Defining criteria to interpret multilocus variable-number tandem repeat analysis to aid Clostridium difficile...

Area of Science:

• Microbiology
• Infectious Diseases
• Epidemiology

Background:

• * Pulsed-field gel electrophoresis (PFGE) is the North American standard for Clostridium difficile surveillance but has limited discriminatory power for outbreak investigations.
• * The high prevalence of the North American pulsotype (NAP) 1 strain in hospitals complicates PFGE-based outbreak analysis.
• * Multilocus variable-number tandem repeat analysis (MLVA) shows promise for improved discriminatory power, but clear criteria for defining clonality are needed.

Purpose of the Study:

• * To evaluate the discriminatory power of MLVA for Clostridium difficile infection (CDI) outbreak investigation.
• * To establish criteria for defining clonality using MLVA within the epidemic NAP1 strain.
• * To compare MLVA with PFGE for surveillance and outbreak detection of CDI.

Main Methods:

• * Conducted surveillance for toxin-positive CDI at an academic health sciences center from September 2009 to April 2010.
• * Subjected 86 CDI episodes from 74 patients to both PFGE and MLVA typing.
• * Analyzed results using Bionumerics software for phylogenetic tree generation and coupled with demographic data; population analysis of multiple colonies and outbreak typing were performed.

Main Results:

• * MLVA identified two, four, or seven distinct clusters among NAP1 strains using 90%, 95%, or 97% similarity cut-offs, respectively, with Manhattan distance-based clustering.
• * Population analysis indicated that a 1-3% difference in MLVA types is typical for isolates from a single individual.
• * Typing in the context of three institutional outbreaks (42 isolates) demonstrated MLVA's ability to delineate outbreak clusters within the NAP1 strain.

Conclusions:

• * MLVA provides superior discriminatory power compared to PFGE for Clostridium difficile outbreak investigations.
• * A 95-97% similarity cut-off using Manhattan distance-based clustering is effective for distinguishing outbreak clones from sporadic CDI isolates.
• * Combining longitudinal surveillance, epidemiological data, population analysis, and outbreak typing enhances the utility of MLVA for CDI epidemiology.