Immunological non-response and low hemoglobin levels are predictors of incident tuberculosis among HIV-infected individuals on Truvada-based therapy in Botswana
Immunological non-response and low hemoglobin levels are predictors of incident tuberculosis among HIV-infected individuals on Truvada-based therapy in Botswana
Lucy Mupfumi 1
Sikhulile Moyo 0 1
Prisca K. Thami 1
Motswedi Anderson 1
Tuelo Mogashoa 1
Thato Iketleng 1
Joseph Makhema 0 1
Ric Marlink 0 1
Max Essex 0 1
Rosemary M. Musonda 0 1
Simani Gaseitsiwe 0 1
Editor: Ann Rawkins
Public Health England
0 Department of Immunology and Infectious Diseases, Harvard TH Chan School of Public Health, Boston, Massachusetts, United States of America, 4 Department of Statistics, University of Botswana , Gaborone , Botswana , 5 Department of Biological Sciences, University of Botswana , Gaborone , Botswana , 6 College of Health Sciences, University of KwaZulu-Natal, Durban, Republic of South Africa, 7 Rutgers Global Health Institute, Robert Wood Johnson Medical School, Rutgers University , New Brunswick, New Jersey , United States of America
1 Botswana Harvard AIDS Institute Partnership, Gaborone, Botswana, 2 Department of Medical Laboratory Sciences, School of Allied Health Professionals, University of Botswana , Gaborone , Botswana
2008 and 2011. We used Cox proportional hazards regressions to determine predictors of
Data Availability Statement: The subset of data
used for these analyses has been cleaned and
extracted from the larger, complete database of the
ªBomolemo studyº. This dataset is ready for
sharing with interested researchers after going
through the appropriate, and mandatory, regulatory
procedures in accordance with Botswana National
guidelines. Requests for data can be made by
applying to the Botswana Ministry of Health and
Wellness Human Research Development
Committee (HRDC). IRB contact: Seeletso
Of 300 participants enrolled, 45 (15%) had a diagnosis of TB at baseline. During 428
person-years (py) of follow-up, the incidence rate of TB was 3.04/100py (95% CI, 1.69±5.06),
with 60% of the cases occurring within 3 months of ART initiation. Incident cases had low
baseline CD4+ T cell counts (153cells/mm3 [Q1, Q3: 82, 242]; p = 0.69) and hemoglobin
levels (9.2g/dl [Q1, Q3: 8.5,10.1]; p<0.01). In univariate analysis, low BMI (HR = 0.73; 95%
CI 0.58±0.91; p = 0.01) and hemoglobin levels <8 g/dl (HR = 10.84; 95%CI: 2.99±40.06;
p<0.01) were risk factors for TB. Time to incident TB diagnosis was significantly reduced in
Funding: This work was supported through the
Sub-Saharan African Network for TB/HIV Research
Excellence (SANTHE), a DELTAS Africa Initiative
[grant # DEL-15-006]. The DELTAS Africa Initiative
is an independent funding scheme of the African
Academy of Sciences (AAS)'s Alliance for
Accelerating Excellence in Science in Africa (AESA)
and supported by the New Partnership for Africa's
Development Planning and Coordinating Agency
(NEPAD Agency) with funding from the Wellcome
Trust [grant # 107752/Z/15/Z] and the UK
government. The views expressed in this
publication are those of the author(s) and not
necessarily those of AAS, NEPAD Agency,
Wellcome Trust or the UK government. LM, KM,
PT, TM& SG received funding from SANTHE. The
Bomolemo study was funded by ACHAP, a
country-led, public±private development
partnership between the Government of Botswana,
the Bill & Melinda Gates Foundation, and the MSD/
Merck Company Foundation. SM was supported by
the Fogarty International Center and National
Institute of Mental Health, of the National Institutes
of Health under Award Number D43 TW010543.
The funders had no role in study design, data
collection and analysis, decision to publish, or
preparation of the manuscript.
Competing interests: R. Marlink indirectly received
funding from MSD/Merck through the ACHAP
grant. This does not alter our adherence to Plos
ONE policies on sharing data and materials.
patients with poor immunological recovery (p = 0.04). There was no association between
baseline viral load and risk of TB (HR = 1.75; 95%CI: 0.70±4.37).
Low hemoglobin levels prior to initiation of ART are significant predictors of incident
tuberculosis. Therefore, there is potential utility of iron biomarkers to identify patients at risk of TB
prior to initiation on ART. Furthermore, additional strategies are required for patients with
poor immunological recovery to reduce excess risk of TB while on ART.
Tuberculosis (TB) is the leading cause of mortality by an infectious disease, ranking above
HIV in current estimates [
]. In 2015, an estimated 1.4 million TB-related deaths occurred, a
third of these among HIV infected individuals. In sub-Saharan Africa, a region that bears the
brunt of these twin epidemics, 11% of incident TB cases occur among HIV-infected people [
]. Strategies recommended by the WHO to reduce the TB burden include intensified case
finding, isoniazid preventative therapy, and early ART initiation for TB-HIV co-infected
The survival benefits of initiation of combination antiretroviral therapy (cART) in
coinfected patients, particularly those with low CD4+ T-cell counts, have been well documented
]. However, although cART has been reported to reduce the risk of developing active TB
among HIV infected persons by 67% , co-infected individuals have a heightened risk of
]. Furthermore, HIV infected individuals on cART remain at a higher risk of
developing active TB than HIV uninfected individuals [9±11]. Studies conducted in both low
and high income countries have determined that the risk of active TB does not return to
background rates despite long term ART [2, 3, 11±13].
