Highlights from the 8th Annual Conference on Retroviruses and Opportunistic Infections
The 8th CROI was sponsored by the Foundation for Retrovirology and Human Health, in scientific collaboration with the U.S. National Institute of Allergy and Infectious Diseases (NIAID) and the U.S. Centers for Disease Control and Prevention (CDC). Additionally, nine pharmaceutical or diagnostic test corporations were financial supporters of various aspects of the conference. Thirty-nine full or partial scholarships were awarded to community participants, and approximately 125 travel grants were awarded to graduate students and fellows, including four from developing countries.
While there were no major breakthroughs reported at the 8th CROI, a number of significant though smaller pieces of information have advanced our knowledge, understanding, and treatment of HIV/AIDS. Encouragingly, the pipeline for anti-HIV drugs under development is quite extensive.
Many presentations, including two oral plenary presentations, addressed entry inhibitors, a newer class of anti-HIV drugs. These drugs can be subdivided into attachment inhibitors and fusion inhibitors. The farthest along in development is T-20, a fusion inhibitor from Trimeris and Roche that binds to gp41 on the envelope of HIV.
Six separate presentations focused on T-20. The most significant clinical presentation was authored by Jay Lalezari, M.D., of Quest Clinical Research in San Francisco in the late-breaker session. Dr. Lalezari's Phase II, randomized study was the first controlled safety and efficacy study of T-20 injections in chronic dosing, with results out to 16 weeks. Eligibility criteria were a detectable viral load (greater than 400 copies/mL), experience with protease inhibitor (PI) drugs, and no prior experience with non-nucleoside reverse transcriptase inhibitor (NNRTI) drugs. A total of 71 subjects (4% women) were randomized to placebo (inactive drug) or T-20 at a twice-daily dose of 50, 75, or 100mg by self-injection. The older formulation of T-20, 50mg/mL, was used. All participants also received abacavir (Ziagen), efavirenz (Sustiva), amprenavir (Agenerase), and low-dose ritonavir (Norvir, 200mg twice daily). The median baseline CD4 cell count was 232 cells/mm3, with a viral load of 4.3 log (18,620) copies/mL. Past anti-HIV drug history and baseline drug resistance testing were not reported.
The 16-week results were as follows. Using a strict intent-to-treat analysis (i.e., all enrolled subjects included), the T-20 arms achieved a 48% rate of viral undetectability (limit 50 copies/mL), compared with 37% for the control arm. The percentages by study arm were not reported. Excluding seven T-20 subjects who apparently had no viral load results, the pooled T-20 undetectability rate was 56%. In an on-treatment analysis (i.e., including only those who remained on drug), the 100mg dosing arm achieved the greatest viral load decrease (-2.8 log copies/mL), compared with -2.2 log for the control and 50mg arms and -2.3 log for the 75mg arms. The CD4 cell count increases were: 10 cells/mm3 (control arm with HAART [highly active antiretroviral therapy]); 37 cells/mm3 (50mg T-20); 74 cells/mm3 (75mg); and 64 cells/mm3 (100mg).
The most common adverse event was a local reaction at the site of skin injection, occurring in 57-77% of those in T-20 arms. The only grade 4 (life-threatening) local reaction occurred in the 100mg arm: a skin abscess, or pocket of pus, due to unsterile injection technique by the participant on the abdomen (stomach) wall. However, grade 2 (moderate) reactions occurred in 18% of T-20 arms, with a similar rate in each dosing arm; grade 3 (severe) reactions occurred in 9% of T-20 arms. Two participants discontinued T-20 as a result. Adverse events that were more frequent in the 100mg arm than in the HAART plus placebo arm were nausea, dizziness, and headache; however, these findings were not consistently related to increasing T-20 dosage. Interestingly, the T-20 arms had a slightly lower rate of insomnia (sleeping difficulties) and hypoasthesia (decreased sensation on skin) than the HAART plus placebo arm. Regarding grade 3-4 laboratory abnormalities, there were no obvious significant trends: they occurred in 21% (control), 38% (50mg), 25% (75mg), and 13% (100mg) among the respective groupings. Each of the T-20 study arms had one case of grade 3-4 liver enzyme increases; however, the placebo arm had two such subjects. Due to adverse events, three participants reduced their T-20 dose to 50mg twice daily. Discontinuation rates due to adverse events were 16% (control), 6% (50mg), 5% (75mg), and 31% (100mg). Dr. Lalezari mentioned that all subjects were allowed to rollover after 16 weeks into the T-20 100mg arm, using an improved formulation of 100mg/mL that allows for one injection twice daily. It will be possible to understand the results of this study better when baseline and 16-week drug resistance information is available. Phase III studies with T-20 are underway and expanded access for T-20 is imminent, though likely to be limited due to production constraints.
