Plasma Immune Biomarkers Predictive of Progression to Active Tuberculosis in Household Contacts of Patients With Tuberculosis

Abstract

Background

The progression from Mycobacterium tuberculosis infection to active tuberculosis disease varies among individuals, and identifying biomarkers to predict progression is crucial for guiding interventions. In this study, we aimed to determine plasma immune biomarker profiles in healthy household contacts of index patients with pulmonary tuberculosis, who either progressed to tuberculosis or remained as nonprogressors.

Methods

A cohort of household contacts of adults with pulmonary tuberculosis was enrolled, consisting of 15 contacts who progressed to tuberculosis disease and 15 nonprogressors. Plasma samples were collected at baseline, 4 months, and 12 months to identify predictive tuberculosis progression markers.

Results

Our findings revealed that individuals in the progressor group exhibited significantly decreased levels of interferon (IFN) γ, tumor necrosis factor α, interleukin 2, IL-1α, IL-1β, and 17A, and interleukin 1 receptor antagonist (IL-1Ra) at baseline, month 4, and month 12. In contrast, the progressor group displayed significantly elevated levels of IFN-α, IFN-β, interleukin 6 and 12, granulocyte-macrophage colony-stimulating factor (GM-CSF), interleukin 10 (IL-10) and 33 (IL-33), CCL2, CCL11, CXCL8, CXCL10, CX3CL1, vascular endothelial growth factor, granzyme B, and programmed death ligand -1 compared to the nonprogressor group at baseline, months 4 and 12. Receiver operating characteristic analysis (ROC) identified IFN-γ, GM-CSF, IL-1Ra, CCL2, and CXCL10 as the most promising predictive markers, with an area under the receiver operating characteristic curve of ≥90. Furthermore, combinatorial analysis demonstrated that GM-CSF, CXCL10, and IL-1Ra, when used in combination, exhibited high accuracy in predicting progression to active tuberculosis disease.

Conclusions

Our study suggests that a specific set of plasma biomarkers, GM-CSF, CXCL10, and IL-1Ra, can effectively identify household contacts at significant risk of developing tuberculosis disease. These findings have important implications for early intervention and preventive strategies in tuberculosis-endemic regions.

Plasma levels of cytokines interferon γ, granulocyte-macrophage colony-stimulating factor (GM-CSF), and interleukin 1 receptor antagonist (IL-1Ra) and chemokines CCL2 and CXCL10 predict tuberculosis progression in contacts. Combinatorial analysis of GM-CSF, CXCL10, and IL-1Ra shows high accuracy in predicting active tuberculosis.

Tuberculosis, caused by Mycobacterium tuberculosis (M.tb), remains a global health challenge, with a significant burden on populations worldwide. In 2021 alone, approximately 1.6 million individuals lost their lives to tuberculosis, highlighting the urgency of addressing this infectious disease [1]. While a substantial portion of the global population is estimated to have been infected with M.tb, only a minority, about 5%–10%, progress from latent M.tb infection to symptomatic tuberculosis disease [2]. Precisely identifying individuals at high risk of developing tuberculosis disease is crucial for achieving the World Health Organization's tuberculosis elimination goals.

Existing diagnostic tools, such as interferon (IFN) γ release assays (IGRAs) and the tuberculin skin test (TST), have limitations in distinguishing between tuberculosis disease and latent M.tb infection [3–11]. Moreover, they have low positive predictive values and cannot be used to predict the progression from latent infection to active tuberculosis disease [3, 4, 12, 13]. Indeed, the identification of biomarkers capable of accurately discerning individuals at the highest risk of transitioning from latent tuberculosis infection (LTBI) to active tuberculosis disease remains a challenging endeavor. This challenge is rooted in the fact that, in individuals with LTBI, the M.tb bacteria are primarily sequestered within lung granulomas and draining lymph nodes, making direct detection of the pathogen exceedingly difficult [14, 15]. However, promising avenues have emerged through the examination of host signals within the blood compartment, particularly inflammatory markers, which have demonstrated their potential to mirror the intricate host-pathogen interactions occurring at the localized sites of disease [16–18]. Leveraging these blood-based biomarkers offers a promising strategy for identifying individuals who are on the trajectory from M.tb infection to active tuberculosis disease [16], but limited data exist in India, a country with the world's largest tuberculosis burden.