The mechanisms driving the heightened risk of TB among HIV infected individuals are not
well understood. The most recognized immune defect caused by HIV is the absolute reduction
in CD4+ T-cells, the mainstay of the immune response to TB [
]. This accounts for the
increased risk of TB in HIV-infected persons, which is strongly associated with the progressive
loss of CD4+ T-cells . However, the fact that the risk of TB remains high even in early HIV
infection or with immune restoration while on cART shows that HIV infection confers
qualitative changes to the functionality of the CD4+ T-cells and other aspects of the immune
response to TB [
Other well recognized factors accounting for this elevated risk include a low hemoglobin,
low BMI, and increasing age [16, 19±22]. Low BMI is a strong independent predictor of
mortality among HIV infected individuals even in the context of ART initiation [
], with higher
BMI levels associated with a protective effect on both mortality and incident TB estimates [
]. HIV-infected individuals are therefore likely to benefit from micronutrient
supplementation, although there have been conflicting reports of their effect on HIV-related co-morbidities
or mortality . Similarly, determinants for TB among HIV-infected patients vary across
Understanding the context-specific incidence and predictors of TB during cART remain
important keys to designing effective interventions to mitigate the ravenous effects of the
syndemic. Furthermore, most studies describing burden of TB in ART programs have been in
patients on zidovudine-containing regimens, there has only been one study conducted in
2 / 13
patients on a Tenofovir-based ART regimen [
]. In Botswana, currently listed as an HIV-TB
high burden country[
], high mortality rates have been reported among co-infected patients
despite provision of ART and anti-tuberculosis therapy (ATT) [
]. It is therefore critical to
determine factors associated with risk of incident TB to allow early identification of patients
that will benefit from isoniazid preventative therapy. We conducted a retrospective analysis of
data collected during an observational study evaluating efficacy of a Truvada-based cART
backbone in Botswana to determine the TB incidence and the associated predictive factors.
Materials and methods
Study design and participants
We conducted a retrospective analysis of data from 300 adults enrolled into an observational
study, evaluating the efficacy and tolerability of a Truvada-based regimen in HIV-1C infected
adultsÐªBomolemo studyºÐconducted in Gaborone between November 2008 and July 2011
]. In 2008, Botswana adopted Tenofovir plus Emtricitabine (Truvada™) combined with
either Efavirenz or Nevirapine as its first line ART regimen. The ªBomolemo study” was
designed to demonstrate the tolerability, virologic and immunologic response of a
Truvadacontaining regimen for HIV-1 C infected adults in Botswana. Participants were HIV infected,
ART-naïve and aged 18 years and older. Additional eligibility criteria included the presence of
an AIDS defining condition or a CD4+ T-cell count <250cells/mm3, consistent with WHO
guidelines at the time. Female participants were asked to provide a urine specimen for
pregnancy testing and were excluded if they were pregnant or had received single dose Nevirapine
as PMTCT within the 6 months preceding enrollment. After study entry and ART initiation,
participants were scheduled for evaluations at 1 month and then every 3 months until the final
study visit at week 96. At these visits, physical examination and medical history was taken
in addition to laboratory tests for CD4, viral load, hematology, and chemistry. This study
received ethical approval from the Botswana Ministry of Health's Health Research
Development Committee (PPME-13/18/1) and the Harvard T.H. Chan School of Public Health IRB
(16470±02). All participants provided a written informed consent including use of their
HIV and chronic hepatitis B viral (HBV) infection was diagnosed by double ELISA and surface
antigen testing respectively. Viral load samples were processed on the Roche Amplicor HIV-1
monitor and CD4 T-cell measurements on the BD FACSCalibur at the Botswana Harvard
HIV reference Laboratory (BHHRL). For this analysis, we defined immunologic non-response
as CD4+ T-cell count gains of less than 20% within 6 months of cART initiation [
virologic suppression as viral load less than 400copies/ml after 6months on cART.
Tuberculosis screening and diagnoses
Tuberculosis was diagnosed and treated through the national programme per the Botswana
National TB treatment guidelines. Briefly, tuberculosis diagnosis was based on either a positive
sputum AFB or culture result, or abnormal chest radiology. Patients on TB treatment with
neither abnormal chest x-ray nor microbiological confirmation were classified as ªclinical TBº.
Information on clinical diagnoses was collected at each study visit and documented on a
case report form. For this analysis, any patient reported as having tuberculosis with a
documented diagnosis date was recorded as a case. We also defined prevalent TB cases as those
3 / 13
who had a TB diagnosis at baseline and incident TB cases as those diagnosed with TB after
Person-time at risk was accrued from the date of cART initiation until death or study closeout
or, in the case of those who developed incident TB, date of first recorded TB diagnosis. Incident
TB rate was calculated as the number of new cases of TB divided by total person-years (PY) at
risk and expressed per 100py. Kaplan-Meier estimator of the survivor function was used to
generate a TB-free survival curve and to determine the impact of poor immunological response on
TB incidence, adjusting for gender. Univariate and multivariate Cox proportional hazard
models were generated adjusting for CD4, gender and age to evaluate risk factors for incident TB
during ART. Variables with p-values < 0.05 were included in the multivariate model. All
statistical analyses were conducted in Stata version 14.2 (College Station, Texas, USA).
The median age of the cohort was 36 years (Q1, Q3: 32, 42), and 64% were females [Table 1].
Overall, participants had advanced immunodeficiency, with median CD4 counts of 168 cells/
mm3 (Q1, Q3; 86, 232) and log HIV viral load of 5.1cps/ml (Q1, Q3: 4.6, 5.6). Participants with
prevalent TB had lower hemoglobin levels (11.2g/dl; Q1, Q3: 8.9, 13), although this was not
significantly different from the rest of the participants.
