| |
Progressive reversion of HIV-1 mutations may have serious consequences
|
| |
| |
By Megan Rauscher
NEW YORK (Reuters Health) - Although multidrug-resistant HIV strains, because they are potentially less virally fit than non-drug resistant strains, were hoped to be less likely to cause disease, a report published this month suggests just the opposite.
After transmission of a resistant HIV strain, the virus is able to evolve to become more fit and cause significant disease, according to the study in the December 15th issue of Clinical Infectious Diseases.
"The frequency of HIV that is resistant to current medications seems to be growing, and this may limit global treatment strategies," Dr. Rajesh T. Gandhi from Partners AIDS Research Center in Boston, who led the study, noted in comments to Reuters Health.
Dr. Gandhi and colleagues monitored viral evolution after transmission of HIV-1 containing multiple reverse transcriptase (RT) and protease mutations (PR) in a 32-year-old previously healthy man.
In the absence of ongoing antiretroviral therapy, 5 of 12 drug resistance mutations reverted in a stepwise fashion to wild type over the course of one year. Of note, reversal of the M184V mutation alone did not produce a change in replicative capacity but it did enhance resistance to zidovudine and tenofovir.
However, "reversions of a second (RT) mutation and 3 PR mutations were associated with an increase in viral replicative capacity and this was temporally correlated with a marked decrease in CD4 cell number," the researchers report.
Dr. Gandhi and colleagues say this study "demonstrates the gradual stepwise back-mutation of certain drug resistance mutations in vivo in the absence of ongoing drug selection pressure."
While larger studies are needed to confirm these findings, Dr. Gandhi told Reuters Health, this report "emphasizes that the recent increase in HIV transmission among men who have sex with men in the United States may result in acquisition of multidrug resistant virus, with very serious consequences."
Clin Infect Dis 2003;37:1693-1698.
ORIGINAL ARTICLE EXCERPTS
"Progressive Reversion of Human Immunodeficiency Virus Type 1 Resistance Mutations In Vivo after Transmission of a Multiply Drug-Resistant Virus"
Rajesh T. Gandhi,1,2,3 Alysse Wurcel,1,2,3 Eric S. Rosenberg,1,2,3 Mary N. Johnston,1,2,3 Nicholas Hellmann,4,a Michael Bates,4 Martin S. Hirsch,1,2,3 and Bruce D. Walker1,2,3
1Infectious Diseases Unit and 2Partners AIDS Research Center, Massachusetts General Hospital, and 3Division of AIDS, Harvard Medical School, Boston; and 4ViroLogic, South San Francisco, California
ABSTRACT
Evolution and transmission of multiply drug-resistant human immunodeficiency virus type 1 (HIV-1) may limit therapeutic options as global treatment efforts expand. However, the stability of these mutants in the absence of drug selection pressure is not known.
We performed a longitudinal analysis of plasma virus from a person who acquired HIV-1 that contained multiple reverse transcriptase (RT) and protease (PR) mutations. In the absence of therapy, 5 of 12 drug resistance mutations reverted in a stepwise fashion to wild type over the course of 52 weeks. Reversion of the M184V mutation alone did not change viral replicative capacity (RC), but it led to enhanced resistance to zidovudine and tenofovir.
However, reversions of a second RT mutation and 3 PR mutations were associated with an increase in viral RC, and this was temporally correlated with a marked decrease in CD4 cell number.
This study demonstrates the gradual stepwise back-mutation of certain drug resistance mutations in vivo in the absence of ongoing drug selection pressure. Moreover, it suggests that, despite initially impaired viral fitness, a transmitted HIV-1 isolate with multiple drug resistance mutations can evolve to develop increased RC and significant pathogenicity.
BACKGROUND
Combination antiretroviral therapy (ART) has substantially decreased the number of deaths resulting from AIDS in the United States and Europe, and its use is now being extended to resource-poor settings. There is concern that evolution and transmission of multiply drug-resistant HIV-1 may limit future therapeutic options. However, the stability of these mutants in the absence of the drug selection pressure under which they were generated is not known. In addition, limited information is available about the natural history of infection with multiply drug-resistant HIV-1. Because resistance mutations can impair viral fitness, this may affect the rate of progression to AIDS.
