Abstract
Background:
Extensively drug-resistant tuberculosis (XDR-TB)/HIV co-infection has been associated with high mortality and poor TB outcomes. We performed a prospective study to comprehensively characterize a cohort of XDR-TB patients.
Methods:
Adult XDR-TB patients were enrolled at treatment initiation at a TB referral hospital in KwaZulu-Natal Province, South Africa and followed through the end of treatment. Clinical data, questionnaires, adherence data, and sputum were collected monthly. Whole genome sequencing (WGS) was performed on baseline M. tuberculosis (MTB) isolates. Treatment outcomes were defined using standard definitions.
Results:
105 XDR-TB patients (76.1%HIV infected) were enrolled from August 2009 through July 2011. Among HIV co-infected patients, 82.5% were on ART initially and 93.8% cumulatively over the study period. At 24 months 31.4% had a successful outcome and 68.6% had an unsuccessful outcome with 41% mortality. ART was associated with improved mortality in HIV co-infected patients (p= 0.05), as was TB culture conversion (p <0.0001). On WGS the majority of strains were LAM4/KZN lineage (68%), with few SNP differences.
Conclusion:
Despite improved HIV care, treatment outcomes and mortality were only modestly improved compared to previous South African XDR-TB/HIV treatment cohorts. Of note this study was completed prior to introduction of new antimycobacterial agents (e.g. bedaquiline, delaminid). As new TB drugs and regimens become available it is important to monitor treatment to ensure benefits seen in clinical trials are reproduced in high-burden, low-resource settings.
Keywords: tuberculosis, drug-resistance, HIV/AIDS, South Africa, treatment outcomes, epidemiology
INTRODUCTION
The most drug-resistant form of tuberculosis, extensively drug resistant tuberculosis (XDR-TB), is defined as resistance to isoniazid and rifampicin, any fluoroquinolone, and any second line injectable antimycobacterial agent [1]. XDR-TB/HIV co-infection was first described in an explosive outbreak in rural KwaZulu-Natal, South Africa in 2006 [2,3]. The initial outbreak was associated with high mortality in the absence of effective treatment and was initially believed to have occurred through primary nosocomial transmission [2,3]. Epidemiologic and microbiologic studies have identified F15/LAM4/KZN as the predominant Mycobacterium tuberculosis (MTB) strain in XDR-TB patients in KwaZulu-Natal. Distinct from other regions of high XDR-TB prevalence, next generation sequencing has shown surprisingly few SNP differences between geographically dispersed MTB isolates with the inference that ongoing transmission is responsible for a large proportion XDR-TB cases in this HIV high burden population [2,4–8].
Previous retrospective studies have shown high mortality and very low rates of cure in the treatment of XDR-TB/HIV [9–13]. Factors associated with favorable XDR-TB treatment outcomes include treatment with regimens to which the MTB strain is susceptible, access to novel and re-purposed agents (including bedaquiline, delaminid, and linezolid), and in patients with HIV co-infection treatment with antiretroviral therapy (ART) [14]. ART has been shown to improve mortality in TB/HIV when given early in the TB treatment course, particularly in patients with low CD4 T-cell count [15,16]. ART is also associated with decreased rates of reactivation in patients with latent TB infection. However, in XDR-TB/HIV, while ART is clearly associated with improved mortality it is less clear that ART improves MTB culture conversion or other TB-specific outcomes [17].
To comprehensively characterize a cohort of XDR-TB patients living both with and without HIV, to avoid bias introduced by retrospective design, and to capture greater detail about treatment regimens, we analyzed data from a prospectively followed cohort of XDR-TB patients in KwaZulu-Natal Province, South Africa. We performed this study to characterize treatment outcomes and the effects of ART on TB treatment response and to describe the results of whole genome sequencing (WGS).
