ABSTRACT
In order to analyze the impact of protease gene polymorphism on response to regimens containing a protease inhibitor, the entire protease coding domain from 58 human immunodeficiency virus type 1 (HIV-1)-infected patients who were protease inhibitor naive was sequenced before therapy was started. Plasma HIV-1 RNA levels were measured at baseline and at month 3 and month 6 after treatment. All patients were treated with a combination of two reverse transcriptase inhibitors and a protease inhibitor (saquinavir EOF [n = 28], ritonavir [n = 16], or indinavir [n = 14]). Before treatment, 30 different positions whose codons differed from the subtype B consensus sequence were observed. Major mutations associated with protease inhibitor resistance were not observed. No statistical correlation between the number of amino acid differences and the treatment efficacy at month 3 (−2.4 log) or month 6 (−2.7 log) was observed. At baseline, genotypic analysis of the HIV-1 protease gene of patients who have never received a protease inhibitor does not allow prediction of the efficacy of regimens containing a protease inhibitor.
Protease inhibitor regimens are associated with a decrease in plasma human immunodeficiency virus type 1 (HIV-1) RNA levels, an increase in CD4 lymphocytes, and prolonged survival of patients infected with HIV-1 (5). The appearance of drug-resistant virus under this treatment is possible. Protease inhibitor-resistant variants of HIV-1 have been described in vitro and in vivo, and numerous mutations in the protease have been determined (12, 13). Two categories of mutations have been reported: major mutations, usually located in the active site of the protease, and accessory mutations, located outside the active site. Both categories are usually required for high-level resistance (1, 2, 11, 25). However, the presence of numerous accessory mutations has been reported for patients who had not been previously exposed to a protease inhibitor (1, 8, 12, 13, 16, 21). In vitro, these accessory amino acid substitutions do not significantly alter the susceptibility of the protease inhibitors (1, 12, 16). However, during therapy, these amino acid substitutions have been described to have an influence on 50 and 90% inhibitory concentration increases (3, 14).
At baseline, the impact of these amino acid substitutions on response to treatment including a protease inhibitor is not known. At baseline, the influence of substitutions in the protease gene of patients naive for protease inhibitors on the virologic response to combinations including a protease inhibitor and two nucleoside reverse transcriptase inhibitors was analyzed.
(This work was presented in part at the 2nd International Workshop on HIV Drug Resistance and Treatment Strategies, Lake Maggiore, Italy, 24 to 27 June 1998, and at the 38th Interscience Conference on Antimicrobial Agents and Chemotherapy, San Diego, Calif., 24 to 27 September 1998.)
MATERIALS AND METHODS
At baseline, the sequence of the protease gene from 58 patients who were protease inhibitor naive was characterized. Among these patients, 12 had been treated previously with a combination of two nucleoside reverse transcriptase inhibitors. Plasma HIV-1 RNA levels were measured by Roche-Monitor assay (Roche Diagnostic Systems, Inc., Branchburg, N.J.), including the ultrasensitive assay, before introduction of the protease inhibitor and at 3 and 6 months afterward (lower limit of quantification, 20 copies/ml).
All patients were treated with two reverse transcriptase inhibitors and a protease inhibitor with usual daily dosages: 28 patients received saquinavir EOF (1,200 mg three times a day), 16 received ritonavir (600 mg twice a day), and 14 received indinavir (800 mg three times a day).
HIV-1 RNA was extracted from 200 μl of plasma by using the HCV specimen preparation kit (Roche Diagnostic Systems, Inc.), reverse transcribed to cDNA, and then directly amplified by nested PCR. The primers used for cDNA synthesis and PCR amplification and cycle conditions have been previously described (4). The entire protease coding domain from the 58 patients before introduction of the protease inhibitor was sequenced on an automated DNA sequencer (Applied Biosystems model 377) and compared to the HIV-1 clade B consensus sequence (15).
Statistical analyses were performed with nonparametric tests (Spearman’s rank correlation and Mann-Whitney U test).
