SUMMARY
SETTING
The impact of the genetic characteristics of Mycobacterium tuberculosis on the clustering of multi-drug-resistant tuberculosis (MDR-TB) has not been analyzed together with clinical and demographic characteristics.
OBJECTIVE
To determine factors associated with genotypic clustering of MDR-TB in a community-based study.
DESIGN
We measured the proportion of clustered cases among MDR-TB patients and determined the impact of clinical and demographic characteristics and that of three M. tuberculosis genetic characteristics: lineage, drug resistance-associated mutations, and rpoA and rpoC compensatory mutations.
RESULTS
Of 174 patients from California and Texas included in the study, the number infected by East-Asian, Euro-American, Indo-Oceanic and East-African-Indian M. tuberculosis lineages were respectively 70 (40.2%), 69 (39.7%), 33 (19.0%) and 2 (1.1%). The most common mutations associated with isoniazid and rifampin resistance were respectively katG S315T and rpoB S531L. Potential compensatory mutations in rpoA and rpoC were found in 35 isolates (20.1%). Hispanic ethnicity (OR 26.50, 95%CI 3.73–386.80), infection with an East-Asian M. tuberculosis lineage (OR 30.00, 95%CI 4.20–462.40) and rpoB mutation S531L (OR 4.03, 95%CI 1.05–23.10) were independent factors associated with genotypic clustering.
CONCLUSION
Among the bacterial factors studied, East-Asian lineage and rpoB S531L mutation were independently associated with genotypic clustering, suggesting that bacterial factors have an impact on the ability of M. tuberculosis to cause secondary cases.
Keywords: community-based study, transmission, molecular epidemiology
Studies have shown that patients with tuberculosis (TB) are heterogeneous when transmitting Mycobacterium tuberculosis.1 Most of this variability has been attributed to features of the index patient, such as the number of organisms expelled on coughing.2 However, some studies have suggested that bacterial characteristics have an impact on the transmission of M. tuberculosis and on its pathogenicity, i.e., its ability to cause secondary cases of TB. In one study, organisms of the East-Asian lineage, including the Beijing family, were five times more likely to cause secondary cases than patients with other M. tuberculosis lineages.3 In other studies, patients with drug-resistant M. tuberculosis were less likely to cause secondary cases than susceptible M. tuberculosis.4,5 Although some studies have shown no differences,6 other studies have suggested that isoniazid (INH) resistant M. tuberculosis with mutations in the inhA promoter or with S315T katG mutations were more likely to be transmitted than those without these mutations.7,8 Furthermore, in vitro fitness studies demonstrated that resistant M. tuberculosis strains with the most common rifampin (RMP) resistance-associated mutation, rpoB S531L, were more fit than strains with less frequent mutations, such as rpoB H526Y.9 Interestingly, mutations in rpoA and rpoC have been observed at high frequency in RMP-resistant M. tuberculosis, and are considered to be compensatory mutations for any fitness loss that could be caused by the RMP resistance-associated mutation.9 It should be noted that the frequency of strains with these compensatory mutations was high in regions with a high burden of multidrug-resistant TB (MDR-TB; i.e., TB resistant to at least INH and RMP),10 and more frequent among RMP-resistant M. tuberculosis isolates that caused secondary cases.11,12 However, these studies did not control for other factors known to be associated with transmission.
In this report, we describe the demographic, clinical and bacterial factors associated with genotypic clustering in MDR-TB cases. Genotypic clustering has been used as an indicator of M. tuberculosis involved in chains of transmission. We include an analysis of three bacterial characteristics: M. tuberculosis lineage, drug resistance-associated mutations, and presence of compensatory mutations in rpoA and rpoC.
STUDY POPULATION AND METHODS
Study population and data sources
We included all patients with pulmonary TB caused by MDR M. tuberculosis organisms (defined by phenotypic drug susceptibility testing) identified between January 2005 and December 2011 from eight health jurisdictions in California (Sacramento, San Mateo, Contra Costa, Alameda, San Diego, Santa Clara, San Francisco and Orange counties), and in five jurisdictions in Texas (Harris, Dallas, Tarrant, Hidalgo and Cameron counties). Some of these patients were enrolled in a parent study to investigate MDR-TB transmission in the United States.13 The Human Research Protection Program of the University of California, San Francisco (UCSF), CA, USA, of the Centers for Disease Control and Prevention (CDC) and of each participating institution approved the study protocol.
Data on demographic and clinical characteristics and the epidemiologic links between patients were collected as part of the standard of care and the MDR-TB transmission study.13 Lineage of the M. tuberculosis isolates was described according to the phylogenetic characterization methodologies previously reported.14 INH resistance-associated mutations in katG, the inhA promoter and the RMP resistance-associated mutations in the RMP-resistance determining region (RRDR) of rpoB were identified using pyrosequencing.15 Only amino acid positions 1–191 of rpoA (Rv3457c) and positions 245–560 of rpoC (Rv0668) (hotspots for possible compensatory mutations10) were sequenced using rpoA, primers (F5 ′GGACGTCGAAAGGAAGAAGA3 ′ and R5′GTCTCCACGTCCAGGATCAG3′) and rpoC primers (F5′CGAAAACCTCTACCGCGAAC3′ and R5′GCGACAGGATGTTGTTGGAG3′), respectively.11 Polymerase chain reaction products were sequenced using an ABI377 automatic DNA sequencer (Perkin Elmer, Applied Biosystems, Carlsbad, CA, USA) at the UCSF genomics core facility. Sequence polymorphisms were identified by comparing the consensus sequence of each isolate to the corresponding gene sequence of the H37Rv genome using the A Plasmid Editor, V2.0.46 (W Davis, University of Utah, Salt Lake City, UT, USA).
