C104-16 Test Performance of Forced Vital Capacity (FVC)/Diffusing Capacity of Carbon Monoxide (DLCO) Ratio as a Marker of Pulmonary Hypertension in Interstitial Lung Disease

Introduction Pulmonary hypertension (PH) is a treatable and prevalent condition in patients with interstitial lung disease (ILD), but optimal timing and frequency of screening has not been defined. Forced vital capacity (FVC)/diffusing capacity of carbon monoxide (DLCO) ratio >1.4 has been proposed as a predictor of PH in patients with idiopathic pulmonary fibrosis (IPF) in prior literature, but this cutoff has not been validated or extended across ILD subtypes. We aimed to test the performance of FVC/DLCO in patients with a wide variety of ILDs to predict the presence of concurrent PH. Methods Patients with transthoracic echocardiogram (TTE) data available from the University of Chicago ILD Natural History registry were included in this analysis. Demographic and pulmonary function data were abstracted from the medical record. Direct echocardiographic PH was defined as notation of PH in the TTE report or right ventricular systolic pressure (RVSP) >40 mmHg , and indirect echocardiographic PH was defined as evidence of dilated RV. PH on right heart catheterization (RHC) was defined as mean pulmonary artery pressure >20 mmHg. Logistic regression controlled for hemoglobin was performed comparing patients with increased FVC/DLCO to those without, stratified by presence or absence of PH. Receiver operating characteristic (ROC) curve analysis was performed to determine the optimal cutoff for FVC/DLCO in the diagnosis of PH. Results Two hundred and sixty-three patients (IPF=116, HP = 13, CTD-ILD=91, unclassifiable=8, other=35) were included in this study. Ninety-eight (37.2%) patients had direct echocardiographic PH. An additional 23 (8.7%) patients had indirect echocardiographic PH. One hundred and sixty-two patients (61.5% of cohort) had FVC/DLCO >1.4. Sixty-nine of these had no direct or indirect echocardiographic PH; of these, 13 received RHC, of whom 8 had PH (62%). Overall, FVC/DLCO >1.4 had a sensitivity of 77% and specificity of 49% for echocardiographic and/or RHC pulmonary hypertension (c-statistic = 0.69). The best-performing cutoff was FVC/DLCO 2.16, with a sensitivity of 43% and a specificity of 86% for pulmonary hypertension, correctly classifying 68% of the cohort (Figure). Conclusions Increased FVC/DLCO was prevalent in patients with ILD of all subtypes in a tertiary care referral center. Some patients with increased FVC/DLCO had evidence of PH on RHC but not echocardiography, indicating that FVC/DLCO can be helpful in predicting PH missed by echocardiography and thus requiring confirmatory RHC. Our data indicate that a higher FVC/DLCO than that used in prior studies is required to more accurately classify PH in patients with ILD. This abstract is funded by: None