Debra L. Beck

TORONTO – Transcatheter mitral valve-in-valve/valve-in-ring implantation is an effective option in most patients with degenerative mitral valves after surgical intervention, according to Dr. Danny Dvir.

In the highest-risk patients (STS scores greater than 20%), however, mortality after this percutaneous technique approached 50% at 1 year, Dr. Dvir reported at the American Association for Thoracic Surgery annual meeting.

"Most of these procedures were clinically effective, with 1-year results that are comparable to native aortic valve transcatheter implantation," said Dr. Dvir of St. Paul’s Hospital, Vancouver, B.C.

"However, we saw safety and efficacy concerns that included relatively high in-hospital mortality, device malposition, sporadic cases with elevated LVOT [left ventricular outflow tract] gradient, and elevated postprocedural gradients in procedures performed inside small surgical valves."

Dr. Joseph E. Bavaria, professor of surgery and director of the thoracic aortic surgery program at the University of Pennsylvania Health System, Philadelphia, the scheduled discussant of the abstract, asked, referring to the STS score greater than 20% group: "Is this really a cohort C or futile situation, where [the intervention] is probably not better than medical management?"

Transcatheter VIV/VIR

The majority of prosthetic heart valves are tissue valves that tend to degenerate and eventually fail over time. Surgical replacement of these valves is associated with significant risks.

Transcatheter valve-in-valve/valve-in-ring (VIV/VIR) implantation within a failed bioprosthesis performed in the mitral position has been reported in only small case series, with the safety and efficacy of the technique being largely unknown. The Global Valve-in-Valve Registry, with 765 patients from 73 centers worldwide, was designed to track the efficacy and safety of this nonsurgical redo in either the aortic or mitral valve position. The registry included 190 patients from 23 centers who underwent percutaneous mitral valve procedures (VIV in 157 patients and VIR in 33).

The mechanism of failure was regurgitation in 37%, stenosis in 25%, and a combination of the two in 38%. Patients were a median of 9 years past their last cardiac surgery and were deemed at high risk, with a mean STS score of 14.4. The mean age was 73.6 years, and 65.2% were female. The median length of hospital stay was 8 days.

A SAPIEN or SAPIEN XT device was used in the majority of cases (93.7%), mostly deployed via transapical access (84.7%).

During the procedure, initial device malpositioning occurred in 5.3% of cases, with a second device implanted in 4.3%. Postimplantation valvuloplasty was utilized in 8%.

"We saw a high rate of malpositioning," Dr. Dvir said. "In some of the cases, this led to a decision to leave the malpositioned device in the left atrium and implant a second device."

The incidence of all-cause death at 30 days was 8.9%, with cardiovascular death seen in 6.8% and major stroke in 2.2%. Importantly, 85.8% of patients were in NYHA function class I or II at 30 days, Dr. Dvir said.

"We were very happy to see this," he commented. "If you survive the valve-in-valve or valve-in-ring procedure, you will most likely have a good functional class.

Overall, the Kaplan-Meier mortality rate was 22.3% at 1 year. When stratified according to STS score, however, mortality in the group of patients with an STS score greater than 20% was 45.6%, falling to 23.2% for an STS score of 10%-20% and 12.9% for an STS score less than 10%. Predictors of 1-year mortality included STS score and renal failure.

"In these patients with STS greater than 20%, there is really a big problem. We need to understand which patients should not undergo a procedure, and maybe should be treated with an alternative approach," Dr. Dvir said. He noted that 28% of cases in the registry were in this highest-risk category.

"An 8.9% mortality at 1 year is not so bad in a group with a mean STS score of 14.1," noted Dr. Bavaria. "[But] this is not really a benign procedure as the length of stay was over 8 days."

Surgical valve sizing gone wrong?

After the VIV/VIR procedure, the mean mitral valve area was 2.1 ± 0.7 cm², with a mean residual gradient of 6.2 mm Hg. Significant mitral regurgitation was noted in 4.2% of patients, and a left ventricular outflow tract mean gradient greater than 20 mm Hg was noted in only 2.1% of cases.

"We saw a trend for elevated gradients especially when we were treating small valves with a valve-in-valve approach," Dr. Dvir said.

Indeed, in the patients with smaller surgical valve label size, one-third had postprocedural mean gradients greater than 10 mm Hg. The rates of mean gradient greater than 10 mm Hg dropped to 7.4% and 5.3% with progressively larger surgical valve label sizes.

"I am very worried about the cardiac surgery and interventional cardiology decision-making process, as over 60% of valve-in-ring cases were 26-mm and 28-mm rings and over 56% of valve-in-valve cases were less than or equal to 27-mm prosthetic valves," said Dr. Bavaria. "There will obviously be significant mitral gradients in these cases, as this study showed. Is this simply a lack of knowledge in the operators, should these small valve-in-valve cases be done, or is this a limit of this technology?"

"I think this issue of residual gradients will haunt valve in valve for many years to come," said Dr. Dvir. "However, we know from the aortic arena, that even if you have a high residual gradient, the results at 1 year are not different than the other groups. So, it doesn’t affect 1-year outcomes, although I predict that it will affect the durability of the device and the long-term outcome."

TORONTO – Transcatheter mitral valve-in-valve/valve-in-ring implantation is an effective option in most patients with degenerative mitral valves after surgical intervention, according to Dr. Danny Dvir.

In the highest-risk patients (STS scores greater than 20%), however, mortality after this percutaneous technique approached 50% at 1 year, Dr. Dvir reported at the American Association for Thoracic Surgery annual meeting.

"Most of these procedures were clinically effective, with 1-year results that are comparable to native aortic valve transcatheter implantation," said Dr. Dvir of St. Paul’s Hospital, Vancouver, B.C.

"However, we saw safety and efficacy concerns that included relatively high in-hospital mortality, device malposition, sporadic cases with elevated LVOT [left ventricular outflow tract] gradient, and elevated postprocedural gradients in procedures performed inside small surgical valves."

Dr. Joseph E. Bavaria, professor of surgery and director of the thoracic aortic surgery program at the University of Pennsylvania Health System, Philadelphia, the scheduled discussant of the abstract, asked, referring to the STS score greater than 20% group: "Is this really a cohort C or futile situation, where [the intervention] is probably not better than medical management?"

Transcatheter VIV/VIR

The majority of prosthetic heart valves are tissue valves that tend to degenerate and eventually fail over time. Surgical replacement of these valves is associated with significant risks.

Transcatheter valve-in-valve/valve-in-ring (VIV/VIR) implantation within a failed bioprosthesis performed in the mitral position has been reported in only small case series, with the safety and efficacy of the technique being largely unknown. The Global Valve-in-Valve Registry, with 765 patients from 73 centers worldwide, was designed to track the efficacy and safety of this nonsurgical redo in either the aortic or mitral valve position. The registry included 190 patients from 23 centers who underwent percutaneous mitral valve procedures (VIV in 157 patients and VIR in 33).

The mechanism of failure was regurgitation in 37%, stenosis in 25%, and a combination of the two in 38%. Patients were a median of 9 years past their last cardiac surgery and were deemed at high risk, with a mean STS score of 14.4. The mean age was 73.6 years, and 65.2% were female. The median length of hospital stay was 8 days.

A SAPIEN or SAPIEN XT device was used in the majority of cases (93.7%), mostly deployed via transapical access (84.7%).

During the procedure, initial device malpositioning occurred in 5.3% of cases, with a second device implanted in 4.3%. Postimplantation valvuloplasty was utilized in 8%.

"We saw a high rate of malpositioning," Dr. Dvir said. "In some of the cases, this led to a decision to leave the malpositioned device in the left atrium and implant a second device."

The incidence of all-cause death at 30 days was 8.9%, with cardiovascular death seen in 6.8% and major stroke in 2.2%. Importantly, 85.8% of patients were in NYHA function class I or II at 30 days, Dr. Dvir said.

"We were very happy to see this," he commented. "If you survive the valve-in-valve or valve-in-ring procedure, you will most likely have a good functional class.

Overall, the Kaplan-Meier mortality rate was 22.3% at 1 year. When stratified according to STS score, however, mortality in the group of patients with an STS score greater than 20% was 45.6%, falling to 23.2% for an STS score of 10%-20% and 12.9% for an STS score less than 10%. Predictors of 1-year mortality included STS score and renal failure.

"In these patients with STS greater than 20%, there is really a big problem. We need to understand which patients should not undergo a procedure, and maybe should be treated with an alternative approach," Dr. Dvir said. He noted that 28% of cases in the registry were in this highest-risk category.

"An 8.9% mortality at 1 year is not so bad in a group with a mean STS score of 14.1," noted Dr. Bavaria. "[But] this is not really a benign procedure as the length of stay was over 8 days."

Surgical valve sizing gone wrong?

After the VIV/VIR procedure, the mean mitral valve area was 2.1 ± 0.7 cm², with a mean residual gradient of 6.2 mm Hg. Significant mitral regurgitation was noted in 4.2% of patients, and a left ventricular outflow tract mean gradient greater than 20 mm Hg was noted in only 2.1% of cases.

"We saw a trend for elevated gradients especially when we were treating small valves with a valve-in-valve approach," Dr. Dvir said.

Indeed, in the patients with smaller surgical valve label size, one-third had postprocedural mean gradients greater than 10 mm Hg. The rates of mean gradient greater than 10 mm Hg dropped to 7.4% and 5.3% with progressively larger surgical valve label sizes.

"I am very worried about the cardiac surgery and interventional cardiology decision-making process, as over 60% of valve-in-ring cases were 26-mm and 28-mm rings and over 56% of valve-in-valve cases were less than or equal to 27-mm prosthetic valves," said Dr. Bavaria. "There will obviously be significant mitral gradients in these cases, as this study showed. Is this simply a lack of knowledge in the operators, should these small valve-in-valve cases be done, or is this a limit of this technology?"

"I think this issue of residual gradients will haunt valve in valve for many years to come," said Dr. Dvir. "However, we know from the aortic arena, that even if you have a high residual gradient, the results at 1 year are not different than the other groups. So, it doesn’t affect 1-year outcomes, although I predict that it will affect the durability of the device and the long-term outcome."

TORONTO – Transcatheter mitral valve-in-valve/valve-in-ring implantation is an effective option in most patients with degenerative mitral valves after surgical intervention, according to Dr. Danny Dvir.

In the highest-risk patients (STS scores greater than 20%), however, mortality after this percutaneous technique approached 50% at 1 year, Dr. Dvir reported at the American Association for Thoracic Surgery annual meeting.

"Most of these procedures were clinically effective, with 1-year results that are comparable to native aortic valve transcatheter implantation," said Dr. Dvir of St. Paul’s Hospital, Vancouver, B.C.

"However, we saw safety and efficacy concerns that included relatively high in-hospital mortality, device malposition, sporadic cases with elevated LVOT [left ventricular outflow tract] gradient, and elevated postprocedural gradients in procedures performed inside small surgical valves."

