QUALITY IMPR TNM Staging of Colorectal Cancer Should be Reconsidered According to Weighting of the T Stage Verification Based on a 25-Year Follow-Up

The gradient monotonicity of existing tumor, node, metastases staging systems for colorectal cancer is unsatisfactory. Our proposed T-plus staging system strengthens weighting of the T stage. In this study, applicability of the T-plus staging system was verified with data of a Chinese colorectal cancer center. Records of 2080 nonmetastatic, advanced cancer patients undergoing colorectal cancer surgery from 1985 to 2011 were reviewed for T, N stage pathology and follow-up information. Using overall and disease-specific survival data, the 7th edition tumor, node, metastases staging system and the T-plus staging system were compared for stage homogeneity and discrimination and gradient monotonicity. For gradient monotonicity, the T-plus staging system was superior for both colon and rectal cancer. With Kaplan–Meier survival curves, the T-plus staging system discriminated among different stages, and the corresponding survival was inversely associated with the stage. However, for the 7th edition tumor, node, metastases staging system, stage IIIa had a better prognosis than stage II for rectal cancer and stage I for colon cancer. For homogeneity within the same stage and discrimination between different stages, the 2 staging systems were similar for color, MM, Jin-Song Li, an, MD, MD, and Ke-Feng Ding, MD lymph node status as the criterion to discriminate colorectal cancer stage II and stage III with greater weighting of the T stage. (Medicine 95(6):e2711) Abbreviations: AIC = Akaike information criteria, CEA = carcinoembryonic antigen, CI = confidence interval, DSS = disease-specific survival, HR = hazard ratio, IQR = inter-quartile range, LR = likelihood ratio, M = median, OS = overall survival, R0 resection = complete resection with no microscopic residual tumor, SAHZU = Second Affiliated Hospital, Zhejiang University School of Medicine, SEER = Surveillance, Epidemiology, and End Results, TNM = tumor, node, metastases. INTRODUCTION T he tumor, node, metastases (TNM) staging system established by the American Joint Committee on Cancer is widely used to predict the prognosis for patients with colorectal cancer, to guide adjuvant therapy after potentially curative surgery, and to classify patients for participation in clinical trials. The ideal prognostic system should provide homogeneity within the same stage, good discrimination between different stages, and monotonicity of gradients that predicts survival outcomes that are consistent with the severity of cancer staging. Currently, the 7th edition of the TNM staging system is used in America and China, whereas some European countries continue to use the 5th edition. However, all of the TNM staging systems share similar anatomical elements and the same key principles that were inherited from the Dukes staging system that was developed in the 1932. The key principle is that patients with lymph node involvement are classified as C in the Dukes staging system and as stage III in the 7th edition TNM staging system. Thus, the principles for staging patients with colorectal cancer have been substantially stable for >80 years. It is well recognized that the Dukes B and the 7th edition TNM staging system stage II are composites of better (T3N0M0) and worse (T4N0M0) prognostic groups. Moreover, stage IIIa patients often have a better prognosis than some in stage II. In summary, the monotonicity of gradients of the existing TNM staging system for colorectal cancer is unsatisfactory. The causes of the defect in gradient monotonicity are not well understood so far. We have reanalyzed the summary survival data of the patients from Surveillance, Epidemiology, and End Results (SEER). We rearranged all of the TN categories according to the observed survival, and then used cluster analysis to revise the staging system. According to the cluster analysis of the TN scores, T1N1a was classified as stage b-2a were classified as stage II for both . However, T4bN0 was classified as IIIa s IIIb in rectal cancer (Table 1). In the www.md-journal.com | 1 TABLE 1. TN Categories of the 7th Edition TNM Staging System and the T-Plus Staging System for Colorectal Cancer Stages 7th Edition TNM Staging System T-Plus Staging System I T1–2N0 T1N0–1a II T3–4bN0 T1N1b-2a T2N0–1b T3N0 IIIa T1–2N1 T1N2b T1N2a T2N2 T3N1a–2a T4aN0–1a T4bN0 (colon cancer) IIIb T3–4aN1 T3N2b T2–3N2a T4aN1b–2b T1–2N2b T4bN0 (rectal cancer) T4bN1 IIIc T4aN2a T4bN2 T3–4aN2b T4bN1–2 For the T-plus staging system, tumor deposit was defined by the criteria used in the 7th edition TNM staging system, and N1c cases were Li et al revised staging system, which we named the ‘‘T-plus staging system,’’ extra emphasis was placed on the weighting of the T stage. The SEER survival data had good monotonicity of gradients when fitted to the T-plus staging system. However, the T-plus staging system is just a proposal, and it has not been verified by individual survival data. In this study, we tested the applicability of the T-plus staging system using the data of 1 major colorectal cancer center in China. PATIENTS AND METHODS