Therefore, in the current study, we assessed the intricate immunological distinctions between individuals who progress to tuberculosis disease (progressors) and those who do not (nonprogressors) in the early stages after contact with an index tuberculosis case. The primary objective is to elucidate the unique immune profiles of progressors and nonprogressors, seeking predictive markers that can identify those at risk of tuberculosis disease progression.

METHODS

Ethical Approval

The parent protocol of this study received ethical approval from the institutional review board/ethics committees of Johns Hopkins University, Byramjee Jeejeebhoy Government Medical College (BJGMC), and the Indian Council of Medical Research (ICMR)–National Institute for Research in Tuberculosis (NIRT) (ICMR-NIRT-NIRT-IEC-2020-021).

Study Cohort

A cohort of healthy household contacts (HHCs) of individuals newly diagnosed with pulmonary tuberculosis was established at 2 study sites in India: ICMR-NIRT in Chennai and BJGMC in Pune. This collaborative effort was part of the Cohort for Tuberculosis Research by the Indo-US Medical Partnership (C-TRIUMPH) study. The enrolled participants (progressors and nonprogressors) were followed up between August 2014 and December 2017. Detailed information regarding the C-TRIUMPH study design and implementation can be found elsewhere [19]. The criteria used for classifying study participants are outlined in Table 1. All HHCs underwent clinical and laboratory assessments for tuberculosis at the beginning of the study and during subsequent follow-up visits. Additional samples were not collected if the recruited participants had active tuberculosis diagnosed.

Table 1.

Definitions Used to Stratify Study Participants

Classification	Definition
Progressors	HHCs who developed tuberculosis at any time ≥2 months after tuberculosis diagnosis in the index case, with tuberculosis diagnosis based on chest radiography, positive sputum smear, culture, and GeneXpert M.tb/RIF assay
Nonprogressors	HHCs who remained healthy and did not develop tuberculosis during the 2-y follow-up period (ie, had negative symptom screen, chest radiography, TST, and IGRA findings)

Abbreviations: HHCs, healthy household contacts; IGRA, interferon γ release assay; TST, tuberculin skin test.

TSTs and IGRA were performed at the start of the study and repeated at each visit if the previous test yielded a negative result. HHCs were considered to have active tuberculosis disease if they had positive chest radiographic, sputum smear and culture, or GeneXpert M.tb/RIF findings, and prevalent tuberculosis cases were excluded. An equal number of HHCs, referred to as “nonprogressors,” were matched with progressors in terms of age and sex. In the nonprogressors, active tuberculosis did not develop during the study period. To minimize variability, we followed a standardized protocol for the collection, transportation, and storage of plasma samples. Samples were collected using venipuncture. They were then transported at ambient temperature to the laboratory facilities for processing. On arrival, samples were centrifuged to separate plasma, which was subsequently aliquoted into single-use tubes to avoid freeze-thaw cycles, and stored at −80°C until further processing. These steps were taken to maintain sample integrity and minimize variation due to preanalytical conditions.

Diagnostic Tests for Active Tuberculosis

Sputum samples were collected from all participants and subjected to the GeneXpert M.tb/RIF assay, followed by culture using Löwenstein-Jensen media and Mycobacterial Growth Indicator Tube liquid culture. Samples testing positive for M.tb by either test were classified as confirmed tuberculosis cases.

Multiplex Immune Analyte Assay

The circulating levels of cytokines, chemokines, and growth factors were assessed in stored plasma samples using the Human XL Cytokine Magnetic Luminex Performance Assay 45-plex Fixed Panel (R & D Systems), following the manufacturer's instructions. The panel included the following analytes: IFN-α, IFN-β, IFN-γ, interleukin 1α (IL-1α), interleukin 1 receptor antagonist (IL-1ra), interleukin 1β (IL-1β), interleukin 2 (IL-2), interleukin 3, 4, and 5, interleukin 6 (IL-6), interleukin 7, interleukin 10 (IL-10), interleukin 12p70 (IL-12p70), interleukin 13 and 15, interleukin 17 (IL-17) A and E, interleukin 33 (IL-33), and tumor necrosis factor (TNF) α. The panel also included granulocyte colony-stimulating factor (G-CSF), granulocyte-macrophage colony-stimulating factor (GM-CSF), CXCL1, CXCL2, interleukin 8/CXCL8, CXCL10/IP10, CCL2, CCL3, CCL4, CCL5/RANTES, CCL11/eotaxin, CCL19, CCL20, CX3CL1/fractalkine, CD40 ligand (CD40L), epidermal growth factor (EGF), fibroblast growth factor basic, FMS-like tyrosine kinase 3 ligand (FLT-3L), granzyme B, programmed death ligand 1 (PDL-1), platelet-derived growth factor (PDGF) AA and AB/BB, transforming growth factor α (TGFα), Tumor Necrosis Factor–Related Apoptosis-Inducing Ligand (TRAIL), and vascular endothelial growth factor (VEGF). The lower detection limits for all parameters are shown in Supplementary Table 3.

