Deep Learning Prediction of Metastasis in Locally Advanced Colon Cancer Using Binary Histologic Tumor Images.

In this study, we developed the Binary ImaGe Colon Metastasis classifier (BIg-CoMet), a semi-guided approach for the stratification of colon cancer patients into two risk groups for the occurrence of distant metastasis, using an InceptionResNetV2-based deep learning model trained on binary images. We enrolled 291 colon cancer patients with pT3 and pT4 adenocarcinomas and converted one cytokeratin-stained representative tumor section per case into a binary image. Image augmentation and dropout layers were incorporated to avoid overfitting. In a validation collective (n = 128), BIg-CoMet was able to discriminate well between patients with and without metastasis (AUC: 0.842, 95% CI: 0.774-0.911). Further, the Kaplan-Meier curves of the metastasis-free survival showed a highly significant worse clinical course for the high-risk group (log-rank test: p < 0.001), and we demonstrated superiority over other established risk factors. A multivariable Cox regression analysis adjusted for confounders supported the use of risk groups as a prognostic factor for the occurrence of metastasis (hazard ratio (HR): 5.4, 95% CI: 2.5-11.7, p < 0.001). BIg-CoMet achieved good performance for both UICC subgroups, especially for UICC III (n = 53), with a positive predictive value of 80%. Our study demonstrates the ability to stratify colon cancer patients via a semi-guided process on images that primarily reflect tumor architecture.

Stroma AReactive Invasion Front Areas (SARIFA)-A New Easily to Determine Biomarker in Colon Cancer-Results of a Retrospective Study.

Many studies have used histomorphological features to more precisely predict the prognosis of patients with colon cancer, focusing on tumor budding, poorly differentiated clusters, and the tumor-stroma ratio. Here, we introduce SARIFA: Stroma AReactive Invasion Front Area(s). We defined SARIFA as the direct contact between a tumor gland/tumor cell cluster (≥5 cells) and inconspicuous surrounding adipose tissue in the invasion front. In this retrospective, single-center study, we classified 449 adipose-infiltrative adenocarcinomas (not otherwise specified) from two groups based on SARIFA and found 25% of all tumors to be SARIFA-positive. Kappa values between the two pathologists were good/very good: 0.77 and 0.87. Patients with SARIFA-positive tumors had a significantly shorter colon-cancer-specific survival (p = 0.008, group A), absence of metastasis, and overall survival (p < 0.001, p = 0.003, group B). SARIFA was significantly associated with adverse features such as pT4 stage, lymph node metastasis, tumor budding, and higher tumor grade. Moreover, SARIFA was confirmed as an independent prognostic indicator for colon-cancer-specific survival (p = 0.011, group A). SARIFA assessment was very quick (<1 min). Because of low interobserver variability and good prognostic significance, SARIFA seems to be a promising histomorphological prognostic indicator in adipose-infiltrative adenocarcinomas of the colon. Further studies should validate our results and also determine whether SARIFA is a universal prognostic indicator in solid cancers.