Analysis of reports of Clostridium difficile infection in phase 2 and 3 studies of tigecycline

Abstract number: 1733_1179

Wilcox M., Freeman J.

Objectives: Antibiotics vary in their propensity to induce C. difficile infection (CDI). Anaerobic and/or broad spectrum activity may be associated with increased risk of CDI, although these associations are not clear. We reviewed the reports of possible CDI in tigecycline pre-registration studies to determine whether this extended broad spectrum activity antibiotic has increased risk of inducing CDI.

Methods: Clinical trial database searches were performed on all completed Phase 2 (n = 4) and 3 (n = 9) studies; 7 of the 9 phase 3 studies were originally blinded. To identify subjects who may have had CDI associated with tigecycline or comparator therapy, searches were performed for adverse event terms, indications for receipt of concomitant antibiotics, and the results of local hospital faecal sample investigations. A diagnosis of CDI required symptoms (diarrhoea, abdominal pain/distension, colitis) to be present for at least 48 h; 2 cases (1 tigecycline and 1 comparator), who also received no antibiotic treatment for CDI, were excluded on this basis. Cases who had symptoms of (n = 1) or were being treated for (n = 5) CDI at the time that study antibiotic was started were also excluded.

Results: Records of 2,402 tigecycline and 1,892 comparator (825 imipenem, 589 vancomycin ± aztreonam, 422 levofloxacin, 52 ciprofloxacin and 4 linezolid) recipients were searched for evidence of CDI. After the above exclusions there were 9 possible cases of CDI. Of these, 5 received tigecycline, 3 comparator and 1 tigecycline (5 days) and then oral comparator (levofloxacin) for 5 days. The latter case developed symptoms of CDI 4 days after stopping levofloxacin, and therefore causality with either antibiotic could not be established. The tigecycline associated CDI cases received 3–6 (median 3) other antibiotics before onset of symptoms. Of the comparator associated CDI cases, 2 (imipenem) each received 3 other antibiotics before onset of symptoms; the remaining case received only the comparator antibiotic (levofloxacin). All of the cases were mild-moderate in severity.

Conclusion: CDI was uncommon following tigecycline administration, was associated with prior exposure to other antibiotics, and was of a similar frequency to that seen with comparator antibiotics (~0.2% of study recipients). The extended broad spectrum activity of tigecycline does not appear to result in increased risk of CDI.