Exome-based genotype-first reverse phenotyping using structured electronic health record data identifies novel...

Although several SERPINA1 genetic variants have been reported for their pathogenicity to induce liver disorders through phenotype-driven approaches, data regarding genotype-driven approaches of the SERPINA1 locus remain unavailable. This study aimed to characterize the clinical and liver biological profiles of patients harboring nonbenign SERPINA1 variants.

We conducted a retrospective, exome-based genotype-first reverse phenotyping study using structured electronic health record data from consecutive patients from January 1, 2015, to January 31, 2022. Statistical associations were assessed using frequentist and Bayesian models, with validation in the UK Biobank cohort.

Among 1377 patients analyzed, 15 SERPINA1 variants classified as nonbenign were identified in 217 (15.7%) patients. Data were available for 126 patients (median age, 41.5 years; 52.4% male). Liver disease, hyperferritinemia, and pulmonary emphysema were observed in 32.5% (41/126), 23% (29/126), and 5.6% (7/126) of the patients, respectively. The median follow-up duration was 1.3 years and encompassed 1085 biological observations. We confirmed associations with well-documented variants of SERPINA1 (p.Glu366Lys, p.Pro393Ser, p.Ala308Ser, p.Glu288Val, and p.Phe76del). We identified three novel genetic associations with liver markers: c.*10G > A, c.1065 + 10C > T, and p.Arg63Cys. The UK Biobank data confirmed significant gene- and variant-level associations, notably for the variants identified in our study, which ranked in the top decile of statistical associations.

This study supports the utility of a genotype-first approach in characterizing hepatic manifestations of nonbenign SERPINA1 variants. The findings highlight novel genotype-biomarker associations and suggest a role for SERPINA1 genetic testing in patients with unexplained liver abnormalities.