Background: Recent advances in endoscopic therapy, including conventional endoscopic resection and endoscopic submucosal dissection (ESD), have led to a large number of patients with early colorectal cancer (CRC) being cured; however, when resected specimens obtained by these procedures manifest risk factors for lymph node metastasis, additional treatments need to be considered. The aim of our study was to evaluate the outcomes of salvage surgery in CRC patients treated initially by advanced therapeutic endoscopy.
Methods: We investigated 145 patients who underwent salvage surgery in our department after endoscopic therapy for CRC between April 2006 and March 2015. Demographic and pathological data, endoscopic procedures, reasons for surgery, and operative outcomes, including perioperative details and recurrence-free and disease-specific survival after surgery, were analyzed. These data were further compared with those of 59 patients with submucosal invasive CRC treated by conventional endoscopic resection/ESD alone and 133 patients treated by surgery alone.
Results: Overall lymph node metastases were observed in 14% of patients who underwent salvage surgery after therapeutic endoscopy and 16% of those who received abdominal surgery alone. In analyses of surgical cases, patients with lymph node metastases more frequently included cases with lymphatic infiltration (63%) and ESD-treated cases (45%) than those without metastases (21%, P < .0001 and 22%, P = .02; respectively). A logistic regression analysis identified lymphatic infiltration as an independent predictive factor for lymph node metastases (odds ratio: 8.77, 95% confidence interval: 2.90-33.31, P < .0001). Long-term outcomes were favorable in both lymphatic infiltration-negative and positive cases. Moreover, survivals were comparable among the different treatment groups.
Conclusion: Because of the high rate of nodal involvement, adequate lymphadenectomy need to be performed in salvage surgery after upfront endoscopic therapy.
Copyright © 2016 Elsevier Inc. All rights reserved.