P1.217. Para-Anastomotic Fluid Analysis as a Predictor for Leak After Transhiatal Esophagectomy: Results From a Pilot Trial

Abstract

Abstract Topic Esophageal Cancer: Surgical Treatment of Esophageal Cancer – early outcomes and complications Background Anastomotic leak after esophagectomy is associated with poor conduit perfusion and cervical location. Peri-anastomotic fluid acidity (pH) has been shown to be a predictor of leak after colonic anastomosis. We hypothesized that fluid acidity, from samples collected from cervical wounds after transhiatal esophagectomy could predict anastomotic leaks. Methods A prospective pilot study of patients undergoing esophagectomy with cervical anastomosis at a single academic institution was performed. A portable, non-invasive biosensor system was attached in line between a cervical drain and standard bulb. The sensor continuously monitored drainage fluid characteristics including pH and electrical conductivity, viewed in real-time on a bedside monitor. The sensor was initiated on postoperative day (POD) 0/1 and remained attached until the clinical removal of the drain (POD 5). Cervical leaks were identified by a barium esophagogram on POD 6/7. Cervical neck infections were identified based on clinical appearance and were managed similarly to those with leaks with cervical wound packing. A Mann-Whitney U Test was utilized to complete preliminary analysis and analyze statistical significance. Results 14 men and five women with a mean age of 68.26 ± 7.77 participated. The average pH of the cervical drains was 8.68 ± 0.39. Seven patients had incomplete data due to technical issues. Of the remaining 12, 8 had no-leak and 4 had a leak and/or infection. All 12 patients had comparable pH levels during the early postoperative period. After approximately 20 hours, the pH of the leak/infection group declined sharply, while the pH for the control group began to gradually increase towards 9.0 and stabilized (Figure). The difference in the mean pH was statistically significant between 48-60 hours postoperatively with p-values of 0.0012 and 0.005 for hours 48 – 54 and 54 – 60, respectively. Between postoperative hours 48 – 54 and 54 – 60, the average pH for the healthy cohort was 8.883 and 8.959, and the average pH for the leak/infection cohort was 7.637 and 8.324, respectively. Conclusion Physiological changes associated with postoperative leaks or infections are associated with an early decline of drainage pH. A statistically significant difference in pH level was recorded between 48 – 60 hours postoperatively, days earlier than a clinical leak or infection was suggested. Further investigation into pH trends could allow for modified post-operative management to minimize the impact of leaks and/or infections on patient outcomes.