Immunotherapy for Infectious Diseases

286 Wallis and Johnson
Several studies have indicated compartmentalization of T-cell responses at the site of
disease (75–78). In addition, intrinsic T-cell refractoriness, possibly associated with a
tendency toward apoptosis (programmed cell death), may be present in the peripheral
blood (79).
Immunologically mediated tissue damage and other toxicities also appear to be
responsible for many of the clinical manifestations of TB. TNF-� appears to be the
cause of much of the fever, wasting, inflammation, and tissue necrosis characteristic of
the disease.
HIV AND TUBERCULOSIS
Coinfection with HIV is the most potent risk factor for active TB in a person latently
infected with M. tuberculosis. TB typically is an early complication of HIV infection,
occurring prior to an AIDS-defining illness in 50–67% of HIV-infected patients (80).
Before the introduction of protease inhibitors, the diagnosis carried an expected
mortality of 21% at 9 months, even in those subjects presenting without other AIDSdefining
conditions (81). Death was infrequently (13%) due to active TB, however.
More often, it resulted from other AIDS-related causes (particularly Pneumocystis
carinii or bacterial pneumonia, or wasting syndrome), which may occur shortly after
the diagnosis of tuberculosis.
Several studies indicate that the adverse interactions of M. tuberculosis and HIV are
bidirectional, i.e., that TB affects HIV disease in addition to the better recognized converse
interaction. TB is characterized by prolonged antigenic stimulation and immune
activation, even in HIV-positive subjects (79,82). Antigen-induced T-cell activation and
expression of the proinflammatory cytokines TNF-� and other inflammatory cytokines
in turn promote HIV expression by latently infected cells (83–89). M. tuberculosis and
its proteins and glycolipids directly stimulate HIV replication by mechanisms involving
monocyte production of TNF-� (90–93). In the lung, TNF-� and HIV-1 RNA are
both increased in bronchoalveolar lavage fluid of involved segments of lungs of patients
with pulmonary TB and HIV-1 infection (94). Phylogenetic analysis of V3 sequences
demonstrated that HIV-1 RNA present in bronchoalveolar fluid had diverged from
plasma, indicating that pulmonary TB enhances local HIV-1 replication in vivo. In this
context, IL-10 and TGF-� expression may be of benefit to the host, in that they inhibit
antigen-induced HIV expression, via inhibitory effects on lymphocyte activation (88).
These interactions appear to have significant clinical consequences. Plasma HIV
viral load increases 5- to 160-fold in HIV-infected persons during the acute phase of
TB (95). Subsequently, new AIDS-defining opportunistic infections occur at a rate 1.4
times that of CD4-matched HIV-infected control subjects without a history of TB (95%
confidence interval: 0.94–2.11) (96). Cases also had a shorter overall survival than did
controls ( p � 0.001), as well as an increased risk for death (odds ratio � 2.17). The
adverse effect on survival is most pronounced in those individuals with the greatest evidence
for macrophage activation (neopterin � 14 ng/mL, or soluble type II TNF-�
receptor � 6.5 ng/mL, or negative tuberculin skin tests; p � 0.01) (97). Thus, although
active TB may be an independent marker of advanced immunosuppression in HIVinfected
patients, it may also act as a cofactor to accelerate the clinical course of HIV
infection.