Dr. Joseph E. Bavaria, professor of surgery and director of the thoracic aortic surgery program at the University of Pennsylvania Health System, Philadelphia, the scheduled discussant of the abstract, asked, referring to the STS score greater than 20% group: "Is this really a cohort C or futile situation, where [the intervention] is probably not better than medical management?"

Transcatheter VIV/VIR

The majority of prosthetic heart valves are tissue valves that tend to degenerate and eventually fail over time. Surgical replacement of these valves is associated with significant risks.

Transcatheter valve-in-valve/valve-in-ring (VIV/VIR) implantation within a failed bioprosthesis performed in the mitral position has been reported in only small case series, with the safety and efficacy of the technique being largely unknown. The Global Valve-in-Valve Registry, with 765 patients from 73 centers worldwide, was designed to track the efficacy and safety of this nonsurgical redo in either the aortic or mitral valve position. The registry included 190 patients from 23 centers who underwent percutaneous mitral valve procedures (VIV in 157 patients and VIR in 33).

The mechanism of failure was regurgitation in 37%, stenosis in 25%, and a combination of the two in 38%. Patients were a median of 9 years past their last cardiac surgery and were deemed at high risk, with a mean STS score of 14.4. The mean age was 73.6 years, and 65.2% were female. The median length of hospital stay was 8 days.

A SAPIEN or SAPIEN XT device was used in the majority of cases (93.7%), mostly deployed via transapical access (84.7%).

During the procedure, initial device malpositioning occurred in 5.3% of cases, with a second device implanted in 4.3%. Postimplantation valvuloplasty was utilized in 8%.

"We saw a high rate of malpositioning," Dr. Dvir said. "In some of the cases, this led to a decision to leave the malpositioned device in the left atrium and implant a second device."

The incidence of all-cause death at 30 days was 8.9%, with cardiovascular death seen in 6.8% and major stroke in 2.2%. Importantly, 85.8% of patients were in NYHA function class I or II at 30 days, Dr. Dvir said.

"We were very happy to see this," he commented. "If you survive the valve-in-valve or valve-in-ring procedure, you will most likely have a good functional class.

Overall, the Kaplan-Meier mortality rate was 22.3% at 1 year. When stratified according to STS score, however, mortality in the group of patients with an STS score greater than 20% was 45.6%, falling to 23.2% for an STS score of 10%-20% and 12.9% for an STS score less than 10%. Predictors of 1-year mortality included STS score and renal failure.

"In these patients with STS greater than 20%, there is really a big problem. We need to understand which patients should not undergo a procedure, and maybe should be treated with an alternative approach," Dr. Dvir said. He noted that 28% of cases in the registry were in this highest-risk category.

"An 8.9% mortality at 1 year is not so bad in a group with a mean STS score of 14.1," noted Dr. Bavaria. "[But] this is not really a benign procedure as the length of stay was over 8 days."

Surgical valve sizing gone wrong?

After the VIV/VIR procedure, the mean mitral valve area was 2.1 ± 0.7 cm², with a mean residual gradient of 6.2 mm Hg. Significant mitral regurgitation was noted in 4.2% of patients, and a left ventricular outflow tract mean gradient greater than 20 mm Hg was noted in only 2.1% of cases.

"We saw a trend for elevated gradients especially when we were treating small valves with a valve-in-valve approach," Dr. Dvir said.

Indeed, in the patients with smaller surgical valve label size, one-third had postprocedural mean gradients greater than 10 mm Hg. The rates of mean gradient greater than 10 mm Hg dropped to 7.4% and 5.3% with progressively larger surgical valve label sizes.

"I am very worried about the cardiac surgery and interventional cardiology decision-making process, as over 60% of valve-in-ring cases were 26-mm and 28-mm rings and over 56% of valve-in-valve cases were less than or equal to 27-mm prosthetic valves," said Dr. Bavaria. "There will obviously be significant mitral gradients in these cases, as this study showed. Is this simply a lack of knowledge in the operators, should these small valve-in-valve cases be done, or is this a limit of this technology?"

"I think this issue of residual gradients will haunt valve in valve for many years to come," said Dr. Dvir. "However, we know from the aortic arena, that even if you have a high residual gradient, the results at 1 year are not different than the other groups. So, it doesn’t affect 1-year outcomes, although I predict that it will affect the durability of the device and the long-term outcome."

TORONTO – Mitral valve repair rates continue to show great variability, ranging from zero to 90% for patients with moderate to severe mitral regurgitation, according to data presented by Dr. Damien J. LaPar, a thoracic surgeon at the University of Virginia, Charlottesville.

Surgeons performing more than 20 mitral cases per year were about three times more likely to perform a repair over a replacement compared with surgeons performing fewer than 20 mitral surgeries per year.

"Average annual surgeon volume appears more significantly associated with an increased likelihood for mitral repair than institutional volume," Dr. LaPar reported at the annual meeting of the American Association for Thoracic Surgery.

Mitral valve repair appears to be underused despite accumulated data favoring repair over replacement in the treatment of patients with moderate to severe mitral regurgitation. Dr. LaPar and colleagues evaluated the relationship between procedure volume and propensity for mitral repair (over replacement) in a multi-institution, regional cohort of patients.

The Virginia Cardiac Surgery Quality Initiative is a voluntary consortium of 17 collaborating cardiac surgery centers in Virginia that captures about 99% of all the operations performed in the state, with each center contributing their data to the Society of Thoracic Surgeons (STS) Adult Cardiac Surgery Database.

Records for 4,178 patients were evaluated from 2001-2012; 2,516 of these patients underwent isolated mitral valve replacement and 1,662 underwent isolated mitral valve repair. To deal with issues of confounding, a propensity matched cohort was developed including 1,661 patients in each group.

Median annual mitral procedure volume was 54 cases per year for hospitals, ranging from 5 to 128, and 13 cases per year for individual surgeons, ranging from zero to 58 cases. Mitral valve repair rates also ranged significantly, from 35% to 70% of all mitral valve procedures for hospitals and from zero to 90% for surgeons.

As expected, mitral valve replacement patients presented with higher STS PROM (5.6% vs. 1.7%, P less than .001), higher age, and a greater number of comorbidities.

When the researchers adjusted for STS PROM, both average annual hospital (P = .04) and surgeon (P less than .0001) mitral procedure volume were associated with probability of mitral repair, with surgeon volume showing more influence in the decision to repair rather than replace the mitral valve.

When the propensity to perform mitral valve repair was plotted against surgeon volume, the researchers saw a clear "inflection point" of increased probability for repair over replacement when the average annual surgeon mitral volume (used as a surrogate for surgeon experience) exceeded 20 operations per year. This finding was validated in the propensity matched cohort.

"We did a post hoc analysis and demonstrated that when you categorize volume by our visible threshold of 20 cases per year, for both hospital and surgeon volume in the overall cohort, hospitals and surgeons performing more than 20 cases per year, more strongly performed mitral valve repair," reported Dr. LaPar.

Indeed, among surgeons and hospitals performing more than 20 mitral operations annually, repairs were done in 62% of cases, compared to 37% for lower-volume centers. Among surgeons performing more than 20 mitral cases per year, repair rates were 73% compared to 26% for lower-volume mitral operators (both P less than .001).

Similarly significant differences were seen in the propensity matched cohort, although the effect was more muted (53% vs. 47% for high- and low-volume hospitals and 67% vs. 39% for high- and low-volume surgeons; P less than .001 for both).

"In the upcoming era of percutaneous mitral valve repair technology, we believe that surgeon volume and expertise should act as a gatekeeper and should dictate not only access to this technology but the role of surgeons and cardiologists in this," concluded Dr. LaPar.

"This will be an excellent contribution to the literature further defining the relationship between surgeon volume and hospital center volume in mitral valve surgical outcomes and choice of operation," said the invited discussant, Dr. David A Fullerton, professor of surgery and cardiothoracic surgery at the University of Colorado at Denver, Aurora.

Dr. Daniel Drake, the current head of the Michigan Society of Thoracic and Cardiovascular Surgeons Mitral Initiative, added that in his state they have started to see a greater number of referrals – based on an assessment of case complexity – from less experienced surgeons to more experienced surgeons.

Dr. LaPar agreed that this would be a good solution, but noted that he hadn’t yet seen this happening widely in Virginia.

Dr. LaPar and Dr. Fullerton reported having no financial disclosures.

TORONTO – Mitral valve repair rates continue to show great variability, ranging from zero to 90% for patients with moderate to severe mitral regurgitation, according to data presented by Dr. Damien J. LaPar, a thoracic surgeon at the University of Virginia, Charlottesville.

Surgeons performing more than 20 mitral cases per year were about three times more likely to perform a repair over a replacement compared with surgeons performing fewer than 20 mitral surgeries per year.

"Average annual surgeon volume appears more significantly associated with an increased likelihood for mitral repair than institutional volume," Dr. LaPar reported at the annual meeting of the American Association for Thoracic Surgery.

Mitral valve repair appears to be underused despite accumulated data favoring repair over replacement in the treatment of patients with moderate to severe mitral regurgitation. Dr. LaPar and colleagues evaluated the relationship between procedure volume and propensity for mitral repair (over replacement) in a multi-institution, regional cohort of patients.

The Virginia Cardiac Surgery Quality Initiative is a voluntary consortium of 17 collaborating cardiac surgery centers in Virginia that captures about 99% of all the operations performed in the state, with each center contributing their data to the Society of Thoracic Surgeons (STS) Adult Cardiac Surgery Database.

Records for 4,178 patients were evaluated from 2001-2012; 2,516 of these patients underwent isolated mitral valve replacement and 1,662 underwent isolated mitral valve repair. To deal with issues of confounding, a propensity matched cohort was developed including 1,661 patients in each group.

Median annual mitral procedure volume was 54 cases per year for hospitals, ranging from 5 to 128, and 13 cases per year for individual surgeons, ranging from zero to 58 cases. Mitral valve repair rates also ranged significantly, from 35% to 70% of all mitral valve procedures for hospitals and from zero to 90% for surgeons.

As expected, mitral valve replacement patients presented with higher STS PROM (5.6% vs. 1.7%, P less than .001), higher age, and a greater number of comorbidities.

When the researchers adjusted for STS PROM, both average annual hospital (P = .04) and surgeon (P less than .0001) mitral procedure volume were associated with probability of mitral repair, with surgeon volume showing more influence in the decision to repair rather than replace the mitral valve.

When the propensity to perform mitral valve repair was plotted against surgeon volume, the researchers saw a clear "inflection point" of increased probability for repair over replacement when the average annual surgeon mitral volume (used as a surrogate for surgeon experience) exceeded 20 operations per year. This finding was validated in the propensity matched cohort.

"We did a post hoc analysis and demonstrated that when you categorize volume by our visible threshold of 20 cases per year, for both hospital and surgeon volume in the overall cohort, hospitals and surgeons performing more than 20 cases per year, more strongly performed mitral valve repair," reported Dr. LaPar.

Indeed, among surgeons and hospitals performing more than 20 mitral operations annually, repairs were done in 62% of cases, compared to 37% for lower-volume centers. Among surgeons performing more than 20 mitral cases per year, repair rates were 73% compared to 26% for lower-volume mitral operators (both P less than .001).

