Letter to Editor: “Gastric Fistula after Laparoscopic Sleeve Gastrectomy: Don't Forget to Treat for Candida”

Abstract

According to recent meta-analyses, the gastric fistula (GF) rate following laparoscopic sleeve gastrectomy (LSG) is about 2 % [1, 2]. Gastric fistula is defined according to the modified UK Surgical Infection Study Group classification [3, 4] and is treated as a function of the time between surgery and the discovery of GF [5]. When combined with surgical or endoscopic procedures, pharmacological and nutritional care consists of the administration of proton pump inhibitors, appropriate antibiotics, and (in some cases) the use of somatostatin analogs and nutritional support (parenteral nutrition or enteral nutrition via a feeding jejunostomy). Peritonitis as a result of upper gastrointestinal tract damage is known to be predictive for the isolation of yeast in the peritoneal fluid [6]. Hence, we decided to establish whether patients presenting post-LSGGF are at risk of this pathological feature—feature that has not previously been described in this context. We retrospectively reviewed a prospective cohort of patients referred to our institution for GF following LSG between November 2004 and April 2013. All patients with GF after LSG were included (n=68). The study population included 55 females (80.8 %) and 13 males and the mean (range) age was 36.4 (22–55). The surgical procedures included primary LSG (n=54), gastric band removal plus LSG in the same procedure (n=6), gastric band removal, and LSG in separate procedures (n=4) and repeat LSG (n=2). Details of the latter two patients have already been published [7]. Peritoneal fluid samples had been collected for 39 of the 68 patients (57.3 %). Candida sp. was isolated from 21 of these 39 samples (53.8 %), with Candida albicans in 17 cases (81 %), Candida glabrata in three cases (14.3 %), and Candida norvegensis in one case (4.7 %). According to our calculation of Dupont's score on admission, 82.3 % of the patients (n=56) had three or more risk factors for the presence of Candida in peritoneal fluid (Fig. 1). These 56 patients received antifungal treatment (83 %), with a loading dose of 800 mg of fluconazole on the first day and then 400 mg/day for 10 days. Only one patient with fluconazole-resistant C. glabrata received a different antifungal treatment (70 mg/day of caspofungin for 10 days). Of the patients with three or four risk factors for Candida isolation, 33 had peritoneal fluid samples and 19 of the latter were positive for Candida (57.5 %). Of the 21 patients with yeast in their peritoneal fluid, 19 (90.4 %) had three or more risk factors (according to Dupont's score; Table 1). The sensitivity and specificity of Dupont's score were respectively 90.4 and 22.2 %, whereas the positive and negative predictive values were respectively 35.2 and 83.3. Of the 68 patients, 14 (20.6 %) had failure of at least one organ and required hospitalization in the intensive care unit. As reported in our study of pseudo-aneurysms following persistent GF, one patient died 161 days after discovery of the GF [8]. None of the patients with yeastpositive samples died. L. Rebibo : C. Cosse :A. Dhahri : P. Verhaeghe : J.<M. Regimbeau Department of Digestive Surgery, Amiens University Medical Center, Amiens, France