At the just completed 6th Conference on Retroviruses and Opportunistic Infections, the significant amount of drug-resistant HIV in treatment-naïve patients became a major point of concern. Two U.S. studies (late-breaker abstracts LB9 and LB10) found at least mild resistance to protease inhibitors and nonnucleoside reverse transcriptase inhibitors, appearing at a 10% to 15% rate. Resistance to nucleoside analogs had a 3% to 10% incidence, depending on the assay used. Not all of this reduced susceptibility to therapy is due to transmission of drug-resistant HIV selected by exposure to antiviral drugs in the source of the infection. The virus's natural genetic variability in the absence of drugs also is at work here.
The threat of drug failure has created a crying need for new agents not affected by resistance to the older ones. The Retrovirus Conference provided a further, and many times a first, look at a second generation of drugs that have potential roles for salvage or rescue therapy in people unable to completely suppress their HIV with the established medications. The most advanced of these is amprenavir, which should be on the market within months. Two more, ABT-378 and T-20, are entering phase II human trials and should be available within two years. These three, and probably their less advanced cousins, too, do not represent major breakthroughs. They all behave like other drugs in terms of potency and vulnerability to resistance-conferring HIV mutations. Even so, they offer advantages that will increase the currently mediocre success rates of second-line and rescue therapies.
Laboratory studies that attempted to generate amprenavir-resistant HIV by exposing the virus to increasing concentrations of this Glaxo Wellcome-Vertex protease inhibitor noted a unique mutational response on HIV's part. Resistance to amprenavir was based on a mutation at position (codon) 50 of the protease gene, a location left unaltered during the emergence of resistance to other protease inhibitors. In people, the picture is somewhat more complicated and less hopeful.
At the 6th Retrovirus Conference (abstract 118), Glaxo researchers presented a resistance analysis of people failing to respond to amprenavir (viral load remaining or returning to greater than 400 copies/ml). In the first place, 80% of 145 people not responding to amprenavir did not have detectable amprenavir resistance, only to other drugs in the combination. This still unexplained phenomenon has been observed with other protease inhibitors, most notably indinavir.
If their HIV did show resistance to amprenavir, volunteers who were protease inhibitor-naïve before the trials most frequently had HIV mutated at protease codon 50 or, one-third of the time, at codon 54. There were also supplemental mutations at codons 46 and 10. Mutations at codons 10, 46 and 54 commonly contribute to broad cross-resistance to protease inhibitors in multidrug-resistant HIV.
The picture was further complicated when the Glaxo investigators looked at trial volunteers who had prior experience with other protease inhibitors. When such people switch to amprenavir, preexisting mutations at codons 84 and 10 were the ones most frequently associated with baseline amprenavir resistance. The codon 84 mutation, like the one at codon 10, frequently arises during treatment with other protease inhibitors and also fosters cross-resistance to protease inhibitors as a class.
In CNA2007, which tested the combination of abacavir, amprenavir and efavirenz in a heavily pretreated cohort (see page 4-5 and 9-10 of this issue), resistance to amprenavir at the beginning of the trial was markedly less than for the other protease inhibitors. HIV resistant to the approved protease inhibitors was present in 72% (for saquinavir) to 84% (for ritonavir) of the trial participants. Only 45% of CNA2007 participants harbored HIV with significant amprenavir resistance. (See 6th Retrovirus Conference, abstract 133.)
Paris researchers obtained analogous results from phenotypic resistance testing in 108 patients at the Hôpital Bichat-Claude Bernard who had failed to suppress HIV replication after protease inhibitor-containing therapy lasting 2 to 30 months (median of 12 months). The patients had been treated with indinavir, nelfinavir, saquinavir and/or ritonavir, but not amprenavir. About 40% of protease inhibitor-resistant HIV isolated from patients' blood samples was also resistant to amprenavir, whereas cross-resistance to other protease inhibitors ranged from 60% to 90% (6th Retrovirus Conference, abstract 133).