Similarly significant differences were seen in the propensity matched cohort, although the effect was more muted (53% vs. 47% for high- and low-volume hospitals and 67% vs. 39% for high- and low-volume surgeons; P less than .001 for both).

"In the upcoming era of percutaneous mitral valve repair technology, we believe that surgeon volume and expertise should act as a gatekeeper and should dictate not only access to this technology but the role of surgeons and cardiologists in this," concluded Dr. LaPar.

"This will be an excellent contribution to the literature further defining the relationship between surgeon volume and hospital center volume in mitral valve surgical outcomes and choice of operation," said the invited discussant, Dr. David A Fullerton, professor of surgery and cardiothoracic surgery at the University of Colorado at Denver, Aurora.

Dr. Daniel Drake, the current head of the Michigan Society of Thoracic and Cardiovascular Surgeons Mitral Initiative, added that in his state they have started to see a greater number of referrals – based on an assessment of case complexity – from less experienced surgeons to more experienced surgeons.

Dr. LaPar agreed that this would be a good solution, but noted that he hadn’t yet seen this happening widely in Virginia.

Dr. LaPar and Dr. Fullerton reported having no financial disclosures.

TORONTO – Mitral valve repair rates continue to show great variability, ranging from zero to 90% for patients with moderate to severe mitral regurgitation, according to data presented by Dr. Damien J. LaPar, a thoracic surgeon at the University of Virginia, Charlottesville.

Surgeons performing more than 20 mitral cases per year were about three times more likely to perform a repair over a replacement compared with surgeons performing fewer than 20 mitral surgeries per year.

"Average annual surgeon volume appears more significantly associated with an increased likelihood for mitral repair than institutional volume," Dr. LaPar reported at the annual meeting of the American Association for Thoracic Surgery.

Mitral valve repair appears to be underused despite accumulated data favoring repair over replacement in the treatment of patients with moderate to severe mitral regurgitation. Dr. LaPar and colleagues evaluated the relationship between procedure volume and propensity for mitral repair (over replacement) in a multi-institution, regional cohort of patients.

The Virginia Cardiac Surgery Quality Initiative is a voluntary consortium of 17 collaborating cardiac surgery centers in Virginia that captures about 99% of all the operations performed in the state, with each center contributing their data to the Society of Thoracic Surgeons (STS) Adult Cardiac Surgery Database.

Records for 4,178 patients were evaluated from 2001-2012; 2,516 of these patients underwent isolated mitral valve replacement and 1,662 underwent isolated mitral valve repair. To deal with issues of confounding, a propensity matched cohort was developed including 1,661 patients in each group.

Median annual mitral procedure volume was 54 cases per year for hospitals, ranging from 5 to 128, and 13 cases per year for individual surgeons, ranging from zero to 58 cases. Mitral valve repair rates also ranged significantly, from 35% to 70% of all mitral valve procedures for hospitals and from zero to 90% for surgeons.

As expected, mitral valve replacement patients presented with higher STS PROM (5.6% vs. 1.7%, P less than .001), higher age, and a greater number of comorbidities.

When the researchers adjusted for STS PROM, both average annual hospital (P = .04) and surgeon (P less than .0001) mitral procedure volume were associated with probability of mitral repair, with surgeon volume showing more influence in the decision to repair rather than replace the mitral valve.

When the propensity to perform mitral valve repair was plotted against surgeon volume, the researchers saw a clear "inflection point" of increased probability for repair over replacement when the average annual surgeon mitral volume (used as a surrogate for surgeon experience) exceeded 20 operations per year. This finding was validated in the propensity matched cohort.

"We did a post hoc analysis and demonstrated that when you categorize volume by our visible threshold of 20 cases per year, for both hospital and surgeon volume in the overall cohort, hospitals and surgeons performing more than 20 cases per year, more strongly performed mitral valve repair," reported Dr. LaPar.

Indeed, among surgeons and hospitals performing more than 20 mitral operations annually, repairs were done in 62% of cases, compared to 37% for lower-volume centers. Among surgeons performing more than 20 mitral cases per year, repair rates were 73% compared to 26% for lower-volume mitral operators (both P less than .001).

Similarly significant differences were seen in the propensity matched cohort, although the effect was more muted (53% vs. 47% for high- and low-volume hospitals and 67% vs. 39% for high- and low-volume surgeons; P less than .001 for both).

"In the upcoming era of percutaneous mitral valve repair technology, we believe that surgeon volume and expertise should act as a gatekeeper and should dictate not only access to this technology but the role of surgeons and cardiologists in this," concluded Dr. LaPar.

"This will be an excellent contribution to the literature further defining the relationship between surgeon volume and hospital center volume in mitral valve surgical outcomes and choice of operation," said the invited discussant, Dr. David A Fullerton, professor of surgery and cardiothoracic surgery at the University of Colorado at Denver, Aurora.

Dr. Daniel Drake, the current head of the Michigan Society of Thoracic and Cardiovascular Surgeons Mitral Initiative, added that in his state they have started to see a greater number of referrals – based on an assessment of case complexity – from less experienced surgeons to more experienced surgeons.

Dr. LaPar agreed that this would be a good solution, but noted that he hadn’t yet seen this happening widely in Virginia.

Dr. LaPar and Dr. Fullerton reported having no financial disclosures.

TORONTO – Routine EEG monitoring after surgery with cardiopulmonary bypass in neonates revealed a seizure incidence of 8% in a recent study. In most cases (85%), seizure activity was detectable only on EEG and would not have been identified or treated without EEG monitoring, reported Dr. Maryam Y. Naim, a critical care pediatrician at the Children’s Hospital of Philadelphia, during the American Association for Thoracic Surgery Annual Meeting.

Of concern, status epilepticus was noted in 62% of neonates with seizures, and mortality was higher in babies with seizures versus those without (38% vs. 3%; P less than .01).

"Postoperative seizures are associated with worse neurological outcomes," said Dr. Naim. In addition to being a biomarker of underlying brain injury, there is some evidence that the seizures themselves may cause secondary brain injury.

The American Clinical Neurophysiology Society (ACNS) recommends routine EEG monitoring following infant cardiac surgery. The CHOP group implemented this recommendation and performed a quality improvement study to determine the incidence of seizures in neonates following surgery with cardiopulmonary bypass (CBP).

A total of 161 neonates had 48-hours of EEG monitoring initiated within 6 hours of cardiac surgery with CPB. The median gestational age of the cohort was 39 weeks, 16% were premature, and 13% had identified genetic defects. The median age at the time of surgery was 5 days. Deep hypothermic circulatory arrest was used in 48% of surgeries (median time, 48 minutes), 16% had open chest with delayed sternal closure, and 9% had a cardiac arrest.

Seizures were detected in 13 (8%), with a median onset of 20 hours after return to the cardiac ICU (CICU) from surgery. Seizures were subclinical, or EEG only, in 11 patients (85%), electroclinical in the other 2 (15%), and status epilepticus was seen in 8 (62%).

When seizures occurred, the CICU team was notified by the EEG technologist, neurology consult was obtained, and the patient was treated with antiseizure medications. As well, abnormal vital signs or movements suggestive of seizure activity were noted at the bedside; although such events were recorded in 32 patients (22%), none had EEG correlates consistent with seizure activity.

"Neonates with all types of heart disease had seizures ... " reported Dr. Naim. " ... with a highest percentage occurring in those with single ventricles and arch obstruction."

Neuroimaging studies were reviewed by a neurologist to determine any association between injury and seizure location. The most commonly detected lesion was periventricular leukomalacia, and four patients had seizures arising from one of their injury sites.

Although neonates with and without seizures had similar CICU lengths of stay, mortality was higher in those with seizures (38% vs. 3%; P less than .01). No predictors of seizures were identified on multivariable analysis.

"Based on these data, we are continuing routine postoperative EEG monitoring in all neonates following surgery with cardiopulmonary bypass," said Dr. Naim.

While not discredited in any way, Dr. Naim’s data were met with a fair amount of pushback from the gathered group of pediatric cardiothoracic surgical experts. An informal poll of the audience showed that 80%-90% do not routinely monitor neonates for seizures post-CPB despite ACNS recommendations, and several of the comments questioned the technical and financial feasibility of routine EEG monitoring.

Said the invited discussant, Dr. Frank Pigula, a cardiac surgeon from Boston Children’s Hospital, "In all the groups that have studied this, in all patients who have had a seizure, there were documented brain abnormalities. So, the ways I picture these data are that a seizure is a sign of an underlying brain injury, much like a fever is a sign of an underlying infection."

"And you’ve made a good case for the routine postoperative surveillance for seizures; I’m sure everyone would agree that treating seizures is a good thing, but is there any evidence showing us that the early identification and treatment of seizures improves outcomes, either developmental delays or mortality?

In response, Dr. Naim noted that all the CHOP seizure sufferers were treated, whether they had EEG only or clinical seizures, and, at the 4-year mark, they are showing fewer neurodevelopmental issues than are previous cohorts of untreated neonates.

"I think one thing that is very concerning is emerging evidence that the seizures themselves cause secondary brain injury," she said. "So I think that the way to move forward, since our emphasis has really changed not to improving neurodevelopmental outcomes in our patients, is to monitor them postoperatively."

Dr. Naim reported having no financial disclosures. Dr. Pigula, the invited discussant, also reported no financial disclosures.

TORONTO – Routine EEG monitoring after surgery with cardiopulmonary bypass in neonates revealed a seizure incidence of 8% in a recent study. In most cases (85%), seizure activity was detectable only on EEG and would not have been identified or treated without EEG monitoring, reported Dr. Maryam Y. Naim, a critical care pediatrician at the Children’s Hospital of Philadelphia, during the American Association for Thoracic Surgery Annual Meeting.

Of concern, status epilepticus was noted in 62% of neonates with seizures, and mortality was higher in babies with seizures versus those without (38% vs. 3%; P less than .01).

"Postoperative seizures are associated with worse neurological outcomes," said Dr. Naim. In addition to being a biomarker of underlying brain injury, there is some evidence that the seizures themselves may cause secondary brain injury.

The American Clinical Neurophysiology Society (ACNS) recommends routine EEG monitoring following infant cardiac surgery. The CHOP group implemented this recommendation and performed a quality improvement study to determine the incidence of seizures in neonates following surgery with cardiopulmonary bypass (CBP).

A total of 161 neonates had 48-hours of EEG monitoring initiated within 6 hours of cardiac surgery with CPB. The median gestational age of the cohort was 39 weeks, 16% were premature, and 13% had identified genetic defects. The median age at the time of surgery was 5 days. Deep hypothermic circulatory arrest was used in 48% of surgeries (median time, 48 minutes), 16% had open chest with delayed sternal closure, and 9% had a cardiac arrest.

Seizures were detected in 13 (8%), with a median onset of 20 hours after return to the cardiac ICU (CICU) from surgery. Seizures were subclinical, or EEG only, in 11 patients (85%), electroclinical in the other 2 (15%), and status epilepticus was seen in 8 (62%).

