The use of simple tests to predict biopsy-proven steatohepatitis in people with type 2 diabetes

Metabolic-associated steatotic liver disease (MASLD) has become the most common chronic liver disease, especially in people with type 2 diabetes mellitus (T2DM). Liver biopsy remains the gold standard method for diagnosis of MASLD subtypes, but prevalences may be under or overestimated when biopsy is performed with selection bias. The aims of this study were to define prevalence of MASLD subtypes by liver biopsy in T2DM participants not selected by abnormal exams, determine variables associated with metabolic-associated steatohepatitis (MASH), and analyze performance of aminotransferases and abdominal ultrasound in diagnosis.

T2DM participants from 18 to 70 years were considered for enrollment. Of the 396 participants, 85 were included and submitted to clinical, laboratory examinations, and ultrasound. Eighty-three performed liver biopsy evaluated by two independent pathologists. Factors independently associated to MASH and significant fibrosis were assessed by hierarchical multivariate logistic regression.

Prevalence of MASLD was 92% (50% simple steatosis, 42% MASH) and kappa = 0.78. Steatosis was mild in 76% of participants with simple steatosis and severe in 65% of MASH (P < 0.001). Presence of MASH or fibrosis was associated with BMI and alanine aminotransferase (ALT) [threshold of 33.5 mg/dl in predicting MASH (area under the curve = 0.82, P = 0.001)].

Prevalence of MASLD by liver biopsy in T2DM regardless of ultrasound or ALT elevation is almost 100%, with 42% of MASH. MASH was associated to severe steatosis on histology. BMI and ALT were independently associated with MASH and ALT close to the upper limit of normal gave the best cutoff point for MASH detection.