Wildlifescienceintl.com

Resistance Mutations in Protease and Reverse Transcriptase Genes of Human
Immunodeficiency Virus Type 1 Isolates from Patients with Combination
Antiretroviral Therapy Failure
Benjamin Young, Steven Johnson, Minoo Bahktiari,
Division of Infectious Diseases, University of Colorado Health David Shugarts, Russell K. Young, Michael Allen,
Sciences Center, VA Medical Center, Denver, and Department of Rob Roy Ramey II, and Daniel R. Kuritzkes
Environmental, Population and Organismic Biology, University of High-density oligonucleotide arrays were used to determine the sequence of the protease
(PR) and reverse transcriptase (RT) genes of human immunodeficiency virus type 1 isolates
from 35 patients in whom combination therapy that included a protease inhibitor had failed.
Isolates had a median of three PR mutations (range, none to six). Three isolates had no known
resistance mutations in PR. Twelve isolates (34%) had two or fewer resistance mutations in
PR. The most commonly observed PR mutations were L10I, V82A/T/F, and L90M. No
mutations were observed at codons 30 or 48. Mutations at RT codons 215 and 184 were
observed in the majority of isolates. These data suggest that therapy can fail in some patients
with relatively few PR resistance mutations. Clinical failure in the absence of resistance mu-
tations implies inadequate drug exposure due to pharmacologic factors or suboptimal patient
adherence to drug therapy.
Advances in our understanding of the dynamics of human such therapies, the population consensus sequence of HIV-1 immunodeficiency virus type 1 (HIV-1) replication [1, 2] and PR and RT genes was determined for HIV-1 isolates from 35 the development of new potent antiretroviral agents have led patients in whom combination antiretroviral therapy that in- to profound changes in the treatment of HIV-1 infection [3].
cluded a protease inhibitor had failed.
Current antiretroviral therapy is directed against two virallyencoded proteins, HIV-1 reverse transcriptase (RT) and pro-tease (PR). Combination antiretroviral therapy with nucleoside RT and PR inhibitors dramatically reduces plasma and tissuelevels of HIV-1 RNA and confers significant clinical benefits Study population and specimen collection. [3–5]. These benefits may be limited, however, by the emergence infected persons from the Denver metropolitan area in whom of drug-resistant variants of HIV-1 [5].
combination therapy with a protease inhibitor was failing. Treat- Although genetic determinants of drug resistance are well- ment failure was defined as a rising plasma HIV-1 RNA level or characterized for individual antiretroviral agents, less is known a plasma HIV-1 RNA level 110,000 copies/mL in the setting ofapparent adherence to prescribed treatment. Isolation and cry- about the emergence of resistance mutations in the context of opreservation of patients’ plasma and peripheral blood mononu- potent combination regimens [6–8]. To assess the contribution clear cells for HIV-1 culture were done according to standard con- of protease inhibitor resistance mutations to clinical failure of Nucleic acid sequencing and analysis. were obtained from proviral DNA of cultured HIV-infected pe- Received 29 January 1998; revised 27 May 1998.
Presented in part: International Workshop on HIV Drug Resistance, ripheral blood mononuclear cells [9] or from plasma HIV RNA Treatment Strategies, and Eradication, St. Petersburg, Florida, 25–28 June that was amplified by an RT-coupled polymerase chain reaction [10]. Hybridization-based nucleic acid sequencing was done with Informed consent was obtained from all subjects participating in this high-density oligonucleotide arrays as described [10]. This study. Conduct of this study conformed to human experimentation guide- method is biased toward the detection of the predominant sequence lines of the US Department of Health and Human Services, the Universityof Colorado Health Sciences Center, and the Department of Veterans in the virus population. The resulting nucleotide sequences were determined by use of the Genechip “rules” algorithm (Affymetrix, Financial support: NIH (training grant AI-07447 support to B.Y.; AI- Santa Clara, CA) and were compared with the HIV-1 clade B 38858 (Virology Advanced Technology Laboratory Development Award).
