1st International AIDS Society (IAS) Conference on HIV Pathogenesis and Treatment
July 7-11, 2001
Buenos Aires, Argentina
Buenos Aires Summary
Written for NATAP by Nancy Shulman, MD, Stanford University
The 1st IAS Conference on HIV Pathogenesis was held in Buenos Aires, Argentina, where the pleasant 50-60 degree winter weather and optimistic mood of the conference hid any signs of impending economic collapse of the hosting country. The program content was very good, particularly for a 1st ever conference, and between 2500-3000 people attended the conference including most of the leaders in HIV research.
--New drugs update: (PI) DMP-450, tipranavir, BMS-232632
--Kaletra update of two studies (60 weeks & 144 weeks)
--Switch studies: to an NNRTI, BEST study from IDV to IDV/RTV (800/100)
--Abacavir vs IDV: final 48 week dara
--STIs: 3 different studies looking at interruptions of varying length-- 1 week off/on, 3 weeks on/off, 2 months on/1 month off
--3-5 year durability studies of EFV, RTV/SQV, IDV
--PK and dosing of APV/RTVtwice or once daily
--Liver disease and HIV
Opening Session (July 8th pm):
David Ho from the Aaron Diamond AIDS Research Center, opened the meeting with a summary of his groups and other groups research on HIV viral dynamics (how fast HIV replicates) and potential for eradicating the virus from the body completely. By measuring how fast the viral load levels decay (decrease) and the changes in the decay rates over time after treatment with potent antiretroviral drugs or by exchanging the plasma (the part of the blood that viral loads are measured from) from HIV-infected patients with plasmaphresis, they determined the life cycle of HIV from entry into the CD4 T-cell to its progeny leaving the T-cell to be about 24 hours. The half-life of activated HIV-infected T-cells (the time it half of these T-cells to die after circulating in the body) was determined to be 1.5 to 2 days. Macrophages (cells HIV infects that are generally found in the tissues) turn over much slower with an average half-life of about 14 days. In addition, there is a small but long-lived population of cells called the latent resevoir that consists of memory T-cells and other cells that HIV can infect and remain for long periods of time. The cells in the latent resevoir have an estimated half life of nearly 4 years.
Based on the above findings and the additional findings that when most patients have "undetectable" viral loads on current HAART regimens, they still have HIV replication in their body that can infect new T cells including the long-lived ones, it does not seem possible to eradicate HIV with our current treatments. Dr. Ho presented his newer data on how more potent regimens with multiple classes of drugs can increase the rates of viral decay and may limit the measurable and unmeasurable low-level replication which can prolong suppression and reduce the emergence of resistance. For eradication to be possible, we would need more potent drugs that people can take consistently and we need a way to increase the turnover of the latent memory T-cells.
Julio Montaner from the British Columbia Center for Excellence followed with a talk about the evolution of our strategy of when to initiate antiretroviral therapy. We have shifted away from treating HIV in its early stages (except perhaps within two weeks of infection, before seroconversion) and are delaying the onset of treatment for many reasons:
Dr. Montaner presented data from his cohort that showed adherence is key in predicting mortality. In his cohort, patients who started therapy at <50 CD4 count who were >95% adherent to their medications had similar outcomes to those that started therapy at higher T-cells. Writers comment: adherence can be more difficult in patients with CD4 counts of <50 because of concurrent treatments of opportunistic infections and increases in medication side affects. Treatment should be initiated in patients with CD4 counts of <200, the range of 200-350 is controversial and is not supported by Dr. Montaners data. Treatment should be deferred in asympomatic patients with T-cells of 350 according to the most recent Treatment guidelines.
Stefano Vella, current president of the IAS presented the last lecture of the evening discussing the huge problem of getting antiretroviral therapy to the bulk of HIV-infected patients in the world who do not have access to them, 70% of them in sub-saharan Africa, where AIDS is the number one cause of death, causing twice as much as the number 2 killer, tuberculosis. Improvements in average life expectancy that were seen in 1950-1980 have been negated. Life expectancies have recently been reduced by 15 years, to 40 years old in some areas, near the life-expectancy of people of Rome 2000 years ago.
