Early Post-Operative Diagnosis of Anastomotic Leak by Local Electrophysiological Parameters: Multicenter Pilot Study.

Background: Morbidity associated with anastomotic leak (AL) can be reduced by timely detection. Local electrophysiological parameters are highly sensitive to pathophysiological processes (inflammation/ischemia) underlying AL and may enable early post-op monitoring ahead of clinical signs. A new approach to early leak detection includes post-operative measurement of bioelectric activity of the GI tract. The method has been implemented in a medical device that uses electrodes embedded in a standard surgical drain to continuously monitor local bowel myoelectric activity following GI surgery. Methods: A multicenter pilot study evaluated the performance of the device to support detection of clinically recognizable anastomotic leaks on post-operative day 3. The device's technology collected data that an algorithm classified to predict AL. Data were collected for a duration of 3-10 days, while the surgical drain was in place (without affecting routine care). Pilot endpoints were safety and performance analyses: sensitivity and specificity. Results: Fifty patients from 6 medical centers (mean age: 63.6 ± 13.3 years; 38% male) undergoing elective anterior resection (laparoscopic (N = 26), robotic (N = 16), open (N = 8)) for cancer (N = 44), diverticulitis (N = 4), and rectal prolapse (N = 2) were included. Average anastomotic height was 9 ± 5 cm from the anal verge. Grade C ALs were observed in 6.0% (N = 3) of the patients, resulting in increased hospitalization time (6.4 ± 2.4 days with no leak vs. 15.0 ± 7.9 days with Grade C leak). In this N = 50 cohort, the device had 100% sensitivity and 100% specificity at POD3 in prediction of Grade C leaks. Given the small number of events (three Grade C leaks, seven leaks overall), these results are preliminary. No device-related serious adverse events were observed, and the standard drain form factor contributed to positive usability feedback. Conclusions: Pilot data suggest that the bioelectric sensing drain device is safe and warrants further evaluation as a potential adjunct for AL detection. The small study sample size and wide confidence intervals for key outcomes require confirmation in a pivotal diagnostic accuracy study with a fixed thresholding algorithm.