Tuberculosis is very prevalent in Spain [8]. Disseminated tuberculosis has been the AIDS-defining disease in 18% of HIV-infected patients in Madrid [9]. In our hospital, more than 60% of patients with AIDS had documented tuberculosis in the course of their HIV infection, and 21% of all culture-proven tuberculosis is diagnosed in HIV-infected patients [10, 11]. Given the magnitude of the problem, we evaluated the risk for developing tuberculosis among anergic and nonanergic HIV-infected persons and assessed the convenience of broadening the indications of chemoprophylaxis to anergic patients in our area.

Methods

Top Methods Results Discussion Author & Article Info References

The study was conducted in a 2200-bed teaching institution that serves a primarily middle-low socioeconomic class population of approximately 800 000. The institution also serves as a referral center for the Madrid prison hospital. From 1985 through mid-1989, patients with HIV infection who were referred for immunologic evaluation underwent delayed-type hypersensitivity skin testing, including a PPD test and a CD 4⁺ cell count. Immunologic evaluation is routine in HIV-infected patients when first seen at the hospital, and all those evaluated for delayed-type hypersensitivity anergy were included in the study. Patients were identified through the Immunology Laboratory registries, and the date of the first immunologic evaluation was considered the date of study entry.

The clinical charts of patients were reviewed for demographic information, risk practice for HIV infection, previous history of tuberculosis, chemoprophylaxis with isoniazid, results of mycobacterial cultures, development of active tuberculosis, time to development and sites of involvement of tuberculosis, and date of death. Delayed-type hypersensitivity testing results and CD4 count at entry and those done during follow-up were recorded. Information was collected until 31 December 1991.

Development of tuberculosis was assessed, in addition to reviewing clinical charts, by crossmatching with the Mycobacteria Laboratory logbook at our hospital. Patients' follow-up was completed through the AIDS Registry of the Comunidad Autonoma of Madrid and by reviewing the clinical records of HIV-infected patients at three city hospitals serving many HIV-infected persons. Much of the information was available from more than one source and could thus be corroborated.

A diagnosis was considered certain when Mycobacterium tuberculosis was isolated from any clinical specimen. Tuberculosis was considered probable in patients who had clinical findings highly suggestive of tuberculosis that resolved with antituberculous therapy alone or by the histologic finding of granuloma in clinical specimens.

Immunologic Tests

To evaluate delayed-type hypersensitivity responsiveness, intradermal injections of 0.1 mL of increasing doses of the following three antigens in physiologic saline solution were administered: streptokinase-streptodornase (4, 40, and 100 units), Candida albicans (at 1:100 and 1:10 dilutions), and tuberculin (2, 5, and 25 units of the RT-23 strain equivalent to 5, 10, and 50 TU of PPD). Responses were read 48 to 72 hours after administration and the maximum induration was recorded. The skin tests were administered and the results were interpreted by trained personnel in the Immunology Laboratory. A positive response to a PPD test was defined as an induration of 5 mm or more in diameter 48 to 72 hours after administration of the 2 units of the antigen. Anergy was defined as the lack of any cutaneous response to increasing doses of all of the above-mentioned antigens.

Peripheral blood mononuclear cells were isolated from whole blood by Ficoll-Hypaque density-gradient centrifugation. T-lymphocyte subsets were quantified by direct immunofluorescence using monoclonal antibodies of the T series (Becton Dickinson; Mountain View, California) and flow cytometric examination (FACScan; Becton Dickinson). Absolute numbers of T cells were calculated using leukocyte and differential counts.

Statistical Analysis

For purposes of analysis, the patients were divided into PPd-positive, PPd-negative with no anergy, and PPd-negative with anergy (anergic patients) groups according to the response to PPD at first evaluation. Comparisons between groups were made using the Newman-Keuls multiple-range test for continuous data, and chi-square or Fisher exact tests for categorical data. Incidence rates in the three groups were calculated as the number of patients in whom tuberculosis develops in each group divided by the total follow-up period of the same group (in patient-years). Incidence rates were compared using the chi-square test [12]. All statistical associations were assessed using two-tailed t-tests. A P value of less than 0.05 was considered indicative of statistical significance.

