Assessment of Clinical Worsening End Points as a Surrogate for Mortality in Pulmonary Arterial Hypertension: A Systematic Review and Meta-Analysis of Randomized Controlled Trials

Abstract

Background: Clinical worsening (CW) is a composite end point commonly used in pulmonary arterial hypertension (PAH) trials. We aimed to assess the trial-level surrogacy of CW for mortality in PAH trials, and whether the various CW components were similar in terms of frequency of occurrence, treatment-related relative risk (RR) reduction, and importance to patients. Methods: We searched MEDLINE, Embase, and the Cochrane Library (January 1990 to December 2020) for trials evaluating the effects of PAH therapies on CW. The coefficient of determination between the RR for CW and mortality was assessed by regression analysis. The frequency of occurrence, RR reduction, and importance to patients of the CW components were assessed. Results: We included 35 independent cohorts (9450 patients). PAH therapies significantly reduced CW events (RR, 0.64 [95% CI, 0.55–0.73]), including PAH-related hospitalizations (RR, 0.61 [95% CI, 0.47–0.79]), treatment escalation (RR, 0.57 [95% CI, 0.38–0.84]) and symptomatic progression (RR, 0.58 [95% CI, 0.48–0.69]), and modestly reduced all-cause mortality when incorporating deaths occurring after a primary CW-defining event (RR, 0.860 [95% CI, 0.742–0.997]). However, the effects of PAH-specific therapies on CW only modestly correlated with their effects on mortality ( R 2 trial , 0.35 [95% CI, 0.10–0.59]; P <0.0001), and the gradient in the treatment effect across component end points was large in the majority of trials. The weighted proportions of CW-defining events were hospitalization (33.5%) and symptomatic progression (32.3%), whereas death (6.7%), treatment escalation (5.6%), and transplantation/atrioseptostomy (0.2%) were infrequent. CW events were driven by the occurrence of events of major (49%) and mild-to-moderate (37%) importance to patients, with 14% of the events valued as critical. Conclusions: PAH therapies significantly reduced CW events, but study-level CW is not a surrogate for mortality in PAH trials. Moreover, components of CW largely vary in frequency, response to therapy, and importance to patients and are thus not interchangeable. Registration: URL: https://www.crd.york.ac.uk/PROSPERO ; Unique identifier: CRD42020178949.