F. Rayburn, professor and chairman of the department of ob.gyn. at the University of New Mexico, Albuquerque. Dr. Rayburn is a maternal and fetal medicine specialist with a national reputation in this area.

Indications and Contraindications

Indications

Abruptio placentae

Chorioamnionitis

Fetal demise

Pregnancy-induced hypertension

Premature rupture of membranes

Postterm pregnancy

Maternal medical conditions (such as diabetes mellitus, renal disease, chronic pulmonary disease, chronic hypertension)

Fetal compromise (such as severe fetal growth restriction, isoimmunization)

Preeclampsia, eclampsia

Contraindications

Vasa previa or complete placenta previa

Transverse fetal lie

Umbilical cord prolapse

Previous transfundal uterine surgery

Special Attention

One or more previous low-transverse cesarean deliveries

Breech presentation

Maternal heart disease

Multifetal pregnancy

Polyhydramnios

Presenting part above the pelvic inlet

Severe hypertension

Abnormal fetal heart rate patterns not necessitating emergent delivery

Source: Adapted from ACOG Practice Bulletin No. 10, “Induction of Labor” (Nov. 1999).

The American College of Obstetricians and Gynecologists (ACOG) has described examples of commonly accepted indications, contraindications, and clinical conditions requiring special attention for an induction of labor. (See box p. 37.) We must remember that indications for labor induction are often not absolute and need to take maternal and fetal conditions, gestational age, and cervical status into account. Many contraindications are the same as those for either spontaneous labor or vaginal delivery; several obstetric conditions are not contraindications, but do necessitate special attention.

In 1988, the National Center for Health Statistics began requiring hospitals to indicate on birth certificates whether labor was induced or not. This requirement has provided us with remarkable insight into labor induction rates—insight that should cause us to pause, to reflect on available data and our own practices, and to demand that the issue receive more widespread attention.

Over a 10-year period beginning in 1989, the rate of labor induction doubled from about 9% to almost 19% of live births. (See chart.) The trend steadily continued into the new millennium, to the point where, in 2003, nearly 23% of all births involved induction of labor. Clearly, labor induction is one of the most common procedures in obstetrics.

Examining the Increase

The reasons for this significant increase over just 15 years relate to the availability of FDA-approved cervical ripening agents; to both the patient's desire and the physician's convenience; to the acceptance of added risks of cesarean delivery; and to increases in marginal or elective inductions for term pregnancies, especially those past 40 weeks. Inductions in which the reason is not evidence based now account for at least half of all term inductions, or up to 10% of all deliveries. The increase in medically indicated inductions was slower than the overall increase, suggesting that inductions for marginal or elective reasons have risen more rapidly.

Also contributing to the rising rate in inductions is our increasing success with cervical ripening and the fact that, in the current era of ultrasound availability and a more accurate dating of gestational age, we have had to worry less about iatrogenic prematurity.

When considering labor induction, we can view “elective” and “marginal” indications as being very similar, or we can differentiate the two, with “elective” meaning there is no plausible medical or obstetric reason for the induction, and “marginal” referring to cases in which obstetricians face or suspect problems but have no data to suggest that the benefits of labor induction outweigh the risks. I believe it is valuable to consider the terms separately as we attempt to understand the changes in induction rates.

Marginal indications include gestational hypertension; unexplained and mild fetal-growth restriction; idiopathic decreased amniotic fluid (which does not pose substantial danger unless it is accompanied by a recognized complication, such as hypertension or a small-for-gestational-age baby); and a pregnancy beyond 40 weeks. Prospective studies to recommend induction for these and other marginal indications are limited in size or design, or are nonexistent.

There is some rationale behind induction for suspected fetal macrosomia in nondiabetic pregnancies. Theoretically, eliminating further fetal growth should reduce the risks of shoulder dystocia and perhaps of cesarean delivery. However, there is no evidence-based justification for labor induction in these patients. Studies have shown, in fact, that the procedure approximately doubles the cesarean delivery risk, does not reduce neonatal morbidity, and does not appear to reduce the risk of shoulder dystocia.

There is also no published evidence to support the induction of labor for preterm mild preeclampsia, prior shoulder dystocia, and prior cephalopelvic disproportion.

ACOG weighed into the issue by approving “logistic reasons” for labor induction, such as a risk of rapid labor, a patient's unacceptable distance from the hospital, and psychosocial indications. This has left ob.gyns. with a substantial amount of latitude. For instance, one could argue that “psychosocial” reasons could include alleviating the concerns of a mother who previously had a stillborn infant, or alleviating the anxiety of a woman whose spouse is scheduled for deployment to Iraq before the delivery date.

In analyzing the increased rate of labor inductions, we can simply and easily make our own justifications for elective and marginal inductions—we are making our patients happy, for one thing—and put on the back burner the lack of evidence favoring non-medically indicated induction. No matter how appealing our justifications might be, however, we cannot ignore the paucity of published data on benefits, nor can we ignore the data that do exist on the risks of labor induction.

Appreciating the Risks

Studies have shown that induced labor is associated with an increase in epidurals, with the greatest risk of uterine rupture in patients with a scarred uterus, with perhaps an increase in instrumental vaginal deliveries, and with an increase in cesarean deliveries, particularly among nulliparas undergoing an induction with an unfavorable cervix.

Investigators of a large study published in 2005 found a 1.5-fold greater risk of diagnosing a nonreassuring fetal heart rate pattern, a twofold increase in the need for epidural anesthesia, and a 1.5-fold increased risk of having a cesarean delivery among women who had elective inductions of labor compared with women who had spontaneous labor.

The risks of oxytocin use are principally dose related. Excess or undesired uterine hyperstimulation and subsequent fetal heart rate decelerations (“hyperstimulation syndrome”) are the most common side effects. In addition, hyperstimulation is associated with a greater risk of abruptio placentae or uterine rupture. There does not appear to be a significant increase in adverse fetal outcomes from uterine tachysystole.

Uterine hyperstimulation is an adverse effect that is also dose-dependent for prostaglandins (misoprostol, dinoprostone) used as cervical-ripening agents. The potential risks associated with amniotomy include prolapse of the umbilical cord, chorioamnionitis, significant umbilical cord compression, and rupture of vasa previa. With close monitoring and proper precaution, these hazards are fortunately uncommon.

Even if no additional risks are found with elective and marginal indications, it is important to consider issues related to personnel and cost. In addition to increasing the primary cesarean rate—and even a small additional risk of cesarean delivery for nulliparous women who have their labor induced translates into a significantly larger number of cesarean deliveries nationally—labor that is induced requires more one-on-one care and thus more nurses or nursing time.

It also independently leads to significantly longer time in labor and delivery, as well as a prolonged maternal length of hospital stay. Investigators have demonstrated significant differences in the admission-to-delivery times and in-hospital costs between patients who have vaginal deliveries after induced labor as compared with those who have spontaneous labor, as well as with patients who have cesarean deliveries in both scenarios.

Other studies have shown that labor inductions can overload the labor and delivery departments of some hospitals during “popular” midweek times. Downstream, labor induction also leads to an excess number of vaginal births after cesarean (VBAC) or repeat cesarean procedures. I am convinced, moreover, that litigation will be a concern in the future, especially with our armamentarium of cervical ripening agents. When there is a negative outcome after induction, I believe we can anticipate an allegation of unnecessary induction due to the lack of a medical indication.

The frequency of elective inductions and inductions for marginal indications appears to be higher in community hospitals than at university hospitals. A study that my colleagues and I published in 2000 found that 5% of all labor inductions at a university hospital were elective or not medically indicated using the ACOG criteria. At two community hospitals, on the other hand, 44% and 57% of inductions were for elective reasons.

Physicians in academic settings—particularly those involved in clinical trials to assess the effectiveness of therapies for labor induction—are more likely to use the Bishop scoring system. The Bishop score, first described in 1964, is based on cervical dilation, effacement, consistency, and position, as well as on fetal station. Although the scale isn't used much outside of academia, the principles should be consistently and universally applied, particularly the assessment of dilation and cervical consistency.

Planning the Future

Investigators have looked and will continue to look for predictors of success and ideal conditions for labor induction, but at this point in time the only known conditions are a favorable cervix and a patient who has had a previous vaginal delivery. Multiparous women at term generally present with a more favorable cervix.

Right now, roughly half of women who have their labor induced—or roughly 10% of the overall pregnancy population—have an unfavorable cervix. Cesarean rates are high for nulliparas who undergo an induction with an unfavorable cervix. This is a picture that needs widespread attention and an awareness of the desirability of evidence-based decisions.

Obstetricians must construct consistent and evidence-based protocols for cervical ripening; formally evaluate physician and patient satisfaction with induction; and design and lead clinical trials to provide answers on the value of marginal indications.

In the meantime, labor induction rates for hospitals and physicians should be monitored, and patients should be educated about the risks of induction so that they can participate in decision making and be better able to balance concerns and benefits. It is quite possible that written consent may become a standard of care before any induction is undertaken.

Until we do so, we should be aware that we may be complicating the uncomplicated.

ELSEVIER GLOBAL MEDICAL NEWS

Induction of Labor

The timing of parturition remains a conundrum in obstetric medicine in that the majority of pregnancies will go to term and enter labor spontaneously, whereas another portion will go post term and often require induction, and still others will enter labor prematurely.

The concept of labor induction, therefore, has become very important in obstetric management, especially in addressing pregnancies that either go post term or pregnancies that require induction because of medical complications in the mother.

Increasingly, however, patients are apt to have labor induced for their own convenience, for personal reasons, for the convenience of the physician, and sometimes for all of these reasons.

This increasingly utilized social option ushers in a whole new perspective on the issue of induction, and the question is raised about whether or not the elective induction of labor brings with it added risk and more complications.

It is for this reason that we decided to develop a Master Class feature on this topic. It gives us the important opportunity to examine and consider the pros and cons of labor induction, the timing of labor induction, and the advisability of the various conditions under which induction can and does occur.

This month's guest professor is Dr. William F. Rayburn, professor and chairman of the department of ob.gyn. at the University of New Mexico, Albuquerque. Dr. Rayburn is a maternal and fetal medicine specialist with a national reputation in this area.

Indications and Contraindications

Indications

Abruptio placentae

Chorioamnionitis

Fetal demise

Pregnancy-induced hypertension

Premature rupture of membranes

Postterm pregnancy

Maternal medical conditions (such as diabetes mellitus, renal disease, chronic pulmonary disease, chronic hypertension)

Fetal compromise (such as severe fetal growth restriction, isoimmunization)

Preeclampsia, eclampsia

Contraindications

Vasa previa or complete placenta previa

Transverse fetal lie

Umbilical cord prolapse

Previous transfundal uterine surgery

Special Attention

One or more previous low-transverse cesarean deliveries

Breech presentation

Maternal heart disease

Multifetal pregnancy

Polyhydramnios

Presenting part above the pelvic inlet

Severe hypertension

Abnormal fetal heart rate patterns not necessitating emergent delivery

Source: Adapted from ACOG Practice Bulletin No. 10, “Induction of Labor” (Nov. 1999).

We have known for some time that infants who are growth restricted are more prone to problems related to oxygen deprivation and have a higher chance of dying in utero, dying during labor and delivery, and dying during the first hours, weeks, and months of life.

We have learned more recently, moreover, that fetal growth restriction has long-term adverse consequences that extend into adult life. Epidemiologic studies in England in particular show that infants who were growth restricted in utero have a higher chance of developing diabetes, hypertension, stroke, and cardiovascular disease. Significant attention has been paid to the Barker hypothesis, which theorizes that the cardiovascular and endocrine systems of growth-restricted fetuses undergo a sort of intrauterine programming caused by a compromising prenatal environment.

Many aspects of the causes and pathophysiology of growth restriction remain unclear, and none of the therapeutic approaches that have been tried to improve fetal condition—from maternal oxygen administration, various nutritional interventions, and pharmacologic agents to plasma volume expansion and abdominal compression—have been consistently successful or valuable.

However, we have made advances in our understanding of the mechanisms and perinatal risks. We also have made significant progress in diagnosis and management and can today follow an evidence-based approach for managing the complications.

We are at the point today where our role as obstetricians can and should be to identify patients at risk of fetal growth restriction, to sonographically diagnose fetal growth restriction in at-risk patients, to monitor growth-restricted fetuses for in utero compromise, and to ensure a timing of delivery that will maximize gestation while minimizing the risks of continuing the pregnancy.

Identifying the Problem

Fetal growth restriction—or intrauterine growth restriction, as it is sometimes called—refers to the failure of the fetus to realize its optimal growth potential. A baby should be considered growth restricted when sonographically measured fetal dimensions—particularly the abdominal circumference or the estimated fetal weight based on head circumference or diameter, abdominal circumference, and femoral length—deviate below the 10th percentile for gestational age.

Some have advocated for a more rigorous threshold of the 5th percentile, or even the 3rd percentile, and others have suggested using the 15th percentile as a cutoff. The 10th percentile is indeed arbitrary, but for now it is the most commonly used threshold and should be considered the current standard of practice. It strikes the right balance.

Since subnormal growth is defined using gestational age-specific standards, we must establish gestational age as early in the pregnancy as possible, preferably in the first trimester. We must also be as accurate as possible, since overestimating or underestimating the gestational age by even a few days can have significant clinical implications for the discovery of fetal growth restriction.

Use of the crown-rump length presents us with a possible 4- to 5-day variation in gestational age, which is significant but still better than a 2-week variation.

Menstrual history in general is not very reliable, but if there is good concordance between gestational age based on menstrual history and that based on crown-rump length, then one can use the menstrual age. If there is more than a 1-week difference, then I advise using the crown-rump length.

One of the major issues we face in dealing with fetal growth restriction is, of course, that not all babies who are small are abnormal; some are just constitutionally small. Similarly, some babies appear to be normal—and may even be of an appropriate weight for their gestational age—but in reality are facing uteroplacental insufficiency and are not realizing their growth potential. There is, therefore, a definite “gray area” in distinguishing those babies who are truly growth restricted.

This is one reason why the best diagnostic test for fetal growth restriction is serial ultrasonography. Once we identify patients at risk for fetal growth restriction—from those who have preeclampsia and other hypertensive conditions to those with perinatal infection and smoking or substance-abuse problems—we can follow that patient and fetus to get a better sense, for instance, of whether the fetus is small but growing normally or small with progressively declining growth. We can also apply the 10th percentile threshold.

In general, routine growth scans in at-risk patients should start at around 28 weeks and be done every 3–4 weeks unless a diagnosis of growth restriction is made. If a problem such as preeclampsia becomes evident at another time, then serial ultrasonography should commence.

Recognizing the Risks

There is a host of disorders—various maternal, fetal, and placental factors—that can interfere with the mechanisms that regulate fetal growth.

Studies show that hypertensive disorders, for one, are present in 30%–40% of pregnancies that involve fetal growth restriction, and that even without proteinuria, elevated diastolic blood pressure in pregnancy is associated with small-for-gestational-age infants. Preeclampsia is associated with a fourfold increase in the risk of having a small-for-gestational-age infant.

Maternal autoimmune disorders (lupus and antiphospholipid syndrome, for instance), various medications (including certain anticonvulsants, particular β-blockers, cancer chemotherapy, and steroids), cigarette smoking, and even moderate alcohol use, have also been implicated in causing fetal growth restriction. Treatment of some of these conditions, such as the hypertensive conditions, is necessary for the health of the mother but, unfortunately, will not necessarily improve fetal growth.

Treatment of other conditions, such as those involving maternal lifestyle, will definitely lower the severity of the complication. If the mother is a smoker, for instance, a smoking cessation program is absolutely critical. Her fetus's drop in birth weight will be significantly less if smoking is stopped after the first trimester than if it continues throughout the pregnancy.

