Skip Navigation

Immune Reconstitution in HIV-Infected Patients

  1. Kenneth H. Mayer, Section Editor
  1. Hans H. Hirsch1,2,
  2. Gilbert Kaufmann1,
  3. Pedram Sendi1,3, and
  4. Manuel Battegay1
  1. 1Division of Infectious Diseases, University Hospital Basel, Switzerland
  2. 2Institute for Medical Microbiology, University Hospital Basel, Switzerland
  3. 3Institute for Clinical Epidemiology, University Hospital Basel, Switzerland
  1. Reprints or correspondence: Dr. Manuel Battegay, Div. of Infectious Diseases, University Hospital Basel, Petersgraben 4, 4031 Basel, Switzerland (mbattegay{at}uhbs.ch).

Abstract

The prognosis of patients infected with human immunodeficiency virus (HIV) type 1 has dramatically improved since the advent of potent antiretroviral therapies (ARTs), which have enabled sustained suppression of HIV replication and recovery of CD4 T cell counts. Knowledge of the function of CD4 T cells in immune reconstitution was derived from large clinical studies demonstrating that primary and secondary prophylaxis against infectious agents, such as Pneumocystis jirovecii (Pneumocystis carinii), Mycobacterium avium complex, cytomegalovirus, and other pathogens, can be discontinued safely once CD4 T cell counts have increased beyond pathogen-specific threshold levels (usually >200 CD4 T cells/mm3) for 3–6 months. The downside of immune reconstitution is an inflammatory syndrome occurring days to months after the start of ART, with outcomes ranging from minimal morbidity to fatal progression. This syndrome can be elicited by infectious and noninfectious antigens. Microbiologically, the possible pathogenic pathways involve recognition of antigens associated with ongoing infection or recognition of persisting antigens associated with past (nonreplicating) infection. Specific antimicrobial therapy, nonsteroidal anti-inflammatory drugs, and/or steroids for managing immune reconstitution syndrome should be considered.

The hallmark of HIV disease is the continuous depletion of CD4 T cells, leading to progressive immunodeficiency, opportunistic diseases, and death [1]. Despite the fact that CD4 T cells function at a pathogen-specific level that is not yet assessable in the clinical routine, a clear inverse relationship exists between the number of CD4 T cells in the patient's peripheral blood and their risk of opportunistic infection. Thus, in asymptomatic HIV-infected patients, total CD4 T cell counts provide a major criterion for initiating primary prophylaxis against opportunist illness, such as Pneumocystis jirovecii (also known as Pneumocystis carinii) pneumonia (if the CD4 T cell count is <200 cells/mm3) and disseminated Mycobacterium avium complex infection (if the CD4 T cell count is <50 cells/mm3) [2]. Conversely, an increase in CD4 T cell count after the suppression of HIV replication by antiretroviral therapy (ART) correlates with improved T cell responses to antigens and mitogens [3, 4] and adequate protection against specific opportunistic infections, allowing discontinuation of primary and secondary prophylaxis in clinical practice [59]. Thus, plasma HIV load and peripheral CD4 T cell count represent surrogate markers for assessing the risk of AIDS, as well as the need for ART and/or prophylaxis [7, 10].

Recovery of CD4 T Cell Counts

Recent studies indicate that CD4 T cell recovery is sustained for ⩾3–4 years in individuals with advanced HIV-1 infection who receive ART [11, 12]. After an initial rapid increase in CD4 T lymphocyte counts within the first 3–6 months of ART, a second phase of slower increase occurs in most patients [13]. However, considerable individual variation has been noted in this response. In fact, CD4 T cell counts may remain below the critical threshold of 200 cells/mm3 in 10%–20% of treated individuals [12]. Several factors are associated with impaired CD4 T cell reconstitution, including older age, nonsuppressed HIV replication, and treatment interruption [7, 12]. In general, the lower the CD4 T cell count at the initiation of ART, the longer it takes to reach desired target levels [14, 15]. Certain investigations even suggest that CD4 T cell recovery may level off before the physiological range is reached, particularly in individuals who begin ART when their CD4 T cell counts are extremely low (i.e., <50 cells/mm3) [15, 16]. Conversely, early initiation of ART during primary HIV-1 infection may not only preserve but also rapidly restore memory and naive CD4 T cells, which are crucial for the rate and the range of the adaptive immune response [15]. These observations should be taken into account when counseling the patient with respect to the start of ART [1719].

