Ann Surg Oncol. 2025 Aug 10. doi: 10.1245/s10434-025-18000-6. Online ahead of print.
ABSTRACT
BACKGROUND: Despite treatment advancements, esophageal cancer survival remains poor due to high recurrence rates. Early detection of recurrence may allow for timely treatment and improved outcomes. This study evaluated recurrence patterns and detection methods using different imaging surveillance strategies after esophagectomy.
METHODS: A retrospective study reviewed patients who underwent esophagectomy for cancer at a high-volume National Cancer Institute (NCI)-designated comprehensive cancer center from 2007 to 2019. Postoperative surveillance followed a protocol based on the CROSS trial incorporating routine computed tomography (CT) imaging and clinical exams. Statistical analyses included independent t tests for continuous variables and chi-square tests for categorical variables. Times to recurrence and survival were calculated using Kaplan-Meier and compared by the log-rank test. Multivariate analysis used binary logistic regression.
RESULTS: Among 368 patients, 302 (82.1 %) received neoadjuvant chemoradiation. Recurrence occurred for 140 (38 %) patients, with more than 80 % of the recurrences detected within 2 years after surgery. In a multivariate analysis, lymphovascular invasion and clinical stage 3 disease were associated with recurrence. Clinically driven imaging discovered 61 (43.6 %) of the 140 recurrence cases, whereas routine surveillance imaging identified 77 (55 %) of the cases. The median time to recurrence was 9.4 months. The patients whose recurrence was detected through routine surveillance had a longer survival than those whose recurrence was detected by clinically driven imaging (median survival, 29.3 vs 17.7 months, respectively; p < 0.05).
CONCLUSIONS: The incidence of early recurrence is high after trimodality therapy (CROSS regimen) for esophageal cancer. Routine imaging surveillance combined with clinical examination as a surveillance protocol is necessary for early detection and timely treatment.
PMID:40784997 | DOI:10.1245/s10434-025-18000-6
