Analyzing 9998 patients with gastric cancer who underwent surgery, Ahn et al compared survival rates assessed by the seventh and sixth editions of the TNM staging system.1 The authors concluded that the seventh edition appears to provide better categorized grouping than the sixth. However, they encouraged the use of the N3b classification.
The inclusion in the same TNM stages of patients with 7 or 16 positive lymph nodes jeopardizes the reported 5-year survival homogeneity of the seventh edition TNM stage IIB (ie, 77% for N3a vs 35% for N3b) and stage IIIA (ie, 59% for N3a vs 40% for N3b). Unfortunately, the authors did not attribute a statistically significant value to these evident differences. Hence, the seventh edition of the TNM has no homogeneous groups (and not 2 as reported). Moreover, the authors stated that each previous TNM stage was divided into the new TNM stages with different survival rates. However, once again this conclusion is not supported by log-rank tests for each stratum (nor by any graph) within the different groups.
Furthermore, in our opinion, the seventh edition of the TNM exposes more patients to stage migration because of reduced pN cutoff values and the small pN ranges. Ahn et al reported a stage migration >20%. However, the actual stage migration (ie, the “Will Rogers phenomenon”) should be calculated depending on the extent of lymph node dissection, but they did not provide details regarding lymphadenectomy. In addition, even if the minimum number of lymph nodes for staging reached 16,2 in the analysis by Ahn et al, patients with 15 examined lymph nodes also were included.
Given the lack of statistical details and an accurate evaluation of stage migration phenomenon, the authors are not convincing regarding the superiority of the seventh edition of the TNM staging system compared with the sixth.