METHODS
Study design
A cohort of consecutive XDR-TB patients was enrolled at a tuberculosis referral hospital in KwaZulu-Natal, South Africa from August 2009 to July 2011 and prospectively followed for greater than 24 months unless death or patient default occurred. Since 2008 KwaZulu-Natal has decentralized the care for drug-resistant TB by initially opening 4 decentralized sites throughout the province, and more recently, 9 decentralized sites for the treatment of drug-resistant TB. Eligible patients were ≥ 18 years of age, had not been previously treated for XDR-TB and/or molecular evidence, were diagnosed with active TB disease according to South African/WHO guidelines, had culture-proven XDR-TB, agreed to start treatment for XDR-TB, and had capacity to give informed consent. Written consent was obtained by the study. Clinical data, questionnaires, and sputum were collected monthly and treatment outcomes assigned according to standard definitions at 24 months after treatment initiation [18].
Culture conversion was defined as two consecutive monthly TB cultures without growth of MTB. Net culture conversion was defined as culture conversion without subsequent reversion to MTB culture positivity. Treatment outcomes included cure, treatment completion, lost to follow-up, death, and failure, and were defined at 24 months after XDR-TB treatment initiation. Cure was defined as a patient who completed 24 months of treatment with 3 or more consecutive negative MTB cultures taken at least 30 days apart in the last 12 months of treatment [19]. Treatment completion was defined as a patient who completed 24 months of treatment with clinical improvement, but without sufficient negative MTB cultures to qualify for cure. Lost to follow up (formerly ‘default’) was defined as any XDR-TB patient whose XDR-TB treatment was interrupted for two or more consecutive months. Treatment failure was defined as two or more of five cultures recorded as positive in the final 12 months of treatment, or if any of the final three cultures were positive. We further used a set of modified MDR-TB treatment outcome criteria suggested by Lange et al. to define treatment outcomes at 36 months [20]. Under the modified criteria, cure is defined as a negative culture status at six months, with no subsequent positive culture or replaces within a year, and failure is defined as a positive culture status at six months or relapse after one year of treatment. Death was defined as a death during the period of observation and loss to follow-up was defined as not being in care at six months. A status of undeclared was defined as an outcome that was not assessed. This study protocol was approved through the Columbia University Review Board as well as the Biomedical Research Ethics Council of the University of KwaZulu-Natal.
Laboratory Methods and Whole Genome Sequencing
Sputum for microscopy and MTB culture was obtained monthly as is routine clinical practice. MTB culture positivity was determined using the BACTEC MGIT 960 fluorometric system (Becton Dickinson Diagnostics, Sparks, MD, USA). Drug susceptibility testing for first-line and second-line drugs was performed at a regional TB reference laboratory in Durban, South Africa. Drug susceptibility testing for isoniazid, rifampin, ethambutol, streptomycin, ethionamide, ofloxacin, and kanamycin was by the modified proportional growth method using standard techniques. Molecular susceptibility testing was not routinely performed during the study period.
Mycobacteria were extracted from solid media agar growth media and genomic MTB DNA isolated using standard protocols [21]. Library preparation and WGS were performed as previously described using the Illumina HiSeq platform [22]. Raw sequencing reads were trimmed for low-quality bases and aligned to the H37Rv MTB reference genome using the Burrows-Wheeler Aligner. Variants calls were obtained using samtools. Variants in the highly-variable PE/PPE gene family, variants in repeat regions or mobile elements, and variants within 15 base pairs of an insertion or deletion were excluded from analysis. High quality SNPs were defined as those with PHRED-scaled quality scores > 100 and > 75% of reads supporting the alternate allele. All data were submitted to the NCBI Sequence Read Archive. To construct the maximum likelihood phylogenetic tree reconstruction in RAxML with ascertainment bias correction, the general time-reversal-gamma distribution model of nucleotide substitution and ascertainment bias was used. 1000 bootstrap pseudoreplicates evaluated node robustness.
