Oral Fluconazole Therapy in Patients With Active Ulcerative Colitis Who Have Detectable Candida in the Stool : A Double-Blind Randomized Placebo-controlled Trial

Background There is an emerging role of fungal dysbiosis in the pathogenesis of inflammatory bowel disease. Prevalence of Candida in patients with active ulcerative colitis (UC) and the effect of fluconazole therapy in reducing disease activity of UC are not known. Patients and methods All consecutive consenting patients with active UC defined as Mayo score ≥3 were evaluated for presence of Candida by stool culture and predictors for presence of Candida were identified. Those who had evidence of Candida in the stool were randomized to receive oral fluconazole 200 mg daily or placebo for 3 weeks along with standard medical therapy. Patients were assessed by clinical, sigmoidoscopy, and laboratory parameters at baseline and at 4 weeks. The primary outcome was clinical and endoscopic response at 4 weeks defined by a 3-point reduction in Mayo score. Secondary outcomes were reduction in fecal calprotectin, histologic response, and adverse events. Results Of the 242 patients with active UC, 68 (28%) patients had Candida in stool culture. Independent predictors for presence of Candida in patients with active UC were partial Mayo score of ≥3 and steroid exposure. Among those with Candida on stool culture (n=68), 61 patients fulfilled eligibility criteria and were randomized to receive fluconazole (n=31) or placebo (n=30). Three-point reduction in Mayo score though was numerically higher in the fluconazole group than the placebo group but was not statistically significant [5 (16.1%) vs. 1 (3.33%); P=0.19]. Postintervention median Mayo score was lower in fluconazole than placebo group [4 (3, 5) vs. 5 (4, 6); P=0.034]. Patients in fluconazole group had more often reduction in fecal calprotectin [26 (83.9%) vs. 11 (36.7%); P=0.001] and histologic scores [23 (74.1%) vs. 10 (33.3%); P=0.001] compared with placebo. All patients were compliant and did not report any serious adverse event. Conclusion Candida colonization is found in 28% of patients with UC. Steroid exposure and active disease were independent predictors for the presence of Candida. There was no statistically significant difference in the number of patients who achieved 3-point reduction in Mayo score between 2 groups. However, clinical, histologic, and calprotectin levels showed significant improvement in fluconazole group.