Recovery of Pathogen-Specific CD4 T Cell Function

In the late 1990s, it became evident that ART resulted not only in a numerical increase in CD4 T cells, but also in a functional immune reconstitution [3, 4]. Qualitative and quantitative recovery of pathogen-specific cellular and humoral responses has been demonstrated for a number of agents, including mycobacteria, cytomegalovirus, Epstein-Barr virus, hepatitis B virus (HBV) and hepatitis C virus (HCV), and Candida albicans [3, 2022]. Improvements in pathogen-specific immune responses to HBV and HCV are illustrated by seroconversion to anti-HCV-positive or anti-Hbe-positive states and by clearance of hepatitis [2325]. With the possible exception of delayed-type hypersensitivity skin testing for particular antigens, the more sophisticated assays that measure cytokine concentration, lymphoproliferation, specific T cell receptors, or cytotoxicity have not yet entered clinical routine and/or are lacking clinical validation. In the clinical arena, randomized, open clinical studies, as well as prospective cohorts, have demonstrated that stopping primary prophylaxis for P. jirovecii is safe in patients with CD4 T cell counts that are stably reconstituted to >200 cells/mm3 for >6 months [9, 2631]. This same conclusion was reached for M. avium complex in patients with CD4 T cell counts of >100 cells/mm3 for >3 months, according to 2 randomized, blinded trials and a cohort study in which disease occurred in only 2 of 863 patients during a follow-up period of ⩾1 year [3234].

Immunologically and clinically more challenging is the discontinuation of secondary prophylaxis, particularly when it concerns pathogens with a propensity for latent infection in vulnerable sites (e.g., cytomegalovirus or Toxoplasma gondii, which have a propensity for latent infection in the CNS). For P. jirovecii, Lopez et al. [30] demonstrated in a prospective trial involving 113 individuals with a median follow-up period of 12 months that interruption of secondary prophylaxis is safe. No case of P. jirovecii pneumonia was observed among the 60 patients who had CD4 T cell counts of >200 cells/mm3 for >3 months and virus loads of <5000 copies/mL (no incidents per 100 person-years; 95% CI, 0–4.57 ) [30]. Similar results were obtained from a large observational study [35] that analyzed the data of 8 European HIV databases (no incidents per 100 person-years; 95% CI, 0–1.3). For Cryptococcus neoformans, a prospective randomized trial confirmed that discontinuation of secondary prophylaxis after adequate treatment for cryptococcal meningitis is safe in patients who had CD4 T cell counts of >100 cells/mm3 and HIV loads of <50 copies/mL for >3 months [36].

For cytomegalovirus retinitis, observational studies and case series indicate that secondary prophylaxis can be safely discontinued provided that the disease is determined to no longer be active by repeated ophthalmologic examination and a sustained CD4 T cell increase to levels ranging from >100 cells/mm3 to >150 cells/mm3 for >6 months has been documented [3743]. In these studies, the inclusion criteria regarding CD4 T cell thresholds and/or increases varied considerably, from 50–150 cells/mm3, with nadirs ranging from 10–42 cells/mm3. However, there were only 2 cases of cytomegalovirus disease among >200 patients studied [3743]. In the comprehensive study by Kirk et al. [43], interruption of maintenance therapy against previous infection with 4 common HIV-associated opportunistic pathogens was investigated. In this observational study, 358 patients who were receiving ART and who had CD4 T cell counts of >50 cells/mm3 interrupted secondary prophylaxis. A total of 379 episodes were analyzed (162 episodes of cytomegalovirus disease, 103 episodes of M. avium complex infection, 75 episodes of toxoplasmosis, and 39 episodes of cryptococcosis). Only 5 patients experienced a relapse during 781 person-years of follow-up; 2 patients had cytomegalovirus disease, 2 patients had M. avium complex infection, and 1 patient had toxoplasmosis [43]. Few of these events occurred at CD4 T cell counts greater than the respective thresholds. The overall incidence of recurrent disease was calculated per 100 person-years as follows: cytomegalovirus, 0.54 (95% CI, 0.07–1.95); M. avium complex, 0.90 (95% CI, 0.11–3.25); toxoplasmosis, 0.84 (95% CI, 0.02–4.68); and cryptococcosis, 0.0 (95% CI, 0.00–5.27) [43]. Discontinuation of antifungal therapy for disseminated histoplasmosis after ART was investigated in 32 patients with CD4 T cell counts of >150 cells/mm3 (median CD4 T cell count, 289 cells/mm3). No relapse of disseminated histoplasmosis was observed (95% CI, 4.6 incidents per 100 person years) [44], indicating that discontinuation of antifungal treatment for disseminated histoplasmosis is safe in this setting. In summary, the current recommendations for discontinuing secondary prophylaxis require completion of initial therapy, inactive disease, and stable CD4 T cell counts greater than pathogen-related thresholds. For patients with P. jirovecii infection and toxoplasmosis, a CD4 T cell count of >200 cells/mm3 for ⩾3 months is recommended; for patients with M. avium complex infection, cytomegalovirus disease, and cryptococcosis, a CD4 T cell count of >100 cells/mm3 for ⩾6 months is recommended (guidelines are available at http://www.aidsinfo.nih.gov/guidelines/). However, CD4 T cell counts have to be monitored so that prophylaxis can be reinitiated if decreasing values are noted. Although there are no data specifically addressing this issue, reinitiation of prophylaxis uses the same thresholds as primary prophylaxis. As a note of caution, CD4 T cell thresholds are not absolute, but are only intended to provide guidance. This is illustrated by data from the Swiss HIV Cohort Study of patients with toxoplasmic encephalitis. Of these patients, 7% were determined to have had CD4 T cell counts of >200 cells/mm3 (which is the threshold recommended by the Swiss guidelines), and 19.4% were determined to have had CD4 T cell counts of >100 cells/mm3 (which is the threshold recommended by the US guidelines) [45].

