170 Molecular Epidemiology of Clostridium difficile Infection (CDI) in Children at a Tertiary Care Hospital - A Pilot Study

Saturday, April 2, 2011
Trinity Ballroom (Hilton Anatole)
Anthony J. Perissinotti, PharmD , Wayne State University, Detroit, MI
Teena Chopra, MD , Wayne State University, Detroit, MI
Ahmed Mohammadieh, BS , Wayne State University, Detroit, MI
Jing J. Zhao, PharmD , Wayne State University, Detroit, MI
Paul Lephart, PhD , Wayne State University, Detroit, MI
Lynn Semproch, BS, MT, MPH , Wayne State University, Detroit, MI
Smitha Sathyaprakash, MD , Wayne State University, Detroit, MI
Keith Kaye, MD, MPH , Wayne State University, Detroit, MI
George Alangaden, MD , Wayne State University, Detroit, MI
Background: Scant data exists regarding the molecular epidemiology of CDI in children.

Objective: To determine the prevalence of the hypervirulent epidemic North American Pulse Field Type-1 (NAP-1) strain of C. difficile (CD) in CD toxin-positive stools of children.

Methods: Stool samples that were CD toxin-positive from all patients hospitalized at the Detroit Medical Center Hospitals (DMC) from January 2008-2010 have been collected for research purposes. From this repository, 22 stool specimens obtained from children hospitalized at a tertiary- care hospital were tested to determine the prevalence of the NAP-1 epidemic strain of CD.  The specimens were processed under anaerobic conditions to isolate CD. PCR typing using Cepheid Xpert CD assay was used to identify the NAP-1 strain of CD. Patient variables collected included demographics, clinical, laboratory, treatment, and outcome data.

Results: During the study period a total of 46 positive CD toxin tests were seen. Four of the 22 (18%) randomly selected isolates were the epidemic NAP-1 strain. 2 (50%) cases of NAP-1 were classified as CO-CA, 1 (25%) as HO-HCFA, and 1 (25%) as CO-HCFA. For the non-NAP-1 group 6 (37.5%) cases were classified as CO-CA, 7 (43.75%) as HO-HCFA, and 3 (18.75%) as CO-HCFA. The median age of the cohort was 5 yr (IQR=1-12): median age of children with NAP-1 strain being 10 yr (range 5-13 yr) vs 2 years (range 0-14 yr) for the non-NAP-1 strain. 17/22 children had prior exposure to antibiotics within 30d of a positive CD toxin test, with ampicillin/amoxicillin being the most common agent. The median duration of antibiotic exposure was 11d (IQR=1-22) for the NAP-1 group compared to 2d (IQR=0-6.5) for the non-NAP-1 group. Overall 19/22 (84%) of the children, including all of those with the NAP-1 strain had clinical disease consistent with CD infection (CDI). The NAP-1 group did not experience severe CDI; however 2 cases of the non-NAP-1 group were admitted to the ICU for CDI. All of the 19 symptomatic children treated for CDI with oral vancomycin, metronidazole, or both, responded to therapy. 1 NAP-1 strain and 2 non-NAP-1 patients experienced recurrent CDI within 8 weeks of the first episode. No deaths related to CDI were noted.

Conclusions: In our cohort, asymptomatic carriage of CD was uncommon, and almost all children with CD toxin-positive stools had symptomatic CDI. NAP-1 strain was found in about a fifth of patients with symptomatic CDI, mainly in older children. The presence of the NAP-1 strains was not associated with severe CDI. Larger studies in children to characterize the molecular epidemiology of CD strains and correlation with severity of CDI are needed.