Tuberculosis (TB) accounts for significant morbidity and mortality in individuals with HIV infection globally, with an estimated 1.1 million incident cases of HIV-associated TB (HIV-TB) in 2012 . In the United Kingdom, the majority of HIV-TB co-infections occur in people with previously unknown HIV status . National and international guidelines recommend starting antiretroviral therapy (ART) in such individuals regardless of CD4+ cell count [3,4]. This has resulted in reduced mortality, particularly in individuals with advanced immunosuppression [5–7].
Concurrent treatment for TB and ART, however, poses various management challenges, including drug–drug interactions, increased pill burden, overlapping drug toxicities and immune reconstitution disease . There are conflicting reports of the impact of TB diagnosis and treatment on CD4+ cell responses following ART initiation [8–13]. Potential mechanisms for impaired CD4+ responses in the presence of HIV-TB co-infection include the practical difficulties associated with concurrent therapy, as well as possible immunological mechanisms due to increased T-cell activation and apoptosis . A recent systematic review suggested that CD4+ cell reconstitution appeared similar in individuals receiving TB treatment at the time of ART initiation compared to those who did not have co-existent TB . However, due to study heterogeneity, a meta-analysis was not possible; considerable debate therefore remains regarding the impact of TB on CD4+ responses to ART.
The present study utilized comprehensive national data from persons attending the National Health Service (NHS) to compare long-term CD4+ responses to ART in HIV-infected adults (>15 years) with TB diagnosed around the time of ART initiation, and HIV-infected, TB-free individuals. Since differential CD4+ responses to ART have been observed by sex , ethnicity , age  and baseline CD4+ cell count , the TB-free population was frequency-matched to the HIV-TB cases on these variables.
We used data from the HIV and AIDS Reporting System (HARS) linked to the Enhanced TB Surveillance (ETS) system for England, Wales and Northern Ireland. HARS comprises reports of new HIV and AIDS diagnoses and deaths, annually updated clinical and demographic information on all individuals accessing HIV care, laboratory reports of CD4+ cell counts and reports of deaths from the Office for National Statistics. These HIV datasets were linked to ETS using probabilistic matching as previously described .
HIV-infected individuals were included in the ‘HIV-TB cohort’ if reported to have active TB (through ETS or HARS as an AIDS-defining illness) less than 6 months before, or 3 months after, starting ART. The ‘TB-free cohort’ comprised of HIV-diagnosed individuals not reported to have TB within this time period. Baseline CD4+ was defined as a CD4+ cell count in the 6 months prior to ART start date. If multiple CD4+ cell counts were available during this time period, the last count prior to commencing ART was used. Individuals were defined as lost to follow-up if there was no evidence of access to HIV care for an entire calendar year. An ART interruption was defined as an individual reported to be not receiving ART any time after their initial ART start date.
Individuals were included if they initiated ART from 2005 to 2009, had a known baseline CD4+ cell count and at least one CD4+ cell count available after initiating ART. The TB-free cohort was frequency-matched to the HIV-TB cohort according to sex; age strata (16–35 vs. 36–55 vs. ≥56 years); ethnicity (white vs. black-African vs. other) and baseline CD4+ count strata (<100 vs. 100–199 vs. 200–349 vs. ≥350 cells/μl), with a ratio of three individuals in the TB-free cohort for every one individual in the HIV-TB cohort. Individuals reported to have non-TB mycobacterial infections around the time of starting ART were excluded from the TB-free cohort.
Median increase in CD4+ cell count from baseline (ΔCD4+) was calculated for HIV-TB and TB-free cohorts at 6-monthly intervals up to 36 months of ART. Only individuals with a CD4+ cell count less than 61 days from each time point were included in the median calculations (count closest to corresponding time point was used if multiple counts were available within this timeframe). Individual follow-up was censored at the earliest of: date of death; date of loss to follow-up; date of ART interruption or 31 December 2011. All analyses were completed using Stata version 12.0 (StataCorp, College Station, Texas, USA). Chi-square tests were used to test proportions and Wilcoxon rank-sum test was used to compare distributions (P < 0.05 threshold for significance).