Several longitudinal studies that have included patients with chronic HIV-1 infection describe the changes in drug resistance mutations that occur after incompletely suppressive ART regimens are discontinued. For example, in patients who have persistent viremia and drug-resistant virus during receipt of a protease inhibitor (PI)containing regimen, the circulating virus generally becomes susceptible to PIs within 12 weeks after all antiretroviral medication is stopped, which suggests that an outgrowth of wild-type virus occurs once drug selection pressure is removed. In such patients, the in vitro replicative capacity (RC) of drug-resistant virus is often low while the patient is receiving a failing treatment regimen; however, after discontinuation of therapy, the viral RC increases coincident with the outgrowth of wild-type virus. Whether the same pattern occurs after transmission of resistant virus is not certain, although there is recent evidence that detectable resistance may persist longer under these conditions.
In this report, we present a detailed longitudinal analysis of viral evolution after transmission of HIV-1 containing multiple genotypic and phenotypic mutations in both reverse transcriptase (RT) and protease (PR). In the absence of ongoing ART, virus containing multiple mutations underwent stepwise reversion to wild type, demonstrating the intrinsic instability of certain drug resistance mutations in vivo. The progressive reversion of mutations was associated with increased viral RC and accelerated CD4 cell count decline. This suggests that, despite initially reduced viral fitness, a multiply drug-resistant HIV-1 isolate can evolve to develop increased RC and significant pathogenicity.
CASE REPORT
A 32-year-old previously healthy man presented with fever and sore throat 3 weeks after participating in unprotected receptive anal intercourse with an anonymous male partner. A pharyngeal culture was positive for group A streptococcus, and the patient received a 10-day course of oral penicillin. In addition, the patient was HIV-1 antibody negative, and his HIV-1 RNA level was >750,000 copies/mL, which is consistent with acute HIV-1 infection. His CD4 cell count was 399 cells/mm3.
The patient was enrolled in an institutional review board approved study of acute HIV-1 infection at Massachusetts General Hospital (Boston). A viral isolate was sent to the laboratory for genotyping (Virologic; South San Francisco, CA), and an ART regimen that included zidovudine, lamivudine, abacavir, and efavirenz was initiated 2 weeks after the diagnosis of HIV-1 infection was made (5 weeks after the unprotected sexual contact). Treatment was discontinued when the results of genotyping revealed that the patient had been infected with a highly resistant isolate of HIV-1. The total duration of the initial therapeutic regimen was 3 weeks. On the day that treatment was discontinued, the patient's plasma HIV-1 RNA level was 12,400 copies/mL. Another isolate was genotyped on the day of treatment discontinuation and did not demonstrate any new drug resistance mutations.
Subsequently, the patient's plasma HIV-1 RNA levels and CD4 cell counts were measured every 13 months; during that time, the patient did not receive ART. After discontinuation of therapy, his plasma HIV-1 RNA level increased to >100,000 copies/mL, and his CD4 cell count initially stabilized between 300 and 490 cells/mm3.
Approximately 40 weeks after the initial infection, the patient's CD4 cell count began to decrease precipitously. Over the course of an 8-week period, his CD4 cell count decreased from 302 to 73 cells/mm3. A regimen that included tenofovir, lamivudine, stavudine, didanosine, and efavirenz was initiated at week 52 after infection. The patient's plasma HIV-1 RNA level decreased from 257,000 to <50 copies/mL and his CD4 cell count increased to 203 cells/mm3 within 24 weeks after initiation of this antiretroviral regimen.