Forty-five (15%) participants had TB at baseline (prevalent TB) and were more likely to be
male (p<0.01), have lower BMI (p<0.01), and higher HIV viral loads (p<0.01) [Table 1].
These participants were excluded from further analysis [Fig 1]. The remaining participants
contributed 428 person-years of follow-up. Incident TB was diagnosed in 13 of 254 [5%, IR
36 (32, 42)
21.6 (19.1, 25.1)
11.5 (9.9, 12.9)
168 (86, 232)
5.13 (4.6, 5.6)
Fig 1. Participant flow diagram of individuals enrolled in the incident TB cohort analysis. Forty-six individuals were excluded from further analysis:
45 had TB at baseline (prevalent TB) and one did not have a recorded date of diagnosis.
3.04/100py (95% CI, 1.69±5.06)] participants within a median of 41 days (Q1, Q3:28, 78) from
enrolment. Bacteriological confirmation was available in 6 (46%) of the incident cases. The
remaining cases were placed on treatment based on abnormal chest radiology (23%) or clinical
Incident TB and CD4 counts
Throughout the follow-up period, males had significantly higher CD4+T-cell counts than
females (p<0.01), although there were proportionally more males with incident TB (Table 2).
The median CD4+T-cell count for the incident TB cases was 153 (Q1, Q3: 82, 242) cells/ mm3
(Table 2). The risk of developing incident TB was not associated with immunologic
nonresponse [HR 2.99 (95% CI 0.98±9.15), p = 0.06] (data not shown), however, immunologic
non-responders had a shorter time to diagnosis [log rank p = 0.02, Fig 2].
Risk factors for incident TB
Incident cases had significantly lower baseline BMI and hemoglobin levels compared to
TB cases (Table 2). In univariate analysis, the risk of incident TB was higher among participants
5 / 13
34 (32, 37)
Fig 2. Kaplan Meier plot of time to incident TB in patients with poor immunological response. Participants who failed to achieve a CD4+T-cell
count increase of at least 20% of the baseline counts by 6months of follow-up were classified as poor immunological responders.
PLOS ONE | https://doi.org/10.1371/journal.pone.0192030
6 / 13
with low BMI [HR 0.73 (95% CI 0.58±0.91) p<0.01] and low baseline hemoglobin levels [ 8g/
dl; HR 10.84 (95% CI 2.93±40.06) p<0.01] (Table 3). In additional analysis, participants with
CD4+T-cell counts below 250cells/mm3 by one year of follow-up were at significant risk of
incident TB [HR 5.48 (95% CI 1.66±18.2) p<0.01 (results not shown)].
In the multivariate model, only baseline hemoglobin levels were significantly associated
with incident TB [adjusted HR 6.88 (95% CI: 1.78±26.58) p = 0.01; Table 3]. We also found a
significant interaction between gender and BMI [adjusted HR 0.61 (95% CI: 0.38±0.96)
p = 0.03] (data not shown). However, there was no interaction between gender and
hemoglobin level [adjusted HR 1.08 (95% CI 0.96±1.22) p = 0.19] (data not shown).
We have shown in this study a TB incidence rate of 3.04/100py over a 2-year follow-up period
of patients on a Truvada-based cART regimen. As is the case with ART programs in
sub-Saharan Africa, 70% of the incident TB cases occurred within the first three months of cART
initiation. Additionally, there was a significant association between anemia and incident TB,
highlighting the need to identify biomarkers that will allow timing of appropriate
interventions. Patients with poor immunologic recovery by one year on cART and those with low CD4
counts had an increased risk of and shorter time to incident TB disease.
The reduction in TB risk attributable to cART varies depending on the background risk of
]. Studies have shown that cART reduces the individual risk of TB disease
among HIV-infected persons by between 60 and 90% irrespective of CD4 count [
20, 31, 32
However, this reduction is time dependent, decreasing from 3.6/100py in the first year to 1.0/
100py by the fifth year on cART [
]. Furthermore, by year three on treatment, patients still
have a 5±10 times higher risk of TB compared to HIV-uninfected [
]. A retrospective
analysis of data from Wales confirmed these findings, concluding that the risk of TB in HIV
infected patients does not reach background rates in black Africans . This is likely due to
the fact that long-term restoration of immune phenotype and function is limited despite a long
duration on cART [
], particularly in patients with low CD4 counts.
In this study, patients with advanced immunodeficiency were recruited, reflecting the
cART eligibility guidelines of the Botswana HIV Treatment Programme at the time. The risk
of development of active TB is associated with lower nadir CD4 counts and those with
suboptimal immunological responses have an increased risk of developing active TB. As we
demonstrated, it is therefore time-updated and not baseline CD4, that is the strongest predictor of
incident TB. In our study, patients who failed to achieve a 20% gain in baseline CD4 counts by
6 months on cART had a 2-fold higher risk of TB. Other studies have reported that patients
with the lowest baseline CD4 counts remain at increased risk of TB for a longer period of time
7 / 13
compared to those with high CD4 counts [
]. Furthermore, updated CD4 counts closer to the
time of diagnosis have also been reported to be stronger predictors of incident TB than
baseline CD4 counts [
Immunological non-recovery is a known predictor of the immune reconstitution syndrome
(IRIS) that complicates management of patients on cART [35±38]. Patients who fail to achieve
at least a 20% increase in CD4 counts on cART are considered to be immunological
]. The reasons for this non-response in the context of viral suppression are not
well understood but may be linked to a lower CD4 nadir [
] or the elevated levels of
T-regulatory cells in circulating blood [
]. Previous work in our setting showed that immunological
non-recovery in TB/HIV co-infected patients was a significant risk factor for TB-IRIS and
early mortality and this was associated with elevated levels of the pro-inflammatory cytokine,
]. Most of the patients analyzed in this study had low CD4 counts prior to cART
initiation and this may account for the high incidence rate reported in this study, although our
estimate is consistent with what has been reported in similar high burden settings [
contrast, studies from other African countries have reported lower incidence rates in patients
on cART [
11, 41, 42
]. This could possibly be a reflection of the lower background rates of
prevalent TB or inadequate screening or documentation of TB cases in these settings .
