In the last two years significant advances have been made in the treatment of HIV disease, and these advances have led to the first decline in AIDS-related mortality since the Centers for Disease Control began tracking the epidemic in the early 1980s. Overall, AIDS deaths dropped 23% last year -- from the previous year's all-time high of 50,000 -- and reductions in AIDS-related mortality have been even more dramatic in some parts of the country. In New York City, for example, the death rate fell almost 30% in 1996 (see "C.D.C. reports substantial drop in AIDS deaths," Vol. 3, No. 2).
Although much of this reduced mortality can be attributed to the introduction of the protease inhibitors, which became widely available in early 1996, AIDS deaths actually began to decline in some locations as early as 1993. This suggests that the improved survival we are now seeing in patients with advanced HIV disease is probably the result of a number of factors, among them improved access to care, better education of infected populations, more successful prophylaxis against the opportunistic infections seen in these patients, and more effective clinical management of OIs.
All of these factors have contributed to improved survival, but their significance is minor compared to the contribution made by highly-active antiretroviral therapies, which combine one or more protease inhibitors with nucleoside and non-nucleoside reverse transcriptase inhibitors. HAART regimens are now the standard of care for patients with advanced HIV infection. They do not work with equal efficacy in all patients, and they require that patients remain strictly compliant with highly demanding dosing schedules, but when they do work they frequently achieve almost complete suppression of viral replication.
In the last two years we have seen significant advances made in the treatment of HIV disease. In the next two years we will see further advances, as new drugs -- and new formulations of old drugs -- are added to the clinician's armamentarium (Table 1). These new formulations include Combivir® -- which, as its name suggests, combines two widely prescribed antiretroviral agents, ZDV and 3TC, in a single tablet. These two drugs are routinely prescribed together because the addition of 3TC to a ZDV-containing regimen results in reversal of high-level ZDV-phenotype resistance and leads to more durable suppression of viral load (see "Treatment with ZDV plus 3TC proves superior to monotherapy with either agent," Vol. 2, No. 1, pages 9-10). The particular advantage of Combivir, which is now on pharmacy shelves, is that it allows patients to take fewer pills -- two tablets a day, instead of the four to eight pills per day that patients assigned ZDV and 3TC must take. This simplified dosing is likely to enhance adherence to combination antiretroviral regimens that include these two nucleoside analogs.
In addition, the clinician's ability to treat HIV infection will be enhanced by potent new drug combinations, and the patient's ability to adhere to therapy will be enhanced by new dosing schedules for older drugs. These new dosages -- all of which are currently being tested in clinical trials -- are designed to increase efficacy or, more often, to improve compliance by reducing the number of daily doses of a particular agent that a patient needs to take to achieve adequate serum levels of that drug (Table 2).
Among the potent new regimens now available to clinicians is the triple-drug combination of ZDV, 3TC, and indinavir. This combination began to generate intense interest when the preliminary data were presented almost two years ago (see "Potent, sustained antiviral activity of ZDV and 3TC combined with indinavir," Vol. 2, No. 1). The results of that trial, released earlier this year, reconfirm last year's preliminary findings and underscore the benefits of multidrug regimens in patients with late-stage HIV infection.
In all, 1,156 patients were enrolled in ACTG 320. Participants were stratified into two cohorts: those with CD4 counts above 50 cells/mm3 and those with counts below that number. The mean baseline CD4 count was 86 cells/mm3; 439 participants (38%) had baseline counts below 50 cells/mm3. As investigators had anticipated, progression rates were higher in the cohort with CD4 counts below 50 cells/mm3. What was more noteworthy, therefore, was that progression rates were higher in patients who did not get indinavir, irrespective of CD4 stratum.
These data are similar to those reported in Abbott 247 and other studies in which ritonavir was added to regimens of one or more nucleoside analogs (see "Two studies establish the safety and efficacy of ritonavir," Vol. 2, No. 1, and "Update: The Protease Inhibitors," Vol. 2, No. 3, pages 52-61). The final data from ACTG 320 show the clear superiority, in patients with advanced HIV disease, of antiretroviral regimens that include a protease inhibitor. This conclusion is reaffirmed by the 100-week data from Merck 035, another clinical trial that assessed the benefits of adding indinavir to the combination of ZDV and 3TC.
