Genomic analysis of Candida auris transmission within an urban region

Abstract

Candida auris emerged in Chicago, IL, USA, in 2016 and has since become endemic. We used whole-genome sequencing (WGS) of 494 isolates, epidemiologic metadata and patient transfer data to describe the transmission of C. auris among Chicago healthcare facilities between 2016 and 2021. In total, 99% of isolates formed a single clade IV phylogenetic lineage, suggesting a single introduction. The remaining isolates were clade III. Isolates were grouped into genomic clusters based on phylogenetic clustering and relatedness. Isolates from 14 facilities were included in 77 genomic clusters with evidence of inter- and intra-facility transmission; 62% of isolates included in genomic clusters clustered with another from the same facility. Patient transfer data corroborated transmission among facilities in large genomic clusters, and patient transfers between facilities correlated with more similar genomes. In total, 30% of individuals who had serial samples collected over time had isolates collected during different facility stays. Most of these individuals carried closely related isolates over time, although 25% carried isolates whose genomes diverged. Integrating WGS data and patient metadata to describe inter- and intra-facility clusters revealed regional transmission patterns within and between healthcare facilities that could be used to guide targeted interventions to prevent transmission. Given that patient transfer data were concordant with WGS analysis, they may serve as a predictive tool for risk of inter-facility transmission.

Keywords: Candida auris; fungal pathogens; genomic epidemiology; healthcare associated infections; phylogenetics; whole-genome sequencing.

Fig. 1.. Chicago C. auris cases by specimen type, 1 August 2018–31 December 2021. New screening cases (grey) and clinical infections (black) are indicated in bar colour. A person may be counted twice, as both a screening and clinical case.

Fig. 2.. Phylogenetic and SNP analysis of Chicago clade IV C. auris isolates. (a) Phylogenetic tree of Illinois sequences with tree tips coloured by isolate facility. Metadata block to the right of tree tips represents isolate collection year. Isolates included in clusters and branches leading to identified clusters are bolded in alternating black and red for clarity. Asterisks indicate clusters with five or more isolates, which were selected for subtree visualization in Fig. 3. Scale bar indicates SNP distance. (b) and (c) Violin plots of pairwise SNP distances with dashed lines showing median (red) and 25th and 75th quartiles (black); individual data points shown when <100. Isolate numbers shown above plots in (b). Dashed horizontal line in panel (c) indicates the maximum SNP distance of 7 used for cluster selection.

Fig. 3.. Illinois C. auris cluster analysis. (a–e) Histograms showing the number of isolates in each cluster (a), the pairwise SNP distances contained in each cluster (b), bootstrap support for each cluster (c) intra/inter-facility relationship between pairwise comparisons of isolates with facility metadata (d) and the intra/inter-facility relationship between pairwise comparisons with facility metadata within clusters containing various numbers of isolates (e). (f–j) Subtrees showing clusters with five or more isolates. Subtree tips are coloured by isolate facility and relevant isolate/patient metadata are located to the right of tree tips. Scale bar indicates SNPs.

Fig. 4.. Patient transfer networks for facilities with sequenced C. auris. Network analysis of direct patient transfers (connections) for facilities (nodes) from (a) all sequenced isolates and (b) facilities included in large clusters corresponding to Fig. 3(f–j). Facilities included in large clusters are coloured according to Fig. 3. Connecting lines are weighted by the number of transfers, with the finest lines indicating fewer than ten transfers. Vertex shape indicates facility type: ACH, LTACH, IRF, vSNF and SNF.

Fig. 5.. More frequent patient transfers are associated with more closely related C. auris genomes. Association between bi-directional patient flow, calculated from facility transfer networks, and (a) the number of inter-facility isolate pairs with three or fewer SNP differences, selected based on the median within-cluster SNP distance or (b) the minimum pairwise SNP distance between inter-facility pairs.

Fig. 6.. Relatedness of isolates collected from the same individual. Phylogenetic tree as in Fig. 2(a) with tree tips coloured by (a) individuals with isolates showing phylogenetic linkage and/or ≤7 SNP differences and (b) individuals with isolates showing divergent clustering and >7 SNP differences. Isolates from one individual highlighted in subtrees in panel (d) are circled. (c) Correlation between pairwise SNP distance and the number of days between pairwise isolates collected longitudinally from one individual with linear regression. (d) Subtrees of one individual with divergent longitudinal isolates [circled in panel (b)] with relevant metadata in blocks to the right of tree tips. Tree tips indicate individuals, as in panel (b).