Noninvasive determinants of pulmonary hypertension in interstitial lung disease

Pulmonary hypertension (PH) in interstitial lung disease (ILD) is associated with increased mortality and impaired exertional capacity. Right heart catheterization is the diagnostic standard for PH but is invasive and not readily available. Noninvasive physiologic evaluation may predict PH in ILD. Forty-four patients with PH and ILD (PH-ILD) were compared with 22 with ILD alone (non-PH ILD). Six-min walk distance (6MWD, 223 ± 131 vs. 331 ± 125 m, p = 0.02) and diffusing capacity for carbon monoxide (DLCO, 33 ± 14% vs. 55 ± 21%, p < 0.001) were lower in patients with PH-ILD. PH-ILD patients exhibited a lower gas-exchange derived pulmonary vascular capacitance (GX_CAP, 251 ± 132 vs. 465 ± 282 mL × mmHg, p < 0.0001) and extrapolated maximum oxygen uptake (VO_2max) (56 ± 32% vs. 84 ± 37%, p = 0.003). Multivariate analysis was performed to determine predictors of VO_{2 max}. GX_CAP was the only variable that predicted extrapolated VO_{2 max} among PH-ILD and non-PH ILD patients. Receiver operating characteristic curve analysis assessed the ability of individual noninvasive variables to distinguish between PH-ILD and non-PH ILD patients. GX_CAP (area under the curve [AUC] 0.85 ± 0.04, p < 0.0001) and delta ETCO₂ (AUC 0.84 ± 0.04, p < 0.0001) were the strongest predictors of PH-ILD. A CART analysis selected GX_CAP, estimated right ventricular systolic pressure (eRVSP) by echocardiogram, and FVC/DLCO ratio as predictive variables for PH-ILD. With this analysis, the AUC improved to 0.94 (sensitivity of 0.86 and sensitivity of 0.93). Patients with a GX_CAP ≤ 416 mL × mmHg had an 82% probability of PH-ILD. Patients with GX_CAP ≤ 416 mL × mmHg and high FVC/DLCO ratio >1.7 had an 80% probability of PH-ILD. Patients with GX_CAP ≤ 416 mL × mmHg and an elevated eRVSP by echocardiogram >43 mmHg had 100% probability of PH-ILD. The incorporation of GX_CAP with either eRVSP or FVC/DLCO ratio distinguishes between PH-ILD and non-PH-ILD with high probability and may therefore assist in determining the need to proceed with a diagnostic right heart catheterization and potential initiation of pulmonary arterial hypertension-directed therapy in PH-ILD patients.