All isolates were genotyped using spoligotyping and 24-locus mycobacterial interspersed repetitive units (MIRU) typing, as part of the CDC National Tuberculosis Genotyping Service surveillance system,16 and with insertion sequence (IS) 6110 restriction fragment length polymorphism (RFLP) using standardized methods.17 Genotypic clustering was defined as two or more MDR-TB isolates from patients from the same state with identical spoligo-type, 24-MIRU type, IS6110-RFLP and known drug resistance-associated mutations. Compensatory mutations in rpoA and rpoC were not included for the definition of clustering. We assumed that patients with clustered MDR-TB isolates within each state had TB due to recent transmission and were part of a chain of transmission. Patients with unique genotypes were considered as having TB due to the reactivation of latent infection. We used the ‘n–1’ method to calculate the transmission index that measures the average number of subsequent cases produced by potential index cases.18
Statistical analysis
We described the demographic, clinical and bacterial characteristics of M. tuberculosis isolates associated with M. tuberculosis genetic clustering (the outcome) using logistic regression and exact logistic regression where expected cell counts were <5. We performed full-fitted exact logistic regression models, first including variables with P < 0.25 in the unadjusted analysis, then including those with P < 0.20. Correlated variables were removed from the model to identify the most stable, parsimonious, and informative model possible. P < 0.05 was considered statistically significant. Statistical analysis was performed using SPSS (Statistical Package for the Social Sciences, version 18.0, IBM Corp, Armonk, NY, USA) and SAS (version 9.4, SAS Institute Inc., Cary, NC, USA).
RESULTS
Patient characteristics
During the study period, 169 patients from California and 38 from Texas had pulmonary TB caused by MDR organisms, including four with XDR-TB (i.e., MDR-TB with additional resistance to fluoroquinolones and second-line injectable drugs) and 29 pre-XDR-TB (MDR-TB with resistance to either fluoroquinolones or second-line injectable drugs). After excluding cases without M. tuberculosis DNA or clinical data, 140 patients (83%) from California and 34 (89%) from Texas were included in the analysis. The characteristics of included vs. excluded patients are shown in Table 1. The mean age of the 174 patients included was 39.1 (±16) years, 80 (46%) were female and most were Asian (n= 104, 59.8%) or White (n = 63, 36.2%) (Table 2).
Table 1.
Clinical characteristics of included and excluded patients
| Included (n = 174 | Excluded (n = 33) | ||
|---|---|---|---|
| n (%) | n (%) | P value | |
| Age, years, mean ± SD | 39.1 ± 16 | 37.7 ± 16.6 | 0.646 |
| Female sex | 80 (46) | 16 (48.5) | 0.881 |
| Race* | 0.497 | ||
| Asian | 104 (59.8) | 20 (60.6) | |
| White | 63 (36.2) | 11 (33.3) | |
| African American | 4 (2.3) | 2 (6.1) | |
| Unknown | 3 (1.7) | 0 | |
| Hispanic | 55 (31.6) | 7 (21.2) | 0.225 |
| Prior anti-tuberculosis treatment | 46 (26.4) | 12 (36.4) | 0.325 |
| HIV infection* | 5 (2.9) | 0 | 0.380 |
| Diabetes mellitus* | 7 (4) | 1 (3) | 0.786 |
| BCG vaccination | 33 (19) | 5 (15.2) | 0.604 |
| Homeless in last 12 months* | 7 (4) | 0 | 0.602 |
| Correctional facility at time of diagnosis* | 5 (2.9) | 3 (9.1) | 0.113 |
| Long-term facility at time of diagnosis* | 2 (1.1) | 1 (3) | 0.392 |
| Injecting drug user in last 12 months* | 2 (1.1) | 0 | 1.000 |
| Excessive alcohol use in last 12 months* | 22 (12.6) | 4 (12.1) | 1.000 |
As expected cell counts were <5, P value was calculated using Fisher’s exact test.
SD = standard deviation; HIV = human immunodeficiency virus; BCG = bacille Calmette-Guérin.
Table 2.
