Dr. Meena Khandelwal

Preterm birth continues to be a major problem for our patients and for our society. Although the preterm birth rate in the United States fell for a fourth consecutive year in 2010, it remains alarmingly high. The incidence of preterm birth was almost 12% in 2009, according to the Centers for Disease Control and Prevention. It has been estimated, moreover, that two-thirds of preterm births in the United States are spontaneous.

Tocolysis can delay preterm birth by up to 48 hours and improve neonatal outcomes by allowing time for administration of corticosteroids. In general, however, tertiary prevention has been ineffective in decreasing the number of preterm births, and our attention has shifted to primary and secondary prevention. Several risk factors – from body mass index to various infections – have been studied, but their sensitivity or positive predictive value has been disappointing. However, two of these factors – a history of prior preterm birth and short cervical length – have proved to be important independent predictors of preterm delivery.

Term labor is preceded by cervical ripening and effacement by several days. It would make logical sense that preterm cervical effacement would similarly predict preterm delivery. Since the 1990s, research from various countries has shown that cervical effacement, detected as a short cervix by transvaginal ultrasound in the midtrimester, is indeed a powerful predictor of spontaneous preterm birth. Detecting women at high risk for preterm birth without proven effective therapies or interventions, however, would only create anxiety for the mother and frustration for the physician. Routine screening, therefore, traditionally was not offered.

Today, however, we have ample evidence from the past decade of the safety and benefit of vaginal progesterone – and possibly other interventions – in women with singleton gestation and cervical shortening. Multiple studies have demonstrated that vaginal progesterone can significantly cut the rate of spontaneous preterm birth among the approximately 2% of women who have asymptomatic short cervix (most commonly defined as less than or equal to 2 cm) in the midtrimester.

Images courtesy Dr. Meena Khandelwal

Transvaginal ultrasound, in the meantime, has been widely recognized as a relatively safe, easy-to-perform, sensitive, and reproducible method for detecting shortened cervical length. It can easily be performed at 18-22 weeks’ gestation in conjunction with the fetal anatomic survey.

We currently have the knowledge and tools, therefore, to further reduce the incidence of preterm birth. By screening for short cervix, we can offer hope to more women that preterm birth can be averted. We also can improve the rates of neonatal and infant morbidity and mortality as well as longer-term prematurity-related medical problems and societal issues. It is time to recommend routine cervical length screening of women with singleton gestation at midtrimester, and to offer prophylactic treatment to those with short cervical length.

Efficacy of Progesterone

The first meaningful advance with respect to progesterone and prevention of preterm birth came in 2003 with the da Fonseca study using daily vaginal progesterone (Am. J. Obstet. Gynecol. 2003;188:419-24) and the Meis study using weekly injections of 17-alpha-hydroxyprogesterone (17P). Both double-blind, placebo-controlled studies looked at prevention of recurrent preterm delivery in women with a singleton pregnancy who had previously delivered prematurely, and both showed efficacy of progesterone. Dr. Meis and colleagues, for instance, reported that progesterone treatment reduced the risk of delivery at less than 37 weeks’ gestation by approximately one-third (N. Engl. J. Med. 2003;348:2379-85).

Images courtesy Dr. Meena Khandelwal

The value of careful history taking and prophylactic intervention for women with a prior preterm birth was recognized and approved by medical organizations and widely implemented in practice.

Subsequently, several other studies demonstrated a benefit of vaginal progesterone, which is less painful than 17P, in women with a short cervix. The Fetal Medicine Foundation Second Trimester Screening Group from London, for instance, reported a 44% lower rate of spontaneous delivery before 34 weeks’ gestation in women who were randomized to receive daily vaginal progesterone than in those randomized to receive a placebo (19.2% vs. 34.4%). The 250 asymptomatic women who participated had a cervical length of 15 mm or less at 20-25 weeks’ gestation; they represented 1.7% of 24,620 women who were screened during routine prenatal care, but accounted for approximately 26% of all spontaneous preterm births (N. Engl. J. Med. 2007;357:462-9).

An even larger National Institutes of Health–collaborated, randomized, double-blind trial conducted at 44 centers in 10 countries documented a 45% reduction in the rate of preterm birth before 33 weeks’ gestation in women who received daily progesterone gel compared with placebo. The study enrolled 458 asymptomatic women with cervical length of 10-20 mm at 19-23 weeks’ gestation.

