背景：现在，解剖切除（AR）对肝细胞癌（HCC）的疗效存在争议。这项研究的目的是在我们的队列中检查它，并检测AR的最佳指标。
方法：本研究包括2000年至2019年接受肝切除术的656例符合米兰标准的原发性肝癌患者。我们的队列分为AR（n=378）和非解剖切除（NAR）（n=278）组，并进行1:1倾向评分匹配(PSM)以最大限度地减少潜在混杂因素的影响.无复发生存率（RFS），总生存期（OS），检查AR的术前指标。
结果：每组210例患者匹配良好，术前混杂因素在两组间平衡。PSM前(RFS;HR=0.89P=0.25,OS;HR=1.08P=0.64)和后(RFS;HR=0.93P=0.60,OS;HR=1.07P=0.75)两组间的RFS和OS差异均无统计学意义。亚组分析显示，在甲胎蛋白（AFP-L3）岩藻糖基化分数>10％且分化较差的情况下，观察到AR的生存改善作用（P为相互作用<0.05）。此外，logistic回归分析显示,术前AFP-L3>10%是低分化的独立预测因子(OR=2.58,P=0.03)。
结论：在米兰标准范围内，AR对原发性肝癌患者的疗效未显示。但建议AFP-L3>10％可能是米兰标准内HCCAR的术前指标。
BACKGROUND: Still now, the efficacy of anatomic resection (AR) for hepatocellular carcinoma (HCC) is controversial. The aim of this study is to examine it in our cohort and detect an optimal indicator for AR.
METHODS: The present study included 656 patients with primary HCC within Milan criteria who underwent hepatectomy from 2000 to 2019. Our cohort was divided into AR (n = 378) and non-anatomic resection (NAR) (n = 278) groups, and 1:1 propensity score matching (PSM) was performed to minimize the effect of potential confounders. Recurrence-free survival (RFS), overall survival (OS), and a preoperative indicator for AR were examined.
RESULTS: 210 patients from each group were well-matched, and preoperative confounding factors were balanced between the two groups. There was no significant difference in RFS and OS between the two groups before (RFS; HR = 0.89 P = 0.25, OS; HR = 1.08 P = 0.64) and after PSM (RFS; HR = 0.93 P = 0.60, OS; HR = 1.07 P = 0.75). Subgroup analysis showed that the survival improvement effect of AR was observed in cases with a fucosylated fraction of alfa-fetoprotein (AFP-L3) > 10% and poorly differentiation (P for interaction <0.05). Moreover, the logistic regression analysis showed that preoperative AFP-L3 > 10% was an independent predictor for poorly differentiation (OR = 2.58, P = 0.03).
CONCLUSIONS: The efficacy of AR for patients with primary HCC within Milan criteria was not shown. But it was suggested that AFP-L3 > 10% might be a preoperative indicator of AR for HCC within Milan criteria.