MERCK 035: This study evaluated 97 patients randomized to receive either indinavir (800 mg q8h), ZDV (200 mg q8h) + 3TC (150 mg b.i.d.), or all three drugs at those doses. To qualify for entry into the trial, patients had to have CD4 counts between 50 and 400 cells/mm3 and plasma HIV RNA levels greater than 20,000 copies/mL. In addition, participants had to have been treated with ZDV for at least six months and be naïve to both 3TC and all protease inhibitors. After 24 weeks of therapy, the triple-drug regimen was shown to be significantly more efficacious in terms of both CD4 cell count increases and suppression of viral replication, and patients in the monotherapy arm (indinavir alone) and the dual-therapy arm (ZDV + 3TC) were given the option of switching to drug-free therapy.
Data from 100 weeks of follow-up were presented at the I.C.A.A.C. meeting held in Toronto this fall. Subjects on the three-drug regimen had a sustained decrease of more than 2 logs in HIV RNA levels, and roughly 80% had viral loads that were less than 500 copies/mL. By contrast, patients in the other arms had decreases of only 1.0 to 1.5 log, and only 40% had viral loads below 500 copies/mL. The addition of indinavir to the combination of ZDV and 3TC did produce dramatic changes in clinical parameters of disease progression, but only 40% of the patients initially treated with the two nucleosides had their HIV RNA levels below 500 copies/mL -- yet another indication that combination therapy should be initiated earlier rather than later.
After 100 weeks of follow-up, CD4 counts averaged 230 cells/mm3 and were continuing to rise in the patients who were originally randomized to the three-drug combination, whereas CD4 counts averaged around 100 cells/mm3 in patients randomized to the other regimens -- even when those patients switched to the triple combination at 24 weeks. These data establish the superiority to combination therapy, and they confirm that this particular three-drug regimen provides durable benefits in most treated patients even after two years of treatment.
GW1592U89 (ABACAVIR): This promising nucleoside analog, now in Phase II/III clinical trials, is a lipophilic, water-soluble guanosine analog that is triphosphorylated in cells by a unique mechanism. GW1592 -- which Glaxo Wellcome has named abacavir -- has demonstrated in vitro synergy with other reverse transcriptase inhibitors, and it is the only drug in its class that can match ZDV's penetration of the central nervous system (see box below). Metabolism is not cytochrome P450-dependent, which reduces the likelihood of adverse drug-drug interactions with phenobarbital, triazolam, dapsone, rifabutin, and azolides.
In general, abacavir has been well tolerated in early clinical trials, and there appears to be no cross-resistance with ZDV or d4T. High-level resistance to 3TC is likely to develop if mutations do occur with chronic use of abacavir, however, because the first RT mutation provoked by this agent occurs at the codon associated with 3TC resistance. The preliminary evidence suggests that this mutation leads to no more than a two- or three-fold decrease in the susceptibility of HIV to abacavir. Even when additional mutations do occur, the drug remains moderately active in vitro, with a maximum 10-fold reduction in susceptibility seen with multiple mutations.
Limited clinical data are now available on 219 patients, including 47 pediatric patients. Individuals with less than 12 weeks' prior exposure to ZDV, CD4 counts between 200 and 500 cells/mm3, and any level of HIV RNA were randomized to receive abacavir for four weeks, followed by the addition of either ZDV or placebo for an additional eight weeks. Viral loads dropped to undetectable levels (<500 copies/mL in this study) in 30% of patients receiving abacavir monotherapy, versus 60% of those on combination therapy. CD4 counts rose by 90 to 145 cells/mm3 on either therapy.
The optimal dose of abacavir appears to be 300 mg b.i.d. The drug is well tolerated, with headache and GI disturbances being the most common adverse events. An allergic reaction to the drug was reported in eight patients (3%). The reaction is characterized by fever, malaise, nausea, vomiting, and an erythematous, generalized, and macular rash. This reaction usually occurred in the first week of therapy. Rechallenge resulted in an exaggerated reaction, so rechallenge with abacavir is contraindicated.
Reverse transcriptase mutations could be detected in 60% of the viral isolates taken from patients treated for 12 weeks with abacavir alone, compared with only 13% of isolates from patients who got both abacavir and ZDV. Abacavir is up to three times more active than ZDV, and when used as monotherapy it produces reductions in HIV RNA of two logs. Expanded trials of abacavir -- in combination with other nucleosides and with protease inhibitors -- should soon give us additional information on this new and potentially valuable RT inhibitor.