Vaginal progesterone was also associated in this trial with a 38% reduction in spontaneous preterm birth before 35 weeks’ gestation, and a 50% reduction before 28 weeks’ gestation. In addition, there was a significantly lower incidence of respiratory distress syndrome (RDS) and a lower rate of any neonatal morbidity or mortality in the vaginal progesterone group (Ultrasound Obstet. Gynecol. 2011;38:18-31).

The publication this year of an individual patient data meta-analysis of randomized controlled trials in asymptomatic women with a midtrimester cervical length of 25 mm or less provides added evidence of progesterone’s benefit (Am. J. Obstet. Gynecol. 2012;206:124.e1-19).

Images courtesy Dr. Meena Khandelwal

The meta-analysis by Dr. Roberto Romero and colleagues covered 775 women (723 singleton pregnancies) with a sonographic short cervix of 25 mm or less in the midtrimester. Overall, treatment with vaginal progesterone was associated with a 42% reduction in the rate of preterm birth at less than 33 weeks’ gestation. Women with singleton gestation and no prior preterm birth had a 40% reduction, and women with a singleton gestation and at least one prior preterm birth had a 46% reduction.

Treatment with vaginal progesterone also was associated in singletons with significant reductions in the rate of preterm birth at less than 35 weeks’ and less than 28 weeks’ gestation, the incidence of RDS, composite neonatal morbidity and mortality, birth weight, and NICU admission.

A beneficial role of pessary use in women with midtrimester short cervical length is also emerging, with benefits seen in the PECEP trial (Lancet 2012;379:1800-6).

Ready for Screening

In the past 3 years, at least several articles advocating universal cervical length screening and vaginal progesterone – and at least two cost-effectiveness analyses supporting such calls – have been published. One, by a team at Yale University, for instance, calculated significant cost-savings and numbers of neonatal deaths and long-term neurologic deficits that would be prevented with universal screening in low-risk pregnancies (Ultrasound Obstet. Gynecol. 2011;38:32-7).

Unfortunately, however, the use of transvaginal ultrasound cervical length screening in singleton gestations without a history of spontaneous preterm birth continues to be a controversial issue.

In a committee opinion on "Incidentally Detected Short Cervical Length" issued in April this year, the American College of Obstetricians and Gynecologists (ACOG) recommended that a cervical length measurement be performed at the time the ultrasound examination is undertaken for fetal anatomic survey at around 18-22 weeks’ gestation, and said "cervical length measured by transvaginal ultrasound examination is a useful screening test for predicting spontaneous preterm birth."

Interestingly, the ACOG opinion was withdrawn shortly after publication and the college issued a Practice Bulletin in October 2012 on "Prediction and Prevention of Preterm Birth." The bulletin reaffirms that "cervical length screening by transvaginal ultrasonography is safe, highly reproducible, and more predictive than transabdominal ultrasound screening" and digital examination.

Despite such conclusions, however, ACOG fell short of mandating routine transvaginal cervical length screening in women without prior preterm birth. The bulletin states instead that "this screening strategy may be considered" and that "practitioners who decide to implement universal cervical length screening should follow one of the protocols for transvaginal measurement of cervical length from the clinical trials" (Obstet. Gynecol. 2012;120:964-73).

There are no recommendations regarding specific management options or strategies for short cervix detected by transabdominal ultrasound or other methods, possibly because published studies have involved diagnoses confirmed by transvaginal ultrasound. It seems fair, therefore, to assume that if a short cervix is detected by any other method, transvaginal ultrasound will need to be performed for accuracy and to determine the prognosis and management plan.

The Society for Maternal-Fetal Medicine also recognized in a recently published guideline that universal cervical screening in singleton pregnancies without prior preterm birth is a "reasonable" option – one that can be considered by individual practitioners but "cannot be mandated" (Am. J. Obstet. Gynecol. 2012;206:376-86).

The SMFM guideline takes a more decisive stance with respect to singletons with prior preterm birth, recommending transvaginal ultrasound cervical length measurements every 2 weeks and providing management decisions for those who choose to offer universal screening in all singleton gestations. All told, the language suggests to me that SMFM is encouraging universal screening without mandating it.

Critics of universal screening have cited a concern that facilities for transvaginal ultrasound screening are not widely enough available. Others have spoken about an uncertainty that screening outcomes in actual practice will compare favorably with the outcomes of controlled trials. There also is concern about the potential misuse or overuse of technology, and concern that many women will undergo treatment unnecessarily.