Clinical and microbiological characteristics of patients with and without M. tuberculosis rpoA or rpoC mutations and ORs of having mutations in rpoA and rpoC genes
|
rpoA or rpo C mutations
|
|||||
|---|---|---|---|---|---|
| All (n = 174) | Yes (n = 35) | No (n = 139) | |||
| n (%) | n (%) | n (%) | OR (95%CI) | P value | |
| Age, years, mean ± SD | 39.1 ± 16 | 35.2 ± 13.9 | 40.1 ± 16.3 | 0.98 (0.95–1.01) | 0.115 |
| Female sex | 80 (46) | 22 (62.9) | 58 (41.7) | 2.25 (1.05–4.83) | 0.038 |
| Race* | |||||
| Asian | 104 (59.8) | 20 (57.1) | 84 (60.4) | 1.00 | — |
| White | 63 (36.2) | 14 (40) | 49 (35.3) | 1.20 (0.51–2.76) | 0.782 |
| African American | 4 (2.3) | 1 (2.9) | 3 (2.2) | 1.40 (0.03–18.50) | 1.000 |
| Unknown | 3 (1.7) | 0 | 3 (2.2) | — | — |
| Hispanic ethnicity | 55 (31.6) | 12 (34.3) | 43 (30.9) | 1.15 (0.53–2.53) | 0.723 |
| Prior anti-tuberculosis treatment | 46 (26.4) | 10 (28.6) | 36 (25.9) | 1.16 (0.50–2.69) | 0.727 |
| HIV infection* | 5 (2.9) | 0 | 5 (3.6) | — | — |
| Diabetes mellitus* | 7 (4) | 0 | 7 (5) | — | — |
| BCG vaccination | 33 (19) | 5 (14.3) | 28 (20.1) | 0.66 (0.24–1.86) | 0.432 |
| Homeless in last 12 months* | 7 (4) | 1 (2.9) | 6 (4.3) | 0.65 (0.01–5.68) | 1.000 |
| Correctional facility at time of diagnosis* | 5 (2.9) | 3 (8.6) | 2 (1.4) | 6.23 (0.69–77.50) | 0.115 |
| Long-term facility at time of diagnosis* | 2 (1.1) | 0 | 2 (1.4) | — | — |
| Injecting drug use in last 12 months* | 2 (1.1) | 1 (2.9) | 1 (0.7) | 4.05 (0.05–323.40) | 0.721 |
| Excessive alcohol use in last 12 months* | 22 (12.6) | 4 (11.4) | 18 (12.9) | 0.91 (0.21–3.04) | 1.000 |
| M. tuberculosis lineages* | |||||
| East-Asian | 70 (40.2) | 12 (34.3) | 58 (41.7) | 1.00 | — |
| Euro-American | 69 (39.7) | 15 (42.9) | 54 (38.8) | 1.34 (0.53–3.45) | 0.638 |
| Indo-Oceanic | 33 (19) | 6 (17.1) | 27 (19.4) | 1.07 (0.30–3.50) | 1.000 |
| East-African Indian | 2 (1.1) | 2 (5.7) | 0 | — | — |
| Isoniazid resistance-associated mutations in:† | |||||
| katG S315T | 124 (71.3) | 31 (88.6) | 93 (66.9) | 3.83 (1.28–11.50) | 0.016 |
| inhA promoter ‡ | 40 (23) | 4 (11.4) | 36 (25.9) | 0.37 (0.12–1.12) | 0.078 |
| Rifampin resistance-associated mutations | |||||
| rpoB S531L | 105 (60.3) | 29 (82.9) | 76 (54.7) | 4.01 (1.56–10.30) | 0.003 |
| No rpoB S531L and wt rpoB§ | 69 (39.7) | 6 (17.1) | 63 (45.3) | 1 | — |
As expected cell counts were <5, OR, (95%CI) and P values were calculated using exact logistic regression.
10 isolates did not have mutations in katG or inhA.
36 isolates had a mutation −15T, 2 isolates −8A, and 2 isolates −17T.
There were six isolates with wt rpoB and 63 cases with 68 rpoB mutations other than S531L (five patients had two mutations each): H526D (n = 18), H526Y (n = 12), D516V (n= 8), H526R (n= 5), L533P (n= 3) Q513L (n= 2), D516Y (n= 2), N519K (n= 2), H526A (n= 2), H526C (n= 2), H526L (n= 2), H526Q (n= 1), Q510L (n = 1), L511P (n = 1), Q513K (n = 1), D516F (n = 1), S522L (n = 1), 525del (n = 1), H526N (n = 1), K527P (n = 1), H526S (n = 1). Five patients had two mutations and the following combinations: D516V and Q510L, D516Y and 525del, H526N and Q513L, H526Q and L533P, H526S and K527P.
OR = odds ratio; CI = confidence interval; SD = standard deviation; HIV = human immunodeficiency virus; BCG = bacille Calmette-Guérin; wt = wild-type.
Genetic characteristics of M. tuberculosis
Of the 174 patients, the number infected by East-Asian, Euro-American, Indo-Oceanic and East-African-Indian lineages were respectively 70 (40.2%), 69 (39.7%), 33 (19%) and 2 (1.1%) (Table 2). The INH resistance-associated mutation, katG S315T, and mutations in the inhA promoter were observed in respectively 124 (71.3%) and 40 (23%) of the isolates. The remaining 10 isolates did not have mutations in these genes, and none had mutations in both genes. The RMP resistance-associated mutation rpoB S531L (n = 105, 60.3%) (Table 2) was the most frequent mutation, followed by H526D (n = 18, 10.3%) and H526Y (n = 12, 6.9%). Five patients had isolates with two mutations each in rpoB (Table 2 footnote). M. tuberculosis from six patients did not have a mutation in rpoB RRDR.
We identified four unique single nucleotide polymorphisms (SNPs) in rpoA and 17 unique SNPs in rpoC, of which we excluded two from the analysis: A542A (1626 C→G), associated with Euro-American lineage but not with resistance,10,11 and the synonymous mutation R480R (1440 C→T). These SNPs were observed in respectively 23 and 2 patients. The 19 non-synonymous potentially compensatory mutations were found in 35 M. tuberculosis isolates (35/174, 20.1%) (Table 3). Five isolates had two mutations each in the rpoC gene. Twelve unique isolates had rpoC mutations in codon 483 (V483G, n = 9 and V483A, n = 3) and six had the A466V mutation. The mutation P434V in rpoC was associated with two haplotypes (1300 C→G and 1301 C→T), and occurred in the same isolate (Table 3). The A466V and D271G mutations have not been reported previously. There were four different rpoA mutations in four isolates (4/174, 2.3%), two of which had not been reported before (G115A, T127A) (Table 3).