Results

Top Methods Results Discussion Author & Article Info References

A total of 706 HIV-infected patients underwent delayed-type hypersensitivity skin testing. Sufficient data were available from 567 (80%) patients. Of these, 43 (7.6%) had a history of previous tuberculosis and 150 (26%) had active tuberculosis at the time of study entry. The remaining 374 (65%) study patients, who had adequate follow-up, and the 139 patients for whom follow-up data were not available did not differ between groups with respect to age, sex, risk practice for HIV infection, and the number of PPd-positive and nonanergic PPd-negative patients. The group that was followed (study group) included statistically more anergic patients (112 versus 25; P = 0.009) and had a lower CD 4⁺ cell count (median, 467 versus 680 cells/mm³; P < 0.001).

The study group included 108 (29%) PPd-positive patients, 154 (41%) nonanergic PPd-negative patients, and 112 (30%) anergic patients. The characteristics of the three groups are shown in Table 1. The groups differed with respect to the CD4 cell count at entry and the duration of follow-up. Completed follow-up (that is, until development of tuberculosis, death, or the end of study) was achieved in 66 (61%) PPd-positive patients, in 92 (60%) nonanergic, PPd-negative patients, and in 89 (79%) anergic patients. During the period of study, death was more frequent in anergic patients (48%) than in PPd-positive patients (5%) or nonanergic, PPd-negative patients (12%) (P < 0.001, each). PPd-positive, true PPd-negative, and anergic patients with incomplete follow-up were observed a mean of 21, 19, and 20 months, respectively.

Of the 374 patients studied, 27 who had completed a course of isoniazid chemoprophylaxis were analyzed separately. Among the 347 patients who had not received previous isoniazid chemoprophylaxis, active tuberculosis developed in 63 (18%), of whom 49 (78%) had positive results of mycobacterial culture. The rate of tuberculosis varied according to the previous PPD status (Table 2). Tuberculosis developed in 24 of 84 (28%) patients with a positive PPD test result (10.4 per 100 person-years), in 19 of 151 (13%) patients with a negative PPD test and no anergy (5.4 per 100 person-years), and in 20 of 112 (18%) anergic patients (12.4 per 100 person-years). The risk for developing tuberculosis was similar between PPd-positive and anergic patients (P > 0.2) but was significantly higher in both groups than in nonanergic, PPd-negative patients (P = 0.005 and P = 0.01, for PPd-positive and anergic patients, respectively). The median time from the performance of delayed hypersensitivity skin testing to the development of active tuberculosis was 22 months (range, 3 to 51 months) for PPd-positive patients, 11 months (range, 2 to 46 months) for nonanergic, PPd-negative patients, and 7.5 months (range, 2 to 38 months) for anergic patients.

If the analysis is restricted to the 49 patients with microbiologically confirmed tuberculosis, the risk for development of active infection in PPd-positive patients; nonanergic, PPd-negative patients; and anergic patients is 10.4, 4.5, and 8.1 per 100 person-years, respectively.

When stratified by risk practices for HIV infection, intravenous drug users had positive results of tuberculin skin tests more frequently than did patients in other HIV-transmission risk categories (31% compared with 13%, respectively; P = 0.03), as well as a significantly higher risk for developing tuberculosis (see Table 2). Sixty-three (22%) of 290 intravenous drug abusers developed tuberculosis, compared with none of 27 homosexual patients and 30 patients with other risk factors.