Statistical Analysis

Statistical analyses were conducted using GraphPad Prism 10.0.3 and JMP 17.0.0 software. To compare the progressor and nonprogressor groups, the Mann-Whitney U test was used, followed by Holm's multiple comparison for post hoc correction. Differences were considered statistically significant at P ≤ .05. To evaluate the diagnostic potential of each biomarker, receiver operating characteristic (ROC) analysis was used, considering sensitivity, specificity, and overall value. In addition, a comprehensive analysis was performed to identify the optimal combination of circulating plasma cytokines among the examined immune biomarkers.

RESULTS

Study Cohort

Between August 2014 and December 2017, a total of 1051 HHCs were recruited for the parent study [19]. Among these participants, 20 individuals (1.9%) were categorized as progressors as they developed active tuberculosis within 2 years. The time between enrollment and the diagnosis of active tuberculosis ranged from 3 to 21 months. Supplementary Figure 1 provides an overview of the study design and participant selection. Notably, 15 known tuberculosis progressors had plasma samples available at recruitment, 9 at month 4, and 4 at month 12. To maintain parity, 30 nonprogressors had samples collected at recruitment, and another 30 had samples collected at months 4 and 12. Figure 1 illustrates the time taken for the recruited progressors to progress to active tuberculosis. Table 2 presents the clinical and demographic characteristics of the study cohort.

Figure 1.

Time to active tuberculosis among the 15 individual recruited progressors.

Table 2.

Clinical and Demographic Characteristics of Participants

Characteristics	Participants, No. (%)a
	Progressors (n = 15)	Nonprogressors (n = 30)
Age, median (IQR), y	30 (23–35)	29 (21–38)
Sex
Male	6 (40)	16 (53)
Female	9 (60)	14 (47)
HIV status
Positive	0 (0)	0 (0)
Negative	15 (93)	30 (100)
History of diabetes
Yes	0 (0)	0 (0)
No	15 (100)	30 (100)
IGRA outcome
Positive	9 (60)	10 (33)
Negative	6 (40)	20 (67)
Basis for patient categorization
Culture positive	33% (5)	…
Xpert positive	6.7 % (1)	…
Both positive	26.7% (4)	…
AFB smear	6.7% (1)	…
Positive by culture, Xpert, and smear	0	…

Abbreviations: AFB, acid-fast bacilli; HIV, human immunodeficiency virus; IGRA, interferon γ release assay; IQR, interquartile range; Xpert, GeneXpert M.tb/RIF assay.

^aData represent no. (%) of participants, unless otherwise specified.

Plasma Cytokine Levels in Progressors and Nonprogressors

Plasma levels of type I IFNs (IFN-γ, IL-2, and TNF-α), type II cytokines (interleukin 4, 5, 13, and 25, IL-10, IL-33, and IL-1Ra), proinflammatory cytokines (IL-6, IL-12p70, interleukin 15, 3, and 7, G-CSF, and GM-CSF), and anti-inflammatory cytokines were estimated in progressors and nonprogressors at baseline, month 4, and month 12. As depicted in Figure 2A, progressors exhibited significantly reduced levels of IFN-γ, IL-2, TNF-α, IL-1α, IL-1β, IL-17A, and GM-CSF, while they had elevated levels of IFN-α, IFN-β, IL-6, IL-12p70, IL-10, and IL-1Ra compared with nonprogressors at all time points, as shown in Figure 2B. Median cytokine levels in plasma samples from both groups are presented in Supplementary Table 1.

Figure 2.