Table 3.
rpo A and rpoC mutations in MDR-TB isolates
| Gene | Mutation site | Isolates n | Nucleotide substitution | Amino acid change | References |
|---|---|---|---|---|---|
| rpoC | 812 | 1 | A→G | D271G | Not reported |
| rpoC | 994 | 1 | G→A | G332S | 19, 20 |
| rpoC | 1297 | 1 | G→A | G433S | 10, 12, 19, 20 |
| rpoC | 1300 | 1 | C→G | P434V | 21 |
| rpoC | 1301 | 1 | C→T | P434V | 21 |
| rpoC | 1301 | 1 | C→A | P434Q | 19, 20 |
| rpoC | 1354 | 3 | T→C | F452L | 21, 22 |
| rpoC | 1397 | 6 | C→T | A466V | Not reported |
| rpoC | 1448 | 3 | T→C | V483A | 11, 12, 19–22 |
| rpoC | 1448 | 9 | T→G | V483G | 10–12, 20, 21 |
| rpoC | 1471 | 3 | A→G | I491V | 10–12, 19, 20 |
| rpoC | 1519 | 3 | T→G | L507V | 12, 21 |
| rpoC | 1547 | 1 | T→C | L516P | 10, 19–21 |
| rpoC | 1562 | 1 | C→A | A521D | 10, 12, 20 |
| rpoC | 1573 | 1 | C→A | H525N | 10, 12 |
| rpoA | 344 | 1 | G→C | G115A | Not reported |
| rpoA | 379 | 1 | A→G | T127A | Not reported |
| rpoA | 559 | 1 | A→G | T187A | 10, 12, 20, 21 |
| rpoA | 569 | 1 | A→G | D190G | 20, 21 |
MDR-TB = multidrug-resistant tuberculosis.
The frequency of M. tuberculosis isolates with rpoA or rpoC mutations varied depending on the rpoB mutation: they were more frequent in isolates with the S531L mutation (82.9% vs. 17.2%, odds ratio [OR] 4.01, 95% confidence interval [CI] 1.56–10.03) (Table 2). The proportion of M. tuberculosis isolates with rpoA and rpoC mutations was different among the different lineages, although this difference was not statistically significant (Table 2).
Genotypic clustering
Of the 174 M. tuberculosis isolates, 23 (13.2%) were placed in eight genotypic clusters (Table 4). Five clusters were composed of East-Asian lineage isolates and three of Euro-American lineage isolates. The transmission indices in East-Asian and Euro-American lineage were respectively 0.143 and 0.072. Isolates in two clusters had different rpoC mutations. In cluster 1, the patient with the F452L mutation was the earliest case in the cluster (December 2005) based on the date of diagnosis. The remaining two patients were diagnosed in December 2007 (G332T mutation) and August 2008 (F452L mutation). In cluster 7, the patient with the double mutation was reported in January 2007 and the patient with the wild-type rpoA and rpoC in May 2009. Patients in cluster 7 reported knowing each other. The only other epidemiologic link reported was among two of the four patients in cluster 3.
Table 4.
Clusters identified in MDR-TB isolates from 140 patients from California and 34 from Texas
| Clusters | Location | Isolates n | Lineage | rpoA mutations | rpoC mutations | rpoB mutations |
|---|---|---|---|---|---|---|
| 1 | California | 3 | East-Asian | None | G332T in 1 | S531L |
| F452L in 2 | S531L | |||||
| 2 | California | 4 | Euro-American | None | wt in 4 | S531L |
| 3 | California | 4 | East-Asian | None | wt in 4 | S531L |
| 4 | California | 4 | East-Asian | None | wt in 4 | H526D |
| 5 | California | 2 | East-Asian | None | wt in 2 | S531L |
| 6 | Texas | 2 | Euro-American | None | wt in 2 | S531L |
| 7 | Texas | 2 | Euro-American | None | A466V and L507V in 1; wt in 1 | H526D |
| 8 | Texas | 2 | East-Asian | None | wt in 2 | S531L |
MDR-TB = multidrug-resistant tuberculosis; wt = wild-type.
In the unadjusted analysis (Table 5), Hispanics (OR 2.72, 95%CI 1.11–6.62), patients with excessive alcohol consumption (OR 6.61, 95%CI 2.11–20.50), and those infected with an East-Asian lineage (OR 3.27, 95%CI 1.30–8.21) were more likely to be in genotypic clusters. The adjusted analysis using all values with P < 0.25 was unstable due to the small number of outcomes (cluster). The more parsimonious model using values with P < 0.20 was unstable due to interactions between excessive alcohol use, Asian race, and East-Asian lineage. Alcohol was reported in one Asian patient, and Asian race was correlated with East-Asian lineage; we therefore removed alcohol and race from the adjusted model. The most stable and parsimonious model showed that Hispanic ethnicity (OR 26.5, 95%CI 3.73–386.80), infection with an East-Asian M. tuberculosis lineage (OR 30.0, 95%CI 4.20–462.40) and the presence of an rpoB S531L mutation (OR 4.03, 95%CI 1.05–23.10) were independently associated with genotypic clustering (Table 5). To explore the role of excessive alcohol use, we performed a similar analysis, stratified by race, and found that among non-Asians, excessive alcohol was associated with genotypic clustering (Table 6).