Statistical Analysis
Characteristics of XDR-TB patients at treatment initiation were stratified by HIV status and by TB lineage. Frequencies and percentages of TB drugs used were described. Comparisons used the chi-square test or Fischer’s exact test for categorical variables. Continuous variables were summarized with the median and interquartile range (IQR). Unfavorable treatment outcomes included death, treatment, failure, or loss to follow-up and favorable treatment outcomes included cure and treatment completion. CD4 count/ mm3 was defined as the first CD4 count recorded for each patient.
Survival analysis for mortality was calculated from time of XDR-TB treatment initiation until time of death or censoring. Duration was calculated in days and then converted into months by dividing by 30.4. Kaplan-Meier curves were stratified by risk factors for mortality. Tests between strata were assessed using the log-rank test.
Hazard ratios and 95% confidence intervals were estimated from Cox-proportional hazards models for mortality and favorable treatment outcomes. Person-time for favorable outcomes analysis was calculated from time of XDR-TB treatment initiation to the favorable outcome, censored events, or end of follow-up at 24 months. Bivariate analysis was used to assess the relationship between exposure variables and time-to-event variables. Multivariable models were adjusted for variables that were of significance according to the literature and proportional hazards assumptions were tested.
Subdistributional hazard models, with death as a competing risk, and Cox proportional hazard models were generated to identify patient characteristics associated with net MTB sputum culture conversion. Statistical analysis was performed by using SAS version 9.4 software (SAS Institute, Cary, NC, USA) and STATA version 12.0 (StataCorp LP, College Station, Texas).
RESULTS
During the study period, 110 patients with XDR-TB aged 18 years or older presented at a public TB specialist hospital in KwaZulu Natal Province, South Africa. Of these, five did not meet study eligibility criteria (one was incarcerated and ineligible by protocol, three were determined to not be XDR-TB, and one was lost to initial treatment) and 105 were enrolled and provided data for analysis. The majority of patients were female (53.3%) and younger than 36 (52.4%) years; 41 (58.1%) patients had a previous history of MDR-TB treatment and 97 (92.4%) had a previous history of drug-sensitive TB. Median CD4 cell count/ mm3 at XDR-TB treatment initiation was 192 (IQR 117–390), 66/ 80 (82.5%) patients with HIV infection were receiving ART at the time of XDR-TB treatment initiation (after treatment initiation an additional 9 patients started ART (93.8%)). Over the course of the study, median time to sputum culture conversion was 65 days (IQR 34–99) [Table 1]. Median follow up time was 18.1 months (IQR: 6.5–24.1).
Table 1. Characteristics of patients enrolled with XDR-TB in KwaZulu-Natal Province, South Africa (n = 105).
| Variable | Patients with HIV n = 80 (%) |
Patients without HIV n = 25 (%) |
All patients n = 105 (%) |
