As reported in the Journal of Clinical Oncology, Weiser et al have developed a "third-generation" clinical calculator at Memorial Sloan Kettering Cancer Center (MSK), which incorporates molecular and clinicopathologic characteristics and provides an accurate prediction of disease recurrence following curative resection of stage I to III colon cancer.
Development of the calculator involved prospectively collected data from 1,095 patients who underwent colectomy between 2007 and 2014 at MSK. Variables tested in development consisted of:
The calculator was tested for prediction of 5-year freedom from recurrence by discrimination (concordance index) and calibration curves (with perfect calibration indicated by alignment between predicted and actual probabilities of freedom from recurrence along the 45° diagonal). External validation of the calculator was performed in a cohort of 299 patients from the Washington University Siteman Cancer Center.
Among 1,095 patients in the MSK cohort, 120 experienced disease recurrence, with a 5-year freedom from recurrence rate of 85%.
The final clinical calculator incorporated six variables: microsatellite genomic phenotype; AJCC T category; number of tumor-involved lymph nodes; presence of high-risk pathologic features such as venous, lymphatic, or perineural invasion; presence of tumor-infiltrating lymphocytes; and use of adjuvant chemotherapy.
The calculator provides 5-year freedom from recurrence estimates based on a Kaplan-Meier curve for patients with tumors with MSI or a nomogram for patients with tumors with MSS or untested microsatellite status.
The concordance index of the clinical calculator for predicting 5-year freedom from recurrence in the MSK cohort was 0.792 (95% confidence interval [CI] = 0.749–0.837) compared with 0.708 (95% CI = 0.671–0.745) for the AJCC 5th edition staging system, 0.757 (95% CI = 0.715–0.799) for the AJCC 8th edition staging system, and 0.756 (95% CI = 0.707–0.805) for the updated (2019) MSK 2008 nomogram.
Among 299 patients in the external validation cohort, 46 experienced disease recurrence, with a 5-year freedom from recurrence rate of 80%. The concordance index for predicting 5-year freedom from recurrence was 0.738 (95% CI = 0.703–0.811). The plot of predicted vs observed freedom from recurrence approached a 45° diagonal, with the calculator overestimating freedom from recurrence for the highest-risk patients.
The investigators concluded, “This third-generation clinical calculator for predicting cancer recurrence following curative colectomy successfully incorporates microsatellite genomic phenotype and the presence of tumor-infiltrating lymphocytes, resulting in improved discrimination and predictive accuracy. This exemplifies an evolution of a clinical calculator to maintain relevance by incorporating emerging variables as they become validated and accepted in the oncologic community.”
Martin R. Weiser, MD, of Memorial Sloan Kettering Cancer Center, is the corresponding author for the Journal of Clinical Oncology article.
Disclosure: The study was supported by the National Cancer Institute and the John and Michelle Martello Research Fund. For full disclosures of the study authors, visit ascopubs.org.The content in this post has not been reviewed by the American Society of Clinical Oncology, Inc. (ASCO®) and does not necessarily reflect the ideas and opinions of ASCO®.