Dr. Vella stressed that prevention strategies are inadequate without treatment of those who are infected and emphasized the moral obligation of the wealthier nations and HIV research/ pharmaceutical community to financially support the countries with less resources to enable access universal access to medication and support infrastructure. Research on cheaper regimens, providing cheaper prices, using pulsed or intermittent therapies to delay cost, and less expensive treatment monitoring methods are urgently needed.
Ho D. PL1
Montaner JG. PL2
Vella S. PL3
Drugs in Clinical Development:
Mozenavir: This is a cyclic urea protease inhibitor that was originally developed by Dupont (DMP-450) and sold to Triangle. This is a highly water soluble protease inhibitor that could be formulated in one pill per dose and is easier to produce than most other protease inhibitors which in a perfect world would mean lower prices. Unfortunately, mozenavir has a short half-life (3 hours) similar to indinavir that is probably going to have to be taken three times a day, and has a resistance pattern that appears to be similar to indinavir (PI mutation 82 develops first).
Results of a small Phase I/II study were presented that evaluated the safety, efficacy, and pharmacokinetics of mozenavir dosed with d4T and 3TC and compared efficacy with d4T, 3TC and indinavir. Fifty antiretroviral naïve patients were randomized to receive d4Tand 3TC with either mozenavir 750mg three times a day, 1250m three times a day, 1250mg twice a day, or indinavir 800mg three times a day for 48 weeks. All arms were roughly the same at suppressing viral load to less than 50 copies at 48 weeks (67-77%), but the 750mg mozenavir arm had less CD4 count rise than the other groups. Of note, there were no changes in the QT intervals (a change in the conduction of the heart seen on an EKG that can rarely lead to a fatal heart rhythm) that were seen in earlier studies of mozenavir in dogs.
Sierra-Madero J, et al. Abstract 2
Tipranavir: Tipranavir, a non-peptidic PI, is one of the most promising PIs in development in that it retains activity to most isolates that are resistant to the currently available PIs, perhaps because of its flexible structure. It has a short half-life, but like other Pis, the half-life is greatly increased by co-administering small doses of ritonavir. In addition, the earlier 300mg hard-filled capsules that were less bioavailable were recently reformulated into a better absorbed 250mg soft-gel capsule.
A phase II dose ranging and efficacy study was presented on tipranavir. Forty-one subjects with a mean CD4 count of 300 and viral load of 4.5 log10 copies/ml who had failed 2 or more PI-containing HAART regimens, but were non-nucleoside reverse transcriptase inhibitor (NNRTI) naïve were randomized to receive either tipranavir 1200mg or 2400mg twice daily in hard-filled capsules in combination with either 100mg or 200mg of ritonavir respectively twice daily. All arms received efavirenz and a new nucleoside. Midway through the study, when the soft-gel capsules were available, the 1200mg and 2400mg tipranavir groups were changed to 500mg and 1000mg respectively and ritonavir was changed to 100mg in both arms. At 48 weeks, 94% of the low dose and 79% of the high dose group suppressed their viral loads to <400 copies in an on treatment analysis (doesnt include the patients who dropped out of the study in the analysis). In an intent-to-treat analysis (where drop-outs are considered treatment failures) these numbers were 78% and 50% respectively. None of these values were statistically different due to the small numbers in each group. Nearly 60% of the patients reported diarrhea, 30% had nausea, 17% had vomiting, but only 2 subjects discontinued the study. Side affects were reportedly higher with the hard-filled capsules.
A pharmacokinetic analysis was performed using the trough levels (lowest levels of drug in the blood, just before taking the next dose. Median trough levels of tipranavir were 46 in the low-dose group while they were taking the hard-filled capsules vs. 79 in the high dose group. During the later portion of the study where the soft-gel capsules were used, the low dose group was 33 and the high dose group was 28. When asked how the higher dose could have generated lower trough levels, the presenter suggested that the patients must not have been adhering as well in the high dose group, which may have been why there was no improved response rates. Analyses of drug resistance have not yet been performed, although they are planned.
The study design had a number of design problems including the change in the formulation and dose midway through the study, the decrease in the ritonavir dose midway through the study and the strange results of the pharmacokinetic analysis. Despite the problems, and the fact that the ideal doses of tipranavir soft-gel capsule and ritonavir have yet to be determined, tipranavir looks promising as a part of salvage therapy in PI-experienced patients.