ADEFOVIR DIPIVOXIL: Adefovir is a novel nucleoside analog that is related to a class of compounds called nucleotides, the best-known of which is cidofovir. Nucleotides differ from other RT inhibitors in that they do not require intracellular monophosphorylation; the diphosphorylated metabolite is the active form of these compounds.
This new drug has broad-spectrum in vitro antiviral activity against HIV-1 and -2, hepadnaviruses, and a wide range of herpesviruses including HSV, HHV-6, and CMV. As importantly, in vitro resistance to this drug has proven difficult to demonstrate, and HIV strains that are resistant to nucleoside RT inhibitors are sensitive to adefovir. Because the parent formulation of adefovir has poor bioavailability, Gilead Sciences has developed a prodrug, adefovir dipivoxil, which is 40% bioavailable and is completely converted to adefovir following absorption. Adefovir diphosphate has an intracellular half-life of roughly 18 hours, which means that patients are required to take it only once a day -- an important consideration given the complexity of the dosing schedules of most multidrug antiretroviral regimens.
Based on these favorable results, a double-blind, multicenter Phase I/II trial was conducted in 72 adult patients with advanced HIV disease. These subjects, all of whom had CD4 counts below 200 cells/mm3 and HIV RNA levels above 10,000 copies/mL, were randomized (in a ratio of 2:1) to receive drug or placebo for six weeks, followed by six weeks of open-label treatment. Two dosage levels -- 125 mg/day and 250 mg/day -- were evaluated, and assessments were made of the drug's immunologic and virologic activity as well as its safety and tolerability.
At the lower of the two doses, adefovir was extremely well tolerated, with no patients withdrawing from the study due to therapy-related toxicities. At the 250-mg dose, however, 4 of 36 patients withdrew because of nausea and vomiting, dysuria and genital ulcers, or fever and chills. The most common dose-related adverse event was nausea (9 of 48 patients), with diarrhea occurring in 4 of 24 patients taking the 250-mg dose.
After the first six weeks of the study, CD4 counts had risen by a mean of 46 cells/mm3 in the 125-mg cohort and 15 cells/mm3 in the 250-mg cohort compared to the placebo arm. These changes persisted at 12 weeks, with mean absolute increases of 57 and 27 cells/mm3, respectively, in the low- and high-dose cohorts. As expected, HIV RNA levels did not change in the placebo group. In the treatment arms, by contrast, viral levels dropped by a median of 0.5 log in the low-dose group and 0.4 log in those getting the higher dose. These benefits endured throughout the 12-week study.
The durability of these changes must now be investigated in larger, longer clinical trials. But the evidence at hand suggests that further studies of the 125-mg/day dose of adefovir -- in combination with other antiretroviral agents -- are warranted.
PMPA: This investigational nucleotide analog has a mechanism of action and intracellular metabolism similar to those of adefovir, and the drug's in vitro activity against the common strains of HIV is also comparable to adefovir's. However, PMPA has significantly less in vitro cellular cytotoxicity than adefovir, which at least theoretically might translate into less in vivo toxicity.
When PMPA was given for 28 days to macaques infected with simian immunodeficiency virus, the monkeys' SIV RNA levels fell by as much as 3.0 logs, as did viremia in peripheral-blood mononuclear cells. No significant toxicities were seen when PMPA was given at a once-daily dose of 30 mg/kg; at a dose of 75 mg/kg/day, reversible anemia and hypophosphatemia were noted.
PMPA has also been used as post-exposure prophylaxis against SIV in macaques, and here it has yielded results that are truly dramatic: in the handful of animal studies done to date, the drug has been 100% effective in preventing infection, even when given as much as 24 hours after exposure. This phenomenon has never before been demonstrated, in this or any other animal model, for any other antiretroviral agent. These preliminary but intriguing findings suggest that PMPA has unusually high antiretroviral activity in simian models and merits study in humans. Phase I trials in HIV-infected adults are now in the early planning stages.
