Highly-active anti-retroviral therapy (HAART) has produced a sustained decrease in the incidence of death and AIDS-defining events including tuberculosis (TB) amongst HIV-positive individuals [1–3]. However, its use has been associated with inflammatory reactions that often mimic infection. Termed the immune reconstitution inflammatory syndrome (IRIS), these events typically occur within weeks of starting therapy [4–6]. In TB they are termed paradoxical reactions (PR) and have been reported in up to 36% of individuals being treated for TB on HAART [7,8]. Similar reactions can be observed in HIV-negative individuals however, HAART appears to promote more rapid and severe PR . We postulated that commencing HAART may precipitate clinically-apparent TB in individuals with latent or sub-clinical disease as an IRIS phenomenon and so undertook a retrospective study to investigate whether there were clinical differences between patients that developed active TB soon after starting HAART and those who developed symptoms later.
Patients who were diagnosed with TB between February 1997 and April 2004 were identified from the Royal Free Hospital TB database. Individuals co-infected with HIV were located and a retrospective case-note review was performed to identify and describe those receiving HAART at the time of TB diagnosis. Additional information was obtained from the Royal Free Hospital HIV database, described elsewhere . TB diagnoses were accepted if the patient had: (1) a positive culture for Mycobacterium tuberculosis; (2) been culture negative but acid fast bacilli (AFB) smear and/or nucleic acid amplification assay positive (TB Strand Displacement Amplification assay (SDA), Becton-Dickinson, Franklin Lakes, New Jersey, USA) with clinico-radiological features and treatment response consistent with TB; (3) histological findings and response to treatment consistent with TB. PR was defined as worsening of clinical or radiological findings following initiation of appropriate anti-tuberculosis therapy without evidence of disease relapse or the presence of another diagnosis. As a comparison for the rate of PR, we compared our study population to a previously described group of 28 unselected patients attending our centre with HIV and TB who during the same time period started anti-TB treatment prior to taking HAART .
Comparisons between groups were made using χ2 tests, Fishers exact tests, and Mann–Whitney U tests as appropriate.
One hundred and eleven HIV/TB co-infected individuals were identified of whom 19 (17%) were receiving HAART when they developed TB. These 19 had a median age of 39 years (range, 23–58 years), and comprised 12 men and seven women. At the time of TB diagnosis median blood CD4 cell count was 241 × 106 cells/l (range, 18–518 × 106 cells/l) and plasma viral load 71 copies/ml (range, < 50–23 000 copies/ml).
The nineteen could be divided into two groups based on time from commencing HAART to the development of TB. Thirteen of 19 (68%) presented with TB symptoms 7 to 109 days after starting HAART (median, 41 days), and this group was defined as ‘early TB’. Eight of 13 were Black-African and the remaining five were Caucasian. Five of 13 had pulmonary disease; five of 13 had miliary disease and 11 of 13 (85%) were TB culture positive. The remaining six patients (32%) developed TB 257 to 598 days after HAART initiation (median, 358 days) and were termed ‘late TB’. Of these, three of six were Black-African and the other three were Caucasian; two of six had pulmonary disease; and none had miliary disease. Four of six (67%) were TB culture positive. In the ‘early TB’ group six of 13 had TB at multiple sites against none of six (0%) in the other group (P = 0.04).
Median baseline CD4 cell count when starting HAART was 87 × 106 cells/l (range, 12–289) in the ‘early TB’ group compared with 218 × 106 cells/l (range, 150–521) in the ‘late-TB’ group (P = 0.04); however, 6 months after starting HAART no difference was seen in the rate of change of blood CD4 cell counts between the two populations [median increase in CD4 cell count: ‘early TB’ group = 119 × 106 cells/l (range:–7 to 349); ‘late TB’ group = 113 × 106 cells/l (range, –15 to 169) (P = 0.5)]. No significant differences in these measures were found at earlier time points. Median viral load when starting HAART was 665 500 copies/ml (range, 47 800–750 000) in the ‘early-TB’ group versus 360 000 (51 831–750 000) in the ‘late-TB’ group. The rate of change in plasma HIV load was also similar within the two groups (data not shown).