Curry R, et al. Abstract 3
Atazanavir: This is the rather difficult-to pronounce name that Bristol Myers Squibb chose for its new PI, BMS-232632. This PI can be dosed once per day and the problems of elevated cholesterol and triglycerides associated with the currently available PIs has not been seen with this drug.
Preliminary 24-week data was presented from a phase II study comparing atazanavir 400mg once daily vs. 600mg once daily vs. nelfinavir 1250mg twice daily, all in combination with d4T and 3TC twice a day. 467 patients from 51 different sites were randomized in a 2:2:1 ratio. The median baseline viral load was 4.74 log10 and median CD4 count was 270. The change in viral load from baseline was comparable in all of the groups, around 2.5 logs. 67%, 70%, and 63% achieved <400 copies in the low dose atazanavir, high dose, and nelfinavir groups respectively. All groups had comparable increases in CD4 counts. The most significant finding were the lipid profiles during the study. The median LDL cholesterol (the bad cholesterol) increased by 37mg/dl in the nelfinavir group and did not change in the atazanavir groups (-4 mg/dl). Median triglycerides increased by 30 mg/dl in the nelfinavir group vs. 3mg/dl in the atazanavir group. These numbers were all statistically significant (p<0.0001).
The most common side affect in the atazanavir was a rise in the bilirubin without affecting the other liver numbers. That was seen in half of the high dose group. 6 patients required dose reduction of their atazanavir. No associated elevations in transaminases were seen. It is thought that atazanavir interferes with bilirubin conjugation similar to indinavir. Because of this and the similar efficacy of the 400 and 600mg doses, the 400mg dose will likely be used in future phase 3 studies.
Cahn P, et al. Abstract 5
Lopinavir/Ritonavir (Kaletra) studies: Updated 60-week data was presented on the M98-983 study comparing lopinavir/ritonavir (LPV/r) 400mg/100mg twice a day with Nelfinavir 750mg three times a day, both in combination with D4T and 3TC in 653 antiretroviral naïve patients. At 60 weeks, in an intent-to-treat analysis where those who discontinued the study for any reason were considered as treatment failures, 74% achieved viral loads of <400 copies in the LPV/r arm vs. 61% in the nelfinavir arm (p<0.001). 63% and 51% were <50 copies respectively (p<0.001).
144-week data (about 3 years) was presented from the M97-720 study of D4T+3TC+LPV/r. The study enrolled 100 antiretroviral naïve subjects. At the beginning, subjects received one of 3 doses of LPV/r (200/100mg BID, 400mg/100mg BID, or 400mg/200mg with D4T and 3TC either given after 3 weeks of monotherapy or from the entry of the study. After 48 weeks, all patients received the 400mg/100mg of LPV/r, the dose that is used now in practice. Median viral load was 4.9 log10 and median CD4 was 326 at baseline. In an intent-to treat analysis, 79% were <400 and 76% were <50 copies. Only 5 subjects discontinued due to drug-related adverse events. The yearly rate of viral load rebound after the first year was calculated to be a only 1.2% per year. This means that this combination should be able to control virus for 30 years in most patients. CD4 counts increased a mean of 384 cells from baseline in three years.
RESISTANCE. In this same study, all those subjects who had any viral loads above 400 were tested for resistance. NO resistance to the LPV/r has been seen in any of the people on the drug with detectable virus. This is quite impressive that at 3 years, over 300 patients have not developed protease resistance either by genotype or phenotype. Of 40 samples who were tested for resistance, only 7 had minor secondary protease mutations (L10I, M36I, L63P, and A71T) and no major mutations were seen. Only 38% had had resistance to 3TC (M184V). In contrast, in the nelfinavir arm, 31 of 84 samples (37%) had resistance to nelfinavir (either a D30N or L90M) and 81% had resistance to 3TC.
