Half of Anastomotic Leaks After Esophagectomy Are Undetected on Initial Postoperative Esophagram

BackgroundThe sensitivity of fluoroscopic esophagography with oral administration of contrast material to exclude anastomotic leak after esophagectomy is not well documented, and the consequences of missing a leak in this setting have not been previously described.MethodsWe performed a retrospective cohort study of a prospectively maintained institutional database of patients undergoing esophagectomy with esophagogastric anastomosis from 2008 to 2020. Relevant details of leaks, management, and outcomes were obtained from the database and formal chart review. Statistical analysis was performed to compare patients with and without leaks and those with false-negative vs positive esophagrams.ResultsThere were 384 patients who underwent esophagectomy with gastric reconstruction; the majority were Ivor-Lewis (82%), and 51% were wholly or partially minimally invasive. By use of a broad definition of leak, 55 patients (16.7%) developed an anastomotic leak. Of the 55 patients, 27 (49%) who ultimately were found to have a leak initially had a normal esophagram result (performed on average on postoperative day 6). Those with a normal initial esophagram result were more likely to have an uncontained leak (81% vs 29%; P < .01), to require unplanned readmission (70% vs 39%; P = .02), and to undergo reoperation (44% vs 11%; P < .01).ConclusionsEarly postoperative esophagrams intended to evaluate anastomotic integrity have a low sensitivity of 51%, and leaks missed on the initial esophagram have greater clinical consequences than those identified on the initial esophagram. These findings suggest that a high index of suspicion must be maintained even after a normal esophagram result and call into question the common practice of using this test to triage patients for diet advancement.