METHODS
Plasma HIV-1 RNA levels were quantified using an RT-PCR method (Amplicor HIV-1 Monitor assay; Roche Diagnostics). Population-based nucleotide sequencing of HIV-1 pol sequences (GeneSeq HIV genotyping assay) was performed by ViroLogic. Cryopreserved plasma samples were also analyzed for drug susceptibility (phenotyping) by a recombinant virus assay (PhenoSense HIV; ViroLogic). The relative RC of the subject's virus was determined by a modification of the drug susceptibility assay. HIV-1 RT and PR sequences from clinical specimens were cloned into recombinant test vectors containing a luciferase indicator gene that allows for quantification of viral replication. The replication of vectors containing patient-derived pol sequences was compared with that of a vector containing wild-type sequence (based on the reference strain HIV NL4-3) in the absence of antiretroviral drug. The RC was calculated as the ratio of luciferase activity from vectors containing patient-derived viral sequences to the activity from the NL4-3 wild-type reference vector. The RC measurement of the patient-derived virus is expressed as a percentage of NL4-3 RC, which is set at 100%. A percentage <100% indicates that the patient's HIV-1 strain has a reduced RC, compared with that of NL4-3. The median RC of wild-type clinical isolates in this assay is about 60%.
RESULTS
Initial resistance testing. The initial HIV-1 genotyping performed at the time of diagnosis demonstrated multiple drug resistance mutations in the viral RT (M41L, M184V, and T215Y) and PR (L10F, L33F, M36I, I54V, A71V, V82A, I84V, and L90M). Phenotypic drug susceptibility testing confirmed high-level resistance to lamivudine and zidovudine. In addition, there was at least 23-fold decreased susceptibility of the plasma virus to all approved PIs, including lopinavir.
Longitudinal assessment of viral genotype. The subject underwent longitudinal monitoring of viral genotype while he was not receiving ART. Serial viral genotyping revealed stepwise loss of mutations in RT. The first reversion was noted at day 126, with the emergence of a mixture of wild-type and mutant virus at codon 184 in RT, which mediates resistance to lamivudine. By day 196, the entire population of detectable virus in plasma encoded the wild-type methionine at this position. On day 301, we first detected a mixture of tyrosine and cysteine at codon 215. By day 343, all detectable plasma virus encoded cysteine at position 215. Thus, 2 of 3 RT mutations reverted over the course of 1 year in the absence of therapy.
Similar genotypic analysis of PR mutations likewise demonstrated gradual detection of mixtures that first began to appear at day 301, when the L10F mutant became a mixture. By day 343, 3 of the 9 PR mutations contained mixtures that included wild-type amino acids (L10L/F/V, I54I/V, and V82V/A).
Phenotypic evolution in the absence of drug selection pressure. To assess the impact of these reversions on drug susceptibility, longitudinal phenotypic resistance testing was performed. As resistance to lamivudine waned (from >105-fold decreased susceptibility at day 28 to 3.7-fold decreased susceptibility at day 175), resistance to zidovudine increased (from 19- to 53-fold decreased susceptibility). An increase in resistance to tenofovir was also seen as lamivudine resistance decreased. Reversion at position 215 was associated with a dramatic reduction in phenotypic resistance to zidovudine. Despite the appearance of mixtures of wild-type and mutant amino acids in 3 of 9 PR mutation sites, the virus continued to demonstrate high-level phenotypic resistance to PIs.
RC(replication capacity). The in vitro RC of the patient's viral isolate was initially low (9%20% of wild type). Between day 301 and day 364 after infection, the RC of the circulating virus progressively increased, to 89% of wild type. This increase in RC paralleled the reversion of T215Y to T215C in RT and the emergence of mixtures at 3 PR codons (L10L/F/V, I54I/V, and V82V/A). In addition, there was a close temporal relationship between the increase in RC and the rapid decrease in the patient's CD4 cell count.
DISCUSSION
This report demonstrates transmission of an HIV-1 isolate with high-level resistance to multiple antiretroviral agents, including lopinavir, and documents longitudinally the changes in genotypic and phenotypic resistance in the absence of antiretroviral drug selection pressure. The patient infected with this virus suffered progressive immunodeficiency within a short time after infection, and a temporal correlation was seen between the rapid decrease in CD4 cell count, the increase in viral RC, and the reversion of specific mutations in viral RT and PR.
Several studies have suggested that transmission of drug-resistant HIV-1 is increasing in frequency. To date, transmission of virus that is resistant to lopinavir has not been documented. However, given that drug resistance to lopinavir in chronically infected individuals has been seen, it is not surprising that transmission of such virus can occur. Because of the increasing frequency of transmission of resistant virus, it has been suggested that resistance testing should be performed on isolates from all patients thought to have recent infection with HIV-1. The present report supports such an approach.