DMP-266: This new non-nucleoside reverse-transcriptase inhibitor has potent activity against HIV. DMP-266 -- which is being developed jointly by Merck and DuPont and will be marketed under the trade name Sustrin® -- has particularly good activity against strains of HIV that are resistant to nucleoside analogs and/or to indinavir. Laboratory-derived strains of HIV that are resistant to nevira- pine and delavirdine, the two currently approved NNRTIs, are also susceptible to DMP-266. Both in vitro and in vivo resistance to DMP-266 have been described, however, with the K130N mutation being the most important. This mutation occurs rapidly with monotherapy -- which means that DMP-266, like the other NNRTIs, should always be used in combination with nucleoside analogs and/or protease inhibitors.
DMP-266 has a very favorable pharmacokinetic profile. Its half-life of 52 to 76 hours allows for daily dosing at 200 mg. Cerebrospinal fluid penetration is roughly 1% of free drug found in the serum. Although this percentage is exceedingly low, the levels attained in CSF are above the IC95 for most clinical isolates of HIV. DMP-266 is slowly metabolized, and coadministration with indinavir reduces serum levels of the latter drug. Blood levels of DMP-266 are not affected by coadministration of indinavir, ZDV, or 3TC, however. Moreover, food does not appear to significantly affect absorption.
The major side effects of DMP-266 therapy are skin rash and CNS symptoms. Rash has been reported in 5% to 40% of treated patients and is similar in character to the rash seen with other NNRTIs. It should be noted that erythema multiforme and Stevens-Johnson reactions have not been reported as yet in patients taking DMP-266. The most common CNS symptoms are light-headedness and a feeling of being "drugged" or disengaged. Some patients have reported that the drug has a stimulant-like effect and makes them feel "high." These symptoms are more common at doses above 200 mg/day and tend to wane over a few days to weeks.
Limited clinical-efficacy data are available on DMP-266. The results of a small study -- which used the new Merck-DuPont drug at 200 mg/day in combination with indinavir at one of two doses (800 mg or 1000 mg q8h) were reported at this year's I.C.A.A.C. meeting. After 24 weeks of therapy, the mean decrease in plasma HIV RNA was 4.75 logs, with 94% of patients having fewer than 500 copies/mL. CD4 counts increased by a mean of 199 cells/mm3. This drug combination was well tolerated, with 82% of patients completing the 24-week study. These very impressive results suggest that this NNRTI is significantly more potent than either nevirapine or delavirdine.
NELFINAVIR MESYLATE: Nelfinavir is the most recent addition to the growing lineup of approved protease inhibitors and it is the first of these powerful antiretroviral agents to be approved for use in children (see "At last, a protease inhibitor for children as well as adults," Vol. 3, No. 1, pages 18-19). The drug has a molecular weight of 586, which makes it the smallest of the approved protease inhibitors, and it is approximately 80% bioavailable, with absorption not inhibited by food.
Although nelfinavir, unlike abacavir, is metabolized through the cytochrome P450 system, selective inhibitors of this system (such as the azole antifungals and macrolides) and moderate inhibitors of the system (such as rifabutin) are unlikely to alter serum concentrations of nelfinavir to a significant degree. On the other hand, nelfinavir does reduce concentrations of ethinylestradiol by 50% -- which means that increased doses of estrogen will be needed for effective contraception. This is an important clinical consideration, given the number of HIV-positive women who become pregnant despite reported use of barrier contraceptives. Adequate ethinylestradiol levels will prevent contraception, but of course they will not prevent sexual transmission of HIV -- a fact that should be emphasized in discussions with HIV-positive patients and their sexual partners.
Administration of nelfinavir with either saquinavir or ritonavir is likely to result in increased serum concentrations of the older drugs, since these protease inhibitors are also metabolized by P450. Indeed, the early in vitro data indicate that coadministration of ritonavir and nelfinavir leads to dramatic increases in plasma levels of the latter drug. (This is not an altogether surprising finding, given that coadministration of these two older protease inhibitors boosts saquinavir concentrations 40-fold.) Although no data are yet available on the pharmacokinetics of combining nelfinavir and the non-nucleoside RT inhibitor nevirapine, the likelihood is that serum levels of nelfinavir will drop roughly 30%, as they do when this NNRTI is given in combination with other protease inhibitors.