When seizures occurred, the CICU team was notified by the EEG technologist, neurology consult was obtained, and the patient was treated with antiseizure medications. As well, abnormal vital signs or movements suggestive of seizure activity were noted at the bedside; although such events were recorded in 32 patients (22%), none had EEG correlates consistent with seizure activity.

"Neonates with all types of heart disease had seizures ... " reported Dr. Naim. " ... with a highest percentage occurring in those with single ventricles and arch obstruction."

Neuroimaging studies were reviewed by a neurologist to determine any association between injury and seizure location. The most commonly detected lesion was periventricular leukomalacia, and four patients had seizures arising from one of their injury sites.

Although neonates with and without seizures had similar CICU lengths of stay, mortality was higher in those with seizures (38% vs. 3%; P less than .01). No predictors of seizures were identified on multivariable analysis.

"Based on these data, we are continuing routine postoperative EEG monitoring in all neonates following surgery with cardiopulmonary bypass," said Dr. Naim.

While not discredited in any way, Dr. Naim’s data were met with a fair amount of pushback from the gathered group of pediatric cardiothoracic surgical experts. An informal poll of the audience showed that 80%-90% do not routinely monitor neonates for seizures post-CPB despite ACNS recommendations, and several of the comments questioned the technical and financial feasibility of routine EEG monitoring.

Said the invited discussant, Dr. Frank Pigula, a cardiac surgeon from Boston Children’s Hospital, "In all the groups that have studied this, in all patients who have had a seizure, there were documented brain abnormalities. So, the ways I picture these data are that a seizure is a sign of an underlying brain injury, much like a fever is a sign of an underlying infection."

"And you’ve made a good case for the routine postoperative surveillance for seizures; I’m sure everyone would agree that treating seizures is a good thing, but is there any evidence showing us that the early identification and treatment of seizures improves outcomes, either developmental delays or mortality?

In response, Dr. Naim noted that all the CHOP seizure sufferers were treated, whether they had EEG only or clinical seizures, and, at the 4-year mark, they are showing fewer neurodevelopmental issues than are previous cohorts of untreated neonates.

"I think one thing that is very concerning is emerging evidence that the seizures themselves cause secondary brain injury," she said. "So I think that the way to move forward, since our emphasis has really changed not to improving neurodevelopmental outcomes in our patients, is to monitor them postoperatively."

Dr. Naim reported having no financial disclosures. Dr. Pigula, the invited discussant, also reported no financial disclosures.

TORONTO – Routine EEG monitoring after surgery with cardiopulmonary bypass in neonates revealed a seizure incidence of 8% in a recent study. In most cases (85%), seizure activity was detectable only on EEG and would not have been identified or treated without EEG monitoring, reported Dr. Maryam Y. Naim, a critical care pediatrician at the Children’s Hospital of Philadelphia, during the American Association for Thoracic Surgery Annual Meeting.

Of concern, status epilepticus was noted in 62% of neonates with seizures, and mortality was higher in babies with seizures versus those without (38% vs. 3%; P less than .01).

"Postoperative seizures are associated with worse neurological outcomes," said Dr. Naim. In addition to being a biomarker of underlying brain injury, there is some evidence that the seizures themselves may cause secondary brain injury.

The American Clinical Neurophysiology Society (ACNS) recommends routine EEG monitoring following infant cardiac surgery. The CHOP group implemented this recommendation and performed a quality improvement study to determine the incidence of seizures in neonates following surgery with cardiopulmonary bypass (CBP).

A total of 161 neonates had 48-hours of EEG monitoring initiated within 6 hours of cardiac surgery with CPB. The median gestational age of the cohort was 39 weeks, 16% were premature, and 13% had identified genetic defects. The median age at the time of surgery was 5 days. Deep hypothermic circulatory arrest was used in 48% of surgeries (median time, 48 minutes), 16% had open chest with delayed sternal closure, and 9% had a cardiac arrest.

Seizures were detected in 13 (8%), with a median onset of 20 hours after return to the cardiac ICU (CICU) from surgery. Seizures were subclinical, or EEG only, in 11 patients (85%), electroclinical in the other 2 (15%), and status epilepticus was seen in 8 (62%).

When seizures occurred, the CICU team was notified by the EEG technologist, neurology consult was obtained, and the patient was treated with antiseizure medications. As well, abnormal vital signs or movements suggestive of seizure activity were noted at the bedside; although such events were recorded in 32 patients (22%), none had EEG correlates consistent with seizure activity.

"Neonates with all types of heart disease had seizures ... " reported Dr. Naim. " ... with a highest percentage occurring in those with single ventricles and arch obstruction."

Neuroimaging studies were reviewed by a neurologist to determine any association between injury and seizure location. The most commonly detected lesion was periventricular leukomalacia, and four patients had seizures arising from one of their injury sites.

Although neonates with and without seizures had similar CICU lengths of stay, mortality was higher in those with seizures (38% vs. 3%; P less than .01). No predictors of seizures were identified on multivariable analysis.

"Based on these data, we are continuing routine postoperative EEG monitoring in all neonates following surgery with cardiopulmonary bypass," said Dr. Naim.

While not discredited in any way, Dr. Naim’s data were met with a fair amount of pushback from the gathered group of pediatric cardiothoracic surgical experts. An informal poll of the audience showed that 80%-90% do not routinely monitor neonates for seizures post-CPB despite ACNS recommendations, and several of the comments questioned the technical and financial feasibility of routine EEG monitoring.

Said the invited discussant, Dr. Frank Pigula, a cardiac surgeon from Boston Children’s Hospital, "In all the groups that have studied this, in all patients who have had a seizure, there were documented brain abnormalities. So, the ways I picture these data are that a seizure is a sign of an underlying brain injury, much like a fever is a sign of an underlying infection."

"And you’ve made a good case for the routine postoperative surveillance for seizures; I’m sure everyone would agree that treating seizures is a good thing, but is there any evidence showing us that the early identification and treatment of seizures improves outcomes, either developmental delays or mortality?

In response, Dr. Naim noted that all the CHOP seizure sufferers were treated, whether they had EEG only or clinical seizures, and, at the 4-year mark, they are showing fewer neurodevelopmental issues than are previous cohorts of untreated neonates.

"I think one thing that is very concerning is emerging evidence that the seizures themselves cause secondary brain injury," she said. "So I think that the way to move forward, since our emphasis has really changed not to improving neurodevelopmental outcomes in our patients, is to monitor them postoperatively."

Dr. Naim reported having no financial disclosures. Dr. Pigula, the invited discussant, also reported no financial disclosures.

MINNEAPOLIS – Sleep apnea is associated with neurocognitive dysfunction in adult Hispanics living in the United States, research presented at the annual meeting of the Associated Professional Sleep Societies shows.

In unadjusted models, sleep apnea, as assessed by the apnea-hypopnea index (AHI), was inversely associated with neurocognitive dysfunction in both men and women.

"However, after adjusting for covariates, these interactions were attenuated," reported Dr. Alberto R. Ramos. "In a fully adjusted model accounting for age, BMI [body mass index], tobacco use, depression and anxiety scores, stroke, diabetes, hypertension, and field center tested, this association was seen only in women, but not in men."

Dr. Ramos of the department of clinical neurology at the University of Miami and his colleagues conducted a cross-sectional analysis of 9,714 Hispanic men and women between the ages of 45 and 74 who were participants in the Hispanic Community Health Study/Study of Latinos (HCHS/SOL). The researchers sought to determine the extent to which sleep apnea contributes to neurocognitive dysfunction in a representative sample of Hispanic men and women residing in the United States.

Subjects were administered multiple neurocognitive tests, assessing learning, recall, word fluency, and digit symbol substitution. Obstructive sleep apnea was determined objectively using the apnea risk evaluation system and defined by the apnea-hypopnea index (AHI), and subjectively using standardized sleep symptom scales.

The mean AHI was 8.9, 11.5 for men and 6.8 for women (P less than 0.001 for difference between genders). With increasing age, the AHI significantly increased, too, from 7.4 in subjects aged 45 to 54, to 11.5 in those aged 65 to 74 (P less than 0.001).

Dementia prevalence is on the rise, particularly among Hispanics/Latinos who are up to 3.3 times more likely to meet diagnostic criteria for advanced dementia, compared with non-Hispanic whites, Dr. Ramos reported.

Dr. Ramos reported having no disclosures. The study received support from multiple National Heart, Lung, and Blood Institute grants.

MINNEAPOLIS – Sleep apnea is associated with neurocognitive dysfunction in adult Hispanics living in the United States, research presented at the annual meeting of the Associated Professional Sleep Societies shows.

In unadjusted models, sleep apnea, as assessed by the apnea-hypopnea index (AHI), was inversely associated with neurocognitive dysfunction in both men and women.

"However, after adjusting for covariates, these interactions were attenuated," reported Dr. Alberto R. Ramos. "In a fully adjusted model accounting for age, BMI [body mass index], tobacco use, depression and anxiety scores, stroke, diabetes, hypertension, and field center tested, this association was seen only in women, but not in men."

Dr. Ramos of the department of clinical neurology at the University of Miami and his colleagues conducted a cross-sectional analysis of 9,714 Hispanic men and women between the ages of 45 and 74 who were participants in the Hispanic Community Health Study/Study of Latinos (HCHS/SOL). The researchers sought to determine the extent to which sleep apnea contributes to neurocognitive dysfunction in a representative sample of Hispanic men and women residing in the United States.

Subjects were administered multiple neurocognitive tests, assessing learning, recall, word fluency, and digit symbol substitution. Obstructive sleep apnea was determined objectively using the apnea risk evaluation system and defined by the apnea-hypopnea index (AHI), and subjectively using standardized sleep symptom scales.

The mean AHI was 8.9, 11.5 for men and 6.8 for women (P less than 0.001 for difference between genders). With increasing age, the AHI significantly increased, too, from 7.4 in subjects aged 45 to 54, to 11.5 in those aged 65 to 74 (P less than 0.001).

Dementia prevalence is on the rise, particularly among Hispanics/Latinos who are up to 3.3 times more likely to meet diagnostic criteria for advanced dementia, compared with non-Hispanic whites, Dr. Ramos reported.

Dr. Ramos reported having no disclosures. The study received support from multiple National Heart, Lung, and Blood Institute grants.

MINNEAPOLIS – Sleep apnea is associated with neurocognitive dysfunction in adult Hispanics living in the United States, research presented at the annual meeting of the Associated Professional Sleep Societies shows.

In unadjusted models, sleep apnea, as assessed by the apnea-hypopnea index (AHI), was inversely associated with neurocognitive dysfunction in both men and women.

"However, after adjusting for covariates, these interactions were attenuated," reported Dr. Alberto R. Ramos. "In a fully adjusted model accounting for age, BMI [body mass index], tobacco use, depression and anxiety scores, stroke, diabetes, hypertension, and field center tested, this association was seen only in women, but not in men."