Other research on T-20 was presented by Dr. M. Greenberg of Duke University in Durham, NC. T-20 and T-1249 (another fusion inhibitor being developed by Trimeris/Roche) will likely be effective in early as well as late HIV infection and would not be dependent upon CCR5 or CXCR4 entry coreceptors. Dr. Greenberg found that in the laboratory, "coreceptor specificity had no impact on sensitivities to T-20 or T-1249." Two other reports at CROI focused on T-20 efficacy among HIV positive children.
Greenberg ML and others. Virus sensitivity to T-20 and T-1249 is independent of coreceptor usage. Abstract and poster 473.
Lalezari J and others. A controlled Phase II trial assessing three doses of T-20 in combination with abacavir, amprenavir, low dose ritonavir and efavirenz in non-nucleoside naive protease inhibitor experienced HIV-1 infected adults. Abstract and Late Breaker presentation LB5.
Joseph Eron, M.D., of the University of North Carolina at Chapel Hill reported Phase I/II study results of T-1249 with 72 subjects (8% women). All HIV isolates with resistance to T-20 are sensitive to T-1249, which binds to a different part of HIV's gp41 than does T-20. In laboratory testing, resistance to T-1249 reportedly has been difficult to generate. T-1249 has a half-life (time until an original amount is decreased by half) that will allow for once-daily dosing by self-injection. In the current 14-day study, T-1249 monotherapy dosing was 6.25-50mg, with most dosing levels divided into once or twice daily. The mean baseline CD4 cell count ranged between 84 and 146 cells/mm3, with 98% having prior anti-HIV treatment with a mean of ten drugs; the baseline viral load ranged between 5 and 5.5 log (89,125-346,736) copies/mL.
The results showed a dose-dependent reduction in HIV viral load, with a maximal reduction of -1.4 log (25-fold) in the 50mg arm. There was also a dose-dependent increase in blood concentration of T-1249. The most common adverse event was local skin injection site reaction (pain, discomfort, and redness) in 40%; all but one was mild (grade 1). Other adverse events included headache (11%), dizziness (8%), fever (8%), and diarrhea (6%). There were two serious adverse events: hypersensitivity (allergic) reaction and neutropenia (low white blood cells). However, there were no dose-dependent patterns of adverse events overall.
Dr. Eron also reported that the maximal tolerated dose has not yet been achieved, and that due to an apparent unique resistance pattern, T-1249 may be an excellent follow-up compound to T-20. Other studies are ongoing.
Eron J and others. A 14-day assessment of the safety, pharmacokinetics, and antiviral activity of T-1249, a peptide inhibitor of membrane fusion. Abstract and oral presentation 14.