consensus [11]. Additional sequence comparisons were made with Reprints or correspondence: Dr. Benjamin Young, Division of Infectious Diseases, University of Colorado Health Sciences Center, 4200 E. Ninth a pairwise distance matrix (PAUP, version 3.1). Drug resistance Ave., B-168, Denver, CO 80262 (Ben.Young@UCHSC.edu).
mutations in PR and RT included for analysis were obtained frompublished reports [6–8]. The sequences have been submitted to The Journal of Infectious Diseases
1998; 178:1497–1501
GenBank (NCBI reference no. 444508; accession nos. AF072936– 1998 by the Infectious Diseases Society of America. All rights reserved.
cultured and plasma virus. To assess for the possibility of lab-oratory contamination, the nucleotide or amino acid sequences Forty-seven patients were enrolled, but this analysis includes of virus isolates were compared with each other and with the only the 35 patients for whom complete clinical data and treat- sequences of 5 strains of HIV-1 commonly studied in our lab- ment history were available. The median CD4 lymphocyte oratory. The sequence difference between virus isolates from count was 157 cells/mm3 (range, 5–515), and the median plasmaHIV-1 RNA level was 5.0 log different patients or laboratory strains was ≥2.0%, consistent Patients had been treated with a median of two protease in- with a lack of contamination (data not shown). Drug resistance hibitors (range, one to three). Thirty-four patients (97%) had mutations in the PR gene were found in HIV-1 isolates from received indinavir, 15 (43%) had received saquinavir, and 18 32 of 35 patients (91%; table 1). The median number of PR (51%) had received ritonavir. Two patients had received pro- resistance mutations per isolate was three (range, none to six).
tease inhibitors in combination. None of the patients had re- Virus isolates from 12 patients (34%) had two or fewer mu- ceived nelfinavir or amprenavir (141W94).
tations in the PR gene. One patient had an isolate with only a The nucleotide sequence of HIV-1 PR and RT genes was single mutation in PR (N37S), and 3 patients had isolates with determined from cultured virus in 27 patients and from plasma no known resistance mutations in PR.
virus in 11 patients. In 3 patients, sequencing was done on both The most commonly observed PR resistance mutations Mutations in HIV-1 protease gene at codons associated with drug resistance in patients with combination antiretroviral treatment Amino acid residues are indicated by single-letter abbreviations. HIV-1 clade B consensus sequence [11] is shown for comparison. — indicates no change from consensus sequence; ? indicates indeterminate sequence. Protease inhibitor use: indinavir (IDV), ritonavir (RTV), saquinavir (SQV); slash indicates dual proteaseinhibitor use. NA, not applicable.
were V82A/T/F (18/35 [51%]), L90M (16/35 [46%]), and L10I with one or more zidovudine resistance mutations (codons 41, (14/35 [40%]). Of the 16 isolates with the L90M mutation, 12 67, 70, 215, 219) [8], and isolates from 28 patients (80%) had (75%) also had mutations at codon 10 and/or 82. Whereas 10 the M184V mutation, which confers resistance to zalcitabine (67%) of 15 isolates from patients with prior saquinavir therapy [8]. One isolate (3%) had a didanosine resistance mutation at had an L90M mutation, this mutation was also found in 6 codon 74 [8], and 4 (11%) had one or more nonnucleoside RT (30%) of 20 isolates from patients who were never treated with inhibitor resistance mutations (codons 103, 106, 108, 181, 188, saquinavir. Mutations at codon 54 were found in 4 (11%) of 190) [8]. The multidrug resistance mutation at codon 151 was 35 isolates and were always found in association with mutations at codon 82. No mutations were observed at codons 30 or 48.