|---|---|---|---|
| Gender | |||
| Male | 34 (42.5) | 15 (60.0) | 49 (46.7) |
| Female | 46 (57.5) | 10 (40.0) | 56 (53.3) |
| Age (y) | |||
| <36 | 38 (47.5) | 17 (68.0) | 55 (52.4) |
| ≥ 36 | 42 (52.5) | 8 (32.0) | 50 (47.6) |
| Median (IQR) | 36 (30–42) | 25 (20–44) | 35 (27–43) |
| Any TB history | |||
| Yes | 72 (90.0) | 25 (100) | 97 (92.4) |
| No | 8 (10.0) | 0 (0) | 8 (7.6) |
| History of MDR-TB treatment | |||
| Yes | 44 (55.0) | 17 (68.0) | 61 (58.1) |
| No | 36 (45.0) | 8 (32.0) | 44 (41.9) |
| CD4 cell count / mm3 at XDR-TB treatment initiation | |||
| ≤ 200 | 40 (52.0) | -- | -- |
| > 200 | 36 (48.0) | -- | -- |
| Median (IQR) | 192 (117–390) | -- | -- |
| ART | |||
| At XDR-TB treatment initiation | 66 (82.5) | -- | -- |
| After XDR-TB treatment initiation | 9 (11.3) | -- | -- |
| Never | 5 (6.2) | ||
| Culture Conversionb | |||
| Yes | 28 (35.0) | 13 (52.0) | 41 (39.1) |
| No | 52 (65.0) | 12 (48.0) | 64 (61.0) |
| Weight at XDR-TB treatment initiation (kg) | 50.4 (46–57) | 45.7 (42–52) | 49.9 (43.6–56.9) |
| Median (IQR) | |||
| Time to net culture conversion (days) | |||
| Median (IQR) | 67 (33–91) | 70 (39–124) | 67 (34–99) |
| Education | |||
| Completed secondary school | 50 (62.5) | 18 (72.0) | 68 (64.8) |
| Did not complete secondary school | 27 (33.8) | 5 (20.0) | 32 (30.5) |
| Unknown | 3 (3.8) | 2 (8.0) | 5 (4.8) |
| Cotrimoxazolec | |||
| Yes | 24 (57.1) | -- | -- |
| No | 18 (42.9) | -- | -- |
One patient’s HIV status was reported as having changed during the 24-month period.
Net culture conversion: 4 patients reverted during the course of the study.
Out of HIV patients with CD4 counts <250 at baseline (n=42).
Treatment Outcomes
Treatment outcomes were determined at 24 months. 33 (31.4%) patients had favorable treatment outcomes (25.7% were cured and 5.7% completed treatment); 72 patients (68.6%) had unfavorable outcomes: 43 patients died (41%), 17 patients were lost to follow up (16.2%), and 12 patients failed to complete treatment (11.4%) [Table 2]. Of the 43 reported deaths, TB was the likely proximate cause for 21 (48.8%) cases (including 3 cases of massive hemoptysis), and TB was excluded as the proximate cause of five deaths (including 3 cases of acute renal failure). In 17 cases, the cause of death was unknown, occurring outside of the hospital setting. One patient seroconverted to HIV infected during 24 months of follow-up. Among those that qualified for Co-trimoxazole treatment (HIV infected and with CD4 count <250 / mm3), 57.1% (24/42) were ever treated with Co-trimoxazole. Among the 5 HIV co-infected patients who never started ART, the median CD4 count at XDR-TB treatment initiation was 415.5 (IQR: 271.5–506.5). The 5 who did not start ART were enrolled in the study over a period of July 2009 to January 2011. All had a previous history of TB. They did not appear to differ from those who started ART by demographic or other factors. Among those who did start ART, the median CD4 count was 187.0 (IQR: 113.5–347.0).
Table 2. Treatment outcomes for XDR-TB patients enrolled in KwaZulu- Natal Province, South Africa (n = 105).
