Drug Resistance Patterns, Genetic Subtypes, Clinical Features, and Risk Factors in Military Personnel with HIV-1 Seroconversion FREE

Genetic variability is a central feature of HIV-1. The high frequency of mutations during HIV-1 replication leads to the development of viral quasi-species in vivo and contributes to genetic heterogeneity among HIV-1 isolates (1 - 2). There is a growing appreciation that HIV-1 genetic diversity, including the existence of distinct genetic subtypes and the evolution of drug-resistant genotypes, can greatly affect the diagnosis and treatment of HIV-1 infection (1 - 2). On the basis of DNA sequence analysis, HIV-1 has been classified into genetic subtypes, with subtypes A through I making up the major HIV-1 group (group M). The more genetically diverse groups, O and N, have also recently been described (1). For reasons that are not clear, HIV-1 subtypes are variably dispersed throughout the world. Some regions, such as central Africa and eastern Europe, have multiple circulating subtypes, whereas the distribution in other regions is more restricted (1,3). Differences among HIV-1 subtypes can affect the sensitivity of some diagnostic assays of antibody (4) and plasma HIV-1 RNA (5 - 6). Although non-subtype B infection has been reported in the United States, the prevalence in the U.S. population is unknown (7 - 8).

Widespread use of antiretroviral drugs has led to transmission of drug-resistant HIV-1, but the prevalence of resistant mutations in treatment-naive persons has not been thoroughly studied. Because all U.S. Navy and U.S. Marine Corps personnel are screened for HIV-1 infection at 1- to 3-year intervals, HIV-1 infection is often detected early and the seroconversion period can be estimated (9). This permits an investigation of the epidemiologic correlates and risk behaviors associated with the acquisition of non-subtype B and drug-resistant infections. Thus, we determined the prevalence of non-subtype B infection and genotypes associated with antiretroviral drug resistance in a well-characterized cohort of military personnel with recently acquired HIV-1 infection.

United States Navy or Marine Corps personnel with HIV-1 seroconversion who are assigned to military bases west of the Mississippi River in the United States or in the Pacific region overseas are referred to the Navy Medical Center San Diego for initial and follow-up HIV-1 evaluations. Between February 1997 and February 1998, 99 of 141 personnel referred were eligible on the basis of documented seroconversion within the past 3 years. Ninety-five of 99 patients enrolled and signed a consent form approved by the institutional review board of the Navy Medical Center San Diego. Thirty-two patients were being referred for their first HIV evaluation, and 63 were enrolled during a follow-up visit. Blood was obtained to determine genetic subtype and the presence of antiretroviral drug resistance.

All patients completed a self-administered risk factor questionnaire labeled with a unique code number. Surveys were sealed in an envelope and placed in a locked drop box that was emptied weekly by off-site data entry personnel. Clinical research staff extracted clinical and laboratory information from the medical record and sent this information, identified by code number, to the data entry site.

Seroconversion was documented by a previous negative result on whole-virus HIV-1 enzyme immunoassay followed by a positive result on enzyme immunoassay and a confirmatory positive result on Western blot assay. Lymphocyte subset analysis was performed by flow cytometry, and plasma HIV-1 RNA (available after July 1996) was measured by quantitative reverse transcription polymerase chain reaction assay (Amplicor HIV-1 Monitor assay, Roche Molecular Systems, Branchburg, New Jersey). Genetic subtyping of HIV-1 was performed by using a two-step algorithm. Sera were screened by using a competitive binding enzyme immunoassay with peptides derived from the third variable loop of the HIV-1 envelope glycoprotein. For samples that were serologically reactive to non-subtype B peptides, DNA was extracted from corresponding peripheral blood mononuclear cells, and sequence analysis was performed over a 640-base pair segment of the HIV-1 envelope gene (10).

Testing for viral drug resistance was successfully performed in 31 of the 32 therapy-naive patients. Plasma-derived viral RNA was reverse transcribed into complementary DNA, amplified by polymerase chain reaction, and directly sequenced by using an automated ABI Sequencer (Applied Biosystems, Foster City, California). Consensus DNA sequences for the protease and reverse transcriptase genes from each participant were examined for mutations associated with HIV-1 antiretroviral resistance (11).

Descriptive analysis of demographics, risk behaviors, and laboratory results were performed. Duration of HIV infection was calculated starting from the time of onset, which we estimated as the midpoint between the last negative and first positive result on HIV enzyme immunoassay. Bivariate analyses using odds ratios and 95% CIs determined by the Fisher exact method were done to compare the treatment-naive patients with drug-resistant infection and those with drug-sensitive infection. Statistical analyses were done by using Epi-Info, version 6.02 (Centers for Disease Control and Prevention, Atlanta, Georgia).

Characteristics and risk exposures of the cohort are summarized in Table 1. Duration of infection and laboratory results are presented for the entire cohort on initial evaluation and for the treatment-naive cohort at the time of enrollment and antiretroviral testing. Risk behaviors known to be associated with acquisition of HIV-1 were reported during the period of seroconversion in all but two patients.