I understand some of these concerns, but I do not believe they are justifiable reasons for not requiring the use of a cost-effective, beneficial screening test. Such concerns are inherent to the introduction of any new test or therapy. In the past, we have successfully implemented many advances and breakthrough findings that have involved similar concerns. Even if the degree of benefit is debated, the risk-benefit ratio balance still tips toward routine screening and the use of vaginal progesterone in women with shortened cervical length – especially because there are no significant side effects. Moreover, if the SMFM can so strongly recommend frequent transvaginal ultrasound measurements for women with prior preterm birth, then the concern regarding availability of technology may not be as significant or insurmountable as some have suggested. The infrastructure and equipment for transvaginal ultrasound already exists in most centers that perform fetal ultrasound.

Certainly, transvaginal ultrasound is preferable over transabdominal ultrasound. Although transabdominal screening would be better than no screening at all, it has been shown to have higher failure rates and lower sensitivity than the transvaginal approach.

Transvaginal ultrasound by far has been the main screening tool in studies on cervical length and the risk of preterm birth as well as studies looking at the impact of intervention. In the one major study that I could identify employing transabdominal ultrasound as the primary screening tool, the percentage of women identified with a cervical length of 15 mm or less was 0.6%, compared with 1.6%-1.7% in studies using transvaginal ultrasound as the front-line screening tool. The transvaginal approach also has been shown in numerous studies to be well accepted by patients.

Some physicians have advocated the use of CerviLenz, an instrument that measures the cervico-portio length without the use of ultrasound. In the one prospective trial of CerviLenz, the instrument diagnosed cervical length less than 3 cm with a sensitivity of 88% and a specificity of 92% (J. Reprod. Med. 2007;52:385-389). Transvaginal ultrasound was needed to confirm the findings. While CerviLenz could possibly be used in areas where transvaginal ultrasound is not available, the gold standard is transvaginal ultrasound. By screening with anything less than the gold standard, we would be providing suboptimal care.

In general, most agree that some form of cervical length screening in the midtrimester is required. The major point of debate, it seems, is whether it should be transabdominal or transvaginal ultrasound. The fact is that transabdominal scan identifies less than 50% of women with short cervix, and a majority of these women have severe cervical shortening (less than 10 mm) – a scenario where vaginal progesterone is minimally effective. Using this method will have little impact on the rate of preterm deliveries.

For optimal care of pregnant women, we have instituted universal transvaginal cervical length screening of all women with singleton gestation at the time of their anatomy scan. If cervical length is 5-20 mm, vaginal progesterone is offered; if the woman had a prior preterm delivery, the choice of cervical cerclage is also offered. If the cervical length is 21-25 mm, transvaginal ultrasound is repeated in 1-2 weeks.

This column, Master Class, appears regularly in Ob.Gyn News. Dr. Khandelwal is professor of obstetrics and gynecology at Cooper Medical School of Rowan University, Camden, N.J. She was an investigator in the NIH-collaborated study on vaginal progesterone that is mentioned in this Master Class (Ultrasound Obstet. Gynecol. 2011:38:18-31), and served as a paid consultant to Watson Pharmaceuticals Inc. in its evaluation of and FDA submission in 2011 for the Prochieve vaginal progesterone gel.

Preterm birth continues to be a major problem for our patients and for our society. Although the preterm birth rate in the United States fell for a fourth consecutive year in 2010, it remains alarmingly high. The incidence of preterm birth was almost 12% in 2009, according to the Centers for Disease Control and Prevention. It has been estimated, moreover, that two-thirds of preterm births in the United States are spontaneous.

Tocolysis can delay preterm birth by up to 48 hours and improve neonatal outcomes by allowing time for administration of corticosteroids. In general, however, tertiary prevention has been ineffective in decreasing the number of preterm births, and our attention has shifted to primary and secondary prevention. Several risk factors – from body mass index to various infections – have been studied, but their sensitivity or positive predictive value has been disappointing. However, two of these factors – a history of prior preterm birth and short cervical length – have proved to be important independent predictors of preterm delivery.

Term labor is preceded by cervical ripening and effacement by several days. It would make logical sense that preterm cervical effacement would similarly predict preterm delivery. Since the 1990s, research from various countries has shown that cervical effacement, detected as a short cervix by transvaginal ultrasound in the midtrimester, is indeed a powerful predictor of spontaneous preterm birth. Detecting women at high risk for preterm birth without proven effective therapies or interventions, however, would only create anxiety for the mother and frustration for the physician. Routine screening, therefore, traditionally was not offered.

