Background: Numeracy and Adjuvant! are 2 web-based calculators that are used widely to estimate the prognosis and potential benefit of adjuvant 5-fluorouracil (5-FU)-based therapy for patients with stage II and III colon cancer. In this study, the authors compared the predicted survival estimates from these models with the actual observed estimates in independent datasets that were derived from a population cohort and from clinical trials.
Methods: The population cohort was derived from the British Columbia Colorectal Cancer Outcomes Unit database, which identified referred patients with stage II and III colon cancer from 1995 to 1996 and from 1999 to 2003. Patients who were enrolled in North Central Cancer Trials Group (NCCTG) trials NCCTG 94651 and NCCTG 914653 were included in the trials dataset. Patient and disease data were used to predict 5-year relapse-free and overall survival using both tools.
Results: In the population-based dataset (N = 2033), Adjuvant! offered more reliable predictions of prognosis for patients who underwent surgery alone, but it had reliability similar to that of Numeracy for predicting the prognosis for patients who received adjuvant 5-FU. Both models tended to overestimate survival for patients with stage II disease who received 5-FU. In the trials dataset of patients who underwent and received 5-FU (N = 1729), Numeracy and Adjuvant! demonstrated similar performance and improved correctness.
Conclusions: This independent validation analysis demonstrated that both Numeracy and Adjuvant! had similar predictive performance and acceptable reliability for patients with stage III disease. Survival outcomes of patients with stage II colon cancer who received adjuvant 5-FU were slightly lower than estimated by either model.
Copyright © 2011 American Cancer Society.