Ruane P, et al. Abstract 6
White C, et al. Abstract 217
Kempf D, et al. Abstract 129
Other head-to-head therapeutic clinical trials: 1-year analysis was presented on the COMBINE study that was a randomized, open-label study comparing nelfinavir 1250mg twice daily with or nevirapine 200mg twice daily in combination with combivir (AZT+3TC) in antiretroviral naïve patients. 142 patients enrolled with a mean baseline viral load of 5.15 log10 and mean CD4 count was 359. Since the numbers in each group were relatively small, most of the differences in the groups did not achieve significance (p<0.05). There was a trend towards improved rates of suppression in the nevirapine arm with 75% achieving <200 in an intent-to-treat analysis vs. 60% in the nelfinavir group (p=0.056). No difference in CD4 change was observed. In a small subgroup of patients with >100,000 viral load at baseline, 16/21 in the nevirapine (76%) achieved <200 vs. 14/26 in the nelfinavir group (54%), p=0.11. There were 7 cases of hepatitis (grade III/IV ALT elevation), 3 with symptoms, 4 without, in the nevirapine group vs. 1 in the nelfinavir group. 14% of patients in the nevirapine arm developed a rash. 36% of the nelfinavir group had diarrhea. This is yet another study that showed that non-nucleoside reverse transcriptase inhibitors are at least as good, if not better first line regimens as protease inhibitors.
48 week data was presented on CNA3014 that compared combivir + abacavir to combivir + indinavir. 342 subjects were enrolled including 33% with >100,000 viral load at baseline. In the intent-to-treat analysis, abacavir was superior to indinavir at suppressing viral load <400 copies (p=0.002). In the <50 analysis, the groups were equal with abacavir at 57% and indinavir at 53%. In the subgroup of subjects with baseline viral loads of >100,000, neither the triple nucleoside group or the indinavir idinavir group faired as well with 48% vs. 46% <50 copies. Not surprisingly, abacavir adherence was better than the indinavir group who had to take medication three times a day on an empty stomach. 72% of the abacavir group were >95% adherent vs. 45% of the indinavir group. The outcome of the study suggested the abacavir with combivir was at least equivalent to indinavir at lower viral loads. Indinavir is probably more potent, but the abacavir regimen was much easier for patients to take. Abacavir does not look as potent at higher viral loads as was previously seen in earlier registration trials, but at lower viral loads, it gives patients a very simple option of one pill twice a day and we know that adherence is crucial to success.
Podamczer D, et al. Abstract 7
Vibhagool A, et al. Abstract 63
Switch Studies: A randomized study switching patients who were suppressed on protease-based regimens to either nevirapine (n=50) or efavirenz (n=47) was presented. 88 subjects have completed 24 weeks of therapy. 7 patients in the nevirapine arm have discontinued treatment, 6 from drug toxicity (rash and liver toxicity) and 1 because of virologic failure vs. 2 subjects in the efavirenz group, both for side affects. In summary, both groups maintained viral load suppression great. This study had a high rate of nevirapine-related toxicities.
Many patients are being switched from indinavir 800mg three times a day to a regimen that is twice a day and includes a combination of indinavir and ritonavir for ease of dosing. A very interesting study was presented called the BEST study. This enrolled 326 subjects who were suppressed on indinavir three times a day plus two nucleosides and randomized them to either continue their current regimen or switch to indinavir 800mg along with 100mg ritonavir twice a day. After one year of follow-up, 42% of the twice-daily arm had stopped that therapy, mainly due to adverse events (30%). These were mostly gastrointestinal symptoms like nausea and diarrhea. Only 24% of the three times a day arm discontinued the regimen, 11% related to adverse events. Kidney stones occurred in 20% of the combination arm vs. 7% of the three times a day arm. Those who remained on treatment in both groups maintained virologic suppression at similar rates 94% in the three times a day vs. 95% in the twice a day combination arm. After prompting from the audience, the consensus was one of the cardinal rules in medicine and in life, "If it aint broke, dont fix it."
Patterson P, et al. Abstract 58
Cahn P, et al. Abstract 60
Structured Treatment Interruptions (STI): A randomized study of 60 patients who received STI with cycles of 3 weeks on followed by 3 weeks off for 36 weeks vs. continuous therapy with D4T, DDI, and indinavir with or without hydroxyurea was presented by Franco Lori. No changes in rebound level of virus was observed in the subjects randomized to STI, but subjects in the STI arms had similar viral load reductions at the end of 36 weeks, and had similar CD4 increases. Data on cholesterol or other toxicities was not presented. [The study data presented was preliminary: 30/60 patients]. It is encouraging to see that intermittent therapy can maintain similar CD4 levels to that of continuous therapy. This is probably because in general, the viral load is still suppressed most of the time. CD4 counts did fall with each interruption, but rose back up. Longer-term follow up is needed as well as how many of the interrupters developed resistance. Data is also needed about the rates of side affects in the two groups.
