| |
Drug-resistant HIV may have diminished fitness
|
| |
| |
NEW YORK (Reuters Health) - Multidrug-resistant HIV-1 variants in patients with long-term exposure to combination therapy and with persistent virological failure show diminished replicative capacity and infectivity, Spanish and US researchers report in the September 1st issue of Clinical Infectious Diseases.
Dr. Javier Martinez-Picado of Hospital Germans Trias i Pujol, Badalona, Spain, and colleagues note during treatment failure, HIV load often remains partially suppressed and CD4 cell counts remain stable, suggesting selection of HIV-1 variants with decreased replication capacity.
To investigate further, the researchers used plasma samples from four chronically HIV-infected patients who showed such characteristics. Recombinant viruses were generated from samples taken before or shortly after the start of monotherapy, as well as some years later during highly active antiretroviral therapy.
In three patients, the investigators found that the multidrug-resistant variants selected during long-term therapy were less fit and infectious than their wild-type or monotherapy-selected counterparts.
The team conclude that this reduced fitness could have contributed to the stability of the CD4 counts seen over the 8 years of the study. Conversely, the eventual decrease seen in the remaining patient may have been related to the improved fitness of one of the recombinant viruses.
Clin Infect Dis 2005;41:729-737.
HIV Type 1 Fitness Evolution in Antiretroviral-Experienced Patients with Sustained CD4+ T Cell Counts but Persistent Virologic Failure
Julia G. Prado,1 Neil T. Parkin,2 Bonaventura Clotet,1 Lidia Ruiz,1 and Javier Martinez-Picado1
1IrsiCaixa Foundation, Hospital Germans Trias i Pujol, Universitat Autònoma de Barcelona, Badalona, Spain; and 2ViroLogic, South San Francisco, California
The aim of this study was to investigate changes in viral fitness associated with the sequential introduction of different antiretroviral drug classes in 4 chronically HIV-1-infected patients, with infection characterized by a sustained CD4 cell count despite the presence of a detectable viral load during long-term antiretroviral treatment. Thus, we generated recombinant viruses from plasma samples obtained before or shortly after monotherapy initiation, as well as a few years later during HAART, and we evaluated changes in the replication capacity profile of these viruses.
ABSTRACT
Background. Over recent years, treatment guidelines for human immunodeficiency virus (HIV) infection have evolved from monotherapy to combination regimens that include ⩾3 active drugs, resulting in a sharp decrease in morbidity and mortality. In the present article, we evaluated changes in HIV type 1 viral fitness associated with the sequential introduction of antiretroviral treatment strategies in 4 chronically infected patients with sustained CD4 cell count despite having a persistently detectable viral load.
Methods. Plasma samples were obtained before and during treatment to construct recombinant virus containing the 3_-end of gag, the protease and the reverse-transcriptase coding region. Drug susceptibility phenotype was evaluated with a panel of multiple reverse-transcriptase and protease inhibitors. Replicative capacity (RC) and infectivity were measured, and production of p24 was monitored after transfection.
Results. Multidrug-resistant (MDR) viruses selected during long-term antiretroviral therapy were less fit and infectious than their wild-type or monotherapy-selected counterparts, with the exception of viruses recovered from patient B. In 3 of 4 cases, p24 kinetics after transfection showed a delay in viral production of recombinant viruses containing MDR mutations. Data from the RC and infectivity assays showed good correlation (P < .03) and corroborated the p24 kinetics data.
Conclusions. This study shows that accumulation of MDR mutations during long-term antiretroviral treatment results, albeit not in all cases, in reductions of viral fitness.
Clinical parameters.
The median duration of clinical follow-up for the patients in the study was 8 years (interquartile range, 6.5-8.0 years). The plasma viral load in these patients decreased a median of 1.7 log HIV-1 RNA copies/mL (interquartile range, 1.2- 1.8 log HIV-1 RNA copies/mL) between the beginning and the end of the study period, but it never reached a level of <50 copies/mL, despite the occurrence of multiple switches of therapy in pursuit of that goal. Nevertheless, the patients had a median increase in the CD4 cell count of 24 cells/mL per year, although this increment was not constant during the study period. These patients initiated zidovudine monotherapy during the first half of the 1990s and were subsequently switched to dual-drug therapy and, finally, to HAART, in accordance with upgrades in the HIV-1 treatment guidelines. All 4 subjects were initially treated with NRTIs. PIs were introduced to their antiretroviral regimens during 1996 and 1997, and the change was mostly associated with significant increases in the CD4 cell counts and reductions in the plasma viral load; later on, NNRTIs were added to the patients' treatments. At the time that the last samples were obtained, the patients had received a median of 10 antiretroviral drugs (range, 9-13 drugs), including the drugs included in the most recent therapeutic regimen.
DISCUSSION
In the present study, we used different methodologies to evaluate the fitness of recombinant viruses that were present before therapy or selected during monotherapy and their counterparts selected during receipt of HAART regimens in patients with sustained CD4+ cell counts but persistent virologic failure.
The relationship between viral fitness and clinical parameters in HIV-1-infected individuals is complex. The presence of multidrug-resistant viruses with impaired fitness has been associated with sustained CD4 cell counts, despite the presence of detectable plasma viremia, suggesting that there is an immunological benefit for the patient [15, 21]. The fitness values of the recombinant viruses used in our study could have contributed to the stability of the CD4 cell counts over the 8 years of the study. Presumably, the decrease in the CD4 cell count in patient B during the last 2 years of follow-up, after an initially sharp increase, may have been related to the improved fitness of the recombinant virus B00.
