Feasibility of Personalized and Tumor-Informed Circulating Tumor DNA Assay for Early Recurrence Detection in Patients With Hepatocellular Carcinoma
Maen Abdelrahim, Alejandro Mejia, Abdullah Esmail, Juan Carlos Barrera Gutierrez, Mahmoud Ouf, Joseph W. Franses, Irun Bhan, Sudha Kodali, Ashish Saharia, Amit Mahipal, Nikolas Naleid, Catherine Bridges, Antony Tin, Chris M. Brewer, Vasily N. Aushev, Arkarachai Fungtammasan, Charuta C. Palsuledesai, J. Bryce Ortiz, Adham Jurdi, Minetta C. Liu, R. Mark Ghobrial, Aiwu Ruth He
Abstract
PURPOSE Hepatocellular carcinoma (HCC) has high relapse rates after standard-of-care (SOC) resection or liver transplantation (LT). We evaluated the utility of circulating tumor DNA (ctDNA) to predict relapse/progression risk in patients with HCC. Materials and Methods This retrospective analysis examined real-world data from ctDNA testing on 125 patients with HCC (721 plasma samples) undergoing curative-intent treatments and SOC management. Patients were divided into four subcohorts: cohort A (n = 64) and B (n = 52) comprised patients under recurrence monitoring after LT or resection, respectively. Cohort C (n = 4) and D (n = 5) comprised patients under treatment response monitoring with known recurrence or inoperable disease, respectively. A personalized, tumor-informed 16-plex polymerase chain reaction next-generation sequencing assay (Signatera, Natera, Inc, Austin, TX) was used for ctDNA testing. The molecular residual disease (MRD) window was defined as 2-12 weeks post-LT/resection (cohorts A/B), before starting adjuvant therapy (AT). Surveillance window was defined as post-MRD window or 2 weeks post-AT (cohort B) or during ongoing treatment (cohorts C/D). RESULTS The median follow-up was 40 (1.5-60) months. In cohort A, 97.2% (35/36) of patients with ctDNA negativity in the MRD window remained negative during surveillance. In cohort B, ctDNA was detected in 29.4% (10/34) of patients within the MRD window, all of whom experienced clinical recurrence (hazard ratio [HR], 7.2 [95% CI, 2.6 to 20]; P < .0001). In the surveillance window (cohort B), the ctDNA detection rate was 32.3% (10/31), and all experienced recurrence (HR, 18.0 [95% CI, 3.9 to 85]; P < .0001). In cohorts C/D, on-treatment ctDNA dynamics were concordant with treatment response as measured by imaging. Compared with alpha-fetoprotein, ctDNA had higher sensitivity and a significantly longer lead time (7.9 v 2.2 months) for recurrence detection. CONCLUSION Serial ctDNA testing effectively identified HCC recurrence early, postresection and post-LT. ctDNA was also useful for treatment response monitoring and could help resolve ambiguous imaging results.