We included several sensitivity analyses. Firstly, we restricted analysis in both cohorts to individuals known to have HIV viral suppression (<50 copies/ml) after 1 year of ART (75.3% of all individuals) in order to assess whether there was any difference in CD4+ cell responses amongst those receiving ‘effective’ ART. Secondly, we excluded individuals with non-TB AIDS from the TB-free cohort (n = 243) to investigate if non-TB AIDS conditions could have blunted CD4+ responses to ART in a similar way to TB. Thirdly, we excluded individuals with TB reported ‘at any time’ from controls (n = 101) to assess whether inclusion of these individuals in the TB-free cohort could have masked any difference in CD4+ responses between the populations. Fourthly, we stratified HIV-TB cases by disease site (pulmonary only vs. any extra-pulmonary disease) to see if this impacted on CD4+ responses to ART. Fifthly, we adjusted the timeframe for inclusion of CD4+ cell counts in median calculations (<61 days of each time point in main analysis; <43 days and <85 days also tested), and the timeframe between TB diagnosis and ART start date defining HIV-TB cases (6 months before to 3 months after ART start date in the main analysis; 6 months before to ART start date, and 3 months before to 3 months after ART start date also tested). Finally, to test the findings of the main analysis, we compared CD4+ responses during ART between the HIV-TB and TB-free cohorts with a different methodology, using a linear mixed model.
In addition, we conducted sub-analyses to assess whether the impact of TB on CD4+ responses varied by age (≤35 vs. >35 years), sex and baseline CD4+ cell count (<100 vs. ≥100 cells/μl).
Baseline characteristics of HIV-tuberculosis and tuberculosis-free cohorts
A total of 593 and 1779 individuals were included in the HIV-TB and TB-free cohorts, respectively (Table 1). In both populations, 49.2% were women, median age was 36 years [inter-quartile range (IQR) 31–42] and the majority were black-African (74.9%). Baseline median CD4+ cell count was similar in HIV-TB [74 cells/μl (IQR 30–166)] and TB-free cohorts [80 cells/μL (IQR 25–170)], as was median baseline viral load [5.8 log copies/ml (IQR 3.9–11.3) vs. 5.6 log copies/ml (IQR 3.9–9.9)].
Characteristics of HIV-TB cases included in the analysis were similar to all HIV-TB cases diagnosed around the time of starting ART during the study period (data not shown). The frequency-matched TB-free cohort differed markedly from all TB-free individuals initiating ART during the study period, with a lower median CD4+ cell count [80 (IQR 25–170) vs. 214 cells/μl (IQR 120–300)], higher proportion of women (49.2 vs. 37.2%), higher proportion of black-Africans (74.9 vs. 40.4%) and lower proportion of white individuals (11.1 vs. 45.5%). The median time from TB treatment to starting ART among individuals in the HIV-TB group diagnosed with TB prior to commencing ART was 54 days (IQR 27–85).
Median follow-up was 3.8 years (IQR 2.4–5.2), during which individuals had a median of 13 CD4+ cell counts (IQR 8–18) reported. A greater proportion of individuals in the HIV-TB cohort died than in the TB-free cohort (4.2 vs. 2.2%; P = 0.01), but similar proportions of both cohorts were lost to follow-up and had ART interruptions (Table 1). Median time from ART initiation to death was 0.9 years (IQR 0.2–2.0) in the HIV-TB cohort and 1.3 years (IQR 0.3–2.0) in the TB-free cohort. Amongst individuals who died during follow-up, median baseline CD4+ cell counts were 60 (IQR 10–66) and 70.5 (20–97) cells/μl in the HIV-TB and TB-free cohorts, respectively.
CD4+ cell responses to antiretroviral therapy
CD4+ cell count responses to ART were similar in both cohorts at all time points [median ΔCD4+ 294 (IQR 198–424) cells/μl in the HIV-TB cohort; and 296 (IQR 196–431) cells/μl in the TB-free cohort after 3 years of ART] (Fig. 1). Individuals who died had worse CD4+ cell count responses after 1 year of ART than those who survived [median ΔCD4+ 126 (IQR 44–172) cells/μl amongst individuals who died; and 170.5 (IQR 103.5–258.5) cells/μl in those who survived; P = 0.0047].
