Salah A Abdelrahim, Reem M Gahtani, Sameer Alqassimi, Mohammed Alsaweed, Mohammad Miraj, Marya Ahsan, Ayaz Khurram Mallick, Atta Ullah Khan
ctDNA detection was associated with recurrence after curative-intent treatment, but heterogeneity, small-study effects, risk-of-bias concerns, and prospective validation restrict clinical applicability. Routine use is not yet supported.
BACKGROUND: Recurrence after curative-intent treatment remains common in HCC. Circulating tumor DNA (ctDNA) may provide prognostic information, but evidence is heterogeneous.
METHODS: We reviewed 12 cohort studies (880 patients) with HCC treated with curative intent. Associations of preoperative or postoperative ctDNA detection with recurrence-free survival (RFS) or overall survival (OS) were synthesized. Treatment-modality subgroup analyses were performed; assay-platform and sampling-time analyses were not performed because studies were insufficiently comparable. Risk of bias was assessed using QUIPS and QUADAS-2, and certainty using GRADE. Meta-analyses used generic inverse-variance pooling in RevMan 5.4; leave-one-out analyses used log hazard ratios and standard errors in Stata 17.
RESULTS: Preoperative ctDNA detection was associated with worse RFS (HR 3.34, 95% CI 1.99-5.63; 4 studies, 291 patients; moderate certainty). Postoperative ctDNA detection was associated with worse RFS (HR 4.95, 95% CI 2.77-8.83; 11 studies, 799 patients; low certainty; I2 = 79%) and OS (HR 3.07, 95% CI 1.80-5.23; 3 studies, 357 patients; low certainty). Exploratory diagnostic accuracy for subsequent recurrence was modest (8 studies, 508 patients; very-low certainty).
CONCLUSION: ctDNA detection was associated with recurrence after curative-intent treatment, but heterogeneity, small-study effects, risk-of-bias concerns, and prospective validation restrict clinical applicability. Routine use is not yet supported.