Dr. Ramos of the department of clinical neurology at the University of Miami and his colleagues conducted a cross-sectional analysis of 9,714 Hispanic men and women between the ages of 45 and 74 who were participants in the Hispanic Community Health Study/Study of Latinos (HCHS/SOL). The researchers sought to determine the extent to which sleep apnea contributes to neurocognitive dysfunction in a representative sample of Hispanic men and women residing in the United States.

Subjects were administered multiple neurocognitive tests, assessing learning, recall, word fluency, and digit symbol substitution. Obstructive sleep apnea was determined objectively using the apnea risk evaluation system and defined by the apnea-hypopnea index (AHI), and subjectively using standardized sleep symptom scales.

The mean AHI was 8.9, 11.5 for men and 6.8 for women (P less than 0.001 for difference between genders). With increasing age, the AHI significantly increased, too, from 7.4 in subjects aged 45 to 54, to 11.5 in those aged 65 to 74 (P less than 0.001).

Dementia prevalence is on the rise, particularly among Hispanics/Latinos who are up to 3.3 times more likely to meet diagnostic criteria for advanced dementia, compared with non-Hispanic whites, Dr. Ramos reported.

Dr. Ramos reported having no disclosures. The study received support from multiple National Heart, Lung, and Blood Institute grants.

MINNEAPOLIS – Using the adult treatment threshold for serum ferritin to guide treatment in children with restless sleep may lead to inappropriate iron supplementation.

In both adults and children, iron deficiency has been linked to the presence and severity of restless legs syndrome (RLS) and periodic limb movements of sleep (PLMS). For adults, a serum ferritin less than 50 mcg/L is the threshold commonly used to guide iron supplementation for patients with RLS or PLMS. A threshold of 40-50 mcg/L also has been used in pediatric studies.

For children, however, a far lower level of 20 mcg/L appears more appropriate, Alyson Connor reported at the annual meeting of the Associated Professional Sleep Societies.

"Our findings raise a question about the best serum ferritin threshold to use when seeing a child with evidence of sleep restlessness, in particular with elevated" periodic limb movement index, Ms. Connor said. "While many children may benefit from iron supplementation for these disorders, this area is worthy of further study as investigation in animals suggests that treatment of iron deficiency with high-dose iron supplementation in early development may differentially affect the development of the brain."

Ms. Connor, a student in the biomedical research program at the University of Michigan, Ann Arbor, and her colleagues conducted a review of 537 children aged 12 months to 18 years who were referred to the University of Michigan pediatric sleep clinic. In this cohort (mean age was 8.9 years; 62% were male), a median serum ferritin level of 23.6 mcg/L was linked with polysomnographic measures of periodic limb movement index (PLMI) per hour of total sleep time of 5 or more, while a median level of 30 mcg/L was linked with PLMI less than 5. About 26% of the subjects had a PLMI of 5 or more.

"There was a significant association between serum ferritin and periodic limb movement index such that, for every 10-mcg/L increase in serum ferritin, it decreased the odds of having an elevated PLMI by 11%," Ms. Connor said. Boys, younger patients, those with lower serum ferritin levels, and those with a shorter time between hematology and polysomnography had significantly increased odds of a PLMI of 5 or more. Only 19% of patients had a serum ferritin above 50 mcg/L, while 50% had below 30 mcg/L.

"The cutoff of 50 is very sensitive, giving few false negatives, but not very specific, whereas a lower cutoff improves specificity," Ms. Connor said.

Studies are needed to assess the link between iron status and sleep measures in a general pediatric population and to determine the best iron dose, timing, and method of delivery, she said.

MINNEAPOLIS – Using the adult treatment threshold for serum ferritin to guide treatment in children with restless sleep may lead to inappropriate iron supplementation.

In both adults and children, iron deficiency has been linked to the presence and severity of restless legs syndrome (RLS) and periodic limb movements of sleep (PLMS). For adults, a serum ferritin less than 50 mcg/L is the threshold commonly used to guide iron supplementation for patients with RLS or PLMS. A threshold of 40-50 mcg/L also has been used in pediatric studies.

For children, however, a far lower level of 20 mcg/L appears more appropriate, Alyson Connor reported at the annual meeting of the Associated Professional Sleep Societies.

"Our findings raise a question about the best serum ferritin threshold to use when seeing a child with evidence of sleep restlessness, in particular with elevated" periodic limb movement index, Ms. Connor said. "While many children may benefit from iron supplementation for these disorders, this area is worthy of further study as investigation in animals suggests that treatment of iron deficiency with high-dose iron supplementation in early development may differentially affect the development of the brain."

Ms. Connor, a student in the biomedical research program at the University of Michigan, Ann Arbor, and her colleagues conducted a review of 537 children aged 12 months to 18 years who were referred to the University of Michigan pediatric sleep clinic. In this cohort (mean age was 8.9 years; 62% were male), a median serum ferritin level of 23.6 mcg/L was linked with polysomnographic measures of periodic limb movement index (PLMI) per hour of total sleep time of 5 or more, while a median level of 30 mcg/L was linked with PLMI less than 5. About 26% of the subjects had a PLMI of 5 or more.

"There was a significant association between serum ferritin and periodic limb movement index such that, for every 10-mcg/L increase in serum ferritin, it decreased the odds of having an elevated PLMI by 11%," Ms. Connor said. Boys, younger patients, those with lower serum ferritin levels, and those with a shorter time between hematology and polysomnography had significantly increased odds of a PLMI of 5 or more. Only 19% of patients had a serum ferritin above 50 mcg/L, while 50% had below 30 mcg/L.

"The cutoff of 50 is very sensitive, giving few false negatives, but not very specific, whereas a lower cutoff improves specificity," Ms. Connor said.

Studies are needed to assess the link between iron status and sleep measures in a general pediatric population and to determine the best iron dose, timing, and method of delivery, she said.

MINNEAPOLIS – Using the adult treatment threshold for serum ferritin to guide treatment in children with restless sleep may lead to inappropriate iron supplementation.

In both adults and children, iron deficiency has been linked to the presence and severity of restless legs syndrome (RLS) and periodic limb movements of sleep (PLMS). For adults, a serum ferritin less than 50 mcg/L is the threshold commonly used to guide iron supplementation for patients with RLS or PLMS. A threshold of 40-50 mcg/L also has been used in pediatric studies.

For children, however, a far lower level of 20 mcg/L appears more appropriate, Alyson Connor reported at the annual meeting of the Associated Professional Sleep Societies.

"Our findings raise a question about the best serum ferritin threshold to use when seeing a child with evidence of sleep restlessness, in particular with elevated" periodic limb movement index, Ms. Connor said. "While many children may benefit from iron supplementation for these disorders, this area is worthy of further study as investigation in animals suggests that treatment of iron deficiency with high-dose iron supplementation in early development may differentially affect the development of the brain."

Ms. Connor, a student in the biomedical research program at the University of Michigan, Ann Arbor, and her colleagues conducted a review of 537 children aged 12 months to 18 years who were referred to the University of Michigan pediatric sleep clinic. In this cohort (mean age was 8.9 years; 62% were male), a median serum ferritin level of 23.6 mcg/L was linked with polysomnographic measures of periodic limb movement index (PLMI) per hour of total sleep time of 5 or more, while a median level of 30 mcg/L was linked with PLMI less than 5. About 26% of the subjects had a PLMI of 5 or more.

"There was a significant association between serum ferritin and periodic limb movement index such that, for every 10-mcg/L increase in serum ferritin, it decreased the odds of having an elevated PLMI by 11%," Ms. Connor said. Boys, younger patients, those with lower serum ferritin levels, and those with a shorter time between hematology and polysomnography had significantly increased odds of a PLMI of 5 or more. Only 19% of patients had a serum ferritin above 50 mcg/L, while 50% had below 30 mcg/L.

"The cutoff of 50 is very sensitive, giving few false negatives, but not very specific, whereas a lower cutoff improves specificity," Ms. Connor said.

Studies are needed to assess the link between iron status and sleep measures in a general pediatric population and to determine the best iron dose, timing, and method of delivery, she said.

MINNEAPOLIS – Caffeine therapy for apnea of prematurity did not prevent the development of persistent and significant obstructive apnea in children by the time they were 9 years old.

"Apnea of prematurity occurs in more than three-quarters of infants born at under 30 weeks’ gestation, and more than half of these apneas are actually obstructive in nature," Dr. Carole Marcus, director of the sleep center at Children’s Hospital of Philadelphia, said at the annual meeting of the Associated Professional Sleep Societies. "Caffeine is now the most commonly used drug [to treat apnea of prematurity] in the NICU in infants less than 32 weeks’ gestation," she said.

The long-term effects of caffeine on sleep in the developing brain are not well understood. It is not known whether neonatal caffeine administration has permanent adverse effects on sleep architecture and ventilatory control, perhaps increasing the risk of later sleep disorders such as insomnia and obstructive sleep apnea (OSA).

In the earlier CAP (Caffeine for Apnea of Prematurity) trial, 793 premature infants with birth weights of 500-1250 g were randomly assigned to caffeine or placebo until therapy for apnea of prematurity was no longer needed. Caffeine significantly improved the rate of survival without neurodevelopmental disability at 18-21 months in these babies versus placebo (N. Engl. J. Med. 2007;357:1893-902).

The subsequent axillary long-term CAP-S (Sleep) trial of 201 CAP subjects looked at whether neonatal caffeine administration resulted in later sleep abnormalities. The researchers assessed the ex-premature children aged 5-11 years (mean age, 9 years) via sleep questionnaires, actigraphy, and full ambulatory (home-based) polysomnography. The patients assessed in CAP-S were from Canada or Australia, with a high proportion of white patients, high maternal education, and high socioeconomic status, "which is relevant to our obstructive sleep apnea outcomes," noted Dr. Marcus.

No significant differences were noted in children who had received caffeine, compared with those who did not, in terms of subjective measures of sleep quality or quantity. Total recording time and total sleep time on polysomnography were somewhat longer in the caffeine arm, but there was no difference seen in sleep efficiency between groups (P = .91).

However, OSA (apnea-hypopnea index of more than two episodes per hour) was common in both groups (8.2% of the caffeine group and 11% of the placebo group; P = .22). In contrast, the prevalence of OSA in the general pediatric population is 1%-4%.

Also, 24% of the caffeine group and 29% of the placebo group had OSA on polysomnography and/or a history of adenoidectomy/tonsillectomy, again with no difference between groups (P = .35).

A large proportion of subjects in both arms had elevated periodic limb movements (17.5% in the caffeine group and 11% in the placebo groups; P = .27), a proportion that was markedly higher than was the normal prevalence in cases in which the child had more than five episodes of periodic limb movement per hour, which lies between 5% and 8%."This study further supports the use of caffeine for apnea of prematurity as it has been shown to have quite a number of beneficial effects and no long-term adverse effects," said Dr. Marcus. "Further study is needed on the mechanisms underlying the high prevalence of sleep disorders, both OSA and [periodic limb movement syndrome], in ex-preterm infants."