A total of 103 patients who had a negative reaction to the PPD test at the first evaluation had new evaluations at varying periods after entering the study. Among 67 nonanergic, PPd-negative patients, 10 (15%) showed PPD conversion from negative to positive, and tuberculosis developed in 3. The PPD converted in 3 of 36 (8%) anergic patients, and tuberculosis developed in 2 of them. The time between the last negative PPD result and the first positive PPD result was at least 1 year in all but one patient.

Among the 108 patients with previously positive PPD test results, 24 (22%) completed a 9- to 12-month course of isoniazid prophylaxis after entering the study, as did 3 (23%) of 13 patients whose PPD converted from negative to positive. No patient had a documented history of having received isoniazid before the date of inclusion. The median length of observation after completing the prophylaxis treatment was 22 months (range, 2 to 46 months). Active tuberculosis developed in only 1 (3.7%) of the 27 patients who had received isoniazid prophylaxis at month 28 of follow-up, compared with 29 (31%) of 94 patients with positive PPD test results who had not received the drug (P = 0.008).

Of the 64 patients who developed tuberculous disease, 40 (62%) had CD4 cell counts done within 3 months of the diagnosis of tuberculosis. Median CD4 cell count was 216 cells/mm³ in 14 PPd-positive patients, 196 cells/mm³ in 11 nonanergic, PPd-negative patients, and 62 cells/mm³ in 15 anergic patients. Fourteen (93%) patients in the anergic group had fewer than 500 CD4 cells/mm³, compared with 11 (78%) PPd-positive patients and 8 (73%) nonanergic, PPd-negative patients (differences not significant, Fisher exact test).

Discussion

Top Methods Results Discussion Author & Article Info References

This study was conducted in an area with a high prevalence of tuberculosis and was designed to assess the risk for the disease to develop among HIV-infected patients with known delayed-type hypersensitivity responsiveness. We found a high rate of positive PPD test results (29% of all study participants, 41% of patients who were not anergic), as well as a high incidence of tuberculosis (9% per year). All three groups considered in the analysis (PPD positive; nonanergic, PPD negative; and anergic patients) have been shown to be at high risk for developing the disease in our area. These findings underscore the need for establishing preventive measures among HIV-infected patients who live in areas where tuberculosis is prevalent.

The cumulative incidence of active tuberculosis in a prospective study among seropositive intravenous drug users with positive tuberculin skin test reactions in New York was 8% per year [7], similar to that found in this study (10.4% per year). This high risk for re-activation of latent tuberculous infection in HIV-infected patients justifies current recommendations that chemoprophylaxis treatment be considered for all HIV-infected patients who show a positive ( 5 mm) reaction to a PPD test [13].

Previous reports have described the rapid progression of newly acquired tuberculous infection in HIV-infected patients. Of 18 HIV-infected patients exposed to M. tuberculosis, Di Perri and colleagues [5] reported the rapid development of tuberculosis in 8 patients with negative tuberculin skin test results. Similarly, Daley and colleagues [6] described an outbreak of tuberculosis in a residential facility for HIV-infected persons, where tuberculosis developed in 37% of the residents exposed to the infection. Progression of newly acquired tuberculosis is the most probable mechanism to explain the high rate of active infection observed among our patients with negative PPD test results and otherwise preserved delayed-type hypersensitivity responsiveness (5.4% per year). The high rate of PPD conversion in this study (15% among nonanergic patients) is consistent with the high prevalence of tuberculosis in our area, especially among prison inmates and patients from lower socioeconomic groups. These findings suggest the need for aggressive tuberculosis control procedures for HIV-infected patients who may have been in contact with recently diagnosed patients with tuberculosis as well as in prisons and other closed institutions where spread is likely.