Evaluation of plasma cytokine levels in progressors and nonprogressors, measured at baseline, at month 4 for progressors (n = 9) and nonprogressors (n = 30), and at month 12 for progressors (n = 4) and nonprogressors (n = 30). A, Plasma levels of interferon (IFN) γ, interleukin 2 (IL-2), tumor necrosis factor (TNF) α, and interleukin 17A, 1α, 1β, and 1 receptor antagonist (IL-17A, IL-1α, IL-1β, and IL-1Ra). B, Plasma levels of IFN-α, IFN-β, interleukin 10, 33, 6, and 12p70 (IL-10, IL-33, IL-6, and IL-12p70), and granulocyte-macrophage colony-stimulating factor (GM-CSF). Data in both A and B are presented as violin plots, with each circle representing an individual; P values were calculated using the Mann-Whitney U test, followed by Holm's multiple correction.

CC and CXC Chemokine Levels in Progressors and Nonprogressors

The study also assessed levels of CC (CCL2, CCL3, CCL4, CCL5, CCL11, CCL19, and CCL20) and CXC chemokines (CXCL1, CXCL2, CXCL8, CXCL10, and CX3CL1) in progressors and nonprogressors at baseline, month 4, and month 12. As demonstrated in Figure 3A, progressors had significantly higher levels of CCL2, CCL11, CXCL8, CXCL10, and CX3CL1 compared with nonprogressors. Other measured chemokines did not show any significant differences between the 2 groups. Median chemokine levels of plasma samples from both groups were shown in Supplementary Table 1.

Figure 3.

Levels of CC and CXC chemokine and growth factors, granzyme B, and programmed death ligand 1 (PDL-1) levels in progressors and nonprogressors, measured at baseline, at month 4 for progressors (n = 9) and nonprogressors (n = 30), and at month 12 for progressors (n = 4) and nonprogressors (n = 30. A, Plasma levels of CCL2, CCL11, CXCL8, CXCL10 and CX3CL1. B, Plasma levels of vascular endothelial growth factor (VEGF), granzyme B, and PDL-1. Data in both A and B are presented as violin plots, with each circle representing an individual; P values were calculated using the Mann-Whitney U test, followed by Holm's multiple correction.

Growth Factor Levels in Progressors and Nonprogressors

Furthermore, the plasma levels of various growth factors (VEGF, EGF, FGF-2, PDGF-AA, PDGF-AABB, transforming growth factor α, Flt3L, granzyme B, PDL-1, TRAIL, and CD40L) were assessed in progressors and nonprogressors at baseline, month 4, and month 12. As shown in Figure 3B, progressors exhibited significantly elevated levels of VEGF, granzyme B, and PDL-1, compared with nonprogressors at these time points.

Determination of Biomarkers Using ROC Analysis

Next, the potential of various parameters to serve as predictive biomarkers for the progression to active tuberculosis was evaluated using ROC analysis. Among the analyzed parameters (Figure 4A), IFN-γ (area under the ROC curve [AUC], 0.9121 [95% confidence interval 0.8356–0.9887]), IL-1Ra (0.9400 [0.8629–1.000]), GM-CSF (0.9200 [0.7930–.1.000]), CCL2 (0.9822 [0.9522–1.000]), and CXCL10 (0.9400 [0.8636–1.000]) demonstrated promise as predictive parameters with AUC values >90 and P values <0.0001. Sensitivity, specificity, P values, and AUC values for baseline, month 4, and month 12 are provided in Supplementary Tables 1, 2, and 3, respectively.

Figure 4.

Determination of biomarkers using receiver operating characteristic (ROC) analysis. A, ROC analysis was performed to assess the discriminatory power of plasma cytokines and chemokines in distinguishing progressors from nonprogressors. Sensitivity, specificity, and area under the ROC curve were estimated using interferon (IFN) γ, interleukin 1 receptor antagonist (IL-1Ra), granulocyte-macrophage colony-stimulating factor (GM-CSF), CCL2, and CXCL-10 at baseline, month 4, and month 12. Abbreviations: Interferon Gamma-induced Protein-10 (IP-10); Monocyte chemoattractant protein-1 (MCP-1). B, CombiROC model analysis identified a combination of cytokines and chemokines (GM-CSF + CXCL-10 + IL-1Ra) as most accurate in discriminating progressors from nonprogressors.

