Efficacy of different therapeutic options for spontaneous abdominal abscesses in Crohn's disease: are antibiotics enough?

BACKGROUND: Management of abdominal abscesses (AA) in Crohn's disease (CD) always includes antibiotics, and some type of drainage is added according to the response and other considerations. Our aim was to assess the efficacy of different therapeutic approaches to spontaneous AA in CD.

METHODS: All cases of AA in CD were identified from the databases of five university hospitals. Postoperative cases were excluded. Therapeutic success was defined as abscess resolution and nonreappearance within 1 year of follow-up.

RESULTS: We identified 128 cases in 2236 patients (cumulative incidence 5.7%). Initial therapy included medical therapy with antibiotics alone (42.2%), antibiotics plus percutaneous drainage (23.4%), and antibiotics plus surgical drainage (34.4%). The highest final efficacy corresponded to surgery (91%) as compared with antibiotic therapy alone (63%) or percutaneous drainage (30%). Failure of initial therapy was predicted by immunomodulators at diagnosis (odds ratio [OR] 8.45; 95% confidence interval [CI] 1.16-61.5; P = 0.03), fistula detectable in imaging techniques (OR 5.43; 95% CI 1.18-24.8; P = 0.02), and abscess size (OR 1.65; 95% CI 1.07-2.54; P = 0.02) only for patients treated with antibiotic therapy alone. Percutaneous drainage was associated with 19% of complications (enterocutaneous fistulas 13%); surgery was associated with 13% of postoperative complications (enterocutaneous fistulas 7.7%). Following abscess resolution, 60% of patients were started on thiopurines, 9% on biologics, and in 31% baseline therapy was not modified.

CONCLUSIONS: Management of spontaneous AA in CD with antibiotics alone seems to be a good option for small abscesses, especially those without associated fistula and appearing in immunomodulator-naïve patients. Surgery offers better results in the remaining clinical settings, although percutaneous drainage can avoid operative treatment in some cases.