Table 5.
Clinical and microbiological factors associated with clustering
| Clustering*
|
Unadjusted
|
Adjusted
|
||||
|---|---|---|---|---|---|---|
| Yes (n = 23) | No (n = 151) | |||||
| n (%) | n (%) | OR (95%CI) | P value | OR (95%CI) | P value | |
| Age ≥65 years† | 2 (8.7) | 9 (6.0) | 1.56 (0.15–8.37) | 0.846 | ||
| Male sex | 11 (47.8) | 79 (52.3) | 0.79 (0.33–1.90) | 0.598 | ||
| Asian race | 11 (47.8) | 95 (62.9) | 0.54 (0.22–1.31) | 0.171 | ||
| Hispanic ethnicity | 12 (52.2) | 43 (28.5) | 2.72 (1.11–6.62) | 0.028 | 26.50 (3.73–386.80) | <0.001 |
| Previous anti-tuberculosis treatment | 3 (13.0) | 43 (28.5) | 0.46 (0.13–1.67) | 0.239 | ||
| HIV infection | 0 | 5 (3.3) | — | — | ||
| Diabetes mellitus† | 3 (13.0) | 4 (2.6) | 5.54 (0.76–35.40) | 0.095 | 3.64 (0.35–35.30) | 0.355 |
| BCG vaccination† | 1 (4.3) | 32 (21.2) | 0.17 (0.004–1.17) | 0.087 | 0.18 (0.004–1.39) | 0.140 |
| Homeless in last 12 months† | 2 (8.7) | 5 (3.3) | 2.72 (0.24–18.00) | 0.473 | ||
| Correctional facility at time of diagnosis† | 1 (4.3) | 4 (2.6) | 1.68 (0.03–18.00) | 1.000 | ||
| Long-term care facility at time of diagnosis | 0 | 2 (1.3) | — | — | ||
| Injecting drug user in last 12 months† | 1 (4.3) | 1 (0.7) | 6.65 (0.08–535.00) | 0.497 | ||
| Excessive alcohol in last 12 months† | 9 (39.1) | 13 (8.6) | 6.61 (2.11–20.50) | <0.001 | ||
| East-Asian lineage infection | 15 (65.2) | 55 (36.4) | 3.27 (1.30–8.21) | 0.012 | 30.00 (4.20–462.40) | <0.001 |
| rpoA or rpoC mutations† | ||||||
| No mutations | 19 (82.6) | 120 (79.5) | 1.00 | — | ||
| Overall mutations | 4 (17.4) | 31 (20.5) | 0.80 (0.19–2.66) | 0.954 | ||
| rpoC mutation | 4 (17.4) | 27 (17.9) | 0.95 (0.22–3.17) | 1.000 | ||
| rpoA mutation | 0 | 4 (2.7) | — | — | ||
| rpoB mutation status | ||||||
| Without S531L mutation | 6 (26.1) | 63 (41.7) | 1.00 | — | ||
| With S531L mutation | 17 (73.9) | 88 (58.3) | 2.03 (0.76–5.43) | 0.159 | 4.03 (1.05–23.10) | 0.040 |
| katG mutation status | ||||||
| Without katG mutation | 4 (17.4) | 46 (30.5) | 1.00 | — | ||
| With katG mutation | 19 (82.6) | 105 (69.5) | 2.08 (0.67–6.46) | 0.205 | ||
Defined as isolates collected in the same state and sharing identical spoligotyping, 24-locus mycobacterial interspersed repetitive units typing, insertion sequence 6110 restriction fragment length polymorphism pattern and drug-resistance-associated mutations.
As expected cell counts were <5, unadjusted ORs and P values were calculated using the exact logistic method.
OR = odds ratio; CI = confidence interval; HIV = human immunodeficiency virus; BCG = bacille Calmette-Guérin.
Table 6.
Most parsimonious model of clinical and microbiological factors associated with clustering in non-Asian patients
| Clustering*
|
Unadjusted
|
Adjusted
|
||||
|---|---|---|---|---|---|---|
| Yes (n = 12) | No (n = 56) | |||||
| n (%) | n (%) | OR (95%CI) | P value | OR (95%CI) | P value | |
| Age ≥65 years† | 0 | 2 (3.7) | Could not calculate | |||
| Hispanic † | 12 (100) | 42 (75.0) | Could not calculate | |||
| Diabetes mellitus† | 3 (25.0) | 1 (1.8) | 17.1 (1.23–978.90) | 0.031 | 10.0 (0.53–751.00) | 0.169 |
| BCG vaccination† | 1 (8.3) | 10 (17.9) | 0.42 (0.01–3.59) | 0.751 | ||
| Injecting drug user in last 12 months† | 1 (8.3) | 1 (1.8) | 4.74 (0.06–392.20) | 0.656 | ||
| Excessive alcohol in last 12 months† | 9 (75.0) | 12 (22.2) | 10.0 (2.09–66.80) | 0.001 | 7.78 (1.55–52.50) | 0.009 |
| East-Asian lineage† | 4 (33.3) | 7 (12.5) | 3.42 (0.60–17.70) | 0.188 | ||
| With S531L mutation† | 6 (50.0) | 36 (65.5) | 0.53 (0.12–2.29) | 0.495 | ||
| With katG mutation† | 8 (66.7) | 40 (72.7) | 0.75 (0.17–3.93) | 0.918 | ||
Defined as isolates collected in the same state and sharing identical spoligotyping, 24-locus mycobacterial interspersed repetitive units typing, insertion sequence 6110 restriction fragment length polymorphism pattern and drug-resistance-associated mutations.