In the ‘early TB’ group eight of 13 (62%) experienced PR after starting anti-tuberculosis therapy against none of six (0%) in the ‘late TB’ group (P = 0.02). Three of the five ‘early TB’ subjects who did not experience PR received prolonged courses of corticosteroids for other indications which may have abrogated this inflammatory reaction (none of six in the ‘late TB’ group received steroids). The rate of PR in the ‘early TB’ group was compared to a further group of HIV-infected individuals who received HAART after starting anti-TB therapy. These 28 subjects had a median age of 36 years (range, 24–62 years) and 77% were from Sub-Saharan Africa. Median CD4 cell count prior to receiving HAART was 57 × 106 cells/l (range, 2–142) and 14 of 30 (47%) had disseminated disease. The rate of PR in this group was lower at 30% (nine of 30) (P = 0.05).
In the ‘early TB’ group four of 13 (31%) were asymptomatic at commencement of HAART, which was started due to falling CD4 cell counts. One of 13 had recurrent Herpes simplex infection, seven complained of fever, seven reported weight loss and one had non-specific malaise (Table 1). Two described cough and had multiple sputum samples that were AFB smear and TB culture negative. Broncho-alveolar lavage was not undertaken. Ten had pre-HAART chest radiographs. Eight revealed no abnormality. One showed a small nodule in the right middle lobe with no inflammatory features and one non-specific right lower lobe shadowing. Two patients with normal chest radiographs had computerized tomography (CT) scans of the chest. Both revealed single, small mediastinal lymph nodes and one showed a small right middle lobe nodule. One individual presented with headache and fever. CT brain scan demonstrated an isolated basal ganglia lesion which on biopsy was Toxoplasma polymerase chain reaction (PCR) positive but TB PCR and culture negative. In the ‘late TB’ group all started HAART as a result of low or falling CD4 cell counts. Two had associated weight loss and one had been treated for Pneumocystis jiroveci pneumonia. In line with local policy, no patients had a tuberculin skin test (TST), as this has poor specificity in our largely BCG-vaccinated population, and low sensitivity with advanced HIV infection .
Of the nine individuals in the ‘early TB’ group who were symptomatic prior to starting HAART, four improved initially and then subsequently deteriorated; three rapidly worsened and two remained symptomatically stable prior to subsequent deterioration. Figure 1 shows the chest radiographs before HAART and at TB diagnosis (10 days later) of a patient with sweats and weight loss prior to starting HAART. Over 10 days, the patient developed left upper lobe consolidation from which TB was cultured.
Our data indicate that almost one-fifth of cases of TB in HIV infected individuals occurred in patients receiving HAART. This is consistent with previous work . Moreover we have found that a large proportion, mainly from social groups with high rates of TB or from TB endemic countries, have active TB diagnosed within a few weeks of starting HAART. The majority had either fever, cough or weight loss prior to commencing antiretrovirals. It is possible therefore that despite investigation we failed to identify active TB pre-HAART. Whatever the case, we suggest that HAART contributed to disease presentation either by unmasking TB in previously asymptomatic individuals, or exaggerating its features in those with symptoms prior to its use .
Strikingly, the 13 patients in whom active TB developed rapidly following HAART initiation (‘early TB’) had a significantly greater rate of paradoxical reaction than those in whom TB presented later (‘late TB’). While possibly reflecting differences in baseline CD4 cell count and rates of disseminated disease between the groups, the PR rate in the ‘early TB’ group was also greater than that seen in a set of HIV/TB co-infected controls with similar pre-HAART CD4 cell counts, ethnicity and identical rates of disseminated disease. This suggests an unusually exaggerated and prolonged ‘HAART-associated’ inflammatory response in the ‘early TB’ population and raises the possibility of an underlying ‘inflammatory phenotype’ as postulated in other settings of IRIS .
