| |
Immune Response Controls HIV- During ART-Drug Resistance
|
| |
| |
"Strong Cell-Mediated Immune Responses Are Associated with the Maintenance of Low-Level Viremia in Antiretroviral Treated Individuals with Drug-Resistant Human Immunodeficiency Virus Type 1"
Journal of Infectious Diseases 2004;189:312-321
Steven G. Deeks,1,4 Jeffrey N. Martin,1,3,4 Elizabeth Sinclair,1,4,5 Jeff Harris,5 Torsten B. Neilands,3 Holden T. Maecker,6 Elilta Hagos,1,4 Terri Wrin,7 Christos J. Petropoulos,7 Barry Bredt,1,4 and Joseph M. McCune1,2,4,5
Departments of 1Medicine and 2Microbiology and Immunology and 3Center for AIDS Prevention Studies, University of California, San Francisco, 4San Francisco General Hospital, 5Gladstone Institute of Virology and Immunology, San Francisco; 6BD Biosciences, San Jose; and 7Virologic, South San Francisco, California
ABSTRACT/SUMMARY
Antiretroviral (ARV)treated patients often maintain low to moderate levels of viremia, despite the emergence of drug-resistant human immunodeficiency virus (HIV).
We studied host and viral factors that may contribute to the control of viral replication in a cohort of 189 adults.
Among ARV-treated patients with detectable viremia, there was a bell-shaped relationship between Gag-specific CD4+ T cell responses and viremia, with the highest cellular immune responses observed in patients with plasma HIV RNA levels of 1000-10,000 copies/mL.
In contrast, there was a negative association between Gag-specific CD4+ T cell responses and viremia among ARV-untreated individuals with wild-type HIV.
Strong cellular immune responses among individuals with drug-resistant HIV predicted subsequent lack of virological progression.
Finally, there was a positive correlation between replicative capacity and viremia.
Collectively, these data suggest that the selection of drug-resistance mutations may reduce the pathogenic potential of HIV, which leads to a balanced state of enhanced cellular immunity and low-level viremia.
BACKGROUND
There is substantial evidence to indicate that the immune system can effectively control viral replication in human immunodeficiency virus (HIV) type 1infected patients, particularly those who receive antiretroviral therapy (ART) during early primary infection and those who are classified as long-term nonprogressors (LTNPs). However, most antiretroviral (ARV)untreated patients eventually experience increased viral replication, accelerated CD4+ T cell depletion, and disease progression. This may occur despite the presence of a large number of CD8+ T cells directed at a variety of HIV-1specific epitopes. Although multiple mechanisms may account for the failure of the cellular immune system to control viral replication, virus-mediated destruction and/or dysregulation of HIV-1specific CD4+ T cells probably is a primary defect that leads to the loss of virus control.
Many HIV-1infected patients are unable to achieve and maintain undetectable levels of plasma viremia with combination ART. As a consequence of incomplete viral suppression, drug-resistant HIV-1 often emerges, thus limiting the virological response to subsequent regimens. Although substantial levels of viremia may occur in treated patients, the steady-state level of drug-resistant viremia during treatment is often lower than the steady-state level of wild-type viremia before treatment. Indeed, when patients with drug-resistant viremia discontinue ART, wild-type HIV often emerges and rapidly outcompetes the drug-resistant variant. This is typically associated with a rapid decrease in CD4+ T cell counts, thus placing patients at high risk for clinical progression. Therefore, many patients with drug-resistant HIV-1 and limited therapeutic options continue to receive stable therapeutic regimens, despite ongoing viral replication.
Several factors probably contribute to the maintenance of a reduced level of viremia in ARV-treated patients with drug-resistant HIV-1. ART often maintains some degree of direct antiviral activity, even when multiple resistance-associated mutations are present. The replicative capacity of the drug-resistant variant is often lower than the wild-type variant, in part because the mutations associated with drug resistance may decrease the enzymatic efficiency of the target proteins (e.g., HIV protease and reverse transcriptase). It also has been found that drug-resistant virus may spare intrathymic T cell production, thus permitting the continued generation of CD4+ and CD8+ T cells with a diverse T cell receptor repertoire, and that the presence of drug-resistant viremia is associated with reduced peripheral CD4+ T cell destruction and turnover. Theoretically, this could result in the generation and maintenance of enhanced HIV-specific cellular immunity. If so, a new steady state may be achieved, in which drug-resistant variants are controlled by a combination of drug pressure and effective antiviral T cell responses.
Because of the observation that drug-resistant variants are less fit and perhaps less virulent in vivo, we hypothesized that the emergence of a drug-resistant variant could be associated with the generation and preservation of HIV-1specific CD4+ and CD8+ T cells and that these cells then would be capable of partially controlling the drug-resistant variant. This hypothesis was tested in a prospective cohort study of ARV-untreated and ARV-treated HIV-infected adults. From this cohort, we studied 189 patients for detailed virological and immunological studies. We excluded patients who had initiated, discontinued, or modified therapy during the preceding 4 months.