Fetal chromosomal abnormalities and congenital malformations are also significantly associated with fetal growth restriction, as is perinatal infection. Malaria may be one of the most significant causes of growth restriction in many countries where this disease is endemic. Even in the United States about 5%–10% of all cases of fetal growth restriction can be attributed to viral or protozoan infections in utero.

Bacterial infections have not traditionally been implicated as causes, but there is emerging evidence that subclinical infection and inflammation, as well as extragenital infection, may be associated with growth restriction.

Experts have long recognized a strong association between fetal growth restriction and prematurity, though it's unclear whether there is a true casual relationship.

Monitoring the Growth-Restricted Baby

When a diagnosis of fetal growth restriction is made, our role then focuses on fetal surveillance and the recognition of fetal stress and compromise.

Ultrasonography, first of all, should be done every 2–4 weeks after the diagnosis is made. Of all the additional modalities that we can use for fetal surveillance, umbilical arterial Doppler, which measures blood-flow impedance in the placenta, is one of the most effective tests we have for detecting a fetus who is getting into trouble. It should be used as our primary test. We now have compelling evidence from more than 20 randomized trials that fetal Doppler surveillance significantly improves outcomes (deaths in utero and other medical outcomes) in well-defined, high-risk pregnancies—most notably those involving fetal growth restriction and preeclampsia.

We can supplement Doppler with traditional tests of fetal heart rate monitoring, namely the nonstress test (NST), and evaluation of amniotic fluid volume. Both nonreactive NST and oligohydramnios have been associated with adverse perinatal outcome.

We also can use the biophysical profile (BPP), which incorporates parameters relating to the heart rate pattern, the fluid levels, umbilical artery Doppler, and examination of growth via ultrasound.

Just as the nonstress test does, the BPP has a low false-negative rate but a high false-positive rate. None of these additional tests is backed by the “level 1” evidence (randomized controlled trials) that Doppler carries, but they have essentially become standards of care. When used once a week, the tests are a valuable part of management, and I have incorporated them into my own evidence-based management guidelines. (See chart below.)

Usually, ominous changes in the fetal heart rate pattern or the BPP will follow nonreassuing Doppler indices—a fact that is indicative not only of the value of umbilical arterial Doppler but the value of these other tests in helping us to assess fetal distress and compromise, and the need for delivery, as completely as possible.

If our umbilical arterial Doppler shows an absence of flow at the end of the cardiac cycle and the other tests are normal, we can—if the pregnancy hasn't reached 34 weeks—step up the frequency of our other tests and attempt to carry the gestation through a bit further. If the end-diastolic flow is reversed, however, we need to intervene promptly. Reversed end-diastolic flow is an ominous sign.

Other ominous signs are a BPP score of 4 or less; an amniotic fluid index of 5 cm or less or a single deepest pocket less than 2 cm; and nonreassuring fetal heart rate patterns such as persistent nonreactive NSTs, continuous deceleration, and poor heart rate variability from one cardiac cycle to another.

The use of venous Doppler sonography is getting more attention today as another back-up test for evaluating fetal well-being when the umbilical arterial Doppler shows absent end-diastolic flow.

Doppler assessment of flow patterns through the inferior vena cava, umbilical vein, and the ductus venosus have all been suggested as supplementary tests—experimentation is underway particularly in Europe—but it is flow through the ductus venosus that may warrant the most attention at this point in time in institutions that have appropriately trained personnel. When flow during atrial contraction is absent or reversed in the ductus venosus, urgent intervention is usually necessary.

Our decisions to deliver, of course, should always be highly individualized, taking into account gestational age, the progression of change, institutional resources and expertise, and other issues. In general, though, once we're at or beyond 34 weeks of gestation, there is no benefit to prolonging the pregnancy if we have any ominous findings.

The absence of end-diastolic flow on the umbilical arterial Doppler, for instance, should prompt delivery once we've reached 34 weeks, whereas before 34 weeks we could instead intensify surveillance and watch for additional ominous findings. (Many, however, would use a cut-off of 32 completed weeks based on outcomes in the intensive care nursery of their institution).

We also should not allow pregnancies involving growth restriction to become postdated. There are no clear-cut guidelines addressing the question of whether we should induce babies who have come to term, but if the baby is in jeopardy—if there are multiple signs of compromise or distress—the baby will have a limited ability to tolerate labor, and a cesarean section is best.

Our most difficult decisions come with gestations of less than 28 weeks. Unfortunately, a recent randomized controlled trial of delivering early vs. delaying delivery (the Growth Restriction Invention Trial) brought us no clear answers.

This means that we have to continue utilizing our clinical judgment about the respective risks of a hostile intrauterine environment and the risk of pulmonary immaturity, and have a compassionate, nonpatronizing discussion with the parents. In general, if multiple parameters are abnormal, too much waiting will deprive the fetus of any chance of survival.

Umbilical arterial Doppler is an effective tool for monitoring fetal growth restriction. The image on the left shows normal end-diastolic flow in the umbilical artery; the image on the right shows absent end-diastolic flow. Photos courtesy Dr. Dev Maulik

ELSEVIER GLOBAL MEDICAL NEWS

Fetal Growth Assessment

On the other hand, in some pregnancies we encounter excessive fetal growth or restrictive fetal growth. Both of these conditions require careful attention, careful assessment and, sometimes, careful intervention.

Fetal growth restriction may occur under certain clinical conditions. Some of these conditions may be nutritional, some may be related to medical conditions such as diabetes or hypertension, and some may be due to a congenital cause or even an environmental cause such as smoking. Regardless of the actual etiology, if indeed fetal growth restriction is suspected or detected, it requires intense fetal surveillance because of the potential complications that can occur either in the short term or the long term. Some of these complications can result in significant comorbidities or even mortality.

It is for this reason that this month's Master Class will provide an in-depth look at fetal growth restriction and some of the diagnostic and management approaches that may be employed. I am pleased to welcome as our guest professor Dev Maulik, M.D., Ph.D., who is currently chair of obstetrics and gynecology at Winthrop University Hospital in Mineola, N.Y., and professor of obstetrics and gynecology at the State University of New York at Stony Brook.

Dr. Maulik has written extensively about low birth weight and prematurity, predicting adverse perinatal outcome, and detecting and managing fetal growth restriction. He has recently accepted a new appointment as professor and chair of obstetrics and gynecology at the University of Missouri-Kansas City.

We have known for some time that infants who are growth restricted are more prone to problems related to oxygen deprivation and have a higher chance of dying in utero, dying during labor and delivery, and dying during the first hours, weeks, and months of life.

We have learned more recently, moreover, that fetal growth restriction has long-term adverse consequences that extend into adult life. Epidemiologic studies in England in particular show that infants who were growth restricted in utero have a higher chance of developing diabetes, hypertension, stroke, and cardiovascular disease. Significant attention has been paid to the Barker hypothesis, which theorizes that the cardiovascular and endocrine systems of growth-restricted fetuses undergo a sort of intrauterine programming caused by a compromising prenatal environment.

Many aspects of the causes and pathophysiology of growth restriction remain unclear, and none of the therapeutic approaches that have been tried to improve fetal condition—from maternal oxygen administration, various nutritional interventions, and pharmacologic agents to plasma volume expansion and abdominal compression—have been consistently successful or valuable.

However, we have made advances in our understanding of the mechanisms and perinatal risks. We also have made significant progress in diagnosis and management and can today follow an evidence-based approach for managing the complications.

We are at the point today where our role as obstetricians can and should be to identify patients at risk of fetal growth restriction, to sonographically diagnose fetal growth restriction in at-risk patients, to monitor growth-restricted fetuses for in utero compromise, and to ensure a timing of delivery that will maximize gestation while minimizing the risks of continuing the pregnancy.

Identifying the Problem

Fetal growth restriction—or intrauterine growth restriction, as it is sometimes called—refers to the failure of the fetus to realize its optimal growth potential. A baby should be considered growth restricted when sonographically measured fetal dimensions—particularly the abdominal circumference or the estimated fetal weight based on head circumference or diameter, abdominal circumference, and femoral length—deviate below the 10th percentile for gestational age.

Some have advocated for a more rigorous threshold of the 5th percentile, or even the 3rd percentile, and others have suggested using the 15th percentile as a cutoff. The 10th percentile is indeed arbitrary, but for now it is the most commonly used threshold and should be considered the current standard of practice. It strikes the right balance.

Since subnormal growth is defined using gestational age-specific standards, we must establish gestational age as early in the pregnancy as possible, preferably in the first trimester. We must also be as accurate as possible, since overestimating or underestimating the gestational age by even a few days can have significant clinical implications for the discovery of fetal growth restriction.

Use of the crown-rump length presents us with a possible 4- to 5-day variation in gestational age, which is significant but still better than a 2-week variation.

Menstrual history in general is not very reliable, but if there is good concordance between gestational age based on menstrual history and that based on crown-rump length, then one can use the menstrual age. If there is more than a 1-week difference, then I advise using the crown-rump length.

One of the major issues we face in dealing with fetal growth restriction is, of course, that not all babies who are small are abnormal; some are just constitutionally small. Similarly, some babies appear to be normal—and may even be of an appropriate weight for their gestational age—but in reality are facing uteroplacental insufficiency and are not realizing their growth potential. There is, therefore, a definite “gray area” in distinguishing those babies who are truly growth restricted.

This is one reason why the best diagnostic test for fetal growth restriction is serial ultrasonography. Once we identify patients at risk for fetal growth restriction—from those who have preeclampsia and other hypertensive conditions to those with perinatal infection and smoking or substance-abuse problems—we can follow that patient and fetus to get a better sense, for instance, of whether the fetus is small but growing normally or small with progressively declining growth. We can also apply the 10th percentile threshold.

In general, routine growth scans in at-risk patients should start at around 28 weeks and be done every 3–4 weeks unless a diagnosis of growth restriction is made. If a problem such as preeclampsia becomes evident at another time, then serial ultrasonography should commence.

Recognizing the Risks

There is a host of disorders—various maternal, fetal, and placental factors—that can interfere with the mechanisms that regulate fetal growth.

Studies show that hypertensive disorders, for one, are present in 30%–40% of pregnancies that involve fetal growth restriction, and that even without proteinuria, elevated diastolic blood pressure in pregnancy is associated with small-for-gestational-age infants. Preeclampsia is associated with a fourfold increase in the risk of having a small-for-gestational-age infant.

Maternal autoimmune disorders (lupus and antiphospholipid syndrome, for instance), various medications (including certain anticonvulsants, particular β-blockers, cancer chemotherapy, and steroids), cigarette smoking, and even moderate alcohol use, have also been implicated in causing fetal growth restriction. Treatment of some of these conditions, such as the hypertensive conditions, is necessary for the health of the mother but, unfortunately, will not necessarily improve fetal growth.

Treatment of other conditions, such as those involving maternal lifestyle, will definitely lower the severity of the complication. If the mother is a smoker, for instance, a smoking cessation program is absolutely critical. Her fetus's drop in birth weight will be significantly less if smoking is stopped after the first trimester than if it continues throughout the pregnancy.

Fetal chromosomal abnormalities and congenital malformations are also significantly associated with fetal growth restriction, as is perinatal infection. Malaria may be one of the most significant causes of growth restriction in many countries where this disease is endemic. Even in the United States about 5%–10% of all cases of fetal growth restriction can be attributed to viral or protozoan infections in utero.

Bacterial infections have not traditionally been implicated as causes, but there is emerging evidence that subclinical infection and inflammation, as well as extragenital infection, may be associated with growth restriction.

Experts have long recognized a strong association between fetal growth restriction and prematurity, though it's unclear whether there is a true casual relationship.

Monitoring the Growth-Restricted Baby

When a diagnosis of fetal growth restriction is made, our role then focuses on fetal surveillance and the recognition of fetal stress and compromise.

Ultrasonography, first of all, should be done every 2–4 weeks after the diagnosis is made. Of all the additional modalities that we can use for fetal surveillance, umbilical arterial Doppler, which measures blood-flow impedance in the placenta, is one of the most effective tests we have for detecting a fetus who is getting into trouble. It should be used as our primary test. We now have compelling evidence from more than 20 randomized trials that fetal Doppler surveillance significantly improves outcomes (deaths in utero and other medical outcomes) in well-defined, high-risk pregnancies—most notably those involving fetal growth restriction and preeclampsia.

We can supplement Doppler with traditional tests of fetal heart rate monitoring, namely the nonstress test (NST), and evaluation of amniotic fluid volume. Both nonreactive NST and oligohydramnios have been associated with adverse perinatal outcome.

We also can use the biophysical profile (BPP), which incorporates parameters relating to the heart rate pattern, the fluid levels, umbilical artery Doppler, and examination of growth via ultrasound.

Just as the nonstress test does, the BPP has a low false-negative rate but a high false-positive rate. None of these additional tests is backed by the “level 1” evidence (randomized controlled trials) that Doppler carries, but they have essentially become standards of care. When used once a week, the tests are a valuable part of management, and I have incorporated them into my own evidence-based management guidelines. (See chart below.)

Usually, ominous changes in the fetal heart rate pattern or the BPP will follow nonreassuing Doppler indices—a fact that is indicative not only of the value of umbilical arterial Doppler but the value of these other tests in helping us to assess fetal distress and compromise, and the need for delivery, as completely as possible.

If our umbilical arterial Doppler shows an absence of flow at the end of the cardiac cycle and the other tests are normal, we can—if the pregnancy hasn't reached 34 weeks—step up the frequency of our other tests and attempt to carry the gestation through a bit further. If the end-diastolic flow is reversed, however, we need to intervene promptly. Reversed end-diastolic flow is an ominous sign.

Other ominous signs are a BPP score of 4 or less; an amniotic fluid index of 5 cm or less or a single deepest pocket less than 2 cm; and nonreassuring fetal heart rate patterns such as persistent nonreactive NSTs, continuous deceleration, and poor heart rate variability from one cardiac cycle to another.

The use of venous Doppler sonography is getting more attention today as another back-up test for evaluating fetal well-being when the umbilical arterial Doppler shows absent end-diastolic flow.

Doppler assessment of flow patterns through the inferior vena cava, umbilical vein, and the ductus venosus have all been suggested as supplementary tests—experimentation is underway particularly in Europe—but it is flow through the ductus venosus that may warrant the most attention at this point in time in institutions that have appropriately trained personnel. When flow during atrial contraction is absent or reversed in the ductus venosus, urgent intervention is usually necessary.

Our decisions to deliver, of course, should always be highly individualized, taking into account gestational age, the progression of change, institutional resources and expertise, and other issues. In general, though, once we're at or beyond 34 weeks of gestation, there is no benefit to prolonging the pregnancy if we have any ominous findings.

The absence of end-diastolic flow on the umbilical arterial Doppler, for instance, should prompt delivery once we've reached 34 weeks, whereas before 34 weeks we could instead intensify surveillance and watch for additional ominous findings. (Many, however, would use a cut-off of 32 completed weeks based on outcomes in the intensive care nursery of their institution).

We also should not allow pregnancies involving growth restriction to become postdated. There are no clear-cut guidelines addressing the question of whether we should induce babies who have come to term, but if the baby is in jeopardy—if there are multiple signs of compromise or distress—the baby will have a limited ability to tolerate labor, and a cesarean section is best.

Our most difficult decisions come with gestations of less than 28 weeks. Unfortunately, a recent randomized controlled trial of delivering early vs. delaying delivery (the Growth Restriction Invention Trial) brought us no clear answers.

