Recursive partitioning for new classification of patients with esophageal cancer treated by chemoradiotherapy.

Abstract

e14500 Background: The 7th edition of the American Joint Committee on Cancerstaging system does not include lymph node size in the guidelines for staging patients with esophageal cancer. The objectives of this study were to determine the prognostic impact of the largest lymph node diameter (ND) on survival and to develop and validate a new staging system for patients with esophageal squamous cell cancer who were treated with definitive chemoradiotherapy (CRT). Methods: Information on 402 patients with esophageal cancer undergoing CRT at 2 institutions was reviewed. Univariate and multivariate analyses of data from 1 institution were used to assess the impact of clinical factors on survival, and recursive partitioning analysis was performed to develop the new staging classification. To assess its clinical utility, the new classification was validated using data from the second institution. Results: According to RPA, ND stages were best when classified as ND0 (the absence of lymph node metastases), ND1 (< 2.8 cm), and ND2 (≥ 2.8 cm). By multivariate analysis, gender, T, N, and ND stages were independently and significantly associated with survival (p<0.05). The resulting new staging classification showed the following: T1-2ND0 as Group I; T3-4ND0 or T1-2ND1 as Group II; T3-4ND1 as Group III; and TanyND2 as Group IV. The 4 new stages led to good separation of survival curves in both the developmental and validation datasets (p<0.05). Conclusions: Our results showed that lymph node size is a strong independent prognostic factor and that the new staging system, which incorporated lymph node size, provided good prognostic power and discriminated effectively for patients with esophageal cancer undergoing CRT.