As expected cell counts were <5, unadjusted ORs and P values were calculated using the exact logistic method.
OR = odds ratio; CI = confidence interval; BCG = bacille Calmette-Guérin.
DISCUSSION
In this study of the clustering of patients with TB caused by MDR M. tuberculosis, we found that Hispanic ethnicity, being infected with an M. tuberculosis strain from the East-Asian lineage and with an rpoB S531L mutation were independent risk factors for genotypic clustering of TB cases. Compensatory mutations in rpoA and rpoC were not associated with clustering. To our knowledge, this is the first study to include a systematic analysis of clinical and epidemiologic data together with drug resistance-associated mutations, compensatory mutations and M. tuberculosis lineage on transmission and pathogenesis (measured by clustering) of MDR-TB.
The distribution of the mutations causing INH and RMP resistance was similar to that noted in previous reports.23,24 The frequency of potential compensatory mutations in rpoA and rpoC mutations was also similar to that in other reports.10–12,22 In all studies, rpoC mutations were more frequent than rpoA mutations. Although most of the rpoA and rpoC mutations have been reported previously,10–12,19–22 we identified four new mutations (G115A and T127A in rpoA and D271G and A466V in rpoC) in the hotspot area that potentially affect the interaction between the rpoA, rpoB and rpoC subunits of the RNA polymerase, and which have not been previously reported.10
The cluster rate in our study population was 13.8%, and was independently associated with being infected with M. tuberculosis strains from the East-Asian lineage. The East-Asian M. tuberculosis lineage has been associated with clustering in many molecular epidemiologic studies in East Asia;25 however, its impact in areas outside Asia is more controversial.26 In vitro and animal model studies have suggested that the East-Asian lineage is more pathogenic and virulent compared with other strains,27 and the production of phenolic glycolipid has been proposed as a possible mechanism.28 Despite the uncertain pathogenesis, our findings suggest a unique role of the East-Asian lineage in MDR-TB transmission. We also found that rpoB S531L was associated with genetic clustering, which supports the findings that this mutation has no fitness cost in in vitro studies,9 causing 40–73% of the M. tuberculosis RMP resistance,24,29,30 and is associated with compensatory mutations in rpoA and rpoC genes.11,22,31
Contrary to recent reports, we did not find any association between rpoA and rpoC mutations and clustering. de Vos and Li analyzed convenience samples from South Africa and China, respectively, and found that mutations in rpoC were significantly associated with clustering of RMP-resistant M. tuberculosis;11,12 however, they did not consider other factors known to be associated with genotypic clustering. We did not include the rpoA and rpoC genotype in the definition of clustering, as the implication for their phenotype and microevolution of M. tuberculosis has not been defined for most of the mutations.
Hispanic patients were more likely to be in a genotypic cluster. This result is similar to the cross-sectional study that evaluated the transmission of MDR-TB in the United States.13 Excessive alcohol use was found to be an independent risk factor associated with genotypic clustering in non-Asian patients. As it was only reported in one Asian patient, we could not evaluate the impact of alcohol consumption in this population. Based on the latest results from the 2010 National Survey on Drug Use and Health, excessive alcohol use is rare among Asian populations.32 Patients with chronic use of alcohol are known to be lymphopenic, with a reduced response to mitogen stimulation and impaired delayed-type hypersensitivity responses.33 Excessive use of alcohol, usually in conjunction with homelessness, injection drug use, and smoking, has been widely reported as an important risk factor for TB transmission in molecular epidemiologic studies, especially in low TB incidence areas.34
The present study had several limitations. First, our sample size was small and, as not all MDR-TB isolates from the study period were included, it is likely that we underestimated the number of genotypic clustered cases. Second, as both Indo-Oceanic and East-African-Indian lineages were underrepresented among the isolates studied, no conclusions can be drawn about their potential for clustering. Third, some factors that have been shown to be associated with transmission, such as the presence of live M. tuberculosis in cough droplets,2 delay in treatment initiation, socio-economic status, or place of exposure, were not measured in our study. Fourth, only partial regions of rpoA and rpoC were sequenced for mutation detection. However, the sequenced area includes the rpoA–rpoC interaction region of the rpoC gene, which has the potential to ameliorate the fitness cost of rpoB resistance mutations.10 Moreover, most of the mutations found outside of the rpoA–rpoC interaction region have not shown convergent evolution as has been observed for drug resistance mutations35 and for compensatory mutations,10 and the likelihood that they are compensatory mutations is therefore lower. Finally, we were not able to perform whole genome sequencing, which has been shown to better delineate the transmission links.
CONCLUSIONS
This community-based study presents a systematic analysis of clinical, epidemiologic and bacterial genetic factors associated with clustering of MDR-TB in two US states. We found that M. tuberculosis from the East-Asian lineage and isolates with the rpoB S531L mutation are bacterial factors independently associated with genetic clustering, suggesting that bacterial factors may have an impact on the ability of M. tuberculosis to cause secondary cases. In addition, we found that Hispanic patients were more likely to be part of a genetic cluster, as were non-Asian patients with excessive use of alcohol.