All sensitivity analyses had little impact on results with no observed differences in CD4+ cell responses between the HIV-TB and the TB-free cohorts (data shown graphically in Appendices, http://links.lww.com/QAD/A688).
The present study found that blood CD4+ cell responses to ART were similar in the HIV-infected individuals with and without TB co-infection. Previous data on this subject have been conflicting. Most studies have suggested that TB co-infection has no impact [8,10,12,13]; however, recent data from Italy, Uganda and South Africa have described blunted CD4+ responses in HIV-infected individuals diagnosed with TB prior to, or soon after, initiation of ART [9,11,20]. One reason for this apparent discrepancy may be the different timeframes used for HIV-TB cases. Some have included TB cases diagnosed before starting ART only [8,10,11,13]; others only incident TB cases during early ART ; whilst some have combined both [9,12]. Here, when applying various timeframes within the sensitivity analysis, we found there to be little difference within our results.
A recent systematic review concluded that individuals receiving TB treatment at the time of ART initiation have similar CD4+ cell count increases during ART . Our findings provide further support for this and indicate that ART can effectively restore CD4+ cell counts in individuals presenting with HIV-TB co-infection, despite the numerous challenges faced in the management of these cases . This indicates that, if individuals survive, long-term immune recovery is likely to be very good. However, median baseline CD4+ cell count of HIV-TB cases was 74 cells/μl. This is far below the median CD4+ cell count of all individuals commencing ART in our cohort, indicating that the vast majority of HIV-TB cases are in individuals presenting late with advanced immunodeficiency, and 4.2% of these individuals died during follow-up (80% of individuals with HIV-TB who died had a baseline CD4+ cell count <100 cells/μl). This highlights a need for routine HIV testing in high-risk groups to facilitate earlier HIV diagnosis and initiation of ART, to allow a reduction in risk of acquiring opportunistic infections including TB and a reduction in subsequent morbidity and mortality associated with these infections.
Strengths of this study include the large and comprehensive dataset available on the HIV-positive cohort and the number of HIV-TB cases, systematically collected through NHS HIV and TB clinics as part of routine national surveillance. This allowed a TB-free population to be frequency-matched to cases to account for differential CD4+ responses by age group, sex, ethnicity and baseline CD4+ cell count. Furthermore, median follow-up was 3.8 years, allowing assessment of the long-term impact of TB on CD4+ cell count recovery. A limitation of this study is that we are unable to account for individuals who died and who appeared to have worst blood CD4+ cell count responses to ART. A higher proportion of HIV-TB cases than TB-free individuals died during follow-up, which may have created an element of survivor bias. However, the purpose of this study was to assess CD4+ recovery in those who survived, and in fact only a relatively small proportion of the sample (4.2%) of HIV-TB cases died during follow-up. Finally, whilst we measured changes in absolute CD4+ cell counts over time, this study did not assess potential deficits in CD4+ cell function that may be ongoing in individuals with and without TB co-infection .
In conclusion, we found that long-term CD4+ cell count responses to ART were not impaired in individuals with HIV-TB co-infection. This highlights the ability of ART to restore CD4+ cell counts despite the array of challenges faced in the management of individuals with HIV-TB co-infection, and indicates that, should individuals presenting with HIV-TB survive, long-term CD4+ cell recovery is likely to be very good.
This study was undertaken by the UK TB-HIV Research, Epidemiology And Development (UK-THREAD) group. We would like to thank all members of the HARS and ETS teams at Public Health England for their excellent data management and for providing the data used in this study. R.K.G., B.R., A.B., H.L.T., Z.Y., D.Z., A.P., I.A., V.D., and M.L. designed the analysis. R.K.G. performed the analysis and wrote the first draft of the manuscript. A.B., H.L.T., Z.Y., I.A., and V.D. collected and maintained the datasets. All authors critically reviewed the manuscript and approved the final version prior to submission.
Conflicts of interest
There are no conflicts of interest.
