187 Differences in Clinical Outcomes of C. difficile by Cytotoxicity Assay and NAP1 Status: Review of 125 Patients at a Cancer Hospital

Saturday, April 2, 2011
Trinity Ballroom (Hilton Anatole)
Anna Kaltsas, MD, MS , Memorial Sloan Kettering Cancer Center, New York, NY
Matt Simon, MD , New York Presbyterian Hospital - Weill Cornell, New York, NY
Larissa Unruh, BS , Columbia University, Mailman School of Public Health, New York, NY
Crystal Son, MPH , Memorial Sloan Kettering Cancer Center, New York, NY
Danielle Wroblewski, PhD , Wadsworth Center, NYS Dept of Health, Albany, NY
Kimberlee A. Musser, PhD , Wadsworth Center, NYS Dept of Health, Albany, NY
Ngolela Esther Babady, PhD , Memorial Sloan Kettering Cancer Center, New York, NY
Kent Sepkowitz, MD , Memorial Sloan Kettering Cancer Center, New York, NY
Mini Kamboj, MD , Memorial Sloan Kettering Cancer Center, New York, NY
Background:

The incidence of CDI has increased in the last decade.  EIA for detection of toxin A & B has low sensitivity and cytotoxin assay (CYT) is labor and time intensive. PCR based methods are likely to be superior in terms of sensitivity and specificity; however data regarding their clinical utility is lacking.

Objective:

To evaluate clinical outcomes including 30 day all cause mortality and various measures of C. difficile infection (CDI) severity in patients with NAP1 as compared to wild-type strains. To compare test results and clinical outcomes of those who were positive by cytotoxicity assay (CYT) vs PCR.

Methods:

This was a retrospective study from February – August 2008 and March – May 2010. Clinical and demographic data were extracted from the MSKCC clinical information systems. Infection was defined as clinical symptoms with a positive CYT and/or PCR. For the 2008 period, NAP1 strains were identified as tcdC variant with 2 real-time PCR methods as described by Wroblewski and Sloan et al. For the 2010 period, we used the Cepheid Xpert C. difficile Epi assay. Severe CDI was defined as evidence for colitis on imaging, need for dual vancomycin and metronidazole therapy, or associated sepsis. Univariate analysis employed the Student’s t, Mann-Whitney U, chi2 and Fisher’s exact tests.

Results:

CDI was diagnosed in 125 patients; 33 were NAP1 strains. NAP1 strains declined from 40.3 to 26% in the 2 study periods. Patients with NAP1 strains were significantly more likely to have had a longer length of stay (LOS) prior to CDI diagnosis (p=0.02) and to have a positive CYT (p=0.014). 50% of patients with wild-type CDI were detected by PCR only, compared to 25% of NAP1 strains.

Patients with NAP1 strains were more likely to have colitis on imaging, and surgery within 60 days prior to diagnosis (p=0.08 and 0.09, respectively). Complications including need for dual therapy, transfer to ICU, and 30 day mortality were similar between the two groups.

In a multivariate logistic regression predicting NAP1 positivity, a positive CYT was associated with an odds ratio of 7.78 (95% CI 1.45-41.73), when adjusting for presence of colitis, LOS, age, and sex.

Patients who tested positive for NAP1 but had a negative CYT were often not treated, since clinicians were only made aware of CYT results. In 4/8 cases where this occurred, clinical outcomes were not significantly different from those with non-NAP1 strains.

Conclusions:

PCR is a more sensitive testing method for CDI, especially in detection of non- NAP1 strains. A longer LOS prior to CDI was also significantly associated with having the NAP1 strain. However presence of this strain did not convey a significantly increased risk for worse clinical outcomes. Larger studies are needed to explore the significance of NAP1 strains and the clinical utility of PCR based methods to detect them.