Consistent with reports from similar high burden settings [
3, 16, 20, 43, 44
], more than
half of the incident cases occurred within 6 months of cART initiation. This possibly
demonstrates the difficulty in screening patients for TB and highlights the need for biomarkers that
can identify patients at increased risk of TB at the time of cART initiation. The TB symptom
screen algorithm proposed by Cain and colleagues [
] may be less sensitive and specific
among individuals taking cART [
]. Cases were identified within a short period of cART
initiation, thus likely represent subclinical disease [
16, 20, 44
] or TB-IRIS [
]. The fact that
this finding has also been demonstrated in a study that recruited patients with CD4 counts
] shows that enhanced diagnostics and intensified case finding, in addition
to early cART initiation, will be required to curb the challenge of incident TB in ART
programs in Africa.
An interesting finding from this study is the heightened risk of incident TB in patients with
severe anemia, defined as a hemoglobin level less than 8g/dl [
]. Anemia is a common
comorbidity reported at TB diagnosis in 32±86% of patients [
]. Patients in our study were on a
Tenofovir-based regimen, which does not induce anemia as is the case with zidovudine. This
suggests there is another mechanism driving the intricate relationship between anemia and
TB. It has been postulated that severe anemia in patients with undiagnosed TB is likely due to
the pro-inflammatory cytokine activation that induces an anemia of chronic disease [
IL-6 upregulates the transcription of hepcidin, the iron regulatory protein, which leads to
reduction in iron absorption and sequestration of iron by macrophages and enterocytes ,
limiting iron delivery to erythroblasts thus causing a functional anemia. This increased iron
storage in macrophages promotes replication of bacilli thus markers of iron homeostasis may
be early markers of risk of progression from latency to active TB. Studies have also shown that
time-updated anemia severity is an independent predictor of mortality in patients on ART [
], which may suggest a role for iron biomarkers in screening strategies for TB prior to cART
initiation. Iron biomarkers also have the potential to contribute to the development of risk
profiles for the progression from latent TB infection to active disease , a current area of
Nutritional status as measured by BMI, was an independent predictor of incident TB in this
cohort, in concordance with reported studies from the region [
11, 20, 21, 43
]. There is a
bidirectional relationship between weight loss and TB. People with TB most often have a loss of
appetite that results in weight loss, a main symptom of TB disease . Individuals with
8 / 13
increased BMI may have higher daily protein and energy intake which could result in a more
robust immune function and drive the reduction in both incident TB and mortality [
category change has also been reported as a marker of favorable treatment outcomes [
HIV induces early aging of the immune system, which likely contributes to reactivation
of latent TB and therefore heightened risk of incident TB in older ages [
]. Other studies
have also reported an association between gender and incident TB [
], although we
did not observe such an effect. We observed significant interaction between gender
and BMI, which possibly masked the true effect of BMI in our multivariate model. This
is consistent with findings from a large cohort study of patients on ART in Nigeria that
showed that gender is an effect modifier of the association between BMI and anemia with
TB, suggesting that males may benefit from nutritional interventions to boost their
immunity to TB .
Incident TB may have been overestimated due to most cases appearing within the first
6 months of ART. Furthermore, some cases may have died before a TB diagnosis which
may bias our estimates as we did not perform a competing risk analysis. However, our
findings are similar to estimates reported from similar high-burden settings[
10, 20, 43
combination of both a high rate of prevalent and incident TB in this study highlights the
magnitude of the HIV-associated TB epidemic in ART programs in low-resource setting.
While ART initiation at high CD4 counts is expected to reduce incident TB rates[
], there is need to strengthen TB screening algorithms and combine with biomarkers to
ensure that patients at increased risk of TB are identified and appropriate interventions
The combination of cART and isoniazid preventive therapy would be useful in reducing
the burden of TB in ART programs in Africa[
]. However, the risk of infection post-IPT
remains considerably higher in high incidence compared to low incidence settings [
when IPT is extended to 36 months [
]. While it is not clear if these incident cases post-IPT
represent a failure to sterilize latent infection at the time of IPT initiation or reinfection, it is
generally accepted that these are a function of TB endemicity [
]. Nonetheless, there is no
disputing the survival benefit of combined IPT and ART in high incidence settings. Multivariate
algorithms that rely on accurate classification of latent infection and better treatment regimens
for IPT are therefore urgently needed[
Our results should be interpreted in light of the strength and weaknesses of the study
design. The main weakness of the study is the retrospective study design that has inherent
weaknesses. Tuberculosis was not actively investigated in the main study and relied on
clinical investigation as per the standard of care through the routine health system. Consequently,
over half the incident cases had no laboratory confirmation, which may have introduced
misclassification bias and overestimation of our incident cases. However, most TB cases are
diagnosed empirically in Africa even in settings provided with the GeneXpert test [
Secondly, this study was conducted at a time when ART guidelines recommended initiation
at CD4 count less than 250cells/mm3 so the findings of this analysis may not be generalizable
to current guidelines. However, studies conducted in patients with CD4 count above
350cells/ mm3 suggest that incident TB is still a problem  thus pointing to the fact that
multiple interventions will be required in order to mitigate the burden of HIV-associated
In conclusion, we have reported a high rate of incident TB among patients on ART in
Botswana and highlight a possible role for iron biomarkers in identifying patients at risk of
TB. Longitudinal studies are warranted in the HIV test and treat era to define true estimates of
incident TB and determine the predictive value of iron biomarkers.
