72 - Target Trial Emulation: A Comparison of Resection, Ablation, and TARE for Early-Stage HCC – a Real-World Analysis

Purpose: Hepatocellular carcinoma (HCC) is increasingly detected at an early stage, but the optimal first-line therapy remains uncertain. We aimed to emulate a target trial to compare outcomes of surgical resection, local ablation, and transarterial radioembolization (TARE) in real-world practice.

Materials and Methods: We conducted a retrospective target trial emulation using data from the Epic EHR system, encompassing over 300 million patient encounters across multiple hospitals. Patients diagnosed with AJCC Stage I HCC between 2016 and 2024 with preserved liver function (no ascites or hepatic encephalopathy, creatinine < 2.0, INR < 1.7, bilirubin < 2.0) were eligible. Only patients whose first-line treatment was resection, ablation, or TARE were included. Inverse probability of treatment weighting (IPTW) was applied to balance baseline covariates, with resection as the comparison group.

Results: A total of 512 patients met eligibility criteria, including 157 who underwent resection, 180 who underwent ablation, and 175 treated with TARE. The median age at diagnosis was 70 years (IQR 64–76; 74% male). Cirrhosis was present in 380 patients, with hepatitis C virus infection (50%) and MASLD (16%) being the most common underlying etiologies. Median overall survival for the entire cohort was 77 months (IQR 62–86 months). When analyzed by treatment type, median survival was 63 months for ablation and 55 months for TARE, while resection did not reach a median survival.
On univariate analyses, both ablation and TARE were associated with worse survival compared to resection. However, this difference was explained by differences in baseline characteristics, as patients treated with ablation or TARE were generally older (median 72 vs. 70 years) and had higher Charlson–Deyo comorbidity scores (median 4 vs. 2). In multivariate models adjusting for these factors, the median survival was 67 months for ablation and 57 months for TARE (p = 0.648). Compared with resection, ablation was associated with a hazard ratio (HR) of 1.86 (95% CI 0.89–3.87, p = 0.097), and TARE with an HR of 2.23 (95% CI 0.96–5.21, p = 0.063). Notably, overall survival did not differ significantly between ablation and resection, resection and TARE, or ablation and TARE (HR for TARE vs. ablation: 1.12, 95% CI 0.54–2.62, p = 0.648).

Conclusion: In this real-world target trial emulation for early-stage HCC (AJCC Stage 1), ablation and TARE achieved similar overall survival to surgical resection. These findings suggest that non-surgical modalities may be appropriate first-line options for selected patients with preserved liver function.