Early in 1995, investigators from the Aaron Diamond Research Center and the University of Alabama examined the kinetics of HIV replication in patients with CD4 cell counts of less than 500 cells/mm³ [2, 3]. Their work had astonishing results. The replication of HIV produced an average of about 10⁹ new virions daily, the half-life of HIV in serum was about 1 day, 30% of the plasma virus burden was turned over daily, and the rate of destruction of CD4 cells was also about 30%. This was accompanied by high rates of viral mutation, with estimates of more than 10⁸ HIV variants by mid-stage disease. Previous concepts about this disease were often based on the assumption that viral latency explained the prolonged period of asymptomatic infection, and the result was an aggressive pursuit of co-factors that might explain the late clinical deterioration and gradual decrease in CD4 cell count. The demonstration of active viral replication throughout the course of HIV infection prompted enthusiasm for more aggressive treatment. Most persons in the field concluded that the recommendations of the Expert Panel, published in 1993, were badly outdated [4], and many adopted the thesis of "hit HIV, early and hard" [5]. The problem was the limited menu of drugs from which to choose.

These observations on viral kinetics have been accompanied by the rapid escalation of interest in a new class of drugs active against HIV: protease inhibitors. Human immunodeficiency virus protease cleaves polyproteins into functional protein products during the late stages of HIV replication [6-8]. The absence of this cleavage results in an immature virus that is incapable of infecting new cells. Protease inhibitors are three-dimensional constructs that are complementary to the active site of HIV protease and result in competitive inhibition of the enzyme. These are "designer drugs" that inhibit HIV in vitro, but the development of products for oral administration has been hampered by poor bioavailability due to reduced absorption and rapid excretion, toxicity, or both. Many of these problems have been conquered, and the result is that one protease inhibitor is now approved by the FDA, and the approval of several more is anticipated in 1996.

Saquinavir is the first protease inhibitor to acquire FDA approval [9]. It is a potent inhibitor of HIV replication in vitro, but in vivo activity (in terms of the decrease in viral RNA concentrations) is modest at the recommended doses because of poor absorption and rapid hepatic metabolism. Oral bioavailability is at best only 4%, even when administered with a "British breakfast" consisting of a 1000-calorie meal containing 60 g of fat to maximize absorption. The antiviral effect of saquinavir appears to be dose-related, as shown by Schapiro and associates in this issue [10]. These investigators found that saquinavir had a greater and more durable quantitative effect on HIV RNA levels and CD4 cell counts at a dose of 7200 mg/d than at a dose of 3600 mg/d. The FDA-approved dose is 1800 mg/d and is substantially less effective.

The observations of Schapiro and associates are not surprising: A dose-dependent antiviral effect occurs with this and other protease inhibitors, and so the limiting factor of these drugs is toxicity. Care providers will also be concerned about the economics and the pill burden associated with the higher dose. The average wholesale price of saquinavir is about $5800 per year, and the recommended regimen is 9 tablets per day. The dose of 7200 mg/d would require the consumption of 36 tablets per day at an annual average wholesale price cost of about $23 000. Saquinavir is recommended only for use in combination with other drugs; this further magnifies both the cost and the pill burden of this treatment. A saquinavir formulation with better oral bioavailability is now being investigated, but an alternative to higher doses is the concurrent use of drugs that inhibit cytochrome p450. Ketoconazole, for example, increases blood levels of saquinavir by 150%. An attractive possible future option is the concurrent use of saquinavir and ritonavir, another protease inhibitor that is also a potent inhibitor of cytochrome p450. Studies in rodents suggest that the concurrent use of these drugs magnifies levels of orally administered saquinavir in the blood several hundred times [11], although this drug interaction should be studied in patients before recommendations can be made.

Saquinavir is the first on what appears to be a long list of protease inhibitors. At least 10 pharmaceutical companies have these agents in various stages of development. The protease inhibitors that are most advanced in terms of clinical trials are ritonavir (Norvir, Abbot Laboratories, North Chicago, Illinois), indinavir (Crixivan, Merck & Co., Inc., West Point, Pennsylvania), and nelfinavir (Viracept, Agouron, La Jolla, California). The preliminary data from phase I and phase II trials show that these newer agents have good oral bioavailability and substantially less toxicity than most nucleoside analogues [12, 13].

To date, the greatest antiviral activity seen with antiretroviral drugs has been achieved with ritonavir and indinavir [2, 3, 12-16]. These drugs are accompanied by an in vivo decrease in viral burden of 1.5 to 2.0 log at 12 to 16 weeks, as measured by quantitative HIV RNA levels and an increase in CD4 cells averaging as high as 250 cells/mm³. By comparison, nucleoside analogues decrease HIV levels by 0.5 to 0.7 log and increase CD4 cell counts by mean values of 20 cells/mm³ to 50 cells/mm³. A study of combination therapy with indinavir, lamivudine, and zidovudine [14] showed that most patients had a reduction in plasma viremia to undetectable levels—something not previously achieved. A study of ritonavir or placebo added to nucleoside analogue treatment in patients with CD4 counts of less than 100 cells/mm³ showed a 58% decrease in AIDS-defining events or death in the group receiving ritonavir, indicating the anticipated clinical benefits of the drug's antiviral effects [16].

Protease inhibitors are the most promising agents developed in the nearly 10 years that antiretroviral therapy has existed; they may also be the most complicated to use. Resistance to protease inhibitors commonly occurs by week 12 when these drugs are used as monotherapy, and so they must be used in combination with other drugs, presumably nucleoside analogues [17]. Resistance is caused by amino acid changes at codons that are idiosyncratic for different protease inhibitors, which suggests that cross-resistance may not be problematic. Nevertheless, one study [18] showed that exposure to indinavir resulted in variants that showed cross-resistance to six structurally distinct protease inhibitors, suggesting the possibility of protease inhibitor "class resistance." It appears that guidelines for the use of protease inhibitors may be stringent and may include strong admonitions against monotherapy, interrupted therapy, and suboptimal doses.

An additional concern about protease inhibitors that may be related to resistance is the durability of antiviral effects. Peak antiviral response is noted after several weeks, and sustained activity persists 1 to 2 years, but some patients show that the reduction in plasma HIV RNA levels is less pronounced and CD4 count is returning toward baseline [12, 13]. The final concern, as suggested above, is cost. Protease inhibitors are difficult to produce, the initial product has an average wholesale price of $5800 per year, and these drugs will be used in combination with one or two additional agents. We may be entering an era in which standard treatment consists of antiretroviral therapy throughout most of the course of disease at costs of $8000 to $12 000 per year.

The protease inhibitors are an exciting addition to the list of drugs currently used for the treatment of HIV infection. A major concern is that overuse or abuse of these drugs could lead to resistance in the individual patient and ultimately in "community strains." It will be important to adhere to strict usage guidelines, always using protease inhibitors in combination with other drugs and always without interruption. No one is saying that these agents will provide a cure. The optimal result could be the conversion of treated patients to the status of long-term nonprogressors who have low viral burdens and high CD4 cell counts for many years, but from many points of view that result is, indeed, progress.