Presently, the evolution of HIV-1 fitness in antiretroviral-experienced patients cannot be predicted on the basis of the genotype of drug-targeted genes. The multidrug-resistant viruses with reduced drug susceptibility that were selected during HAART from patients A, C, and D had lower replication capacity and infectivity than did the corresponding zidovudine-resistant recombinant viruses. For patient D, the selection of the substitution M184V in the reverse transcriptase resulted in a severe fitness loss [22, 23]. Additional accumulation of new mutations as treatment progressed made it impossible to obtain a viral stock for in vitro fitness evaluation, despite persistent plasma viremia. In patient B, however, the replication capacity of the pretreatment recombinant virus (B95; 27%) was significantly impaired (44% would correspond to the 10th percentile of the wild-type population; median replication capacity, 100%) [24]. Although it is in general accepted that drug-resistance mutations emerge at the expense of a loss in viral fitness, the replication capacity of virus B00 increased up to 104%, despite progressive accumulation of drug-resistance mutations. However, we cannot rule out the effect of non-drug resistance-associated mutations in replication capacity changes. Therefore, intrapatient comparisons are far more interesting, as shown here, than is comparison of the relative replication capacity with that of viral reference strains to assess viral fitness evolution during antiretroviral treatment.
The discrepancy between patient B's pair viruses and those obtained from the rest of the patients was consistent among the different methods used in the study. Indeed, data obtained with the single-cycle assay significantly correlated with replication constant and infectivity results. Because there are many factors that are relevant in the development of appropriate comparative fitness assays, consistency among the different type of fitness experiments employed is important. Because we exclusively used chimeric viruses, we cannot exclude that multidrug-resistant isolates had altered infectivity and growth kinetics, compared with drug-susceptible viruses, as shown in the context of primary HIV-1 infection [25].
The accumulation of mutations might result in not only highly crippled but also nonviable viruses. Thus, in the case of patient A, we observed a reduction in the efficiency of single-clone transfection concomitantly with the replication capacity impairment. Some changes in the clones transfected could be linked to genotypic determinants associated with severe defects in infectivity and fitness (e.g., the presence of N88S mutation in the protease gene [20] or the presence of a stop codon in the reverse-transcriptase gene, either as a result of viral polymerases or PCR in vitro-based polymerases). However, most of the clonal sequences had a complex mutational pattern in the protease and the reverse-transcriptase, making it very difficult to associate the lack of virus growing with specific amino acid changes. The accumulation of mutations during treatment failure might result in fewer viable viral clones after transfection, suggesting a diminution of the effective viral population after many years of treatment. This reduction of viral viability might also be associated with an increase of the mutagenesis rate through continuous drug pressure, reducing fitness and infectivity, and also contributing to a reduction in the CD4 cell-infected pool. Reductions of the effective virus population associated with drug pressure have been reported before for HIV-1 and foot-and-mouth disease virus [26, 27].
Although the patterns of mutations involved in replication capacity changes are not clearly defined, it has been proposed that pol and gag coevolution has an important role in patients receiving antiretroviral therapy [28-33]. Selection of mutations in gag (both CSM and non-CSM) has been previously described as a means to improve viral fitness in highly resistant viruses [25, 31-35]. The CSM A431V has been associated with protease inhibitor-resistance mutations at codons 46 and 82 [30, 36], whereas P453L may direct the protease resistance pathway through I84V instead of V82A mutation [36], in agreement with the emergence of these CSM in patients C and D. The 3-amino acid duplication (APP) in the proline-rich p6Gag PTAP motif, which interacts with cellular proteins crucial for viral budding, has been linked to an increased infectivity and resistance to NRTIs [28], which is consistent with its emergence before the introduction of protease inhibitors in the treatment regimens of patients B and D. This insertion, as well as an insertion similar to the SRLE developed in patient C, have been recently associated with restoration of reduced replication capacity in multidrug-resistant viruses [37]. Although the presence of the APP insertion in p6gag in the recombinant viruses from patient B might be related to the increase in replication capacity observed in the multidrug-resistant virus (B00) respect to the pretreatment virus (B95), the emergence of the same insertion in the recombinant virus D98, did not seem to compensate for the selection of 5 thymidine-associated mutations and the reverse-transcriptase M184V. Multidrug-resistant recombinant viruses recovered from patients A, C, and D did not recover replication capacity to levels similar to or higher than those for earlier viruses, despite an accumulation of CSMs and non-CSMs. By contrast, multidrug-resistant virus from patient B had an improved replication capacity, with respect to the pretreatment virus. Only 2 non-CSMs (K418G and G466R) emerged during treatment, but it has not been reported whether these mutations might play a role as replication capacity compensatory changes.
Although we focused on the role of gag and pol in replication capacity, overall, evolution of CD4 cell count and viral load in these patients could have been, in part, a consequence of complex virus-host interactions as changes in viral tropism or adaptive immune response, which go beyond the scope of this study.
Our findings challenge the idea that the accumulation of drug-resistance mutations during long-term antiretroviral treatment inevitably results in viruses with impaired fitness. However, multidrug-resistant viruses with reduced fitness are related to sustained CD4 cell counts in patients with persistently detectable plasma viremia. Changes in replication capacity and infectivity would be modulated by drug pressure and virus-host interactions playing an important role in the maintenance of the immunologic status. Therefore, the therapeutic benefits of locking into the virus genome drug-resistance mutations, which impair viral fitness, remains highly speculative.
|
|
| |
| |
|
|
|