Today, however, we have ample evidence from the past decade of the safety and benefit of vaginal progesterone – and possibly other interventions – in women with singleton gestation and cervical shortening. Multiple studies have demonstrated that vaginal progesterone can significantly cut the rate of spontaneous preterm birth among the approximately 2% of women who have asymptomatic short cervix (most commonly defined as less than or equal to 2 cm) in the midtrimester.

Images courtesy Dr. Meena Khandelwal

Transvaginal ultrasound, in the meantime, has been widely recognized as a relatively safe, easy-to-perform, sensitive, and reproducible method for detecting shortened cervical length. It can easily be performed at 18-22 weeks’ gestation in conjunction with the fetal anatomic survey.

We currently have the knowledge and tools, therefore, to further reduce the incidence of preterm birth. By screening for short cervix, we can offer hope to more women that preterm birth can be averted. We also can improve the rates of neonatal and infant morbidity and mortality as well as longer-term prematurity-related medical problems and societal issues. It is time to recommend routine cervical length screening of women with singleton gestation at midtrimester, and to offer prophylactic treatment to those with short cervical length.

Efficacy of Progesterone

The first meaningful advance with respect to progesterone and prevention of preterm birth came in 2003 with the da Fonseca study using daily vaginal progesterone (Am. J. Obstet. Gynecol. 2003;188:419-24) and the Meis study using weekly injections of 17-alpha-hydroxyprogesterone (17P). Both double-blind, placebo-controlled studies looked at prevention of recurrent preterm delivery in women with a singleton pregnancy who had previously delivered prematurely, and both showed efficacy of progesterone. Dr. Meis and colleagues, for instance, reported that progesterone treatment reduced the risk of delivery at less than 37 weeks’ gestation by approximately one-third (N. Engl. J. Med. 2003;348:2379-85).

Images courtesy Dr. Meena Khandelwal

The value of careful history taking and prophylactic intervention for women with a prior preterm birth was recognized and approved by medical organizations and widely implemented in practice.

Subsequently, several other studies demonstrated a benefit of vaginal progesterone, which is less painful than 17P, in women with a short cervix. The Fetal Medicine Foundation Second Trimester Screening Group from London, for instance, reported a 44% lower rate of spontaneous delivery before 34 weeks’ gestation in women who were randomized to receive daily vaginal progesterone than in those randomized to receive a placebo (19.2% vs. 34.4%). The 250 asymptomatic women who participated had a cervical length of 15 mm or less at 20-25 weeks’ gestation; they represented 1.7% of 24,620 women who were screened during routine prenatal care, but accounted for approximately 26% of all spontaneous preterm births (N. Engl. J. Med. 2007;357:462-9).

An even larger National Institutes of Health–collaborated, randomized, double-blind trial conducted at 44 centers in 10 countries documented a 45% reduction in the rate of preterm birth before 33 weeks’ gestation in women who received daily progesterone gel compared with placebo. The study enrolled 458 asymptomatic women with cervical length of 10-20 mm at 19-23 weeks’ gestation.

Vaginal progesterone was also associated in this trial with a 38% reduction in spontaneous preterm birth before 35 weeks’ gestation, and a 50% reduction before 28 weeks’ gestation. In addition, there was a significantly lower incidence of respiratory distress syndrome (RDS) and a lower rate of any neonatal morbidity or mortality in the vaginal progesterone group (Ultrasound Obstet. Gynecol. 2011;38:18-31).

The publication this year of an individual patient data meta-analysis of randomized controlled trials in asymptomatic women with a midtrimester cervical length of 25 mm or less provides added evidence of progesterone’s benefit (Am. J. Obstet. Gynecol. 2012;206:124.e1-19).

Images courtesy Dr. Meena Khandelwal

The meta-analysis by Dr. Roberto Romero and colleagues covered 775 women (723 singleton pregnancies) with a sonographic short cervix of 25 mm or less in the midtrimester. Overall, treatment with vaginal progesterone was associated with a 42% reduction in the rate of preterm birth at less than 33 weeks’ gestation. Women with singleton gestation and no prior preterm birth had a 40% reduction, and women with a singleton gestation and at least one prior preterm birth had a 46% reduction.

Treatment with vaginal progesterone also was associated in singletons with significant reductions in the rate of preterm birth at less than 35 weeks’ and less than 28 weeks’ gestation, the incidence of RDS, composite neonatal morbidity and mortality, birth weight, and NICU admission.

A beneficial role of pessary use in women with midtrimester short cervical length is also emerging, with benefits seen in the PECEP trial (Lancet 2012;379:1800-6).

