BACKGROUND: The role of Mandard's tumor regression grade (TRG) classification is still controversial in defining the prognostic role of patients who have undergone neoadjuvant chemoradiation (CRT) and total mesorectal excision. The present study evaluated multiple correspondence analysis (MCA) as a tool to better cluster variables, including TRG, for a homogeneous prognosis. PATIENTS AND METHODS: A total of 174 patients with a minimum follow-up period of 10 years were stratified into 2 groups: group A (TRG 1-3) and group B (TRG 4-5) using Mandard's classification. Overall survival and disease-free survival were analyzed using univariate and multivariate analysis. Subsequently, MCA was used to analyze TRG plus the other prognostic variables. RESULTS: The overall response to CRT was 55.7%, including 13.2% with a pathologic complete response. TRG group A correlated strictly with pN status (P = .0001) and had better overall and disease-free survival than group B (85.1% and 75.6% vs. 71.1% and 67.3%; P = .06 and P = .04, respectively). The TRG 3 subset (about one third of our series) showed prognostically heterogeneous behavior. In addition to multivariate analysis, MCA separated TRG 1 and TRG 2 versus TRG 4 and TRG 5 well and also allocated TRG 3 patients close to the unfavorable prognostic variables. CONCLUSION: TRG classification should be used in all pathologic reports after neoadjuvant CRT and radical surgery to enrich the prognostic profile of patients with an intermediate risk of relapse and to identify patients eligible for more conservative treatment. Thus, MCA could provide added value.
- Mandard's classification
- Neoadjuvant CRT
- Prognostic profile
- Rectal cancer
Mancini, R., Pattaro, G., Diodoro, M. G., Sperduti, I., Garufi, C., Stigliano, V., Perri, P., Grazi, G. L., & Cosimelli, M. (2018). Tumor Regression Grade After Neoadjuvant Chemoradiation and Surgery for Low Rectal Cancer Evaluated by Multiple Correspondence Analysis: Ten Years as Minimum Follow-up. Clinical Colorectal Cancer, 17(1), e13-9. https://doi.org/S1533-0028(16)30276-6 [pii]