For patients with nonalcoholic steatohepatitis (NASH), the GALAD score may accurately predict hepatocellular carcinoma (HCC) as early as 560 days before diagnosis, according to investigators.
The GALAD score, which combines sex, age, alpha-fetoprotein-L3 (AFP-L3), alpha-fetoprotein, and des-gamma-carboxyprothrombin (DCP), could improve cancer surveillance among NASH patients whose obesity limits sensitivity of ultrasound, reported lead author Jan Best, MD, of the University Hospital Magdeburg in Germany, and colleagues.
“The limitations of ultrasound surveillance alone for early detection of HCC are particularly evident in patients with NASH,” the investigators wrote in Clinical Gastroenterology and Hepatology. “Serum-based biomarkers might be more effective, with or without ultrasound surveillance, for HCC surveillance in NASH patients, although data in this patient population are currently lacking. The current study assessed the performance of the GALAD score for early HCC detection in patients with NASH-related liver disease.”
The study consisted of two parts: first, a retrospective case-control analysis, and second, a phase 3 prospective trial that implemented the GALAD score in a real-world population.
The retrospective component of the study involved 126 NASH patients with HCC (cases) and 231 NASH patients without HCC (controls), all of whom were treated at eight centers in Germany. The median GALAD score was significantly higher among NASH patients with HCC than in those without (2.93 vs. –3.96; P less than .001). At an optimal cutoff of –1.334, the GALAD score predicted HCC with a sensitivity of 91.2% and a specificity of 95.2%. Each component of the GALAD score aligned with previously published findings, as patients with HCC were predominantly older men with elevated serum AFP-L3, AFP, and DCP. But a closer look at the data showed that the GALAD score more accurately predicted HCC than any of its constituent serum measurements in isolation. For any stage of HCC, GALAD had an area under the curve (AUC) of 0.96, compared with significantly lower values for AFP (0.88), AFP-L3 (0.86), and DCP (0.87). Similarly, for early-stage HCC, GALAD score AUC was 0.92, compared with significantly lower values for AFP (0.77), AFP-L3 (0.74), and DCP (0.87).
The accuracy of the GALAD score – for detection of both any-stage and early-stage HCC — remained high regardless of cirrhosis status. Among patients with cirrhosis, the AUC for any-stage HCC was 0.93, and 0.85 for early-stage HCC. For patients without cirrhosis, GALAD was slightly more predictive, based on AUC’s of 0.98 and 0.94 for detection of any-stage and early-stage HCC, respectively. Again, these accuracy values significantly outmatched each serum measurement in isolation.
“These data on NASH-HCC patients demonstrate that GALAD can detect HCC independent of cirrhosis or stage of HCC,” the investigators wrote. “Indeed, even early noncirrhotic NASH-HCC seems clearly separable from NASH controls, as even small groups resulted in robust performance.”
The prospective component of the study involved screening 392 patients with NASH at a single treatment center in Japan. From this cohort, 28 patients developed HCC after a median of 10.1 years. Many patients in this group had significantly higher GALAD scores for 5 or more years before being diagnosed with HCC, and scores rose sharply in the months preceding diagnosis. Depending on selected cutoff value, the GALAD score predicted HCC from 200 to 560 days prior to diagnosis.
“While this specific result has to be confirmed in further prospective studies, it is a promising observation for potential use of GALAD as a screening tool in NASH patients,” the investigators wrote.
“In conclusion, our data confirm that the GALAD score is superior to individual serum markers for detection of HCC in NASH, independent of tumor stage or cirrhosis,” the investigators wrote. “The findings suggest that GALAD should be investigated as a potential tool for screening of NASH individuals to detect HCC at a resectable stage in a sufficiently large prospective study to identify a cutoff.”
The study was funded by Deutsche Forschungsgemeinschaft, the Wilhelm-Laupitz Foundation, and the Werner Jackstaedt Foundation. The investigators declared no conflicts of interest.
SOURCE: Best J et al. Clin Gastro Hepatol. 2019 Nov 8. doi: 10.1016/j.cgh.2019.11.012.