Enhanced discrimination of <i>Clostridioides difficile</i> transmission using whole-genome sequencing and <i>in silico</i> multi-locus variable number tandem repeat analysis

Whole-genome sequencing (WGS) is used to establish genetic relatedness of bacteria and track outbreaks in healthcare settings. While WGS provides sufficient discriminatory power to make inferences about genetic relatedness and transmission for most bacterial species, WGS for <i>Clostridioides difficile</i> often fails to do so, even at low single-nucleotide polymorphism (SNP) differences. Multi-locus variable number tandem repeat analysis (MLVA), which analyzes rapidly mutating tandem repeat loci, has previously been shown to be useful for this purpose for <i>C. difficile</i>. We investigated whether <i>in silico</i> MLVA can further elucidate genetic relatedness of <i>C. difficile</i> clusters identified by short-read WGS but lacking epidemiological links. Potential healthcare-associated toxin-positive <i>C. difficile</i> isolates were collected at our hospital from November 2016 to November 2019. Short-read WGS was performed on the Illumina platform to cluster isolates with ≤2 SNPs, and Nanopore long-read sequencing was used to resolve MLVA loci within these clustered isolates. Among 666 isolates, 62 unique patient isolates met the ≤2 SNP criterion and underwent MinION sequencing. Of the 105 pairs with 0-2 SNP differences, 79.0% had a summed tandem-repeat difference (STRD) of 0-5, 10.5% had an STRD of 6-10, and 10.5% had an STRD of 11-20. A significant correlation was found between a lower STRD value and the presence of a unit/procedure-based epidemiological link within low SNP clusters (odds ratio: 0.45; 95% CI: 0.29-0.70). Our findings demonstrate that MLVA provides additional genomic discrimination for closely related <i>C. difficile</i> isolates identified by WGS, enhancing outbreak investigation precision.IMPORTANCE<i>Clostridioides difficile</i> is a leading cause of healthcare-associated infections, often spreading undetected within hospitals. To track its transmission, hospitals increasingly rely on bacterial genetic sequencing, but this approach is often not discriminatory enough for identifying the spread of this organism between patients. In this study, we applied an additional genetic method that looks at highly changeable regions of the bacteria's DNA to improve the detection of likely transmission events. By combining two sequencing techniques, we were able to separate seemingly related infections that were not actually linked in the hospital. This enhanced resolution can help infection prevention teams focus their investigations and stop real outbreaks more efficiently, improving patient safety. Our findings support the use of this combined sequencing strategy in routine hospital surveillance and show how it can fill important gaps when standard methods are not sufficient.