Baseline pulmonary artery pressure (PAP) and six-month changes in PAP helped predict heart failure hospitalization (HFH) within two years in patients with chronic heart failure (HF), according to research published Aug. 6 in JACC: Heart Failure. The findings add to evidence linking PAP with HFH risk and mortality.

Investigators Michael R. Zile, MD, FACC, et al., analyzed retrospective data from 4,019 patients with chronic HF who participated in four different studies that used the CardioMEMS PAP monitor: CHAMPION, GUIDE-HF, US PAS and MEMS-HF. All patients were included in the present analysis regardless of treatment assignment; patients with precapillary pulmonary disease were excluded from all four trials. Among patients with a known LVEF, 2,561 had HF with reduced EF (HFrEF) and 1,453 had HF with preserved EF (HFpEF).

Over a mean follow-up of 1.2 years, baseline diastolic PAP was an independent predictor of HFH (hazard ratio [HR], 1.043; p<0.0001). Change in diastolic PAP from baseline to six months, assessed as a continuous variable, also predicted HFH risk (landmark analysis; HR, 1.047; p<0.0001).

When assessed as categorical variables, a decrease or increase of >2 mm Hg in diastolic PAP from baseline to six months predicted a 16% decrease and 50% increase in HFH, respectively (p<0.0001 for both), compared with no change. These results were similar for systolic and mean PAPs and across LVEFs, although HFH risk was higher among patients with HFrEF than with HFpEF at comparable PAP levels.

"The results of the analyses presented in this study support novel, clinically actionable insights into the relationship between PAPs and the risk of HFHs," the authors write. "These data support the central pivotal role that pressures play in the prognosis of all patients with chronic HF and the opportunity to prevent HF outcomes by applying hemodynamically guided care to patients with chronic HF."

Taken together, the findings reinforce the potential role of PAP-guided care in identifying and reducing HFH risk across HF phenotypes.