1094. Performance of toxin enzyme immunoassays and PCR cycle threshold for differentiating Clostridium difficile infection from colonization in children with diarrhea
Session: Poster Abstract Session: Diarrhea Diagnostic Dilemmas
Friday, October 5, 2018
Room: S Poster Hall
Posters
  • Kociolek ID Week 2018 Poster FINAL.pdf (4.1 MB)
  • Background: C. difficile colonization is common in children. PCR does not distinguish infection (CDI) from colonization. Toxin enzyme immunoassay (EIA) and PCR cycle threshold (Ct) may predict CDI in PCR+ adults, but assay performance in children is poorly understood.

    Methods: Stools from children 2-21 years old with laboratory-identified (labID) CDI (tcdB PCR+; GeneXpert) underwent: toxin EIA (QUIK CHEK Complete [QCC] and Immunocard [IC]); cell culture cytotoxicity neutralization assay (CCCNA); and C. difficile stool culture (Cx). Children were determined to have clinical CDI (cCDI) by chart review and/or parent communication if all were noted: at least 3 unformed stools (Bristol type 5-7) in 24h; response to CDI treatment within 5d; and no other likely diarrheal etiology. EIA and PCR Ct performance were measured for various reference standards (RefStd) based on stool assay results and/or cCDI classification.

    Results: 253 PCR+ stools were included. All stools underwent QCC; 218 (86%) were quantity sufficient for IC. Discordant EIA results occurred in 19/218 (8.7%) stools. Table 1 lists EIA sensitivity (Sn), EIA specificity (Sp), and median PCR Ct for each RefStd. Fig. 1 shows the receiver operating characteristic (ROC) curve for PCR Ct to identify PCR+/CCCNA+/cCDI+ children (area under curve = 0.76). The difference between sensitivity (71%) and specificity (72%) was minimized at Ct < 23.5.

    Conclusion: Only a minority of PCR+ children meets strict clinical and laboratory CDI criteria. More stringent CDI definitions are associated with increasing toxin EIA Sn and lower PCR Ct (i.e., greater stool C. difficile inoculum). However, both toxin EIA and PCR Ct perform suboptimally as stand-alone tests to distinguish CDI from colonization in PCR+ children.

    Table 1: Toxin EIA and PCR Ct Performance

    RefStd (n for QCC)

    QCC EIA (n=253)

    IC EIA (n=218)

    Median PCR Ct

    (Ref+ / Ref-)

    Sn

    Sp

    Sn

    Sp

    PCR+ only (253)

    0.36

    0.34

    24.3

    PCR+ / Cx+ (211)

    0.41

    0.93

    0.39

    0.94

    23.7 / 29.1**

    PCR+ / CCCNA+ (128)

    0.69

    0.99

    0.65

    0.99

    22.2 / 28.5**

    PCR+ / cCDI+ (103)

    0.46

    0.71

    0.47

    0.74

    23.6 / 25.2*

    PCR+ / Cx+ / cCDI+ (89)

    0.51

    0.73

    0.51

    0.76

    23.2 / 26.1*

    PCR+ / CCCNA+ / cCDI+ (63)

    0.73

    0.77

    0.72

    0.80

    21.8 / 26.3**

    Ref+ vs. Ref- (Wilcoxon rank-sum): *P<0.05; **P< 0.0001


    Figure 1: ROC Curve of PCR Ct to Identify PCR+ / CCCNA+ / cCDI+ Children

    Aakash Balaji, BS, Northwestern University Feinberg School of Medicine, Chicago, IL, Robyn Espinosa, MPH, Ann & Robert H. Lurie Children's Hospital of Chica, Chicago, IL, Kathleen Todd, BS, MT(ASCP), Microbiology, Ann & Robert H. Lurie Children's Hospital of Chicago, Chicago, IL, Kerry Steed, Research Assistant, Northwestern University, Chicago, IL and Larry Kociolek, MD, MSCI, Pediatric Infectious Diseases, Ann & Robert H. Lurie Children's Hospital of Chicago, Chicago, IL; Pediatrics, Northwestern University Feinberg School of Medicine, Chicago, IL

    Disclosures:

    A. Balaji, None

    R. Espinosa, None

    K. Todd, None

    K. Steed, None

    L. Kociolek, Alere/Techlab: Investigator , Research support .

    Findings in the abstracts are embargoed until 12:01 a.m. PDT, Wednesday Oct. 3rd with the exception of research findings presented at the IDWeek press conferences.