A multi‐analyte cell‐free DNA–based blood test for early detection of hepatocellular carcinoma

Abstract

Abstract The limited performance of guideline‐recommended abdominal ultrasound and serum alpha‐fetoprotein (AFP) highlights the urgent, unmet need for new biomarkers for more accurate detection of early hepatocellular carcinoma (HCC). To this end, we have conducted a prospective clinical validation study to evaluate the performance of the HelioLiver Test, a multi‐analyte blood test combining cell‐free DNA methylation patterns, clinical variables, and protein tumor markers. A blinded, multicenter validation study was performed with 247 subjects, including 122 subjects with HCC and 125 control subjects with chronic liver disease. The performance of the HelioLiver Test was compared with AFP and the GALAD score as established HCC surveillance blood tests. The performance of the HelioLiver Test (area under the receiver operating characteristic curve [AUROC] = 0.944) was superior to both AFP (AUROC = 0.851; p < 0.0001) and GALAD (AUROC = 0.899; p < 0.0001). Using a prespecified diagnostic algorithm, the HelioLiver Test showed sensitivities of 85% (95% confidence interval [CI], 78%–90%) for HCC of any stage and 76% (95% CI, 60%–87%) for early stage (American Joint Committee on Cancer [AJCC] I and II) HCC. In contrast, AFP (≥20 ng/mL) alone and the GALAD score (≥−0.63) showed lower sensitivities of 62% (95% CI, 54%–70%) and 75% (95% CI, 67%‐82%) for HCC overall, and 57% (95% CI, 40%–71%) and 65% (95% CI, 49%–79%) for early stage (AJCC I and II) HCC, respectively. The specificities of the HelioLiver Test (91%; 95% CI, 85%–95%), AFP (97%; 95% CI, 92%–99%), and the GALAD score (94%; 95% CI, 88%–97%) were similar for control subjects. The HelioLiver Test showed superior performance for HCC detection compared to with both AFP and the GALAD score and warrants further evaluation in HCC surveillance settings.