The Case for Conservative Management of Early HIV Disease

THE LAST 3 years have been a remarkable time to be involved in the care of patients with human immunodeficiency virus (HIV) infection. Improvements in antiretroviral therapy have decreased the rates of death and opportunistic infections by about 70% in our clinic (W.J.B., R.R.R., and D.L.C., unpublished data, 1998) and brought an atmosphere of hope to patients and clinicians. However, we read with concern the 1997 recommendations from the International AIDS Society–USA Panel regarding the use of therapy for all patients with an HIV RNA level greater than 5000 to 10000 copies/mL, regardless of CD4 cell count.¹ Aggressive treatment of early HIV infection was also recommended by a panel convened by the Department of Health and Human Services.² The Department of Health and Human Services panel was more conservative and presented the potential advantages and disadvantages of aggressive, early therapy, but eventually recommended treatment of any patient with a CD4 cell count less than 0.50×10⁹/L (<500/µL) or a viral load greater than 10000 to 20000 copies/mL.

These expert panels recognized that the strategy of early aggressive therapy has not been evaluated in controlled clinical trials but argued that there was a compelling scientific rationale for advocating this approach. Active replication of HIV occurs throughout the course of infection and causes progressive damage to the immune system. The panels postulated that improvements in immune function related to antiretroviral therapy are incomplete and that early therapy is necessary to prevent irreversible immune deficits. Finally, the panels emphasized that the high mutation rate of HIV mandates multidrug therapy in order to prevent therapeutic failure due to emergence of drug-resistant viral strains.

Two additional conditions are necessary for the recommendation of aggressive therapy for early HIV disease: (1) that therapy is relatively nontoxic (ie, few serious adverse effects), and (2) that adherence to the currently available 3-drug regimens will be sufficient to prevent emergence of drug resistance. We will argue that there is growing evidence of immune reconstitution following antiretroviral therapy, that antiretroviral drug safety has not been adequately evaluated to recommend long-term use in persons with early HIV disease, and that nonadherence to the complex regimens recommended by the panels will result in high rates of acquired drug resistance.

The linchpin of the case for early therapy is the assertion that continued viral replication causes irreversible immune damage at CD4 cell counts well above 0.20×10⁹/L. Supporters of early therapy cite the results of in vitro studies showing an incomplete restoration of T-cell receptor repertoire in patients responding to antiretroviral therapy.³ However, others have reported substantial treatment-related improvements in a number of in vitro assays of lymphocyte function.^{4 - 5} Currently, the in vitro evidence regarding reversibility of HIV-associated immunodeficiency is inconclusive.

Clinical data, however, offer clear evidence of the potency of the immune response to antiretroviral therapy. Dramatic declines in opportunistic infections have been reported⁶ among patients with CD4 cell counts less than 0.10×10⁹/L, showing that antiretroviral therapy results in marked improvement in the immune function of persons with advanced acquired immunodeficiency syndrome (AIDS). Further evidence of immune restoration with therapy are early reports suggesting that maintenance therapy for opportunistic infections may not be needed for patients responding to antiretroviral therapy.⁷ Given this clinical evidence, it is premature to conclude that ongoing HIV replication in persons with CD4 cell counts well above 0.20×10⁹/L causes irreversible immune damage.

A key factor in the discussion of management of early HIV infection is the variability in clinical course of this infection. Although median time to development of AIDS is 10 to 12 years, some persons progress to death within 2 years and others appear healthy after 15 years.⁸ For those with advanced HIV disease, the possibility of drug adverse effects has limited relevance in decisions about initiating therapy; dramatic improvements have been documented in clinical trials, despite drug adverse effects.⁹ However, in patients with low short-term risks of HIV complications, risks of drug adverse effects become critical determinants in overall risk-benefit assessment of early treatment.

There is insufficient evidence to conclude that initiation of life-long therapy with antiretroviral drugs in persons at low risk of disease progression carries an acceptably low risk of long-term adverse effects. Nucleoside analogues have been associated with a rare, often fatal syndrome of intractable lactic acidosis and hepatic steatosis, particularly in women¹⁰ (a group that did not comprise a substantial percentage of those evaluated in initial studies of these drugs). Gastrointestinal adverse effects of protease inhibitors are common, and there are now indications of metabolic adverse effects as well, including hyperlipidemia, lipodystrophy, and insulin resistance. These metabolic adverse effects are particularly notable; they are common¹¹ and may predispose patients to accelerated atherosclerosis,¹² yet were not appreciated in prelicensing studies. There are numerous examples of medications that were efficacious in reducing a disease manifestation in early trials, only to be shown to result in substantial morbidity or mortality from an unexpected adverse effect when extended to patient populations other than those evaluated in formal trials.¹³

Aggressive antiretroviral therapy using multidrug regimens markedly decreases emergence of drug resistance.¹⁴ Thus, the panels recommended 3-drug regimens including a protease inhibitor for all patients starting therapy. However, there is increasing evidence that sustained viral suppression requires close adherence to therapy, and the recommended regimens require ingestion of 8 to 16 pills divided into 2 to 5 different doses per day. The question of adherence to long-term, complex medical regimens is crucial in evaluating the strategy of aggressive early therapy.