9 / 13
We are grateful to Dr Nicola Zetola for providing critical edits to this manuscript. We wish to
thank the participants of the ªBomolemoº study as well as the clinic and laboratory staff at
Botswana Harvard AIDS Institute Partnership.
Conceptualization: Lucy Mupfumi, Sikhulile Moyo, Rosemary M. Musonda, Simani
Data curation: Joseph Makhema, Ric Marlink.
Formal analysis: Lucy Mupfumi, Sikhulile Moyo, Kesaobaka Molebatsi, Prisca K. Thami.
Supervision: Sikhulile Moyo, Ishmael Kasvosve, Rosemary M. Musonda, Simani Gaseitsiwe.
Writing ± original draft: Lucy Mupfumi.
Writing ± review & editing: Prisca K. Thami, Motswedi Anderson, Tuelo Mogashoa, Thato
Iketleng, Joseph Makhema, Ric Marlink, Ishmael Kasvosve, Max Essex, Rosemary M.
Musonda, Simani Gaseitsiwe.
10 / 13
11 / 13
12 / 13
therapy in South Africa: a cohort study. BMC medicine. 2015; 13(1):70.
https://doi.org/10.1186/s12916015-0320-9 PMID: 25889688
1. http://www.who.int/tb/publications/global_report/en/. Global TB report2015 Accessed [27 October 2016 ].
2. Lawn SD , Wood R . Tuberculosis control in South AfricaÐwill HAART help? South African medical journal = Suid-Afrikaanse tydskrif vir geneeskunde . 2006 ; 96 ( 6 ): 502 ± 4 . Epub 2006/07/15. PMID: 16841129 .
3. Stephen D. Lawn LM , Linda-Gail Bekker , Robin Wood . Burden of tuberculosis in an antiretroviral treatment programme in sub-Saharan Africa: impact on treatment outcomes and implications for tuberculosis control . AIDS . 2006 ; 20 : 1605 ± 12 . https://doi.org/10.1097/01.aids. 0000238406 .93249. cd PMID : 16868441
4. Abdool Karim SS , Naidoo K , Grobler A , Padayatchi N , Baxter C , Gray AL , et al. Timing of Initiation of Antiretroviral Drugs during Tuberculosis Therapy . The New England journal of medicine . 2010 ; 362 : 697 ± 706 . https://doi.org/10.1056/NEJMoa0905848 PMID: 20181971
5. Karim SSA , Karim QA . Antiretroviral prophylaxis: a defining moment in HIV control . The Lancet . 2011 ; 378 ( 9809 ):e23± e5 .
6. Lawn SD , Bekker LG , Calmy A , Wood R . Monitoring of antiretroviral therapy in low-resource settings . Lancet . 2008 ; 372 ( 9635 ): 287 ± 8; author reply 9 . Epub 2008/07/29. https://doi.org/10.1016/S0140- 6736 ( 08 ) 61101 - 0 PMID: 18657697 .
7. Lawn SD , Harries AD , Anglaret X , Myer L , Wood R . Early mortality among adults accessing antiretroviral treatment programmes in sub-Saharan Africa . AIDS . 2008 ; 22 ( 15 ): 1897 ± 908 . Epub 2008/09/12. https://doi.org/10.1097/QAD.0b013e32830007cd PMID: 18784453 .
8. Lawn SD , Harries AD . Reducing tuberculosis-associated early mortality in antiretroviral treatment programmes in sub-Saharan Africa . AIDS . 2011 ; 25 ( 12 ): 1554 ± 5; author reply 6 . Epub 2011/07/13. https:// doi.org/10.1097/QAD.0b013e328348fb61 PMID: 21747238 .
9. Wood R , Middelkoop K , Myer L , Grant AD , Whitelaw A , Lawn SD , et al. Undiagnosed tuberculosis in a community with high HIV prevalence: implications for tuberculosis control . American journal of respiratory and critical care medicine . 2007 ; 175 . https://doi.org/10.1164/rccm.200606-759OC PMID: 16973982
10. Gupta A , Wood R , Kaplan R , Bekker LG , Lawn SD . Tuberculosis Incidence Rates during 8 Years of Follow-Up of an Antiretroviral Treatment Cohort in South Africa: Comparison with Rates in the Community . PloS one . 2012 ; 7 ( 3 ): e34156 . Epub 2012 /04/06. https://doi.org/10.1371/journal.pone. 0034156 PMID: 22479548
11. Chang CA , Meloni ST , Eisen G , Chaplin B , Akande P , Okonkwo P , et al. Tuberculosis Incidence and Risk Factors Among Human Immunodeficiency Virus (HIV)-Infected Adults Receiving Antiretroviral Therapy in a Large HIV Program in Nigeria . Open Forum Infect Dis . 2015 ; 2 ( 4 ): ofv154 . Epub 2015 /11/ 28. https://doi.org/10.1093/ofid/ofv154 PMID: 26613097
12. Gupta RK , Rice B , Brown AE , Thomas HL , Zenner D , Anderson L , et al. Does antiretroviral therapy reduce HIV-associated tuberculosis incidence to background rates? A national observational cohort study from England, Wales, and Northern Ireland . The Lancet HIV . 2015 ; 2(6):e243±e51 . https://doi. org/10.1016/S2352- 3018 ( 15 ) 00063 - 6 PMID: 26423197
13. Lawn SD , Myer L , Wood R . Efficacy of antiretroviral therapy in resource-poor settings: are outcomes comparable to those in the developed world? Clinical infectious diseases: an official publication of the Infectious Diseases Society of America . 2005 ; 41 ( 11 ): 1683 ± 4; author reply 4 . Epub 2005/11/04. https:// doi.org/10.1086/498030 PMID: 16267745 .
