A meta-analysis of exercise hemodynamics in heart failure with preserved ejection fraction: the relevance of PAWP/CO slope

Abstract Background Exercise right heart catheterization (RHC) is considered the gold-standard test to diagnose heart failure with preserved ejection fraction (HFpEF). However, exercise RHC is an insufficiently standardized technique, and current hemodynamic thresholds to define HFpEF are not universally accepted. We sought to describe the exercise hemodynamics profile of HFpEF cohorts reported in literature, as compared with control subjects. Methods We performed a systematic literature review following the PRISMA statement until December 2020. Studies reporting pulmonary artery wedge pressure (PAWP) at rest and peak exercise were extracted. Summary estimates of all hemodynamic variables were evaluated, stratified according to body position (supine/upright exercise). The PAWP / cardiac output (CO) slope during exercise was extrapolated. Results Twenty-seven studies were identified, providing data for 2180 HFpEF patients and 682 controls. At peak exercise, HfpEF cohorts showed a summary estimate of PAWP at peak which was twice as high as compared with control cohorts (30; 95% CI: 29–31 mmHg and 16; 95% CI: 15–17 mmHg, respectively), as well as of delta PAWP (15; 95% CI: 14–16 mmHg and 7; 95% CI: 6–8 mmHg, respectively), and of right atrial pressure (18; 95% CI: 16–19 mmHg and 8; 95% CI: 8–9 mmHg, respectively). These differences persisted after adjustment for age, sex, body mass index, body position. Additionally, summary estimates of PAWP at peak performed during supine exercise was slightly higher than that obtained in upright position only for HFpEF cohorts (supine position: 31; 95% CI: 30–32 mmHg vs upright position; 26; 95% CI: 25–27 mmHg, respectively, p-value<0.01). However, peak PAWP values were highly heterogeneous among the cohorts (I2=93%), with a relative overlap with controls (Figure 1). HFpEF had a significantly larger impairment in the hemodynamic response to exercise, witnessed by a steeper summary PAWP/CO slope than controls (3.75; 95% CI: 3.20–4.28 mmHg/L/min and 0.95; 95% CI: 0.30–1.59 mmHg/L/min, p-value <0.0001), even after adjustment for covariates (p=0.007) (Figure 2). Finally, summary estimates of PAWP/CO slope were higher in HFpEF cohorts performing exercise in the supine position compared with those in upright position (p<0.0001 and p=0.0002 at non-adjusted and adjusted analysis, respectively), but not in control cohorts (p=0.135 and p=0.966 at non-adjusted and adjusted analysis, respectively). Conclusions Despite methodological heterogeneity across centers, the hemodynamic profile of HFpEF patients is consistent across studies and characterized by a higher left and right filling pressure at rest compared with controls, enhanced by physical exercise. A PAWP/CO slope cut-off >2 mmHg/L/min seems to retain validity also for studies conducted in the supine position, potentially overcoming the need of different supine and upright PAWP cut-offs. Funding Acknowledgement Type of funding sources: Private grant(s) and/or Sponsorship.