Selection of Patients for Prophylaxis

In 1990, the American Thoracic Society and the Centers for Disease Control and Prevention (CDC) recommended that the definition of a reactive PPD test be adjusted according to the likelihood of exposure and the ability to mount DTH [4]. For persons infected with HIV, the proposed cutoff value of 5 mm of induration has been supported by studies showing that using a lower threshold of PPD reactivity, as evidence of tuberculous infection, preserves sensitivity [5, 6]. In 1991, the CDC further advised that 1 year of isoniazid prophylaxis be considered for HIV-infected patients who are either PPD positive ( 5 mm) or who are anergic to control recall antigens and who come from a group with 10% prevalence of active or latent tuberculous infection [7]. This recommendation to treat anergic patients is supported by a decision analysis [8] and a prospective study of intravenous drug users in New York City. The drug users had a 20% prevalence of tuberculin reactivity and 9.7 active tuberculosis cases/100 patient-years were observed in those who did not receive isoniazid [9].

The study of intravenous drug users in Madrid, in this issue of Annals [1], describes the striking vulnerability to tuberculosis (>10 cases/100 patient-years) of HIV-infected patients whether they are PPD positive or anergic [7]. Moreno and colleagues [1] also found a relatively high rate of active tuberculosis in PPd-negative, nonanergic patients (5.5 cases/100 patient-years), a group that might be expected to be at much lower risk. This finding suggests that either the 5-mm cutoff of PPD reactions misses some patients with latent infection or that exogenous infection is common.

Whether infection is exogenous or results from reactivation is of practical relevance, because it determines the duration of prophylaxis necessary to protect HIV-infected patients over long periods. If reactivation of latent infection accounts for most disease, then 1 year of prophylaxis should be adequate. The importance of reactivation is supported by the apparent lasting effectiveness of 1 year of isoniazid in preventing active disease in HIV-infected patients; however, only limited, retrospective data are available [1, 9]. On the other hand, if exogenous infection predominates, then long-term prophylaxis may be required, because re-exposure may be frequent in areas with high prevalence of active tuberculosis. Regardless of the mechanism, the high incidence of active tuberculosis in PPd-negative, nonanergic persons suggests that all HIV-infected persons from groups with a high prevalence of tuberculous infection should receive prophylaxis [8]. Two emerging issues may require changes in this recommendation. First, the prevalence of isoniazid-resistant Mycobacterium tuberculosis is increasing; therefore, new prophylactic regimens must be devised and tested. Second, rifabutin is being used for prophylaxis of Mycobacterium avium infection. Although rifabutin has in-vitro activity against M. tuberculosis [10], its clinical effectiveness as single-agent prophylaxis has not yet been shown.

Prophylaxis for PPd-negative and nonanergic patients may not be justified in some groups of patients such as urban gay men in developed countries. Markowitz and colleagues [2] report, in this issue of Annals, a study of PPD reactivity and DTH in a group of predominantly homosexual men in the United States. The overall prevalence of PPD reactivity in HIV-negative controls was 10% (6.8% in homosexual men and 19.1% in intravenous drug users). More than one half of these homosexual patients were PPD negative and nonanergic. Most were not infected with M. tuberculosis and, thus, can be spared unnecessary treatment. This approach, which is concordant with current CDC recommendations, needs prospective evaluation with provision for regular DTH and PPD testing.

The management of PPd-negative patients who convert from DTH reactivity to anergy is unclear. Such patients may well be at lesser risk for active tuberculosis than those with anergy at first evaluation, but this assumes that most tuberculosis in HIV-infected patients is due to reactivation. On the other hand, HIV-induced anergy probably increases susceptibility to exogenous infection.

Active tuberculosis develops over a broader range of immunosuppression than most other opportunistic infections. What value then does CD4 enumeration have in interpreting PPD testing? Anergy to control recall antigens and PPD is more common with advanced immunosuppression, but PPD reactivity is preserved in some infected patients even at low CD4 counts. Thus, we suggest that PPD testing should begin as early as possible, should be repeated at least annually, and should not be omitted in patients with AIDS just because reactions are expected to be lower.

Prognostic Value of Delayed-Type Hypersensitivity Testing

Delayed-type hypersensitivity testing to recall antigens has potential as a simple, inexpensive, functional assay of cell-mediated immunity (although it does require a return visit in 48 to 72 hours and is not easily quality controlled). The loss of DTH might reflect the progression from HIV infection to AIDS. To understand why, it is useful to review the immunology of the DTH response. Biopsies of tuberculin reactions showed accumulations of polymorphonuclear cells followed by lymphocytes and macrophages [11]. Vascular leakage of plasma proteins and the subsequent formation of an intradermal fibrin clot accounts for most of the induration, which constitutes the measurable reaction [12]. The formation of fibrin is initiated by a unique lymphokine that induces macrophages to express tissue factor and thereby cause clotting [13, 14]. This lymphokine is produced by type 1 CD4 T cells [15] that are known to control DTH responses in rodents [16]. Thus, DTH may reflect both the number and functional capacity of the type 1 CD4 subset of T lymphocytes.

Although the CD4 cell count has been included in the current definition of AIDS, its use for the prediction of opportunistic infections and for the evaluation of therapies is increasingly controversial. For example, in a recent placebo-controlled European study (the "Concorde" trial), patients treated immediately with zidovudine (AZT, Retrovir) had a slower decrease in CD4 cells over time compared with control patients receiving delayed treatment, but the clinical outcomes were similar [17]. One explanation of the apparent lack of clinical benefit despite increased numbers of CD4 cells is variation in the functional capacities of these cells. Delayed-type sensitivity might detect differences in function not reflected in CD4 cell counts.

The role of DTH in prognostic staging of HIV infection is addressed by Blatt and colleagues [13], also in this issue of Annals. The reaction to four recall antigens was classified into zero (anergy), one (partial anergy), or two or more (reactivity) and was used to predict the risk for AIDS during the 5 years after the initial assessment of asymptomatic U.S. Air Force personnel. Delayed-type hypersensitivity reactions estimated the risk for progression better than CD4 or Walter-Reed staging when all three were included in a proportional-hazards model. Before DTH testing is recommended for staging HIV disease, other, well-established and more easily measured prognostic markers, such as serum ß-2 µglobulin, should be compared with DTH in future studies.

Why might loss of DTH predict progression of HIV infection? During the period of clinical latency, when prediction is most difficult, cell-mediated immune responses control viral replication at levels that produce only slow erosion of the number and function of CD4 cells. Progression to advanced immunosuppression is characterized by loss of control of viral replication by CD8 lymphocytes, increased plasma viremia, and more rapid loss of CD4 cell numbers in lymph nodes and in blood. Clerici and Shearer [18] have proposed that this deterioration reflects a shift from a type 1 to a type 2 CD4 lymphocyte response to HIV. Type 1 CD4 cells secrete lymphokines (for example, interleukin-2 and -interferon) that support both cytotoxic and DTH responses, whereas type 2 CD4 cells produce lymphokines (for example, interleukin-4 and interleukin-10) that promote antibody synthesis and suppress DTH. Anergy may signal the early phases of this shift that indicate a poor prognosis. Thus, monitoring DTH to control and HIV-related antigens in clinical trials of antiretroviral agents and vaccines could provide a useful measure of their effect.

5 Graham NM, Nelson KE, Solomon L, Bonds M, Rizzo RT, Scavotto J, et al. Prevalence of tuberculin positivity and skin test anergy in HIV-1-seropositive and -seronegative intravenous drug users. JAMA. 1992; 267:369-73.