Bektas et al., JCO Clin Cancer Inform 2024 • C-index 0.812
N0 — No nodal involvement
Eligibility:AJCC Stage I–III MSI/dMMR colon cancer (cecum to rectosigmoid, >12cm from anal verge) after complete (R0) resection. Excludes: concurrent malignancy, metastatic disease, neoadjuvant therapy.
MSI Predictor
Enter T category and positive node count to compute 3- and 5-year freedom from recurrence.
Guidelines & Evidence
Clinical Details
Section 1
When to Use
When to Use
Estimating individualized 5- and 10-year freedom from recurrence (FFR) after curative resection of non-metastatic (AJCC Stage I–III) colon cancer.
MSI-specific model: predicting 3- and 5-year FFR specifically in patients with microsatellite instability (MSI/dMMR) colon cancer, who have generally favourable outcomes but with a high-risk subgroup.
Counselling patients about adjuvant chemotherapy decisions by providing a more refined recurrence risk than conventional AJCC staging alone.
Clinical trial design: identifying high-risk patients suitable for novel adjuvant immunotherapy or chemotherapy trials.
Informing post-operative surveillance intensity and frequency based on individualised recurrence risk.
Section 2
Formula & Logic
General Nomogram (Weiser et al. 2008)
Developed at Memorial Sloan Kettering Cancer Center from 1,320 patients (1990–2000, AJCC Stage I–III).
Cox proportional-hazards regression with restricted cubic splines for continuous variables.
Inputs: patient age, preoperative CEA, tumor site, T stage, number of positive lymph nodes, number of negative lymph nodes, lymphovascular invasion (LVI), perineural invasion (PNI), adjuvant chemotherapy.
Outputs: 5-year and 10-year probability of freedom from recurrence (FFR).
Concordance index: 0.77 (vs. 0.73–0.74 for AJCC 5th/6th edition).
MSI-Specific Nomogram (Bektas et al. 2024)
Developed at MSK from 384 patients with resected AJCC Stage I–III MSI/dMMR colon cancer (2014–2021).
Two significant predictors in multivariable analysis: T category (T1-2 vs T3-4) and number of positive lymph nodes (modelled as restricted cubic spline).
Outputs: 3-year and 5-year freedom from recurrence.
Outperforms AJCC 8th edition staging (C-index 0.759) for MSI patients.
MSI Nomogram Formula (Approximate)
Cox PH model with restricted cubic spline on positive lymph nodes.
T category HR: 10.3 (T3-4 vs T1-2), p=0.023
Positive LN linear component HR: 2.82 per unit, p<0.001
Positive LN quadratic component HR: 0.51, p<0.001
Base FFR at 5 years (overall cohort): ~89%
Section 3
Pearls/Pitfalls
MSI Biology & Prognosis
MSI/dMMR tumours represent ~15% of colon cancers and arise via defective DNA mismatch repair (MMR), caused by epigenetic silencing (sporadic) or germline MMR mutation (Lynch syndrome).
MSI tumours have generally favourable prognosis, but advanced T category (T3-4) and multiple positive lymph nodes markedly increase recurrence risk.
MSI etiology (germline/sporadic/undesignated) was not independently prognostic in multivariable analysis — staging factors dominate.
VELPI (venous, lymphatic, or perineural invasion) was excluded from the final MSI model due to collinearity with nodal involvement.
Adjuvant Therapy Implications
Adjuvant doublet chemotherapy (CAPOX/FOLFOX) may benefit high-risk MSI stage III patients, though immune checkpoint blockade (nivolumab + ipilimumab) has shown exceptional pathologic complete response rates in the NICHE-2 trial (67% pCR in cT3 node-positive MSI colon cancer). The MSK nomogram can help identify candidates for these novel adjuvant strategies.
Lynch Syndrome Monitoring
Lynch syndrome (germline MSI) was present in 19% of the MSI training cohort.
While Lynch syndrome was not independently prognostic for recurrence, 15 of 72 Lynch patients (21%) developed metachronous cancers.
Germline MSI diagnosis warrants enhanced surveillance for metachronous colorectal and extracolonic malignancies regardless of recurrence risk score.
Section 4
Evidence Appraisal
Primary References
Clinical Calculator for Predicting Freedom From Recurrence After Resection of Stage I–III Colon Cancer in Patients With Microsatellite Instability.
Bektas AB et al. • JCO Clinical Cancer Informatics. 2024;Vol. 8, e2300233. [https://doi.org/10.1200/CCI.23.00233](https://doi.org/10.1200/CCI.23.00233). C-index 0.812 (training), 0.744 (validation).
Individualized Prediction of Colon Cancer Recurrence Using a Nomogram.
Weiser MR et al. • Journal of Clinical Oncology. 2008;Vol. 26(3):380–385. [https://doi.org/10.1200/JCO.2007.14.1291](https://doi.org/10.1200/JCO.2007.14.1291). C-index 0.77.
Predicting survival after curative colectomy for cancer: Individualizing colon cancer staging.
Weiser MR et al. • Journal of Clinical Oncology. 2011;Vol. 29(36):4796–4802.
Contemporary validation of a nomogram predicting colon cancer recurrence, revealing all-stage improved outcomes.
Konishi T et al. • JNCI Cancer Spectrum. 2019;Vol. 3(1):pkz015.
