The presence of lymph node metastasis (LNM) affects treatment strategy decisions in T1NxM0 colorectal cancer (CRC), but the currently used clinicopathological-based risk stratification cannot predict LNM accurately. In this study, we detected proteins in formalin-fixed paraffin-embedded (FFPE) tumor samples from 143 LNM-negative and 78 LNM-positive patients with T1 CRC and revealed changes in molecular and biological pathways by label-free liquid chromatography tandem mass spectrometry (LC-MS/MS) and established classifiers for predicting LNM in T1 CRC. An effective 55-proteins prediction model was built by machine learning and validated in a training cohort (N=132) and two validation cohorts (VC1, N=42; VC2, N=47), achieved an impressive AUC of 1.00 in the training cohort, 0.96 in VC1 and 0.93 in VC2, respectively. We further built a simplified classifier with nine proteins, and achieved an AUC of 0.824. The simplified classifier was performed excellently in two external validation cohorts. The expression patterns of 13 proteins were confirmed by immunohistochemistry, and the IHC score of five proteins was used to build an IHC predict model with an AUC of 0.825. RHOT2 silence significantly enhanced migration and invasion of colon cancer cells. Our study explored the mechanism of metastasis in T1 CRC and can be used to facilitate the individualized prediction of LNM in patients with T1 CRC, which may provide a guidance for clinical practice in T1 CRC.
Keywords: T1 colorectal cancer; cancer biology; human; lymph nodes metastasis; machine learning; proteomics.
Most patients with early-stage colorectal cancer can be treated with a minimally invasive procedure. Surgeons use a flexible tool to remove precancerous or cancerous cells, cutting the risk of death from colorectal cancer in half. But a small number of early-stage colorectal cancer patients are at risk of their cancer spreading to the lymph nodes. These patients need more extensive surgery. Clinicians use risk stratification tools to decide which patients need more extensive surgery. Unfortunately, the existing risk stratification tools are not very accurate. The current approach, which analyzes colon tissue for cancerous changes, classifies 70% to 80% of early-stage colorectal cancer patients as high risk for cancer spread. But only about 8% to 16% of patients in the high risk group have lymph node metastasis. As a result, many patients undergo unnecessary, invasive surgery. Zhuang, Zhuang, Chen, Qin, et al. developed a more accurate way to predict which patients are at risk of lymph node metastasis using proteins. In the experiments, the team analyzed the proteins in tumor samples from 143 patients with early colorectal cancer who did not have lymph node metastases and 78 patients with metastases. Zhuang et al. then used machine learning to build a prediction tool that used 55 proteins to identify patients at risk of metastases. The new approach was more accurate than existing tools and simplified versions with only nine or five proteins also performed better than existing tools. This work provides preliminary evidence that protein-based models using as few as five proteins can more accurately identify which patients are at risk of metastasis. These models may reduce the number of patients who undergo unnecessary invasive surgery. The experiments also identified potential targets for therapies to prevent or treat lymph metastases. For example, they showed that low levels of the RHOT2 protein predict metastasis.
© 2023, Zhuang, Zhuang, Chen et al.