Immune Containment and Consequences of Measles Virus Infection in Healthy and Immunocompromised Individuals

The pathogenicity of measles virus (MV) is intimately linked to the immune status of the infected individual. Measles is typically a self-limiting disease; however, individuals who are immunocompromised (56, 79), malnourished (26, 78, 113), or at the extremes of age (23) are at increased risk for severe measles. At the present time, with human immunodeficiency virus (HIV) and immunosuppressive drug therapy affecting the immune competence of large numbers of individuals, MV is reemerging as an important pathogen worldwide. Elucidating the mechanism by which the immune system controls MV infection and prevents reinfection will be crucial for our understanding of disease pathogenesis and transmission as well as the development of novel MV vaccination strategies.

Despite reaching global measles vaccination coverage of >80% of individuals, MV remains the fifth leading cause of death and the most common cause of vaccine-preventable death in children under 5 years of age (84). An estimated 31 million cases of measles occurred globally in 2001, resulting in 777,000 deaths, the majority (452,000) of which occurred in sub-Saharan Africa (84). MV is one of the most contagious pathogens known to humans, and large measles outbreaks, facilitated by overcrowding in poor communities, continue to occur even in countries that have achieved high vaccine coverage.

MV is a 15-kilobase, enveloped, negative-strand RNA virus and a member of the Morbillivirus genus of the Paramyxoviridae family. MV infects only humans and nonhuman primates. It is a stable, monotypic virus, making it an excellent candidate for eventual eradication. In culture, MV can be adapted to grow in nonprimate cells, a process that has facilitated the attenuation of viral strains for the development of a safe vaccine. Measles is transmitted via direct person-to-person contact and causes a symptomatic viral prodrome culminating in the hallmark maculopapular measles rash. The measles rash is preceded by systemic viremia and lymphopenia, and clearance of the rash is followed by a transient suppression of T-lymphocyte responses that lasts several weeks, leaving the infected individual susceptible to other infections.

While nonspecific innate immune mechanisms may be important in the control of MV during the first days following infection, adaptive MV-specific immune responses mediate viral clearance and provide protection against subsequent MV infections. Natural MV infection generates long-lasting immunity that includes both MV-specific antibody (14) and memory T-lymphocyte (50, 123, 136) responses. Long-term protection from reinfection occurs without a requirement for reexposure (98).

This article reviews our current understanding of the immune control of MV and the consequences of MV infection on the immune system of the immunocompetent host. We then discuss the clinical consequences of MV infection in immunocompromised individuals. Finally, we describe recent studies that have raised questions regarding the outcome of MV infection in the immunocompromised host.

IMMUNE CONTROL OF MV

Humoral immunity.

Our understanding of the immune control of MV replication is evolving. Antibodies are certainly important in preventing MV infection, as MV-specific neutralizing antibody titers at the time of exposure to virus correlate with protection from disease (21). It is unclear, however, whether antibodies have a role in MV clearance after an infection is initiated. In 1956, Good and Zak described the clinical outcome of MV infection in children with congenital or acquired immune defects. In their report of these “experiments of nature,” children with cellular immune defects developed a particularly severe clinical syndrome, whereas children with hypogammaglobulinemia had a typical clinical course (37). These observations suggested a primary role for cell-mediated rather than humoral immunity in recovery from MV infections.

Recent studies with nonhuman primates have addressed the contribution of humoral immunity to the clearance of MV infection. Rhesus monkeys were depleted of B lymphocytes by monoclonal antibody infusion and infected with MV. These B-cell-depleted animals had a virologic and clinical outcome that was indistinguishable from that of healthy monkeys (100). This study suggests that humoral immunity has only a limited role in the clearance of replicating MV in the immunologically naïve individual. While neutralizing antibody is a strong predictor of protection from disease following MV exposure (21), there is not an established correlation of neutralizing antibody with clearance of virus (103). However, other recent in vivo and in vitro studies suggest that antibodies might contribute to the control of MV replication. In one study done to evaluate antibody-dependent cellular cytotoxicity in measles patients, the appearance of an antibody that mediates this function correlated with the end of infectious viremia (31). Moreover, the failure of an infected individual to mount a robust anti-MV antibody response is associated with a poor clinical prognosis (134). Finally, binding of antibody to infected cells in vitro downregulates intracellular viral replication (34, 35, 117). All of these observations support a role for antibodies in MV control.

