Abstract
Background
Understanding relationships between human immunodeficiency virus (HIV) and multidrug-resistant tuberculosis (MDR-TB) is crucial for ensuring successful MDR-TB outcomes.
Methods
We used a cross-sectional analysis to evaluate sociodemographic and clinical characteristics as correlates of antiretroviral therapy (ART) use, having an HIV viral load result, and HIV viral suppression in a cross-sectional sample of people with HIV (PWH) and MDR-TB enrolled in a cluster-randomized trial of nurse case management to improve MDR-TB outcomes.
Results
Among 1479 PWH, mean age was 37.1 years, 809 (54.7%) were male, and 881 (59.6%) were taking ART. Housing location, employment status, and CD4 count differed significantly between those taking versus not taking ART. Among the 881 taking ART, 681 (77.3%) had available HIV viral load results. Housing location, CD4 count, and prior history of TB differed significantly between those with and without a viral load result. Among the 681 with a viral load result, 418 (61.4%) were virally suppressed. Age, education level, CD4 count, TB history, housing location, and ART type differed significantly between those with and without viral suppression.
Conclusion
PWH presenting for MDR-TB treatment with a history of TB, taking a protease inhibitor, or living in a township may risk poor MDR-TB outcomes.
Keywords: multidrug-resistant tuberculosis, human immunodeficiency virus, South Africa
Multidrug-resistant tuberculosis (MDR-TB), a severe form of TB resistant to the first-line drugs isoniazid and rifampin, is a growing problem, with some countries and regions particularly affected1. In South Africa, a high-burden TB country, there were an estimated 21,000 incident cases of MDR- or rifampicin-resistant TB in 2021, of which 66% were reported to the World Health Organization as successfully treated according to World Health Organization treatment outcome definitions2,3. HIV co-infection complicates MDR-TB treatment due to overlapping treatment side effects4,5, drug-drug interactions6,7, and increased pill burden8. Despite these challenges, many people with HIV (PWH) are successfully treated for MDR-TB. Taking antiretroviral therapy (ART) to treat HIV and having better immune function indicated by a higher CD4 count have been shown to increase odds of MDR-TB treatment success among PWH, and HIV viral load may also play a role9-13.
The World Health Organization (WHO) has recommended integrated care for people co-infected with HIV and MDR-TB since 201414. Given the importance of controlling HIV during MDR-TB treatment, a thorough understanding of a patient’s HIV status, treatment, and virologic control is essential. HIV viral load is a widely accepted marker for HIV disease and ART status, and the WHO has endorsed at least yearly HIV viral load monitoring for all PWH since 201315. South African HIV and MDR-TB treatment guidelines recommend HIV viral load monitoring at the time of MDR-TB diagnosis and every six months during treatment16,17. Understanding frequency of viral load measurement and viral suppression rates among PWH initiating MDR-TB treatment may clarify the importance of viral load measurement and its relationship to MDR-TB outcomes.
This analysis will investigate associations between demographic and clinical characteristics with ART status, HIV viral load testing, and HIV viral suppression in PWH at the time of presentation for MDR-TB treatment initiation. Our goal is to identify pre-treatment factors that may prompt clinicians to intensify HIV follow-up and adherence interventions to optimize HIV viral suppression and ultimately MDR-TB treatment outcomes.
METHODS
Parent Study
This secondary data analysis used data from a cluster-randomized trial comparing intensive nurse case management to improve MDR-TB outcomes in South Africa. A total of 2839 participants 13 years and older receiving care for MDR-TB at one of the participating South African TB hospitals were enrolled in the parent study, whose methods have been thoroughly described elsewhere18. The parent study did not attempt to balance recruitment based on gender or age group, but rather attempted to recruit any eligible participants presenting for MDR-TB initiation. Data collection took place between 2014 through 2021 at 13 hospital sites in KwaZulu-Natal and Eastern Cape provinces of South Africa. At intervention sites, a nurse case manager assisted treating clinicians with patient management including adherence support, symptom evaluation and management, and strategies to reduce loss to follow-up. At control sites, research assistants extracted study data from medical records with no intervention.
This cross-sectional secondary analysis conducted at baseline of the parent study included PWH enrolled in the parent study who had one of the WHO-defined MDR-TB outcomes: treatment success, treatment failure, death, or loss to follow-up3. Parent study participants were excluded if they were found to have additional resistance beyond MDR-TB or drug-sensitive TB or if they transferred out of the study prior to MDR-TB treatment outcome. Data for this analysis was abstracted from REDCap, a web-based secure data storage platform19. A diagram of the study sample is presented in Figure 1.
Figure 1. Study Sample.
HIV, human immunodeficiency virus; TB, tuberculosis; ART, antiretroviral therapy; VL, viral load; transfer out, parent study participant was transferred to a different MDR-TB treatment site prior to MDR-TB outcome which did not partcipate in the parent study.
*At the time of this analysis, data was available for 2545 of 2890 participants in the parent study.
