Discordance Between Radiologic and Pathologic Response After Preoperative Therapy for Hepatocellular Carcinoma: A Systematic Review and Meta-Analysis

Background & Aims: Radiologic-pathologic discordance remains a significant challenge in managing hepatocellular carcinoma (HCC) after conversion therapy. With the shift from locoregional therapy (LRT) to immunotherapy-based (IO) regimens, the accuracy of mRECIST has been called into question. We performed a systematic review and meta-analysis to quantify discordance rates and diagnostic performance across different treatment eras.Methods: We searched PubMed, Embase, and Cochrane for studies (2011-2026) reporting both mRECIST evaluation and pathologic response (pCR/major necrosis) in HCC. A random-effects model was used to pool discordance rates. Diagnostic accuracy was assessed via sensitivity, specificity, and Area Under the Summary Receiver Operating Characteristic (SROC) curve.Results: Twenty unique studies (N=3,462) were included. The overall pooled discordance rate was 27.8% (95% CI: 24.1–31.5%). Subgroup analysis revealed a significant era-dependent shift: discordance was significantly higher in the IO-based group (29.4%) compared to the LRT/TACE group (18.2%). While mRECIST showed high overall specificity (0.84), its sensitivity was significantly lower in the IO subgroup (0.54 vs. 0.74 in LRT; p < 0.001), primarily due to massive immune cell infiltration mimicking viable tumor (false negatives). The pooled AUC was 0.78. No significant publication bias was detected (p = 0.42).Conclusions: Radiologic-pathologic discordance has nearly doubled in the immunotherapy era. mRECIST significantly underestimates pathologic response in almost half of IO-treated responders. These findings suggest that stable disease on imaging should not be a contraindication for surgery, and there is an urgent need for adjunctive biomarkers to refine surgical decision-making in the era of modern conversion therapy.