Cellular immunity.

While these conflicting findings raise questions concerning the role of humoral immunity in recovery from MV infection, the importance of cellular immunity in the control of MV infection was implicated by the clinical observation of severe measles in children with cellular immunodeficiencies (37, 56). Cellular immunity, especially the CD8⁺ T-lymphocyte cytolytic response, has been associated with MV clearance in several studies (42, 53, 123). Levels of soluble CD8 and β2 microglobulin are increased in the plasmas of MV-infected individuals during the time of measles rash (42). An MV-specific memory CD8⁺ T-lymphocyte pool is established after infection (53, 87, 123), with specificity for several MV proteins (45, 54, 126). Most recently, CD8⁺ lymphocyte-depleted rhesus monkeys infected with MV developed viremia with both a peak and duration that were greater than those observed in immunocompetent macaques (101). Importantly, the magnitude and duration of this viremia in the CD8⁺ lymphocyte-depleted monkeys did not increase further when monkeys were depleted of both CD8⁺ and B lymphocytes (100).

Studies have also evaluated the role of the CD4⁺ T-helper cell response in MV immunity. CD4⁺ T lymphocytes are activated and expand during the effector phase of the immune response to MV (126, 131). Moreover, soluble CD4 is detectable in plasmas of infected individuals from the time of onset to several weeks after resolution of the rash (40). Major histocompatibility complex (MHC) class II-restricted T-lymphocyte proliferation in response to MV proteins has been demonstrated (13, 83, 104, 107, 126). However, a role for CD4⁺ effector T lymphocytes has not been established for the immune containment of MV. While a Th1 cytokine profile, with elevations in peripheral blood gamma interferon (IFN-γ) and interleukin-2 (IL-2), is apparent during the measles prodrome (41, 42), measles convalescence is characterized by elevations in Th2 cytokines, such as IL-4 and IL-5. These changes in the cytokine profile may last for several weeks and may, in part, account for the susceptibility of individuals to secondary infections following MV infection (43). Nevertheless, in spite of a virally induced cellular immune suppression, the otherwise healthy host is able to control MV viremia and clear the virus.

PROPOSED MECHANISMS OF MV IMMUNOSUPPRESSION

MV infection causes profound and prolonged immunosuppression characterized by suppression of lymphocyte proliferation. The amelioration of autoimmune diseases (63, 64, 138) has been noted following clinical measles, presumably as a result of generalized immune dysfunction. The extinction of the tuberculin skin reaction and the reactivation of clinical tuberculosis have both been described in this setting (23). Importantly, the majority of measles deaths are due to secondary infections that occur as a consequence of immune suppression (10, 77). Despite compelling evidence for immunosuppression following MV infection, the mechanisms underlying this immunosuppression continue to be poorly understood.

All viruses studied to date that cause immunosuppression infect cells of the immune system (74). Infectious MV can be demonstrated in peripheral blood mononuclear cells (PBMC) of infected individuals for 1 week or more following the appearance of the rash (96, 135). CD4⁺ and CD8⁺ T lymphocytes, B lymphocytes, and monocytes can be infected by MV in vitro (48, 55, 121), and monocytes comprise the major population of infected unstimulated cells (111). The spontaneous proliferation of mononuclear cells sampled from the peripheral blood of infected individuals occurs just before the rash appears and persists for several weeks (39, 42), providing multiple targets for productive MV infection (131). Nevertheless, MV infects only a small proportion of circulating PBMC in the infected host (49, 96, 116, 135), and the major cell type infected in vivo seems to be in the monocyte/macrophage lineage (29).