Immune Reconstitution Syndrome (IRS)

IRS is recognized as a potential complication that can occur after potent ART [46]. The frequency of IRS has not been reported conclusively, but it may be estimated to occur in 10%–25% of patients who receive ART [47]. However, the first cases of M. avium-associated IRS were reported to have occurred after administration of zidovudine monotherapy [48]. Mycobacterial antigens are frequently implicated in IRS, and together they represent almost one-third of all reported cases [20]. Other infectious agents that are frequently associated with IRS include C. neoformans, cytomegalovirus, and HBV and HCV. More recently, it has been proposed that HIV-mediated IRS mediates a novel form of severe demyelinating leucoencephalopathy in the CNS [49]. Also, a role for BK virus-targeted IRS might account for cases of hemorrhagic cystitis, as discussed elsewhere [50, 51]. Table 1 summarizes manifestations associated with different microbes and the reported management of those manifestations.

Table 1

Review of immune reconstitution syndrome, by pathogen.

The onset of IRS is characterized by paradoxical worsening of clinical or laboratory parameters despite “favorable” development of the HIV surrogate markers [63]. In the affected patients, CD4 T cell counts are often <50 cells/mm3 at the start of ART and subsequently increase >2–4 fold during the 12 months after initiation of ART, and a significant decrease in HIV load of >2 log10 copies/mL is often noted [4, 20]. However, initial experience of IRS occurring after zidovudine monotherapy suggests that small breaks of HIV replication may already be sufficient for IRS to occur in response to some particular antigens [48]. The interval between the start of ART and the onset of IRS is highly variable, ranging from <1 week to several months, but the majority of events occur within the first 8 weeks after ART initiation [4, 20, 47] and occur in patients with low mean CD4 T cell counts [47]. In vitro characterization suggested that reconstitution occurred earlier in response to antigens of ubiquitous or abundant exposure, compared with antigens of low exposure (such as tetanus toxoid), and that earlier reconstitution correlated with an increase in memory CD4 T cells [3, 91]. Of note, HIV-specific immunity is not adequately reconstituted in most patients [92], a fact underscored by recent reports of HIV superinfection [9395].

IRS may arise in 2 different settings, depending on whether ART was started in a patient treated for ongoing opportunistic infection or in a clinically stable patient with or without requiring primary prophylaxis. For patients with active infection and antimicrobial treatment for P. jirovecii or Mycobacterium tuberculosis, for example, some clinicians may choose to defer ART by 1–3 weeks to avoid high pill counts, cumulating toxicity, and possibly IRS. If IRS occurs in such patients, the differential diagnosis includes failure of the antimicrobial therapy; manifestation of a new opportunistic infection; unmasking of an ongoing, previously undiagnosed infection; or manifestation of a diagnosed, ongoing infection in a previously unrecognized site of involvement. Although discontinuing ART has to be considered (table 1), most clinicians prefer to continue ART if the CD4 T cell counts are <100 cells/mm3 or if IRS manifests months after ART has been initiated. There are no clear guidelines on when to stop ART. As gathered from anecdotal reports, ART is more readily discontinued if inflammatory responses are considered life-threatening, the responses are not amenable to steroids, and the pathogens involved are not controllable by specific antimicrobial treatments (as exemplified by IRS in response to parvovirus B19 or polyomavirus JC [78, 96]), or if ART toxicity was the main differential diagnosis (e.g., as in hepatitis).

