Purpose The role of tumor-infiltrating lymphocytes (TILs) in predicting lymph node metastasis (LNM) in patients with T1 colorectal cancer (CRC) remains unclear. Furthermore, clinical utility of a machine learning–based approach has not been widely studied.Materials and Methods Immunohistochemistry for TILs against CD3, CD8, and forkhead box P3 in both center and invasive margin of the tumor were performed using surgically resected T1 CRC slides. Three hundred and sixteen patients were enrolled and categorized into training (n=221) and validation (n=95) sets via random sampling. Using clinicopathologic variables including TILs, the least absolute shrinkage and selection operator (LASSO) regression model was applied for variable selection and predictive signature building in the training set. The predictive accuracy of our model and the Japanese criteria were compared using area under the receiver operating characteristic (AUROC), net reclassification improvement (NRI)/integrated discrimination improvement (IDI), and decision curve analysis (DCA) in the validation set.Results LNM was detected in 29 (13.1%) and 12 (12.6%) patients in training and validation sets, respectively. Nine variables were selected and used to generate the LASSO model. Its performance was similar in training and validation sets (AUROC, 0.795 vs. 0.765; p=0.747). In the validation set, the LASSO model showed better outcomes in predicting LNM than Japanese criteria, as measured by AUROC (0.765 vs. 0.518, p=0.003) and NRI (0.447, p=0.039)/IDI (0.121, p=0.034). DCA showed positive net benefits in using our model.Conclusion Our LASSO model incorporating histopathologic parameters and TILs showed superior performance compared to conventional Japanese criteria in predicting LNM in patients with T1 CRC.

We investigated the clinical significance of a vascular growth pattern of hepatocellular carcinoma (HCC), the vessels that encapsulate tumor clusters (VETC), previously linked to HCC metastatic dissemination. VETC was assessed in a large multi‐institutional cohort of 541 resected HCCs from Italy, Korea and Japan, and matched against a full spectrum of clinical and pathological variables. The VETC phenotype (defined as ≥ 55% tumor area by CD34 immunostaining) was easily reproducible and reliably detectable in whole sections and small‐sized tissues of tissue microarray. VETC HCCs represented 18.9% of the whole series, the lowest proportion occurring in the cohort with smallest tumors (8.7%, Japanese series). VETC was significantly associated with several clinical and pathological features such as high alfa‐fetoprotein (AFP) level, tumor size greater than 5 cm, poor differentiation, macrotrabecular pattern, less compact pattern, less inflammatory infiltrates, and frequent microvascular invasion. VETC was associated with early recurrence (hazard ratio [HR]: 1.52 [1.06‐2.19], P = 0.023), disease‐free survival (HR: 1.66 [1.21‐2.27], P = 0.002), and overall survival (HR: 2.26 [1.37‐3.72], P = 0.001) at multivariable analysis. VETC affected the survival in HCC patients stratified for etiology (hepatitis C virus/hepatitis B virus), vascular invasion, and specific molecular phenotypes (β‐catenin/GS+). This distinct vascular pattern was enriched in the recently reported macrotrabecular massive HCC subtype, which was seen in 7.8% (42 of 541) of patients and associated with high AFP levels and poor differentiation. Conclusion: The VETC pattern was found to be easily detectable in a consistent fraction of HCC and a powerful pathological finding affecting survival. This study suggests that the heterogeneous pattern of angiogenesis is involved in HCC behavior.