Statistical Analyses
We conducted a cross-sectional analysis of baseline data from participants enrolled in the parent study. Descriptive statistics were examined and compared among PWH with or without our three identified outcomes. We then built three multiple logistic regression models to determine factors associated with each outcome. In each analysis, multiple models were considered using backward selection and the model with the lowest AIC was chosen.
Variable Definitions
We had three outcome variables of interest. First, we modeled the odds of taking ART at the time of MDR-TB treatment initiation, defined as either documented evidence of taking ART before the date of MDR-TB treatment initiation or participant self-report of taking ART prior to MDR-TB treatment initiation with documented evidence of ART use on the date of MDR-TB initiation, versus not taking ART. Next, we modeled the odds of having an HIV viral load result at the time of MDR-TB treatment initiation. As annual viral load assessment is the standard of care and electronic gate keeping within the NHLS prevents clinicians from ordering a new viral load within a 12-month period, HIV viral load results taken within 12 months prior to or four weeks after the date of MDR-TB treatment outcome were included in this analysis. The primary outcome in our third model was HIV viral suppression at MDR-TB treatment initiation, defined as an HIV viral load less than 400 copies per cubic millimeter in accordance with South African national guidelines15.
Demographic and clinical covariates included participant age in years at the time of MDR-TB diagnosis; sex (male or female); history of prior TB episodes (none, one, or two or more); education level (less than primary school, primary school complete, or beyond primary school); employment status (unemployed, employed part-time, or employed full-time); housing location (rural, township, or urban/suburban); baseline body mass index (BMI) in kilograms per meter squared; baseline CD4 count (categorized to less than 50, 50-199, 200-499, and greater than or equal to 500 cells per cubic millimeter [cells/mm3]); and ART regimen (categorized into efavirenz-based regimens, nevirapine-based regimens, ritonavir-boosted lopinavir-based regimens, and other or unknown regimens). Newer ART regimens such as integrase strand transfer inhibitors (INSTIs), including the fixed-dose regimen of tenofovir, emtricitabine, and dolutegravir, were not widely available in South Africa during the time of the parent study, and none of our participants took these regimens. Baseline CD4 counts were included if they were taken within 12 months prior to or four weeks after MDR-TB treatment initiation. Dolutegravir-based ART were not widely available in South Africa during the parent study period and therefore did not appear in the parent study data set. MDR-TB treatment initiation was defined as the first day that the first antitubercular medication was taken.
Missing Data
To ensure completeness of HIV viral load and CD4 count data, the authors searched the NHLS database, a nation-wide registry of laboratory results from every public facility in South Africa which has been shown a complete and reliable source of laboratory data in South Africa. Inclusion in NHLS has been used as a marker for engagement in care among PWH in South Africa20,21. Those without an available HIV viral load result after the NHLS search were included in the first and second models as a separate category but excluded from the third model, the purpose of which was to determine correlates of viral suppression among those with a known viral load, while CD4 count was imputed using multiple imputation for the 12.6% of individuals for whom a CD4 result was clinically unavailable using multiple imputation before being divided into categories described above. For all other variables, the extent and pattern of missing values were examined, and multiple imputation was used to impute missing values for employment status, education level, housing location, BMI, and prior TB episodes.
Ethical Approval
The parent study was approved by the Johns Hopkins School of Medicine Institutional Review Board (Application #NA_00078899), the province-level research committees in Eastern Cape and KwaZulu-Natal, and the IRB at the University of KwaZulu-Natal (Application #BE530/14). The clinicaltrials.gov registration number for the parent study is NCT02129244. This sub-study was approved as a change in research protocol to the original parent study.
RESULTS
Study Sample
Of the 1,479 PWH, mean age was 37.1 (SD 10.3) years, 809 (54.7%) were male, and 881 (59.5%) were taking ART, 681 of whom (77.3% of those taking ART or 46.0% of all PWH) had a viral load result, and 418 had known viral suppression (61.4% of those with a viral load result or 28.2% of all PWH). Of the 598 participants who were not taking ART pre-MDR-TB treatment, 412 (68.9% of those not taking ART or 27.9% of all PWH) were ART naïve. Additional demographic and clinical characteristics are presented in Table 1. Gender was almost evenly balanced in our sample. Though there are fewer incident cases of TB in South African women than men2, cohorts of PWH in sub-Saharan Africa, including South Africa, tend to have more women due to widespread perinatal HIV testing.
Table 1.
