Abstract 328: An Outcomes Assessment of Vasodilator Therapy in Pulmonary Arterial Hypertension through Meta-analysis of 31 Randomized Controlled Trials

Abstract

Rationale: Pulmonary Arterial Hypertension (PAH), a rapidly progressive disease w/high mortality, has an annual therapy cost up to $100,000/yr. We sought to review the RCTs conducted to evaluate the effect of pulmonary vasodilators (PVDs) on outcomes in PAH. Methods: We conducted a systematic review and meta-analysis using MEDLINE, EMBASE, Cochrane Library, Web of Science, and Scopus (inceptions-January 2014) of RCTs in PAH. Data on all-cause mortality and changes in mean pulmonary artery pressure (mPAP), cardiac output (CO), and pulmonary vascular resistance (PVR) were abstracted when available. Summary relative risks (RR) and 95% confidence intervals (CI) for each outcome were calculated using a random effects model comparing various PVDs with placebo/control. Results: The analysis included 31 RCTs (6,271 patients, 77% female, 70% Caucasian, mean age 47±6yr). Median f/u was 12 weeks (18 trials had f/u of ≤12 weeks, and only 5 followed patients ≥6mos). Although no trial evaluated mortality as the primary endpoint, mortality was reported in 27 of the 31 trials. When analyzed overall PVDs demonstrated a mortality benefit with RR = 0.77 (95% CI = 0.60 - 0.98, I 2 = 0%), however prostacyclins were the only class of agents which independently demonstrated a significant mortality benefit, RR = 0.62 (95% CI 0.39-0.99, I 2 = 0%). Oral PVDs demonstrated no benefit on PAH mortality, RR = 0.85 (95% CI 0.65-1.12, I 2 = 0.0%). Many studies reported ≤1 death in at least one treatment arm, which produced unstable RR estimates and wide CIs. As a result, the mortality benefit and lack of heterogeneity observed in pooled estimates may be driven by these RCTs. When we limited the analysis to only studies with at least 2 deaths in both treatment arms (N=9), the mortality benefit for PVDs was no longer significant (RR 0.86, CI 0.66-1.12, I 2 =0%). Mean change in mPAP with all PVDs was -3.5 mmHg, (95% CI -4.2 - -2.8), compared to placebo 0.6 mmHg (95% CI 0.1 - 1.1), p<0.001. Mean increase in CO with PVDs was 0.6 L/min (95% CI 0.4 - 0.9 L/min) compared to placebo of -0.02 L/min (95% CI -0.15 - 0.1 L/min), p<0.001. PVR decreased by 212 dynes/s/cm 5 (95% CI 178 - 244) in PVD arms and increased by 43 dynes/s/cm 5 (95% CI 12 - 74) in placebo arms, p<0.001. No surrogate marker was predictive of changes in mortality. Patient reported outcomes were reported in 80% of RCTs but there was inconsistency in the measurements reported. Conclusion: The RCT literature on PAH medications focuses on short-term studies and surrogate endpoints. None was designed for mortality or patient-reported outcome measures. Only prostacyclins showed a mortality benefit in PAH, however, the mortality benefit for prostacyclins is based on secondary endpoints in small studies. There remains substantial uncertainty about the benefits of these expensive medications based on the existing trial evidence.