RESULTS
Only one patient had a protease gene identical to the HIV-1 clade B consensus sequence. The protease sequences of the other 57 patients (98%) carried 1 to 9 amino acid differences from the consensus sequence. The median number of substitutions was 4: 34 patients harbored between 1 and 4 amino acid changes, and 23 had between 5 and 8 changes. A total of 30 positions differed from the consensus sequence, while the most frequent changes (prevalence, >20%) were located at positions 15, 35, 37, 41, 63, 77, and 93 (Table1). The substitution at position 63 was particularly frequent, as it was observed in 58% of the cases. The 63P substitution (44%) represented the most common amino acid change in the protease.
Incidence of amino acid substitutions in the protease (n = 58)
The major amino acid mutations associated with reduced sensitivity to protease inhibitors, at positions 30, 48, 50, 82, 84, and 90, were not observed.
The median plasma HIV-1 RNA levels were 47.479 copies/ml before introduction of the protease inhibitor, 162 copies/ml (−2.4 log10) at month 3, and 89 copies/ml (−2.7 log10) at month 6 after treatment. At months 3 and 6, 7 (12%) and 26 (45%) of 58 patients, respectively, achieved complete suppression of HIV-1 RNA in plasma (<20 copies/ml).
With Spearman’s rank correlation test, no statistical correlations between the number of protease amino acid substitutions and the decrease in HIV-1 RNA levels at months 3 and 6 were observed. Moreover, the presence of the 63P mutation did not influence the evolution of HIV-1 RNA levels.
The same results were observed when the tests were performed for each drug separately (saquinavir, ritonavir, and indinavir). Moreover, the numbers of amino acid substitutions were not significantly different between patients with complete therapeutic success (<20 copies/ml) and those without complete therapeutic success (>20 copies/ml) (Mann-Whitney U test).
DISCUSSION
No association between the number of amino acid substitutions present at baseline and the response to treatment including a protease inhibitor was observed.
Many mutations in the protease gene that cause inhibitor resistance have been described in vitro and in vivo (1-14, 16-26). The most common major mutations associated in vivo with resistance to indinavir and ritonavir are at position 82, those associated with resistance to nelfinavir are at position 30, those associated with resistance to saquinavir are at positions 48 and 90, and those associated with resistance to amprenavir are at position 50 (2, 4, 17, 20, 23). These major mutations usually occur first in the presence of a protease inhibitor. With the exception of the mutation at position 90, they are located in the active site of the protease. These mutations are associated with phenotypic resistance by reducing the binding affinity of the protease inhibitor to HIV protease (6). In this study, these kinds of mutations in the patient’s baseline protease sequence were not observed. However, many substitutions that have been described as accessory mutations, at positions 10, 33, 35, 36, 63, 71, and 77, were found in the protease from untreated subjects (12, 18, 20).
These accessory mutations are mainly located outside of the active site. Usually, molecular clones containing these changes do not reduce significantly the susceptibility to protease inhibitors in vitro. In vivo, rare cases of significant phenotypic change associated with these substitutions have been described (3). However, the combination of several accessory mutations with a major mutation is usually responsible for an increase in phenotypic resistance (1, 2, 11, 12, 25, 26).
Some observations of sequential early patient samplings have demonstrated that polymorphic changes precede the occurrence of resistance mutations, suggesting that polymorphic changes may serve as early markers of development of antiretroviral resistance (14). In this study, no statistical correlation between the number of substitutions in the protease gene from untreated patients at baseline and the response to treatment including a protease inhibitor was observed. Moreover, the number of accessory substitutions did not influence the rate of complete HIV-1 RNA viral load suppression (HIV-1 RNA level, <20 copies/ml).
At baseline, genotypic analysis of the HIV-1 protease gene from patients who have never received a protease inhibitor does not allow prediction of the efficacy of regimens containing a protease inhibitor.
FOOTNOTES
- Received 4 January 1999.
- Returned for modification 19 April 1999.
- Accepted 14 June 1999.
- Copyright © 1999 American Society for Microbiology
