RESUMO
INTRODUCTION: A subset of patients with celiac disease (CeD) has liver involvement in the form of hypertransaminasemia, liver cirrhosis, and autoimmune hepatitis. We conducted a systematic review with meta-analyses to determine the pooled prevalence of CeD in patients with cryptogenic cirrhosis, all-cause cirrhosis, cryptogenic hypertransaminasemia, and all-cause hypertransaminasemia. METHODS: We searched PubMed and EMBASE up to January 2022. Cross-sectional, case-control, and prospective cohort studies performing serological tests and/or intestinal biopsy for CeD on patients with cryptogenic cirrhosis, all-cause cirrhosis, cryptogenic hypertransaminasemia, and all-cause hypertransaminasemia were included to calculate pooled estimates of seroprevalence and the prevalence of biopsy-confirmed CeD in these 4 groups. RESULTS: Of 6,871 articles screened, 20 articles were included finally in 3 meta-analyses for cryptogenic cirrhosis, all-cause cirrhosis, and cryptogenic hypertransaminasemia. For the all-cause hypertransaminasemia group, a qualitative review of 4 studies was conducted instead of a meta-analysis due to significant differences in studies. The pooled prevalence (95% confidence interval) of biopsy-confirmed CeD in cryptogenic cirrhosis was 4.6% (2.2%-7.5%) while the pooled prevalence of biopsy-confirmed CeD in all-cause cirrhosis was 0.8% (0%-3.4%). The pooled prevalence of biopsy-confirmed CeD in cryptogenic hypertransaminasemia was 5.7% (3.2%-8.8%). DISCUSSION: Nearly 1 in 20 patients each with cryptogenic cirrhosis and cryptogenic hypertransaminasemia have CeD; hence, they should both be considered high-risk groups for CeD. While the prevalence of CeD in those with all-cause cirrhosis is similar to that in general population, it may be worth screening them for CeD because liver pathology has the potential for reversal in them.