This means that we have to continue utilizing our clinical judgment about the respective risks of a hostile intrauterine environment and the risk of pulmonary immaturity, and have a compassionate, nonpatronizing discussion with the parents. In general, if multiple parameters are abnormal, too much waiting will deprive the fetus of any chance of survival.

Umbilical arterial Doppler is an effective tool for monitoring fetal growth restriction. The image on the left shows normal end-diastolic flow in the umbilical artery; the image on the right shows absent end-diastolic flow. Photos courtesy Dr. Dev Maulik

ELSEVIER GLOBAL MEDICAL NEWS

Fetal Growth Assessment

On the other hand, in some pregnancies we encounter excessive fetal growth or restrictive fetal growth. Both of these conditions require careful attention, careful assessment and, sometimes, careful intervention.

Fetal growth restriction may occur under certain clinical conditions. Some of these conditions may be nutritional, some may be related to medical conditions such as diabetes or hypertension, and some may be due to a congenital cause or even an environmental cause such as smoking. Regardless of the actual etiology, if indeed fetal growth restriction is suspected or detected, it requires intense fetal surveillance because of the potential complications that can occur either in the short term or the long term. Some of these complications can result in significant comorbidities or even mortality.

It is for this reason that this month's Master Class will provide an in-depth look at fetal growth restriction and some of the diagnostic and management approaches that may be employed. I am pleased to welcome as our guest professor Dev Maulik, M.D., Ph.D., who is currently chair of obstetrics and gynecology at Winthrop University Hospital in Mineola, N.Y., and professor of obstetrics and gynecology at the State University of New York at Stony Brook.

Dr. Maulik has written extensively about low birth weight and prematurity, predicting adverse perinatal outcome, and detecting and managing fetal growth restriction. He has recently accepted a new appointment as professor and chair of obstetrics and gynecology at the University of Missouri-Kansas City.

We have known for some time that infants who are growth restricted are more prone to problems related to oxygen deprivation and have a higher chance of dying in utero, dying during labor and delivery, and dying during the first hours, weeks, and months of life.

We have learned more recently, moreover, that fetal growth restriction has long-term adverse consequences that extend into adult life. Epidemiologic studies in England in particular show that infants who were growth restricted in utero have a higher chance of developing diabetes, hypertension, stroke, and cardiovascular disease. Significant attention has been paid to the Barker hypothesis, which theorizes that the cardiovascular and endocrine systems of growth-restricted fetuses undergo a sort of intrauterine programming caused by a compromising prenatal environment.

Many aspects of the causes and pathophysiology of growth restriction remain unclear, and none of the therapeutic approaches that have been tried to improve fetal condition—from maternal oxygen administration, various nutritional interventions, and pharmacologic agents to plasma volume expansion and abdominal compression—have been consistently successful or valuable.

However, we have made advances in our understanding of the mechanisms and perinatal risks. We also have made significant progress in diagnosis and management and can today follow an evidence-based approach for managing the complications.

We are at the point today where our role as obstetricians can and should be to identify patients at risk of fetal growth restriction, to sonographically diagnose fetal growth restriction in at-risk patients, to monitor growth-restricted fetuses for in utero compromise, and to ensure a timing of delivery that will maximize gestation while minimizing the risks of continuing the pregnancy.

Identifying the Problem

Fetal growth restriction—or intrauterine growth restriction, as it is sometimes called—refers to the failure of the fetus to realize its optimal growth potential. A baby should be considered growth restricted when sonographically measured fetal dimensions—particularly the abdominal circumference or the estimated fetal weight based on head circumference or diameter, abdominal circumference, and femoral length—deviate below the 10th percentile for gestational age.

Some have advocated for a more rigorous threshold of the 5th percentile, or even the 3rd percentile, and others have suggested using the 15th percentile as a cutoff. The 10th percentile is indeed arbitrary, but for now it is the most commonly used threshold and should be considered the current standard of practice. It strikes the right balance.

Since subnormal growth is defined using gestational age-specific standards, we must establish gestational age as early in the pregnancy as possible, preferably in the first trimester. We must also be as accurate as possible, since overestimating or underestimating the gestational age by even a few days can have significant clinical implications for the discovery of fetal growth restriction.

Use of the crown-rump length presents us with a possible 4- to 5-day variation in gestational age, which is significant but still better than a 2-week variation.

Menstrual history in general is not very reliable, but if there is good concordance between gestational age based on menstrual history and that based on crown-rump length, then one can use the menstrual age. If there is more than a 1-week difference, then I advise using the crown-rump length.

One of the major issues we face in dealing with fetal growth restriction is, of course, that not all babies who are small are abnormal; some are just constitutionally small. Similarly, some babies appear to be normal—and may even be of an appropriate weight for their gestational age—but in reality are facing uteroplacental insufficiency and are not realizing their growth potential. There is, therefore, a definite “gray area” in distinguishing those babies who are truly growth restricted.

This is one reason why the best diagnostic test for fetal growth restriction is serial ultrasonography. Once we identify patients at risk for fetal growth restriction—from those who have preeclampsia and other hypertensive conditions to those with perinatal infection and smoking or substance-abuse problems—we can follow that patient and fetus to get a better sense, for instance, of whether the fetus is small but growing normally or small with progressively declining growth. We can also apply the 10th percentile threshold.

In general, routine growth scans in at-risk patients should start at around 28 weeks and be done every 3–4 weeks unless a diagnosis of growth restriction is made. If a problem such as preeclampsia becomes evident at another time, then serial ultrasonography should commence.

Recognizing the Risks

There is a host of disorders—various maternal, fetal, and placental factors—that can interfere with the mechanisms that regulate fetal growth.

Studies show that hypertensive disorders, for one, are present in 30%–40% of pregnancies that involve fetal growth restriction, and that even without proteinuria, elevated diastolic blood pressure in pregnancy is associated with small-for-gestational-age infants. Preeclampsia is associated with a fourfold increase in the risk of having a small-for-gestational-age infant.

Maternal autoimmune disorders (lupus and antiphospholipid syndrome, for instance), various medications (including certain anticonvulsants, particular β-blockers, cancer chemotherapy, and steroids), cigarette smoking, and even moderate alcohol use, have also been implicated in causing fetal growth restriction. Treatment of some of these conditions, such as the hypertensive conditions, is necessary for the health of the mother but, unfortunately, will not necessarily improve fetal growth.

Treatment of other conditions, such as those involving maternal lifestyle, will definitely lower the severity of the complication. If the mother is a smoker, for instance, a smoking cessation program is absolutely critical. Her fetus's drop in birth weight will be significantly less if smoking is stopped after the first trimester than if it continues throughout the pregnancy.

Fetal chromosomal abnormalities and congenital malformations are also significantly associated with fetal growth restriction, as is perinatal infection. Malaria may be one of the most significant causes of growth restriction in many countries where this disease is endemic. Even in the United States about 5%–10% of all cases of fetal growth restriction can be attributed to viral or protozoan infections in utero.

Bacterial infections have not traditionally been implicated as causes, but there is emerging evidence that subclinical infection and inflammation, as well as extragenital infection, may be associated with growth restriction.

Experts have long recognized a strong association between fetal growth restriction and prematurity, though it's unclear whether there is a true casual relationship.

Monitoring the Growth-Restricted Baby

When a diagnosis of fetal growth restriction is made, our role then focuses on fetal surveillance and the recognition of fetal stress and compromise.

Ultrasonography, first of all, should be done every 2–4 weeks after the diagnosis is made. Of all the additional modalities that we can use for fetal surveillance, umbilical arterial Doppler, which measures blood-flow impedance in the placenta, is one of the most effective tests we have for detecting a fetus who is getting into trouble. It should be used as our primary test. We now have compelling evidence from more than 20 randomized trials that fetal Doppler surveillance significantly improves outcomes (deaths in utero and other medical outcomes) in well-defined, high-risk pregnancies—most notably those involving fetal growth restriction and preeclampsia.

We can supplement Doppler with traditional tests of fetal heart rate monitoring, namely the nonstress test (NST), and evaluation of amniotic fluid volume. Both nonreactive NST and oligohydramnios have been associated with adverse perinatal outcome.

We also can use the biophysical profile (BPP), which incorporates parameters relating to the heart rate pattern, the fluid levels, umbilical artery Doppler, and examination of growth via ultrasound.

Just as the nonstress test does, the BPP has a low false-negative rate but a high false-positive rate. None of these additional tests is backed by the “level 1” evidence (randomized controlled trials) that Doppler carries, but they have essentially become standards of care. When used once a week, the tests are a valuable part of management, and I have incorporated them into my own evidence-based management guidelines. (See chart below.)

Usually, ominous changes in the fetal heart rate pattern or the BPP will follow nonreassuing Doppler indices—a fact that is indicative not only of the value of umbilical arterial Doppler but the value of these other tests in helping us to assess fetal distress and compromise, and the need for delivery, as completely as possible.

If our umbilical arterial Doppler shows an absence of flow at the end of the cardiac cycle and the other tests are normal, we can—if the pregnancy hasn't reached 34 weeks—step up the frequency of our other tests and attempt to carry the gestation through a bit further. If the end-diastolic flow is reversed, however, we need to intervene promptly. Reversed end-diastolic flow is an ominous sign.

Other ominous signs are a BPP score of 4 or less; an amniotic fluid index of 5 cm or less or a single deepest pocket less than 2 cm; and nonreassuring fetal heart rate patterns such as persistent nonreactive NSTs, continuous deceleration, and poor heart rate variability from one cardiac cycle to another.

The use of venous Doppler sonography is getting more attention today as another back-up test for evaluating fetal well-being when the umbilical arterial Doppler shows absent end-diastolic flow.

Doppler assessment of flow patterns through the inferior vena cava, umbilical vein, and the ductus venosus have all been suggested as supplementary tests—experimentation is underway particularly in Europe—but it is flow through the ductus venosus that may warrant the most attention at this point in time in institutions that have appropriately trained personnel. When flow during atrial contraction is absent or reversed in the ductus venosus, urgent intervention is usually necessary.

Our decisions to deliver, of course, should always be highly individualized, taking into account gestational age, the progression of change, institutional resources and expertise, and other issues. In general, though, once we're at or beyond 34 weeks of gestation, there is no benefit to prolonging the pregnancy if we have any ominous findings.

The absence of end-diastolic flow on the umbilical arterial Doppler, for instance, should prompt delivery once we've reached 34 weeks, whereas before 34 weeks we could instead intensify surveillance and watch for additional ominous findings. (Many, however, would use a cut-off of 32 completed weeks based on outcomes in the intensive care nursery of their institution).

We also should not allow pregnancies involving growth restriction to become postdated. There are no clear-cut guidelines addressing the question of whether we should induce babies who have come to term, but if the baby is in jeopardy—if there are multiple signs of compromise or distress—the baby will have a limited ability to tolerate labor, and a cesarean section is best.

Our most difficult decisions come with gestations of less than 28 weeks. Unfortunately, a recent randomized controlled trial of delivering early vs. delaying delivery (the Growth Restriction Invention Trial) brought us no clear answers.

This means that we have to continue utilizing our clinical judgment about the respective risks of a hostile intrauterine environment and the risk of pulmonary immaturity, and have a compassionate, nonpatronizing discussion with the parents. In general, if multiple parameters are abnormal, too much waiting will deprive the fetus of any chance of survival.

Umbilical arterial Doppler is an effective tool for monitoring fetal growth restriction. The image on the left shows normal end-diastolic flow in the umbilical artery; the image on the right shows absent end-diastolic flow. Photos courtesy Dr. Dev Maulik

ELSEVIER GLOBAL MEDICAL NEWS

Fetal Growth Assessment

On the other hand, in some pregnancies we encounter excessive fetal growth or restrictive fetal growth. Both of these conditions require careful attention, careful assessment and, sometimes, careful intervention.

Fetal growth restriction may occur under certain clinical conditions. Some of these conditions may be nutritional, some may be related to medical conditions such as diabetes or hypertension, and some may be due to a congenital cause or even an environmental cause such as smoking. Regardless of the actual etiology, if indeed fetal growth restriction is suspected or detected, it requires intense fetal surveillance because of the potential complications that can occur either in the short term or the long term. Some of these complications can result in significant comorbidities or even mortality.

It is for this reason that this month's Master Class will provide an in-depth look at fetal growth restriction and some of the diagnostic and management approaches that may be employed. I am pleased to welcome as our guest professor Dev Maulik, M.D., Ph.D., who is currently chair of obstetrics and gynecology at Winthrop University Hospital in Mineola, N.Y., and professor of obstetrics and gynecology at the State University of New York at Stony Brook.

Dr. Maulik has written extensively about low birth weight and prematurity, predicting adverse perinatal outcome, and detecting and managing fetal growth restriction. He has recently accepted a new appointment as professor and chair of obstetrics and gynecology at the University of Missouri-Kansas City.

Dr. Roberto Romero: The common pathway of parturition consists of the anatomical, physiologic, biochemical, and clinical events that occur in the mother and/or fetus in both term and preterm labor. The uterine components of this pathway include increased uterine contractility; cervical ripening (dilatation and effacement); and membrane/decidual activation. In most women, these components are typically activated in a synchronous manner in spontaneous labor at term. Indeed, most women admitted in labor have uterine contractions, cervical changes and—sometimes—rupture of membranes.

However, in some cases the activation of the common pathway may be asynchronous. For example, a patient may have increased uterine contractility, but the cervix undergoes very little change over time. This is what is called a prolonged latent phase of labor. In 10% of cases, patients have spontaneous rupture of membranes without myometrial contractility. This is evidence that membrane/decidual activation has occurred without recruitment of the myometrium.

Asynchronous activation is more common in the preterm gestation. Many patients with suspected preterm labor will present with increased uterine contractility without cervical changes. Others will present with the clinical picture of cervical insufficiency (which used to be called cervical incompetence). Finally, some women will present with preterm premature rupture of the membranes (pPROM), which is premature membrane/decidual activation.

EAR: What is the importance of the concept of the common pathway?

RR: Much of the clinical management and research in understanding the causes of premature labor, treatment, and prevention has been focused on the elements of the common terminal pathway. For example, we have used uterine monitors to detect an increase in uterine contractility and tocolytic agents to treat increased myometrial contractility. We use ultrasound to identify patients with a short cervix who are at risk for preterm delivery. In some cases, we have placed cervical cerclage in patients at risk. Finally, we have used fetal fibronectin to detect decidual/membrane deactivation. A positive fetal fibronectin is an indicator that disruption of the choriodecidual interface has occurred. Yet these interventions aim to treat preterm labor as a symptom, without first identifying and understanding the underlying pathology that sets it in motion. Progress on this front is now being made.

EAR: What is the difference between spontaneous labor at term and preterm labor?

RR: We propose that spontaneous labor at term is the inevitable process that occurs when the capacity of the mother to support the fetus in utero has been reached. In other words, when the fetus has achieved maturity and is ready to face extrauterine life, it signals the onset of labor and engages the cooperation of the mother in this process.

In contrast, we propose that premature labor results from a pathologic insult that activates the common pathway of parturition. Before the development of newborn special care units, extreme prematurity was nearly always lethal. Thus, being born preterm is likely to result from such a severe pathologic process that it threatens the survival of the mother and/or baby.

In summary, spontaneous labor at term results from physiologic activation of the common pathway, whereas preterm labor would result from pathologic activation of the pathway.

EAR: What is the evidence that premature labor is a heterogeneous condition?

RR: My laboratory and other groups have generated evidence that the pattern of uterine gene expression—also known as the transcriptional profile—is different in patients with different causes of premature labor. A transcriptional profile is a snapshot of genes that are being upregulated or downregulated at a particular point in time.

We have demonstrated experimentally that the transcriptional profile in the uteri of mice that go into labor as a result of infection is different from the transcriptional profile of mice that go into labor because of ovariectomy (a model of progesterone withdrawal).

