Work

Reply

Year: 2022

Type: letter

Abstract: We thank Drs. Sachan and Kushwah for their interest in our study evaluating GALAD for early detection of HCC.[1] Their letter raises some points that warrant further discussion. We agree that ultrasound alone has suboptimal accuracy for HCC surveillance when used alone and its performance is improved in combination with alpha‐fetoprotein (AFP). However, they cite a meta‐analysis in which the two tests had a sensitivity and specificity for resectable HCC of 89% and 87%, respectively—markedly higher than observed in many practice settings. Indeed, another recent meta‐analysis reported the sensitivity and specificity of ultrasound with AFP for early‐stage HCC (defined by Milan Criteria or BCLC stage A) were substantially lower at 63% and 84%, respectively.[2] Although we caution against cross‐study comparisons of test performance given differences in patient populations, the performance of GALAD appears promising compared to these historical thresholds. In our cohort, GALAD achieved a sensitivity and specificity of 74% and 86.5% for early‐stage HCC when assessed within 6 months before diagnosis. The authors discuss screening‐related harms from false‐positive results, but specificity of GALAD in our cohort was equal to, if not higher than, that of ultrasound with AFP. Further, specificity does not necessarily equate to screening‐related harms, given that screening harms are often lower with biomarker‐based surveillance than ultrasound‐based surveillance given the incorporation of longitudinal changes in biomarker values to inform provider decision making.[3] Finally, we note that when incorporating patient preferences for surveillance parameters, at‐risk patients prioritize sensitivity over screening‐related harms.[4] The authors also raise the issues of cost, particularly in resource‐poor settings. However, cost‐effectiveness of biomarker‐based surveillance is more complicated than adding costs of the biomarkers. One must account for differences in other factors, including: (1) surveillance adherence, which would presumably be higher for biomarkers than ultrasound; (2) diagnostic testing performed for false‐positive results, which we contend would be less with biomarker‐based surveillance, as discussed above; and (3) early detection and quality‐adjusted life‐years. With increasing proportions of HCC related to alcohol‐associated disease and NAFLD, ultrasound‐based surveillance will be more prone to poor visualization and surveillance failure, whereas performance of biomarkers, including GALAD, appears preserved.[5] Dedicated cost‐effectiveness analyses are required to address this question. Finally, the authors cite a study in which PIVKA‐II and AFP had a sensitivity of 75% for HCC detection in HBV‐infected patients, suggesting AFP‐L3 may not be needed; however, less than one third of patients had early‐stage HCC, thereby overestimating biomarker performance. Several other studies demonstrate GALAD’s promising performance for early‐stage HCC detection, including early results from cohort studies such as ours.[6] Although GALAD still requires validation in well‐powered, large phase III cohorts before routine use in clinical practice, available data suggest that GALAD is promising to fill the proposed tenets of availability, affordability, and accuracy. CONFLICT OF INTEREST Amit Singal has served as a consultant or on advisory boards for Bayer, FujiFilm Medical Sciences, Exact Sciences, Roche, Glycotest, and GRAIL. Neehar Parikh has served as a consultant or on advisory boards for Bayer, FujiFilm Medical Sciences, Exact Sciences, and Freenome. He has received research funding from Exact Sciences, Glycotest, Bayer, and Target RWE. AUTHOR CONTRIBUTIONS Conception and interpretation of data: All authors; Drafting of the manuscript: Amit G. Singal; Critical revision of the manuscript for important intellectual content: All authors; Obtained funding: Amit G. Singal.

Source: Hepatology

DOI: 10.1002/hep.32462

Authors: Amit G. Singal, Neehar D. Parikh

Institutions: The University of Texas Southwestern Medical Center, University of Michigan

Language: English

Topic: Hepatocellular Carcinoma Treatment and Prognosis

Subfield: Hepatology

Field: Medicine

Domain: Health Sciences

SDG: Good health and well-being

Open Access status: green

Funder: National Cancer Institute