Resistance is a matter of critical concern in the clinical management of HIV infection, and the available data suggest that nelfinavir works better in combination with other antiretrovirals than alone, and it is more effective as initial than as salvage therapy. For example, it has been shown that taking nelfinavir with ZDV and 3TC reduces the risk of nelfinavir resistance nine-fold when this triple-drug combination is compared with nelfinavir monotherapy. HIV strains that have become resistant to indinavir and ritonavir will be cross-resistant to nelfinavir, but nelfinavir-resistant viral strains may remain susceptible to the other protease inhibitors -- making nelfinavir an appealing candidate for first-line therapy.
The major side effect of nelfinavir therapy is mild-to-moderate diarrhea, which can be readily controlled with over-the-counter antimotility drugs. No significant laboratory abnormalities were attributable to nelfinavir, and this finding, in combination with the other safety and efficacy data from these studies, indicates that nelfinavir is a very useful addition to the current roster of antiretroviral agents.
GW141W94: GW141 is a non-peptide protease inhibitor with a molecular weight of 505.6, making it the smallest of the protease inhibitors now being studied. The molecule has a backbone similar to the sulfonamides -- which may be of particular importance, given what we know about the increased reactogenicity of these drugs in patients with HIV disease. There is no reported cross-resistance with other protease inhibitors, and studies have shown that GW141 has good in vitro synergy with two other antiretroviral agents developed by Glaxo Wellcome, ZDV and abacavir. CNS penetration is limited, but no more than it is with all protease inhibitors.
Preliminary Phase I pharmacokinetic data on GW141 indicate that efficacy is dose-related. When administered to nine patients for four weeks at a dose of 1200 mg b.i.d., GW141 produced a median maximum decrease in viral load of 1.69 log, but only one of the nine subjects had his HIV RNA level drop below 400 particles/mL.
This investigational protease inhibitor has also been administered in combination with abacavir. In a small pilot study, seven patients were given 900 mg of GW141 and 600 mg of abacavir twice daily for four weeks. This two-drug regimen resulted in a median maximal decrease in viral load of 2.08 logs, and five of the subjects had their HIV RNA levels drop below the lower limit of detection in this study. The median maximum increase in CD4 count in this pilot study was 114 cells/mm3 for patients treated with GW141 alone at a dose of 2400 mg/day, compared with 79 cells/mm3 for participants in the GW141/abacavir combination arm. The most common adverse events reported by patients taking GW141, alone or in combination with abacavir, were diarrhea, nausea, headache, rash, and paresthesias. There were no significant hematologic or biochemical abnormalities noted in any of the study subjects, and there was no clear correlation between the dose given and the incidence of side effects.
It is now well established that protease inhibitors and nucleoside RT inhibitors work better in combination than alone, so the early data reported above come as no surprise. Indeed, clinicians should anticipate using both GW141 and abacavir in combination -- either with one another or with other antiretroviral agents (Table 3).
ABBOTT 378: ABT-378 is a new and novel protease inhibitor currently in Phase I development. Tested to date only in vitro and in rats, this compound is roughly ten times more active in human serum than ritonavir. ABT-378 maintains potent in vitro activity against viral strains resistant to ritonavir. Indeed, this investigational drug is active in the face of multiple mutations, including the mutation at codon 82, which denotes the initial stage of resistance to both ritonavir and indinavir. As a result, patients who have been extensively treated with these older protease inhibitors should remain partially sensitive to ABT-378, and patients treated with saquinavir should remain wholly sensitive to Abbott's new drug. Moreover, this new protease inhibitor is more active than ritonavir in the presence of the cluster of mutations that confer high-level resistance to Abbott's older drug.
Investigators predict that ABT-378 will achieve and maintain high plasma concentrations, and it is anticipated that this new protease inhibitor will require dosing only once or twice a day. These investigators note that there appears to be a high degree of pharmacologic effect between Abbott's protease inhibitors. The positive impact of ritonavir on ABT-378 plasma levels is greater than that of ritonavir on any other protease inhibitor studied to date: the coadministration of ABT-378 and ritonavir in rats produced plasma levels of the former drug that were 50-fold higher than the level of drug needed to inhibit the drug in vitro.
These findings suggest that doses of ritonavir as low as 50 mg/day will enhance the plasma concentrations of ABT-378 -- which would make this combination a valuable addition to the clinician's antiretroviral armamentarium. The results of the Phase I trials now under way should give us a better sense of ABT-378's utility before the end of the year.