Tony Fauci presented data from his group during a plenary lecture on treatment interruptions. In his study of long cycle STI with efavirenz and 2 nucleosides for 2 months followed by 1 month off. 4 patients of about 30 or so developed a K103N associated with efavirenz resistance over the course of the observation period in addition to several with M184V associated with 3TC resistance. This has been the main concern about the strategy of treatment interruptions, particularly with drugs like efavirenz and Nevirapine that hang around in the blood for long periods after the drug is stopped, exposing the virus to low levels of drug. That is a perfect set up for resistance.
He presented a study of short cycle STI in 11 patients who received D4T, 3TC, ritonavir and indinavir in cycles of 7 days on and 7 days off. These patients have remained suppressed, have no observed resistance out one year, no change in their viral reservoir and most have reduced their cholesterol and LDL from baseline to some degree. Longer follow up is needed (Editorial note: its only 11 patients in this study. One week on and one week off may create adherence difficulties & confuse patients).
Lori F, et al. Abstract 56
Fauci T. PL-4
Durability/long-term therapy studies: 5-year data (260 weeks) was presented on the original 33 AZT-experienced subjects in the Merck 035 study who received AZT+3TC+ indinavir 800mg three times a day. These patients had median viral load of 41,900 and CD4 count of 133. Subjects had received a median of 28 months of AZT monotherapy. 2 subjects who discontinued therapy for reasons unrelated to treatment with viral loads <500 were not included in the analysis. 21/31 (68%) had viral loads of <500 at 3 years. 3 of those 21 subjects discontinued therapy for adverse events, all related to kidney stones during the 4th year of therapy. The remaining 18/31 (58%) are <500 after 5 years. T-cells have increased a median of 252 over their baseline in the group.
4-year data (192 weeks) on ritonavir-saquinavir (RTV-SQV) therapy was presented. This was a study of 141 protease-naïve subjects who were randomized to receive RTV-SQV at different doses (400/400, 400/600, 600/400, or 600/600 all twice daily). Nucleosides were used to intensify when viral load rebounded or was >200 at 12 weeks. All subjects could add nucleosides after the first year. Most patients eventually changed to the 400/400 of RTV/SQVdose from the higher dose groups. Of the 76 patients who made it to 4 years, 95% were <200 copies. Intent-to treat analysis not done, but that makes over half of the original group <200 at 4 years.
3-year data (144 weeks) was presented on the DMP-006 study that randomized 450 antiretroviral naïve subjects to receive efavirenz (EFV)+AZT+3TC vs. indinavir (IDV) +AZT+3TC vs. IDV + EFV. At week 144, in an intent-to treat analysis, 55% of the EFV+AZT+3TC arm vs. 34% of the other two arms (p<0.05) were <500 copies and similar numbers were <50 copies (52% vs. 30% vs. 35%). Virologic failure rate in later years appears to be about 1.5-2% per year in the EFV+AZT+3TC arm, similar to the Kaletra data.
Gulick R, et al. Abstract 215
Farthing C, et al. Abstract 223
Tashima K, et al. Abstract 224
Pharmacokinetics: In a study of ritonavir/amprenavir in healthy volunteers, amprenavir 600mg dosed with 50mg of ritonavir twice daily had similar pharmacokinetics as the same dose of amprenavir with 100mg of ritonavir twice daily. In once-daily dosing with 1200mg of amprenavir, 100mg and 200mg once again, were very similar in regards to Cmax (highest level in the blood), Cmin (lowest level in the blood), and AUC (a measure of the average levels in the blood over time). These studies need to be done in HIV-positive patients, but suggests we can get away with less ritonavir and probably less side affects in this combination.