Dr. Marcus said there is a long-term behavioral and neurocognitive outcomes study ongoing on the CAP cohort, but they were unable to combine those results with the sleep study results because the data were collected too far apart.

MINNEAPOLIS – Caffeine therapy for apnea of prematurity did not prevent the development of persistent and significant obstructive apnea in children by the time they were 9 years old.

"Apnea of prematurity occurs in more than three-quarters of infants born at under 30 weeks’ gestation, and more than half of these apneas are actually obstructive in nature," Dr. Carole Marcus, director of the sleep center at Children’s Hospital of Philadelphia, said at the annual meeting of the Associated Professional Sleep Societies. "Caffeine is now the most commonly used drug [to treat apnea of prematurity] in the NICU in infants less than 32 weeks’ gestation," she said.

The long-term effects of caffeine on sleep in the developing brain are not well understood. It is not known whether neonatal caffeine administration has permanent adverse effects on sleep architecture and ventilatory control, perhaps increasing the risk of later sleep disorders such as insomnia and obstructive sleep apnea (OSA).

In the earlier CAP (Caffeine for Apnea of Prematurity) trial, 793 premature infants with birth weights of 500-1250 g were randomly assigned to caffeine or placebo until therapy for apnea of prematurity was no longer needed. Caffeine significantly improved the rate of survival without neurodevelopmental disability at 18-21 months in these babies versus placebo (N. Engl. J. Med. 2007;357:1893-902).

The subsequent axillary long-term CAP-S (Sleep) trial of 201 CAP subjects looked at whether neonatal caffeine administration resulted in later sleep abnormalities. The researchers assessed the ex-premature children aged 5-11 years (mean age, 9 years) via sleep questionnaires, actigraphy, and full ambulatory (home-based) polysomnography. The patients assessed in CAP-S were from Canada or Australia, with a high proportion of white patients, high maternal education, and high socioeconomic status, "which is relevant to our obstructive sleep apnea outcomes," noted Dr. Marcus.

No significant differences were noted in children who had received caffeine, compared with those who did not, in terms of subjective measures of sleep quality or quantity. Total recording time and total sleep time on polysomnography were somewhat longer in the caffeine arm, but there was no difference seen in sleep efficiency between groups (P = .91).

However, OSA (apnea-hypopnea index of more than two episodes per hour) was common in both groups (8.2% of the caffeine group and 11% of the placebo group; P = .22). In contrast, the prevalence of OSA in the general pediatric population is 1%-4%.

Also, 24% of the caffeine group and 29% of the placebo group had OSA on polysomnography and/or a history of adenoidectomy/tonsillectomy, again with no difference between groups (P = .35).

A large proportion of subjects in both arms had elevated periodic limb movements (17.5% in the caffeine group and 11% in the placebo groups; P = .27), a proportion that was markedly higher than was the normal prevalence in cases in which the child had more than five episodes of periodic limb movement per hour, which lies between 5% and 8%."This study further supports the use of caffeine for apnea of prematurity as it has been shown to have quite a number of beneficial effects and no long-term adverse effects," said Dr. Marcus. "Further study is needed on the mechanisms underlying the high prevalence of sleep disorders, both OSA and [periodic limb movement syndrome], in ex-preterm infants."

Dr. Marcus said there is a long-term behavioral and neurocognitive outcomes study ongoing on the CAP cohort, but they were unable to combine those results with the sleep study results because the data were collected too far apart.

MINNEAPOLIS – Caffeine therapy for apnea of prematurity did not prevent the development of persistent and significant obstructive apnea in children by the time they were 9 years old.

"Apnea of prematurity occurs in more than three-quarters of infants born at under 30 weeks’ gestation, and more than half of these apneas are actually obstructive in nature," Dr. Carole Marcus, director of the sleep center at Children’s Hospital of Philadelphia, said at the annual meeting of the Associated Professional Sleep Societies. "Caffeine is now the most commonly used drug [to treat apnea of prematurity] in the NICU in infants less than 32 weeks’ gestation," she said.

The long-term effects of caffeine on sleep in the developing brain are not well understood. It is not known whether neonatal caffeine administration has permanent adverse effects on sleep architecture and ventilatory control, perhaps increasing the risk of later sleep disorders such as insomnia and obstructive sleep apnea (OSA).

In the earlier CAP (Caffeine for Apnea of Prematurity) trial, 793 premature infants with birth weights of 500-1250 g were randomly assigned to caffeine or placebo until therapy for apnea of prematurity was no longer needed. Caffeine significantly improved the rate of survival without neurodevelopmental disability at 18-21 months in these babies versus placebo (N. Engl. J. Med. 2007;357:1893-902).

The subsequent axillary long-term CAP-S (Sleep) trial of 201 CAP subjects looked at whether neonatal caffeine administration resulted in later sleep abnormalities. The researchers assessed the ex-premature children aged 5-11 years (mean age, 9 years) via sleep questionnaires, actigraphy, and full ambulatory (home-based) polysomnography. The patients assessed in CAP-S were from Canada or Australia, with a high proportion of white patients, high maternal education, and high socioeconomic status, "which is relevant to our obstructive sleep apnea outcomes," noted Dr. Marcus.

No significant differences were noted in children who had received caffeine, compared with those who did not, in terms of subjective measures of sleep quality or quantity. Total recording time and total sleep time on polysomnography were somewhat longer in the caffeine arm, but there was no difference seen in sleep efficiency between groups (P = .91).

However, OSA (apnea-hypopnea index of more than two episodes per hour) was common in both groups (8.2% of the caffeine group and 11% of the placebo group; P = .22). In contrast, the prevalence of OSA in the general pediatric population is 1%-4%.

Also, 24% of the caffeine group and 29% of the placebo group had OSA on polysomnography and/or a history of adenoidectomy/tonsillectomy, again with no difference between groups (P = .35).

A large proportion of subjects in both arms had elevated periodic limb movements (17.5% in the caffeine group and 11% in the placebo groups; P = .27), a proportion that was markedly higher than was the normal prevalence in cases in which the child had more than five episodes of periodic limb movement per hour, which lies between 5% and 8%."This study further supports the use of caffeine for apnea of prematurity as it has been shown to have quite a number of beneficial effects and no long-term adverse effects," said Dr. Marcus. "Further study is needed on the mechanisms underlying the high prevalence of sleep disorders, both OSA and [periodic limb movement syndrome], in ex-preterm infants."

Dr. Marcus said there is a long-term behavioral and neurocognitive outcomes study ongoing on the CAP cohort, but they were unable to combine those results with the sleep study results because the data were collected too far apart.

MINNEAPOLIS – Increased exposure to ozone appeared to increase cardiorespiratory events in healthy sleeping infants, based on preliminary data.

Dr. Chana I.C. Chin, a pediatric pulmonology fellow at the Children’s Hospital Los Angeles, and her associates found that for each 10-part-per-billion (ppb) elevation in ozone, there was a 22% higher risk of an apneic or bradycardic event (P = .0012), after adjustment for seasonality, weekday, and individual random effects.

"Air pollution likely increases infant mortality and likely impacts infants to a greater degree than adults," said Dr. Chin. "Infants have small body mass ratio, narrower airways, higher particulate deposition rates, greater volume of air breathed per body weight, and underdeveloped defense mechanisms."

"Infant sleep offers a great way to study the impact of air pollutants. And [infants] sleep 16-18 hours a day and depend on healthy sleep for appropriate growth and development," she added.

Disturbed sleep has been seen in rats exposed to ozone (O₃), but human data are scant. Dr. Chin and her associates studied the association between air pollution and sleep-associated cardiopulmonary events in the hopes of elucidating a novel pathway linking early-life exposure to adverse health effects in childhood.

Researchers studied 196 healthy term infants, preterm infants, siblings of SIDS infants, and infants experiencing apparent life-threatening events who were residing in Southern California and recruited in 1994-1998 for the Collaborative Home Infant Monitoring Evaluation Study (CHIME). Home-based monitoring of sleep measured central, obstructive, and mixed apnea and bradycardia aggregated over a 24-hour period.

About 60% of the cohort were preterm, with an overall median gestational age of 33.7 weeks. Historical Environmental Protection Agency air quality data on daily O₃, nitrogen dioxide, and particulate matter exposure was assigned to each residential zip code.

A greater adverse effect for ozone was found in the term babies (odds ratio, 1.35), compared with the preterm babies (OR, 1.01; P=.0005 for between-group difference), Dr. Chin reported at the annual meeting of the Associated Professional Sleep Societies.

Nitrogen dioxide and particulate matter exposure were not significantly associated with sleep apnea or bradycardia in these infants.

"We speculate that the observed O₃ apnea and bradycardia association implies a possible mechanistic pathway for subsequent adverse morbidity and mortality and merits further investigation," Dr. Chin said.

In response to a query on the relative age of the data (1994-1998), she commented that the CHIME data are so valuable since "we don’t really monitor healthy infants anymore."

"There’s no large trial now where a parent would say, "Yes, put my infant on an apnea monitor when there’s no indication," Dr. Chin added.

This study was supported by a pilot project grant from the National Institute of Environmental Health Sciences–funded USC/UCLA Southern Environmental Health Sciences Center and a seed grant from the American Medical Association. Dr. Chin reported having no disclosures.

MINNEAPOLIS – Increased exposure to ozone appeared to increase cardiorespiratory events in healthy sleeping infants, based on preliminary data.

Dr. Chana I.C. Chin, a pediatric pulmonology fellow at the Children’s Hospital Los Angeles, and her associates found that for each 10-part-per-billion (ppb) elevation in ozone, there was a 22% higher risk of an apneic or bradycardic event (P = .0012), after adjustment for seasonality, weekday, and individual random effects.

"Air pollution likely increases infant mortality and likely impacts infants to a greater degree than adults," said Dr. Chin. "Infants have small body mass ratio, narrower airways, higher particulate deposition rates, greater volume of air breathed per body weight, and underdeveloped defense mechanisms."

"Infant sleep offers a great way to study the impact of air pollutants. And [infants] sleep 16-18 hours a day and depend on healthy sleep for appropriate growth and development," she added.

Disturbed sleep has been seen in rats exposed to ozone (O₃), but human data are scant. Dr. Chin and her associates studied the association between air pollution and sleep-associated cardiopulmonary events in the hopes of elucidating a novel pathway linking early-life exposure to adverse health effects in childhood.

Researchers studied 196 healthy term infants, preterm infants, siblings of SIDS infants, and infants experiencing apparent life-threatening events who were residing in Southern California and recruited in 1994-1998 for the Collaborative Home Infant Monitoring Evaluation Study (CHIME). Home-based monitoring of sleep measured central, obstructive, and mixed apnea and bradycardia aggregated over a 24-hour period.

About 60% of the cohort were preterm, with an overall median gestational age of 33.7 weeks. Historical Environmental Protection Agency air quality data on daily O₃, nitrogen dioxide, and particulate matter exposure was assigned to each residential zip code.

A greater adverse effect for ozone was found in the term babies (odds ratio, 1.35), compared with the preterm babies (OR, 1.01; P=.0005 for between-group difference), Dr. Chin reported at the annual meeting of the Associated Professional Sleep Societies.