Studies of adherence to other long-term medical therapies suggest that adherence to antiretroviral therapy will be problematic, particularly in the context of early HIV disease. Adherence is inversely correlated with regimen complexity. Using highly sensitive assessment techniques, adherence with once-daily medications ranged from 62% to 87%, but dropped to 37% to 39% with medications prescribed for 3 or 4 times per day.^{15 - 16} Symptom severity correlates with adherence; it is difficult for asymptomatic persons to adhere to even simple drug treatment regimens for long periods, eg, only 64% of tuberculin-positive physicians completed a recommended course of once-daily isoniazid preventive therapy.¹⁷ Finally, clinicians poorly predict patient adherence,¹⁸ so it will be difficult to select patients who will have excellent adherence to complex medical regimens based on routinely available clinical information.

The available data suggest that these lessons regarding adherence will apply to HIV treatment, eg, only 65% of patients prescribed a protease inhibitor were reported adherent, despite the generous definition of nonadherence as failing to fill 2 or more monthly prescriptions within 6 to 12 months of follow-up.¹⁹ Another telling example of the difficulties in adhering to complex therapy regimens that may have adverse effects for asymptomatic persons comes from studies of health care workers given triple-drug postexposure prophylaxis; 61% to 92% reported adverse effects and only 22% to 56% completed a 1-month course of prophylaxis.^{20 - 21}

Nonadherence leads to resistance to the treatment regimen. There is also growing evidence of cross-resistance between drugs in a specific class; nonadherence to 1 protease inhibitor, for example, may markedly decrease activity of any subsequent protease inhibitor.²² Therefore, nonadherence with an initial course of antiretroviral therapy may markedly limit subsequent treatment options. Acquired drug resistance may have another adverse consequence as well. Transmission of drug-resistant HIV is occurring and appears to be increasing.²³ Rising rates of primary drug resistance in HIV may reverse some of the achievements of antiretroviral therapy and must be considered in decisions regarding the treatment of early HIV disease. There are already disturbing signs that availability of potent antiretroviral therapy may decrease perceived risk associated with becoming HIV infected and, thereby, increase unsafe behavior.²⁴

The example of tuberculosis treatment has been cited to illustrate the need for multidrug antiretroviral regimens to prevent drug resistance. There is another equally important lesson from tuberculosis treatment. Over 20 years ago, clinical trials showed that multidrug tuberculosis treatment regimens could cure more than 95% of patients and virtually eliminate acquired drug resistance. One might expect then, that rates of drug-resistant tuberculosis would have decreased in subsequent years as these regimens were adopted in clinical practice. Quite the opposite occurred; around the world, rates of drug resistance increased.²⁵ The lesson of tuberculosis is crucial—regimens that are highly efficacious in clinical trials may not prevent the emergence of drug resistance in clinical practice; the application of potent antimicrobial regimens in clinical practice without aggressive attention to adherence may promote, rather than prevent, drug resistance.

How should clinicians counsel patients with HIV about treatment initiation? Patients should be told that combination therapy including a protease inhibitor provides major benefits⁹ for those with CD4 cell counts less than 0.20×10⁹/L. Patients with higher CD4 cell counts should be informed that the optimal time for initiating therapy is unknown. Given the uncertainties about long-term adverse effects of drugs, consequences of nonadherence, and cross-resistance, we agree with the more conservative guidelines of the British HIV Association²⁶ that therapy should be recommended to patients with CD4 cell counts less than 0.30×10⁹/L as well as those with higher CD4 cell counts and HIV RNA levels greater than 10000 to 50000 copies/mL. For patients whose CD4 cell count and viral load indicate a good intermediate prognosis (CD4 cell count, >0.30×10⁹/L; HIV RNA, <10000-50000 copies/mL), there may be significant advantages to the strategy of deferred therapy. Deferred therapy preserves therapeutic options during a phase of HIV disease when adherence to complex medical regimens is particularly problematic; deferred therapy delays the risk of troublesome and serious adverse drug effects. During discussions of antiretroviral therapy, we show patients their individualized risk of progression to AIDS, based on CD4 cell count and viral load,² and discuss the potential benefits and hazards of therapy. For those patients interested in starting antiretroviral therapy, we recommend a self-evaluation of potential adherence to complex therapy using a "placebo pillbox," filled to match the schedule of the regimen being considered. All of this can and should take time; lifelong is a long time.

Finally, it should be clear as clinicians and patients that we need more data. Rather than promote early aggressive therapy as the standard of care, we suggest that long-term studies be undertaken that will evaluate not only the short-term virologic and immunologic effects of early therapy, but also the clinical effectiveness, adverse effects, adherence, and quality of life.