Cirrhosis Risk in Sarcoidosis Without Documented Hepatic Involvement—A Large Nationwide Population‐Based Cohort Study

Abstract

ABSTRACT Background & Aims Given the likely underestimation of hepatic sarcoidosis, the aims were to evaluate long‐term cirrhosis risk in patients with sarcoidosis without documented hepatic involvement at baseline and to assess the impact of multi‐organ disease and immunomodulatory therapies on cirrhosis risk. Methods A retrospective cohort study using the Clalit Health Services database in Israel included adults with sarcoidosis without documented hepatic involvement or prevalent cirrhosis and age‐ and sex‐matched controls. The primary outcome was incident cirrhosis, analysed with inverse‐probability‐of‐treatment‐weighted (IPTW) Fine‐Gray modelling with death as a competing risk. Secondary analyses evaluated multi‐organ involvement using landmark methods and immunomodulatory therapies as time‐varying exposures. Results Among 5489 sarcoidosis patients and 6016 controls (142 002 person‐years), cirrhosis occurred in 112 (2.0%) and 38 (0.6%), respectively (incidence rate 1.61 vs. 0.53 per 1000 person‐years; rate ratio 3.05). In the primary IPTW‐weighted Fine‐Gray analysis, sarcoidosis was associated with incident cirrhosis (subdistribution hazard ratio [sHR] 2.84, 95% confidence interval [CI] 1.97–4.11, p < 0.001); the weighted 20‐year cumulative incidence was 2.8% versus 1.1%. The association was consistent across IPTW‐weighted cause‐specific, covariate‐adjusted and follow‐up‐truncated analyses (hazard ratio [HR]/sHR range 2.71–2.84). Among patients with confirmed organ involvement, 1057 had single‐organ and 87 had multi‐organ disease. In the 5‐year landmark analysis ( n = 940; 18 events), multi‐organ involvement was associated with cirrhosis in the parsimonious model (HR 5.33, 95% CI 1.90–15.00, p = 0.002). No immunomodulatory therapy was significantly associated with cirrhosis. Conclusions Sarcoidosis was independently associated with increased cirrhosis risk despite no hepatic involvement, particularly in multiorgan disease. This supports a risk‐stratified spectrum from heightened clinical awareness to closer hepatic surveillance in selected patients.