Top Line: What factors are predictive of recurrence risk among patients with colon cancer?
The Study: Many of us use the MSK nomograms to help individualize estimates of recurrence risk for various malignancies. This paper outlines development and external validation of an updated nomogram to predict recurrence using data from >1000 patients with stage I-III colon cancer (sadly excluding rectal cancers). Development included standard clinical variables in addition to newer molecular and histopathologic characteristics. Five variables drove the final model: T stage, number of involved nodes, VELPI (venous, lymphatic, or perineural invasion), TIL (tumor infiltrating lymphocytes), and receipt of adjuvant chemotherapy. This model was more concordant with 5-year freedom from recurrence than the 8th edition AJCC staging system. In addition, it provides a continuous (rather than discrete) estimate of recurrence risk based on patient-specific variables. A few important factors didn’t make the cut including tumor location (i.e. right side vs. left side), number of nodes examined, and preoperative CEA level. The first seemed to have a less strong association with recurrence than when used for metastatic disease. With respect to nodes, it’s the number of positive nodes, not just the number dissected, that’s truly predictive of outcome.
TBL: This updated nomogram performs better than AJCC staging at predicting the risk of recurrence for patients with stage I-III colon cancer. | Weiser, J Clin Oncol 2021