Attrition After Neoadjuvant Therapy in Patients Planned for Curative Esophagectomy: A Single-Institution Retrospective Cohort Study

Background: Neoadjuvant therapy, followed by esophagectomy, is a standard treatment for locally advanced, resectable cancer of the esophagus and gastroesophageal junction (GEJ).Not every patient who starts neoadjuvant therapy reaches surgery.In trials reporting outcomes on an intention-to-treat basis, patients who do not reach surgery after neoadjuvant therapy remain included in survival estimates but are often absent from surgical series; attrition therefore reflects the gap between the intended multimodal treatment strategy and the treatment actually received.We examined how often this occurred in a tertiary surgical oncology unit in India, why it occurred, and which patient characteristics were associated with it.Methods: We retrospectively reviewed a departmental database for consecutive patients planned for curative esophagectomy after either neoadjuvant chemoradiotherapy (nCRT) with the Chemoradiotherapy for Oesophageal Cancer Followed by Surgery Study (CROSS) regimen (weekly carboplatin and paclitaxel with 41.4 Gy of radiotherapy) or perioperative chemotherapy with fluorouracil, leucovorin, oxaliplatin, and docetaxel (FLOT).Baseline characteristics, treatment received, toxicity, and radiological response were compared between resected and non-resected patients using the Mann-Whitney U test, the Fisher exact test, and the chi-square test, as appropriate.The comparisons were unadjusted.Results: The cohort included 21 (100.0%)patients; the median age was 60 years.There were 16 (76.2%)men, 11 (52.4%)patients with squamous cell carcinoma (SCC), and 18 (85.7%)with clinical stage III disease.Esophagectomy was performed in 14 (66.7%)patients; seven (33.3%) did not undergo surgery (95% confidence interval for attrition: 15-57%).Among non-resected patients, the reasons were disease progression in three (42.9%),decline in performance status in one (14.3%),acute treatment-related toxicity in one (14.3%),medical decompensation in one (14.3%),and patient refusal in one (14.3%).Compared with resected patients, non-resected patients were older (median = 67 vs. 57.5 years; p = 0.003) and more often had baseline Eastern Cooperative Oncology Group (ECOG) performance status 2 (3 (42.9%) vs. 0 (0.0%); p = 0.026).SCC (7 (100.0%) vs. 4 (28.6%);p = 0.004) and nCRT (7 (100.0%) vs. 5 (35.7%); p = 0.007) were more common, whereas radiological response was less common (1 (14.3%) vs. 14 (100.0%);p < 0.001).Among resected patients, 13 (92.9%)had an R0 resection, two (14.3%) had a pathological complete response, and six (42.9%) had a major pathological response.All resected patients, 14 (100.0%),were alive at a median followup of 16.5 months; four (57.1%) non-resected patients had died at a median follow-up of nine months (p = 0.006).Conclusions: In this small cohort, attrition after neoadjuvant therapy was substantial, and short-term outcomes in non-resected patients were poor.Older age, poor baseline performance status, thoracic SCC treated with nCRT, and absence of a radiological response were associated with attrition; the effects of histology and regimen could not be separated.Early prehabilitation and the reporting of outcomes on an intention-to-treat basis deserve more attention.