Increased Variance in Oral and Gastric Microbiome Correlates With Esophagectomy Anastomotic Leak.

BACKGROUND

Anastomotic leak after esophagectomy remains a significant source of morbidity and mortality. The gastrointestinal (GI) microbiome has been found to play a significant role in tumor oncogenesis and postoperative bowel anastomotic leak. We hypothesized that the GI microbiome could differentiate between esophageal cancer histologies and predict postoperative anastomotic leak.

METHODS

A prospective study of esophagectomy patients was performed from May 2013 to August 2014, with the collection of oral saliva, intraoperative esophageal and gastric mucosa, and samples of postoperative infections (neck swab or sputum). The presence and level for each bacterial probe as end points were used to analyze correlations with tumor histology, tumor stage, and presence of postoperative complications by unequal variances t tests for multiple comparisons and principal coordinate analysis.

RESULTS

Esophagectomy was successful in 55 of 66 patients who were enrolled. Among those, the diagnosis was adenocarcinoma in 44 (80%) squamous cell carcinoma in (13%), and benign disease in 4 (7%). The 30-day mortality was 1.8% (1 of 55). Complications included anastomotic leak requiring local drainage in 18% (10 of 55) and postoperative pneumonia in 2% (1 of 55). No correlation was noted between GI microbiome flora and tumor histology or tumor stage. A significant difference (p = 0.015) was found when the variance in bacterial composition between the preoperative oral flora was compared with intraoperative gastric flora in patients who had a leak but not in patients with pneumonia.

CONCLUSIONS

Patients with anastomotic leaks had increased variance in their preoperative oral and gastric flora. Microbiome analysis could help identify patients at higher risk for leak after esophagectomy.