Demographic, Disease, and Laboratory Characteristics of PWH (N=1479)
| Total number of participants | All PWH | Not Taking ART | Taking ART (n=881) | |||
|---|---|---|---|---|---|---|
| VL not done^ | VL done (n=681) | |||||
| VL undetectable^^ | VL detectable^^^ | |||||
| 1479 (100%) | 598 (40.4%) | 200 (22.7%) | 418 (61.4%) | 263 (38.6%) | ||
| Age | Mean (SD) | 37.1 (10.3) | 36.2 (9.5) | 36.9 (11.7) | 39.4 (10.7) | 35.9 (9.7) |
| Sex | Male | 809 (54.7%) | 342 (57.2%) | 112 (56.3%) | 205 (49.0%) | 150 (57.0%) |
| Female | 670 (45.3%) | 256 (42.8%) | 87 (43.7%) | 213 (51.0%) | 113 (43.0%) | |
| Number of Prior TB Episodes | None | 649 (43.9%) | 337 (56.4%) | 81 (40.5%) | 159 (38.0%) | 72 (27.4%) |
| One | 657 (44.4%) | 204 (34.1%) | 86 (43.0%) | 207 (49.5%) | 160 (60.8%) | |
| Two or more | 117 (7.9%) | 35 (5.9%) | 19 (9.5%) | 37 (8.9%) | 26 (9.9%) | |
| Unknown | 56 (3.8%) | 22 (3.7%) | 14 (7.0%) | 15 (3.6%) | 5 (1.9%) | |
| Education Level | Less than primary school | 263 (17.8%) | 91 (15.2%) | 31 (15.5%) | 96 (23.0%) | 45 (17.1%) |
| Primary school complete | 795 (53.8%) | 334 (55.9%) | 111 (55.5%) | 199 (47.6%) | 151 (57.4%) | |
| Beyond primary school | 407 (27.5%) | 168 (28.1%) | 55 (27.5%) | 118 (28.2%) | 66 (25.1%) | |
| Unknown | 14 (1.0%) | 5 (0.8%) | 3 (1.5%) | 5 (1.2%) | 1 (0.4%) | |
| Employment Status | Unemployed | 972 (65.7%) | 378 (63.2%) | 134 (67.0%) | 267 (63.9%) | 193 (73.4%) |
| Employed part-time | 196 (13.3%) | 93 (15.6%) | 26 (13.0%) | 53 (12.7%) | 24 (9.1%) | |
| Employed full-time | 308 (20.8%) | 126 (21.1%) | 40 (20.0%) | 96 (23.0%) | 46 (17.5%) | |
| Unknown | 3 (0.2%) | 1 (0.2%) | 0 (0.0%) | 2 (0.5%) | 0 (0.0%) | |
| Housing | Rural or farm | 879 (59.4%) | 301 (50.3%) | 117 (58.8%) | 293 (70.1%) | 167 (63.5%) |
| Township | 544 (36.8%) | 271 (45.3%) | 75 (37.5%) | 110 (26.3%) | 88 (33.5%) | |
| Urban/CBD or suburban | 52 (3.5%) | 24 (4.0%) | 7 (3.5%) | 14 (3.4%) | 7 (2.6%) | |
| Unknown | 4 (0.3%) | 2 (0.3%) | 0 (0.0%) | 1 (0.2%) | 1 (0.4%) | |
| Baseline BMI | Mean (SD) | 20.7 (4.9) | 20.4 (4.7) | 21.1 (5.3) | 21.3 (5.3) | 20.0 (4.2) |
| Baseline CD4 (cells/mm3) | <50 | 240 (16.2%) | 131 (21.9%) | 24 (12.0%) | 13 (3.1%) | 72 (27.4%) |
| 50-199 | 445 (30.1%) | 216 (36.1%) | 50 (25.0%) | 80 (19.1%) | 99 (37.6%) | |
| 200-499 | 414 (28.0%) | 139 (23.2%) | 44 (22.0%) | 171 (40.9%) | 60 (22.8%) | |
| ≥500 | 194 (13.1%) | 54 (9.0%) | 19 (9.5%) | 110 (26.3%) | 11 (4.2%) | |
| Unknown | 186 (12.6%) | 58 (9.7%) | 63 (31.5%) | 44 (10.5%) | 21 (8.0%) | |
| ART Regimen | EFV-based | 161 (80.5%) | 342 (81.8%) | 197 (74.9%) | ||
| NVP-based | 21 (10.5%) | 40 (9.6%) | 16 (6.1%) | |||
| LPVr-based | 11 (5.5%) | 30 (7.2%) | 41 (15.6%) | |||
| Other or unknown | 7 (3.5%) | 6 (1.4%) | 9 (3.4%) | |||
PWH, people with HIV; ART, antiretroviral therapy; TB, tuberculosis; HIV, human immunodeficiency virus; VL, human immunodeficiency viral load; SD, standard deviation; CBD, central business district; BMI, body mass index; mm3, cubic millimeters; ^VL available within 12 months prior to or four weeks after MDR-TB treatment initiation; ^^VL <400 copies/mm3; ^^^VL ≥400 copies/mm3. Note: Some percentages do not add to 100.0% due to rounding.
Model 1: ART Use at MDR-TB Baseline
Table 2 displays the odds of taking ART prior to MDR-TB treatment initiation by demographic and clinical characteristics. In multivariable analysis, having a history of TB significantly increased the odds of taking ART. Those who were employed part-time had lower odds of pre-treatment ART, along with persons residing in either a township or the city/suburb. Those presenting with a CD4 count less than 200 cells/mm3 were less likely to be taking ART.
Table 2.