Acknowledgments
The authors would like to thank to P Oh, J Medinilla, AWood, and K Shiao from the California Department of Public Health, Tuberculosis Control Branch, Richmond, CA, USA, and the staff of the participating local health jurisdictions who facilitated the collection of data and whose high quality of service and cooperation have made this work possible.
This work was supported by the National Institutes of Health, Bethesda, MD, USA (NIH AI076476-01A2). PN is supported by the National Institutes of Health (NIAID R01AI104589) and the Centers for Disease Control and Prevention’s Tuberculosis Trials Consortium (TBTC), Atlanta, GA, USA. KS, JF, EAG and PKM were funded by the Centers for Disease Control and Prevention’s Tuberculosis Epidemiologic Studies Consortium (TBESC) Task Order #8 (200-2001-00080/0008). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Conflicts of interest: none declared.
References
- 1.Morrison J, Pai M, Hopewell PC. Tuberculosis and latent tuberculosis infection in close contacts of people with pulmonary tuberculosis in low-income and middle-income countries: a systematic review and meta-analysis. Lancet Infect Dis. 2008;8:359–368. doi: 10.1016/S1473-3099(08)70071-9. [DOI] [PubMed] [Google Scholar]
- 2.Jones-Lopez EC, Namugga O, Mumbowa F, et al. Cough aerosols of Mycobacterium tuberculosis predict new infection: a household contact study. Am J Respir Crit Care Med. 2013;187:1007–1015. doi: 10.1164/rccm.201208-1422OC. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Niemann S, Diel R, Khechinashvili G, Gegia M, Mdivani N, Tang YW. Mycobacterium tuberculosis Beijing lineage favors the spread of multidrug-resistant tuberculosis in the Republic of Georgia. J Clin Microbiol. 2010;48:3544–3550. doi: 10.1128/JCM.00715-10. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Grandjean L, Gilman RH, Martin L, et al. Transmission of multidrug-resistant and drug-susceptible tuberculosis within households: a prospective cohort study. PLOS MED. 2015;12:e1001843. doi: 10.1371/journal.pmed.1001843. discussion e1001843. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Burgos M, DeRiemer K, Small PM, Hopewell PC, Daley CL. Effect of drug resistance on the generation of secondary cases of tuberculosis. J Infect Dis. 2003;188:1878–1884. doi: 10.1086/379895. [DOI] [PubMed] [Google Scholar]
- 6.Teixeira L, Perkins MD, Johnson JL, et al. Infection and disease among household contacts of patients with multidrug-resistant tuberculosis. Int J Tuberc Lung Dis. 2001;5:321–328. [PubMed] [Google Scholar]
- 7.Gagneux S, Burgos MV, DeRiemer K, et al. Impact of bacterial genetics on the transmission of isoniazid-resistant Mycobacterium tuberculosis. PLOS Pathog. 2006;2:e61. doi: 10.1371/journal.ppat.0020061. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Fenner L, Egger M, Bodmer T, et al. Effect of mutation and genetic background on drug resistance in Mycobacterium tuberculosis. Antimicrob Agents Chemother. 2012;56:3047–3053. doi: 10.1128/AAC.06460-11. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Gagneux S, Long CD, Small PM, Van T, Schoolnik GK, Bohannan BJ. The competitive cost of antibiotic resistance in Mycobacterium tuberculosis. Science. 2006;312:1944–1946. doi: 10.1126/science.1124410. [DOI] [PubMed] [Google Scholar]
- 10.Comas I, Borrell S, Roetzer A, et al. Whole-genome sequencing of rifampicin-resistant Mycobacterium tuberculosis strains identifies compensatory mutations in RNA polymerase genes. Nat Genet. 2012;44:106–110. doi: 10.1038/ng.1038. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.de Vos M, Muller B, Borrell S, et al. Putative compensatory mutations in the rpoC gene of rifampin-resistant Mycobacterium tuberculosis are associated with ongoing transmission. Antimicrob Agents Chemother. 2013;57:827–832. doi: 10.1128/AAC.01541-12. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Li QJ, Jiao WW, Yin QQ, et al. Compensatory mutations of rifampicin resistance are associated with transmission of multidrug resistant Mycobacterium tuberculosis Beijing genotype strains in China. Antimicrob Agents Chemother. 2016;60:2807–2812. doi: 10.1128/AAC.02358-15. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Moonan PK, Teeter LD, Salcedo K, et al. Transmission of multidrug-resistant tuberculosis in the USA: a cross-sectional study. Lancet Infect Dis. 2013;13:777–784. doi: 10.1016/S1473-3099(13)70128-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Gagneux S, DeRiemer K, Van T, et al. Variable host-pathogen compatibility in Mycobacterium tuberculosis. Proc Natl Acad Sci USA. 2006;103:2869–2873. doi: 10.1073/pnas.0511240103. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Lin SY, Rodwell TC, Victor TC, et al. Pyrosequencing for rapid detection of extensively drug-resistant Mycobacterium tuberculosis in clinical isolates and clinical specimens. J Clin Microbiol. 2014;52:475–482. doi: 10.1128/JCM.01821-13. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Ghosh S, Moonan PK, Cowan L, Grant J, Kammerer S, Navin TR. Tuberculosis genotyping information management system: enhancing tuberculosis surveillance in the United States. Infect Genet Evol. 2012;12:782–788. doi: 10.1016/j.meegid.2011.10.013. [DOI] [PubMed] [Google Scholar]