| Treatment Outcomes 24 months (WHO Definitions) | Patients with HIV at XDR-TB treatment initiation n = 80 (%) | Patients without HIV at XDR-TB treatment initiation n =25 (%) | Total n = 105 (%) |
p-value |
|---|---|---|---|---|
| Favorable | 32 (27.5) | 11 (44.0) | 33 (31.4) | 0.08 |
| Cure | 16 (20.0) | 11 (44.0) | 27 (25.7) | |
| Completed | 6 (7.5) | 0 (0) | 6 (5.7) | |
| Unfavorable | 58 (72.5) | 14 (56.0) | 72 (68.6) | |
| Lost to follow up | 12 (15.0) | 5 (20.0) | 17 (16.2) | |
| Failure | 11 (13.8) | 1 (4.0) | 12 (11.4) | |
| Died | 35 (43.8) | 8 (32.0) | 43 (41.0) | |
| Treatment Outcomes 36 months (Lange Definitions) | ||||
| Favorable | 12 (15.0) | 5 (20.0) | 17 (16.2) | 0.52 |
| Cure | 12 (15.0) | 5 (20.0) | 17 (16.2) | |
| Undeclared | 4 (5.0) | 3 (12.0) | 7 (6.7) | |
| Unfavorable | 64 (80.0) | 17 (68.0) | 81 (76.2) | |
| Lost to follow up | 12 (15.0) | 5 (20.0) | 17 (16.2) | |
| Failure | 16 (20.0) | 3 (12.0) | 19 (18.1) | |
| Died | 36 (45.0) | 9 (36.0) | 45 (42.9) |
We additionally followed patients out to 36 months (12 months post-treatment completion) using the criteria suggested by Lange et al. to assign outcomes. Using these criteria, 6 additional patients were re-classified as failure since they did not convert within 6 months or reverted to culture positive after 24 months, 1 additional patient died, and 7 patients were considered undeclared since we did not have sufficient post-24 month follow-up. The median number of XDR-TB drugs in the treatment regimen was 9 (IQR 7–11), and the most commonly used drugs among all patients (n=105) were Pyrazinamide (96.2%), Ethambutol (91.4%), Moxifloxacin (91.4%), Para-aminosalicylic Acid (90.5%), and Terizidone (90.5%) and Capreomycin (88.6%) [see Table 1, Supplemental Digital Content].
Whole Genome Sequencing
The majority of the XDR-TB isolates that were sequenced (n=50) were of the LAM4/KZN lineage (68%), followed by Beijing (10%), PGG1,LIN-2 (10%), S,PGG2,LIN-4 (10%), and ‘Other’ isolates (12%), which include CAS1-Kili, PGG1, LIN-3; T3,PGG3,LIN-4; and X3,PGG2,LIN-4 [see Table 2, Supplemental Digital Content]. Among those with whole genome sequencing available, 74% had a previous history of MDR-TB compared to 44% of those who did not have whole genome sequencing available (p < 0.01). In whole genome sequencing, mean pairwise SNP difference between LAM4 isolates was 50.37. When we excluded two outlier genomes the mean pairwise SNP difference between the LAM4 isolates was 17.41 and when we used the original Tugela Ferry MTB isolate as a reference the mean pairwise SNP difference was 10.71.
Statistical Analysis
In survival analysis, XDR-TB/HIV co-infected patients receiving ART had an improved mortality compared to XDR-TB/HIV patients who never received ART (HR: 0.28, 95% CI: 0.08–0.98, p value = 0.05) [Figure 1]. Mean time to event was 17.34 months and for patients not receiving ART median time to event was 6.61 [Figure 1B]. Among those who did not culture convert, median time to event was 13.82 months [Figure 1C]. In univariate analysis, patients who had MTB sputum culture conversion had a 95% lower hazard for mortality (HR: 0.05, 95% CI 0.02–0.17, p value < 0.0001) than patients who did not [see Table 3, Supplemental Digital Content]. In an analysis of risk factors associated with favorable treatment outcomes (defined as cure and completed treatment), no variables were significantly associated with the outcome [Table 4, Supplemental Digital Content].
Figure 1.
Kaplan-Meier survival estimates (A) Probability of survival stratified by HIV status at XDR-TB treatment initiation (B) Probability of survival stratified by ART status (C) Probability of survival stratified by culture conversion
In a competing risk analysis, with death as the competing risk, there were no predictor variables that were significantly associated with MTB sputum culture conversion [S5–S6]. In multivariable analysis of competing risks, XDR-TB/ HIV patients who received ART at XDR-TB treatment initiation were more likely to experience net MTB sputum culture conversion than patients not receiving ART [adjusted subdistribution hazards ratio (aSHR): 1.45, 95% CI: 0.44–4.75, p = 0.37]; however, this was not statistically significant see (Table 8, Supplemental Digital Content).