Odds Ratios for Taking ART MDR-TB Initiation (n=1479)
| Univariable Model | Multivariable Model | ||||
|---|---|---|---|---|---|
| OR | 95% CI | aOR | 95% CI | ||
| Age (years) | 1.02** | 1.01-1.03 | 1.01 | 0.99-1.02 | |
| Female (ref. male) | 1.18 | 0.96-1.46 | 1.18 | 0.93-1.51 | |
| Prior TB episodes (ref. none) | One | 2.35*** | 1.87-2.96 | 2.58*** | 2.03-3.28 |
| Two or more | 2.45*** | 1.58-3.78 | 2.82*** | 1.78-4.48 | |
| Education level (ref. less than primary school) | Primary school complete | 0.72* | 0.54-0.96 | 0.91 | 0.66-1.27 |
| More than primary school | 0.75 | 0.54-1.03 | 1.02 | 0.70-1.48 | |
| Housing (ref. rural or farm) | Township | 0.52*** | 0.42-0.65 | 0.49*** | 0.38-0.62 |
| City/CBD or suburban | 0.60 | 0.34-1.06 | 0.52* | 0.29-0.95 | |
| CD4 count (cells/mm3, ref. ≥500) | <50 | 0.32*** | 0.22-0.47 | 0.33*** | 0.21-0.50 |
| 50-199 | 0.40*** | 0.28-0.58 | 0.40*** | 0.27-0.58 | |
| 200-499 | 0.70 | 0.49-1.01 | 0.69 | 0.47-1.01 | |
| Employment status (ref. unemployed) | Employed part-time | 0.70* | 0.52-0.96 | 0.65* | 0.47-0.91 |
| Employed full-time | 0.92 | 0.71-1.20 | 0.96 | 0.72-1.29 | |
| Baseline BMI | 1.02 | 0.99-1.04 | 1.02 | 0.99-1.05 | |
HIV, human immunodeficiency virus; PWH, people with HIV; ART, antiretroviral therapy; VL, viral load; MDR-TB, multidrug-resistant tuberculosis; OR, odds ratio; CI, confidence interval; aOR, adjusted odds ratio; ref., reference; TB, tuberculosis; BMI, body mass index; mm3, cubic millimeters; *p<0.05; **p<0.01; ***p<0.001
Model 2: Having an HIV Viral Load Among PWH Taking ART
Among the 881 participants taking ART at the time of MDR-TB diagnosis, 681 (77.3%) had a viral load result. Table 3 shows the odds of having a viral load according to demographic and clinical characteristics. In multivariable analysis, having one prior episode of TB raised odds of having a viral load result, while living in a township or having a CD4 count of less than 200 cells/mm3 lowered odds of having a viral load result.
Table 3.
Odds Ratios for Having an HIV Viral Load Among PWH Taking ART MDR-TB Initiation (n=881)
| Univariable Model | Multivariable Model | ||||
|---|---|---|---|---|---|
| OR | 95% CI | aOR | 95% CI | ||
| Age (years) | 1.01 | 0.99-1.03 | 1.01 | 0.99-1.03 | |
| Female (ref. male) | 1.17 | 0.85-1.60 | 1.29 | 0.89-1.85 | |
| Prior TB episodes (ref. none) | One | 1.48* | 1.05-2.08 | 1.59* | 1.11-2.28 |
| Two or more | 1.18 | 0.66-2.11 | 1.24 | 0.67-2.26 | |
| Education level (ref. less than primary school) | Primary school complete | 0.73 | 0.47-1.13 | 0.87 | 0.54-1.39 |
| More than primary school | 0.78 | 0.48-1.27 | 0.97 | 0.57-1.65 | |
| Housing (ref. rural or farm) | Township | 0.68* | 0.48-0.95 | 0.68* | 0.48-0.95 |
| City/CBD or suburban | 0.77 | 0.32-1.86 | 0.73 | 0.30-1.81 | |
| CD4 count (cells/mm3, ref. ≥500) | <50 | 0.50* | 0.26-0.96 | 0.48* | 0.24-0.96 |
| 50-199 | 0.51* | 0.30-0.93 | 0.48* | 0.35-0.90 | |
| 200-499 | 0.72 | 0.41-1.24 | 0.66 | 0.37-1.16 | |
| Employment status (ref. unemployed) | Employed part-time | 0.86 | 0.53-1.40 | 0.84 | 0.51-1.38 |
| Employed full-time | 1.03 | 0.69-1.54 | 1.13 | 0.74-1.74 | |
| Baseline BMI | 0.99 | 0.96-1.02 | 0.98 | 0.94-1.01 | |
HIV, human immunodeficiency virus; PWH, people with HIV; ART, antiretroviral therapy; VL, viral load; MDR-TB, multidrug-resistant tuberculosis; OR, odds ratio; CI, confidence interval; aOR, adjusted odds ratio; ref., reference; TB, tuberculosis; BMI, body mass index; mm3, cubic millimeters; *p<0.05; **p<0.01; ***p<0.001
Model 3: Viral Suppression Among PWH Taking ART for Whom a Viral Load Result was Available
Among the 681 people taking ART for whom a viral load result was available, 418 (61.4%) were virally suppressed. Those with viral suppression were significantly older and had higher educational obtainment and CD4 counts. Living in a township, having only one prior TB episode, and taking second-line ART, specifically LPVr, significantly lowered odds of viral suppression. Table 4 displays the results of the regression predicting viral suppression in this group.
