Abstract
OBJECTIVE: To determine predictors of adverse outcomes in peripartum cardiomyopathy (PPCM). METHODS AND RESULTS: We conducted a multi-center cohort study across four centers to identify subjects with PPCM with the following criteria: LVEF <40%, development of heart failure within the last month of pregnancy or within 5 months of delivery and no other identifiable cause of heart failure with reduced ejection fraction. Outcomes included 1) survival free from major adverse events (need for extra-corporeal membrane oxygenation, ventricular assist device, orthotopic heart transplantation or death) and 2) LVEF recovery ≥ 50%. Using a univariate logistic regression analysis, we identified significant clinical predictors of these outcomes, which were then used to create multivariable models. NT-proBNP at the time of diagnosis was examined both as a continuous variable (log transformed) in logistic regression and as a dichotomous variable (values above and below the median) using the log-rank test. In all, 237 women (1993 to 2017) with 736.4 person-years of follow-up, met criteria for PPCM. Participants had a mean age of 32.4 ± 6.7 years, mean BMI 30.6 ± 7.8 kg/m2; 63% were White. After median follow-up of 3.6 years (IQR 1.1-7.8), 113 (67%) had LVEF recovery, and 222 (94%) had survival free from adverse events. Significant predictors included gestational age, gravidity, systolic blood pressure, smoking, heart rate, initial LVEF, and diuretic use. In a subset of 110 patients with measured NTproBNP levels, we found a higher event free survival for women with NTproBNP <2585 pg/ml (median) as compared to women with NTproBNP ≥2585 pg/ml (log-rank test p-value 0.018). CONCLUSION: Gestational age, gravidity, current or past tobacco use, systolic blood pressure, heart rate, initial LVEF and diuretic requirement at the time of diagnosis were associated with survival free from adverse events and LVEF recovery. Initial NT-proBNP was significantly associated with event free survival.