Glaxo Wellcome submitted amprenavir's New Drug Application to the FDA in October. Amprenavir's application has not raised any nettlesome issues at the FDA, and approval will not require an advisory committee hearing. The drug should reach the market in early spring. In the meantime, the drug is available through an expanded access program that distributes it to people unable to construct effective treatment combinations with the antiviral drugs on the market. A subprogram is open to people who are not failing their current therapy in virologic terms but are experiencing adverse effects on fat metabolism. Fat metabolism defects include high blood lipids and redistribution of body fat deposits, the latter referred to as "protease paunch," "buffalo hump" or, more generally, "lipodystrophy." Although Glaxo concedes that the meager available data are not complete and surely not conclusive, its researchers note that there have been just four reports of fat redistribution among the 1,477 participants in amprenavir trials. (See 6th Retrovirus Conference, abstract 386.)
For more information on expanded access, doctors may call 800/248-9757. Demand has been high: The program enrolled 1,200 persons from September through January. A January 11 letter to doctors warned of a two- to three-week delay in shipping amprenavir to new enrollees. The company says that this delay has been shortened to several days, but some doctors report that their patients are still waiting three weeks.
In the test tube, ABT-378 is one of the most powerful protease inhibitors discovered so far. ABT-378 in the body achieves very high and stable blood concentrations well above the minimum necessary therapeutic levels. It is being tested as a one-pill, twice-a-day drug.
Abbott Laboratories, ABT-378's sponsor, has claimed that there is little cross-resistance between its new compound and other protease inhibitors. Indeed, the molecular structure of ABT-378 was selected specifically because it is not very sensitive to the key mutation at protease codon 82. This is the most common mutational linchpin for resistance to protease inhibitors, especially to indinavir and ritonavir. ABT-378 retained activity against three ritonavir-resistant HIV isolates from human beings when tested in the lab (see H.L. Sham et al. Antimicrobial Agents and Chemotherapy, Dec. 1998, pages 3218-24). This is not the whole story, though, for ABT-378's activity is hobbled by many of the same protease mutations that affect amprenavir in persons with prior protease inhibitor treatment (namely the mutations at codons 10, 46 and 84 -- see A. Carrillo et al., Journal of Virology, Sept. 1998, pages 7532-41).
Abbott is carrying out one pilot study in 101 treatment-naïve volunteers taking ABT-378 plus d4T and 3TC. By week 20 to 24, about 95% of those still on treatment had achieved viral loads below 400 copies/ml. Counting the dropouts as failures reduces this figure to about 85%. Diarrhea and loose stools, usually transient, were the sole notable side effect, appearing in about one-third of trial volunteers.
Trials for people with refractory HIV have been slow to get off the ground. A trial that recruited 70 volunteers with experience on a single other protease inhibitor is currently ongoing, nearly two years after ABT-378's resistance profile was first revealed. Trial participants switch their protease inhibitor to ABT-378 at week 0 and then switch their other drugs at week 2 to nevirapine plus the most active two nucleoside analogs that can be selected considering their personal history.
Sources at Abbott privately concede that this trial so far has had more successes than failures, although it is much too early to tally up the results in a definitive manner. At a community meeting held in conjunction with the 6th Retrovirus Conference, Eugene Sun, Abbott's head of antiviral research, took the highly unusual tack of emphasizing two examples of poor outcome -- participants in the single protease failure trial whose HIV rebounded at day 30 or day 50. Dr. Sun's presentation may have reflected the simmering controversy over accelerating research on ABT-378 as salvage therapy and instituting an expanded access program as soon as appropriate. In January, a national consensus statement signed by over 100 individuals and organizations complained that Abbott's delays in this area "have been unacceptable when people with HIV continue to be at serious risk of illness and death."
"[T]he community hereby demands that a broad expanded access program be commenced immediately, upon first evidence of antiviral activity [in a population unresponsive to established agents]," the statement continued.