Comparison of the sequences derived from plasma and cul- The I93L substitution previously noted as a naturally occurring tured virus isolates in the same patient revealed limited differ- polymorphism [10] was observed in 14 (40%) of 35 isolates.
ences. In 2 patients, these differences occurred at positions Multiple mutations in RT were observed in patient iso- associated with protease inhibitor resistance (tables 1 and 2).
lates, with a median of four mutations per isolate (range, none Virus cultured from patient 4 showed an A71T substitution, to six; table 2). The distribution of RT resistance mutations but the plasma virus remained wild type at this position, reflected the patterns of antiretroviral therapy in the Denver whereas plasma virus from patient 26 had an L10I substitu- metropolitan area. Thirty-four (97%) of 35 patients had isolates tion, but the cultured virus remained wild type. Discordance Mutations in HIV-1 reverse transcriptase gene at codons associated with drug resistance in patients with combination antiretroviral Amino acid residues are indicated by single-letter abbreviations. HIV-1 clade B consensus [11] is shown for comparison. — indicates no change from consensus sequence; ? indicates indeterminate sequence. No substitutions were observed at codons 106 and 108. NA, not applicable.
between cultured and plasma virus was also observed in the quencies in the population. These rare variants may have the RT sequence of patient 26. The M41L and T215Y substitutions potential to emerge rapidly as the predominant quasispecies in were present in cultured virus but not in plasma virus.
response to the appropriate selective pressure [6]. Althoughsuch variants might not have been detected by the methodsused in our study, we do not believe that they can account for Discussion
therapeutic failure in our patients. Given the rapid replicationrate of HIV-1, the circulating virus population should be rep- We have determined the population consensus sequence for resentative of the actively replicating population.
HIV-1 PR and RT genes in HIV-1 isolates from 35 patients in Four patients in our study had HIV-1 isolates with no known whom combination antiretroviral therapy that included a pro- resistance mutations in the PR gene. Clinical failure in the ab- tease inhibitor had failed. Although resistance mutations in PR sence of drug resistance implies inadequate drug exposure due and RT were common in patients with clinical failure of com- to pharmacologic factors or suboptimal adherence to drug ther- bination therapy, the number of PR mutations was lower than apy. These results are qualitatively similar to those of another expected. These data suggest that treatment failure can be as- study in which nearly 50% of patients with protease inhibitor sociated with relatively few PR mutations.
therapy failure had a wild type HIV-1 PR gene [14]. The dif- It is possible that those patients with protease inhibitor ther- ferent proportions of patients with wild type isolates in these apy failure who had two or fewer mutations may have had two studies may reflect differences in patient adherence to drug lower levels of drug exposure because of pharmacologic factors regimens or differences in clinician assessment of adherence.
or poor adherence to their prescribed treatment regimen. Ex- An additional study of patients in whom PR inhibitor therapy tragenic mutations, such as those at protease cleavage sites in failed revealed that genotypic evidence of resistance to each the gag-pol precursor, may also be associated with protease component of a multidrug regimen was not required for ther- inhibitor resistance [12]. The genetic analysis done in this study apeutic failure [15]. Collectively, these studies make the point that not all treatment failure is due to viral drug resistance and Mutations at codon 90 were common, particularly in isolates reinforce the importance of assessing treatment adherence.
from patients who were saquinavir-experienced. The L90M mu-tation also was detected in isolates from 4 saquinavir-naivepatients treated with indinavir or ritonavir, usually in associ- Acknowledgments
ation with other PR resistance mutations (e.g., V82A/T/F orI84V). This finding suggests that L90M may contribute to fail- We thank David Poticha and Elizabeth Brown (University of Col- ure of indinavir or ritonavir treatment. One implication of these orado) and Victoria Johnson (University of Alabama, Birmingham)for helpful discussions and Karen Kohler for expert editorial assistance.
results is that selection of the L90M substitution by saquinavirmay impair the subsequent response to indinavir or ritonavir.
In this regard, it is noteworthy that the virologic response to References
indinavir following long-term treatment with the hard gel cap-sule formulation of saquinavir was relatively modest, compared 1. Ho DD, Neumann AU, Perelson AS. Rapid turnover of plasma virions and CD4 lymphocytes in HIV-1 infection. Nature 1995; 373:123–6.
with the ≥2 log reduction in plasma HIV-1 RNA level reported 2. Wei X, Ghosh SK, Taylor ME, et al. Viral dynamics in human immuno- in protease inhibitor–naive subjects [4, 13].
deficiency virus type 1 infection. Nature 1995; 337:117–22.