DISCUSSION
The findings from our study are important as they provide strong evidence that despite improved ART coverage, and reasonable adherence to ART, XDR-TB/ HIV mortality remains unacceptably high and therefore improved HIV care and ART treatment alone may not be sufficient to improve XDR-TB/HIV outcomes. In this study, we found a high percentage of XDR-TB/ HIV patients receiving ART at baseline, with high rates of HIV testing prior to XDR-TB treatment. Despite high ART coverage, mortality rates and treatment outcomes remained poor and loss to follow-up (previously default) during treatment remained high. Failure of XDR-TB to respond to treatment, as measured by culture conversion, was the single greatest risk factor for mortality. Although there was improved TB culture conversion with ART from XDR-TB treatment initiation in a competing risk analysis, the absolute difference was modest and not statistically significant on multivariable analysis through 24 months after treatment initiation.
Whole genome sequencing revealed that MTB of a single lineage and with relatively few SNP differences comprised the majority of cases of XDR-TB in our study. The isolates were taken from patients in different parts of the province over the span of three years and they represent MTB isolates from only the fraction of XDR-TB patients who successfully present to care. Given that these isolates are linked genetically, but not linked in terms of conventional epidemiology, this reinforces the idea that there is a large, uncontrolled XDR-TB epidemic in KwaZulu-Natal, South Africa.
Low rates of successful treatment outcomes among XDR-TB/ HIV patients have been previously described. In one study previously reported by this group, only 61% of XDR-TB/HIV co-infected patients were receiving ART, compared to 82.5% within this cohort, and the rate of mortality was 42%, compared to 41% in this cohort [9]. Dheda et al. (2010) reported that 62% of XDR-TB patients in their study who were living with HIV were on highly active antiretroviral therapy, and the mortality was 36% among the entire cohort, and 55% among those living with HIV [11]. A more recent paper by Olayanju et al. (2018) compared outcomes between two cohorts of XDR-TB patients, of which 49.3% were HIV infected. In the non-bedaquiline cohort, 85.8% of the cohort had an unfavorable outcome and 33.8% died [23], compared to the bedaquiline cohort, of which 33.8% had an unfavorable outcome and 14.7% were deceased. Adherence for this cohort has been previously described, and six-month self-reported adherence was found to be 88.2% for ARVs and 67.7% for TB [24]. The findings from our study are important as they provide strong evidence that despite improved ART coverage, and reasonable adherence to ART, XDR-TB/ HIV mortality remains unacceptably high.
Irrespective of HIV status, outcomes of XDR-TB are poor. A systematic review of treatment outcomes conducted by Jacobson et al (2010) found that among XDR-TB patients who were not HIV infected, 43.7% experienced favorable treatment outcomes (compared to 31.4% in the PROX study) and approximately 20.8% died, although this figure varied according to the setting, with mortality as high as 98% among patients identified in the Tugela Ferry outbreak [25,2]. In this study, we found that HIV status was not associated with mortality.
One of the primary limitations to this study was a small sample size, which subsequently may have led to underpowered statistical analysis for the modeling. Whole genome sequencing was available for only 50 patients due to sample viability. Additionally, as this study was conducted at a tertiary referral hospital, many patients were not referred, or treated and did not start therapy, resulting in an initial selection bias. It is likely that patients who were successfully referred for treatment initiation differed in important ways from those who were not, so in addition to selection bias, there may have been a survivor bias inherent within the selection process. However, during the study period, this was the only center in the province to start XDR-TB treatment and so the study population was representative of those seeking care in that regard. Additionally, although this was a prospective study, HIV viral load data were incomplete and there were limited data on comorbid conditions and opportunistic infections.
