129 Antigen Based Assays Show Decreased Sensitivity for Several Clostridium difficile Ribotypes Including the Epidemic 027 Strain

Friday, March 19, 2010
Grand Hall (Hyatt Regency Atlanta)
Fred C. Tenover, Ph.D. , Cepheid, Sunnyvale, CA
Susan Novak-Weekley, PhD , SCPMG Regional Reference Laboratory, North Hollywood, CA
Thomas Akerlund, PhD , Swedish Institute for Infectious Diseases, Solna, Sweden
Ferric Fang, MD , University of Washington School of Medicine, Seattle, WA
Alice S. Weissfeld, PhD , Microbiology Specialists Inc, Houston, TX
Thomas Davis, MD., PhD , Wishard Memorial Hospital, Indianapolis, IN
Lance R. Peterson, MD , NorthShore University HealthSystem, Evanston, IL
Christopher Woods, MD , Durham Veterans Affairs Medical Center, Durham, NC
Paul Schreckenberger, PhD , Loyola University Medical Center, Maywood, IL
Beth Marlowe, PhD , SCPMG Regional Reference Laboratory, North Hollywood, CA
Ellen Jo Baron, PhD , Cepheid, Sunnyvale, CA
David H. Persing, MD, PhD , Cepheid, Sunnyvale, CA
Background: Clostridium difficile infections (CDI) are a significant cause of healthcare associated diarrhea worldwide.  Antigen detection assays for toxins A and B and glutamate dehydrogenase (GDH) are commonly used for diagnosis of CDI.  GDH is often used to screen samples for C. difficile, although it detects both toxigenic and non-toxigenic strains. Several recent studies have shown lower sensitivity of enzyme immunoassays (EIAs) for toxin A and B when compared to the results of toxigenic culture.  PCR assays targeting the toxin B gene of C. difficile are alternative methods of detecting CDI.  PCR and GDH are often presumed to have equivalent sensitivity; however, the influence of strain type on assay sensitivity has not been explored.

Objective: The goal of this study was to compare the sensitivity of EIA and GDH assays to the sensitivity of a commercial PCR assay for the toxin B gene for a variety of C. difficile ribotypes.

Methods: Specimens were obtained for culture, antigen testing (EIA for all and GDH for a subset), and PCR from a total of 2296 eligible patients with suspected CDI; 338 yielded C. difficile isolates that produced cytotoxin by in vitro testing. All isolates underwent PCR ribotyping. We compared the results of two subsets of isolates for which a known ribotype was determined; commercial EIA and PCR assays run in parallel for 151 isolates of C. difficile, and GDH and PCR for 47 isolates.  Fisher’s Exact test was used for statistical analysis.

Results: Detection of C. difficile using GDH was significantly less sensitive than PCR (72.2% vs 91.7%, respectively, p=0.001) for ribotypes other than 027.   There was no difference between the sensitivities of the two methods for detection of 027 strains (both 90.1%). Toxin A/B EIA assays were less sensitive overall than PCR and only detected 15.4%, 78.4% and 18.8% of ribotypes 002, 027, and 106, respectively, compared to 100%, 100%, and 75% for PCR (p=0.0001, p=0.0001, and p=0.004), respectively. 

Conclusions: PCR showed significantly higher levels of sensitivity than EIA for ribotypes 002, 027, and 106, and overall higher sensitivity than GDH for ribotypes other than the 027 epidemic strain. Laboratories using GDH and EIA may not detect major increases of CDI caused specifically by ribotypes 002, 027, and 106, all of which are currently common in Europe.