14. Diedrich CR , O'Hern J , Gutierrez MG , Allie N , Papier P , Meintjes G , et al. Relationship Between HIV Coinfection, Interleukin 10 Production, and Mycobacterium tuberculosis in Human Lymph Node Granulomas . The Journal of infectious diseases . 2016 ; 214 ( 9 ): 1309 ± 18 . https://doi.org/10.1093/infdis/jiw313 PMID: 27462092
15. Diedrich CR , Flynn JL . HIV-1/Mycobacterium tuberculosis Coinfection Immunology: How Does HIV-1 Exacerbate Tuberculosis? Infection and immunity . 2011 ; 79 ( 4 ): 1407 ± 17 . https://doi.org/10.1128/IAI. 01126-10 PMID: 21245275
16. Kufa T , Chihota V , Mngomezulu V , Charalambous S , Verver S , Churchyard G , et al. The incidence of tuberculosis among hiv-positive individuals with high CD4 counts: implications for policy . BMC infectious diseases . 2016 ; 16 : 266 . https://doi.org/10.1186/s12879-016 -1598-8 PMID: 27286814
17. Thobakgale C , Naidoo K , McKinnon LR , Werner L , Samsunder N , Karim SA , et al. Interleukin 1-Beta (IL-1beta) Production by Innate Cells Following TLR Stimulation Correlates With TB Recurrence in ART-Treated HIV-Infected Patients . J Acquir Immune Defic Syndr . 2017 ; 74 ( 2 ): 213 ± 20 . Epub 2016/09/ 23. PMID: 27654812
18. Ahmed A , Rakshit S , Vyakarnam A . HIV-TB co-infection: mechanisms that drive reactivation of Mycobacterium tuberculosis in HIV infection . Oral diseases . 2016 ; 22 Suppl 1 : 53 ± 60 . https://doi.org/10.1111/ odi.12390 PMID: 27109273 .
19. Liu E , Makubi A , Drain P , Spiegelman D , Sando D , Li N , et al. Tuberculosis incidence rate and risk factors among HIV-infected adults with access to antiretroviral therapy in Tanzania . AIDS (London, England) . 2015 ; 29 ( 11 ): 1391 ± 9 . PMID: 26091295
20. Lawn SD , Badri M , Wood R . Tuberculosis among HIV-infected patients receiving HAART: long term incidence and risk factors in a South African cohort . AIDS . 2005 ; 19 ( 18 ): 2109 ± 16 . Epub 2005/11/15. PMID: 16284460 .
21. Brennan AT , Bonawitz R , Schnippel K , Berhanu R , Maskew M , Long L , et al. Incident tuberculosis in HIV-positive children, adolescents and adults on antiretroviral therapy in South Africa . The International Journal of Tuberculosis and Lung Disease . 2016 ; 20 ( 8 ): 1040 ±5. https://doi.org/10.5588/ijtld.15.0488 PMID: 27393537
22. Minchella PA , Donkor S , Owolabi O , Sutherland JS , McDermid JM . Complex Anemia in Tuberculosis: The Need to Consider Causes and Timing When Designing Interventions . Clinical Infectious Diseases . 2015 ; 60 ( 5 ): 764 ± 72 . https://doi.org/10.1093/cid/ciu945 PMID: 25428413
23. Hanrahan CF , Golub JE , Mohapi L , Tshabangu N , Modisenyane T , Chaisson RE , et al. Body mass index and risk of tuberculosis and death . AIDS (London, England) . 2010 ; 24 ( 10 ): 1501 ±8. https://doi.org/ 10.1097/QAD.0b013e32833a2a4a PMID: 20505496
24. Benova L , Fielding K , Greig J , Nyang'wa B-T , Casas EC , da Fonseca MS , et al. Association of BMI Category Change with TB Treatment Mortality in HIV-Positive Smear-Negative and Extrapulmonary TB Patients in Myanmar and Zimbabwe . PloS one . 2012 ; 7 ( 4 ):e35948. https://doi.org/10.1371/journal. pone. 0035948 PMID: 22545150
25. Visser ME , Durao S , Sinclair D , Irlam JH , Siegfried N. Micronutrient supplementation in adults with HIV infection . Cochrane Database Syst Rev . 2017 ; 5:Cd003650 . Epub 2017/05/19. https://doi.org/10.1002/ 14651858.CD003650. pub4 PMID: 28518221
26. Alemu YM , Awoke W , Wilder-Smith A . Determinants for tuberculosis in HIV-infected adults in Northwest Ethiopia: a multicentre case±control study . BMJ open . 2016 ; 6 ( 4 ). PMID: 27084271
27. Bisson GP , Zetola N , Collman RG . Persistent high mortality in advanced HIV/TB despite appropriate antiretroviral and antitubercular therapy: an emerging challenge . Current HIV/AIDS reports . 2015 ; 12 ( 1 ): 107 ± 16 . Epub 2015/03/17. https://doi.org/10.1007/s11904-015-0256 -x PMID : 25772785
28. Anderson M , Gaseitsiwe S , Moyo S , Thami KP , Mohammed T , Setlhare D , et al. Slow CD4+ T-Cell Recovery in Human Immunodeficiency Virus/Hepatitis B Virus-Coinfected Patients Initiating TruvadaBased Combination Antiretroviral Therapy in Botswana . Open Forum Infect Dis . 2016 ; 3(3):ofw140 . PMID: 27800524