Ready for Screening

In the past 3 years, at least several articles advocating universal cervical length screening and vaginal progesterone – and at least two cost-effectiveness analyses supporting such calls – have been published. One, by a team at Yale University, for instance, calculated significant cost-savings and numbers of neonatal deaths and long-term neurologic deficits that would be prevented with universal screening in low-risk pregnancies (Ultrasound Obstet. Gynecol. 2011;38:32-7).

Unfortunately, however, the use of transvaginal ultrasound cervical length screening in singleton gestations without a history of spontaneous preterm birth continues to be a controversial issue.

In a committee opinion on "Incidentally Detected Short Cervical Length" issued in April this year, the American College of Obstetricians and Gynecologists (ACOG) recommended that a cervical length measurement be performed at the time the ultrasound examination is undertaken for fetal anatomic survey at around 18-22 weeks’ gestation, and said "cervical length measured by transvaginal ultrasound examination is a useful screening test for predicting spontaneous preterm birth."

Interestingly, the ACOG opinion was withdrawn shortly after publication and the college issued a Practice Bulletin in October 2012 on "Prediction and Prevention of Preterm Birth." The bulletin reaffirms that "cervical length screening by transvaginal ultrasonography is safe, highly reproducible, and more predictive than transabdominal ultrasound screening" and digital examination.

Despite such conclusions, however, ACOG fell short of mandating routine transvaginal cervical length screening in women without prior preterm birth. The bulletin states instead that "this screening strategy may be considered" and that "practitioners who decide to implement universal cervical length screening should follow one of the protocols for transvaginal measurement of cervical length from the clinical trials" (Obstet. Gynecol. 2012;120:964-73).

There are no recommendations regarding specific management options or strategies for short cervix detected by transabdominal ultrasound or other methods, possibly because published studies have involved diagnoses confirmed by transvaginal ultrasound. It seems fair, therefore, to assume that if a short cervix is detected by any other method, transvaginal ultrasound will need to be performed for accuracy and to determine the prognosis and management plan.

The Society for Maternal-Fetal Medicine also recognized in a recently published guideline that universal cervical screening in singleton pregnancies without prior preterm birth is a "reasonable" option – one that can be considered by individual practitioners but "cannot be mandated" (Am. J. Obstet. Gynecol. 2012;206:376-86).

The SMFM guideline takes a more decisive stance with respect to singletons with prior preterm birth, recommending transvaginal ultrasound cervical length measurements every 2 weeks and providing management decisions for those who choose to offer universal screening in all singleton gestations. All told, the language suggests to me that SMFM is encouraging universal screening without mandating it.

Critics of universal screening have cited a concern that facilities for transvaginal ultrasound screening are not widely enough available. Others have spoken about an uncertainty that screening outcomes in actual practice will compare favorably with the outcomes of controlled trials. There also is concern about the potential misuse or overuse of technology, and concern that many women will undergo treatment unnecessarily.

I understand some of these concerns, but I do not believe they are justifiable reasons for not requiring the use of a cost-effective, beneficial screening test. Such concerns are inherent to the introduction of any new test or therapy. In the past, we have successfully implemented many advances and breakthrough findings that have involved similar concerns. Even if the degree of benefit is debated, the risk-benefit ratio balance still tips toward routine screening and the use of vaginal progesterone in women with shortened cervical length – especially because there are no significant side effects. Moreover, if the SMFM can so strongly recommend frequent transvaginal ultrasound measurements for women with prior preterm birth, then the concern regarding availability of technology may not be as significant or insurmountable as some have suggested. The infrastructure and equipment for transvaginal ultrasound already exists in most centers that perform fetal ultrasound.

Certainly, transvaginal ultrasound is preferable over transabdominal ultrasound. Although transabdominal screening would be better than no screening at all, it has been shown to have higher failure rates and lower sensitivity than the transvaginal approach.

Transvaginal ultrasound by far has been the main screening tool in studies on cervical length and the risk of preterm birth as well as studies looking at the impact of intervention. In the one major study that I could identify employing transabdominal ultrasound as the primary screening tool, the percentage of women identified with a cervical length of 15 mm or less was 0.6%, compared with 1.6%-1.7% in studies using transvaginal ultrasound as the front-line screening tool. The transvaginal approach also has been shown in numerous studies to be well accepted by patients.

Some physicians have advocated the use of CerviLenz, an instrument that measures the cervico-portio length without the use of ultrasound. In the one prospective trial of CerviLenz, the instrument diagnosed cervical length less than 3 cm with a sensitivity of 88% and a specificity of 92% (J. Reprod. Med. 2007;52:385-389). Transvaginal ultrasound was needed to confirm the findings. While CerviLenz could possibly be used in areas where transvaginal ultrasound is not available, the gold standard is transvaginal ultrasound. By screening with anything less than the gold standard, we would be providing suboptimal care.

