Photo Credit: Natali_Mis

A recent study compared the predictive performance of three imaging models in order to determine the most accurate noninvasive method for diagnosing MASH.

Among patients with metabolic dysfunction-associated steatotic liver disease (MASLD), metabolic dysfunction-associated steatohepatitis (MASH) is a progressive stage of the disease that increases the risk of serious complications such as cirrhosis, hepatocellular carcinoma, and liver-related mortality. Early identification and intervention in treating MASH could improve prognosis in MASLD patients. Currently, guidelines recommend liver biopsy as the gold standard for identifying and diagnosing MASH. However, this screening pathway is invasive, raises potential complications, opens the door to sample variability, and is costly. The consideration of alternative, less invasive screening methods would, therefore, significantly contribute to improving care going forward.

To this end, Jing Zhang, MD, and colleagues developed a study to compare three noninvasive methods of detecting MASH and patients at risk for MASH. As Dr. Zhang and colleagues wrote in a study published in the Journal of Gastroenterology and Hepatology, “This study aimed to explore the potential performance of imaging-based models in identifying patients with MASH and ‘at-risk’ MASH and to attempt to differentiate the whole course of MASLD.”

Noninvasive Models

The three models chosen for examination in the study were:

• MRI aspartate aminotransferase (AST [MAST]) score: combines MRI-proton density fat fraction (PDFF), magnetic resonance elastography (MRE), and AST, a noninvasive blood panel.
• FibroScan-aspartate aminotransferase (FAST) score. FAST is another composite score that includes controlled attenuation parameter (CAP), and liver stiffness measurement (LSM) measured by vibration-controlled transient elastography (VCTE) and AST.
• MRE plus fibrosis-4 (FIB [MEFIB]): combines MRE and FIB, a noninvasive blood panel that estimates the level of liver fibrosis.

The study included 108 patients with biopsy-confirmed MASLD in the study cohort. The median age of the patient group was 38 years, and 55.6% were male. Of the patients studied, 64.8% had MASH, and 25.9% were deemed as being at risk for MASH. In terms of the model scores registered, the median MAST score was 0.09 (95% CI, 0.04-0.21), the median FAST score was 0.58 (95% CI, 0.39-0.73), the median MRE score was 2.76 (95% CI, 2.45-3.10), and the median FIB-4 score was 0.91 (95% CI, 0.62-1.42).

Predictive Performance and Accuracy

In examining the predictive performance of the three models for MASH detection, MAST’s area under the curve (AUC) value was 0.803 (95% CI, 0.719-0.886). FAST’s AUC value was 0.799 (95% CI, 0.707-0.891). MEFIB’s AUC value was 0.671 (95% CI, 0.571-0.772).

Regarding diagnostic accuracy, the researchers recalibrated the optimal rule-out and rule-in cut-off for the three models to determine sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV). MAST showed a PPV of 88.6%, FAST was 87.8%, and MEFIB was 83.3%. MAST showed a NPV of 69.6%, FAST was 76.0%, and MEFIB was 55.2%. MAST placed 46.3% in the gray zone, and MEFIB placed 45.4% compared to FAST, which placed 38.9% in the gray zone (P=0.271 for MAST vs FAST and P=0.891 for MAST vs MEFIB). The number of patients accurately identified as MASH was higher when the researchers applied FAST (50.93%) and MAST (43.52%) compared with MEFIB (37.96%).

Patients at Risk for MASH

The ability to predict which patients are at risk for MASH was also examined. The AUC value for MAST was 0.810 (95% CI, 0.719-0.900); for FAST, it was 0.782 (95% CI, 0.689-0.874); and for MEFIB, it was 0.729 (95% CI, 0.619-0.838).

As with the examination of diagnostic accuracy in the identification of MASH, the researchers recalibrated the optimal rule-out and rule-in cut-off for the three models to determine sensitivity, specificity, PPV, and NPV for patients at risk for MASH. MAST showed a PPV of 63.6%, FAST was 48.5%, and MEFIB was 87.5%. As for NPV, MAST was 89.2%, FAST was 100.0%, and MEFIB was 84.9%. Regarding the gray zone, MAST was 11.1%, MEFIB was 25%, and FAST was 49.1% (P<0.001 between MAST vs FAST and MAST vs MEFIB). MAST correctly identified at-risk MASH patients at a rate of 74.07%, MEFIB at 63.89%, and FAST at 33.33%.