Nitrogen dioxide and particulate matter exposure were not significantly associated with sleep apnea or bradycardia in these infants.

"We speculate that the observed O₃ apnea and bradycardia association implies a possible mechanistic pathway for subsequent adverse morbidity and mortality and merits further investigation," Dr. Chin said.

In response to a query on the relative age of the data (1994-1998), she commented that the CHIME data are so valuable since "we don’t really monitor healthy infants anymore."

"There’s no large trial now where a parent would say, "Yes, put my infant on an apnea monitor when there’s no indication," Dr. Chin added.

This study was supported by a pilot project grant from the National Institute of Environmental Health Sciences–funded USC/UCLA Southern Environmental Health Sciences Center and a seed grant from the American Medical Association. Dr. Chin reported having no disclosures.

MINNEAPOLIS – Increased exposure to ozone appeared to increase cardiorespiratory events in healthy sleeping infants, based on preliminary data.

Dr. Chana I.C. Chin, a pediatric pulmonology fellow at the Children’s Hospital Los Angeles, and her associates found that for each 10-part-per-billion (ppb) elevation in ozone, there was a 22% higher risk of an apneic or bradycardic event (P = .0012), after adjustment for seasonality, weekday, and individual random effects.

"Air pollution likely increases infant mortality and likely impacts infants to a greater degree than adults," said Dr. Chin. "Infants have small body mass ratio, narrower airways, higher particulate deposition rates, greater volume of air breathed per body weight, and underdeveloped defense mechanisms."

"Infant sleep offers a great way to study the impact of air pollutants. And [infants] sleep 16-18 hours a day and depend on healthy sleep for appropriate growth and development," she added.

Disturbed sleep has been seen in rats exposed to ozone (O₃), but human data are scant. Dr. Chin and her associates studied the association between air pollution and sleep-associated cardiopulmonary events in the hopes of elucidating a novel pathway linking early-life exposure to adverse health effects in childhood.

Researchers studied 196 healthy term infants, preterm infants, siblings of SIDS infants, and infants experiencing apparent life-threatening events who were residing in Southern California and recruited in 1994-1998 for the Collaborative Home Infant Monitoring Evaluation Study (CHIME). Home-based monitoring of sleep measured central, obstructive, and mixed apnea and bradycardia aggregated over a 24-hour period.

About 60% of the cohort were preterm, with an overall median gestational age of 33.7 weeks. Historical Environmental Protection Agency air quality data on daily O₃, nitrogen dioxide, and particulate matter exposure was assigned to each residential zip code.

A greater adverse effect for ozone was found in the term babies (odds ratio, 1.35), compared with the preterm babies (OR, 1.01; P=.0005 for between-group difference), Dr. Chin reported at the annual meeting of the Associated Professional Sleep Societies.

Nitrogen dioxide and particulate matter exposure were not significantly associated with sleep apnea or bradycardia in these infants.

"We speculate that the observed O₃ apnea and bradycardia association implies a possible mechanistic pathway for subsequent adverse morbidity and mortality and merits further investigation," Dr. Chin said.

In response to a query on the relative age of the data (1994-1998), she commented that the CHIME data are so valuable since "we don’t really monitor healthy infants anymore."

"There’s no large trial now where a parent would say, "Yes, put my infant on an apnea monitor when there’s no indication," Dr. Chin added.

This study was supported by a pilot project grant from the National Institute of Environmental Health Sciences–funded USC/UCLA Southern Environmental Health Sciences Center and a seed grant from the American Medical Association. Dr. Chin reported having no disclosures.

MINNEAPOLIS – Caffeine therapy for apnea of prematurity did not prevent the development of persistent and significant obstructive apnea in children by the time they were 9 years old.

"Apnea of prematurity occurs in more than three-quarters of infants born at under 30 weeks’ gestation, and more than half of these apneas are actually obstructive in nature," reported Dr. Carole Marcus, who is director of the sleep center at the Children’s Hospital of Philadelphia, at the annual meeting of the Associated Professional Sleep Societies. "Caffeine is now the most commonly used drug [to treat apnea of prematurity] in the NICU in infants less than 32 weeks’ gestation," she said.

Apnea of prematurity is commonly treated with therapeutic caffeine administration, but the long-term effects of caffeine on sleep in the developing brain are not well understood. In particular, it is not known whether neonatal caffeine administration has permanent adverse effects on sleep architecture and ventilatory control, perhaps increasing the risk of later sleep disorders such as insomnia and obstructive sleep apnea.

In the earlier CAP (Caffeine for Apnea of Prematurity) trial, 793 premature infants with birthweights of 500-1250 g were randomly assigned to receive either caffeine or placebo until therapy for apnea of prematurity was no longer needed.

Caffeine therapy was shown to improve the rate of survival without neurodevelopmental disability at 18 to 21 months in these babies compared to placebo (P = .008) (N. Engl. J. Med. 2007;357:1893-902).

The subsequent axillary long-term CAP-S (Sleep) trial involved 201 CAP subjects and looked at whether neonatal caffeine administration resulted in later sleep abnormalities. The investigators assessed the ex-premature children between ages 5-11 years (mean age, 9 years) using sleep questionnaires, actigraphy, and full ambulatory (home-based) polysomnography.

Of note, the patients assessed in CAP-S were from either Canada or Australia, with a high proportion of white patients, high maternal education, and high socioeconomic status, "which is relevant to our obstructive sleep apnea outcomes," noted Dr. Marcus (SLEEP 2014 abstract supplement;37:abst.#0862).

No significant differences were noted in children who had received caffeine, compared with those who did not in terms of subjective measures of sleep quality or quantity. Total recording time and total sleep time on polysomnography were somewhat longer in the caffeine arm, but there was no difference seen in sleep efficiency between groups (P = .91).

However, obstructive sleep apnea (apnea-hypopnea index of more than two episodes/hour) was common in both groups (8.2% of the caffeine group and 11% of the placebo group; P = .22). In contrast, the prevalence of obstructive sleep apnea in the general pediatric population is between 1%-4%.

Also, 24% of the caffeine group and 29% of the placebo group had either obstructive sleep apnea on polysomnography and/or a history of adenoidectomy/tonsillectomy, again with no difference between groups (P = .35).

A large proportion of subjects in both arms had elevated periodic limb movements (17.5% in the caffeine group and 11% in the placebo groups; P = .27), a proportion that was markedly higher than was the normal prevalence in cases in which the child had more than five episodes of periodic limb movement/hour, which lies between 5% and 8%.

"We think that this study further supports the use of caffeine for apnea of prematurity as it has been shown to have quite a number of beneficial effects and no long-term adverse effects," said Dr. Marcus. "However, further study is needed on the mechanisms underlying the high prevalence of sleep disorders, both [obstructive sleep apnea] and [periodic limb movement syndrome], in ex-preterm infants."

In response to a comment, Dr. Marcus added that there is also a long-term behavioral and neurocognitive outcomes study ongoing on the CAP cohort, but they were unable to combine those results with the sleep study results because the data were collected too far apart.

The study was supported by an National Institutes of Health R01 grant and by the Canadian Institute for Health Research. Philips Respironics provided actigraphy equipment. Dr. Marcus reported that she gets unrelated research support from Philips Respironics and Ventus.

MINNEAPOLIS – Caffeine therapy for apnea of prematurity did not prevent the development of persistent and significant obstructive apnea in children by the time they were 9 years old.

"Apnea of prematurity occurs in more than three-quarters of infants born at under 30 weeks’ gestation, and more than half of these apneas are actually obstructive in nature," reported Dr. Carole Marcus, who is director of the sleep center at the Children’s Hospital of Philadelphia, at the annual meeting of the Associated Professional Sleep Societies. "Caffeine is now the most commonly used drug [to treat apnea of prematurity] in the NICU in infants less than 32 weeks’ gestation," she said.

Apnea of prematurity is commonly treated with therapeutic caffeine administration, but the long-term effects of caffeine on sleep in the developing brain are not well understood. In particular, it is not known whether neonatal caffeine administration has permanent adverse effects on sleep architecture and ventilatory control, perhaps increasing the risk of later sleep disorders such as insomnia and obstructive sleep apnea.

In the earlier CAP (Caffeine for Apnea of Prematurity) trial, 793 premature infants with birthweights of 500-1250 g were randomly assigned to receive either caffeine or placebo until therapy for apnea of prematurity was no longer needed.

Caffeine therapy was shown to improve the rate of survival without neurodevelopmental disability at 18 to 21 months in these babies compared to placebo (P = .008) (N. Engl. J. Med. 2007;357:1893-902).

The subsequent axillary long-term CAP-S (Sleep) trial involved 201 CAP subjects and looked at whether neonatal caffeine administration resulted in later sleep abnormalities. The investigators assessed the ex-premature children between ages 5-11 years (mean age, 9 years) using sleep questionnaires, actigraphy, and full ambulatory (home-based) polysomnography.

Of note, the patients assessed in CAP-S were from either Canada or Australia, with a high proportion of white patients, high maternal education, and high socioeconomic status, "which is relevant to our obstructive sleep apnea outcomes," noted Dr. Marcus (SLEEP 2014 abstract supplement;37:abst.#0862).

No significant differences were noted in children who had received caffeine, compared with those who did not in terms of subjective measures of sleep quality or quantity. Total recording time and total sleep time on polysomnography were somewhat longer in the caffeine arm, but there was no difference seen in sleep efficiency between groups (P = .91).

However, obstructive sleep apnea (apnea-hypopnea index of more than two episodes/hour) was common in both groups (8.2% of the caffeine group and 11% of the placebo group; P = .22). In contrast, the prevalence of obstructive sleep apnea in the general pediatric population is between 1%-4%.

Also, 24% of the caffeine group and 29% of the placebo group had either obstructive sleep apnea on polysomnography and/or a history of adenoidectomy/tonsillectomy, again with no difference between groups (P = .35).

A large proportion of subjects in both arms had elevated periodic limb movements (17.5% in the caffeine group and 11% in the placebo groups; P = .27), a proportion that was markedly higher than was the normal prevalence in cases in which the child had more than five episodes of periodic limb movement/hour, which lies between 5% and 8%.

"We think that this study further supports the use of caffeine for apnea of prematurity as it has been shown to have quite a number of beneficial effects and no long-term adverse effects," said Dr. Marcus. "However, further study is needed on the mechanisms underlying the high prevalence of sleep disorders, both [obstructive sleep apnea] and [periodic limb movement syndrome], in ex-preterm infants."

In response to a comment, Dr. Marcus added that there is also a long-term behavioral and neurocognitive outcomes study ongoing on the CAP cohort, but they were unable to combine those results with the sleep study results because the data were collected too far apart.

The study was supported by an National Institutes of Health R01 grant and by the Canadian Institute for Health Research. Philips Respironics provided actigraphy equipment. Dr. Marcus reported that she gets unrelated research support from Philips Respironics and Ventus.

