Text

Background and objective: Resmetirom was recently approved in Europe as the first pharmacologic therapy for patients with metabolic dysfunction–associated steatohepatitis (MASH) and fibrosis stages F2–F3. Recently, a combination of NITs has been proposed to identify treatment-eligible patients, including controlled attenuation parameter (CAP) ≥280 dB/m, aspartate transferase (AST)> 17 U/L for women and >20 U/Lfor men, liver stiffness measurement (LSM) 10–20 kPa, and platelet count ≥140 × 10³/µL after exclusion of cirrhosis. In the present study, we aimed to evaluate this algorithm and compare its diagnostic performance across different centers using liver biopsy as the reference standard.

Method: Data were obtained from consecutive patients with biopsy-confirmed MASLD across four tertiary care centers. Liver biopsy specimens were histologically evaluated using the Steatosis, Activity, Fibrosis / Fatty Liver Inhibition of Progression (SAF/FLIP) algorithm and the NASH Clinical Research Network (NASH-CRN) scoring system. Patients considered histologically eligible were those with MASH (NAS ≥ 4) and fibrosis stages F2–F3. Agreement between the diagnostic tests and liver biopsy findings was assessed using Cohen’s kappa statistic. The included centers were as follows: Center 1 in Germany; Center 2 and 3in the United States and Center 4 in Türkiye.

Result: Prospectively collected data from 949 biopsy-proven MASLD patients were included in the analysis and retrospectively analyzed (Age: 55 [19-84] years, 353 males (37.2%)). A total of 322 patients (33.9%) met the histological criteria for resmetirom use and 308 patients (32.4%) the NIT criteria. Cohen’s kappa indicated poor diagnostic performance across all centers, with values of 0.069, 0.044, 0.410, and 0.162, respectively, showing a slightly better performance for the second US center. The combination of the NIT criteria was able to identify the patients with accurate histological features with a sensitivity of 18%, 69%, 90%, 44% and specificity of 89%, 36%. 52%, 72%, respectively (positive predictive value: 50%, 75%, 63%, 59%; negative predictive value: 62%, 29%, 85% and 58%, respectively). Similarly, the NIT criteria had poor performance in predicting the presence of F2-F3 on biopsy regardless of the presence of MASH (Kappa: -0.019, -0.009, 0.412, 0.164, respectively).

Summary: The NIT criteria demonstrated poor performance compared with histological criteria, with only minor differences observed across countries.