Impact of pulmonary disease on the prognosis in heart failure with preserved ejection fraction: the TOPCAT trial

Abstract

Chronic obstructive pulmonary disease (COPD) is highly prevalent and predictive of worse outcomes in heart failure (HF) with preserved ejection fraction (HFpEF).1,2 Severe COPD can result in cor pulmonale3 and worse outcomes in HF,2 while less severe obstructive lung disease is associated with impaired left ventricular filling and lower cardiac output despite preserved left ventricular ejection fraction (LVEF).4 We investigated the influence of milder obstructive lung disease – defined as the absence of use of steroids or supplemental O2 – on cardiovascular (CV) outcomes among patients with HFpEF enrolled in the Treatment of Preserved Cardiac Function Heart Failure with an Aldosterone Antagonist (TOPCAT) trial in the Americas. TOPCAT was a multicentre, randomized, double-blind, placebo-controlled trial that tested the efficacy of spironolactone to reduce CV morbidity and mortality in 3445 adults ≥50 years of age with HFpEF (LVEF ≥45%).5 Key exclusion criteria relevant to this analysis included severe lung disease requiring home O2 or systemic steroid therapy, moderate or severe pulmonary hypertension, and directed therapy or biologics for lung disease. Given the significant differences in population characteristics and outcomes by region,6 we studied the 1767 patients recruited in the Americas. All patients provided written informed consent, and the study was approved by local institutional review boards. Outcomes included the composite of CV death, aborted sudden death, or HF hospitalization (the TOPCAT primary outcome), the individual components of this composite, all-cause mortality, non-CV mortality, and all-cause hospitalization.5 Pulmonary disease was based on the report by the site investigator of any diagnosis of COPD or asthma at enrolment. Of 1765 patients enrolled in the Americas and with data on pulmonary disease status, 653 (37%) were included in the TOPCAT echocardiographic study.7 Multivariable Cox proportional hazards models were employed to relate pulmonary disease at baseline to each outcome, adjusted for age, female gender, white race, treatment group, enrolment strata, percutaneous coronary intervention, use of beta-blockers, smoking status, body mass index, and heart rate. We further adjusted for New York Heart Association (NYHA) class in separate models. Interaction between pulmonary disease and randomized treatment assignment (spironolactone vs. placebo) on clinical outcomes was assessed using a multiplicative interaction term. The mean age was 72± 10 years, 50% were women, and 22% were non-white. The prevalence of COPD or asthma was 24%. Patients with prevalent lung disease were younger and more frequently non-white, had higher prevalence of current smoking, obesity, prior percutaneous coronary intervention, and NYHA III/IV functional class, and lower prevalence of beta-blocker use (online supplementary Table S1). At a median follow-up of 2.4 years, the primary composite outcome occurred in 522 (30%), CV death in 223 (13%), HF hospitalization in 400 (23%), all-cause mortality in 385 (22%), and all-cause hospitalization in 1059 (60%). Prevalent pulmonary disease was associated with a higher risk of the primary composite endpoint, related to higher risk of HF hospitalization but not of CV death (Table 1). After adjustment for demographics and co-morbidities, associations persisted with the primary composite endpoint, HF hospitalization and all-cause hospitalization (Table 1). In a post hoc exploratory analysis, pulmonary disease at enrolment modified the relationship between treatment with spironolactone and subsequent CV mortality (interaction P = 0.01) and all-cause mortality (interaction P = 0.02), such that the risk reduction associated with spironolactone was greater among patients compared to those without pulmonary disease (Table 1). No significant effect modification was observed for the primary endpoint, HF hospitalization, or all-cause hospitalization. Among patients with pulmonary disease, those randomized to spironolactone demonstrated a lower prevalence of prior myocardial infarction and higher prevalence of beta-blocker use (online supplementary Table S2). Results remained unchanged in models adjusting for randomization strata, and further adjusting for prior myocardial infarction and beta-blocker use (online supplementary Table S3). Among 653 patients in the echocardiographic study, 159 (24%) had pulmonary disease (online supplementary Table S4). Pulmonary disease was associated with greater left ventricular wall thickness and left ventricular hypertrophy prevalence, higher LVEF and tissue Doppler imaging s’, and smaller left atrial volume index in unadjusted analysis. Only associations with LVEF, tissue Doppler imaging s’, and left atrial volume index persisted after accounting for age, sex, and race (online supplementary Table S5). In this analysis of HFpEF patients enrolled in TOPCAT in the Americas, obstructive lung disease was independently associated with a heightened risk of the primary composite outcome, HF hospitalization alone, and allcause hospitalization. Despite this, pulmonary disease was associated with higher LVEF and smaller left atrial volume index, without differences in right ventricular function or pulmonary pressure, suggesting an important role for extracardiac factors in mediating the observed increase in risk. In an exploratory post hoc analysis, obstructive lung disease modified the relationship of randomized treatment with all-cause and CV mortality, but not with the TOPCAT primary endpoint. Similar findings were observed in the I-PRESERVE trial, where COPD prevalence was an independent predictor of HF death or hospitalization.8 Potential mechanisms linking COPD to increased risk of HF hospitalization in HFpEF include misdiagnosis of less severe COPD as a HF exacerbation due to overlapping signs and symptoms,1 or to lower cardiopulmonary reserve in patients with combined HFpEF and obstructive pulmonary disease leading to a lower threshold for HF or respiratory decompensation resulting in an increased likelihood of hospitalization. One possible explanation for the finding of effect modification of baseline pulmonary disease on treatment effect for CV and allcause mortality is chance, given the post hoc nature of this analysis. However, pulmonary