MINNEAPOLIS – Caffeine therapy for apnea of prematurity did not prevent the development of persistent and significant obstructive apnea in children by the time they were 9 years old.

"Apnea of prematurity occurs in more than three-quarters of infants born at under 30 weeks’ gestation, and more than half of these apneas are actually obstructive in nature," reported Dr. Carole Marcus, who is director of the sleep center at the Children’s Hospital of Philadelphia, at the annual meeting of the Associated Professional Sleep Societies. "Caffeine is now the most commonly used drug [to treat apnea of prematurity] in the NICU in infants less than 32 weeks’ gestation," she said.

Apnea of prematurity is commonly treated with therapeutic caffeine administration, but the long-term effects of caffeine on sleep in the developing brain are not well understood. In particular, it is not known whether neonatal caffeine administration has permanent adverse effects on sleep architecture and ventilatory control, perhaps increasing the risk of later sleep disorders such as insomnia and obstructive sleep apnea.

In the earlier CAP (Caffeine for Apnea of Prematurity) trial, 793 premature infants with birthweights of 500-1250 g were randomly assigned to receive either caffeine or placebo until therapy for apnea of prematurity was no longer needed.

Caffeine therapy was shown to improve the rate of survival without neurodevelopmental disability at 18 to 21 months in these babies compared to placebo (P = .008) (N. Engl. J. Med. 2007;357:1893-902).

The subsequent axillary long-term CAP-S (Sleep) trial involved 201 CAP subjects and looked at whether neonatal caffeine administration resulted in later sleep abnormalities. The investigators assessed the ex-premature children between ages 5-11 years (mean age, 9 years) using sleep questionnaires, actigraphy, and full ambulatory (home-based) polysomnography.

Of note, the patients assessed in CAP-S were from either Canada or Australia, with a high proportion of white patients, high maternal education, and high socioeconomic status, "which is relevant to our obstructive sleep apnea outcomes," noted Dr. Marcus (SLEEP 2014 abstract supplement;37:abst.#0862).

No significant differences were noted in children who had received caffeine, compared with those who did not in terms of subjective measures of sleep quality or quantity. Total recording time and total sleep time on polysomnography were somewhat longer in the caffeine arm, but there was no difference seen in sleep efficiency between groups (P = .91).

However, obstructive sleep apnea (apnea-hypopnea index of more than two episodes/hour) was common in both groups (8.2% of the caffeine group and 11% of the placebo group; P = .22). In contrast, the prevalence of obstructive sleep apnea in the general pediatric population is between 1%-4%.

Also, 24% of the caffeine group and 29% of the placebo group had either obstructive sleep apnea on polysomnography and/or a history of adenoidectomy/tonsillectomy, again with no difference between groups (P = .35).

A large proportion of subjects in both arms had elevated periodic limb movements (17.5% in the caffeine group and 11% in the placebo groups; P = .27), a proportion that was markedly higher than was the normal prevalence in cases in which the child had more than five episodes of periodic limb movement/hour, which lies between 5% and 8%.

"We think that this study further supports the use of caffeine for apnea of prematurity as it has been shown to have quite a number of beneficial effects and no long-term adverse effects," said Dr. Marcus. "However, further study is needed on the mechanisms underlying the high prevalence of sleep disorders, both [obstructive sleep apnea] and [periodic limb movement syndrome], in ex-preterm infants."

In response to a comment, Dr. Marcus added that there is also a long-term behavioral and neurocognitive outcomes study ongoing on the CAP cohort, but they were unable to combine those results with the sleep study results because the data were collected too far apart.

The study was supported by an National Institutes of Health R01 grant and by the Canadian Institute for Health Research. Philips Respironics provided actigraphy equipment. Dr. Marcus reported that she gets unrelated research support from Philips Respironics and Ventus.

MINNEAPOLIS – Using the adult treatment threshold for serum ferritin to guide treatment in children with restless sleep may lead to inappropriate iron supplementation.

In both adults and children, iron deficiency has been linked to the presence and severity of restless legs syndrome (RLS) and periodic limb movements of sleep (PLMS). For adults, a serum ferritin less than 50 mcg/L is the threshold commonly used to guide iron supplementation for patients with RLS or PLMS. A threshold of 40-50 mcg/L also has been used in pediatric studies.

For children, however, a far lower level of 20 mcg/L appears more appropriate, Alyson Connor and her colleagues reported at the annual meeting of the Associated Professional Sleep Societies.

"Our findings raise a question about the best serum ferritin threshold to use when seeing a child with evidence of sleep restlessness, in particular with elevated" periodic limb movement index, Ms. Connor said, adding a caution that, "while many children may benefit from iron supplementation for these disorders, this area is worthy of further study as investigation in animals suggests that treatment of iron deficiency with high-dose iron supplementation in early development may differentially effect the development of the brain."

Ms. Connor, a student in the biomedical research program at the University of Michigan, Ann Arbor, and her colleagues conducted a review of 537 children between the ages of 12 months and 18 years who were referred to the University of Michigan pediatric sleep clinic.

In this young cohort (mean age was 8.9 years; 62% were male), a median serum ferritin level of 23.6 mcg/L was associated with polysomnographic measures of periodic limb movement index (PLMI) per hour of total sleep time of 5 or more, while a median level of 30 mcg/L was associated with PLMI less than 5. Overall, about 26% of the subjects had a PLMI of 5 or more.

"There was a significant association between serum ferritin and periodic limb movement index such that, for every 10 mcg/L increase in serum ferritin, it decreased the odds of having an elevated PLMI by 11%, with a P value of .01," Ms Connor said.

Boys, younger patients, those with lower serum ferritin levels, and those with a shorter time between hematology and polysomnography all had significantly increased odds of a PLMI of 5 or more.

Only 19% of patients studied had a serum ferritin above 50 mcg/L, while 50% had a serum ferritin below 30 mcg/L.

"Essentially, the cutoff of 50 is very sensitive, giving few false negatives, but not very specific, whereas a lower cutoff improves specificity," Ms. Connor said.

Future studies are needed to assess the association between iron status and sleep measures in a general pediatric population and to determine the best iron dose, timing, and method of delivery for optimal developmental outcomes, she said.

The study was supported by the Charles Woodson Fund for Clinical Research.

MINNEAPOLIS – Using the adult treatment threshold for serum ferritin to guide treatment in children with restless sleep may lead to inappropriate iron supplementation.

In both adults and children, iron deficiency has been linked to the presence and severity of restless legs syndrome (RLS) and periodic limb movements of sleep (PLMS). For adults, a serum ferritin less than 50 mcg/L is the threshold commonly used to guide iron supplementation for patients with RLS or PLMS. A threshold of 40-50 mcg/L also has been used in pediatric studies.

For children, however, a far lower level of 20 mcg/L appears more appropriate, Alyson Connor and her colleagues reported at the annual meeting of the Associated Professional Sleep Societies.

"Our findings raise a question about the best serum ferritin threshold to use when seeing a child with evidence of sleep restlessness, in particular with elevated" periodic limb movement index, Ms. Connor said, adding a caution that, "while many children may benefit from iron supplementation for these disorders, this area is worthy of further study as investigation in animals suggests that treatment of iron deficiency with high-dose iron supplementation in early development may differentially effect the development of the brain."

Ms. Connor, a student in the biomedical research program at the University of Michigan, Ann Arbor, and her colleagues conducted a review of 537 children between the ages of 12 months and 18 years who were referred to the University of Michigan pediatric sleep clinic.

In this young cohort (mean age was 8.9 years; 62% were male), a median serum ferritin level of 23.6 mcg/L was associated with polysomnographic measures of periodic limb movement index (PLMI) per hour of total sleep time of 5 or more, while a median level of 30 mcg/L was associated with PLMI less than 5. Overall, about 26% of the subjects had a PLMI of 5 or more.

"There was a significant association between serum ferritin and periodic limb movement index such that, for every 10 mcg/L increase in serum ferritin, it decreased the odds of having an elevated PLMI by 11%, with a P value of .01," Ms Connor said.

Boys, younger patients, those with lower serum ferritin levels, and those with a shorter time between hematology and polysomnography all had significantly increased odds of a PLMI of 5 or more.

Only 19% of patients studied had a serum ferritin above 50 mcg/L, while 50% had a serum ferritin below 30 mcg/L.

"Essentially, the cutoff of 50 is very sensitive, giving few false negatives, but not very specific, whereas a lower cutoff improves specificity," Ms. Connor said.

Future studies are needed to assess the association between iron status and sleep measures in a general pediatric population and to determine the best iron dose, timing, and method of delivery for optimal developmental outcomes, she said.

The study was supported by the Charles Woodson Fund for Clinical Research.

MINNEAPOLIS – Using the adult treatment threshold for serum ferritin to guide treatment in children with restless sleep may lead to inappropriate iron supplementation.

In both adults and children, iron deficiency has been linked to the presence and severity of restless legs syndrome (RLS) and periodic limb movements of sleep (PLMS). For adults, a serum ferritin less than 50 mcg/L is the threshold commonly used to guide iron supplementation for patients with RLS or PLMS. A threshold of 40-50 mcg/L also has been used in pediatric studies.

For children, however, a far lower level of 20 mcg/L appears more appropriate, Alyson Connor and her colleagues reported at the annual meeting of the Associated Professional Sleep Societies.

"Our findings raise a question about the best serum ferritin threshold to use when seeing a child with evidence of sleep restlessness, in particular with elevated" periodic limb movement index, Ms. Connor said, adding a caution that, "while many children may benefit from iron supplementation for these disorders, this area is worthy of further study as investigation in animals suggests that treatment of iron deficiency with high-dose iron supplementation in early development may differentially effect the development of the brain."

Ms. Connor, a student in the biomedical research program at the University of Michigan, Ann Arbor, and her colleagues conducted a review of 537 children between the ages of 12 months and 18 years who were referred to the University of Michigan pediatric sleep clinic.

In this young cohort (mean age was 8.9 years; 62% were male), a median serum ferritin level of 23.6 mcg/L was associated with polysomnographic measures of periodic limb movement index (PLMI) per hour of total sleep time of 5 or more, while a median level of 30 mcg/L was associated with PLMI less than 5. Overall, about 26% of the subjects had a PLMI of 5 or more.

"There was a significant association between serum ferritin and periodic limb movement index such that, for every 10 mcg/L increase in serum ferritin, it decreased the odds of having an elevated PLMI by 11%, with a P value of .01," Ms Connor said.

Boys, younger patients, those with lower serum ferritin levels, and those with a shorter time between hematology and polysomnography all had significantly increased odds of a PLMI of 5 or more.

Only 19% of patients studied had a serum ferritin above 50 mcg/L, while 50% had a serum ferritin below 30 mcg/L.

"Essentially, the cutoff of 50 is very sensitive, giving few false negatives, but not very specific, whereas a lower cutoff improves specificity," Ms. Connor said.

Future studies are needed to assess the association between iron status and sleep measures in a general pediatric population and to determine the best iron dose, timing, and method of delivery for optimal developmental outcomes, she said.

The study was supported by the Charles Woodson Fund for Clinical Research.