We have also examined the transcriptional profile of the chorioamniotic membranes in women with preterm labor and intact membranes vs. women with pPROM, in each case studying women with and without histologic chorioamnionitis. These studies have demonstrated that the patterns of gene expression are very different in these four groups, even though the clinical presentations are similar.

Collectively, the observations in patients and animals suggest that premature labor is a heterogeneous condition. Although they share a common pathway (uterine contractility and cervical dilatation and/or membrane rupture or activation), there are multiple causes for the preterm parturition syndrome.

Using the tools of high-dimensional biology, we have learned that premature labor is not simply labor before its time, but rather a disease state. It is caused by different pathologic processes with both environmental components and genetic components. For example, in infection-induced preterm labor (OB.GYN. NEWS, July 1, 2006, p. 42), the environmental component is represented by the microorganisms that cause the infection. The genetic component is the factor that predisposes some women to have an intrauterine infection or to respond more severely to that infection.

A prime example of the importance of genetics lies with the fetus. In pregnancy, we have not one patient, but two. Accumulating evidence indicates that the fetus plays a central role in the initiation of labor in both animals and humans. In humans, mothers of fetuses who have mounted a severe fetal inflammatory response are more likely to go into labor than are mothers of fetuses who have not mounted a fetal inflammatory response to infection. The magnitude and severity of the inflammatory response are under genetic control.

Thus, in pregnancy, the genetic makeup of two hosts—mother and child—plays a role in determining the susceptibility and response to infection. It is a unique situation in medicine.

EAR: Why do you call preterm parturition a syndrome?

RR: The current taxonomy of disease in obstetrics is largely based on the symptoms and signs exhibited by the mother, not the mechanisms of disease responsible for the clinical presentations. A syndrome is a combination of symptoms and signs that form a distinct clinical picture indicative of a particular disorder. Implicit in this definition is the fact that a syndrome can have multiple causes. Our emphasis in referring to premature parturition as a syndrome is that it will help both patients and physicians readjust the unreasonable expectation that one test can diagnose or identify all women at risk, and that one treatment will be effective for all women in premature labor, regardless of the cause.

EAR: In the last Master Class, we spoke about the role of infection and inflammation as a cause of premature labor. You have shared with us a figure that shows other causes of preterm parturition. I would like to discuss the other causes with you. Can you tell us how uterine ischemia and/or vascular pathology cause premature labor?

RR: Placental histology suggests that vascular lesions involving maternal or fetal circulation constitute the second most common apparent cause of preterm labor and pPROM, after inflammation.

Research in this area has focused on elevated rates of vascular abnormalities in women with spontaneous preterm labor with intact membranes and pPROM, compared with women who deliver at term. The vascular lesions have been found on both the maternal and fetal side. For example, thrombosis of the decidual vessels attached to the placenta and failure of physiologic transformation in the myometrial segment of the spiral arteries are generally considered maternal lesions. Abnormalities in the fetal circulation linked to preterm labor include a decreased number of arterioles in the villi and fetal arterial thrombosis.

Whereas uteroplacental ischemia is the driver of vascular events, the leading candidate to explain the molecular mechanisms responsible is the renin-angiotensin system; in severe uteroplacental ischemia, the enzyme thrombin is emerging as an important activator of preterm labor associated with vaginal bleeding.

The work of Dr. Mark Phillippe and Dr. Michal Elovitz has demonstrated in animal models and in vivo investigations that whole blood—but not heparinized blood—raises contractile activity of the uterine muscle, and that such increased uterine activity can be blocked with a thrombin antagonist.

Moreover, women with preterm labor have higher concentrations of thrombin-antithrombin (TAT) complexes in amniotic fluid and in maternal plasma than do women without preterm labor. Similarly, Dr. Todd Rosen and Dr. Charles Lockwood have provided evidence that women destined to develop pPROM have higher concentrations of thrombin weeks before the development of complications. Therefore, thrombin is a potential initiator not only of the rupture of membranes (via stromal cells and matrix metalloproteinase 1), but also of uterine contractility and preterm labor.

The thrombin connection would help to explain why retroplacental hematomas in early pregnancy are associated with preterm delivery, and why vaginal bleeding in the first or second trimester is a risk factor for preterm birth with intact or ruptured membranes.

EAR: What is the evidence that uterine overdistention is a cause of preterm parturition?

RR: Obstetricians and midwives know that multiple gestation is a risk factor for preterm delivery, and that the higher the order of multiple gestation, the greater the risk for preterm birth. Patients with polyhydramnios resulting from a congenital fetal anomaly are also at risk for spontaneous preterm labor and delivery. These two conditions are probably mediated by uterine overdistention. It is likely that the same is the case for patients who have müllerian duct abnormalities in which the uterine cavity is small. One example is congenital hypoplastic uterus and another is the abnormal uterus resulting from diethylstilbestrol exposure.

EAR: What is the mechanism linking uterine overdistention with premature labor?

RR: This is an area that requires further work. In general terms, mechanical signals triggered by uterine stretch may lead to preterm labor—as in multiple gestations and polyhydramnios—although the precise mechanisms involved remain unknown.

Experimental distention of the uterus with a saline-filled balloon rapidly prompted regular uterine contractions in women carrying live term fetuses or dead fetuses. All were delivered within 21 hours.

Stephen Lye, Ph.D., and his group at the University of Toronto have found increased expression of oxytocin receptor, connexin 43, and the c-fos mRNA in the myometrium. Gillian Bryant-Greenwood, Ph.D., at the University of Hawaii, Honolulu, has shown that stretching of the chorioamniotic membranes in experimental models can increase the expression of interleukin-8 and also another cytokine called visfatin, which can have important effects on the integrity of the membranes.

EAR: I noticed that cervical pathology is also a potential cause of the preterm parturition syndrome. Is this the same as cervical insufficiency?

RR: Yes. Cervical insufficiency is also a cause of the preterm parturition syndrome and can result from congenital disorders, surgical trauma, trauma, or infection. Although cervical length has been touted as a predictor of preterm birth, it is important to remember that a short cervix is not always a ripe cervix. Once again, more research is needed to identify the events that occur as a result of cervical insufficiency, and the ways these affect the initiation of preterm and term labor.

EAR: What about abnormal allograft reaction?

RR: The fetoplacental unit has been called “nature's most successful transplant,” or—more accurately—a “semiallograft.” The mechanisms that bring about tolerance of this semiallograft are poorly understood. However, transplants of solid organs are tolerated through the establishment of microchimerism in the transplanted organ as well as in the host. Therefore, we consider that microchimerism in pregnancy is probably important for tolerance of the fetoplacental unit. However, I anticipate that under pathologic conditions—just as in the case of transplants—tolerance of the fetoplacental unit may break down, which in turn may lead to a unique form of rejection of the fetoplacental unit. Unraveling the mechanisms of this rejectionlike process is a fascinating challenge. However, ob.gyns. know that the frequency of adverse pregnancy outcomes is higher in mothers who have pregnancies after embryo or egg donation. Under these circumstances, the placenta and fetus are totally foreign because they do not have the normal 50% genetic endowment from the mother. The complications noticed in these pregnancies include not only preeclampsia and growth restriction, but also preterm labor.

EAR: What is the evidence that an allergic phenomenon could be associated with premature labor, and could antihistamines be a treatment for premature labor?

RR: This idea came to me after seeing the relative of an obstetrician who went into premature labor after eating shellfish to which she was allergic. Around that time, we observed that a group of women in premature labor had eosinophils in the amniotic cavity. Eosinophils are associated with an allergiclike phenomenon and are not normally present in the amniotic cavity. The evidence to support biologic plausibility was generated in our work with Dr. Robert Garfield and Dr. Egle Bytautiene. We were able to demonstrate that guinea pigs allergic to egg proteins (ovalbumin) went into premature labor when challenged with the allergen during pregnancy. Premature labor could be prevented by the administration of an antihistamine in these animals.

We have interpreted this as evidence that some patients with premature labor may have an allergiclike reaction or type I hypersensitivity reaction. The nature of the allergen may vary. However, we know that allergens can cross the placenta. For example, dust mite antigen has been demonstrated in the amniotic fluid of women in the midtrimester of pregnancy. Also, there is evidence that the fetus is able to recognize and mount an immune response to allergens in utero.

In response to your question, we have treated with antihistamines some patients who have a typical allergiclike history along with premature labor, and this has resulted in the disappearance of uterine contractions. However, these are anecdotal reports, and I would like to stress that our interest in this mechanism of disease stems from the importance of demonstrating that a common mechanism of disease, such as allergy, can be a cause of the preterm parturition syndrome. This should not be surprising, because the uterus is bestowed with mast cells, the key effector cells of an allergic response, and the uterus has all the components required to generate an allergiclike immune response.

EAR: What is the evidence and importance of hormonal dysfunction in premature parturition?

RR: Progesterone withdrawal and/or deficiency has not yet been specifically demonstrated as an initiator of spontaneous parturition in humans. However, the role of progesterone in maintaining pregnancy is unquestionable.

Indeed, suspension of progesterone action through the administration of progesterone receptor inhibitors and progesterone receptor antagonists can induce activation of the components of the common pathway of parturition in animals and humans.

The progesterone/estradiol and progesterone/estriol ratios in amniotic fluid are lower in laboring than in nonlaboring subjects, and the progesterone/estriol ratio is lower in preterm labor followed by preterm delivery compared with preterm labor followed by term delivery, suggesting that these hormones are important in determining the duration of pregnancy.

Trials of progesterone administration to prevent preterm delivery have shown interesting results. Specifically, two recent randomized clinical trials demonstrated that vaginal suppositories containing natural progesterone, or injections of a progesterone, 17 α-hydroxyprogesterone caproate, to women at risk for preterm delivery seem to reduce the rate of spontaneous preterm delivery. Moreover, in the trial using 17 α-hydroxyprogesterone caproate, infants of mothers treated with this compound had a lower rate of necrotizing enterocolitis, intraventricular hemorrhage, and the need for supplemental oxygen. Further research is needed to identify the ideal progesterone regimen and the patients who may benefit from this intervention.

EAR: What about stress as a cause of premature labor?

RR: Epidemiologic studies have indicated that women exposed to stressful conditions during pregnancy have a mild increase in the rate of spontaneous preterm labor. The work of Dr. Pathik Wadhwa and Dr. Cal Hobel has been seminal in this area.

The precise mechanisms whereby stress causes premature labor implicate corticotropin-releasing hormone (CRH), which is produced by the hypothalamus and—importantly—by the placenta. Dr. Roger Smith's work in Australia has proposed that CRH is the regulator of a placental clock. Dr. Felice Petraglia in Italy has also contributed significantly to establish a link between CRH and premature labor.

The clinical implications of this work are related to the epidemiologic observations reported by Dr. Emile Papiernik in France, noting that women who are prescribed rest during pregnancy had a lower frequency of preterm delivery. This interesting experience has not been explored in the United States.

However, a targeted intervention to the patient at risk—such as the woman who must stand or do significant physical work during pregnancy—may be beneficial. However, bed rest per se is not an effective treatment to prevent all causes of premature labor. It is easy to understand that if the cause of preterm parturition is infection, then bed rest will not cure it.

EAR: What is the final message that you would like the readers of OB.GYN. NEWS take with them?

RR: Preterm labor is not just labor before its time, but is the result of a pathologic process. The challenge for health care providers is to try to identify why a women is in premature labor, what specific mechanism of disease may be involved, how sick the fetus is, and whether the benefits of pregnancy prolongation outweigh the risk of prematurity. The administration of steroids has been demonstrated to reduce the rate of adverse neonatal outcomes, and it is indicated when the patient is at risk for preterm delivery.

Preterm Labor Has Multiple Causes

In July's Master Class, the specific role of infection in preterm labor was discussed at length by Dr. Roberto Romero, chief of the Perinatology Research Branch at the National Institute of Child Health and Human Development, and professor of obstetrics and gynecology at Wayne State University in Detroit.

Although infection is a leading—and perhaps the best understood—cause of spontaneous preterm labor and delivery, it is not the only cause. Research from Dr. Romero's group and others increasingly point to several disease mechanisms with genetic and environmental components that can be responsible for what we now know as the preterm parturition syndrome.

I am pleased to welcome back Dr. Romero, an international authority on the syndrome.

In this month's discussion, Dr. Romero provides an overview of evolving knowledge about the syndrome and details current scientific progress in the understanding of the noninfectious causes that may be important in the process of preterm labor.

Dr. Roberto Romero: The common pathway of parturition consists of the anatomical, physiologic, biochemical, and clinical events that occur in the mother and/or fetus in both term and preterm labor. The uterine components of this pathway include increased uterine contractility; cervical ripening (dilatation and effacement); and membrane/decidual activation. In most women, these components are typically activated in a synchronous manner in spontaneous labor at term. Indeed, most women admitted in labor have uterine contractions, cervical changes and—sometimes—rupture of membranes.

However, in some cases the activation of the common pathway may be asynchronous. For example, a patient may have increased uterine contractility, but the cervix undergoes very little change over time. This is what is called a prolonged latent phase of labor. In 10% of cases, patients have spontaneous rupture of membranes without myometrial contractility. This is evidence that membrane/decidual activation has occurred without recruitment of the myometrium.

Asynchronous activation is more common in the preterm gestation. Many patients with suspected preterm labor will present with increased uterine contractility without cervical changes. Others will present with the clinical picture of cervical insufficiency (which used to be called cervical incompetence). Finally, some women will present with preterm premature rupture of the membranes (pPROM), which is premature membrane/decidual activation.

EAR: What is the importance of the concept of the common pathway?

RR: Much of the clinical management and research in understanding the causes of premature labor, treatment, and prevention has been focused on the elements of the common terminal pathway. For example, we have used uterine monitors to detect an increase in uterine contractility and tocolytic agents to treat increased myometrial contractility. We use ultrasound to identify patients with a short cervix who are at risk for preterm delivery. In some cases, we have placed cervical cerclage in patients at risk. Finally, we have used fetal fibronectin to detect decidual/membrane deactivation. A positive fetal fibronectin is an indicator that disruption of the choriodecidual interface has occurred. Yet these interventions aim to treat preterm labor as a symptom, without first identifying and understanding the underlying pathology that sets it in motion. Progress on this front is now being made.

EAR: What is the difference between spontaneous labor at term and preterm labor?

RR: We propose that spontaneous labor at term is the inevitable process that occurs when the capacity of the mother to support the fetus in utero has been reached. In other words, when the fetus has achieved maturity and is ready to face extrauterine life, it signals the onset of labor and engages the cooperation of the mother in this process.

In contrast, we propose that premature labor results from a pathologic insult that activates the common pathway of parturition. Before the development of newborn special care units, extreme prematurity was nearly always lethal. Thus, being born preterm is likely to result from such a severe pathologic process that it threatens the survival of the mother and/or baby.

In summary, spontaneous labor at term results from physiologic activation of the common pathway, whereas preterm labor would result from pathologic activation of the pathway.

EAR: What is the evidence that premature labor is a heterogeneous condition?

RR: My laboratory and other groups have generated evidence that the pattern of uterine gene expression—also known as the transcriptional profile—is different in patients with different causes of premature labor. A transcriptional profile is a snapshot of genes that are being upregulated or downregulated at a particular point in time.

We have demonstrated experimentally that the transcriptional profile in the uteri of mice that go into labor as a result of infection is different from the transcriptional profile of mice that go into labor because of ovariectomy (a model of progesterone withdrawal).

We have also examined the transcriptional profile of the chorioamniotic membranes in women with preterm labor and intact membranes vs. women with pPROM, in each case studying women with and without histologic chorioamnionitis. These studies have demonstrated that the patterns of gene expression are very different in these four groups, even though the clinical presentations are similar.