A profound lymphopenia occurs during the acute phase of MV infection, which equally affects CD4⁺ T lymphocytes, CD8⁺ T lymphocytes, and B lymphocytes (4, 39, 110, 133, 135). It has been suggested that direct viral killing of infected immune cells may occur through syncytium formation, leading to immune suppression (9), although an inhibition of T-lymphocyte proliferation in vitro is observed well before extensive syncytium formation in infected lymphocyte cultures (51, 112, 137). The lymphopenia associated with measles is probably not entirely explained by direct viral killing, as the number of circulating infected mononuclear cells is low, and therefore may reflect trafficking of lymphocytes from the periphery into secondary lymphoid tissue (88, 110) and/or bystander apoptosis (1, 110, 128). MV also infects primary and secondary lymphoid tissues, including the thymus. Infection of thymic epithelial cells leads to apoptosis of uninfected thymocytes, which may have a profound effect on T-lymphocyte function (5, 122). However, Zambian children with measles did not have depressed thymic output compared to healthy control children (102). This finding raises questions about the likelihood that depressed thymic function plays a central role in the immune dysfunction associated with MV infection.

A role for monocytes in MV-induced immune suppression was suggested by the observation that lymphocyte proliferation was more strongly suppressed in cultures enriched for monocytes than in unfractionated PBMC (112). Abnormalities in monocyte function due to MV infection have been demonstrated, including reduced tumor necrosis factor alpha production (38, 130) and suppression of IL-12 production (33, 57). Also, MV-infected dendritic cells lose the ability to stimulate allogeneic lymphocyte proliferation and rapidly undergo apoptosis (33, 44).

There is much evidence to suggest that MV has direct effects on lymphocyte function. MV infection suppresses T-lymphocyte proliferation to mitogens (71, 140), IL-2 responses (12, 39, 129), and associated downstream signaling (7, 8), antigen presentation (69), and cytotoxic function (17). MV-induced suppression of B-lymphocyte maturation (73) and impaired immunoglobulin production (17, 52, 71, 120) have been described, previously. T- and B-lymphocyte cell cycle arrest in the late G₁ phase has also been shown (70, 89). Cell-cell interactions mediated through MV glycoproteins expressed on host cell membranes may initiate signaling that inhibits lymphocyte proliferation (36, 44, 91, 115). This is suggested by studies demonstrating that anti-MV antibodies and peptides that block MV fusion can prevent the MV-induced inhibition of T-lymphocyte proliferation (66, 137). Moreover, cocultivation of irradiated, MV-infected PBMC (114) and dendritic cells (25) can lead to T-lymphocyte proliferation defects. This mechanism could explain why direct infection of a small percentage of circulating immune cells can have a global impact on immune function.

Finally, there is evidence that MV infection is associated with global, long-term changes in cytokine secretion by immune cells. Although lymphocyte activation occurs during the acute phase of MV infection (42, 43), a predominant Th2 CD4⁺ T-lymphocyte cytokine profile has been described for MV-infected individuals during the convalescent stage of illness, characterized by increased IL-4 and IL-6 levels as well as low levels of IL-2 and IFN-γ in both plasma and mitogen-stimulated PBMC supernatant fluid (40, 130). Moreover, both neutralization of IL-4 (40) and the addition of recombinant IL-2 (39, 129) to MV-infected lymphocyte cultures improve their proliferation responses. Preferential activation of Th2 CD4⁺ T lymphocytes could lead to long-term depression of both IL-2-dependent T-lymphocyte proliferation and delayed-type hypersensitivity. However, whether a Th2 response is universally induced after MV infection remains controversial. More recent in vitro studies showed that infection of dendritic cells with the MV vaccine promotes Th1 differentiation of naive T lymphocytes, whereas infection with wild-type virus induces a mixed Th1/ThZ (IFN-γ and IL-4 production) response (59).

This diversity of data on immune dysfunction in MV-infected individuals suggests that the cause of MV-induced immune suppression may well be multifactorial. Whatever the causes of the immune suppression may be, this resulting state clearly plays a contributory role in the clinical complications seen in individuals with measles.

