Early last month, TAG members Spencer Cox, Mark Harrington, Tim Horn and Michael Marc headed home after attending the 35th Interscience Conference on Antimicrobial Agents and Chemotherapy (ICAAC) which convened in San Francisco, CA, September 17-20. Upon their return they assembled a 62-page summary of what they found to be the more interesting new reports ("TAG Does ICAAC: AIDS Research Highlights from the 35th Interscience Conference on Antimicrobial Agents and Chemotherapy"). A sampling of their observations appears below, and they warn that "as always, everything needs to be taken with a grain of salt."

The 35th ICAAC was a more interesting conference than usual, with results from many important AIDS clinical trials being reported, perhaps the biggest antiretroviral event being the results from ACTG 175, which were reported on at length in last month's TAGline. The ICAAC meeting benefitted partly from the lack of an international AIDS meeting in 1995, and partly perhaps from its location in San Francisco--a favorite junket site for physicians and activists alike.

Survival Trends, 1990-1994

J.W. Ward and colleagues from the CDC correlated survival with treatment patterns by reviewing medical records of 8,622 persons diagnosed with AIDS from 1990 to 1994. The median survival was 35 months. After controlling for CD4 count, age, risk and antiretroviral treatment; PCP prophylaxis with either Bactrim or dapsone--but not with aerosolized pentamidine--and MAI prophylaxis with either rifabutin or clarithromycin were associated with improved survival. Persons who were prescribed antiretroviral treatment--and prophylaxis against both PCP and MAI--had the longest survival: 47 months.

Later, R.M. Selik and colleagues from the CDC presented data on mortality trends in AIDS from 1987 to 1992. It was clear that the introduction of PCP prophylaxis has had a dramatic effect on overall causes of AIDS mortality, as PCP dropped from 32.5% in 1987 to 13.8% as the cause of AIDS deaths in 1993. Similarly, the introduction of fluconazole appears to have reduced mortality from candida and cryptococcus. Most alarmingly, perhaps, non-pneumocystis pneumonias are now the second most common cause of AIDS-related deaths, while bacterial septicemia comes in third.

AZT+d4T

Because in vitro experiments had suggested that AZT and d4T compete for the same binding site, many people thought that these drugs could not be used in combination with each other, but a French study of 27 HIV-infected patients (CD4 <100cells/mm3) reported a reduced rate of new opportunistic infections and deaths for the AZT+d4T combination group compared to the d4T monotherapy group.

Saquinavir (Invirase )

Ann Collier of the University of Washington presented long-term follow-up data from ACTG 229, the 302-patient study which compared AZT+ddC to AZT+saquinavir to AZT+ddC+saquinavir. She showed that while throughout the first six months there were greater CD4 and viral load changes in the the triple-therapy arm compared to the two double-drug arms, clinically the three groups fared the same. [Yet another nail in the surrogate marker coffin.]

Jonathon Schapiro of Stanford presented an open-label study of two higher doses of saquinavir. Twenty patients each were randomized to receive 3,600 or 7,200 mg saquinavir daily for 24 weeks. Mild liver function test elevations, confusion and gastrointestinal upset were reported at the 7,200 mg dose. CD4 counts rose by an average of 72 cells/mm3 at 4 weeks and by 121 cells at 20 weeks. Thirteen of the forty patients developed either the 48 or 90 mutation over the course of the study; suggesting, but not proving, that the higher doses delay the emergence of resistance. (In a European study of regular dose saquinavir, 50% show low level resistance at week 25, and the rest develop it by week 52.) Curiously though, while the resistance-conferring mutations did not emerge until week 20, plasma HIV-RNA levels bottomed out and began to rebound after only 4 weeks of high-dose saquinavir therapy.

Indinavir sulfate/MK-639 (Crixivan )

In an oral presentation, researchers from Merck presented data on 16 patients who crossed over after a 24-week comparison of the Merck protease inhibitor (MK-639) vs. AZT. These 16 patients had constituted the AZT arm, and were given MK-639 (600 mg every four hours) following cessation of the randomized study. All patients were p24 antigen positive at baseline, with median CD4+ T cell counts of 110 and a viral load over 50,000 (later changed to over 20,000); the actual average viral load was around 100,000. At weeks 12, 24 and 36, median CD4+ T cell counts increased from baseline by 104, 105 and 126 cells respectively. The median CD4+ T cell count was sustained above baseline at 52 weeks. At 8 weeks, there was a median 1.98 log reduction in serum HIV RNA levels. One third of the patients had a greater than 2 log RNA reduction and about one third sustained this reduction through week 24. Median viral load, however, returned to baseline at 24 weeks.

Four patients developed urolithiasis (stones in the urinary tract) and were rechallenged. At a dose of 400 mg every four hours, transient increases in bilirubin occurred in about 15% of patients, and one patient developed crystalized drug in the kidney.

There are nine amino acid substitutions associated with phenotypic resistance to MK-639, of which at least three are required for in vitro resistance (M46->I, L63->P and V28->T). In some treated patients, as many as 23 residues have been shown to have shifted. By week 24, most patients have at least one of the nine amino acid substitutions associated with resistance. Appearance of the resistance mutations correlates with increasing viral burden. However, "simply counting the viral copies cannot predict the extent of the immune damage: maybe we're selecting for less virulent virus" [or virus that is more virulent, see Dave Gilden's excellent review, "Roaches of Inner Space," in the October issue of GMHC's Treatment Issues ). This would explain the apparent dissociation between the return of RNA copy numbers to baseline at 24 weeks with the relatively sustained CD4 rises at 52 weeks.