- 17.Kato-Maeda M, Metcalfe JZ, Flores L. Genotyping of Mycobacterium tuberculosis: application in epidemiologic studies. Future Microbiol. 2011;6:203–216. doi: 10.2217/fmb.10.165. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Small PM, Hopewell PC, Singh SP, et al. The epidemiology of tuberculosis in San Francisco. A population-based study using conventional and molecular methods. N Engl J Med. 1994;330:1703–1709. doi: 10.1056/NEJM199406163302402. [DOI] [PubMed] [Google Scholar]
- 19.Zhang H, Li D, Zhao L, et al. Genome sequencing of 161 Mycobacterium tuberculosis isolates from China identifies genes and intergenic regions associated with drug resistance. Nat Genet. 2013;45:1255–1260. doi: 10.1038/ng.2735. [DOI] [PubMed] [Google Scholar]
- 20.Casali N, Nikolayevskyy V, Balabanova Y, et al. Microevolution of extensively drug-resistant tuberculosis in Russia. Genome Res. 2012;22:735–745. doi: 10.1101/gr.128678.111. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Song T, Park Y, Shamputa IC, et al. Fitness costs of rifampicin resistance in Mycobacterium tuberculosis are amplified under conditions of nutrient starvation and compensated by mutation in the β subunit of RNA polymerase. Mol Microbiol. 2014;91:1106–1119. doi: 10.1111/mmi.12520. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Eldholm V, Monteserin J, Rieux A, et al. Four decades of transmission of a multidrug-resistant Mycobacterium tuberculosis outbreak strain. Nat Commun. 2015;6:7119. doi: 10.1038/ncomms8119. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Shubladze N, Tadumadze N, Bablishvili N. Molecular patterns of multidrug resistance of in Georgia. Int J Mycobacteriol. 2013;2:73–78. doi: 10.1016/j.ijmyco.2013.02.002. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Poudel A, Nakajima C, Fukushima Y, et al. Molecular characterization of multidrug-resistant Mycobacterium tuberculosis isolated in Nepal. Antimicrob Agents Chemother. 2012;56:2831–2836. doi: 10.1128/AAC.06418-11. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Yang C, Luo T, Sun G, et al. Mycobacterium tuberculosis Beijing strains favor transmission but not drug resistance in China. Clin Infect Dis. 2012;55:1179–1187. doi: 10.1093/cid/cis670. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Marais BJ, Hesseling AC, Schaaf HS, Gie RP, van Helden PD, Warren RM. Mycobacterium tuberculosis transmission is not related to household genotype in a highly endemic setting. J Clin Microbiol. 2009;47:1338–1343. doi: 10.1128/JCM.02490-08. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Tsenova L, Ellison E, Harbacheuski R, et al. Virulence of selected Mycobacterium tuberculosis clinical isolates in the rabbit model of meningitis is dependent on phenolic glycolipid produced by the bacilli. J Infect Dis. 2005;192:98–106. doi: 10.1086/430614. [DOI] [PubMed] [Google Scholar]
- 28.Reed MB, Domenech P, Manca C, et al. A glycolipid of hypervirulent tuberculosis strains that inhibits the innate immune response. Nature. 2004;431:84–87. doi: 10.1038/nature02837. [DOI] [PubMed] [Google Scholar]
- 29.Ajbani K, Lin SY, Rodrigues C, et al. Evaluation of pyrosequencing for detecting extensively drug-resistant Mycobacterium tuberculosis among clinical isolates from four high-burden countries. Antimicrob Agents Chemother. 2015;59:414–420. doi: 10.1128/AAC.03614-14. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Rahmo A, Hamdar Z, Kasaa I, Dabboussi F, Hamze M. Genotypic detection of rifampicin-resistant M. tuberculosis strains in Syrian and Lebanese patients. J Infect Public Health. 2012;5:381–387. doi: 10.1016/j.jiph.2012.07.004. [DOI] [PubMed] [Google Scholar]
- 31.Brandis G, Hughes D. Genetic characterization of compensatory evolution in strains carrying rpoB Ser531Leu, the rifampicin resistance mutation most frequently found in clinical isolates. J Antimicrob Chemother. 2013;68:2493–2497. doi: 10.1093/jac/dkt224. [DOI] [PubMed] [Google Scholar]
- 32.Substance Abuse and Mental Health Services Administration. Results from the 2010 National Survey on Drug Use and Health: summary of national findings. Rockville, MD, USA: Substance Abuse and Mental Health Services Administration; 2011. HHS Publication No. (SMA) 11-4658. [Google Scholar]
- 33.Happel KI, Nelson S. Alcohol, immunosuppression, and the lung. Proc Am Thorac Soc. 2005;2:428–432. doi: 10.1513/pats.200507-065JS. [DOI] [PubMed] [Google Scholar]
- 34.Fok A, Numata Y, Schulzer M, FitzGerald MJ. Risk factors for clustering of tuberculosis cases: a systematic review of population-based molecular epidemiology studies. Int J Tuberc Lung Dis. 2008;12:480–492. [PubMed] [Google Scholar]
- 35.Hazbon MH, Motiwala AS, Cavatore M, Brimacombe M, Whittam TS, Alland D. Convergent evolutionary analysis identifies significant mutations in drug resistance targets of Mycobacterium tuberculosis. Antimicrob Agents Chemother. 2008;52:3369–3376. doi: 10.1128/AAC.00309-08. [DOI] [PMC free article] [PubMed] [Google Scholar]