Using pretreatment tumor depth and length to select esophageal squamous cell carcinoma patients for nonoperative treatment after neoadjuvant chemoradiotherapy

Background: Two randomized trials have shown that in patients with good response to neoadjuvant chemoradiotherapy (nCRT), a nonoperative approach (additional CRT) had equal survival to scheduled esophagectomy. However, controversy exists because of the high locoregional recurrence (LR) following a nonoperative approach. Endoscopic complete response (e-CR) determined by endoscopic finding is a good criterion for predicting local control after definitive CRT. We evaluated whether e-CR could also be used to select patients for nonoperative treatment after nCRT. Methods: We retrospectively analyzed esophageal squamous cell carcinoma (SCC) patients with e-CR after nCRT between 1999 and 2006. Patients were divided into two groups by the type of treatment given after e-CR (group A, scheduled esophagectomy; group B, no scheduled surgery and continued CRT). Results: There were 71 and 79 patients in groups A and B, respectively with similar pre/post-nCRT characteristics. Despite similarity in survival and recurrence between groups, the recurrence site differed significantly. LR occurred more frequently in group B, whereas systemic recurrence was the predominant failure pattern in group A (P <.001). With use of multivariate analysis on group B, we determined that pretreatment depth of tumor invasion ≥T3 [odds ratio (OR), 11.19; 95 % CI, 1.4-89; unfavorable, P =.023] and tumor length ≥6 cm (OR, 3.069; 95 % CI, 1.17-8.1; unfavorable, P =.023) were predictors for LR. Patients with initial clinical T2 and <6 cm tumor had comparable LR (5 %) to the surgery group; these patients were candidates for nonoperative treatment after nCRT. Conclusion: In esophageal SCC patients who achieved e-CR after nCRT, pretreatment tumor depth and length were good indicators to select candidates for nonoperative treatment.