The findings from our study are important as they provide strong evidence that despite improved HIV care and substantially higher rates of ART, and co-trimoxazole prophylaxis, mortality remains high and improved HIV care and ART treatment alone may not be sufficient to improve XDR-TB/ HIV outcomes. Fortunately, new medications for drug-resistant TB are being actively rolled out and new regimens are being studied. Although our study was conducted prior to the introduction of these new drugs, the prospective design allows for our study to act as a baseline to which future studies can be compared. As new drugs such as bedaquline and new regimens such as the NiX regimen (bedaquiline, pretomanid, and linezolid given orally for six months) become available it will be important to monitor treatment carefully to ensure that the benefits seen in clinical trials are translated to and reproduced in high burden, low-resource settings [26].
Supplementary Material
Figure 2.
Dendrogram of the subset of XDR-TB isolates sequenced using whole genome sequencing. The majority of strains (35/50 68%) were KZN LAM4 with an average of 10.71 SNP differences between these strains (n = 50).
Acknowledgements:
We would like to thank the patients and clinicians at King Dinizulu Medical Complex, Sydenham, KwaZulu-Natal, South Africa.
Conflicts of Interest and Sources of Funding:
The authors have no conflicts of interest to declare. This work was supported by the National Institutes of Health/ National Institute of Allergy and Infectious Disease K23 [AI098479–01A1]; Gerstner Family Foundation, Doris Duke Clinical Scientist Career Development Award to M.O; and CAPRISA MRC TB-HIV Pathogenesis Unit.
REFERENCES
- 1.World Health Organization. Tuberculosis (TB) Fact Sheet. Drug-resistant TB: Totally drug-resistant TB FAQ. 2017. http://www.who.int/tb/areas-of-work/drug-resistant-tb/totally-drug-resistant-tb-faq/en/ Accessed March 7, 2017.
- 2.Gandhi NR, Moll A, Sturm AW, et al. Extensively drug-resistant tuberculosis as a cause of death in patients co-infected with tuberculosis and HIV in a rural area of South Africa. Lancet. 2006;368:1575–80. [DOI] [PubMed] [Google Scholar]
- 3.Andrews JR, Shah NS, Gandhi N, et al. Multidrug-resistant and extensively drug-resistant tuberculosis: implications for the HIV epidemic and antiretroviral therapy rollout in South Africa. J Infect Dis. 2007. 1;196 Suppl 3: S482–90. [DOI] [PubMed] [Google Scholar]
- 4.Gandhi NR, Brust JC, Moodley P, et al. Minimal diversity of drug-resistant Mycobacterium tuberculosis strains, South Africa. Emerg Infect Dis. 2014. March;20(3):426–33. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Johnson R, Warren RM, van der Spuy GD, et al. Drug-resistant tuberculosis epidemic in the Western cape driven by a virulent Beijing genotype strain. Int J Tuberc Lung Dis. 2010. January;14(1):119–21 [PubMed] [Google Scholar]
- 6.Gandhi NR, Weissman D, Moodley P, et al. Nosocomial transmission of extensively drug-resistant tuberculosis in a rural hospital in South Africa. J Infect Dis. 2013. January 1;207(1)9–17. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Andrews JR, Gandhi NR, Moodley P, et al. Exogenous reinfection as a cause of multidrug-resistant and extensively drug-resistant tuberculosis in rural South Africa. J Infect Dis. 2008. 198(11): 1582–9. [DOI] [PubMed] [Google Scholar]
