Purpose: Neoadjuvant chemoradiation (CRT) is increasingly used in locally advanced esophageal cancer. Some studies have suggested that CRT results in increased surgical morbidity. We assessed the influence of CRT on anastomotic complications in a cohort of patients who underwent CRT followed by Ivor Lewis esophagectomy.
Patients and methods: Clinical and pathologic data were collected from all patients treated with neoadjuvant CRT (36 Gy combined with 5-fluorouracil and cisplatin) followed by Ivor Lewis esophagectomy. On the radiotherapy (RT) planning computed tomography scans, normal tissue volumes were drawn encompassing the proximal esophageal region and the gastric fundus. Within these volumes, dose-volume histograms were analyzed to generate the total dose to 50% of the volume (D(50)). We studied the ability of the D(50) to predict anastomotic complications (leakage, ischemia, or stenosis). Dose limits were derived using receiver operating characteristics analysis.
Results: Fifty-four patients were available for analysis. RT resulted in either T or N downstaging in 51% of patients; complete pathologic response was achieved in 11%. In-hospital mortality was 5.4%, and major morbidity occurred in 36% of patients. Anastomotic complications (AC) developed in 7 patients (13%). No significant influence of the D(50) on the proximal esophagus was noted on the anastomotic complication rate. The median D(50) on the gastric fundus, however, was 33 Gy in patients with AC and 18 Gy in patients without AC (p = 0.024). Using receiver operating characteristics analysis, the D(50) limit on the gastric fundus was defined as 29 Gy.
Conclusions: In patients undergoing neoadjuvant CRT followed by Ivor Lewis esophagectomy, the incidence of AC is related to the RT dose on the gastric fundus but not to the dose received by the proximal esophagus. When planning preoperative RT, efforts should be made to limit the median dose on the gastric fundus to 29 Gy with a V(30) below 40%.
Copyright © 2012 Elsevier Inc. All rights reserved.