Fourteen of 35 isolates in this study carried a mutation at 3. Carpenter CCJ, Fischl MA, Hammer SM, et al. Antiretroviral therapy for PR codon 93. Although the I93L substitution has been de- HIV infection in 1997. JAMA 1997; 227:1962–9.
scribed as a naturally occurring polymorphism, the frequency 4. Gulick RM, Mellors JW, Havlir D, et al. Treatment with indinavir, zido- vudine, and lamivudine in adults with human immunodeficiency vi- of this substitution appears to be substantially higher in our rus infection and prior antiretroviral therapy. N Engl J Med 1997; 337:
treated population than in protease inhibitor–naive patients 5. Hammer SM, Squires KE, Hughes MD, et al. A controlled trial of two Previous studies have shown that drug resistance mutations nucleoside analogues plus indinavir in persons with human immunode- may be detected in plasma virus sooner than in cultured vi- ficiency virus infection and CD4 cell counts of 200 per cubic millimeter or less. N Engl J Med 1997; 337:725–33.
rus [2]. In the present study, the frequency of drug resistance 6. Condra JH, Holder DJ, Schleif WA, et al. Genetic correlates of in vivo viral mutations could have been underestimated, since many of the resistance to indinavir, a human immunodeficiency virus type 1 protease sequences were derived from cultured virus. However, plasma inhibitor. J Virol 1996; 70:8270–6.
virus sequences from 4 patients also showed insignificant 7. Molla A, Korneyeva M, Gao Q, et al. Ordered accumulation of mutations changes in PR (patients 5, 12, 38, 47), providing greater con- in HIV protease confers resistance to ritonavir. Nature Med 1996; 2:760–6.
8. Schinazi RF, Larder BA, Mellors JW. Mutations in retroviral genes associated with drug resistance. Int Antiviral News 1997; 5:129–42.
Population-based sequencing methods are limited by their 9. Division of AIDS, National Institute of Allergy and Infectious Diseases.
inability to detect virus variants that are present at low fre- DAIDS virology manual for HIV laboratories. Washington, DC: US Department of Health and Human Services, 1997; publication NIH-97-
from saquinavir hard capsule (SQVhc) to saquinavir soft gelatin capsule (SQVsgc) vs. switching to indinavir (IDV) after prior saquinavir [abstract].
10. Kozal M, Shah N, Shen N, et al. Extensive polymorphisms observed in HIV- In: Program and abstracts of the 35th annual meeting of the Infectious 1 clade B protease gene using high-density oligonucleotide arrays. Nature Diseases Society of America (San Francisco). Alexandria, VA: IDSA, Med 1996; 2:753–9.
1997:21.
11. Myers G, Wain-Hobson S, Henderson LE, Korber B, Jeang KT, Pavlakis 14. Mayers DL, Gallahan DL, Martin GL, et al. Drug resistance genotypes from GN. Human retroviruses and AIDS 1993. Los Alamos, NM: Los Alamos plasma virus of HIV-infected patients failing combination drug therapy National Laboratory, 1993.
[abstract 80]. Antiviral Ther 1997; 2(suppl 5):72.
12. Zhang YM, Imamichi H, Imamichi T, et al. Drug resistance during indinavir 15. Johnson VA, Nail CD, Hazelwood JD, et al. Multidrug-resistant HIV-1 iso- therapy is caused by mutations in the protease gene and in its gag substrate lates from patients with rising plasma viral burden during double or triple cleavage sites. J Virol 1997; 71:6662–70.
combination therapy in clinical practice [abstract 77]. Antiviral Ther 13. Para MF, Collier A, Coombs T, et al. ACTG 333: antiviral effects of switching 1997; 2(suppl 5):72.

Source: http://www.wildlifescienceintl.com/WSI,_Inc/Published_Works_files/HIVresistance.pdf