BMS-232632 is an experimental once-daily PI drug under development by Bristol-Myers Squibb. This PI has little cross-resistance with the marketed PI drugs (due to a unique "N88S" mutation) and does not cause the significant increases in total cholesterol or triglycerides (fats) that commonly occur with other PIs and efavirenz. Phase II study (AI424-007) results were presented by Kathleen E. Squires, M.D., of the University of Southern California (USC). A total of 420 treatment-naive (i.e., without any previous treatment) participants (36% women, 44% non-European-American) were enrolled, with a median CD4 cell count of 305-386 cells/mm3 and a viral load of approximately 4.7 log (51,000) copies/mL. Participants were randomized to receive BMS-232632 at a once-daily dose of 200, 400, or 500mg, or nelfinavir (Viracept, a PI drug) 750mg three times daily, with all treatment arms adding the two nucleoside reverse transcriptase inhibitor (NRTI) drugs ddI (Videx) and d4T (Zerit) after two weeks.
The interim results for Stage II subjects (77% of entire group), using a strict intent-to-treat analysis, were as follows after 24 weeks. Approximately 40% of the BMS-232632 arms and 30% of the nelfinavir arm achieved viral undetectability (lower limit 50 copies/mL), while the CD4 cell count increase was approximately 100 cells/mm3 in all four study arms.
Approximately 15% of the two higher-dose BMS-232632 arms and the nelfinavir arm had grade 3-4 clinical adverse events. The most common overall adverse events were diarrhea, infection, nausea, and abdominal pain. Grade 3-4 laboratory events included increased liver enzymes (ALT or AST) in 6-12% of the two higher BMS-232632 dosing arms and 1-4% in the nelfinavir arm, with a very low rate of very high total cholesterol (greater than 300mg/dL) in the BMS-232632 arms (1% each), compared with 9% in the nelfinavir arm. Interestingly, Dr. Squires reported that all four arms sustained an increase in the HDL (high-density lipoprotein, or "good") cholesterol of approximately 10mg/dL. Yet while the LDL (low-density lipoprotein, or "bad") cholesterol increased by approximately 25mg/dL in the nelfinavir arm, the three BMS-232632 arms had no significant change. Regarding triglycerides, there was essentially no change in the BMS-232632 arms, yet there was an increase of approximately 25mg/dL in the nelfinavir arm. There was a dose-dependent increase in unconjugated bilirubin (bile pigment not bound to blood protein) in the BMS-232632 arms without liver abnormalities or symptoms, except for icterus (yellowing of the eye whites) in some. This is somewhat similar to a phenomenon that occurs with indinavir (Crixivan) use and with the benign condition called Gilbert's syndrome. Grade 3-4 increases of bilirubin occurred in 28% of the 400mg BMS-232632 arm and 42% of the 500mg arm. Yet dose reduction due to grade 4 elevations was uncommon, occurring in less than 5% of the 500mg arm. Researchers from Bristol-Myers have shown previously that the increase in bilirubin is associated with subjects' genotype for UDPGT (uridine diphosphate glucuronosyltransferase).
The discontinuation rate due to adverse events was 3% in each of the two higher dosing arms of BMS-232632 and 5% in the nelfinavir arm. Dr. Squires reported that the 400mg once-daily dose of BMS-232632 is being used for Phase III studies.
In a separate presentation, Edward O'Mara, M.D., of Bristol-Myers presented the open-label results of a study of 32 healthy volunteers who took BMS-232632 once daily (200 or 400mg) as monotherapy for six days with a light meal and then added ritonavir (100 or 200mg) once daily for ten days. The results showed a general dose-dependent increase in the blood concentrations of BMS-232632, including total exposure (area-under-the-curve, or AUC), maximal, and minimal concentrations. Yet in the 400mg dosing arms of BMS-232632, the various drug levels were not dramatically different when compared with the 100 or 200mg ritonavir arms. Reversible increases in total bilirubin (mostly unbound) occurred in approximately 93% of 30 subjects (levels ranged from 1.2 to 7.7), yet returned to normal after no more than four days off therapy for 89% of them. Previously unreported, 30% of 30 subjects had mild increases in conjugated (bound to protein) bilirubin that returned to normal within four days after drugs were stopped. Icterus occurred in 22%, yet no liver abnormalities occurred with this or with any of the bilirubin increases. However, 6% of participants did develop jaundice (yellowing of the skin) that resolved after the drugs were stopped. No serious adverse events were reported, and no subjects discontinued due to adverse events of study drugs. Similar studies will need to be carried out in HIV positive subjects.