The CD4 cell counts, plasma HIV RNA levels, and results of syphilis and hepatitis serologic testing in study patients are summarized in Table 1. Combined serologic and genetic analysis revealed that 7 of 95 patients were infected with HIV-1 genetic subtype E (7.4%); the remaining patients were infected with subtype B. Eight of 31 treatment-naive patients (26% [95% CI, 12% to 46%]) had one or more primary mutations that have been associated with phenotypic drug resistance (11) (Table 2). All reverse transcriptase mutations were associated with resistance to zidovudine, lamivudine, or nevirapine-delavirdine. Of the 4 patients with reverse transcriptase mutations, 2 (patients 1 and 2) had mutations that could confer resistance to both nucleoside and non-nucleoside reverse transcriptase inhibitors.

Infection with HIV-1 subtype E was documented in 6 men and 1 woman; all 6 men reported sexual contact during short deployments in Thailand. The female patient reported having sex with a man on active duty in the United States. Compared with subtype B-infected patients, subtype E-infected patients were more likely to be heterosexual (100% and 38%), to have had overseas exposure (86% and 27%), and to report sex with commercial sex workers (86% and 15%). Comparisons between patients with wild-type (n = 23) and treatment-resistant (n = 8) genotypes are listed in Table 1.

We found a high prevalence of non-B genetic subtypes and antiretroviral drug-resistant mutations among treatment-naive military personnel with recently acquired HIV-1 infection. Seven of 95 patients (7.4%) were infected with HIV-1 subtype E. Eight of 31 treatment-naive patients (26%) had primary drug-resistant mutations; of these 8 patients, 3 had mutations in the reverse transcriptase gene only, 4 had mutations in the protease gene only, and 1 had mutations in both the reverse transcriptase and protease genes. Although our study is limited to a single clinical referral center and the sample is small, the results are generally similar to those of other studies that focused on selected drug-resistant mutations. Among several recent studies, 6% to 13% of treatment-naive patients had mutations for zidovudine resistance (12 - 14). Although data on the prevalence of mutations in the protease gene have not yet been published, transmission of HIV-1 that is resistant to multiple reverse transcriptase and protease inhibitors has been reported (15). Our data suggest that patients infected with drug-resistant virus were more likely to have acquired HIV in the United States and to report sexual contact with a person who is known to be infected with HIV; they were also less likely to be heterosexual.

Prospective studies of the treatment responses of drug-naive patients with resistant genotypes have not been performed, but some data suggest that baseline nucleoside analogue mutations can diminish the potency and duration of viral suppression by commonly used nucleoside analogue combinations (16). In addition, several reports suggest an inherent decrease in susceptibility to reverse transcriptase or protease inhibitors in some HIV-1 genetic subtypes (17). Thus, both the transmission of acquired mutations and the genetic subtype of HIV-1 may have important implications for treatment of HIV-1 disease.

Although we (7) and others (8) have previously described the introduction of non-subtype B infections into the United States, the current study is the first use of a large cohort of recently infected persons to describe the prevalence and associated risk factors for non-subtype B infection. Six of the seven patients infected with HIV-1 subtype E reported sexual contact during short deployments to Thailand; in contrast, most cases of subtype B infection were acquired in the United States. The acquisition of sexually transmitted diseases and HIV during overseas travel is not unique to the military; 5% to 20% of travelers report having sex with a new partner while abroad (18). The lack of available methods to determine HIV subtypes may contribute to underreporting of non-subtype B infection in the U.S. population (1 - 2).

The genetic variation among HIV-1 subtypes can affect both serologic testing for diagnosis of HIV-1 and the measurement of plasma viral load. Although current HIV-1 antibody assays based on whole-virus lysates seem to detect infection with all HIV-1 subtypes, previous antibody screening assays that were based on synthetic peptides or recombinant antigens had variable reactivity, particularly for detection of the most divergent HIV-1 groups: N and O (3 - 4). Similarly, because of primer mismatch, the Amplicor HIV-1 Monitor assay can underestimate plasma HIV-1 RNA levels in persons with non-subtype B infection, particularly subtypes A, E, F, and G (5 - 6). In some cases, the assay may fail to detect viremia despite plasma viral load in excess of 10⁵ copies/mL (5).

Potentially modifiable risk behaviors for HIV-1 were reported by all but two of the study patients. Possible targets for intervention programs, such as pre-deployment and sexually transmitted disease clinics, as well as specific risk behaviors to guide intervention content, were defined. Almost 70% of patients reported drinking alcohol during sexual encounters, and 50% reported that alcohol use contributed to failure to use a condom. Efficacy in HIV preventive strategies has been linked to formative research that defines relevant behaviors for risk and protection in selected populations (19 - 20).

In summary, we describe the prevalence, epidemiologic correlates, and risk behaviors associated with the acquisition of non-subtype B HIV-1 and drug-resistant genotypes in a cohort of military personnel with recently acquired HIV-1 infection. The findings emphasize the need for larger prospective studies to estimate transmission of drug-resistant virus in order to make recommendations for initial treatment regimens. Furthermore, improved access to laboratory technologies for identification of HIV-1 subtypes is needed.