Clinical characterization and survival of patients with borderline elevation in pulmonary artery pressure

Abstract

Normal resting mean pulmonary artery pressure (PAP) is 8-20 mmHg. Pulmonary hypertension is defined as mean PAP of ≥25 mmHg. Borderline PAP levels of 21-24 mmHg are of unclear significance. We sought to determine the clinical characteristics and survival of subjects with mean PAP of 21-24 mmHg. We examined 1,491 patients enrolled in the Cleveland Clinic Pulmonary Hypertension Registry between February 1990 and May 2012 with baseline right heart catheterization. The relationship between PAP and all-cause mortality was assessed by Cox models and a tree-based analysis. Sixty-three patients had borderline PAP (underlying conditions: 12 left heart disease, 20 respiratory disease, 17 connective-tissue disease, 4 others, and 10 none). We then compared 3 groups: borderline PAP without heart or lung disease ([Formula: see text]), normal PAP without heart or lung disease ([Formula: see text]), and category 1 pulmonary arterial hypertension (PAH; [Formula: see text]). Borderline-PAP patients had levels of hemodynamic and functional compromise between those for normal-PAP patients and those for patients with PAH. Borderline PAP was associated with increased mortality compared to normal PAP (hazard ratio: 4.03 [95% confidence interval: 0.78-20.80], [Formula: see text]). A tree-based analysis demonstrated almost identical cut points in mean PAP (≤20, 21-26, and ≥27 mmHg) associated with differential survival ([Formula: see text]). Connective-tissue disease and an elevated transpulmonary gradient were predictors of worse survival in the borderline-PAP population. Borderline PAP elevation is associated with decreased survival, particularly in the context of connective-tissue disease and an elevated transpulmonary gradient.

Keywords: pulmonary artery pressure; pulmonary heart disease; pulmonary hypertension; survival.

Figure 1

Study population flow chart. RHC: right heart catheterization; PAP: pulmonary artery pressure; WHO: World Health Organization; PH: pulmonary hypertension.

Figure 2

Survival of patients with normal pulmonary artery pressure (PAP), borderline PAP, and group 1 pulmonary arterial hypertension (PAH). for the pairwise comparison between normal and borderline PAP; for the pairwise comparison between group 1 PAH and borderline PAP.

Figure 3

Survival of patients according to the tree-based analysis–identified cut points. for the pairwise comparison between mean pulmonary artery pressures (PAPs) of ≤20 and 21–26 mmHg; for the pairwise comparison between mean PAPs of 21–26 and ≥27 mmHg.

Figure 4

Receiver operating characteristic (ROC) curves showing that the optimal split in mean pulmonary artery pressure (PAP) associated with survival was between ≤26 and ≥27 mmHg (A), with the next optimal split at ≤20 versus 21–26 mmHg in the subset of patients with mean PAP ≤26 mmHg (B).

Figure 5

Predictors of mortality in the borderline–pulmonary artery pressure (PAP) population. Underlying connective-tissue disease (CTD), compared to any other or no associated (Assoc.) condition (A), elevated transpulmonary gradient (TPG; the cutoff of 15 mmHg was selected by recursive partitioning, with an imposed minimum group size of 5 after splitting; B), a lower 6-minute walk distance (6MWD in meters; C), and the presence of right atrial (RA) dilation (D) were associated with decreased survival.