O'Mara E and others. Steady-state pharmacokinetic interaction study between BMS-232632 and ritonavir in healthy subjects. Abstract and poster 740.
Squires KE and others. AI424-007: 48-week safety and efficacy results from a phase II study of a once daily HIV-1 protease inhibitor, BMS-232632. Abstract and oral presentation 15.
GW433908 is a water-soluble prodrug of amprenavir. The formulation of 465mg per tablet will allow for twice-daily dosing of three or four tablets that are smaller than amprenavir capsules. The drug is under development by Vertex Pharmaceuticals and GlaxoSmithKline. The results of a 28-day, double-blind, randomized crossover study (APV20001) were presented by Dr. R. Wood of Somerset Hospital in Cape Town, South Africa. A total of 85 treatment-naive subjects were randomized to 1,395 or 1,860mg (3-4 tablets) of GW433908 twice daily or standard amprenavir 1,200mg (eight capsules) twice daily, with all three arms also receiving the NRTI drugs abacavir and 3TC (Epivir). The baseline median CD4 cell count was 177-348 cells/mm3, with a viral load of approximately 4.6 log (39,810) copies/mL. The pharmacokinetic (drug concentration) results showed that the total drug exposure of GW433908 was nearly identical in both study arms (and to the amprenavir level in that arm), while the maximal concentration of GW433908 was 18-28% lower than that of amprenavir and the minimum concentration was 24-29% higher than amprenavir. Other results after 28 days revealed an approximate -1.9 log copies/mL reduction of HIV RNA in both study arms. The mean CD4 cell count increase was approximately 108 cells/mm3 in the GW433908 arms and 91 cells/mm3 in the amprenavir arm.
Adverse events that were moderate or worse were similar among study arms. However, the GW433908 arms tended to have a slightly higher rate of headache and sleep problems, while the amprenavir arm had slightly more nausea, diarrhea, abdominal pain, and gas symptoms. Adverse laboratory tests that were moderate or worse were also similar in the study arms, with the GW433908 subjects having a somewhat lower rate of increased liver enzymes. No participant discontinued during the 28-day period due to adverse events. After this study, all subjects were given the option to roll into open-label amprenavir with abacavir and 3TC. It would seem that the next step for this drug would be boosting with low-dose ritonavir.
Wood R and others. GW433908, a novel prodrug of the HIV protease inhibitor amprenavir: safety, efficacy and pharmacokinetics (APV20001). Abstract and poster 333.
DPC 681 and DPC 684
DPC 681 and 684 are experimental, second-generation PI drugs under development by DuPont that are currently in Phase I studies. The drugs are effective against 30 selected HIV isolates with resistance to currently marketed PI drugs and to wild-type (nonmutated) HIV in laboratory tests. Dr. Susan Erickson-Viitanen of DuPont described how the protein-adjusted IC90 (inhibitory concentration to block 90% of HIV growth) for these two drugs is extremely favorable, when compared with other PI drugs.
Erickson-Viitanen S and others. DPC 681 and DPC 684: resistance and cross-resistance profiles of second-generation HIV protease inhibitors. Abstract and oral presentation 11.
TMC126 is a prototype of a new class of so-called resistant-repellant PI drugs, according to Dr. John Erickson of Tibotec in Rockville, M.D. Currently in preclinical development (laboratory testing only), the drug has anti-HIV activity at femtomolar (exceedingly low) concentrations against wild-type HIV and various isolates that are resistant to currently marketed PI drugs, including multidrug-resistant isolates. TMC126 was specifically designed to block pathways toward mutations that cause resistance.
Erickson J and others. A femtomolar HIV-1 protease inhibitor with subnanomolar activity against multidrug resistant HIV-1 strains. Abstract and oral presentation 12.
