Drug-Resistant HIV-1: Title and subTitle BreakThe Virus Strikes Back

The human immunodeficiency virus (HIV) has developed elaborate mechanisms to escape the immune system. Since the reverse transcriptase of HIV makes, on average, 1 error per 10000 bases copied, and the virus has a 9200-base genome, each virus produced is slightly different from its forebear. Over time each patient develops a swarm of virus variants (quasi species) with all possible 1-base and most 2-base variants represented. Additionally, if 2 viruses infect 1 cell or 2 infected cells get fused by HIV infection to form a syncytium, 2 different viral genomes can be packaged into the virus produced. When these viruses with 2 different HIV genomes infect the next cell, the reverse transcriptase enzyme can switch back and forth between the 2 genomes to form a recombinant virus with portions of genes from both parent viruses. The mechanisms of mutation and recombination, which the virus uses to escape immune pressure, are readily used to escape pressure exerted by antiviral drugs, resulting in emergence of drug-resistant virus.¹

Two years after the introduction of zidovudine into clinical practice, virus with 100-fold decreased susceptibility to the drug was detected in clinical samples.² Subsequently, drug-resistant HIV has been detected for every antiretroviral drug introduced into clinical practice. Current drug regimens using drugs targeted at the viral reverse transcriptase and protease enzymes can reduce viral replication by 300- to 1000-fold and reduce circulating virus to below current limits of detection in most drug-naive patients starting therapy. The current goal of therapy for infected patients is to maintain the patient's health by suppressing circulating virus below the limits of detection indefinitely. Studies of HIV from lymph nodes of patients with undetectable circulating virus while on combination drug regimens have shown that current drug regimens, while not able to eradicate HIV, can prevent effective viral replication and development of drug resistance.³ This has led to increasing use of combination antiretroviral drug regimens in many patients soon after they are found to be seropositive.

Adherence is a major problem for any disease in which patients must take medications for prolonged periods. For chronic diseases, such as hypertension, studies have shown that typically 50% of patients are not taking medications as prescribed at any point in time.⁴ These problems are compounded for antiretroviral combination drug regimens because of the numerous pills, substantial drug toxicity, many drug and food interactions, and complicated schedules. Because the population of patients eligible for treatment now includes those with earlier stages of HIV disease, clinicians have moved from treating only symptomatic patients to treating asymptomatic patients, who may feel worse while receiving medication than without it, leading to problems with medication adherence.⁵ Patients who stop taking all of their medications will experience a rebound in their viral load with drug-sensitive virus.⁶ Those patients taking their pills intermittently or inconsistently will eventually also have a rebound in circulating virus but it will be drug resistant.⁶ With currently available antiretroviral drugs, there is substantial cross-resistance between agents in each class of drugs; thus, realistically, there may be only 2 possible sequential regimens for each patient.

Current data indicate that a growing number of patients is failing combination drug regimens and that finding effective salvage therapy for them is difficult.⁷ Also, transmission of drug-resistant virus is occurring via sexual contact, intravenous drug use, perinatal transmission, and accidental exposure of health care workers or public safety personnel to blood containing HIV.^{8 - 10}

The clinician with a patient in whom antiretroviral combination drug therapy fails may wish to use antiretroviral drug–resistance testing for help in patient care management. Advances in molecular technologies now allow rapid sequencing of circulating virus genes to detect mutations associated with HIV-drug resistance or the generation of recombinant viruses from patient samples to rapidly obtain a drug-resistance profile in vitro. The consensus report by Hirsch et al¹¹ in this issue of THE JOURNAL provides a good review of the technologies available to assess HIV drug resistance, their strengths and weaknesses, and potential utility for clinical management of HIV infection. As stated in that report, these commercially available technologies have yet to be clinically validated. The use of HIV drug-resistance testing for clinical decisions is at the stage of evaluation similar to that seen for HIV RNA levels 2 to 3 years ago. Clinical validation trials are in progress in the United States and Europe with 1 trial (the Terry Beirn Community Programs for Clinical Research in AIDS 046 study, the Genotypic Analysis Resistance Testing study) having more than 50% of patients enrolled.

The problem for the clinician is whether to use these tests now to manage care of patients in whom combination-drug regimens failed. While these tests are relatively expensive, based on my experience, costs for either phenotypic or genotypic drug-resistance testing are less than the cost of 1 month of combination antiretroviral therapy. Using available information, guidance can be provided for clinicians who choose to use HIV drug-resistance testing for patient evaluation along with CD4⁺ cell count and plasma HIV RNA level. Since drug-sensitive virus will overgrow drug-resistant virus in absence of drug pressure, drug-resistance testing should not be done using samples from patients who have been HIV positive for an unknown period and are starting antiretroviral therapy, or patients who have been off therapy for several weeks. As stated in the consensus report,¹¹ serial plasma HIV RNA levels and CD4⁴ cell counts should be used to decide when to start or change antiretroviral drug regimens. When there is evidence of virologic failure, drug-resistance testing offers information in addition to treatment history, plasma HIV RNA levels, and history of patient adherence to drug regimens. It is critical that the sample for resistance testing be obtained while the patient is still taking the failing drug regimen, prior to stopping use of the drugs or switching to a new regimen. Drug resistance is more likely if the patient had an initially good clinical response to the regimen and later failed than if the patient's viral load never decreased with a regimen that should have been effective.