Collectively, the observations in patients and animals suggest that premature labor is a heterogeneous condition. Although they share a common pathway (uterine contractility and cervical dilatation and/or membrane rupture or activation), there are multiple causes for the preterm parturition syndrome.

Using the tools of high-dimensional biology, we have learned that premature labor is not simply labor before its time, but rather a disease state. It is caused by different pathologic processes with both environmental components and genetic components. For example, in infection-induced preterm labor (OB.GYN. NEWS, July 1, 2006, p. 42), the environmental component is represented by the microorganisms that cause the infection. The genetic component is the factor that predisposes some women to have an intrauterine infection or to respond more severely to that infection.

A prime example of the importance of genetics lies with the fetus. In pregnancy, we have not one patient, but two. Accumulating evidence indicates that the fetus plays a central role in the initiation of labor in both animals and humans. In humans, mothers of fetuses who have mounted a severe fetal inflammatory response are more likely to go into labor than are mothers of fetuses who have not mounted a fetal inflammatory response to infection. The magnitude and severity of the inflammatory response are under genetic control.

Thus, in pregnancy, the genetic makeup of two hosts—mother and child—plays a role in determining the susceptibility and response to infection. It is a unique situation in medicine.

EAR: Why do you call preterm parturition a syndrome?

RR: The current taxonomy of disease in obstetrics is largely based on the symptoms and signs exhibited by the mother, not the mechanisms of disease responsible for the clinical presentations. A syndrome is a combination of symptoms and signs that form a distinct clinical picture indicative of a particular disorder. Implicit in this definition is the fact that a syndrome can have multiple causes. Our emphasis in referring to premature parturition as a syndrome is that it will help both patients and physicians readjust the unreasonable expectation that one test can diagnose or identify all women at risk, and that one treatment will be effective for all women in premature labor, regardless of the cause.

EAR: In the last Master Class, we spoke about the role of infection and inflammation as a cause of premature labor. You have shared with us a figure that shows other causes of preterm parturition. I would like to discuss the other causes with you. Can you tell us how uterine ischemia and/or vascular pathology cause premature labor?

RR: Placental histology suggests that vascular lesions involving maternal or fetal circulation constitute the second most common apparent cause of preterm labor and pPROM, after inflammation.

Research in this area has focused on elevated rates of vascular abnormalities in women with spontaneous preterm labor with intact membranes and pPROM, compared with women who deliver at term. The vascular lesions have been found on both the maternal and fetal side. For example, thrombosis of the decidual vessels attached to the placenta and failure of physiologic transformation in the myometrial segment of the spiral arteries are generally considered maternal lesions. Abnormalities in the fetal circulation linked to preterm labor include a decreased number of arterioles in the villi and fetal arterial thrombosis.

Whereas uteroplacental ischemia is the driver of vascular events, the leading candidate to explain the molecular mechanisms responsible is the renin-angiotensin system; in severe uteroplacental ischemia, the enzyme thrombin is emerging as an important activator of preterm labor associated with vaginal bleeding.

The work of Dr. Mark Phillippe and Dr. Michal Elovitz has demonstrated in animal models and in vivo investigations that whole blood—but not heparinized blood—raises contractile activity of the uterine muscle, and that such increased uterine activity can be blocked with a thrombin antagonist.

Moreover, women with preterm labor have higher concentrations of thrombin-antithrombin (TAT) complexes in amniotic fluid and in maternal plasma than do women without preterm labor. Similarly, Dr. Todd Rosen and Dr. Charles Lockwood have provided evidence that women destined to develop pPROM have higher concentrations of thrombin weeks before the development of complications. Therefore, thrombin is a potential initiator not only of the rupture of membranes (via stromal cells and matrix metalloproteinase 1), but also of uterine contractility and preterm labor.

The thrombin connection would help to explain why retroplacental hematomas in early pregnancy are associated with preterm delivery, and why vaginal bleeding in the first or second trimester is a risk factor for preterm birth with intact or ruptured membranes.

EAR: What is the evidence that uterine overdistention is a cause of preterm parturition?

RR: Obstetricians and midwives know that multiple gestation is a risk factor for preterm delivery, and that the higher the order of multiple gestation, the greater the risk for preterm birth. Patients with polyhydramnios resulting from a congenital fetal anomaly are also at risk for spontaneous preterm labor and delivery. These two conditions are probably mediated by uterine overdistention. It is likely that the same is the case for patients who have müllerian duct abnormalities in which the uterine cavity is small. One example is congenital hypoplastic uterus and another is the abnormal uterus resulting from diethylstilbestrol exposure.

EAR: What is the mechanism linking uterine overdistention with premature labor?

RR: This is an area that requires further work. In general terms, mechanical signals triggered by uterine stretch may lead to preterm labor—as in multiple gestations and polyhydramnios—although the precise mechanisms involved remain unknown.

Experimental distention of the uterus with a saline-filled balloon rapidly prompted regular uterine contractions in women carrying live term fetuses or dead fetuses. All were delivered within 21 hours.

Stephen Lye, Ph.D., and his group at the University of Toronto have found increased expression of oxytocin receptor, connexin 43, and the c-fos mRNA in the myometrium. Gillian Bryant-Greenwood, Ph.D., at the University of Hawaii, Honolulu, has shown that stretching of the chorioamniotic membranes in experimental models can increase the expression of interleukin-8 and also another cytokine called visfatin, which can have important effects on the integrity of the membranes.

EAR: I noticed that cervical pathology is also a potential cause of the preterm parturition syndrome. Is this the same as cervical insufficiency?

RR: Yes. Cervical insufficiency is also a cause of the preterm parturition syndrome and can result from congenital disorders, surgical trauma, trauma, or infection. Although cervical length has been touted as a predictor of preterm birth, it is important to remember that a short cervix is not always a ripe cervix. Once again, more research is needed to identify the events that occur as a result of cervical insufficiency, and the ways these affect the initiation of preterm and term labor.

EAR: What about abnormal allograft reaction?

RR: The fetoplacental unit has been called “nature's most successful transplant,” or—more accurately—a “semiallograft.” The mechanisms that bring about tolerance of this semiallograft are poorly understood. However, transplants of solid organs are tolerated through the establishment of microchimerism in the transplanted organ as well as in the host. Therefore, we consider that microchimerism in pregnancy is probably important for tolerance of the fetoplacental unit. However, I anticipate that under pathologic conditions—just as in the case of transplants—tolerance of the fetoplacental unit may break down, which in turn may lead to a unique form of rejection of the fetoplacental unit. Unraveling the mechanisms of this rejectionlike process is a fascinating challenge. However, ob.gyns. know that the frequency of adverse pregnancy outcomes is higher in mothers who have pregnancies after embryo or egg donation. Under these circumstances, the placenta and fetus are totally foreign because they do not have the normal 50% genetic endowment from the mother. The complications noticed in these pregnancies include not only preeclampsia and growth restriction, but also preterm labor.

EAR: What is the evidence that an allergic phenomenon could be associated with premature labor, and could antihistamines be a treatment for premature labor?

RR: This idea came to me after seeing the relative of an obstetrician who went into premature labor after eating shellfish to which she was allergic. Around that time, we observed that a group of women in premature labor had eosinophils in the amniotic cavity. Eosinophils are associated with an allergiclike phenomenon and are not normally present in the amniotic cavity. The evidence to support biologic plausibility was generated in our work with Dr. Robert Garfield and Dr. Egle Bytautiene. We were able to demonstrate that guinea pigs allergic to egg proteins (ovalbumin) went into premature labor when challenged with the allergen during pregnancy. Premature labor could be prevented by the administration of an antihistamine in these animals.

We have interpreted this as evidence that some patients with premature labor may have an allergiclike reaction or type I hypersensitivity reaction. The nature of the allergen may vary. However, we know that allergens can cross the placenta. For example, dust mite antigen has been demonstrated in the amniotic fluid of women in the midtrimester of pregnancy. Also, there is evidence that the fetus is able to recognize and mount an immune response to allergens in utero.

In response to your question, we have treated with antihistamines some patients who have a typical allergiclike history along with premature labor, and this has resulted in the disappearance of uterine contractions. However, these are anecdotal reports, and I would like to stress that our interest in this mechanism of disease stems from the importance of demonstrating that a common mechanism of disease, such as allergy, can be a cause of the preterm parturition syndrome. This should not be surprising, because the uterus is bestowed with mast cells, the key effector cells of an allergic response, and the uterus has all the components required to generate an allergiclike immune response.

EAR: What is the evidence and importance of hormonal dysfunction in premature parturition?

RR: Progesterone withdrawal and/or deficiency has not yet been specifically demonstrated as an initiator of spontaneous parturition in humans. However, the role of progesterone in maintaining pregnancy is unquestionable.

Indeed, suspension of progesterone action through the administration of progesterone receptor inhibitors and progesterone receptor antagonists can induce activation of the components of the common pathway of parturition in animals and humans.

The progesterone/estradiol and progesterone/estriol ratios in amniotic fluid are lower in laboring than in nonlaboring subjects, and the progesterone/estriol ratio is lower in preterm labor followed by preterm delivery compared with preterm labor followed by term delivery, suggesting that these hormones are important in determining the duration of pregnancy.

Trials of progesterone administration to prevent preterm delivery have shown interesting results. Specifically, two recent randomized clinical trials demonstrated that vaginal suppositories containing natural progesterone, or injections of a progesterone, 17 α-hydroxyprogesterone caproate, to women at risk for preterm delivery seem to reduce the rate of spontaneous preterm delivery. Moreover, in the trial using 17 α-hydroxyprogesterone caproate, infants of mothers treated with this compound had a lower rate of necrotizing enterocolitis, intraventricular hemorrhage, and the need for supplemental oxygen. Further research is needed to identify the ideal progesterone regimen and the patients who may benefit from this intervention.

EAR: What about stress as a cause of premature labor?

RR: Epidemiologic studies have indicated that women exposed to stressful conditions during pregnancy have a mild increase in the rate of spontaneous preterm labor. The work of Dr. Pathik Wadhwa and Dr. Cal Hobel has been seminal in this area.

The precise mechanisms whereby stress causes premature labor implicate corticotropin-releasing hormone (CRH), which is produced by the hypothalamus and—importantly—by the placenta. Dr. Roger Smith's work in Australia has proposed that CRH is the regulator of a placental clock. Dr. Felice Petraglia in Italy has also contributed significantly to establish a link between CRH and premature labor.

The clinical implications of this work are related to the epidemiologic observations reported by Dr. Emile Papiernik in France, noting that women who are prescribed rest during pregnancy had a lower frequency of preterm delivery. This interesting experience has not been explored in the United States.

However, a targeted intervention to the patient at risk—such as the woman who must stand or do significant physical work during pregnancy—may be beneficial. However, bed rest per se is not an effective treatment to prevent all causes of premature labor. It is easy to understand that if the cause of preterm parturition is infection, then bed rest will not cure it.

EAR: What is the final message that you would like the readers of OB.GYN. NEWS take with them?

RR: Preterm labor is not just labor before its time, but is the result of a pathologic process. The challenge for health care providers is to try to identify why a women is in premature labor, what specific mechanism of disease may be involved, how sick the fetus is, and whether the benefits of pregnancy prolongation outweigh the risk of prematurity. The administration of steroids has been demonstrated to reduce the rate of adverse neonatal outcomes, and it is indicated when the patient is at risk for preterm delivery.

Preterm Labor Has Multiple Causes

In July's Master Class, the specific role of infection in preterm labor was discussed at length by Dr. Roberto Romero, chief of the Perinatology Research Branch at the National Institute of Child Health and Human Development, and professor of obstetrics and gynecology at Wayne State University in Detroit.

Although infection is a leading—and perhaps the best understood—cause of spontaneous preterm labor and delivery, it is not the only cause. Research from Dr. Romero's group and others increasingly point to several disease mechanisms with genetic and environmental components that can be responsible for what we now know as the preterm parturition syndrome.

I am pleased to welcome back Dr. Romero, an international authority on the syndrome.

In this month's discussion, Dr. Romero provides an overview of evolving knowledge about the syndrome and details current scientific progress in the understanding of the noninfectious causes that may be important in the process of preterm labor.

Dr. Roberto Romero: The common pathway of parturition consists of the anatomical, physiologic, biochemical, and clinical events that occur in the mother and/or fetus in both term and preterm labor. The uterine components of this pathway include increased uterine contractility; cervical ripening (dilatation and effacement); and membrane/decidual activation. In most women, these components are typically activated in a synchronous manner in spontaneous labor at term. Indeed, most women admitted in labor have uterine contractions, cervical changes and—sometimes—rupture of membranes.

However, in some cases the activation of the common pathway may be asynchronous. For example, a patient may have increased uterine contractility, but the cervix undergoes very little change over time. This is what is called a prolonged latent phase of labor. In 10% of cases, patients have spontaneous rupture of membranes without myometrial contractility. This is evidence that membrane/decidual activation has occurred without recruitment of the myometrium.

Asynchronous activation is more common in the preterm gestation. Many patients with suspected preterm labor will present with increased uterine contractility without cervical changes. Others will present with the clinical picture of cervical insufficiency (which used to be called cervical incompetence). Finally, some women will present with preterm premature rupture of the membranes (pPROM), which is premature membrane/decidual activation.

EAR: What is the importance of the concept of the common pathway?

RR: Much of the clinical management and research in understanding the causes of premature labor, treatment, and prevention has been focused on the elements of the common terminal pathway. For example, we have used uterine monitors to detect an increase in uterine contractility and tocolytic agents to treat increased myometrial contractility. We use ultrasound to identify patients with a short cervix who are at risk for preterm delivery. In some cases, we have placed cervical cerclage in patients at risk. Finally, we have used fetal fibronectin to detect decidual/membrane deactivation. A positive fetal fibronectin is an indicator that disruption of the choriodecidual interface has occurred. Yet these interventions aim to treat preterm labor as a symptom, without first identifying and understanding the underlying pathology that sets it in motion. Progress on this front is now being made.

EAR: What is the difference between spontaneous labor at term and preterm labor?

RR: We propose that spontaneous labor at term is the inevitable process that occurs when the capacity of the mother to support the fetus in utero has been reached. In other words, when the fetus has achieved maturity and is ready to face extrauterine life, it signals the onset of labor and engages the cooperation of the mother in this process.

In contrast, we propose that premature labor results from a pathologic insult that activates the common pathway of parturition. Before the development of newborn special care units, extreme prematurity was nearly always lethal. Thus, being born preterm is likely to result from such a severe pathologic process that it threatens the survival of the mother and/or baby.

In summary, spontaneous labor at term results from physiologic activation of the common pathway, whereas preterm labor would result from pathologic activation of the pathway.

EAR: What is the evidence that premature labor is a heterogeneous condition?

RR: My laboratory and other groups have generated evidence that the pattern of uterine gene expression—also known as the transcriptional profile—is different in patients with different causes of premature labor. A transcriptional profile is a snapshot of genes that are being upregulated or downregulated at a particular point in time.

We have demonstrated experimentally that the transcriptional profile in the uteri of mice that go into labor as a result of infection is different from the transcriptional profile of mice that go into labor because of ovariectomy (a model of progesterone withdrawal).

We have also examined the transcriptional profile of the chorioamniotic membranes in women with preterm labor and intact membranes vs. women with pPROM, in each case studying women with and without histologic chorioamnionitis. These studies have demonstrated that the patterns of gene expression are very different in these four groups, even though the clinical presentations are similar.