- 8.Shah NS, Auld SC, Brust JC et al. Transmission of extensively drug-resistant tuberculosis in South Africa. N Engl J Med. 2017. 376(3):243–253. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.O’Donnell MR, Padayatchi N, Kvasnovsky C, et al. Treatment outcomes for extensively drug-resistant tuberculosis and HIV Co-infection. Emerg Inf Dis. 2013;19(3) [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Pietersen E, Ignatius E, Streicher E, et al. High frequency of resistance, lack of clinical benefit, and poor outcomes in capreomycin treated South African patients with extensively drug-resistant tuberculosis. PLoS One. 2015:10(4); e0123655. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Dheda K, Shean K, Zumla A, et al. Early treatment outcomes and HIV status of patients with extensively drug-resistant tuberculosis in south Africa: a retrospective cohort study. Lancet. 2010; 375(9728):1798–807. doi: 10.1016/S0140-6736(10)60492-8. [DOI] [PubMed] [Google Scholar]
- 12.O’Donnell MR, Pillay M, Pillay M, et al. Primary capreomycin resistance is common and associated with early mortality in patients with extensively drug-resistant tuberculosis in KwaZulu-Natal, South Africa. J Acquir Immune Defic Syndr. 2015;69(5):536–43. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Pietersen E, Ignatius E, Streicher EM, et al. Long-term outcomes of patients with extensively drug-resistant tuberculosis in South Africa: a cohort study. Lancet. 2014;383(9924):1230–9. [DOI] [PubMed] [Google Scholar]
- 14.Curry International Tuberculosis Center and California Department of Public Health, 2016, Drug-Resistant Tuberculosis, A Survival Guide for Clinicians. 3rd edition. [Google Scholar]
- 15.Karim SSA, Naidoo K, Grobler A et al. Timing of initiation of antiretroviral drugs during tuberculosis therapy. N Engl J Med 2011;365:1492–1501. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Havlir DV, Kendall MA, Ive P et al. Timing of antiretroviral therapy for HIV-1 infection and tuberculosis. N Engl J Med 2011;365:1482–1491. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Abdool Karim SS, Naidoo K, Grobler A, et al. Integration of antiretroviral therapy with tuberculosis treatment. N Engl J Med. 2011;365(16):1492–501. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Laserson KF, Thorpe LE, Leimane V, et al. Speaking the same language: treatment outcome definitions for multidrug resistant tuberculosis. Int J Tuberc Lung Dis. 2005;9:640–5. [PubMed] [Google Scholar]
- 19.World Health Organization. Definitions and reporting framework for tuberculosis- 2013 revision. Updated December 2014. http://apps.who.int/iris/bitstream/10665/79199/1/9789241505345_eng.pdf Accessed April 17, 2016.
- 20.Gunther G, Lange C, Alexandru S, Altet N, Avsar K, Bang D, et al. Treatment Outcomes in Multidrug-Resistant Tuberculosis. N Engl J Med. 2016;375(11):1103–1105. [DOI] [PubMed] [Google Scholar]
- 21.Larsen MH, Biermann K, Tandberg S, Hsu T, Jacobs WR. Genetic Manipulation of Mycobacterium tuberculosis. Curr Protoc Microbiol. 2007;Chapter 10: Unit 10A.2 [DOI] [PubMed] [Google Scholar]
- 22.Metzker ML. Sequencing technologies- the next generation. Nat Rev Genet. 2010; 11(1):31–46. Epub 2009 December 8. [DOI] [PubMed] [Google Scholar]
- 23.Olayanju O, Limberis J, Esmail A, et al. Long term bedaquiline-related treatment outcomes in patients with extensively drug resistant tuberculosis from South Africa. Eur Respir J. 2018. [Epub ahead of print]. [DOI] [PubMed] [Google Scholar]
- 24.O’Donnell MR, Wolf A, Werner L, Horsburgh CR, Padayatchi N. Adherence in the treatment of patients with extensively drug-resistant tuberculosis and HIV in South Africa: A prospective cohort study. J Acquir Immune Defic Syndr. 2014. Spet 1;67(1):22–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Jacobson KR, Tierney DB, Jeon CY, et al. Treatment outcomes among patients with extensively drug-resistant tuberculosis: systematic review and meta-analysis. Clin Infect Dis. 2010. July 1;51(1):6–14. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Conradie F, Diacon AH, Everitt D, et al. The NIX-TB trial of pretomanid, bedaquiline and linezolid to treat XDR-TB. [80LB] Presented at: Conference on Retroviruses and Opportunistic Infections (CROI); 2017; Boston. [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.