Evaluation of automated repetitive-sequence-based PCR (DiversiLab) compared to PCR ribotyping for rapid molecular typing of community- and nosocomial-acquired Clostridium difficile.

Automated repetitive PCR (rep-PCR; DiversiLab) was compared to PCR ribotyping of the 16S-23S RNA intergenic spacer of Clostridium difficile (CD) as the "gold standard" method for CD typing. PCR products were separated on DiversiLab LabChips (bioMérieux, St. Laurent, Quebec, Canada) utilizing a 2100 Bioanalyzer (Agilent Technologies, Santa Clara, CA) operating the DiversiLab v1.4 assay. Bioanalyzer data were exported to a secure DiversiLab website and analyzed with DiversiLab v3.4 software. Replicability of each method was verified by confirming that the 5 CD reference strains (RS) formed distinct clusters (CD4, CD6, VL0047, VL0013 [ribotype 027], VL0018 [ribotype 001]) by both typing methods. Ninety randomly selected clinical isolates (CS) were analyzed by both

Methods: 49 from community-acquired and 41 from hospital-acquired cases. A similarity index (SI) of ≥90% was used to define clusters when comparing the known RS cluster to the PCR ribotyping and rep-PCR patterns of CS. Fourteen different PCR-ribotype clusters were identified, but most CS formed 4 major clusters (i.e., CD4 [15/90; 17%], CD6 [17%], 027 [12%], and 001 [9%]). A total of 7 rep-PCR types were identified, but most CS formed 2 major rep-PCR clusters (i.e., CD4 [29/90; 32%] and CD6 [23%]); several PCR ribotypes occurred within a single rep-PCR cluster. Rep-PCR did not distinguish 027 or 001 isolates; i) 027 RS strain did not cluster, ii) eleven 027 CS strains clustered as CD4, iii) no 027 CS strains clustered with the 027 RS, and iv) only 2 001 CS clustered with the RS. Agreement between the PCR-ribotype and rep-PCR clusters only occurred for 35/90 (39%) of the CS using a rep-PCR SI of ≥90%. Rep-PCR time to results was similar, but the annual costs of routinely using this method are 32% higher than PCR ribotyping. Routine use of rep-PCR for CD typing is limited by its lack of definitive separation of the hypertoxigenic 027 or 001 outbreak CD strains.