Table 4.
Odds Ratios for HIV Viral Suppression Among PWH Taking ART With a Known VL at MDR-TB Initiation (n=681)
| Univariable Model | Multivariable Model | ||||
|---|---|---|---|---|---|
| OR | 95% CI | aOR | 95% CI | ||
| Age (years) | 1.03*** | 1.02-1.05 | 1.04*** | 1.01-1.06 | |
| Female (ref. male) | 1.38* | 1.01-1.88 | 0.94 | 0.62-1.43 | |
| Prior TB episodes (ref. none) | One | 0.59** | 0.42-0.83 | 0.64* | 0.42-0.99 |
| Two or more | 0.63 | 0.36-1.12 | 0.63 | 0.30-1.30 | |
| Education level (ref. less than primary school) | Primary school complete | 0.62* | 0.41-0.94 | 1.17 | 0.69-2.01 |
| More than primary school | 0.84 | 0.53-1.34 | 1.96* | 1.05-3.65 | |
| Housing (ref. rural or farm) | Township | 0.71 | 0.51-1.00 | 0.59* | 0.38-0.91 |
| City/CBD or suburban | 1.15 | 0.46-2.91 | 1.04 | 0.90-2.42 | |
| CD4 count (cells/mm3, ref. ≥500) | <50 | 0.02*** | 0.01-0.05 | 0.02*** | 0.01-0.04 |
| 50-199 | 0.08*** | 0.04-0.15 | 0.06*** | 0.03-0.13 | |
| 200-499 | 0.27*** | 0.14-0.54 | 0.23*** | 0.11-0.46 | |
| ART regimen (ref. efavirenz-based) | Nevirapine-based | 1.44 | 0.79-2.64 | 1.66 | 0.84-3.31 |
| Ritonavir-boosed Lopinavir-based | 0.42** | 0.26-0.70 | 0.38** | 0.19-0.73 | |
| Other or Unknown ART | 0.38 | 0.13-1.09 | 0.42 | 0.11-1.60 | |
| Employment status (ref. unemployed) | Employed part-time | 1.59 | 0.95-2.67 | 1.34 | 0.72-2.49 |
| Employed full-time | 1.51* | 1.01-2.24 | 1.48 | 0.90-2.42 | |
| Baseline BMI | 1.05** | 1.01-1.09 | 1.00 | 0.96-1.05 | |
HIV, human immunodeficiency virus; PWH, people with HIV; ART, antiretroviral therapy; VL, viral load; MDR-TB, multidrug-resistant tuberculosis; OR, odds ratio; CI, confidence interval; aOR, adjusted odds ratio; ref., reference; TB, tuberculosis; BMI, body mass index; mm3, cubic millimeters; *p<0.05; **p<0.01 ; ***p<0.001
DISCUSSION
This study examines pre-treatment factors correlated with taking ART, having an HIV viral load result, and having viral suppression among PWH initiating MDR-TB treatment in South Africa. Although others established that HIV co-infection does not necessarily predict poor MDR-TB outcomes, evidence suggests that MDR-TB treatment success requires treating and controlling HIV in co-infected patients8-12. South Africa has adopted the WHO 90-90-90 strategy for HIV control and elimination. However, in this population of PWH initiating MDR-TB treatment, the overall HIV care cascade demonstrated only 59.5% of newly diagnosed MDR-TB cases among PWH were taking ART, 46.0% had a viral load obtained in the last 12 months, and only 28.2% had evidence of viral suppression. It is possible, however, that our findings represent a channeling bias, as those diagnosed with TB may have lower CD4 counts and may not have been fully engaged in HIV treatment prior to MDR-TB diagnosis22-24.
In South Africa, same-day HIV viral load testing is not widely available, and electronic gate keeping by the NHLS, a cost-containment strategy, prohibits repeat HIV viral load testing within a 12-month period. The impact of limited access to HIV viral load monitoring was clearly demonstrated in this data set with only 681 (46.0%) of 1,479 PWH having a viral load result during the year prior to or immediately after MDR-TB diagnosis, despite demonstrating engagement with the health system due to taking ART and the recent MDR-TB diagnosis. Unknown HIV viral load limits clinicians’ ability to rapidly identify adherence challenges which may lead to virologic failure and prevents the optimization of ART regimens.
Prior TB episodes raised the odds of taking ART and having a viral load result but decreased the odds of viral suppression, though the significance of the number of prior TB episodes (one versus two or more) varied across the models. While having a prior TB episode could serve as a marker for interaction with the health system, published literature has also linked HIV status, lower CD4 count, and detectable HIV viral load to TB recurrence and reinfection22-24. Further, TB of any type is an AIDS-defining co-infection in PWH, and TB/HIV co-infection would be sufficient reason to initiate ART even under previous clinical guidelines which otherwise required meeting CD4 thresholds. It follows that a prior episode of TB would both raise odds both of taking ART and having a viral load measure, either due to meeting a guideline-based requirement for ART or due to engagement with the health system, and of having a detectable viral load while on ART.