Dr. Sun contends that salvage therapy research and expanded access is premature. He says, "We can't release the drug on top of other useless drugs, or it is likely to fail. The research has to proceed step by step." The company so far has a series of Phase 3 concept sheets that include a trial in single and multiple protease inhibitor-experienced patients, but a firm timetable is lacking as yet. The company does not envision an expanded access program until early next year, and that assumes the trials turn out well. An FDA New Drug Application for ABT-378 cannot be expected for another 18 months at the earliest.
One clear partner for combining with ABT-378 is Trimeris Corporation's T-20. T-20 is the most advanced of a whole new drug family. It combines with part of the viral envelope to stop HIV particles from fusing with new cells and infecting them (see Treatment Issues, April 1998). Because its mechanism is unique, T-20's activity is undisturbed by alterations in HIV that make it immune to other drugs. Trimeris has been quick to test T-20 as a salvage therapy for this reason. While Abbott faces growing complaints over its slow pace with ABT-378, T-20 provides an exemplary case of rapid salvage therapy testing.
The results of Trimeris' salvage protocol, TRI-003, were reported on February 4 at a late breaker session concluding the 6th Retrovirus Conference (abstract LB13). The eligibility criteria for TRI-003 were very loose: merely a viral load over 5,000 on current, stable therapy. Nonetheless, the actual mean baseline viral load among the trial's 78 participants was 100,000, and the group had been exposed to an average of nine previous drugs, including three protease inhibitors. Trial participants had to remain on their current therapy while adding T-20 through the study's initial 28-day phase.
At week 2, viral load drops in this heavily experienced population averaged 0.3 logs to 1.6 logs (50% to 97.5%), depending on dose. The dose response confirms that T-20 was affecting the viral load, rather than some confounding variable, such as patients surreptitiously switching their background therapies. The negative part is that by week four, viral loads in all the trial arms had returned to within 60% of baseline.
As virtual monotherapy, T-20 starts off strong but begins to lose its effect after just two weeks. Its behavior is similar to such drugs as efavirenz or 3TC, potent agents whose activity is curtailed by a single mutation on HIV's part. By contrast, it takes up to six months of protease inhibitor monotherapy before significant HIV rebound occurs because these drugs' resistance profiles are much more complex.
A genetic analysis performed on participants in a preliminary T-20 trial (6th Retrovirus Conference, abstract 611) found mutations in a region previously identified in cell culture as key to T-20 resistance (see L.T. Rimksy et al., Journal of Virology, February 1998, pages 986-93). In response to these indications of T-20 resistance, Trimeris is developing a second-generation compound that would bind to a nearby area of the HIV envelope. This compound would be useful for rescuing people whose HIV had become insensitive to T-20 (6th Retrovirus Conference, abstracts 616 and 617). And so the same old race to counter HIV's mutations begins anew.
One should not dismiss T-20's accomplishments nonetheless. The recent trial tested the compound in the population that is hardest of all to treat. T-20 effected reductions in HIV that none of the established antivirals could have accomplished in this heavily pretreated, high viral load cohort. Another positive outcome of this trial was that the simpler mode of administering T-20, twice-daily subcutaneous injections, appeared as effective as a portable infusion pump that delivers a constant T-20 dribble under the skin. (T-20 is a peptide, a short protein fragment, and cannot be taken orally because it would be broken down in the digestive tract.) So far, the major side effect has been rash and bumps at the injection site.
T-20, like other anti-HIV drugs, obviously has to be used as part of an effective combination to prevent the rapid emergence of drug resistance, and Trimeris researchers say they are interested in making arrangements to combine T-20 with other companies' experimental agents. For now, though, the next trial will test T-20 for 16 weeks in combination with up to six standard drugs for persons with single or multiple protease inhibitor failure and naïve to nonnucleoside reverse transcriptase inhibitors (delavirdine, efavirenz, nevirapine). The principal stumbling block at present is synthesizing this 36-amino acid molecule at reasonable cost. Trimeris claims to be slowly expanding production and attaining economies of scale, but the compound will always cost substantially more than any of the anti-HIV agents now in pharmacies.