Once it was clear that the patient had been infected with a highly resistant virus, we discontinued ART. Although early therapy may be associated with preservation of important HIV-specific immune responses, we felt that this theoretical benefit was outweighed by the likelihood that selection for further drug resistance would occur, because options for a fully suppressive regimen were limited (tenofovir was not available when the patient was initially evaluated). Thus, we were able to longitudinally monitor changes in drug resistance genotype and phenotype in the absence of drug selection pressure. We found that this individual's plasma virus underwent slow stepwise reversion of mutations in RT and PR, rather than rapid outgrowth of wild-type virus.
This pattern of stepwise reversion of PR mutations contrasts with what is generally seen after discontinuation of ART in individuals chronically infected with HIV-1 for whom this class of therapy is failing. Such patients generally have outgrowth of previously latent wild-type virus within 48 weeks after discontinuation of the PI-containing regimen. This observation is consistent with out-competition of resistant virus by wild-type virus when drug selection pressure is removed, perhaps because the wild-type virus is more replication competent. One likely explanation for the different pattern seen in our patient is that he may have been infected only with resistant virus, rather than harboring a mixture of resistant and wild-type virus, as is the case for individuals who develop resistance during treatment. The persistence of resistance mutations in the absence of drug selection pressure has been found in other instances of transmission of resistant HIV-1These observations support the theory that new infections involve transmission of a single predominant viral strain, rather than the entire diverse quasi species generally present in chronically infected individuals.
The reversion of particular mutations appeared to be associated with an increase in viral RC, which implies that virus evolution to improve replication occurred in the absence of antiretroviral drugs. In vitro competition experiments suggest that T215Y and M184V mutations in RT diminish replication competence relative to the wild type in the absence of antiretroviral drugs. Similar studies indicate that particular PI mutations, such as V82T, D30N, and L90M, result in decreased viral replication capacity. In this study, viral fitness was assessed using a recombinant assay of RC, which correlates well with other in vivo and in vitro measures of viral fitness. In our patient, an increase in the viral RC was associated with the reversion of T215Y to T215C in RT and with the emergence of mixtures at 3 PR sites previously present as pure mutants (L10L/F/V, I54I/V, and V82V/A). This finding suggests that mutations at 1 or more of these particular codons may decrease viral RC.
The correlation between genotypic and phenotypic resistance in this patient also illustrates the fact that the presence of particular mutations that confer resistance to one antiretroviral agent may result in increased sensitivity to another drug. For example, a mutation at codon 184 (M184V) that is associated with decreased susceptibility to lamivudine has been found in vitro to sensitize virus to inhibition by zidovudine. In our patient, as resistance to lamivudine decreased, resistance to zidovudine increased, which demonstrates reciprocal interaction between susceptibility to these 2 drugs in a clinical specimen. A similar sensitization to tenofovir by the 184 mutation has also been suggested by in vitro studies and was seen in our patient's viral samples.
In this patient, infection with an HIV-1 isolate with multiple drug resistance mutations led to rapid progression to AIDS. Despite multiple mutations and impaired RC, the virus load set point was high (>100,000 copies/mL). The patient experienced a rapid decrease in CD4 cell count (temporally associated with reversion of drug resistance mutations) and an increase in in vitro viral RC. The decrease in CD4 cell count and concomitant increase in RC were associated with only a minor increase in plasma virus load, which suggests that the immunologic deterioration was the result of an intrinsic change in the pathogenicity of the virus, rather than of a change in the virus burden. Additional explanations for this individual's rapid progression to AIDS may include host immunologic factors and changes in viral coreceptor use.
In summary, longitudinal analysis of viral evolution after transmission of drug-resistant virus demonstrated stepwise back-mutation at specific sites, rather than outgrowth of wild-type virus. Reversion of specific mutations in viral RT and PR was associated with an increase in HIV-1 RC. The rapid clinical progression seen in this patient demonstrates that, despite initially impaired viral fitness, an HIV-1 isolate with multiple drug-resistance mutations can evolve to develop increased RC and significant pathogenicity.
|
|
| |
| |
|
|
|