In general, most agree that some form of cervical length screening in the midtrimester is required. The major point of debate, it seems, is whether it should be transabdominal or transvaginal ultrasound. The fact is that transabdominal scan identifies less than 50% of women with short cervix, and a majority of these women have severe cervical shortening (less than 10 mm) – a scenario where vaginal progesterone is minimally effective. Using this method will have little impact on the rate of preterm deliveries.

For optimal care of pregnant women, we have instituted universal transvaginal cervical length screening of all women with singleton gestation at the time of their anatomy scan. If cervical length is 5-20 mm, vaginal progesterone is offered; if the woman had a prior preterm delivery, the choice of cervical cerclage is also offered. If the cervical length is 21-25 mm, transvaginal ultrasound is repeated in 1-2 weeks.

This column, Master Class, appears regularly in Ob.Gyn News. Dr. Khandelwal is professor of obstetrics and gynecology at Cooper Medical School of Rowan University, Camden, N.J. She was an investigator in the NIH-collaborated study on vaginal progesterone that is mentioned in this Master Class (Ultrasound Obstet. Gynecol. 2011:38:18-31), and served as a paid consultant to Watson Pharmaceuticals Inc. in its evaluation of and FDA submission in 2011 for the Prochieve vaginal progesterone gel.

Preterm birth continues to be a major problem for our patients and for our society. Although the preterm birth rate in the United States fell for a fourth consecutive year in 2010, it remains alarmingly high. The incidence of preterm birth was almost 12% in 2009, according to the Centers for Disease Control and Prevention. It has been estimated, moreover, that two-thirds of preterm births in the United States are spontaneous.

Tocolysis can delay preterm birth by up to 48 hours and improve neonatal outcomes by allowing time for administration of corticosteroids. In general, however, tertiary prevention has been ineffective in decreasing the number of preterm births, and our attention has shifted to primary and secondary prevention. Several risk factors – from body mass index to various infections – have been studied, but their sensitivity or positive predictive value has been disappointing. However, two of these factors – a history of prior preterm birth and short cervical length – have proved to be important independent predictors of preterm delivery.

Term labor is preceded by cervical ripening and effacement by several days. It would make logical sense that preterm cervical effacement would similarly predict preterm delivery. Since the 1990s, research from various countries has shown that cervical effacement, detected as a short cervix by transvaginal ultrasound in the midtrimester, is indeed a powerful predictor of spontaneous preterm birth. Detecting women at high risk for preterm birth without proven effective therapies or interventions, however, would only create anxiety for the mother and frustration for the physician. Routine screening, therefore, traditionally was not offered.

Today, however, we have ample evidence from the past decade of the safety and benefit of vaginal progesterone – and possibly other interventions – in women with singleton gestation and cervical shortening. Multiple studies have demonstrated that vaginal progesterone can significantly cut the rate of spontaneous preterm birth among the approximately 2% of women who have asymptomatic short cervix (most commonly defined as less than or equal to 2 cm) in the midtrimester.

Images courtesy Dr. Meena Khandelwal

Transvaginal ultrasound, in the meantime, has been widely recognized as a relatively safe, easy-to-perform, sensitive, and reproducible method for detecting shortened cervical length. It can easily be performed at 18-22 weeks’ gestation in conjunction with the fetal anatomic survey.

We currently have the knowledge and tools, therefore, to further reduce the incidence of preterm birth. By screening for short cervix, we can offer hope to more women that preterm birth can be averted. We also can improve the rates of neonatal and infant morbidity and mortality as well as longer-term prematurity-related medical problems and societal issues. It is time to recommend routine cervical length screening of women with singleton gestation at midtrimester, and to offer prophylactic treatment to those with short cervical length.

Efficacy of Progesterone

The first meaningful advance with respect to progesterone and prevention of preterm birth came in 2003 with the da Fonseca study using daily vaginal progesterone (Am. J. Obstet. Gynecol. 2003;188:419-24) and the Meis study using weekly injections of 17-alpha-hydroxyprogesterone (17P). Both double-blind, placebo-controlled studies looked at prevention of recurrent preterm delivery in women with a singleton pregnancy who had previously delivered prematurely, and both showed efficacy of progesterone. Dr. Meis and colleagues, for instance, reported that progesterone treatment reduced the risk of delivery at less than 37 weeks’ gestation by approximately one-third (N. Engl. J. Med. 2003;348:2379-85).