2. Rice B, Elford J, Yin Z, Kruijshaar M, Abubakar I, Lipman M, et al. Decreasing incidence of tuberculosis among heterosexuals living with diagnosed HIV in England and Wales
4. Pozniak AL, Coyne KM, Miller RF, Lipman MCI, Freedman a R, Ormerod LP, et al. British HIV Association guidelines for the treatment of TB/HIV coinfection 2011
. HIV Med
5. Abdool Karim SS, Naidoo K, Grobler A, Padayatchi N, Baxter C, Gray AL, et al. Integration of antiretroviral therapy with tuberculosis treatment
. N Engl J Med
6. Blanc F-X, Sok T, Laureillard D, Borand L, Rekacewicz C, Nerrienet E, et al. Earlier versus later start of antiretroviral therapy in HIV-infected adults with tuberculosis
. N Engl J Med
7. Havlir DV, Kendall Ma, Ive P, Kumwenda J, Swindells S, Qasba SS, et al. Timing of antiretroviral therapy for HIV-1 infection and tuberculosis
. N Engl J Med
8. Schomaker M, Egger M, Maskew M, Garone D, Prozesky H, Hoffmann CJ, et al. Immune recovery after starting ART in HIV-infected patients presenting and not presenting with tuberculosis in South Africa
. J Acquir Immune Defic Syndr
9. Cingolani A, Cozzi Lepri A, Castagna A, Goletti D, De Luca A, Scarpellini P, et al. Impaired CD4 T-cell count response to combined antiretroviral therapy in antiretroviral-naive HIV-infected patients presenting with tuberculosis as AIDS-defining condition
. Clin Infect Dis
10. Siika AM, Yiannoutsos CT, Wools-Kaloustian KK, Musick BS, Mwangi AW, Diero LO, et al. Active tuberculosis is associated with worse clinical outcomes in HIV-infected African patients on antiretroviral therapy
. PLoS One
11. Bassett IV, Chetty S, Wang B, Mazibuko M, Giddy J, Lu Z, et al. Loss to follow-up and mortality among HIV-infected people co-infected with TB at ART initiation in Durban, South Africa
. J Acquir Immune Defic Syndr
12. Lawn SD, Myer L, Bekker L-G, Wood R. Burden of tuberculosis in an antiretroviral treatment programme in sub-Saharan Africa: impact on treatment outcomes and implications for tuberculosis control
13. Breen RAM, Miller RF, Gorsuch T, Smith CJ, Ainsworth J, Ballinger J, et al. Virological response to highly active antiretroviral therapy is unaffected by antituberculosis therapy
. J Infect Dis
14. Hermans SM, van Leth F, Kiragga AN, Hoepelman AIM, Lange JMA, Manabe YC. Unrecognised tuberculosis at antiretroviral therapy initiation is associated with lower CD4+ T cell recovery
. Trop Med Int Health
15. Soeters HM, Napravnik S, Patel MR, Eron JJ, Van Rie A. The effect of tuberculosis treatment on virologic and CD4+ cell count response to combination antiretroviral therapy: a systematic review
16. Giordano TP, Wright JA, Hasan MQ, White AC, Graviss EA, Visnegarwala F. Do sex and race/ethnicity influence CD4 cell response in patients who achieve virologic suppression during antiretroviral therapy?
. Clin Infect Dis
17. Smith CJ, Sabin CA, Youle MS, Kinloch-de Loes S, Lampe FC, Madge S, et al. Factors influencing increases in CD4 cell counts of HIV-positive persons receiving long-term highly active antiretroviral therapy
. J Infect Dis
18. Kaufmann GR, Bloch M, Finlayson R, Zaunders J, Smith D, Cooper DA. The extent of HIV-1-related immunodeficiency and age predict the long-term CD4 T lymphocyte response to potent antiretroviral therapy
19. Lawn SD, Myer L, Bekker L-G, Wood R. CD4 cell count recovery among HIV-infected patients with very advanced immunodeficiency commencing antiretroviral treatment in sub-Saharan Africa
. BMC Infect Dis
20. Hermans SM, Kiragga AN, Schaefer P, Kambugu A, Hoepelman AIM, Manabe YC. Incident tuberculosis during antiretroviral therapy contributes to suboptimal immune reconstitution in a large urban HIV clinic in sub-Saharan Africa
. PLoS One
21. Lawn SD, Bekker L-G, Wood R. How effectively does HAART restore immune responses to Mycobacterium tuberculosis? Implications for tuberculosis control