SAQUINAVIR SGC: The oldest of the protease inhibitors is now available in a new formulation, a soft-gel capsule that has nine times the bioavailability of the old formulation when given at the doses used in clinical trials. Low bioavailability has been the chief limitation of saquinavir, which has an acceptable side-effects profile even at doses substantially higher than the standard therapeutic dose.
A study presented at this year's I.C.A.A.C. meeting showed that this new formulation of Roche Laboratories' protease inhibitor, when used in combination with ZDV and 3TC, reduced viral load below the limit of detection of the assay used in this study (400 copies/mL) in 81% of patients and increased CD4 counts by 170 cells/mm3 after 16 weeks of therapy.
This open-label trial enrolled 41 treatment-naïve patients with viral loads greater than 10,000 copies/mL (mean: 63,408) and CD4 counts greater than 100 cells/mm3 (mean: 412). Saquinavir SGC was given at 1200 mg t.i.d., and ZDV and 3TC were given at the currently recommended doses. Saquinavir SGC was generally well tolerated, with only two patients withdrawing from the study due to adverse effects. The most common complaints were nausea, diarrhea, vomiting, and headache.
Data from the first-ever head-to-head trial of two protease inhibitors were also presented at the I.C.A.A.C. meeting. This study, which compared saquinavir SGC plus ZDV and 3TC to indinavir plus those two nucleoside analogs, found that the two triple-drug combinations had similar antiretroviral activity, and the regimen that contained saquinavir SGC produced greater increases in CD4 cell counts.
This open-label study, which was conducted by a team of Dutch researchers, enrolled 44 subjects with little or no previous exposure to ZDV (<12 months) and viral loads greater than 10,000 copies/mL and/or CD4 counts less than 500 cells/mm3. Patients were randomized to receive ZDV (200 mg t.i.d.) and 3TC (150 mg b.i.d.) plus either saquinavir SGC (1200 mg t.i.d.) or indinavir (800 mg q8h). Patients in both treatment arms had baseline viral loads of nearly 80,000 copies/mL; baseline CD4 counts were 310 and 296 cells/mm3, respectively. Both regimens were generally well tolerated, and the adverse events reported by the Dutch team were mostly mild in intensity.
These two antiretroviral combinations were equally efficacious in reducing viral burden, with reductions of 2.3 logs seen in both cohorts. Viral load fell below the limit of detection in all patients for whom 12-week data were available. These data showed that the saquinavir SGC-containing regimen produced a greater rebound in CD4 counts than the indinavir-containing regimen at the end of this small three-month trial: 124 cells/mm3 versus 49 cells/mm3.
Ongoing clinical trials are testing this new formulation of saquinavir in combination with nucleoside analogs, NNRTIs, and other protease inhibitors (Table 3). A number of studies have already established that the addition of ritonavir to a saquinavir-containing regimen increases serum levels of the latter drug -- a synergistic effect that translates into increased efficacy (see "Enhancing saquinavir levels with ritonavir," Vol. 3, No. 3). It is reasonable to assume that regimens that combine ritonavir and the soft-gel capsule formulation of saquinavir will provide even higher serum concentrations of active drug, leading to even greater antiretroviral activity.
Glaxo Wellcome, which developed the nucleoside analog known as GW1592 or abacavir, has made this investigational drug available to patients suffering from severe, clinically diagnosed AIDS dementia complex. The company based its decision on pharmacokinetic data which indicate that abacavir is the only drug in its class that can match ZDV's exceptional penetration of the central nervous system (see "The neuroprotective benefit of ZDV: Drug reduces incidence of AIDS dementia complex," Vol. 1, No. 6, page 110). To qualify for enrollment in this new program, patients must be at least 14 years of age and must have evidence of HIV-associated dementia as defined by the American Academy of Neurology (and/or a Memorial Sloan Kettering score of 2 to 4). In addition, applicants must have experienced progressive dementia despite treatment with ZDV or must have documented intolerance to ZDV. For further information on this open-label program, clinicians should call 1-800-501-4672.
Harold A. Kessler, M.D., is Professor of Medicine and Immunology/Microbiology, Rush Medical College, Chicago, IL.