Study design. This is an analysis of patients enrolled in an ongoing prospective cohort study aimed at describing the long-term outcomes of drug-resistant viremia (Study of the Consequences of the Protease Inhibitor Era [SCOPE]). Four nonoverlapping treatment groups were recruited: (1) "incomplete" virus suppression, continuous combination ART and a plasma HIV-1 RNA level >50 copies/mL; (2) "complete" virus suppression, continuous combination ART and a plasma HIV-1 RNA <50 copies/mL; (3) ARV untreated, no ART received during the preceding 24 weeks and evidence of progressive CD4+ T cell loss; and (4) LTNPs, self-reported duration of HIV infection for at least 10 years and a median CD4+ T cell count >500 cells/mm3, despite the absence of ART. All enrolled subjects had a CD4+ T cell count >50 cells/mm3.
Subjects in SCOPE are evaluated every 4 months. Plasma, serum, and peripheral blood mononuclear cell samples are archived at each visit. An interviewer-administered questionnaire also is done at each visit. All patients provided informed consent.
For participants who met the definition of "incomplete" virus suppression, "complete" virus suppression, or untreated progressive disease, we studied HIV-1specific T cell responses in the first 175 consecutively enrolled participants. To assure an adequate number of LTNPs, we sampled such individuals separately and studied the first 14 enrolled in the cohort. Studies were generally performed at the second 4-month study visit, using freshly collected blood samples. We excluded patients who had initiated, discontinued, or modified therapy during the 4 months preceding the time that HIV-1specific T cell responses were assessed.
DISCUSSION by Authors
The frequency of HIV-1specific T cells has been investigated in ARV-untreated patients and in patients who achieve a durable virological response to combination ART (plasma HIV-1 RNA level <50 copies/mL). Here, we extend these observations to include patients who remained on ART, despite incomplete viral suppression and the presence of a drug-resistant variant (incomplete virological responders).
Several observations relevant to this patient population were made.
First, using the equilibrium or steady-state virus load as a measure of virological control in these ARV-treated patients, we observed a curvilinear bell-shaped association between the steady-state virus load and the frequency of virus-specific CD4+ T cells.
At low levels of viremia (<3.3. log10 copies RNA/mL), an increasing level of viral replication was associated with increasing levels of Gag-specific CD4+ T cell responses. At higher levels of viremia (>3.3. log10 copies RNA/mL), an increasing level of viral replication was associated with decreasing levels of Gag-specific CD4+ T cell responses.
Second, ARV-treated patients with low levels of viremia generally had low replicative capacity, whereas ARV-treated patients with high levels of viremia generally had high replicative capacity.
Third, high numbers of Gag-specific CD4+ and CD8+ T cells were associated with the maintenance of a stable virus load during short-term observation. Finally, we observed no consistent relationship between viremia and Gag-specific CD8+ T cells. Collectively, these data suggest that viral replicative fitness may affect the capacity of the immune system to generate and maintain HIV-specific CD4+ T cells.
The relationship between viremia and the frequency of HIV-1specific T cells has been controversial. Some studies have reported a positive relationship, which suggests that the level of antigenic exposure drives the immune response. The common observation that HIV-specific T cell numbers decline with highly active ART and increase with its interruption also indicate that the level of viral replication determines, in part, the strength of the immune response. However, other studies have reported a negative correlation between the steady-state level of viremia and the immunological response, which suggests that the immune response determines the virus load set point and/or that high levels of viral replication deplete HIV-1specific T cells (notably, most of these studies have focused on CD8+ T cell responses, and few have specifically addressed the relationship between viremia and CD4+ T cell responses).
Our data in treated patients with detectable viremia (incomplete virological responders) indicate that the relationship among ART, viral replication, and the HIV-1specific CD4+ T cell immune response is complex and nonlinear, with a positive association at low steady-state levels of viremia and a negative association at high steady-state levels of viremia. This bell-shaped relationship between viremia and CD4+ T cell responses is consistent with mathematical models and some empirical observations. These data suggest that a positive correlation between the steady-state level of viremia and the magnitude of the cellular immune response exists in settings when the immune response contributes to the control of viral replication (<3.3 log10 copies RNA/mL in our study), whereas a negative correlation exists when the virus actively inhibits or depletes virus-specific cellular immune responses.
It is tempting to compare the incomplete virological responders in our study with patients who immunologically control wild-type viremia (i.e., LTNPs). By use of similar assays to measure T cell responses, the level of Gag-specific CD4+ T cells observed in our cohort of patients with low to moderate levels drug-resistant viremia was relatively high (especially for patients with a history of advanced stages of HIV-1 disease) and was similar to that found in a cohort of LTNPs. If one assumes that the pretherapy virus/host relationships in our patients partially controlling for drug-resistant HIV-1 was similar to that observed in our untreated patients with advanced disease (i.e., that the levels of viremia and of Gag-specific T cells were similar in the 2 groups), then our data suggest that the introduction of treatment and the emergence of drug resistance resulted in lower levels of viremia and higher levels of Gag-specific CD4+ and CD8+ T cells. In such patients, in the absence of ARV drugs that are more effective and/or better tolerated, the continuation of a "failing" regimen would appear to be a viable treatment option. Certainly, discontinuing therapy in patients with drug-resistant virus typically results in the rapid emergence of a more replication-competent, wild-type virus and a rapid depletion of HIV-1specific CD4+ T cells.
