Survival in pulmonary hypertension due to chronic lung disease: Influence of low diffusion capacity of the lungs for carbon monoxide

Abstract

Background: Patients with pulmonary hypertension (PH) due to chronic lung disease (Group 3 PH) have poor long-term outcomes. However, predictors of survival in Group 3 PH are not well described.

Methods: We performed a cohort study of Group 3 PH patients (n = 143; mean age 65 ± 12 years, 52% female) evaluated at the University of Minnesota. The Kaplan-Meier method and Cox regression analysis were used to assess survival and predictors of mortality, respectively. The clinical characteristics and survival were compared in patients categorized by PH severity based on the World Health Organization (WHO) classification and lung disease etiology.

Results: After a median follow-up of 1.4 years, there were 69 (48%) deaths. The 1-, 3-, and 5-year survival rates were 79%, 48%, and 31%. Age, coronary artery disease, atrial fibrillation, Charlson comorbidity index, serum N-terminal pro‒brain natriuretic peptide (NT-proBNP), creatinine, diffusion capacity of carbon monoxide (DLCO), total lung capacity, left ventricular ejection fraction, right atrial and right ventricular enlargement on echocardiography, cardiac index, and pulmonary vascular resistance (PVR) were univariate predictors of survival. On multivariable analysis, DLCO was the only predictor of mortality (adjusted hazard ratio [HR] for every 10% decrease in predicted value: 1.31 [95% confidence interval 1.12 to 1.47]; p = 0.003). The 1-/5-year survival by tertiles of DLCO was 84%/56%, 82%/44%, and 63%/14% (p = 0.01), respectively. On receiver-operating characteristic curve analysis, DLCO <32% of predicted had the highest sensitivity and specificity for predicting survival. The 1- and 5-year survival in patients with a DLCO ≥32% predicted was 84% and 60% vs 68% and 13% in patients with a DLCO <32% predicted (adjusted HR: 2.5 [95% confidence interval 1.3 to 5.0]; p = 0.007). Lung volumes and DLCO were not related, but higher PVR was strongly associated with reduced DLCO. There was increased mortality in interstitial lung disease‒PH as compared with chronic obstructive pulmonary disease‒PH, but PH severity based on the WHO classification did not alter survival.

Conclusions: Low DLCO is a predictor of mortality and should be used to risk-stratify Group 3 PH patients.

Keywords: COPD; Cor pulmonale; Emphysema; Interstitial lung disease; pulmonary function test; right ventricle.

Figure 1

Survival stratified by DLCO in Group 3 PH. (A) There was a linear relationship between survival and tertiles of DLCO in Group 3 PH. (B) DLCO of 32% predicted provides significant separation in survival in Group 3 PH. Age, gender, and Charlson comorbidity index adjusted HR: 2.5 (95% CI 1.3 to 5.0); p = 0.007.

Figure 2

Comparison of survival between mild and severe Group 3 PH. There are no differences in survival after adjustment for age, gender, and Charlson comorbidity score between patients with mild and severe Group 3 PH.

Figure 3

Comparison of survival between ILD-PH and COPD-PH. There is reduced survival in ILD-PH as compared with COPD-PH.

Figure 4

Survival in Group 3 PH by tertiles (A) and quartiles (B) of mean pulmonary artery pressure (mPAP). There was no difference in survival when the Group 3 PH cohort was divided into tertiles and quartiles of mPAP.

Figure 5

Elevated PVR is associated with increased mortality in Group 3 PH. A PVR of > 7.5 Wood units resulted in increased mortality as compared with a PVR of < 7.5 Wood units.

Figure 6

Low cardiac index was associated with increased mortality in Group 3 PH. A cardiac index of < 2 liters/min/m² resulted in increased mortality as compared with a cardiac index of > 2 liters/min/m².