Patients taking second-line ART may have been in HIV care longer, as the use of a second-line regimen may indicate a history of adherence challenges lead to the development of ART failure in many patients. Clinicians may wish to consider the adherence factors that led to this prior failure to prevent a similar failure of second-line ART or of MDR-TB treatment. Taking a second-line ART regimen, specifically a boosted protease inhibitor like LPV/r, also complicates integrated MDR-TB/HIV treatment due to increased pill burden and potential for adverse effects of medication. Because of these issues and the relationship between second-line ART and a lack of HIV viral suppression, the significance of taking a second-line ART regimen at the time of MDR-TB diagnosis should be a flag for clinicians to enhance adherence support mechanisms.
INSTIs, including a fixed-dose combination containing dolutegravir, were not widely available in South Africa prior to 2020. Since the completion of the parent study, South Africa has rolled out a once-daily, fixed-dose combination containing the INSTI dolutegravir as the first-line treatment for all PWH. Given the tolerability of this newer ART regimen, its ability to rapidly suppress viral replication, and its higher burden to viral resistance, we anticipate that future cohorts of PWH who are diagnosed with MDR-TB may be more likely to have viral suppression at MDR-TB baseline if they are taking these newer regimens. We recommend repeating this analysis once sufficient data from the post-INSTI era is available.
The association between education and HIV viral load in our sample has rarely been investigated or has not reached significance in other research25. While a higher level of education does not guarantee greater health literacy, recent research indicates that higher education may increase health literacy and adherence related self-efficacy26. This association could be meaningful for clinicians, as those with less formal education require counseling and education strategies that are appropriately tailored for their level of knowledge and understanding, which may enhance adherence.
Living in a township decreased the odds of taking ART, having a viral load result, and achieving viral suppression. Township living is a marker of poverty in South Africa, and townships continue to have poorer social services, more crowded conditions, and less robust infrastructure nearly 30 years after the end of apartheid. These difficult living conditions could contribute to reduced access to ART as well as reduced ART adherence, increased stigma, poorer overall health, or a variety of health system and personal factors that make it difficult to achieve and maintain HIV viral suppression.
The association between older age and a higher likelihood of HIV viral suppression is well-known, and CD4 cell count is expected to recover in PWH taking ART whose HIV viral load is suppressed, and this association is widely documented25,27-30. These correlations may be used to facilitate targeted programmatic interventions. Because obtaining frequent HIV viral load measures is difficult, it may be easier and faster to obtain a CD4 count due to the wide availability of point-of-care CD4 testing and there is no electronic gate keeping restriction on this test. Stable CD4 count and older age may reassure clinicians, while younger patients or those with immunosuppression should be targeted to receive HIV viral load testing and additional support.
South Africa has several programs designed to support people with MDR-TB, including those with HIV and those high-risk individuals identified in this analysis such as youth and people who live in townships. For example, our study team has seen each of the following examples implemented in South African TB hospitals as part of the standard of care. South Africa has a network of community health workers who can make home visits and search for people who miss appointments to support adherence31,32. Tuberculosis hospitals employ social workers who can help identify and offer suggestions to overcome barriers to treatment success, including assisting people undergoing MDR-TB treatment with applying for a disability grant, designed to assist those living in poverty to overcome the cost of treatment and the missed opportunities for work32. Adolescents can benefit from age-specific ART support groups33. The parent study intervention of a nurse case manager may also prove effective in supporting people identified as high-risk18. In practice, however, the actual support interventions available at each facility vary. In order to ensure effective support for high-risk individuals, the entire clinical team including physicians, clinical officers, nurses, community health workers, and other professionals must work to identify needs and tailor interventions to ensure their effectiveness. Using a combination of multiple interventions may be necessary to achieve the best outcomes.
Recently, further revisions to MDR-TB treatment guidelines have introduced newer, shorter treatment regimens using additional novel antitubercular agents34. This analysis considered factors affecting HIV indicators at the time of MDR-TB diagnosis and treatment initiation, so the type of MDR-TB treatment regimen that was later started would not have affected these results. However, the effects of the covariates identified in this analysis on MDR-TB and HIV treatment outcomes may differ with these newer regimens. Importantly, as newer MDR-TB treatment regimens continue to decrease the length of treatment, the opportunity for achieving HIV viral suppression during the MDR-TB treatment period decreases. Thus, rapidly identifying PWH, starting ART as soon as possible, and ensuring good adherence and frequent viral load monitoring become even more important. The risk factors for not taking ART, not having an HIV viral load result, and not having HIV viral suppression identified in this analysis should be considered red flags for rapid action, especially for those PWH and MDR-TB whose treatment periods will be reduced.
There are several limitations to this study. First, achieving HIV viral suppression requires good adherence to an ART regimen to which an individual has not developed resistance. Our analysis did not include any measures of treatment adherence as it used data obtained at enrollment in the parent trial. As we were also not able to rely on patient self-report of HIV diagnosis date, we do not have an estimate of the number of years living with HIV. Further, we did not have access to HIV resistance testing as it is not routinely performed in South Africa. This information would inform ART regimen optimization and viral suppression. The results of this analysis may not be generalizable to populations for whom rapid HIV viral load or HIV genotype results are available, as these would be better indicators of ART efficacy and prior adherence. However, we feel that these results may reflect associations in similar low-resource settings with high MDR-TB/HIV co-infection rates and will be useful to identify patients in need of intensive monitoring and adherence support.
