High incidence of tuberculin skin test conversion among HIV-infected individuals who have a favourable immunological response to highly active antiretroviral therapy

Abstract

A total of 129 HIV-infected individuals who were tuberculin skin test (TST) negative in 1997 and started highly active antiretroviral therapy (HAART) after skin testing, received a second TST during 2000. Seven patients converted to a positive TST. In multivariate analysis, an increase in CD4 T lymphocyte count more than 200 × 106/l over baseline was significantly associated with TST conversion. TST should be repeated in TST-negative HIV-infected individuals who experience an immunological response to HAART. The treatment of latent tuberculosis infection in HIV-infected individuals greatly reduces the risk of active tuberculosis [1]. Tuberculin skin test (TST) is at present the only standardized method of identifying individuals with latent tuberculosis infection [2]. However, the sensitivity of this test is greatly reduced in HIV-infected individuals [3]. Highly active antiretroviral therapy (HAART) may determine the restoration of in-vitro T cell response and of delayed-type hypersensitivity response to mycobacterial antigens [4,5]. Moreover, a conversion from negative to positive TST has been recorded in patients with active tuberculosis who initiated effective HAART [6]. On the basis of these observations, current guidelines recommend that clinicians should consider repeating TST for individuals whose initial skin test was negative and whose immune function has improved in response to HAART [7]. However, scant data are available on the rate of conversion from negative to positive TST reaction and on factors predictive of this conversion after the initiation of HAART in patients without active tuberculosis. To address this issue, we performed a cohort study on 129 HIV-infected individuals who started HAART after a first negative TST in 1997 [8] and received a second TST during 2000. At the time of initial TST, delayed-type hypersensitivity to other recall antigens was also assessed. Baseline and follow-up data were collected from patients’ charts and included: age, sex, HIV transmission category, diagnosis of AIDS-defining illnesses, CD4 lymphocyte count, plasma HIV-RNA level, and information on antiretroviral therapy. Differences in categorical variables were analysed with the use of the χ2 test or Fisher's exact test, as appropriate. Continuous variables were compared by using the non-parametric Mann–Whitney U test. Logistic regression analysis was used to identify determinants of conversion to a positive TST. The median age of the 129 patients included in the study was 39 years, 94 patients (73%) were men and 73 (57%) were injecting drug users. The median CD4 lymphocyte count was 295 × 106/l (range 5–963) at baseline and 461 × 106/l (range 54–1310) at follow-up. The median time between the initial and follow-up TST was 36.6 months (range 29.0–43.4), and the median duration of HAART was 31.3 months (range 13.9–42.5). Ninety-six patients (74%) received an antiretroviral regimen containing a protease inhibitor, whereas 33 patients (26%) were treated with a regimen including a non-nucleoside reverse transcriptase inhibitor. There was at least one change in antiretroviral regimen in 79 patients (61%). During follow-up there were no mycobacterial diseases among patients enrolled. Overall, seven patients (5.4%) converted to a positive TST at follow-up. There were no significant differences in the time elapsed between the initial and follow-up TST, in the duration of HAART and in the median CD4 lymphocyte count at baseline between patients converting to a positive TST and those remaining negative. In contrast, patients converting to a positive TST had a significantly higher median CD4 lymphocyte count at follow-up (592 versus 452 × 106/l; P = 0.029). Table 1 summarizes TST conversions by the characteristics of the patients. There was no TST conversion among patients who had an increase in CD4 T lymphocyte count less than 100 × 106/l over baseline; one TST conversion (4.3%) among 23 patients whose increase was between 100 and 200 × 106/l, whereas there were six TST conversions (11.5%) among the 52 patients who had an increase in CD4 T lymphocyte count greater than 200 × 106/l (P = 0.011 versus patients with increases < 100 × 106/l). In multivariate analysis (Table 1), an increase in CD4 T lymphocyte count greater than 200 × 106/l remained significantly associated with TST conversion.Table 1: Tuberculin skin test conversions by characteristics of 129 HIV-infected patients.The overall rate of conversion to a positive TST observed in this study is higher than that observed in previous studies conducted in industrialized countries before HAART came into widespread use [9,10] (E. Girardi, unpublished data). Moreover, all but one of the TST conversions were observed in patients who had an increase in CD4 T lymphocyte count greater than 200 × 106/l after HAART. These findings suggest that in our cohort conversion to a positive TST may be attributable to HAART-induced immune reconstitution. Previous studies have shown that HAART-treated patients frequently develop specific T cell reactivity and delayed-type hypersensitivity response to antigens that may be commonly encountered or may be present in the organism in a latent form (such as cytomegalovirus or mycobacterial antigens) [4]. It is therefore possible that our patients already had a latent tuberculosis infection at the time of first TST, and that HAART had restored their ability to mount a delayed-type hypersensitivity response to purified protein derivative. On the other hand, it has been shown that HAART can also induce a delayed-type hypersensitivity response to neoantigens administered after the initiation of antiretroviral therapy [11]. It is therefore also possible that some patients in our study acquired latent tuberculosis infection while on HAART, and those who had an efficient immune restoration were capable of mounting an immune response. This study has a series of limitations. The small sample size of the study precluded a precise estimation of the rate of TST conversion and the study of the association with TST conversion of other potentially relevant determinants, such as a decrease in HIV viraemia [4]. Moreover, the study was performed in a single clinical centre and the observed cuticonversion rate may well be different in settings with a different prevalence or incidence of tuberculosis infection. Finally, our results cannot be entirely extrapolated to patients starting therapy at a very advanced stage of disease for whom immune recovery may frequently be incomplete even in presence of a marked increase in CD4 T lymphocyte count [12]. The risk of developing active tuberculosis is clearly reduced for HIV-infected patients taking HAART [13]. However, HIV-associated tuberculosis continues to occur in the context of the wide use of HAART, also among patients who have an immunological response to antiretroviral therapy [14]. Our results show that TST should be repeated in TST-negative HIV-infected individuals who experience an immunological response to HAART, in order to identify those at high risk of developing tuberculosis who would benefit from the treatment of latent tuberculosis infection. Acknowledgements Sponsorship: This study was partly supported by the Ministero della Salute ‘Ricerca corrente e finalizzata degli IRCCS’ and by the Istituto Superiore di Sanità ‘Progetto Nazionale Tubercolosi'.