MV INFECTION IN THE IMMUNOCOMPROMISED HOST

Individuals with defects in cell-mediated immunity as a result of lymphoma, chronic infection, malnutrition, or immunosuppressive therapy are at risk for severe, progressive MV infection (56). Such individuals can develop atypical measles rashes. These rashes can be severe and desquamating, although sometimes they are absent (28, 56, 80, 93). Complications of measles that occur in individuals with impaired cellular immunity include MV giant-cell pneumonia (75) and measles inclusion body encephalitis (56). In addition, prolonged MV shedding has been reported for malnourished children (56).

MV and HIV coinfection.

The most common immunosuppressive condition affecting individuals who are exposed to MV is HIV infection. Reports of measles in HIV-infected persons describe unusual clinical manifestations and severe complications, often leading to death (24, 30, 56, 61, 68, 86, 93, 97, 109). Several of the reported HIV-infected children with measles coinfections in the United States presented with either no rash or an uncharacteristic rash (18, 20, 61, 86, 97, 109), and 6 of the 19 reported cases resulted in death during the acute illness, usually caused by giant-cell pneumonia (32, 79). Several cases of MV encephalitis in HIV-infected individuals have also been reported (11, 16, 60, 67, 85, 108, 127). In fact, in an autopsy study of MV-infected patients in the Ivory Coast, MV was identified in the central nervous system in 3 of 13 HIV-infected patients and none of 5 non-HIV-infected patients (76).

In sub-Saharan Africa, where both MV and HIV are endemic, children have been studied to determine whether they are at risk for severe clinical complications resulting from coinfection with these viruses. Particularly high measles case fatality rates were reported for HIV-infected children in Zambia (95), whereas case fatality rates were not significantly different between HIV-infected and uninfected children in Zaire (118). More recent data from Zambia found HIV-infected children to represent a large proportion of hospitalized children under 9 months of age with measles, but higher rates of atypical clinical manifestations, complications, or mortality were not seen in this patient population (80). However, some cohort studies in which measles cases were identified by clinical manifestations may have excluded severely immunosuppressed children who presented with unusual or absent clinical signs of disease. In addition, HIV-infected children are more likely to have prolonged MV shedding, indicating a delay in viral clearance (103), presumably as a result of a deficiency in MV-specific cell-mediated immunity.

Although measles vaccination appears to be less effective at preventing disease in the HIV-infected than the uninfected population, it may provide protection from severe disease and accordingly reduce the mortality rate of measles. In Zambia, 50% of cases of HIV-infected children who presented with measles had a history of vaccination (80); however, unvaccinated children were more likely to have unusual clinical complications such as diarrhea, thrush, and pneumonia and had a higher case fatality rate, regardless of age or HIV status (80, 95). In Zaire, there was a trend towards a lower mortality rate for MV- and HIV-coinfected children who had a previous vaccination history than for those who were unvaccinated (118). Importantly, profound cellular immune suppression may be a risk factor for developing disease after vaccination with the live attenuated MV vaccine, as a case of fatal measles has been reported for an HIV-infected individual after MV vaccination (19). In fact, the safety of routine administration of the MV vaccine in regions with high HIV prevalence has recently been called into question (3). Currently, the World Health Organization and the American Academy of Pediatrics recommend MV vaccination for all HIV-infected children, except those children defined as severely immunosuppressed on the basis of age-specific CD4⁺ T-lymphocyte limits (3).

MV INFECTION IN HEALTHY AND IMMUNOCOMPROMISED NONHUMAN PRIMATES

Studies with animal models may provide an effective tool for resolving some of the important unanswered questions concerning MV biology in the setting of immunosuppression. Small laboratory animals have not been useful for studying MV infections because the virus does not replicate in rodents (81, 82, 90). Rodents do not express either of the known receptors for MV, CD150 and CD46, and therefore are not susceptible to MV infection. Cotton rats can be infected with MV by the intranasal route and show evidence of consequent immune suppression, but viral replication is not detectable in these inoculated animals (90, 91). Moreover, transgenic mice (94) and transgenic rats (92) that express human CD46 do not develop clinical manifestations of measles that approximate human disease and do not shed infectious virus (15, 47, 82, 92, 94). While CD150-expressing mice show evidence of viral replication and spread after MV inoculation, the clinical pathology associated with MV infection and recovery of infectious virus are not demonstrable (119, 132). The absence of an inexpensive, readily available animal model for MV infection has therefore slowed progress in resolving many important questions in MV pathogenesis.