Images courtesy Dr. Meena Khandelwal

The value of careful history taking and prophylactic intervention for women with a prior preterm birth was recognized and approved by medical organizations and widely implemented in practice.

Subsequently, several other studies demonstrated a benefit of vaginal progesterone, which is less painful than 17P, in women with a short cervix. The Fetal Medicine Foundation Second Trimester Screening Group from London, for instance, reported a 44% lower rate of spontaneous delivery before 34 weeks’ gestation in women who were randomized to receive daily vaginal progesterone than in those randomized to receive a placebo (19.2% vs. 34.4%). The 250 asymptomatic women who participated had a cervical length of 15 mm or less at 20-25 weeks’ gestation; they represented 1.7% of 24,620 women who were screened during routine prenatal care, but accounted for approximately 26% of all spontaneous preterm births (N. Engl. J. Med. 2007;357:462-9).

An even larger National Institutes of Health–collaborated, randomized, double-blind trial conducted at 44 centers in 10 countries documented a 45% reduction in the rate of preterm birth before 33 weeks’ gestation in women who received daily progesterone gel compared with placebo. The study enrolled 458 asymptomatic women with cervical length of 10-20 mm at 19-23 weeks’ gestation.

Vaginal progesterone was also associated in this trial with a 38% reduction in spontaneous preterm birth before 35 weeks’ gestation, and a 50% reduction before 28 weeks’ gestation. In addition, there was a significantly lower incidence of respiratory distress syndrome (RDS) and a lower rate of any neonatal morbidity or mortality in the vaginal progesterone group (Ultrasound Obstet. Gynecol. 2011;38:18-31).

The publication this year of an individual patient data meta-analysis of randomized controlled trials in asymptomatic women with a midtrimester cervical length of 25 mm or less provides added evidence of progesterone’s benefit (Am. J. Obstet. Gynecol. 2012;206:124.e1-19).

Images courtesy Dr. Meena Khandelwal

The meta-analysis by Dr. Roberto Romero and colleagues covered 775 women (723 singleton pregnancies) with a sonographic short cervix of 25 mm or less in the midtrimester. Overall, treatment with vaginal progesterone was associated with a 42% reduction in the rate of preterm birth at less than 33 weeks’ gestation. Women with singleton gestation and no prior preterm birth had a 40% reduction, and women with a singleton gestation and at least one prior preterm birth had a 46% reduction.

Treatment with vaginal progesterone also was associated in singletons with significant reductions in the rate of preterm birth at less than 35 weeks’ and less than 28 weeks’ gestation, the incidence of RDS, composite neonatal morbidity and mortality, birth weight, and NICU admission.

A beneficial role of pessary use in women with midtrimester short cervical length is also emerging, with benefits seen in the PECEP trial (Lancet 2012;379:1800-6).

Ready for Screening

In the past 3 years, at least several articles advocating universal cervical length screening and vaginal progesterone – and at least two cost-effectiveness analyses supporting such calls – have been published. One, by a team at Yale University, for instance, calculated significant cost-savings and numbers of neonatal deaths and long-term neurologic deficits that would be prevented with universal screening in low-risk pregnancies (Ultrasound Obstet. Gynecol. 2011;38:32-7).

Unfortunately, however, the use of transvaginal ultrasound cervical length screening in singleton gestations without a history of spontaneous preterm birth continues to be a controversial issue.

In a committee opinion on "Incidentally Detected Short Cervical Length" issued in April this year, the American College of Obstetricians and Gynecologists (ACOG) recommended that a cervical length measurement be performed at the time the ultrasound examination is undertaken for fetal anatomic survey at around 18-22 weeks’ gestation, and said "cervical length measured by transvaginal ultrasound examination is a useful screening test for predicting spontaneous preterm birth."

Interestingly, the ACOG opinion was withdrawn shortly after publication and the college issued a Practice Bulletin in October 2012 on "Prediction and Prevention of Preterm Birth." The bulletin reaffirms that "cervical length screening by transvaginal ultrasonography is safe, highly reproducible, and more predictive than transabdominal ultrasound screening" and digital examination.

Despite such conclusions, however, ACOG fell short of mandating routine transvaginal cervical length screening in women without prior preterm birth. The bulletin states instead that "this screening strategy may be considered" and that "practitioners who decide to implement universal cervical length screening should follow one of the protocols for transvaginal measurement of cervical length from the clinical trials" (Obstet. Gynecol. 2012;120:964-73).