Collectively, the observations in patients and animals suggest that premature labor is a heterogeneous condition. Although they share a common pathway (uterine contractility and cervical dilatation and/or membrane rupture or activation), there are multiple causes for the preterm parturition syndrome.

Using the tools of high-dimensional biology, we have learned that premature labor is not simply labor before its time, but rather a disease state. It is caused by different pathologic processes with both environmental components and genetic components. For example, in infection-induced preterm labor (OB.GYN. NEWS, July 1, 2006, p. 42), the environmental component is represented by the microorganisms that cause the infection. The genetic component is the factor that predisposes some women to have an intrauterine infection or to respond more severely to that infection.

A prime example of the importance of genetics lies with the fetus. In pregnancy, we have not one patient, but two. Accumulating evidence indicates that the fetus plays a central role in the initiation of labor in both animals and humans. In humans, mothers of fetuses who have mounted a severe fetal inflammatory response are more likely to go into labor than are mothers of fetuses who have not mounted a fetal inflammatory response to infection. The magnitude and severity of the inflammatory response are under genetic control.

Thus, in pregnancy, the genetic makeup of two hosts—mother and child—plays a role in determining the susceptibility and response to infection. It is a unique situation in medicine.

EAR: Why do you call preterm parturition a syndrome?

RR: The current taxonomy of disease in obstetrics is largely based on the symptoms and signs exhibited by the mother, not the mechanisms of disease responsible for the clinical presentations. A syndrome is a combination of symptoms and signs that form a distinct clinical picture indicative of a particular disorder. Implicit in this definition is the fact that a syndrome can have multiple causes. Our emphasis in referring to premature parturition as a syndrome is that it will help both patients and physicians readjust the unreasonable expectation that one test can diagnose or identify all women at risk, and that one treatment will be effective for all women in premature labor, regardless of the cause.

EAR: In the last Master Class, we spoke about the role of infection and inflammation as a cause of premature labor. You have shared with us a figure that shows other causes of preterm parturition. I would like to discuss the other causes with you. Can you tell us how uterine ischemia and/or vascular pathology cause premature labor?

RR: Placental histology suggests that vascular lesions involving maternal or fetal circulation constitute the second most common apparent cause of preterm labor and pPROM, after inflammation.

Research in this area has focused on elevated rates of vascular abnormalities in women with spontaneous preterm labor with intact membranes and pPROM, compared with women who deliver at term. The vascular lesions have been found on both the maternal and fetal side. For example, thrombosis of the decidual vessels attached to the placenta and failure of physiologic transformation in the myometrial segment of the spiral arteries are generally considered maternal lesions. Abnormalities in the fetal circulation linked to preterm labor include a decreased number of arterioles in the villi and fetal arterial thrombosis.

Whereas uteroplacental ischemia is the driver of vascular events, the leading candidate to explain the molecular mechanisms responsible is the renin-angiotensin system; in severe uteroplacental ischemia, the enzyme thrombin is emerging as an important activator of preterm labor associated with vaginal bleeding.

The work of Dr. Mark Phillippe and Dr. Michal Elovitz has demonstrated in animal models and in vivo investigations that whole blood—but not heparinized blood—raises contractile activity of the uterine muscle, and that such increased uterine activity can be blocked with a thrombin antagonist.

Moreover, women with preterm labor have higher concentrations of thrombin-antithrombin (TAT) complexes in amniotic fluid and in maternal plasma than do women without preterm labor. Similarly, Dr. Todd Rosen and Dr. Charles Lockwood have provided evidence that women destined to develop pPROM have higher concentrations of thrombin weeks before the development of complications. Therefore, thrombin is a potential initiator not only of the rupture of membranes (via stromal cells and matrix metalloproteinase 1), but also of uterine contractility and preterm labor.

The thrombin connection would help to explain why retroplacental hematomas in early pregnancy are associated with preterm delivery, and why vaginal bleeding in the first or second trimester is a risk factor for preterm birth with intact or ruptured membranes.

EAR: What is the evidence that uterine overdistention is a cause of preterm parturition?

RR: Obstetricians and midwives know that multiple gestation is a risk factor for preterm delivery, and that the higher the order of multiple gestation, the greater the risk for preterm birth. Patients with polyhydramnios resulting from a congenital fetal anomaly are also at risk for spontaneous preterm labor and delivery. These two conditions are probably mediated by uterine overdistention. It is likely that the same is the case for patients who have müllerian duct abnormalities in which the uterine cavity is small. One example is congenital hypoplastic uterus and another is the abnormal uterus resulting from diethylstilbestrol exposure.

EAR: What is the mechanism linking uterine overdistention with premature labor?

RR: This is an area that requires further work. In general terms, mechanical signals triggered by uterine stretch may lead to preterm labor—as in multiple gestations and polyhydramnios—although the precise mechanisms involved remain unknown.

Experimental distention of the uterus with a saline-filled balloon rapidly prompted regular uterine contractions in women carrying live term fetuses or dead fetuses. All were delivered within 21 hours.

Stephen Lye, Ph.D., and his group at the University of Toronto have found increased expression of oxytocin receptor, connexin 43, and the c-fos mRNA in the myometrium. Gillian Bryant-Greenwood, Ph.D., at the University of Hawaii, Honolulu, has shown that stretching of the chorioamniotic membranes in experimental models can increase the expression of interleukin-8 and also another cytokine called visfatin, which can have important effects on the integrity of the membranes.

EAR: I noticed that cervical pathology is also a potential cause of the preterm parturition syndrome. Is this the same as cervical insufficiency?

RR: Yes. Cervical insufficiency is also a cause of the preterm parturition syndrome and can result from congenital disorders, surgical trauma, trauma, or infection. Although cervical length has been touted as a predictor of preterm birth, it is important to remember that a short cervix is not always a ripe cervix. Once again, more research is needed to identify the events that occur as a result of cervical insufficiency, and the ways these affect the initiation of preterm and term labor.

EAR: What about abnormal allograft reaction?

RR: The fetoplacental unit has been called “nature's most successful transplant,” or—more accurately—a “semiallograft.” The mechanisms that bring about tolerance of this semiallograft are poorly understood. However, transplants of solid organs are tolerated through the establishment of microchimerism in the transplanted organ as well as in the host. Therefore, we consider that microchimerism in pregnancy is probably important for tolerance of the fetoplacental unit. However, I anticipate that under pathologic conditions—just as in the case of transplants—tolerance of the fetoplacental unit may break down, which in turn may lead to a unique form of rejection of the fetoplacental unit. Unraveling the mechanisms of this rejectionlike process is a fascinating challenge. However, ob.gyns. know that the frequency of adverse pregnancy outcomes is higher in mothers who have pregnancies after embryo or egg donation. Under these circumstances, the placenta and fetus are totally foreign because they do not have the normal 50% genetic endowment from the mother. The complications noticed in these pregnancies include not only preeclampsia and growth restriction, but also preterm labor.

EAR: What is the evidence that an allergic phenomenon could be associated with premature labor, and could antihistamines be a treatment for premature labor?

RR: This idea came to me after seeing the relative of an obstetrician who went into premature labor after eating shellfish to which she was allergic. Around that time, we observed that a group of women in premature labor had eosinophils in the amniotic cavity. Eosinophils are associated with an allergiclike phenomenon and are not normally present in the amniotic cavity. The evidence to support biologic plausibility was generated in our work with Dr. Robert Garfield and Dr. Egle Bytautiene. We were able to demonstrate that guinea pigs allergic to egg proteins (ovalbumin) went into premature labor when challenged with the allergen during pregnancy. Premature labor could be prevented by the administration of an antihistamine in these animals.

We have interpreted this as evidence that some patients with premature labor may have an allergiclike reaction or type I hypersensitivity reaction. The nature of the allergen may vary. However, we know that allergens can cross the placenta. For example, dust mite antigen has been demonstrated in the amniotic fluid of women in the midtrimester of pregnancy. Also, there is evidence that the fetus is able to recognize and mount an immune response to allergens in utero.

In response to your question, we have treated with antihistamines some patients who have a typical allergiclike history along with premature labor, and this has resulted in the disappearance of uterine contractions. However, these are anecdotal reports, and I would like to stress that our interest in this mechanism of disease stems from the importance of demonstrating that a common mechanism of disease, such as allergy, can be a cause of the preterm parturition syndrome. This should not be surprising, because the uterus is bestowed with mast cells, the key effector cells of an allergic response, and the uterus has all the components required to generate an allergiclike immune response.

EAR: What is the evidence and importance of hormonal dysfunction in premature parturition?

RR: Progesterone withdrawal and/or deficiency has not yet been specifically demonstrated as an initiator of spontaneous parturition in humans. However, the role of progesterone in maintaining pregnancy is unquestionable.

Indeed, suspension of progesterone action through the administration of progesterone receptor inhibitors and progesterone receptor antagonists can induce activation of the components of the common pathway of parturition in animals and humans.

The progesterone/estradiol and progesterone/estriol ratios in amniotic fluid are lower in laboring than in nonlaboring subjects, and the progesterone/estriol ratio is lower in preterm labor followed by preterm delivery compared with preterm labor followed by term delivery, suggesting that these hormones are important in determining the duration of pregnancy.

Trials of progesterone administration to prevent preterm delivery have shown interesting results. Specifically, two recent randomized clinical trials demonstrated that vaginal suppositories containing natural progesterone, or injections of a progesterone, 17 α-hydroxyprogesterone caproate, to women at risk for preterm delivery seem to reduce the rate of spontaneous preterm delivery. Moreover, in the trial using 17 α-hydroxyprogesterone caproate, infants of mothers treated with this compound had a lower rate of necrotizing enterocolitis, intraventricular hemorrhage, and the need for supplemental oxygen. Further research is needed to identify the ideal progesterone regimen and the patients who may benefit from this intervention.

EAR: What about stress as a cause of premature labor?

RR: Epidemiologic studies have indicated that women exposed to stressful conditions during pregnancy have a mild increase in the rate of spontaneous preterm labor. The work of Dr. Pathik Wadhwa and Dr. Cal Hobel has been seminal in this area.

The precise mechanisms whereby stress causes premature labor implicate corticotropin-releasing hormone (CRH), which is produced by the hypothalamus and—importantly—by the placenta. Dr. Roger Smith's work in Australia has proposed that CRH is the regulator of a placental clock. Dr. Felice Petraglia in Italy has also contributed significantly to establish a link between CRH and premature labor.

The clinical implications of this work are related to the epidemiologic observations reported by Dr. Emile Papiernik in France, noting that women who are prescribed rest during pregnancy had a lower frequency of preterm delivery. This interesting experience has not been explored in the United States.

However, a targeted intervention to the patient at risk—such as the woman who must stand or do significant physical work during pregnancy—may be beneficial. However, bed rest per se is not an effective treatment to prevent all causes of premature labor. It is easy to understand that if the cause of preterm parturition is infection, then bed rest will not cure it.

EAR: What is the final message that you would like the readers of OB.GYN. NEWS take with them?

RR: Preterm labor is not just labor before its time, but is the result of a pathologic process. The challenge for health care providers is to try to identify why a women is in premature labor, what specific mechanism of disease may be involved, how sick the fetus is, and whether the benefits of pregnancy prolongation outweigh the risk of prematurity. The administration of steroids has been demonstrated to reduce the rate of adverse neonatal outcomes, and it is indicated when the patient is at risk for preterm delivery.

Preterm Labor Has Multiple Causes

In July's Master Class, the specific role of infection in preterm labor was discussed at length by Dr. Roberto Romero, chief of the Perinatology Research Branch at the National Institute of Child Health and Human Development, and professor of obstetrics and gynecology at Wayne State University in Detroit.

Although infection is a leading—and perhaps the best understood—cause of spontaneous preterm labor and delivery, it is not the only cause. Research from Dr. Romero's group and others increasingly point to several disease mechanisms with genetic and environmental components that can be responsible for what we now know as the preterm parturition syndrome.

I am pleased to welcome back Dr. Romero, an international authority on the syndrome.

In this month's discussion, Dr. Romero provides an overview of evolving knowledge about the syndrome and details current scientific progress in the understanding of the noninfectious causes that may be important in the process of preterm labor.

Dr. E. Albert Reece: How important is infection as a mechanism of disease in premature labor?

Dr. Roberto Romero: Intrauterine and systemic infections are a leading cause of spontaneous preterm labor and delivery. Indeed, infections and inflammation are the only mechanisms of disease for which a clear causal link with prematurity has been established. Moreover, a clear molecular pathophysiology has been described.

EAR: How frequent is intrauterine infection in spontaneous preterm birth?

RR: It has been estimated that one of every four preterm births occurs to mothers with intraamniotic infection (defined as a positive amniotic fluid culture for microorganisms). Under normal circumstances, the amniotic cavity does not contain bacteria, just as cerebrospinal fluid does not. However, approximately 12% of women presenting with an episode of preterm labor will have a positive amniotic fluid culture for microorganisms. The organisms most frequently isolated are genital mycoplasmas, particularly Ureaplasma urealyticum. Among women with preterm premature rupture of membranes (PROM), one of every three will have a positive amniotic fluid culture for microorganisms at the time of presentation. U. urealyticum is the most common microorganism isolated from the amniotic cavity.

EAR: Is there a particular subgroup of women in whom intrauterine infection is more prevalent?

RR: The earlier the gestational age at which a patient presents with preterm labor and intact membranes or preterm PROM, the higher the likelihood of a positive amniotic fluid culture. For example, infection and/or inflammation is present in close to 70% of women presenting around 24 weeks of gestation, but is much rarer in patients presenting after 34 weeks.

EAR: Among women who present with a clinical picture of acute cervical insufficiency, also known as an “incompetent cervix,” how common are infections?

RR: Studies by our group and others indicate that approximately 50% of women presenting with a dilated cervix and bulging membranes before 24 weeks of gestation will have a positive amniotic fluid culture for microorganisms. It is important to realize that rupture of membranes after a cervical cerclage may be the result of a subclinical infectious process, rather than the consequence of cerclage placement.

EAR: What proportion of intrauterine infections manifest themselves with clinical chorioamnionitis?

RR: Most intrauterine infections are subclinical in nature. Our work indicates that among women with intraamniotic infection and preterm labor with intact membranes, only 12% will have a positive amniotic fluid culture. Among women with preterm PROM, only 20% will have clinical signs of chorioamnionitis when a positive amniotic fluid culture is present.

EAR: If most infections are subclinical in nature, how can they be detected?

RR: The most accurate method to detect the presence of intraamniotic infection is analysis of amniotic fluid. Amniotic fluid is normally sterile for bacteria. It is possible to isolate some viruses from amniotic fluid, but generally, cultures for viruses are not performed in patients with preterm labor and preterm PROM. Amniotic fluid should be cultured for aerobic and anaerobic bacteria, as well as for Mycoplasma species. Detection of mycoplasmas is important, because they are the most common organisms found in the amniotic fluid. Commercially available systems exist that can be implemented in U.S. laboratories.

EAR: The results of cultures take several days to become available. Thus, rapid tests are required to assess the likelihood of infection or inflammation. What tests would you recommend for this purpose?

RR: A positive Gram stain has 99% specificity, but 20% sensitivity in the detection of intraamniotic infection. The low sensitivity is because the Gram stain cannot detect mycoplasmas since these organisms are too small to be seen with light microscopy. However, the take-home message is that a positive Gram stain is virtually always associated with a positive culture and that false positives are rare. The current approach to the detection of infection and/or inflammation in the amniotic fluid includes other tests that are routinely performed for the analysis of cerebrospinal fluid in all hospitals in the United States. Such tests include a white blood cell (WBC) count of amniotic fluid, and a glucose determination.

EAR: How can the clinician interpret the results of an amniotic fluid WBC and an amniotic fluid glucose determination?

