Abstract Rationale Adverse events (AEs) associated with pulmonary arterial hypertension (PAH) medications are common and may blunt the therapeutic benefit. Few studies have examined differences in AE risk of PAH therapies by patient phenotype. Understanding AE risk is essential to guide personalized therapy, as this requires balancing benefits and risks for individual patients. We aimed to characterize heterogeneity in risk of treatment-associated AEs across multiple comorbidities. Methods We performed an individual participant data meta-analysis of 18 phase III randomized clinical trials of PAH therapy submitted to the U.S. Food and Drug Administration from 2000 to 2015. Patients with chronic thromboembolic pulmonary hypertension were excluded. We evaluated the effect of comorbidities (obesity, pulmonary disease, and chronic kidney disease CKD) on risk of AEs for participants randomized to active therapy versus placebo. Zero-inflated negative binomial regression models with interaction terms were used to determine the odds of experiencing treatment-associated AEs by comorbidity status, adjusting for age, sex, PAH etiology, and study follow-up time. Results The study included 6,448 adults with PAH. Mean age was 50.0 years, 78.6% were female, and 61.1% had idiopathic/heritable PAH. 34.4% were obese, 19.8% had pulmonary disease (16.0% asthma/COPD and 4.9% interstitial lung disease), and 12.5% had CKD. Most participants were WHO functional class II/III (96.1%), with a mean six-minute walk distance of 347.4 ± 83.5 meters. Comorbidities modified the odds of experiencing multiple treatment-associated AEs (Table). Active treatment was associated with 45% higher odds of heart failure or volume overload compared to placebo in participants with CKD (OR 1.45, 95% CI 1.08-1.94, p = 0.01); this relationship was not observed in those without CKD (OR 1.04, 95% CI 0.92-1.17, p = 0.53) (p for interaction=0.04). The probability of major bleeding while receiving active treatment was significantly higher versus placebo for those with pulmonary disease compared to those without (p for interaction=0.03). Conversely, participants with pulmonary disease had lower odds of treatment-associated nausea/vomiting and diarrhea (p for interaction ≤ 0.01 for both). Obesity did not modify risk for any AE outcome. There were no significant differences in the odds of treatment-associated pulmonary AEs, chest pain, or headache across comorbidity groups (p for interaction ≥ 0.05 for all). Conclusions In patients with PAH, the risk of treatment-associated AEs varies significantly by patient comorbidity. Future studies will examine additional AEs and comorbidities. Our findings highlight the importance of understanding heterogeneity in treatment-associated AE risk by patient phenotype for guiding individualized risk assessment and treatment decisions. This abstract is funded by: Cardiovascular Medical Research and Education Fund, National Institutes of Health T32GM075766, R01HL173533
Mccarthy et al. (Fri,) studied this question.