Clinical decision aids in colon cancer: a comparison of two predictive nomograms

Abstract

This study used 2 available nomograms to compare the predicted outcome for patients treated for early colon cancers and compared the predicted and actual outcomes to assess the benefit of the nomograms in the clinical setting. The study included 134 patients. The actual and predicted progression-free survivals were similar, which suggests that this may be used to predict patient’s outcomes. BACKGROUND: The risk of recurrence of colon cancer after curative surgery can be estimated by using decision aids. These aids use pathologic and patient factors to predict recurrence risk after adjuvant chemotherapy and have been validated when using clinical trial populations; however, the performance of 2 decision aids were compared by using a cohort of patients treated at a single center. PATIENTS AND METHODS: Patient data were used to estimate the risk of recurrence when using both the Adjuvant! for colon cancer and Memorial Sloan Kettering Cancer Center (MSKCC) decision aids. A receiver operator characteristic (ROC) curve analyzed the predicted chance of being disease free at 5 years against the actual outcome for each patient. This curve was then used to define cutoff points at a chosen sensitivity and specificity to stratify patients into risk groups, and survival curves for each group calculated. RESULTS: Data on 134 patients were analyzed. The Pearson correlation between the 2 nomograms was 0.848 (P< .01). The ROC curve for the MSKCC nomogram had an area under the curve of 0.638. At a sensitivity and a specificity of 0.8, the MSKCC curve has a risk recurrence score of 69% and 84%, respectively. By using these cutoffs to stratify patients into 3 risk groups, a statistically significant difference in survival was found between high risk and low risk (P = .025). CONCLUSION: Tools to predict risk or recurrence and estimate benefit from therapy may be enhanced in the future by using genetic profiling, but use of existing tools can help deliver a personalized approach to adjuvant therapy.