A high incidence of tuberculosis among HIV-infected patients who are anergic on multiple skin tests has been found as well (12.4% per year). No previous studies have directly addressed the risk for developing tuberculosis among anergic HIV-infected patients, although some reports offer indirect information. Selwyn and coworkers [7] reported the development of tuberculosis in only 1 of 166 HIV-infected intravenous drug users with negative tuberculin skin tests, with an estimated risk of 0.3% per year. However, the two studies are not strictly comparable. Selwyn's study included intravenous drug users enrolled in a methadone-maintenance program; skin anergy was documented in only 17 individuals (only 64 had skin anergy testing evaluated); and information about the degree of immunosuppression was lacking. The patients in our hospital-based study had HIV-related symptoms or complications and were probably at more advanced states of immunosuppression, which could explain the higher incidence of tuberculosis among anergic patients. In another study from Spain [14], tuberculosis developed in approximately 5% of participants with negative PPD test results, followed for a median of 14 months, but no specific data were presented regarding the risk among patients who were anergic.

Our findings support recommendations of testing all patients who are infected with HIV for skin anergy and call for a consideration of preventive therapy in anergic, HIV-infected patients who live in a high-prevalence area or belong to high-risk groups. Studies documenting the development of tuberculosis in areas or groups with various prevalence rates of tuberculous infection are necessary to establish definite recommendations. On the basis of a decision analysis model of isoniazid as preventive therapy in HIV-infected intravenous drug users [15], the Division of Tuberculosis Elimination of the Centers for Disease Control and Prevention recommends that preventive therapy should be considered for anergic persons in groups in whom the prevalence of tuberculous infection is 10% or more [16].

Most experts agree that preventive therapy with isoniazid be recommended when a positive tuberculin test result is documented [13, 17]. The situation should not differ in the case of anergic patients if chemoprophylaxis is to be given. Latent tuberculous infection may never be detected after HIV-induced anergy has developed. It seems reasonable to initiate prophylaxis in anergic patients, regardless of their CD4 count. Potential advantages of this approach include fewer interactions with drugs used later in the course of HIV infection and a better rate of compliance in patients who take fewer drugs.

It can be argued that reactivation of tuberculous infection would be possible in patients who take the drug at relatively early stages of HIV infection: Isoniazid is unlikely to eradicate M. tuberculosis and the risk for reactivation of the disease increases with the degree of immunosuppression. On the basis of these arguments, some investigators favor delaying initiation of preventive therapy until a certain degree of immunosuppression has been reached (a CD4 count less than 500 cells/mm³ has been suggested) [18]. Our finding that approximately 25% of HIV-infected patients with positive PPD test results develop tuberculosis with a CD4 count greater than 500 cells/mm³ militates against this recommendation. All but one anergic patient in this study had fewer than 500 CD4+ cells/mm³ when developing tuberculosis (approximately 10% of anergic patients in this study had more than 500 CD 4⁺ cells/mm³. Data not shown). The possibility of delaying initiation of preventive therapy in these patients will depend on whether the theoretic risk for reactivating the disease after chemoprophylaxis therapy outweighs the potential benefits of early administration of the drug. Failure of tuberculous preventive therapy does not seem to be prevalent in HIV-infected patients [7, 19]. Although this study was not designed to evaluate the effectiveness of preventive therapy, 9 to 12 months of isoniazid administration seems to be effective in preventing the development of tuberculosis among our patients coinfected with HIV and M. tuberculosis: only 1 of 27 patients developed tuberculosis. Patients with positive tuberculin skin test results who completed a course of isoniazid prophylaxis developed tuberculosis less frequently than did those who did not take isoniazid.

Development of tuberculosis was significantly more frequent in intravenous drug users than in patients with other risk factors for HIV infection. Despite a high rate of positive PPD test results among patients who did not use intravenous drugs, all cases of active infection among PPd-positive patients, as well as among nonanergic, PPd-negative patients, occurred in intravenous drug users in this study. This relationship has long been recognized and, in countries with a low prevalence of tuberculosis such as the United States, intravenous drug users are a risk group for whom control programs should be targeted [20].

Presented at the 32nd Interscience Conference on Antimicrobial Agents and Chemotherapy, 11-14 October 1992 in Anaheim, California.