Plasma Signature of Cytokines and Chemokines as Precise Biomarkers for Active Tuberculosis Progression

Combinatorial analysis of multiple immune biomarkers was performed to identify ideal biomarker combinations among the tested circulating plasma cytokines and chemokines using the CombiROC method. As shown in Figure 4B, dual combinations of cytokines, such as GM-CSF + CXCL10, GM-CSF + IL-1Ra, and CXCL10 + IL-1Ra, as well as the triple combination of GM-CSF + CXCL10 + IL-1Ra, exhibited significant discriminatory power with high AUC values, sensitivity, and specificity in distinguishing between progressors and nonprogressors. The combinations of GM-CSF, CXCL10, and IL-1Ra exhibited remarkable predictive performance in distinguishing progressors from nonprogressors.

DISCUSSION

Identifying individuals with LTBI who are at the highest risk of progressing to active tuberculosis is essential for the effective implementation of preventive treatment strategies [20]. Current diagnostic tools for LTBI, including TST and IGRA, have limitations, particularly in immunocompromised individuals, in whom it is challenging to distinguish LTBI from active tuberculosis [21, 22]. Consequently, there is a critical need for sensitive biomarkers or host indicators that can differentiate LTBI individuals at risk of developing active disease. In the current study, we evaluated 45 systemic immunological characteristics to identify potential predictive markers of tuberculosis progression.

Several cytokines have emerged as pivotal players in the host response to Mycobacterium tuberculosis (Mtb) infection [23, 24]. T-helper 1 (Th1) cytokines, including IFN-γ, IL-2, and TNF-α, are central to antimycobacterial immune responses and granuloma formation [25, 26]. Our findings align with those of previous research [27–30] showing that individuals at risk of progressing to active tuberculosis had significantly reduced levels of these Th1 cytokines. The diminished Th1 response suggests that individuals with impaired Th1 cytokine responses are more susceptible to the development or progression of active tuberculosis disease [5–7, 9, 31] and that Th1 cytokines are immune correlates of tuberculosis disease and promising biomarkers of active tuberculosis [32–36]. In addition to Th1 cytokines, other cytokines, such as IL-1α, IL-1β, IL-6, and IL-12p70, were found to differ significantly between progressors and nonprogressors. Decreased levels of IL-1α and IL-1β, crucial for host defense against M.tb, as found in our study, have also been reported in human and animal studies, indicating their importance in tuberculosis immunity [37–41]. These findings are consistent with those of studies highlighting the role of IL-1Ra as a competitive inhibitor of IL-1α and IL-1β [40]. IL-1Ra may serve as a protective immunological and physical barrier, confining M.tb infection within granulomas [40, 42].

Type I IFNs, including IFN-α and IFN-β, have been implicated in tuberculosis pathogenesis, with elevated levels shown to negatively affect infection control [43, 44]. Our observation of increased levels of IFN-α and IFN-β in progressors corroborates these findings, further underscoring their potential role in tuberculosis progression. IL-17A is known to play a crucial role in the formation of mature granulomas in M.tb infection [45] and can therefore serve as a marker of disease severity and bacterial burden [6, 45]. Reduced IL-17 levels were observed in progressors, suggesting that compromised IL-17 cytokine responses may contribute to tuberculosis progression in these individuals.

Chemokines are pivotal players in the immune response to tuberculosis infection, particularly in the context of granuloma formation and maintenance [46]. Their precise regulation is essential to ensure an effective, controlled immune response that contains the infection [47]. However, an imbalance in chemokine expression can have detrimental effects on host defense mechanisms and may contribute to tuberculosis pathogenesis [5, 8, 10, 48–50]. The evaluation of chemokines revealed elevated levels of CCL2, CXCL8, CXCL10, and CX3CL1 in progressors. Dysregulated chemokines have been linked to the progression of active tuberculosis from LTBI, emphasizing their importance in granuloma formation immune cell recruitment, and tuberculosis pathogenesis.

The comparison of cytokine and chemokine profiles between progressors and nonprogressors provides further insights into the immunological factors associated with the progression of LTBI to active tuberculosis disease. The study by Daniel et al [30] demonstrated that individuals who progressed to active tuberculosis had significantly higher levels of certain cytokines and chemokines in QuantiFERON supernatants compared to nonprogressors. Our findings align with those of Daniel et al, which identified significantly higher levels of IFN-α, IFN-β, IL-6, GM-CSF, and IL-1β in progressors [30]; those author also observed higher IL-1β levels in QFT assays compared with plasma samples from patients with tuberculosis, likely reflecting the localized and antigen-specific nature of the immune response measured by QFT [30]. These observations collectively emphasize the complex interplay of cytokines and chemokines in the pathogenesis of tuberculosis. It suggests that an exaggerated proinflammatory response, coupled with an imbalance in regulatory cytokines such as IL-1Ra, may contribute to the progression of LTBI to active tuberculosis. In the context of growth factors, elevated levels of VEGF, granzyme B, and PDL-1 were observed in progressors. These findings may indicate a potential role for these growth factors in tuberculosis pathogenesis.