An increase in transaminase levels after the initiation of ART is observed in 1%–5% of patients and is associated with coinfection with HBV and/or HCV [9799]. A recent study of 915 adverse events that occurred among 755 patients who were receiving ART identified severe hepatotoxicity in 26 (3.4%) of the patients (incidence, 4.2 adverse events per 100 person-years), with risk factors being injection drug use, chronic alcohol abuse, and chronic viral hepatitis (including HBV and HCV, as confirmed by histology) [100]. The incidence of liver failure was 1.1%. Liver failure was associated with a lower CD4 T cell count at baseline (median, 94 cells/mm3), compared with the CD4 T cell count at baseline of the 19 patients without liver failure (median, 260 cells/mm3), suggesting a role for IRS [100]. Drug hepatotoxicity is associated with ART regimens that contain nevirapine, efavirenz, didanosine, zalcitabine, ritonavir, or saquinavir [97, 101]. However, discontinuation of ART may be associated with rebounding HBV replication in patients coinfected with HIV-HBV who are treated with lamivudine- or tenofovir-containing regimens. In these cases, continuation of monotherapy with lamivudine or tenofovir should be considered. On the other hand, persistent HBV replication and the emergence of lamivudine-resistant HBV has been observed in ∼10% of patients after long-term ART with regimens containing lamivudine [102].

Specific treatments may be required for unusual manifestations of IRS (table 1), such as surgical intervention in a case of bowel perforation due to M. tuberculosis-mediated IRS [65]. Acute renal failure has been observed in a case of M. tuberculosis-associated IRS in a patient whose renal function was initially normal, but who had documented prior miliary spread and who had urine cultures positive for M. tuberculosis [103]. The unmasking, as a result of IRS, of previously undiagnosed ongoing infections, such as Cryptococcus-associated meningitis, has been previously described [68]. In fact, IRS in response to the simultaneous occurrence of 2 entities (i.e., cytomegalovirus and Cryptococcus species) has been reported [104].

The management of symptoms, particularly through the use of nonsteroidal anti-inflammatory drugs, has been recommended. Immune modulation is warranted, but specific drugs and protocols are lacking. Thalidomide, intravenous immunoglobulins, and steroids have been discussed [20]. Steroids represent an important, although double-edged, measure in IRS treatment. Systemic steroids are often considered when inflammatory damage at the site of involvement severely impairs organ function (e.g., eye, liver, brain, or lungs) and becomes life-threatening. For progressive multifocal leucoencephalopathy mediated by polyomavirus JC, transient improvement of the patient's condition on administration of steroids that then showed lethal progression has been reported. The unfavorable course may be the result of several factors, including the advanced stage of the disease, the patient's inability to mount a complete immune response, and the protracted administration of steroids (which adversely affected immune effectors), in addition to the lack of effective pathogen-specific antivirals. However, judicious use of steroids and further immune reconstitution by continued ART remain the only option [77, 105].

Regarding the management of IRS, the question of whether IRS is elicited by active infection or by persisting antigens arises (table 2). If active infection is suspected, antimicrobial therapy is warranted, and modification of treatment should be considered in cases in which the patient is already receiving antimicrobials. On the other hand, reconstitution of immune reaction to persisting antigens of microbes that no longer replicate may make antimicrobial therapy unnecessary and may, in fact, add to toxicity. In clinical practice, it may be impossible to dissect these entities, and hands-on antimicrobial therapy may be administered even if an ongoing infection was not proven and if the differential diagnosis includes antimicrobial treatment failure [8].

Table 2

Recommended management of immune reconstitution syndrome.

Conclusions

Immune reconstitution, marked by an increase in CD4 T cell count, is the most crucial process that takes place in a patient after initiation of ART. Success of ART translates into a sustained recovery of CD4 T cells, which is the primary surrogate marker used in clinical praxis. Recommendations for discontinuing prophylaxis involve well-defined CD4 T cell count thresholds. Nevertheless, clinical judgment may require adaptation of these recommendations, particularly in the first phase of immune recovery. Because IRS may be more frequent in patients with advanced HIV disease who are initiating ART, this complication of immune recovery may decrease in frequency if ART is initiated in patients whose CD4 T cell counts have not already decreased to <200 cells/mm3.

  • Received August 14, 2003.
  • Accepted December 15, 2003.

References

| Table of Contents