Genomic epidemiology of Mycobacterium abscessus at an adult cystic fibrosis programme reveals low potential for healthcare-associated transmission

Abstract

Rationale: Nontuberculous mycobacteria (NTM) has been reported to be transmitted between people with cystic fibrosis (CF) attending CF centres. A suspected Mycobacterium abscessus outbreak was investigated at the University of Texas Southwestern (UTSW) Adult CF Program using a combination of pathogen genomic sequencing and epidemiologic methods. The objectives of the present study were to apply the Healthcare-Associated Links in Transmission of NTM (HALT NTM) study to investigate the occurrence of potential healthcare-associated transmission and/or acquisition of NTM among people with CF infected with genetically similar NTM isolates.

Methods: Whole-genome sequencing of respiratory M. abscessus isolates from 50 people with CF receiving care at UTSW was performed to identify genetically similar isolates. Epidemiologic investigation, comparison of respiratory and environmental isolates, and home residence watershed mapping were studied.

Measurements and main results: Whole-genome sequencing analysis demonstrated seven clusters of genetically similar M. abscessus (four ssp. abscessus and three ssp. massiliense). Epidemiologic investigation revealed potential opportunities for healthcare-associated transmission within three of these clusters. Healthcare environmental sampling did not recover M. abscessus, but did recover four human disease-causing species of NTM. No subjects having clustered infections lived in the same home residence watershed. Some subjects were infected with more than one M. abscessus genotype, both within and outside of the dominant circulating clones.

Conclusions: Healthcare-associated person-to-person transmission of M. abscessus appears to be rare at this centre. However, polyclonal infections of M. abscessus species and subspecies, not originating from the endemic hospital environment, suggest multiple shared modes of acquisition outside the healthcare setting.

FIGURE 1

Overview of the healthcare-associated transmission and/or environmental acquisition of nontuberculous mycobacteria (NTM) investigation. Investigational outcomes include healthcare environmental sample collection of the clinic and hospital settings demonstrating NTM diversity; the number of clusters with overlapping encounters in the healthcare environment; and pangenome analysis of paired respiratory isolates. Whole-genome sequencing (WGS) criteria were defined as more than one positive NTM culture within the abstraction timeframe. pwCF: people with cystic fibrosis; M. abscessus: Mycobacterium abscessus. ^#: largest number identified between 2013 and 2018; ^¶: defined as having more than one NTM isolate and available for research purposes.

FIGURE 2

Phylogenetic analysis of Mycobacterium abscessus isolates among people with cystic fibrosis (CF) at the University of Texas Southwestern Medical Center (UTSW) Adult CF Program. Reference isolates, including type strains and previously published isolate genomes, are shown in grey circles with labels. Isolates from subjects not identified in a cluster are represented in unlabelled grey circles. The sets of clustered nontuberculous mycobacteria isolates are represented as coloured circles, with each letter representing a cluster. ATCC: American Type Culture Collection; ^T: type strain; DCC: dominant circulating clone; SNP: single nucleotide polymorphism.

FIGURE 3

Cluster network analysis. Nontuberculous mycobacteria species are identified by colour, and the letter identifies the cluster. Nodes represent each subject within a cluster. The number in the node identifies the within-cluster subject order based on the first positive culture. Four subjects are identified in one or more Mycobacterium abscessus clusters (M. abscessus polyclonal clustered infections). 10 subjects identified with clustered M. abscessus infections also grew M. avium complex. Core genome single nucleotide polymorphism (SNP) distances are shown as solid lines connecting the nodes.

FIGURE 4

Timeline overlap analysis of people with cystic fibrosis identified in Mycobacterium abscessus clusters demonstrating clinic visits, hospitalisation days, same-day clinic overlap and hospitalisation overlap. Clusters that did not have contact overlaps are not shown. NTM: nontuberculous mycobacteria. ^#: subjects with more than one isolate.