One area of potential utility for drug-resistance testing is helping clinicians separate patients in whom therapy is failing due to HIV drug resistance from those who simply cannot take the medications for whatever reason. Studies indicate that a substantial minority of patients in whom combination drug regimens are failing will fail with virus remaining sensitive to the protease inhibitor in the regimen.⁶ On close questioning, many of the patients found the agents too difficult to take but did not relate this to their health care practitioner. It is important to separate adherence failure from resistance failure since one group of patients needs a more user-friendly regimen and the other needs a complex salvage regimen. Patients in whom therapy fails without having drug-resistant virus may have previously developed drug resistance but stopped taking their drugs prior to testing; thus, treatment history is still critical for choosing a new combination regimen even in patients with apparently sensitive virus in whom therapy fails.

For patients in whom therapy fails because of drug-resistant virus, knowledge of cross-resistance patterns may be useful in selecting drugs that may provide reasonably good antiretroviral activity.¹² As stated by the International AIDS Society–USA Panel,¹¹ a major effort is needed to standardize and validate these assays, develop standardized reporting formats easily understood by practicing clinicians, develop better correlates between drug-resistance mutations and phenotypic resistance, and relate drug-resistance mutations or phenotypic drug-resistance levels to subsequent virologic responses to combination drug regimens.

From a public health perspective, the rapidly increasing number of patients in whom combination therapy regimens fail raises the specter of increasing transmission of drug-resistant HIV to the next generation of patients. The article by Wainberg and Friedland⁵ in this issue of THE JOURNAL reviews the current state of knowledge regarding spread of drug-resistant HIV-1. The original guidelines for use of zidovudine that restricted the drug to patients with late-stage HIV disease were associated with low levels of transmission (1 [1.4%] of 69) of drug-resistant virus to new seroconverters in a study of Swiss, Australian, and US patients.¹³ Subsequently, with the release of the results of the AIDS Clinical Trials Group 016 and 019 studies in 1990, use of zidovudine was expanded to patients with a CD4⁺ cell count of less than 0.50×10⁹/L. Two years later, the rate of zidovudine resistance had risen in new seroconverters to 4 (7.5%) of 53 in 1992 and then to 17 (10.4%) of 164 in 1993 and has remained at roughly that level for the past 5 years.¹³ Lamivudine resistance was detected in new seroconverters 1 year after introduction of the drug into clinical practice.¹⁴ There are anecdotal reports of nonnucleoside reverse transcriptase inhibitor–resistance mutations in virus from new seroconverters and surveillance for protease inhibitor resistance in new seroconverters has recently been initiated.¹⁵ It is critical to emphasize to our patients with HIV infection on an ongoing basis the need for practicing safe sex or abstinence and avoiding needle sharing.

The use of zidovudine to prevent perinatal HIV-1 transmission has been remarkably effective.¹⁶ For the small number of infected infants whose mothers took antiretroviral medications during pregnancy (especially if taken near the time of delivery), risk of drug resistance may be increased. While there is conflicting data about the incidence of drug-resistance transmission with zidovudine prophylaxis, rates of up to 29% for maternal drug resistance at term have been described with use of zidovudine prophylaxis (especially for women receiving zidovudine prior to pregnancy).^{17 - 19} These rates will only increase as women have increased access to antiretroviral drugs early in the course of their disease prior to pregnancy.

Similar concerns arise for health care workers and public safety personnel who are exposed to blood containing HIV.²⁰ The choice of drugs to offer a health care worker exposed to blood from a patient in whom several courses of antiretroviral therapy have failed has become problematic. The same drug regimens are used to treat patients that are used for prophylaxis of blood exposures.

Drug-resistant HIV-1 will increasingly challenge clinicians and public health agencies in ways similar to that of drug-resistant bacteria such as Staphylococcus aureus, pneumococci, and Mycobacterium tuberculosis. The challenge to pharmaceutical companies is to develop drug combinations with once-a-day or twice-a-day dosing to improve adherence. Management of salvage therapy for patients failing combinations will remain difficult. The potential role of HIV drug-resistance testing in patient care management will be rapidly determined by ongoing clinical trials and the experience of the clinical community. The public health impact of drug-resistant HIV-1 will likely increase for the foreseeable future and the rapid implementation of active surveillance systems similar to those developed for drug-resistant bacteria are needed. These problems highlight the need to develop an effective HIV vaccine. Drug-resistant HIV-1 is here to stay.