Data were also presented on MK-639 in combination with IL-2 in patients with <300 CD4+ cells/mm3. The study had three arms: A) IL-2 12MU for 5 days every 2 months + MK-639 at 600 mg qid; b) IL-2 12MU 5 days every 2 months + MK-639 600 mg qid for 10 days every 2 months and c) MK-639 600 mg qid alone.

Ritonavir (a.k.a. ABT-538)

Monotherapy data indicates that plasma viral levels rebound at 8 weeks with doses from 300-400 mg. With higher doses, such as 600-800 mg, the reduction in plasma HIV-RNA is sustained at 0.8 log for up to 28 weeks. The emergence of resistance to ritonavir is a multi-step process. Seven mutations have been correlated with the development of reduced in vitro sensitivity to ritonavir (at codons 82, 20, 36, 46, 54, 71 and 84). The codon 82 mutation seems to be the primary resistance mutation, while the other codon changes merely enhance resistance.

Ritonavir+AZT+ddC

D. Norbeck of Abbott presented data from a French study where patients were given triple therapy using AZT+ddC+ritonavir. At week 20, CD4 cell counts had risen an average of 110 cells and plasma HIV-RNA levels had fallen an average of 2.5 logs. He then went on to make the provocative (and unsubstantiated) claim that some study participants had become "culture negative" over the subsequent weeks, suggesting that the "viral reservoirs had been emptied."

Agouron protease inhibitor (a.k.a. AG-1343 or Viracept )

Results from an English phase II pilot study (20 antiretroviral naïve participants) of Agouron's protease inhibitor AG-1343 showed a drop in viral load of around 1.2 logs at day 20 at the 1,030 mg/d dose. Nausea and diarrhea occurred in 65%; fatigue in 30%. Marty Markowitz, of the Aaron Diamond AIDS Research Center, had less spectacular results with his 30 antiretroviral-naïve patients where twice daily doses of 500 mg, 600 mg or 750 mg of AG-1343 resulted in average maximum reductions in plasma HIV RNA levels of 90%, 94% and 93% (that is, considerably less than 1 log), respectively. Forty percent of the participants experienced diarrhea. The Agouron protease inhibitor is now feared to be cross-resistant with the Roche inhibitor, saquinavir (Invirase ).

Upjohn protease inhibitor (U-103017)

Results from a phase I pharmacokinetic study of the lead Upjohn protease inhibitor concluded that all three formulations were well-tolerated and that two of the three had good bio-availability. Half-life was reported to be between 8-15 hours.

Vertex (Glaxo-Wellcome) protease inhibitor (VX-478)

In vitro studies suggest that an I50->V mutation is necessary but not sufficient to induce resistance to VX-478. (Other amino acid substitutions seen in vitro are L10->F, M46->I, I47->V and I84->V.) The 84 and 46 mutations are also seen with the Merck and Abbott protease inhibitors.

tat inhibitor

Allelix Biopharmaceuticals presented results from a phase I study of its new tat inhibitor ALX40-4C. In a dose-ranging study of 28 HIV-infected volunteers, the top two doses (0.56 mg/kg and 0.70 mg/kg) achieved blood levels above the IC50. No dose-limiting toxicities were seen.

GEM91 antisense

Long-awaited results from the European study of Hybridon's lead antisense therapeutic were presented by David Sereni. The drug was given by infusion (two hours every other day for 27 days) to 30 patients at three different doses. At the highest dose (2.0 mg/kg/day), unacceptable nausea was reported, but so far the lower doses have been well-tolerated. No changes in HIV RNA levels, however, were observed.

Other antiretrovirals

Pre-clinical data was presented on new protease inhibitors by Ciba-Geigy (CGP-57813 and CGP-53437) and the National Cancer Institute (KNI-272); several metabolites of Merck's Crixivan, including one compound with more potent protease inhibition in vitro than the parent compound; two nucleoside analogues reverse transcriptase inhibitors from Glaxo Wellcome, 158U89 and 54W91; and a new non-nucleoside reverse transcriptase inhibitor from Eli Lilly (LY300046-HC1).

Other new agents included a new compound, PRO 2000, that disrupts CD4/gp120 binding in vitro from Procept; a dextran sulfate analogue called D2S that doesn't yet seem to have a sponsor; several "zinc finger" inhibitors being studies by the NCI intramural program; a tat peptide-DNA that may inhibit TAR RNA; an antisense drug AS S-ODN; a naturally occurring antifungal/antibiotic from Parke Davis called Cerulenin which may also be able to inhibit protease cleavage; a foscarnet analogue; an aminoglycoside called Hygromycin B; several acyclic nucleoside analogues with boronic acid moieties; and finally, a quinolone derivative called Vesnarinone. TAG's coverage of ICAAC presentations on the oral ganciclovir prophylaxis trials (the Syntex study and CPCRA 027), new rifabutin and clarithromycin studies for MAI, and other opportunistic infection information will be covered in the forthcoming issue.