Acknowledgements:
The parent study was funded through the National Institutes of Health/National Institute of Allergy and Infectious Disease (NIAID R01AI104488-02; ClinicalTrials.gov Identifier: NCT02129244; PI Jason Farley). Coauthor Keri Geiger's research and training is funded by the National Center for Advancing Translational Sciences grant number TL1 TR003100. Coauthor Alanna Bergman’s research and training is funded by the National Institute of Nursing Research grant number 1F31NR020588-01, as well as the Johns Hopkins University School of Nursing Discovery and Innovation Award. Coauthor Kathrine McNabb’s research is supported by the National Institute of Allergy and infectious disease (F30AI165167). Coauthor Omeid Heidari's research is supported by the National Institutes of Health/National Center for Advancing Translational Sciences (KL2TR002317). The authors have no conflicts of interest to disclose.
References
- 1.World Health Organization. Tuberculosis. WHO.int Web site. https://www.who.int/news-room/fact-sheets/detail/tuberculosis. Updated 2020. Accessed Apr 15, 2021. [Google Scholar]
- 2.World Health Organization. Tuberculosis profile: South Africa. 2022. WHO.int. Web site. https://www.who.int/teams/global-tuberculosis-programme/data#profiles. [Google Scholar]
- 3.World Health Organization. Definitions and reporting framework for tuberculosis – 2013 revision (updated January 2020). Geneva: World Health Organization, 2013. [Google Scholar]
- 4.Schnippel K, Firnhaber C, Berhanu R, Page-Shipp L, Sinanovic E. Adverse drug reactions during drug-resistant TB treatment in high HIV prevalence settings: A systematic review and meta-analysis. J Antimicrob Chemother. 2017;72(7):1871–1879. doi: 10.1093/jac/dkx107. [DOI] [PubMed] [Google Scholar]
- 5.Schnippel K, Berhanu RH, Black A, et al. Severe adverse events during second-line tuberculosis treatment in the context of high HIV co-infection in South Africa: A retrospective cohort study. BMC Infect Dis. 2016;16(1):593–0. doi: 10.1186/s12879-016-1933-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.van Heeswijk RP, Dannemann B, Hoetelmans RM. Bedaquiline: A review of human pharmacokinetics and drug-drug interactions. J Antimicrob Chemother. 2014;69(9):2310–2318. doi: 10.1093/jac/dku171. [DOI] [PubMed] [Google Scholar]
- 7.O’Donnell MR, Padayatchi N, Daftary A, et al. Antiretroviral switching and bedaquiline treatment of drug-resistant tuberculosis HIV co-infection. Lancet HIV. 2019;6(3):e201. doi: 10.1016/s2352-3018(19)30035-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Daftary A, Padayatchi N, O'Donnell M. Preferential adherence to antiretroviral therapy over tuberculosis treatment: A qualitative study of drug-resistant TB/HIV co-infected patients in South Africa. Glob Public Health. 2014;9(9):1107–1116. doi: 10.1080/17441692.2014.934266. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Bastard M, Sanchez-Padilla E, Cros Pd, et al. Outcomes of HIV-infected versus HIV-non-infected patients treated for drug-resistance tuberculosis: Multicenter cohort study. PLOS ONE. 2018;13(3):e0193491. doi: 10.1371/journal.pone.0193491. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Umanah T, Ncayiyana J, Pandanilam X, Nyasulu PS. Treatment outcomes in multidrug resistant tuberculosis-human immunodeficiency virus co-infected patients on anti-retroviral therapy at sizwe tropical disease hospital Johannesburg, South Africa. BMC Infectious Diseases. 2015;15. doi: 10.1186/s12879-015-1214-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Chem ED, Van Hout MC, Hope V. Treatment outcomes and antiretroviral uptake in multidrug-resistant tuberculosis and HIV co-infected patients in sub-Saharan Africa: A systematic review and meta-analysis. BMC Infectious Diseases. 2019;19(1):723. doi: 10.1186/s12879-019-4317-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Isaakidis P, Casas EC, Das M, Tseretopoulou X, Ntzani EE, Ford N. Treatment outcomes for HIV and MDR-TB co-infected adults and children: Systematic review and meta-analysis. Int J Tuberc Lung Dis. 2015;19(8):969–978. doi: 10.5588/ijtld.15.0123. [DOI] [PubMed] [Google Scholar]
- 13.Brust JCM, Shah NS, Mlisana K, et al. Improved survival and cure rates with concurrent treatment for multidrug-resistant tuberculosis-human immunodeficiency virus coinfection in South Africa. Clin Infect Dis. 2018;66(8):1246–1253. doi: 10.1093/cid/cix1125. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.World Health Organization. Companion handbook to the WHO guidelines for the programmatic management of drug-resistant tuberculosis. 2014. http://www.who.int/iris/handle/10665/130918. [PubMed]
- 15.World Health Organization. The use of antiretroviral drugs for treating and preventing HIV infection: Recommendations for a public health approach. 2013. [PubMed]
- 16.National Department of Health, Republic of South Africa. National consolidated guidelines for the prevention of mother-to-child transmission of HIV (PMTCT) and the management of HIV in children, adolescents, and adults. 2015.