FIGURE 5

Integrated pangenome analysis comparing single nucleotide polymorphism (SNP) differences between pairs of Mycobacterium abscessus isolates within clusters versus percentage shared accessory genome. Cluster G as well as M. abscessus ssp. abscessus clusters C and E demonstrated episodes of healthcare-associated opportunities for transmission. Cut-off points of ≤10 SNPs and ≥95% shared accessory genome are noted in the dashed-line box, representing the most stringent degree of clustered isolate similarity. Multiple clustered isolate pairs fall within the stringent genetic similarity cut-offs. Two paired isolates from M. abscessus ssp. massiliense cluster G with healthcare-associated opportunity for transmission from G1 to G7 and G3 to G6 fall within the stringent similarity, while the remaining isolate pairs that fall within the stringent similarity do not have observed healthcare overlaps. M. abscessus ssp. abscessus cluster D, consisting of two subjects with no evidence of healthcare-associated opportunity for transmission, also falls within the stringent similarity cut-offs. Cluster C, consisting of three subjects, has one pair without observed overlaps falling into the stringent similarity cut-offs, while two isolate pairs with healthcare overlaps fall just outside the SNP cut-off for stringent similarity, but meet the shared accessory genome cut-off.

FIGURE 6

Ordinal levels of transmission. Ordinal levels for Mycobacterium abscessus transmission among subjects identified with a clustered infection were created in the following manner. Each subject was paired with each of the other subjects in the cluster. Single nucleotide polymorphism (SNP) distances and shared accessory genome values were used to create the pangenome heatmap. An “unlikely” event was defined as >30 SNPs difference or <90% shared genome. A “possible” event was defined as >15–30 SNPs difference and >90–95% shared genome. A “likely” event was defined as 0–15 SNPs or >95% shared genome. A “very likely” event was defined as ≤10 SNPs difference and ≥95% shared genome. Overlaps were described as the number of days two subjects were in the same healthcare space, the location they occurred, and if they were of interest. The determination of overlap of interest was assigned “yes” if there was biological plausibility for a potential transmission event, either acquisition or transmission from another subject based on epidemiologic review of the healthcare record. An overlap of interest was assigned “no” if both subjects were M. abscessus positive at the time the overlap was observed or the overlap occurred, but the locations where subjects received care was in two different healthcare spaces. The epidemiological investigation heatmap was estimated based on biological plausibility of transmission based on the strength of the epidemiological findings, and categories included no evidence, weak evidence, plausible evidence and strong evidence. Composite variables were determined in the following manner. The composite SNP variable for a subject is the percentage of pairs with <10 SNP difference among all possible pairs between the given subject and others in the same cluster. The composite shared accessory genome variable is defined similarly, but considering pairs with ≥95% shared data. H: hospital; C: clinic.

FIGURE 6

Ordinal levels of transmission. Ordinal levels for Mycobacterium abscessus transmission among subjects identified with a clustered infection were created in the following manner. Each subject was paired with each of the other subjects in the cluster. Single nucleotide polymorphism (SNP) distances and shared accessory genome values were used to create the pangenome heatmap. An “unlikely” event was defined as >30 SNPs difference or <90% shared genome. A “possible” event was defined as >15–30 SNPs difference and >90–95% shared genome. A “likely” event was defined as 0–15 SNPs or >95% shared genome. A “very likely” event was defined as ≤10 SNPs difference and ≥95% shared genome. Overlaps were described as the number of days two subjects were in the same healthcare space, the location they occurred, and if they were of interest. The determination of overlap of interest was assigned “yes” if there was biological plausibility for a potential transmission event, either acquisition or transmission from another subject based on epidemiologic review of the healthcare record. An overlap of interest was assigned “no” if both subjects were M. abscessus positive at the time the overlap was observed or the overlap occurred, but the locations where subjects received care was in two different healthcare spaces. The epidemiological investigation heatmap was estimated based on biological plausibility of transmission based on the strength of the epidemiological findings, and categories included no evidence, weak evidence, plausible evidence and strong evidence. Composite variables were determined in the following manner. The composite SNP variable for a subject is the percentage of pairs with <10 SNP difference among all possible pairs between the given subject and others in the same cluster. The composite shared accessory genome variable is defined similarly, but considering pairs with ≥95% shared data. H: hospital; C: clinic.

FIGURE 7

Environmental collection summary at University of Texas Southwestern Medical Center. M.: Mycobacterium; NTM: nontuberculous mycobacteria. ^#: locations with multiple NTM isolates recovered.