LAUSR

Reply: We appreciate Dr. Xu and colleagues for their interest in our recent manuscript titled ‘‘Subclassification of microscopic vascular invasion in hepatocellular carcinoma.’’ Their letter brings up important issues on microscopic vascular invasion (MiVI) of hepatocellular carcinoma (HCC) in clinical practice. Anatomic liver resection (ALR) is usually recommended in HCC patients with preserved liver function. In HCC with MiVI, more extensive ALR is required to remove the occult intrahepatic metastasis. Previous studies have demonstrated that ALR or major liver resection can improve long-term outcomes in HCC patients with MiVI. In our 200 HCC patients with MiVI, ALR was performed in 126 and non-ALR in 74. However, ALR did not improve disease-free (P 1⁄4 0.590) or overall survival (P 1⁄4 0.262) compared to non-ALR. This result might be attributed to heterogeneity of the tumor stages in our data. More practically important is to predict pathologic MiVI before operation using non-invasive modalities. In addition, it can be helpful to decide the type or extent of liver resection tailored to each individual HCC patient. Recently, we reported that non-boundary gross pattern and high preoperative serum level of desgamma carboxyprothrombin ( 118 mAU/ mL) were independent predictive factors for MiVI, and tumor size 3 cm was an independent prognostic factor for recurrence-free survival in 359 cases of single HCC 5 cm. The high-risk group with 2 or more of these factors showed lower recurrence-free survival than the low-risk group, and major liver resection provided higher recurrence-free survival than minor liver resection in the high-risk group. These factors (gross pattern, tumor size, and serum des-gamma carboxyprothrombin level) can be detected by imaging and serologic studies before operation. Our new scoring system suggests that the type or extent of liver resection can affect oncologic outcomes,