If HAART-associated immune reconstitution influences the presentation of the early cases of TB, this might be reflected in changes in circulating lymphocyte populations . In our study, subjects presenting rapidly with TB did have lower baseline CD4 cell counts; however there appeared to be no difference in the rate of change of either blood CD4 cell count or HIV load on HAART. Conceivably HAART promotes the return of TB-specific lymphocytes to the site of latent or low-grade infection causing a rapid increase in local inflammation with consequent symptoms of clinical disease. Hence measurements of blood-borne markers such as CD4 cell counts may miss local inflammatory reactions .
These local events are of undoubted clinical importance as 85% of subjects with ‘early TB’ were culture positive. Thus the episode would be unlikely to resolve without specific anti-TB therapy – a phenomenon documented with less virulent mycobacteria .
We are unaware of previous reports of ‘early TB’ after starting HAART. This may be due to a number of factors including: (1) the ‘miscounting’ of early events on HAART as arising as a direct consequence of HIV-associated immunosuppression and thus being regarded as unrelated to drug therapy; and (2) the current poor availability of HAART globally. Interestingly, recent reports of HAART introduction in high prevalence TB areas does suggest a significant incidence of TB in HAART-treated populations, although the temporal relationship to HAART and associated clinical features were not fully described [18,19].
Although our study is limited by its size and retrospective nature, the data suggest that this phenomenon of disease exacerbation may substantially impact upon high-burden TB countries where HAART is to be widely introduced . Whereas there is no denying the great benefit of antiretrovirals in reducing death and opportunistic disease, our data flag up a potential negative consequence of HAART. Assuming that it will be introduced into settings with limited diagnostics and targeted at symptomatic individuals with advanced immunosuppression, our data may be tentatively extrapolated as follows. During the study period we commenced 267 individuals from Sub-Saharan Africa on HAART. Thus the observed overall frequency of TB after starting HAART in this group was 11 of 267 [4.12%; 95% confidence interval (CI), 2.07–7.26%], and of ‘early TB’ was eight of 267 (3.00%; 95% CI, 1.30–5.82%). If our observed rate of development of TB rapidly after commencing HAART (3%) were reproduced in the three million on the ‘3 by 5’ programme, then 90 000 cases of TB would be unmasked, rising substantially if HAART were extended to the estimated global figure of 11.4 million with both HIV and unknown M. tuberculosis between the ages of 15 and 49 . It is uncertain how many of these subjects would have remained asymptomatic without HAART. We believe that our findings warrant further, careful prospective investigation involving a larger cohort of patients.
The authors are grateful to Nicholas Price and Felicity Perrin for critical appraisal of the manuscript.
Sponsorship: This study received funding from the Royal Free Hospital Special Trustees.
1. Palella FJ Jr, Delaney KM, Moorman AC, Loveless MO, Fuhrer J, Satten GA, et al
. Declining morbidity and mortality among patients with advanced human immunodeficiency virus infection. N Engl J Med 1998; 338:853–860.
2. Santoro-Lopes G, de Pinho AM, Harrison LH, Schechter M. Reduced risk of tuberculosis among Brazilian patients with advanced human immunodeficiency virus infection treated with highly active antiretroviral therapy. Clin Infect Dis 2002; 34:543–546.
3. Girardi E, Antonucci G, Vanacore P, Libanore M, Errante I, Matteelli A, et al
. Impact of combination antiretroviral therapy on the risk of tuberculosis among persons with HIV infection. AIDS 2000; 14:1985–1991.
4. Barry SM, Lipman MCI, Deery AR, Johnson MA, Janossy G. Immune reconstitution
pneumonitis following Pneumocystis carinii pneumonia in HIV-infected subjects. HIV Medicine 2002; 3:207–211.
5. Race EM, Adelson-Mitty J, Kriegel GR, Barlam TF, Reimann KA, Letvin NL, et al
. Focal mycobacterial lymphadenitis following initiation of protease-inhibitor therapy in patients with advanced HIV-1 disease. Lancet 1998; 351:252–255.