29. Gaardbo JC , Hartling HJ , Gerstoft J , Nielsen SD . Incomplete immune recovery in HIV infection: mechanisms, relevance for clinical care, and possible solutions . Clinical & developmental immunology . 2012 ; 2012 : 670957 . Epub 2012/04/05. https://doi.org/10.1155/ 2012 /670957 PMID: 22474480
30. del Amo J , Moreno S , Bucher HC , Furrer H , Logan R , Sterne J , et al. Impact of antiretroviral therapy on tuberculosis incidence among HIV-positive patients in high-income countries. Clinical infectious diseases: an official publication of the Infectious Diseases Society of America . 2012 ; 54 ( 9 ): 1364 ± 72 . Epub 2012/03/31. https://doi.org/10.1093/cid/cis203 PMID: 22460971
31. Middelkoop K , Bekker L-G , Myer L , Johnson LF , Kloos M , Morrow C , et al. Antiretroviral therapy and TB notification rates in a high HIV prevalence South African community . Journal of acquired immune deficiency syndromes ( 1999 ). 2011 ; 56 ( 3 ): 263 ±9. https://doi.org/10.1097/QAI.0b013e31820413b3 PMID: 21317585
32. Lawn SD , Wood R . Incidence of tuberculosis during highly active antiretroviral therapy in high-income and low-income countries. Clinical infectious diseases: an official publication of the Infectious Diseases Society of America . 2005 ; 41 ( 12 ): 1783 ± 6 . Epub 2005/11/17. https://doi.org/10.1086/498308 PMID: 16288404 .
33. Lawn SD , Myer L , Bekker LG , Wood R . Burden of tuberculosis in an antiretroviral treatment programme in sub-Saharan Africa: impact on treatment outcomes and implications for tuberculosis control . AIDS . 2006 ; 20 . https://doi.org/10.1097/01.aids. 0000238406 .93249. cd PMID : 16868441
34. Lawn SD , Badri M , Wood R . Risk factors for tuberculosis among HIV-infected patients receiving antiretroviral treatment . American journal of respiratory and critical care medicine . 2005 ; 172 ( 10 ) :1348; author reply -9 . Epub 2005/11/09. https://doi.org/10.1164/ajrccm.172.10.1348 PMID: 16275740 .
35. Lawn SD , Myer L , Edwards D , Bekker LG , Wood R . Short-term and long-term risk of tuberculosis associated with CD4 cell recovery during antiretroviral therapy in South Africa . Aids . 2009 ; 23 ( 13 ): 1717 ± 25 . Epub 2009/05/23. https://doi.org/10.1097/QAD.0b013e32832d3b6d PMID: 19461502
36. Dhasmana DJ , Dheda K , Ravn P , Wilkinson RJ , Meintjes G . Immune reconstitution inflammatory syndrome in HIV-infected patients receiving antiretroviral therapy: pathogenesis, clinical manifestations and management . Drugs . 2008 ; 68 ( 2 ): 191 ± 208 . Epub 2008/01/17. PMID: 18197725 .
37. Tieu HV , Ananworanich J , Avihingsanon A , Apateerapong W , Sirivichayakul S , Siangphoe U , et al. Immunologic markers as predictors of tuberculosis-associated immune reconstitution inflammatory syndrome in HIV and tuberculosis coinfected persons in Thailand . AIDS research and human retroviruses . 2009 ; 25 ( 11 ): 1083 ± 9 . Epub 2009/11/06. https://doi.org/10.1089/aid. 2009 .0055 PMID: 19886838
38. Hermans SM , Kiragga AN , Schaefer P , Kambugu A , Hoepelman AI , Manabe YC . Incident tuberculosis during antiretroviral therapy contributes to suboptimal immune reconstitution in a large urban HIV clinic in sub-Saharan Africa . PloS one . 2010 ; 5 ( 5 ): e10527 . Epub 2010 /05/19. https://doi.org/10.1371/journal. pone. 0010527 PMID: 20479873
39. Gaardbo JC , Nielsen SD , Vedel SJ , Ersboll AK , Harritshoj L , Ryder LP , et al. Regulatory T cells in human immunodeficiency virus-infected patients are elevated and independent of immunological and virological status, as well as initiation of highly active anti-retroviral therapy . Clinical and experimental immunology . 2008 ; 154 ( 1 ): 80 ± 6 . Epub 2008/09/30. https://doi.org/10.1111/j.1365- 2249 . 2008 . 03725 . x PMID : 18821942
40. Ravimohan S , Tamuhla N , Steenhoff AP , Letlhogile R , Makutu DK , Nfanyana K , et al. Early Immunologic Failure is Associated With Early Mortality Among Advanced HIV±Infected Adults Initiating Antiretroviral Therapy With Active Tuberculosis . The Journal of infectious diseases . 2013 ; 208 ( 11 ): 1784 ± 93 . https://doi.org/10.1093/infdis/jit368 PMID: 23908475
41. Pathmanathan I , Dokubo EK , Shiraishi RW , Agolory SG , Auld AF , Onotu D , et al. Incidence and predictors of tuberculosis among HIV-infected adults after initiation of antiretroviral therapy in Nigeria, 2004 ± 2012 . PloS one . 2017 ; 12 ( 3 ):e0173309. https://doi.org/10.1371/journal.pone. 0173309 PMID: 28282390
