134 Differentiation between C. difficile reinfection and recurrence by PCR ribotpying

Friday, March 19, 2010
Grand Hall (Hyatt Regency Atlanta)
Perminder Khosa , Memorial Sloan Kettering Cancer Center, New York, NY
Crystal Son , Memorial Sloan Kettering Cancer Center, New York, NY
Kent Sepkowitz , Memorial Sloan Kettering Cancer Center, New York, NY
Mini Kamboj , Memorial Sloan Kettering Cancer Center, New York, NY
Background: For patients with 2 episodes of C. difficile associated diarrhea (CDAD), the CDC distinguishes between recurrent versus second infection based on the interval between positive isolates. Symptomatic patients with 2 positive samples within 2-8 weeks are considered to have recurrence due to the same isolate whereas those with episodes at least 8 weeks apart are considered to have a second acute infection.

Objective: National surveillance for CDAD is based on the above definitions. We used molecular typing of paired samples from individual patients to examine the appropriateness of the definition.

Methods: All patients between 1/18/08 and 12/28/08 at MSKCC with 2 positive samples for CDAD collected at least one week apart were identified. PCR ribotyping was performed on these isolates. Results for individual patients were compared for homology, suggesting infection with the same, rather than a new, isolate.

Results:During the period, 330 patients had an isolate positive for CDAD, including 94 patients with at least two positive isolates. Of these isolates, 36 pairs, corresponding to 24 patients,  at least 1 week apart were collected. Of the 36 typed pairs, homology was found in 28 (78%). Among those with homologous pairs, the interval between collected specimens ranged from 1.7 to 45.7 weeks (6.6 weeks). The 8 pairs which did not demonstrate homology were collected 3.4 to 23.6 weeks apart (median 13.5 weeks). According to national surveillance guidelines, 17 out of 36 (47%) pairs would be classified as “New Cases” of CDAD, but we found that 12 (71%) of these pairs are actually homologous.

Conclusions: Our findings do not support the current CDC definitions of relapse and second infection. Our study suggests that persons with a second episode occurring 8 weeks from the first are more likely to recur with the same strain than be infected by a second strain. This finding, if demonstrated in larger series, has important implications for surveillance and control efforts.