6. Zegans ME, Walton RC, Holland GN, O'Donnell JJ, Jacobson MA, Margolis TP. Transient vitreous inflammatory reactions associated with combination antiretroviral therapy in patients with AIDS and cytomegalovirus retinitis. Am J Ophthalmol 1998; 125:292–300.
7. Narita M, Ashkin D, Hollender ES, Pitchenik AE. Paradoxical worsening of tuberculosis following anti-retroviral therapy in patients with AIDS. Am J Respir Crit Care Med 1998; 158:157–161.
8. Breen RAM, Smith CJ, Bettinson H, Dart S, Bannister B, Johnson MA, et al
. Paradoxical reactions during tuberculosis treatment in patients with and without HIV co-infection. Thorax 2004; 59:704–707.
9. Goldsack NR, Allen S, Lipman MCI. Adult respiratory distress syndrome as a severe immune reconstitution
disease following the commencement of highly active antiretroviral therapy. Sex Transm Infect 2003; 79:337–338.
10. Sabin CA, Lampe F, Chaloner C, Madge SJ, Lipman MC, Youle M, et al
. An audit of antiretroviral treatment use in HIV-infected patients in a London clinic. HIV Medicine 2003; 4:87–93.
11. Jasmer RM, Nahid P, Hopewell PC. Clinical practice. Latent tuberculosis infection. N Engl J Med 2002; 347:1860–1866.
12. Girardi E, Palmieri F, Cingolani A, Ammassari A, Petrosillo N, Gillini L, et al
. Changing clinical presentation and survival in HIV-associated tuberculosis after highly active antiretroviral therapy. J Acquir Immune Defic Syndr 2001; 26:326–331.
13. Mohammed A, Ehrlich R, Wood R, Cilliers F, Maartens G. Screening for tuberculosis in adults with advanced HIV infection prior to preventative therapy. Int J Tuberc Lung Dis 2004; 8:792–795.
14. Price P, Morahan G, Huang D, Stone E, Cheong KY, Castley A, et al
. Polymorphisms in cytokine genes define subpopulations of HIV-1 patients who experienced immune restoration diseases. AIDS 2002; 16:2043–2047.
15. Cooney EL. Clinical indicators of immune restoration following highly active anti-retroviral therapy. Clin Infect Dis 2002; 34:224–233.
16. Barry SM, Lipman MCI, Bannister B, Johnson MA, Janossy G. Purified protein derivivative-activated type 1 cytokine-producing CD4+ T lymphocytes in the lung: a characteristic feature of active pulmonary and non-pulmonary TB. J Infect Dis 2003; 187:243–250.
17. Kerbiriou L, Ustianowski A, Johnson MA, Gillespie SH, Miller RF, Lipman MCI. Human immunodeficiency virus type 1-related pulmonary Mycobacterium xenopi infection: a need to treat? Clin Infect Dis 2003; 37:1250–1254.
18. Sungkanuparph S, Vibhagool A, Mootsikapun P, Chetchotisakd P, Tansuphaswswadikul S, Bowonwatanuwong C. Opportunistic infections after the initiation of highly active antiretroviral therapy in advanced AIDS patients in an area with a high prevalence of tuberculosis. AIDS 2003; 17:2129–2130.
19. Tassie J-M, Szumilin E, Calmy A, Goemare E. Highly active antiretroviral in resource-poor settings: the experience of Médecins Sans Frontières. AIDS 2003; 17:1995–1997.
20. World Health Organisation. The WHO and UNAIDS Global Initiative to Provide Antiretroviral Therapy to 3 million people with HIV/AIDS in Developing Countries by the End of 2005
. Geneva, Switzerland: World Health Organisation; 2003. Available at: http://www.who.int/3by5/en
21. Corbett EL, Watt CJ, Walker N, Maher D, Williams BG, Raviglione MC, et al
. The growing burden of tuberculosis. Arch Intern Med 2003; 163:1009–1021.