42. Auld AF , Mbofana F , Shiraishi RW , Alfredo C , Sanchez M , Ellerbrock TV . Incidence and determinants of tuberculosis among adults initiating antiretroviral therapyÐMozambique, 2004 ± 2008 . PloS one . 2013 ; 8 . https://doi.org/10.1371/journal.pone. 0054665 PMID: 23349948
43. Van Rie A , Westreich D , Sanne I. Tuberculosis in patients receiving antiretroviral treatment: incidence, risk factors and prevention strategies . Journal of Acquired Immune Deficiency Syndromes ( 1999 ). 2011 ; 56 ( 4 ): 349 ± 55 . PMID: 20926954
44. Mupfumi L , Makamure B , Chirehwa M , Sagonda T , Zinyowera S , Mason P , et al. Impact of Xpert MTB/ RIF on antiretroviral therapy-associated tuberculosis and mortality: a pragmatic randomized controlled trial . Open Forum Infectious Diseases . 2014 . https://doi.org/10.1093/ofid/ofu038 PMID: 25734106
45. Cain KP , McCarthy KD , Heilig CM , Monkongdee P , Tasaneeyapan T , Kanara N , et al. An algorithm for tuberculosis screening and diagnosis in people with HIV . The New England journal of medicine . 2010 ; 362 ( 8 ): 707 ± 16 . Epub 2010/02/26. https://doi.org/10.1056/NEJMoa0907488 PMID: 20181972 .
46. http://www.who.int/vmnis/indicators/haemoglobin.pdf. Haemoglobin concentrations for the diagnosis of anaemia and assessment of severity 2011 [[Accessed 27 April 2017 ]].
47. Kerkhoff AD , Wood R , Cobelens FG , Gupta-Wright A , Bekker L-G , Lawn SD . The predictive value of current haemoglobin levels for incident tuberculosis and/or mortality during long-term antiretroviral
48. Minchella PA , Armitage AE , Darboe B , Jallow MW , Drakesmith H , Jaye A , et al. Elevated hepcidin at HIV diagnosis is associated with incident tuberculosis in a retrospective cohort study . The international journal of tuberculosis and lung disease: the official journal of the International Union against Tuberculosis and Lung Disease . 2014 ; 18 ( 11 ): 1337 ± 9 . Epub 2014/10/10. https://doi.org/10.5588/ijtld.14.0143 PMID: 25299867 .
49. Getahun H , Kittikraisak W , Heilig CM , Corbett EL , Ayles H , Cain KP , et al. Development of a standardized screening rule for tuberculosis in people living with HIV in resource-constrained settings: individual participant data meta-analysis of observational studies . PLoS medicine . 2011 ; 8 ( 1 ): e1000391 . Epub 2011 /01/27. https://doi.org/10.1371/journal.pmed. 1000391 PMID: 21267059
50. Brinkhof MW , Egger M , Boulle A , May M , Hosseinipour M , Sprinz E , et al. Tuberculosis after initiation of antiretroviral therapy in low-income and high-income countries. Clinical infectious diseases: an official publication of the Infectious Diseases Society of America . 2007 ; 45 ( 11 ): 1518 ± 21 . Epub 2007/11/09. https://doi.org/10.1086/522986 PMID: 17990236 .
51. Rangaka MX , Wilkinson RJ , Boulle A , Glynn JR , Fielding K , van Cutsem G , et al. Isoniazid plus antiretroviral therapy to prevent tuberculosis: a randomised double-blind, placebo-controlled trial . Lancet . 2014 ; 384 ( 9944 ): 682 ± 90 . Epub 2014/05/20. https://doi.org/10.1016/S0140- 6736 ( 14 ) 60162 - 8 PMID: 24835842
52. Sumner T , Houben RM , Rangaka MX , Maartens G , Boulle A , Wilkinson RJ , et al. Post-treatment effect of isoniazid preventive therapy on tuberculosis incidence in HIV-infected individuals on antiretroviral therapy . AIDS . 2016 ; 30 ( 8 ): 1279 ± 86 . Epub 2016/03/08. https://doi.org/10.1097/QAD. 0000000000001078 PMID: 26950316 .
53. Samandari T , Agizew TB , Nyirenda S , Tedla Z , Sibanda T , Mosimaneotsile B , et al. Tuberculosis incidence after 36 months' isoniazid prophylaxis in HIV-infected adults in Botswana: a posttrial observational analysis . AIDS . 2015 ; 29 ( 3 ): 351 ± 9 . PMID: 25686683
54. Rangaka MX , Wilkinson RJ . Isoniazid prevention of HIV-associated tuberculosis . The Lancet infectious diseases . 13 ( 10 ): 825 ±7. https://doi.org/10.1016/S1473- 3099 ( 13 ) 70218 - 4 PMID: 23954451
55. Theron G , Peter J , Dowdy D , Langley I , Squire SB , Dheda K. Do high rates of empirical treatment undermine the potential effect of new diagnostic tests for tuberculosis in high-burden settings? The Lancet infectious diseases . 2014 ; 14 ( 6 ): 527 ± 32 . Epub 2014/01/21. https://doi.org/10.1016/S1473- 3099 ( 13 ) 70360 - 8 PMID: 24438820 .