Integration of PNPLA3 and TM6SF2 genotypes provides incremental improvement in advanced fibrosis prediction among MASLD patients with type 2 diabetes mellitus.

BACKGROUND & AIMS: Genetic information is not yet used for the clinical diagnosis of advanced fibrosis in patients with metabolic dysfunction-associated steatotic liver disease (MASLD). Herein, we investigated whether incorporating genetic information regarding PNPLA3 and TM6SF2 into existing non-invasive fibrosis scoring systems could improve their predictive accuracy, particularly in patients with type 2 diabetes mellitus (T2DM), a high-risk population for MASLD-related complications.

METHODS: Data were collected from a cohort of 637 patients with biopsy-proven MASLD. All participants underwent liver stiffness measurement (LSM), serum marker analysis, and genotyping for PNPLA3 (rs738409), TM6SF2 (rs58542926), and other relevant single nucleotide polymorphisms. We evaluated the benefit of adding genetic information to existing non-invasive tests (NITs) - including the Agile 3+, Fibrosis-4 (FIB-4) index, and NAFLD fibrosis score (NFS).

RESULTS: Decision curve analysis in the validation cohort (n = 238) demonstrated that incorporating PNPLA3 and TM6SF2 genetic information marginally enhanced net clinical benefit across all three models over a range of threshold probabilities (10-50%). At a 30% threshold probability, the net benefit of genotype-enhanced models increased from 22.0 to 22.8 per 100 patients for Agile 3+, from 17.0 to 18.4 for NFS, and from 13.0 to 16.9 for FIB-4. In the T2DM subgroup (n = 121), genotype incorporation led to small but statistically significant improvements in discrimination for NFS (AUROC increase: 0.053, p = 0.001) and FIB-4 (AUROC increase: 0.058, p = 0.010), while Agile 3+ showed a favorable trend (AUROC increase: 0.016, p = 0.058).

CONCLUSIONS: Incorporating PNPLA3 and TM6SF2 genetic information into non-invasive fibrosis scoring systems for MASLD provides limited but measurable improvements, with statistically significant AUROC gains for NFS and FIB-4, particularly among patients with T2DM. Further validation is required before routine clinical implementation can be recommended.

IMPACT AND IMPLICATIONS: Our research demonstrates that incorporating PNPLA3 and TM6SF2 genetic information into non-invasive fibrosis tests provides modest but measurable improvements in clinical utility for patients with metabolic dysfunction-associated steatotic liver disease, particularly those with type 2 diabetes mellitus - a high-risk population prone to accelerated fibrosis progression and liver-related complications. These findings are relevant for clinicians managing these patients, as genotype-enhanced models (particularly NAFLD fibrosis score and Fibrosis-4 index) showed statistically significant improvements in diagnostic accuracy, enabling better identification of patients requiring closer monitoring. In practice, genetic information could be integrated into routine risk stratification, especially in settings where elastography is unavailable, allowing serum-based tests to more accurately identify high-risk patients who warrant referral for biopsy or specialist evaluation. However, given the modest absolute improvements and tertiary care derivation, further validation in diverse populations is essential before routine clinical implementation can be recommended.