We conducted a retrospective analysis of 451 women with an arteriovenous malformation (AVM) of the brain to determine whether pregnancy is a risk factor for cerebral hemorrhages. A total of 540 pregnancies occurred among our patient population, resulting in 438 live births and 102 abortions. There were 17 pregnancies complicated by a cerebral hemorrhage. The hemorrhage rate during pregnancy for women with an unruptured AVM was 0.035 ±0.005 per person-year. The hemorrhage rate for nonpregnant women of childbearing age with an unruptured AVM was 0.031 ± 0.002 per person-year. Pregnancy did not increase significantly the rate of first cerebral hemorrhage from an AVM (P = 0.35). We found that women with an AVM face a 3.5% risk of hemorrhage during pregnancy. Pregnancy is not a risk factor for hemorrhage in women without a previous hemorrhage. This conclusion assumes no selection bias exists in our study population; a bias would be introduced if the risk of fatal outcome after a hemorrhage were greater in pregnant women than in nonpregnant women.

Background “Complete Extrapolation” of efficacy from adult or other pediatric data, to the pediatric population, is an important scientific tool that reduces the need for pediatric efficacy trials. Dose finding and safety studies in pediatrics are still needed. “No Extrapolation” requires 2 pediatric efficacy trials. “Partial Extrapolation” eliminates the need to conduct 2 pediatric efficacy trials; 1 efficacy or exposure/response study may be sufficient. We examined pediatric extrapolation from 2009 to 2014 evaluating any changes in extrapolation assumptions and the causes for these changes since a prior analysis published in 2011. Methods We reviewed all 157 products with 388 pediatric studies submitted to the FDA from 2009 through 2014. We assessed whether efficacy was extrapolated from adult or other pediatric data and categorized extrapolation as Complete, Partial, or No, and identified the reasons for the changes. Results Partial extrapolation decreased, whereas use of No and Complete extrapolation noticeably increased. Complete, Partial, or No extrapolations changed from 14%, 68%, and 18% in the 2011 study to 34%, 29%, and 37% respectively in the current study. The changes were mostly due to a better understanding of pediatric pathophysiology, why trials have failed, and improved endpoints. Conclusions Evolving science and data obtained from clinical trials increases the certainty of extrapolation assumptions and drives decisions to utilize extrapolation. Lessons learned from the conduct of these trials are critical to improving evidence-based medicine. Extrapolation of Efficacy is a powerful scientific tool that streamlines pediatric product development. Increased knowledge and evolving science inform utilization of this tool.

No abstract

We conducted a retrospective analysis of 451 women with an arteriovenous malformation (AVM) of the brain to determine whether pregnancy is a risk factor for cerebral hemorrhages. A total of 540 pregnancies occurred among our patient population, resulting in 438 live births and 102 abortions. There were 17 pregnancies complicated by a cerebral hemorrhage. The hemorrhage rate during pregnancy for women with an unruptured AVM was 0.035 +/- 0.005 per person-year. The hemorrhage rate for nonpregnant women of childbearing age with an unruptured AVM was 0.031 +/- 0.002 per person-year. Pregnancy did not increase significantly the rate of first cerebral hemorrhage from an AVM (P = 0.35). We found that women with an AVM face a 3.5% risk of hemorrhage during pregnancy. Pregnancy is not a risk factor for hemorrhage in women without a previous hemorrhage. This conclusion assumes no selection bias exists in our study population; a bias would be introduced if the risk of fatal outcome after a hemorrhage were greater in pregnant women than in nonpregnant women.