Our study has unveiled IL-1Ra, GM-CSF, and CXCL-10 as promising predictive markers for the progression of LTBI to active tuberculosis disease. The CombiROC analysis further strengthened the utility of these markers in combination, showcasing high sensitivity and specificity in discriminating between individuals who progress to tuberculosis disease (progressors) and those who do not (nonprogressors). While these findings provide valuable insights into potential biomarkers for tuberculosis progression, it is essential to acknowledge the study's limitations, including a relatively small sample size and variations in progression times among participants. One notable limitation is the absence of a correlation between our identified markers and sputum bacterial loads, which could have offered additional insights into the markers' relationship with disease severity. Incorporating this aspect in future studies may contribute to a more comprehensive understanding of the biomarkers' clinical implications. In addition, the small sample size may influence the generalizability of our findings, emphasizing the need for further validation in larger cohorts.

Despite these limitations, our study significantly adds to the accumulating evidence supporting the use of plasma biomarkers for predicting the progression of LTBI to active tuberculosis disease. The identification of IL-1Ra, GM-CSF, and CXCL10 as potential biomarkers holds promise for enhancing tuberculosis diagnosis and treatment strategies. The synergistic combination of these markers, as indicated by CombiROC analysis, underscores their potential as a robust tool in distinguishing individuals at risk of tuberculosis progression. Looking ahead, future research endeavors should prioritize the validation of these biomarkers in more extensive cohorts, encompassing diverse populations and geographic regions. The assessment of their utility in real-world clinical settings is crucial for establishing their practicality and effectiveness in informing early intervention strategies. Furthermore, prospective studies are warranted to longitudinally evaluate the identified markers and their role in reducing tuberculosis transmission rates. Ultimately, the potential of these biomarkers to contribute to tuberculosis vaccine development should be explored, aligning with global efforts to curb the prevalence of this infectious disease.

Supplementary Data

Supplementary materials are available at The Journal of Infectious Diseases online (http://jid.oxfordjournals.org/). Supplementary materials consist of data provided by the author that are published to benefit the reader. The posted materials are not copyedited. The contents of all supplementary data are the sole responsibility of the authors. Questions or messages regarding errors should be addressed to the author.

Notes

Author contributions. Study design: A. R. and S. B. Conduct of experiments: A. R., E. A. D., and B. D. Data acquisition: A. R., E. A. D., and B. D. Data analysis: A. R., E. A. D., and K. T. Sample collection and processing and data collection: S. P., B. B., and A. S. Contribution of reagents: P. C., A. G., and S. B. Responsibility for enrollment of the participants and contribution to acquisition and interpretation of clinical data: P. C., S. G., A. S., V. K., R. K., M. P., S. V. B. Y. S., V. M., A. G., and L. E. H. Resources, validation and supervision: A. R., V. M., A. G., L. E. H., and S. B. Coordination of data management: A. R. and A. S. Access to and verification of the data: S. B. Writing of the manuscript: A. R. Revision of subsequent manuscript drafts: A. R. and S. B. All authors read and approved the final manuscript.

Disclaimer. The contents of this publication are solely the responsibility of the authors and do not represent the official views of the Government of India’s Department of Biotechnology, the ICMR, the National Institutes of Health, or CRDF Global.

Data availability. The data generated for this study are available within the article and in the Supplementary materials.

Financial support. This work was supported by the Government of India's Department of Biotechnology; the Indian Council of Medical Research (ICMR); the Division of Intramural Research (support to S. B.), National Institute of Allergy and Infectious Diseases, National Institutes of Health; Office of AIDS Research (OAR), and distributed in part by CRDF Global (grant USB1-31149-XX-13 to P. C. and A. G. and RePORT India Consortium supplemental funding [grant OISE-17-62911-1 to L. E. H.]).

Potential conflicts of interest . All authors: No reported conflicts. All authors have submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Conflicts that the editors consider relevant to the content of the manuscript have been disclosed.