Why would the drug-resistant variant be more likely to preserve effective HIV-1specific T cell responses than the wild-type variant? HIV-1 appears to preferentially infect and deplete HIV-1specific CD4+ T cells, which presumably causes a progressive loss of immunological control. Because the drug-resistant variants are replication deficient, compared to wild-type virus, HIV-1specific T cells may be spared in their presence. In addition, we have observed that drug-resistant variants spare intrathymic T cell production and are associated with lower levels of T cell turnover and activation in vivo, thereby permitting the continued production and survival of HIV-1specific CD4+ and CD8+ T cells. Under these conditions, the presence of a less pathogenic viral strain may shift the delicate balance between the HIV-1mediated destruction of CD4+ T cells and the ability of HIV-1specific CD4+ T cells to control viral replication toward a new steady state, favoring the host while still allowing for viral replication to proceed. This may be analogous to previous reports that have involved the natural infection of humans with attenuated HIV-1 and the experimental infection of macaques with attenuated simian immunodeficiency virus. Finally, preliminary data from our group indicate that HLA-restricted CD8+ T cell response may constrain viral evolution during treatment. Theoretically, this could result in an inability of the virus to evolve toward a state of optimal fitness (defined here as the capacity of the virus to replicate in the presence of a drug).
Our data regarding the relationship between viremia and Gag-specific T cells in ARV-untreated versus ARV-treated patients provide important insights into the relative contributions of immune pressure and drug pressure in maintaining a steady-state level of viral replication. Similar to others, we observed very high levels of Gag-specific CD4+ and CD8+ T cell responses among LTNPs, with very low levels of viremia in the absence of therapy (<100 copies RNA/mL). These levels were generally higher than those observed in ARV-treated patients with similar levels of viremia. We interpret this to mean that continued drug pressure in the face of drug resistance and/or the presence of reduced replicative capacity greatly contribute to the maintenance of very low-level viremia during treatment. The relative contribution of HIV-specific T cells responses during partially effective therapy probably is most relevant in patients with a steady-state level of viremia between 100 and 10,000 copies RNA/mL, as was recently suggested by others.
The clinical significance of these data remains to be determined. If, as suggested here, some patients are able to control a drug-resistant variant immunologically, then maintaining patients on a stable regimen may be a reasonable treatment option. However, continued therapy in the face of ongoing viral replication may select for variants that replicate to high titer. This could occur though several mechanisms, including increased drug resistance, increased viral replicative capacity, the gradual depletion of HIV-1specific T cells, and/or selection for viral escape mutants that are no longer recognized by the immune system.
Conclusions drawn from the present study must be tempered by limitations inherent in the design and implementation of human studies and by limitations in the assays used. First, patients with drug-resistant HIV-1 modify medications frequently, which makes it difficult to analyze the effect of HIV-1specific immune responses on long-term outcomes. Second, the hypothesis that the emergence of drug-resistant viremia results in altered or enhanced HIV-1specific cellular immunity would ideally be tested by using longitudinal samples obtained before the administration of highly active ART and treatment failure. Such a study is, however, impractical: patients fail treatment at varied times, and multidrug resistance typically occurs only after several regimen switches. Third, replicative capacity was measured by use of a recombinant vector that contained only a few patient-derived viral gene sequences (reverse transcriptase, protease, and the 3 end of gag). The long-term evolution of HIV in the presence of PI therapy results in the emergence of compensatory mutations in gag and perhaps other genes; thus, our assay probably overestimates the fitness defects associated with drug resistance. Finally, we measured the response to only 1 peptide pool (Gag), studied only 1 tissue (blood), and used only 1 assay (intracellular cytokine production). Therefore, although our data certainly underestimate the total cellular immune response, it is likely that the responses detected in the present study correlate with total response. Notwithstanding these limitations, further analysis of these observations appears to be warranted by the robust nature of the data sets and the relatively large sample size from which they were obtained.
In conclusion, we have observed a protective association between the presence of HIV-1specific CD4+ and CD8+ T cells and the partial control of drug-resistant viral replication. These observations support the following model regarding the pathogenesis of drug resistance in vivo. Initially, T cell immunological function improves as a consequence of a treatment-mediated reduction in viral replication. Because combination therapy eventually fails to fully control viral replication, low-level replication of the drug-resistant variant occurs, thus providing the antigenic stimulus necessary to generate greater numbers of HIV-1specific CD4+ and CD8+ T cells. The partial control of viral replication by ART and/or by the emergence of a less fit, less virulent HIV-1 variant prevents the rapid depletion of HIV-1specific CD4+ T cells. Ultimately, a new steady state is achieved, whereby the wild-type virus is controlled for by therapy and the drug-resistant variants are partially controlled for by the combination of an effective cellular immune response, low replicative capacity, and continued partially effective drug pressure. Further prospective studies are needed to test this hypothetical model.
|
|
| |
| |
|
|
|
