Predictors of mortality in fibrosing pulmonary sarcoidosis

Introduction: Pulmonary fibrosing sarcoidosis is associated with increased mortality. This study was aimed to explore the prognosis value of a panel of parameters for predicting mortality. Methods: This retrospective study included 216 patients with confirmed stage 4 pulmonary sarcoidosis. Stage 4 diagnosis date served as baseline. The following information was systematically present at baseline: epidemiological characteristics; treatments; pulmonary function; composite physiologic index (CPI); systolic pulmonary artery pressure at echocardiography; pulmonary fibrosis extent, main pulmonary artery/ascending aorta diameters ratio (MPAD/AAD) and MPAD/body surface area (BSA) measured and calculated using computed tomography, Walsh's algorithm based on CPI, lung fibrosis extent and MPAD/AAD ratio, and modified Walsh's algorithm with MPAD/BSA replacing MPAD/AAD allowed to estimate good or bad prognosis profiles. The primary outcome of the study was all cause mortality and lung transplantation. The value of baseline parameters was tested as predictors of mortality using univariate and multivariate analyses. Results: Median follow-up was 105 months. There were 41 deaths and 5 transplantations. At multivariate analysis, survival was independently predicted by several parameters including CPI, lung fibrosis extent, pulmonary hypertension at echography or MPAD/BSA ratio, Walsh's algorithm, and geographic origin. The modified Walsh's algorithm was most highly predictive. Conclusion: Survival was best predicted by geographic origin, lung fibrosis extent, PH at echography or MPAD/BSA ratio, as well as by various scores among them the modified Walsh's algorithm had very high predictive value thanks to MPAD/BSA ratio which accurately predicted mortality.