RR: White blood cells—such as neutrophils, monocytes, or eosinophils—are not normally present in the amniotic fluid. Therefore, a high count of white blood cells is an indicator that intraamniotic inflammation is present. We recommend that an amniotic fluid sample be sent to the clinical hematology laboratory, that a WBC count be performed in the standard hemacytometer chamber, and that this be followed by a differential count. When the WBC count is greater than 50 cells/L in patients with intact membranes—or greater than 30 in patients with preterm PROM—the likelihood of a positive amniotic fluid culture is high.

In terms of the amniotic fluid glucose concentration, under normal circumstances glucose is present in the amniotic fluid. The lower the amniotic fluid glucose value, the higher the likelihood of intraamniotic infection or inflammation. For example, glucose values less than 14 mg/dL in women with intact membranes—or less than 10 mg/dL in women with preterm PROM—suggest that intraamniotic infection and/or inflammation is present.

EAR: Are there other tests, such as measurements of cytokines or other proteins, that can be used to detect inflammation in the amniotic fluid?

RR: The concentrations of a cytokine, such as interleukin-6, can be used to detect inflammation. Similarly, we have developed a rapid test that can be used at the bedside to detect inflammation by detecting the concentration of an enzyme produced by neutrophils. This enzyme is MMP-8 (matrix metalloproteinase-8).

These tests can be valuable in the context of midtrimester amniocentesis. The rate of pregnancy loss after a midtrimester amniocentesis has been estimated to be 0.5%–1%; such losses have mistakenly been thought to be always procedure related. However, we have found that among women who have midtrimester amniocenteses, those who have an elevated IL-6 or MMP-8 concentration are more likely to lose their pregnancy or have a spontaneous abortion shortly after the procedure. In these circumstances, determination of IL-6 or MMP-8 in the amniotic fluid stored by the genetic laboratories may be helpful for the patient and physician to identify that intraamniotic inflammation was a cause of the pregnancy loss. This may also have medicolegal implications.

EAR: How common is intraamniotic infection and/or inflammation in women having genetic amniocenteses in the midtrimester of pregnancy?

RR: The frequency of intraamniotic infection has been estimated to be 0.9%; the frequency of intraamniotic inflammation is about 1.2%. The most common organism found in the amniotic fluid is U. urealyticum. Intraamniotic infection and/or inflammation is more common in women who have discolored amniotic fluid at the time of genetic amniocentesis. It is important to realize that these infections are subclinical and that sometimes, patients with these infections rupture their membranes within hours or days of the procedure.

EAR: Can treatment be offered to these patients with midtrimester intraamniotic infections?

RR: Recent evidence from Dr. Sonia Hassan in our group indicates that the administration of antibiotics to the mother can eradicate intraamniotic infection in the midtrimester. Women with a short cervix detected by ultrasound were found to have microorganisms in 9% of cases. Patients were offered treatment with antibiotics and a repeat amniocentesis was performed to be sure that the infections were eradicated. Most women treated in this fashion had eradication of their intraamniotic infection and their pregnancy went to term.

EAR: How frequently are intrauterine infections confined to the amniotic fluid, and how often is the fetus involved?

RR: A study conducted in the United Kingdom in women with preterm PROM indicated that approximately 30% of patients had microorganisms in the amniotic fluid. Of these, 30% had positive fetal blood cultures. This means that 10% of all fetuses with preterm PROM will have fetal bacteremia. Clearly, this represents a minimum estimate of the frequency of fetal infection, a result of the limitations of standard techniques and the difficulties in isolating relevant microorganisms from fetal blood.

EAR: What is the importance of congenital neonatal infections?

RR: Sepsis is a more serious disease in neonates than in adults. Neonates have been generally considered immunosuppressed hosts, and the lethality of sepsis in neonates is high. There is now accumulating evidence that neonates with sepsis are more likely to develop cerebral palsy and bronchopulmonary dysplasia or chronic lung disease.

EAR: How important is intrauterine infection as a cause for cerebral palsy?

RR: It has been estimated that as many as 20% of all cases of cerebral palsy result from infection. Moreover, this applies to term neonates as well as to preterm neonates. Therefore, the traditional paradigm—that intrapartum asphyxia was the leading cause of cerebral palsy—is probably not correct. Obstetricians need to be aware that undiagnosed infections can be a cause for cerebral palsy because this has medicolegal implications.

EAR: What is the link between infection and the brain injury associated with cerebral palsy?

RR: Microorganisms involved in cases of intraamniotic infection can invade the human fetus. When the fetus breathes or swallows infected amniotic fluid, microorganisms may be entering the fetal compartment. Once microorganisms invade the fetus, they elicit a fetal inflammatory response syndrome (FIRS), which is the counterpart of the systemic inflammatory response syndrome (SIRS) in the adult. In FIRS one of the most critical organs affected is the brain. Microorganisms or their products that gain access to the fetal brain can induce damage of neurons or white matter in utero. Damage to the white matter in utero is also known as periventricular leukomalacia (PVL) and is the most important predictor of cerebral palsy. There is evidence that cytokines, chemokines, and other inflammatory products—such as reactive oxygen metabolites—can cause damage to the glia or to neurons, which is responsible for the cognitive abnormalities, including mental retardation.

EAR: If most intrauterine infections are subclinical, how can an obstetrician determine whether or not a neonate had intrauterine infection after birth?

RR: One possibility is to look at the placenta. Inflammation in the placenta can be of maternal origin or of fetal origin. Histologic chorioamnionitis is inflammation of the chorioamniotic membranes caused by maternal cells and is, therefore, a maternal inflammatory response. By contrast, funisitis—inflammation of the umbilical cord—is a fetal inflammatory response. Therefore, the presence of funisitis, diagnosed by examination of the placenta, indicates that the fetus was exposed to microorganisms before birth, or that the fetus mounted a FIRS. This is the reason why we call funisitis the hallmark of FIRS. The practical implication of this is that the examination of the placenta may be helpful in understanding what happened before birth. This is particularly important, given that funisitis has been associated with the subsequent development of cerebral palsy. The medicolegal implications of this are apparent. Because there is no known treatment for funisitis, there is no evidence that any intervention by obstetricians can prevent cerebral palsy associated with or induced by intrauterine infection.

EAR: Do systemic infections cause premature labor?

RR: Systemic clinical infections—such as pyelonephritis, pneumonia, malaria, and appendicitis—have been associated with premature labor and delivery. However, there is recent evidence that subclinical distant infections may also be a cause of premature labor and delivery. Specifically, periodontal disease, which is a chronic inflammatory process, has been associated with the subsequent development of preterm labor as well as with small-for-gestational-age infants.

The Changing Approach to Preterm Labor

Preterm delivery accounts for a significant component of infant mortality in the world. In this, the United States has not been spared; in fact, our country ranks a dismal 21st internationally in infant mortality, with prematurity as a major contributor.

Historically, obstetrics has approached this problem from a therapeutic perspective: If you see a contraction, try to stop it. In large measure, we have been unsuccessful, staving off delivery by a mean of approximately 2 days despite our best efforts. Although this window can allow for the stabilization of patients, arranging for their transfer to appropriate delivery sites, and initiating required medications, it does not solve the problem of prematurity.

By focusing on the symptoms of premature labor, we have not paused sufficiently to ask basic questions about potential causes and triggers that could help us to develop preventive strategies and targeted treatments for what is clearly a multifactorial syndrome.

This is all changing. The National Institutes of Health saw this as such a priority that it formed a Perinatology Research Branch within the National Institute of Child Health and Development and chose international authority Dr. Roberto Romero to lead it. The Centers for Disease Control and Prevention and the March of Dimes have similarly launched research initiatives and made the prevention of prematurity a priority.

Much progress has been made, with Dr. Romero's team taking a role in the forefront. I am very pleased to welcome Dr. Romero, professor of ob.gyn. at Wayne State University in Detroit and chief of the Perinatology Research Branch at NIH, as this month's guest professor. We will review advances in our understanding of the biology of prematurity as a syndrome and offer potential treatment implications, beginning with a focus on infection.

In September, we will similarly review the other contributors to the prematurity syndrome.

Dr. E. Albert Reece: How important is infection as a mechanism of disease in premature labor?

Dr. Roberto Romero: Intrauterine and systemic infections are a leading cause of spontaneous preterm labor and delivery. Indeed, infections and inflammation are the only mechanisms of disease for which a clear causal link with prematurity has been established. Moreover, a clear molecular pathophysiology has been described.

EAR: How frequent is intrauterine infection in spontaneous preterm birth?

RR: It has been estimated that one of every four preterm births occurs to mothers with intraamniotic infection (defined as a positive amniotic fluid culture for microorganisms). Under normal circumstances, the amniotic cavity does not contain bacteria, just as cerebrospinal fluid does not. However, approximately 12% of women presenting with an episode of preterm labor will have a positive amniotic fluid culture for microorganisms. The organisms most frequently isolated are genital mycoplasmas, particularly Ureaplasma urealyticum. Among women with preterm premature rupture of membranes (PROM), one of every three will have a positive amniotic fluid culture for microorganisms at the time of presentation. U. urealyticum is the most common microorganism isolated from the amniotic cavity.

EAR: Is there a particular subgroup of women in whom intrauterine infection is more prevalent?

RR: The earlier the gestational age at which a patient presents with preterm labor and intact membranes or preterm PROM, the higher the likelihood of a positive amniotic fluid culture. For example, infection and/or inflammation is present in close to 70% of women presenting around 24 weeks of gestation, but is much rarer in patients presenting after 34 weeks.

EAR: Among women who present with a clinical picture of acute cervical insufficiency, also known as an “incompetent cervix,” how common are infections?

RR: Studies by our group and others indicate that approximately 50% of women presenting with a dilated cervix and bulging membranes before 24 weeks of gestation will have a positive amniotic fluid culture for microorganisms. It is important to realize that rupture of membranes after a cervical cerclage may be the result of a subclinical infectious process, rather than the consequence of cerclage placement.

EAR: What proportion of intrauterine infections manifest themselves with clinical chorioamnionitis?

RR: Most intrauterine infections are subclinical in nature. Our work indicates that among women with intraamniotic infection and preterm labor with intact membranes, only 12% will have a positive amniotic fluid culture. Among women with preterm PROM, only 20% will have clinical signs of chorioamnionitis when a positive amniotic fluid culture is present.

EAR: If most infections are subclinical in nature, how can they be detected?

RR: The most accurate method to detect the presence of intraamniotic infection is analysis of amniotic fluid. Amniotic fluid is normally sterile for bacteria. It is possible to isolate some viruses from amniotic fluid, but generally, cultures for viruses are not performed in patients with preterm labor and preterm PROM. Amniotic fluid should be cultured for aerobic and anaerobic bacteria, as well as for Mycoplasma species. Detection of mycoplasmas is important, because they are the most common organisms found in the amniotic fluid. Commercially available systems exist that can be implemented in U.S. laboratories.

EAR: The results of cultures take several days to become available. Thus, rapid tests are required to assess the likelihood of infection or inflammation. What tests would you recommend for this purpose?

RR: A positive Gram stain has 99% specificity, but 20% sensitivity in the detection of intraamniotic infection. The low sensitivity is because the Gram stain cannot detect mycoplasmas since these organisms are too small to be seen with light microscopy. However, the take-home message is that a positive Gram stain is virtually always associated with a positive culture and that false positives are rare. The current approach to the detection of infection and/or inflammation in the amniotic fluid includes other tests that are routinely performed for the analysis of cerebrospinal fluid in all hospitals in the United States. Such tests include a white blood cell (WBC) count of amniotic fluid, and a glucose determination.

EAR: How can the clinician interpret the results of an amniotic fluid WBC and an amniotic fluid glucose determination?

RR: White blood cells—such as neutrophils, monocytes, or eosinophils—are not normally present in the amniotic fluid. Therefore, a high count of white blood cells is an indicator that intraamniotic inflammation is present. We recommend that an amniotic fluid sample be sent to the clinical hematology laboratory, that a WBC count be performed in the standard hemacytometer chamber, and that this be followed by a differential count. When the WBC count is greater than 50 cells/L in patients with intact membranes—or greater than 30 in patients with preterm PROM—the likelihood of a positive amniotic fluid culture is high.

In terms of the amniotic fluid glucose concentration, under normal circumstances glucose is present in the amniotic fluid. The lower the amniotic fluid glucose value, the higher the likelihood of intraamniotic infection or inflammation. For example, glucose values less than 14 mg/dL in women with intact membranes—or less than 10 mg/dL in women with preterm PROM—suggest that intraamniotic infection and/or inflammation is present.

EAR: Are there other tests, such as measurements of cytokines or other proteins, that can be used to detect inflammation in the amniotic fluid?

RR: The concentrations of a cytokine, such as interleukin-6, can be used to detect inflammation. Similarly, we have developed a rapid test that can be used at the bedside to detect inflammation by detecting the concentration of an enzyme produced by neutrophils. This enzyme is MMP-8 (matrix metalloproteinase-8).

These tests can be valuable in the context of midtrimester amniocentesis. The rate of pregnancy loss after a midtrimester amniocentesis has been estimated to be 0.5%–1%; such losses have mistakenly been thought to be always procedure related. However, we have found that among women who have midtrimester amniocenteses, those who have an elevated IL-6 or MMP-8 concentration are more likely to lose their pregnancy or have a spontaneous abortion shortly after the procedure. In these circumstances, determination of IL-6 or MMP-8 in the amniotic fluid stored by the genetic laboratories may be helpful for the patient and physician to identify that intraamniotic inflammation was a cause of the pregnancy loss. This may also have medicolegal implications.

EAR: How common is intraamniotic infection and/or inflammation in women having genetic amniocenteses in the midtrimester of pregnancy?

RR: The frequency of intraamniotic infection has been estimated to be 0.9%; the frequency of intraamniotic inflammation is about 1.2%. The most common organism found in the amniotic fluid is U. urealyticum. Intraamniotic infection and/or inflammation is more common in women who have discolored amniotic fluid at the time of genetic amniocentesis. It is important to realize that these infections are subclinical and that sometimes, patients with these infections rupture their membranes within hours or days of the procedure.

EAR: Can treatment be offered to these patients with midtrimester intraamniotic infections?

RR: Recent evidence from Dr. Sonia Hassan in our group indicates that the administration of antibiotics to the mother can eradicate intraamniotic infection in the midtrimester. Women with a short cervix detected by ultrasound were found to have microorganisms in 9% of cases. Patients were offered treatment with antibiotics and a repeat amniocentesis was performed to be sure that the infections were eradicated. Most women treated in this fashion had eradication of their intraamniotic infection and their pregnancy went to term.

EAR: How frequently are intrauterine infections confined to the amniotic fluid, and how often is the fetus involved?

RR: A study conducted in the United Kingdom in women with preterm PROM indicated that approximately 30% of patients had microorganisms in the amniotic fluid. Of these, 30% had positive fetal blood cultures. This means that 10% of all fetuses with preterm PROM will have fetal bacteremia. Clearly, this represents a minimum estimate of the frequency of fetal infection, a result of the limitations of standard techniques and the difficulties in isolating relevant microorganisms from fetal blood.

EAR: What is the importance of congenital neonatal infections?

RR: Sepsis is a more serious disease in neonates than in adults. Neonates have been generally considered immunosuppressed hosts, and the lethality of sepsis in neonates is high. There is now accumulating evidence that neonates with sepsis are more likely to develop cerebral palsy and bronchopulmonary dysplasia or chronic lung disease.

EAR: How important is intrauterine infection as a cause for cerebral palsy?