- 17.Department of Health, Republic of South Africa. Management of rifampicin-resistant tuberculosis: A clinical reference guide. 2019.
- 18.Farley JE, Kelly AM, Reiser K, et al. Development and evaluation of a pilot nurse case management model to address multidrug-resistant tuberculosis (MDR-TB) and HIV in South Africa. PloS one. 2014;9(11):e111702. doi: 10.1371/journal.pone.0111702. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Harris PA, Taylor R, Thielke R, Payne J, Gonzalez N, Conde JG. Research electronic data capture (REDCap)—A metadata-driven methodology and workflow process for providing translational research informatics support. Journal of biomedical informatics. 2009;42(2):377–381. doi: 10.1016/j.jbi.2008.08.010. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Bassett IV, Huang M, Cloete C, et al. Using national laboratory data to assess cumulative frequency of linkage after transfer to community-based HIV clinics in South Africa. J Int AIDS Soc. 2019;22(6):e25326. doi: 10.1002/jia2.25326 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Bassett IV, Huang M, Cloete C, et al. Assessing the completeness and accuracy of south african national laboratory CD4 and viral load data: A cross-sectional study. BMJ Open. 2018;8(8). doi: 10.1136/bmjopen-2018-021506. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Moodley Y, Govender K. A systematic review of published literature describing factors associated with tuberculosis recurrence in people living with HIV in Africa. Afr Health Sci. 2015;15(4):1239–1246. doi: 10.4314/ahs.v15i4.24. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Hawken M, Nunn P, Godfrey-Faussett P, et al. Increased recurrence of tuberculosis in HIV-1-infected patients in Kenya. The Lancet. 1993;342(8867):332–337. doi: 10.1016/0140-6736(93)91474-Z. [DOI] [PubMed] [Google Scholar]
- 24.Glynn JR, Murray J, Bester A, Nelson G, Shearer S, Sonnenberg P. High rates of recurrence in HIV-infected and HIV-uninfected patients with tuberculosis. The Journal of Infectious Diseases. 2010;201(5):704–711. doi: 10.1086/650529. [DOI] [PubMed] [Google Scholar]
- 25.Mbengue MAS, Chasela C, Onoya D, Mboup S, Fox MP, Evans D. Clinical predictor score to identify patients at risk of poor viral load suppression at six months on antiretroviral therapy: Results from a prospective cohort study in Johannesburg, South Africa. Clin Epidemiol. 2019;11:359–373. Accessed Jun 28, 2022. doi: 10.2147/CLEP.S197741. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Wolf MS, Davis TC, Osborn CY, Skripkauskas S, Bennett CL, Makoul G. Literacy, self-efficacy, and HIV medication adherence. Patient Education and Counseling. 2007;65(2):253–260. doi: 10.1016/j.pec.2006.08.006. [DOI] [PubMed] [Google Scholar]
- 27.George S, McGrath N, Oni T. The association between a detectable HIV viral load and non-communicable diseases comorbidity in HIV positive adults on antiretroviral therapy in Western Cape, South Africa. BMC Infect Dis. 2019;19(1):348. doi: 10.1186/s12879-019-3956-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Jiamsakul A, Kariminia A, Althoff KN, et al. HIV viral load suppression in adults and children receiving antiretroviral therapy-results from the IeDEA collaboration. J Acquir Immune Defic Syndr. 2017;76(3):319–329. doi: 10.1097/QAI.0000000000001499. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Joseph Davey D, Abrahams Z, Feinberg M, et al. Factors associated with recent unsuppressed viral load in HIV-1-infected patients in care on first-line antiretroviral therapy in South Africa. Int J STD AIDS. 2018;29(6):603–610. doi: 10.1177/0956462417748859. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Pillay T, Cornell M, Fox MP, et al. Recording of HIV viral loads and viral suppression in South African patients receiving antiretroviral treatment: A multicentre cohort study. Antivir Ther. 2020;25(5):257–266. doi: 10.3851/IMP3371.stylefix [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Department of Health, Republic of South Africa. Multi-drug resistant tuberculosis: A policy framework on decentralized and deinstitutionalized management for South Africa. 2019.
- 32.Department of Health, Republic of South Africa. Mangement of drug-resistant tuberculosis: Policy guidelines. 2011.
- 33.United Nations Children’s’ Fund (UNICEF). HIV treatment, care, and support for adolescents living with HIV in eastern and southern Africa: A review of interventions for scale. 2011.
- 34.World Health Organization. WHO consolidated guidelines on tuberculosis. Module 4: treatment – drug-resistant tuberculosis treatment. 2022. https://www.who.int/publications/i/item/9789240063129. [PubMed] [Google Scholar]