Hemorrhage from arteriovenous malformations during pregnancy

Abstract

Background: Previous hemorrhage, deep venous drainage, and deep location are established risk factors for arteriovenous malformation (AVM) hemorrhage. Although pregnancy is an assumed risk factor, there is a relative paucity of data to support this neurosurgical tenet.

Objective: To elucidate the hemorrhage rate of AVMs during pregnancy.

Methods: We reviewed the records of 54 women with an angiographic diagnosis of an AVM at our institution. Annual hemorrhage rates were calculated as the ratio of the number of bleeds to total number of patient-years of follow-up. Patient-years of follow-up were tallied assuming lesion presence from birth until AVM obliteration. The Cox proportional hazards model for hemorrhage with pregnancy as the time-dependent variable was used to calculate the hazard ratio.

Results: Five hemorrhages in 4 patients occurred over 62 pregnancies, yielding a hemorrhage rate of 8.1% per pregnancy or 10.8% per year. Over the remaining 2461.3 patient-years of follow-up, only 28 hemorrhages occurred, yielding an annual hemorrhage rate of 1.1%. The hazard ratio for hemorrhage during pregnancy was 7.91 (P = 2.23 × 10(-4)), increasing to 18.12 (P = 7.31 × 10(-5)) when limiting the analysis to patient follow-up up to age 40.

Conclusion: Because of the increased risk of hemorrhage from AVMs during pregnancy, we recommend intervention in women who desire to bear children, particularly if the AVM has bled. If the AVM is discovered during pregnancy, we recommend early intervention if it has ruptured; if it is unruptured, we recommend comprehensive counseling, weighing risks of intervention against continuation of pregnancy without intervention.