RR: It has been estimated that as many as 20% of all cases of cerebral palsy result from infection. Moreover, this applies to term neonates as well as to preterm neonates. Therefore, the traditional paradigm—that intrapartum asphyxia was the leading cause of cerebral palsy—is probably not correct. Obstetricians need to be aware that undiagnosed infections can be a cause for cerebral palsy because this has medicolegal implications.

EAR: What is the link between infection and the brain injury associated with cerebral palsy?

RR: Microorganisms involved in cases of intraamniotic infection can invade the human fetus. When the fetus breathes or swallows infected amniotic fluid, microorganisms may be entering the fetal compartment. Once microorganisms invade the fetus, they elicit a fetal inflammatory response syndrome (FIRS), which is the counterpart of the systemic inflammatory response syndrome (SIRS) in the adult. In FIRS one of the most critical organs affected is the brain. Microorganisms or their products that gain access to the fetal brain can induce damage of neurons or white matter in utero. Damage to the white matter in utero is also known as periventricular leukomalacia (PVL) and is the most important predictor of cerebral palsy. There is evidence that cytokines, chemokines, and other inflammatory products—such as reactive oxygen metabolites—can cause damage to the glia or to neurons, which is responsible for the cognitive abnormalities, including mental retardation.

EAR: If most intrauterine infections are subclinical, how can an obstetrician determine whether or not a neonate had intrauterine infection after birth?

RR: One possibility is to look at the placenta. Inflammation in the placenta can be of maternal origin or of fetal origin. Histologic chorioamnionitis is inflammation of the chorioamniotic membranes caused by maternal cells and is, therefore, a maternal inflammatory response. By contrast, funisitis—inflammation of the umbilical cord—is a fetal inflammatory response. Therefore, the presence of funisitis, diagnosed by examination of the placenta, indicates that the fetus was exposed to microorganisms before birth, or that the fetus mounted a FIRS. This is the reason why we call funisitis the hallmark of FIRS. The practical implication of this is that the examination of the placenta may be helpful in understanding what happened before birth. This is particularly important, given that funisitis has been associated with the subsequent development of cerebral palsy. The medicolegal implications of this are apparent. Because there is no known treatment for funisitis, there is no evidence that any intervention by obstetricians can prevent cerebral palsy associated with or induced by intrauterine infection.

EAR: Do systemic infections cause premature labor?

RR: Systemic clinical infections—such as pyelonephritis, pneumonia, malaria, and appendicitis—have been associated with premature labor and delivery. However, there is recent evidence that subclinical distant infections may also be a cause of premature labor and delivery. Specifically, periodontal disease, which is a chronic inflammatory process, has been associated with the subsequent development of preterm labor as well as with small-for-gestational-age infants.

The Changing Approach to Preterm Labor

Preterm delivery accounts for a significant component of infant mortality in the world. In this, the United States has not been spared; in fact, our country ranks a dismal 21st internationally in infant mortality, with prematurity as a major contributor.

Historically, obstetrics has approached this problem from a therapeutic perspective: If you see a contraction, try to stop it. In large measure, we have been unsuccessful, staving off delivery by a mean of approximately 2 days despite our best efforts. Although this window can allow for the stabilization of patients, arranging for their transfer to appropriate delivery sites, and initiating required medications, it does not solve the problem of prematurity.

By focusing on the symptoms of premature labor, we have not paused sufficiently to ask basic questions about potential causes and triggers that could help us to develop preventive strategies and targeted treatments for what is clearly a multifactorial syndrome.

This is all changing. The National Institutes of Health saw this as such a priority that it formed a Perinatology Research Branch within the National Institute of Child Health and Development and chose international authority Dr. Roberto Romero to lead it. The Centers for Disease Control and Prevention and the March of Dimes have similarly launched research initiatives and made the prevention of prematurity a priority.

Much progress has been made, with Dr. Romero's team taking a role in the forefront. I am very pleased to welcome Dr. Romero, professor of ob.gyn. at Wayne State University in Detroit and chief of the Perinatology Research Branch at NIH, as this month's guest professor. We will review advances in our understanding of the biology of prematurity as a syndrome and offer potential treatment implications, beginning with a focus on infection.

In September, we will similarly review the other contributors to the prematurity syndrome.

Dr. E. Albert Reece: How important is infection as a mechanism of disease in premature labor?

Dr. Roberto Romero: Intrauterine and systemic infections are a leading cause of spontaneous preterm labor and delivery. Indeed, infections and inflammation are the only mechanisms of disease for which a clear causal link with prematurity has been established. Moreover, a clear molecular pathophysiology has been described.

EAR: How frequent is intrauterine infection in spontaneous preterm birth?

RR: It has been estimated that one of every four preterm births occurs to mothers with intraamniotic infection (defined as a positive amniotic fluid culture for microorganisms). Under normal circumstances, the amniotic cavity does not contain bacteria, just as cerebrospinal fluid does not. However, approximately 12% of women presenting with an episode of preterm labor will have a positive amniotic fluid culture for microorganisms. The organisms most frequently isolated are genital mycoplasmas, particularly Ureaplasma urealyticum. Among women with preterm premature rupture of membranes (PROM), one of every three will have a positive amniotic fluid culture for microorganisms at the time of presentation. U. urealyticum is the most common microorganism isolated from the amniotic cavity.

EAR: Is there a particular subgroup of women in whom intrauterine infection is more prevalent?

RR: The earlier the gestational age at which a patient presents with preterm labor and intact membranes or preterm PROM, the higher the likelihood of a positive amniotic fluid culture. For example, infection and/or inflammation is present in close to 70% of women presenting around 24 weeks of gestation, but is much rarer in patients presenting after 34 weeks.

EAR: Among women who present with a clinical picture of acute cervical insufficiency, also known as an “incompetent cervix,” how common are infections?

RR: Studies by our group and others indicate that approximately 50% of women presenting with a dilated cervix and bulging membranes before 24 weeks of gestation will have a positive amniotic fluid culture for microorganisms. It is important to realize that rupture of membranes after a cervical cerclage may be the result of a subclinical infectious process, rather than the consequence of cerclage placement.

EAR: What proportion of intrauterine infections manifest themselves with clinical chorioamnionitis?

RR: Most intrauterine infections are subclinical in nature. Our work indicates that among women with intraamniotic infection and preterm labor with intact membranes, only 12% will have a positive amniotic fluid culture. Among women with preterm PROM, only 20% will have clinical signs of chorioamnionitis when a positive amniotic fluid culture is present.

EAR: If most infections are subclinical in nature, how can they be detected?

RR: The most accurate method to detect the presence of intraamniotic infection is analysis of amniotic fluid. Amniotic fluid is normally sterile for bacteria. It is possible to isolate some viruses from amniotic fluid, but generally, cultures for viruses are not performed in patients with preterm labor and preterm PROM. Amniotic fluid should be cultured for aerobic and anaerobic bacteria, as well as for Mycoplasma species. Detection of mycoplasmas is important, because they are the most common organisms found in the amniotic fluid. Commercially available systems exist that can be implemented in U.S. laboratories.

EAR: The results of cultures take several days to become available. Thus, rapid tests are required to assess the likelihood of infection or inflammation. What tests would you recommend for this purpose?

RR: A positive Gram stain has 99% specificity, but 20% sensitivity in the detection of intraamniotic infection. The low sensitivity is because the Gram stain cannot detect mycoplasmas since these organisms are too small to be seen with light microscopy. However, the take-home message is that a positive Gram stain is virtually always associated with a positive culture and that false positives are rare. The current approach to the detection of infection and/or inflammation in the amniotic fluid includes other tests that are routinely performed for the analysis of cerebrospinal fluid in all hospitals in the United States. Such tests include a white blood cell (WBC) count of amniotic fluid, and a glucose determination.

EAR: How can the clinician interpret the results of an amniotic fluid WBC and an amniotic fluid glucose determination?

RR: White blood cells—such as neutrophils, monocytes, or eosinophils—are not normally present in the amniotic fluid. Therefore, a high count of white blood cells is an indicator that intraamniotic inflammation is present. We recommend that an amniotic fluid sample be sent to the clinical hematology laboratory, that a WBC count be performed in the standard hemacytometer chamber, and that this be followed by a differential count. When the WBC count is greater than 50 cells/L in patients with intact membranes—or greater than 30 in patients with preterm PROM—the likelihood of a positive amniotic fluid culture is high.

In terms of the amniotic fluid glucose concentration, under normal circumstances glucose is present in the amniotic fluid. The lower the amniotic fluid glucose value, the higher the likelihood of intraamniotic infection or inflammation. For example, glucose values less than 14 mg/dL in women with intact membranes—or less than 10 mg/dL in women with preterm PROM—suggest that intraamniotic infection and/or inflammation is present.

EAR: Are there other tests, such as measurements of cytokines or other proteins, that can be used to detect inflammation in the amniotic fluid?

RR: The concentrations of a cytokine, such as interleukin-6, can be used to detect inflammation. Similarly, we have developed a rapid test that can be used at the bedside to detect inflammation by detecting the concentration of an enzyme produced by neutrophils. This enzyme is MMP-8 (matrix metalloproteinase-8).

These tests can be valuable in the context of midtrimester amniocentesis. The rate of pregnancy loss after a midtrimester amniocentesis has been estimated to be 0.5%–1%; such losses have mistakenly been thought to be always procedure related. However, we have found that among women who have midtrimester amniocenteses, those who have an elevated IL-6 or MMP-8 concentration are more likely to lose their pregnancy or have a spontaneous abortion shortly after the procedure. In these circumstances, determination of IL-6 or MMP-8 in the amniotic fluid stored by the genetic laboratories may be helpful for the patient and physician to identify that intraamniotic inflammation was a cause of the pregnancy loss. This may also have medicolegal implications.

EAR: How common is intraamniotic infection and/or inflammation in women having genetic amniocenteses in the midtrimester of pregnancy?

RR: The frequency of intraamniotic infection has been estimated to be 0.9%; the frequency of intraamniotic inflammation is about 1.2%. The most common organism found in the amniotic fluid is U. urealyticum. Intraamniotic infection and/or inflammation is more common in women who have discolored amniotic fluid at the time of genetic amniocentesis. It is important to realize that these infections are subclinical and that sometimes, patients with these infections rupture their membranes within hours or days of the procedure.

EAR: Can treatment be offered to these patients with midtrimester intraamniotic infections?

RR: Recent evidence from Dr. Sonia Hassan in our group indicates that the administration of antibiotics to the mother can eradicate intraamniotic infection in the midtrimester. Women with a short cervix detected by ultrasound were found to have microorganisms in 9% of cases. Patients were offered treatment with antibiotics and a repeat amniocentesis was performed to be sure that the infections were eradicated. Most women treated in this fashion had eradication of their intraamniotic infection and their pregnancy went to term.

EAR: How frequently are intrauterine infections confined to the amniotic fluid, and how often is the fetus involved?

RR: A study conducted in the United Kingdom in women with preterm PROM indicated that approximately 30% of patients had microorganisms in the amniotic fluid. Of these, 30% had positive fetal blood cultures. This means that 10% of all fetuses with preterm PROM will have fetal bacteremia. Clearly, this represents a minimum estimate of the frequency of fetal infection, a result of the limitations of standard techniques and the difficulties in isolating relevant microorganisms from fetal blood.

EAR: What is the importance of congenital neonatal infections?

RR: Sepsis is a more serious disease in neonates than in adults. Neonates have been generally considered immunosuppressed hosts, and the lethality of sepsis in neonates is high. There is now accumulating evidence that neonates with sepsis are more likely to develop cerebral palsy and bronchopulmonary dysplasia or chronic lung disease.

EAR: How important is intrauterine infection as a cause for cerebral palsy?

RR: It has been estimated that as many as 20% of all cases of cerebral palsy result from infection. Moreover, this applies to term neonates as well as to preterm neonates. Therefore, the traditional paradigm—that intrapartum asphyxia was the leading cause of cerebral palsy—is probably not correct. Obstetricians need to be aware that undiagnosed infections can be a cause for cerebral palsy because this has medicolegal implications.

EAR: What is the link between infection and the brain injury associated with cerebral palsy?

RR: Microorganisms involved in cases of intraamniotic infection can invade the human fetus. When the fetus breathes or swallows infected amniotic fluid, microorganisms may be entering the fetal compartment. Once microorganisms invade the fetus, they elicit a fetal inflammatory response syndrome (FIRS), which is the counterpart of the systemic inflammatory response syndrome (SIRS) in the adult. In FIRS one of the most critical organs affected is the brain. Microorganisms or their products that gain access to the fetal brain can induce damage of neurons or white matter in utero. Damage to the white matter in utero is also known as periventricular leukomalacia (PVL) and is the most important predictor of cerebral palsy. There is evidence that cytokines, chemokines, and other inflammatory products—such as reactive oxygen metabolites—can cause damage to the glia or to neurons, which is responsible for the cognitive abnormalities, including mental retardation.

EAR: If most intrauterine infections are subclinical, how can an obstetrician determine whether or not a neonate had intrauterine infection after birth?

RR: One possibility is to look at the placenta. Inflammation in the placenta can be of maternal origin or of fetal origin. Histologic chorioamnionitis is inflammation of the chorioamniotic membranes caused by maternal cells and is, therefore, a maternal inflammatory response. By contrast, funisitis—inflammation of the umbilical cord—is a fetal inflammatory response. Therefore, the presence of funisitis, diagnosed by examination of the placenta, indicates that the fetus was exposed to microorganisms before birth, or that the fetus mounted a FIRS. This is the reason why we call funisitis the hallmark of FIRS. The practical implication of this is that the examination of the placenta may be helpful in understanding what happened before birth. This is particularly important, given that funisitis has been associated with the subsequent development of cerebral palsy. The medicolegal implications of this are apparent. Because there is no known treatment for funisitis, there is no evidence that any intervention by obstetricians can prevent cerebral palsy associated with or induced by intrauterine infection.

EAR: Do systemic infections cause premature labor?

RR: Systemic clinical infections—such as pyelonephritis, pneumonia, malaria, and appendicitis—have been associated with premature labor and delivery. However, there is recent evidence that subclinical distant infections may also be a cause of premature labor and delivery. Specifically, periodontal disease, which is a chronic inflammatory process, has been associated with the subsequent development of preterm labor as well as with small-for-gestational-age infants.

The Changing Approach to Preterm Labor

Preterm delivery accounts for a significant component of infant mortality in the world. In this, the United States has not been spared; in fact, our country ranks a dismal 21st internationally in infant mortality, with prematurity as a major contributor.

Historically, obstetrics has approached this problem from a therapeutic perspective: If you see a contraction, try to stop it. In large measure, we have been unsuccessful, staving off delivery by a mean of approximately 2 days despite our best efforts. Although this window can allow for the stabilization of patients, arranging for their transfer to appropriate delivery sites, and initiating required medications, it does not solve the problem of prematurity.

By focusing on the symptoms of premature labor, we have not paused sufficiently to ask basic questions about potential causes and triggers that could help us to develop preventive strategies and targeted treatments for what is clearly a multifactorial syndrome.

This is all changing. The National Institutes of Health saw this as such a priority that it formed a Perinatology Research Branch within the National Institute of Child Health and Development and chose international authority Dr. Roberto Romero to lead it. The Centers for Disease Control and Prevention and the March of Dimes have similarly launched research initiatives and made the prevention of prematurity a priority.

Much progress has been made, with Dr. Romero's team taking a role in the forefront. I am very pleased to welcome Dr. Romero, professor of ob.gyn. at Wayne State University in Detroit and chief of the Perinatology Research Branch at NIH, as this month's guest professor. We will review advances in our understanding of the biology of prematurity as a syndrome and offer potential treatment implications, beginning with a focus on infection.

In September, we will similarly review the other contributors to the prematurity syndrome.