There are no recommendations regarding specific management options or strategies for short cervix detected by transabdominal ultrasound or other methods, possibly because published studies have involved diagnoses confirmed by transvaginal ultrasound. It seems fair, therefore, to assume that if a short cervix is detected by any other method, transvaginal ultrasound will need to be performed for accuracy and to determine the prognosis and management plan.

The Society for Maternal-Fetal Medicine also recognized in a recently published guideline that universal cervical screening in singleton pregnancies without prior preterm birth is a "reasonable" option – one that can be considered by individual practitioners but "cannot be mandated" (Am. J. Obstet. Gynecol. 2012;206:376-86).

The SMFM guideline takes a more decisive stance with respect to singletons with prior preterm birth, recommending transvaginal ultrasound cervical length measurements every 2 weeks and providing management decisions for those who choose to offer universal screening in all singleton gestations. All told, the language suggests to me that SMFM is encouraging universal screening without mandating it.

Critics of universal screening have cited a concern that facilities for transvaginal ultrasound screening are not widely enough available. Others have spoken about an uncertainty that screening outcomes in actual practice will compare favorably with the outcomes of controlled trials. There also is concern about the potential misuse or overuse of technology, and concern that many women will undergo treatment unnecessarily.

I understand some of these concerns, but I do not believe they are justifiable reasons for not requiring the use of a cost-effective, beneficial screening test. Such concerns are inherent to the introduction of any new test or therapy. In the past, we have successfully implemented many advances and breakthrough findings that have involved similar concerns. Even if the degree of benefit is debated, the risk-benefit ratio balance still tips toward routine screening and the use of vaginal progesterone in women with shortened cervical length – especially because there are no significant side effects. Moreover, if the SMFM can so strongly recommend frequent transvaginal ultrasound measurements for women with prior preterm birth, then the concern regarding availability of technology may not be as significant or insurmountable as some have suggested. The infrastructure and equipment for transvaginal ultrasound already exists in most centers that perform fetal ultrasound.

Certainly, transvaginal ultrasound is preferable over transabdominal ultrasound. Although transabdominal screening would be better than no screening at all, it has been shown to have higher failure rates and lower sensitivity than the transvaginal approach.

Transvaginal ultrasound by far has been the main screening tool in studies on cervical length and the risk of preterm birth as well as studies looking at the impact of intervention. In the one major study that I could identify employing transabdominal ultrasound as the primary screening tool, the percentage of women identified with a cervical length of 15 mm or less was 0.6%, compared with 1.6%-1.7% in studies using transvaginal ultrasound as the front-line screening tool. The transvaginal approach also has been shown in numerous studies to be well accepted by patients.

Some physicians have advocated the use of CerviLenz, an instrument that measures the cervico-portio length without the use of ultrasound. In the one prospective trial of CerviLenz, the instrument diagnosed cervical length less than 3 cm with a sensitivity of 88% and a specificity of 92% (J. Reprod. Med. 2007;52:385-389). Transvaginal ultrasound was needed to confirm the findings. While CerviLenz could possibly be used in areas where transvaginal ultrasound is not available, the gold standard is transvaginal ultrasound. By screening with anything less than the gold standard, we would be providing suboptimal care.

In general, most agree that some form of cervical length screening in the midtrimester is required. The major point of debate, it seems, is whether it should be transabdominal or transvaginal ultrasound. The fact is that transabdominal scan identifies less than 50% of women with short cervix, and a majority of these women have severe cervical shortening (less than 10 mm) – a scenario where vaginal progesterone is minimally effective. Using this method will have little impact on the rate of preterm deliveries.

For optimal care of pregnant women, we have instituted universal transvaginal cervical length screening of all women with singleton gestation at the time of their anatomy scan. If cervical length is 5-20 mm, vaginal progesterone is offered; if the woman had a prior preterm delivery, the choice of cervical cerclage is also offered. If the cervical length is 21-25 mm, transvaginal ultrasound is repeated in 1-2 weeks.

This column, Master Class, appears regularly in Ob.Gyn News. Dr. Khandelwal is professor of obstetrics and gynecology at Cooper Medical School of Rowan University, Camden, N.J. She was an investigator in the NIH-collaborated study on vaginal progesterone that is mentioned in this Master Class (Ultrasound Obstet. Gynecol. 2011:38:18-31), and served as a paid consultant to Watson Pharmaceuticals Inc. in its evaluation of and FDA submission in 2011 for the Prochieve vaginal progesterone gel.