Immune Restoration Disease in HIV-Infected Patients after Antiretroviral Therapy

Sir—We were interested to read the paper on “immunorestitution disease” by Cheng et al. [1], since we, too, have argued for several years that infectious and/or inflammatory disease in HIV-infected patients who are responding to antiretroviral therapy (ART) is a manifestation of immune restoration. When zidovudine monotherapy was introduced into clinical practice, our group observed that Mycobacterium avium complex (MAC) disease developed in a small proportion of patients during the first few weeks of therapy and that this development was associated with the restoration of a delayed-type hypersensitivity (DTH) response to mycobacterial antigens [2]. We argued that this reflected restoration of an immune response against subclinical MAC infection [3]. Disease presentation was atypical; in particular, the infection was usually localized to tissues rather than disseminated. After the introduction of highly active ART (HAART), we again observed cases of MAC disease presenting in a similar way, and we also observed infectious and/or inflammatory disease caused by other pathogens [4–6]. As argued elsewhere [4, 7], we believe that these disease events also reflected the restoration of pathogen-specific immune responses, and we have proposed the term “immune restoration disease” (IRD) to designate them [4, 7–9]. We agree with Cheng et al. that IRD must be studied further, if diagnostic methods and clinical management are to be improved. In particular, the immunopathogenesis of IRD should be clarified, because it may be different for different pathogens. Immune reconstitution may be a factor, but it is notable that disease episodes usually occur during the first 12 weeks of therapy [4, 7], before substantial immune reconstitution has taken place. Redistribution of antigen-specific T cells (reviewed in [10]) and/or reversal of T cell immunosuppressive factors, such as the effect of Th2 cytokines [11], might have a more immediate effect. Disease episodes are associated with the restoration of pathogen-specific immune responses, as exemplified by MAC or Mycobacterium tuberculosis IRD (often referred to as paradoxical reactions), and possibly by cytomegalovirus or hepatitis B and C virus IRD (reviewed in [7]). Therefore, pathogen-specific immune responses, may be the critical factor and could occur without substantial changes in CD4 T-cell counts. It is likely that other factors are also involved, since many patients experience restoration of pathogen-specific immune responses, as indicated by their ability to cease prophylactic therapy for opportunistic infections [10], without the occurrence of IRD. Preliminary evidence from our studies suggests that proinflammatory cytokines may be implicated (S.F. Stone, P. Price, N.M. Keane, R.J. Murray, M. A. French, unpublished data). Furthermore, we have demonstrated that particular HLA haplotypes correlate with herpesvirus IRD [12] which suggests that immune mechanisms are different for different pathogens. A case of IRD that develops after a patient had begun receiving antiretroviral therapy may be misinterpreted either as an opportunistic infection resulting from persistent immunodeficiency or as the effect of drug toxicity. In either case, therapy might be stopped prematurely. It is therefore important to establish diagnostic criteria and methods of managing this complication of antiretroviral therapy. It is also important to recognize that mycobacterial IRDs may be a complication of ART consisting of only 1 or 2 nucleoside analogues [7] and that they may become a disease-management problem in developing countries that have a high prevalence of mycobacterial coinfections.