Kurowski, et al. Abstract 351
Liver disease and HIV: A small preliminary study was presented of treatment of hepatitis C with Pegylated interferon (Schering-Plough) plus ribavirin in HIV-infected subjects with CD4 counts >300, and HIV viral load <5000. 51 interferon-naïve subjects from Spain have begun treatment with Peg-Intron 150mg injected subcutaneously every week for 3 months followed by 100mg per week for 3 months. Ribavirin 400mg twice daily (lower than usual dose of RBV) was started at the beginning of treatment. About 50% of these subjects had gentoype 1a or 1b hepatitis C types, which have a poorer response to interferon-based therapies. This is in contrast to an 80% prevalence of genotype 1 in HCV-infected people in the US. 60% of the subjects had "high HCV viral loads" defined as >800,000 copies/ml. This is also associated with a poorer response to treatment.
31 subjects have reached the end of treatment. Only 4 subjects have discontinued treatment. Of the 31, 20 (65%) have cleared their HCV from the blood at the end of treatment and 82% had a normalization of their ALT (which has been shown to correlate with improvements of inflammation in liver biopsies in some studies). Data was not presented about how many subjects had advanced fibrosis or cirrhosis in the subjects, which has also been associated with a reduced response and poorer tolerance to therapy. Data was not presented on the breakdown of response based on genotype.
Although very preliminary, these results are encouraging, Particularly the low drop out rate. Prior small studies of standard interferon and ribavirin in HIV patients have had high drop out rates (20-30%). This may be due to the improved tolerability of long-acting interferons vs. standard interferons. We await the follow up of these patients once they have been off therapy for 6 months (the "sustained response").
Liver disease associated with antiretrovirals is a growing concern with deaths occurring in clinical trials. The roles of the individual antiretrovirals is unclear. The NIH DAIDS group presented a retrospective analysis of 21 ACTG studies that included 10,611 patients on studies ranging from nucleoside mono- and dual therapies to NNRTI- and PI-based regimens. Data on concurrent hepatitis infection and alcohol intake were not available [limiting the interpretation of the data]. Liver toxicity was defined as a grade 3 or 4 elevation in ALT which is 5 times the upper limit of normal, or any liver-related deaths.
6.2% of the patients met the definition for liver toxicity. This led to discontinuation of study in 23% of these 6.2%. 0.4% of subjects had liver related deaths. Most of these deaths were in patients receiving only nucleosides. Older monotherapy trials had higher rates than dual therapy. DDI dosed at 500mg-750mg/day monotherapy had the highest rate of 10.3%, compared with 6.2% who took DDI 400mg/day, all AZT doses 5.5%, and D4T 3.7%. NNRTIs with nucleosides had higher rates than PIs with nucleosides (8.2% vs. 5%). In a breakdown of the specific NNRTIs, efavirenz had a rate of 10.8% vs. 8.9% in nevirapine, and 3.6% in delavirdine.
Boehringer-Ingelheim, the manufacturer of nevirapine, which has received the most publicity about liver toxicity, presented a review of 4 of their key trials of nevirapine with nucleosides that had a nevirapine arm and a nevirapine placebo arm. The studies included were BI 1090 in which 2299 patients with advanced disease (median CD4 93) received AZT or D4T with 3TC and nevirapine or placebo. 3 other trials were BI 1037, 1038, and the INCAS enrolled altogether 456 antiretroviral naïve patients with higher CD4 counts (median 385). Concurrent hepatitis infection status was known in included in the analysis. Again, liver toxicity was defined as a 5-fold or grade III/IV elevation in ALT. In BI 1090, the largest study, the nevirapine group rate was 3.4% vs. 2.2% in the placebo group. In the naïve studies, the rate in the nevirapine group was 8.1% vs. 3.0% in the placebo group. When stratifying this naïve group based on CD4 < or >350, those with lower CD4 counts had had only a 2.9% rate of hepatotoxicity vs. 0% in the placebo. In the higher CD4 count group >350, rates of hepatotoxicity were much increased with nevirapine group 13.1% vs. 5% in the placebo group. When controlling for concurrent hepatitis B or C, the risk of nevirapine hepatotoxicity was only increased in the group without hepatitis B or C.
Therefore, nevirapine hepatotoxicity appears to be more frequent in patients with high CD4 counts. This is in keeping with the rates of Nevirapine related liver toxicity in other studies: ACTG 193 median CD4 20 rate 3.1%; Atlantic Study median CD4 406-rate 9.1%; FTC302 - median CD4 360-rate 17%.
Soriano, et al. Abstract 42
Riesler, et al. Abstract 43
Dieterich, et al. Abstract 44