Nonhuman primate models provide significant advantages over small-animal models of MV infection. Marmosets (2, 62) and tamarins (65) are highly susceptible to MV infection, but the disease they develop is typically fatal, and the pathological changes seen in these nonhuman primates are quite different from those seen in human MV-induced disease. MV infections in cynomolgus (27, 99, 125) and rhesus (6, 27, 72, 124, 139) monkeys cause a disease with similarities to that in infected humans. Experimentally infected cynomolgus monkeys develop MV viremia, cellular immune responses to the virus, and evidence of immune suppression but often lack clinical signs of measles (27, 99, 125). Macaques experimentally infected with MV by respiratory inoculation develop a maculopapular rash, infection of PBMC, lymphopenia, and immunosuppression (6, 27, 72, 99, 124, 125, 139).

The impact of immune suppression has been assessed in MV-infected monkeys. While fatal measles outbreaks have been reported for macaques, with mortality rates between 10 and 20% (22, 105, 106), experimental infection of immunocompromised macaques has not been associated with fatal MV-induced disease. In rhesus monkeys treated with antithymocyte globulin and then inoculated with wild-type MV, a prolongation of viremia was observed compared to that in MV-infected immunocompetent monkeys. However, the antithymocyte globulin-treated animals did not develop atypical clinical manifestations of MV infection (46). In anti-CD8 monoclonal antibody-treated rhesus monkeys infected with MV, the duration of viremia was prolonged, but the course of disease was not severe, and no significant pathologic findings were apparent upon necropsy of the animals 6 weeks after inoculation (101). In a similarly performed study exploring the role of CD20⁺ lymphocytes in MV containment, CD20⁺ lymphocyte-depleted monkeys did not have a more severe course of illness or an altered viremic profile compared to normal monkeys. However, CD20⁺ and CD8⁺ lymphocyte-depleted monkeys displayed a more severe rash and prolonged viremia but did not develop giant-cell pneumonia or MV encephalitis (100). However, necropsy studies of macaques that died of an AIDS-like syndrome following experimental inoculation of simian immunodeficiency virus (SIV)-infected tissue found multinucleated giant cells with intranuclear and intracytoplasmic inclusion bodies, consistent with paramyxovirus infection, in multiple organs in nearly half of the evaluated monkeys. In those studies, one animal with fatal giant-cell pneumonia was vaccinated with a live attenuated MV vaccine only 10 days prior to death. Studies were not performed to confirm that this monkey died as a result of infection with the vaccine virus (58).

The discrepancy in clinical outcomes between monkeys infected as a result of natural exposure to MV and those experimentally infected with an MV isolate may be explained by an attenuated virulence of laboratory MV strains. It also raises the question of whether cellular immunosuppression due to an AIDS virus infection predisposes the host to severe measles (95) or whether AIDS virus-infected individuals merely develop prolonged MV viremia without significant clinical consequences. The severity of MV-induced disease is likely dependent on the degree of immunosuppression of the infected individual. In macaque models of MV infection in the setting of immunosuppression, the depletion of lymphocytes was temporary, and repopulation of lymphocytes was associated with the extinction of viremia (46, 101). A more chronic immunosuppressed state, such as that induced by HIV/SIV infection, may predispose the host to more uncontrolled viral replication, leading to severe sequelae of infection.

CONCLUSION

The elimination of MV from human populations hinges upon adequate control of the disease in HIV-infected individuals. Attaining this goal will be possible only if we learn more about MV infection in the immunocompromised host. Further clinical studies of MV infection in HIV-infected populations and the use of primate models of MV infection in immunosuppressed hosts will facilitate the exploration of MV pathogenesis and vaccination in the face of CD4⁺ T-lymphocyte loss and dysfunction.