Other PI drugs under clinical development that had no new data presented at CROI but that were mentioned during a plenary session called "New Antiretroviral Drugs" by Roy M. Gulick, M.D., M.P.H., of Cornell University in New York include: AG 1776, mozenavir (DMP-450), and tipranavir.
Gulick RM. New antiretroviral drugs. Abstract and oral presentation S25 at session 95.
TMC120 is a highly potent, second-generation NNRTI under development by Tibotec, based in Belgium. Dr. Rudy Pauwels presented the results of a seven-day monotherapy study that was randomized, placebo-controlled, and double-blinded (neither subjects nor treating physicians knew who was receiving drug or placebo). A total of 43 treatment-naive subjects (19% women) from Moscow and Warsaw were randomized to receive TMC120 twice daily at a dose of 50 or 100mg, or placebo. The mean baseline CD4 cell count was 571 cells/mm3 with a viral load of approximately 4.6 log (40,258) copies/mL. Both active drug arms achieved a significant viral load reduction of approximately -1.5 log (32-fold) copies/mL with a significant CD4 cell count increase of 121 cells/mm3. The only serious adverse event was an increase of liver enzymes in one subject with hepatitis C virus (HCV) coinfection. Other adverse events were uncommon and not serious. One subject discontinued prematurely due to acute HIV infection. Two-thirds of the participants were tested and had no genotypic resistance mutations at the end of the study. Dr. Marie-Pierre de Bethune presented an overall resistance profile of TMC120 showing that the drug has activity against most HIV isolates with single mutation resistance to the three currently marketed NNRTIs, including the common K103N mutation. Double-NNRTI mutation isolates were not reported.
Gruzdev B, Pauwels R and others. TMC120, a new non-nucleoside reverse transcriptase inhibitor, is a potent antiretroviral in treatment-naive, HIV-1 infected subjects. Abstract and oral presentation 13.
De Bethune M-P and others. TMC120 (R147681), a next-generation NNRTI, has potent in vitro activity against NNRTI-resistant HIV variants. Abstract and poster 304.
Capravirine is an experimental NNRTI under development by Agouron Pharmaceuticals. Since the drug appears to have caused vasculitis (blood vessel inflammation) in dogs, its development has been placed on hold. However, the results of a Phase II study with capravirine were presented at CROI. This drug has activity against the common K103N mutation that causes resistance to the currently marketed NNRTI drugs. In the study, persons with a detectable viral load (at least 2,000 copies/mL) on a current NNRTI regimen who were naive to PI drugs were randomized to placebo or capravirine twice daily (1,400 or 2,100mg) with nelfinavir 1,250mg twice daily plus two new NRTIs. Preliminary results after 12 weeks showed that greater than 50% of subjects randomized to capravirine achieved an undetectable viral load (limit 400 copies/mL). The most common side effects were gastrointestinal, which led to study discontinuation for some participants. Dr. P. Wolfe of Pacific Oaks Research in Beverly Hills, CA, concluded that the interim results showed anti-HIV efficacy and safety for capravirine, and that the 1,400mg twice-daily dose was supported for future studies.
Wolfe P and others. Safety and efficacy of capravirine versus placebo in HIV-infected patients failing a non-nucleoside reverse transcriptase inhibitor-containing regimen: results of a phase II, double-blind, placebo-controlled study. Abstract 323.
No new information was presented at CROI about DPC 083, an experimental NNRTI under development by DuPont. This drug has activity against the common K103N NNRTI resistance mutation as well as selected single- and double-NNRTI mutation combinations. DPC 083 requires more than one mutation for significant resistance. DuPont has other related NNRTI drugs in development, including DPC 961, DPC 963, and DPC 082. Other NNRTI drugs in the pipeline include emivirine (Coactinon, developed by Triangle Pharmaceuticals) and calanolide A.
Gulick RM. New antiretroviral drugs. Abstract and oral presentation S25 at session 95.
L-Fd4C is an experimental NRTI under development by Achillion Pharmaceuticals. Dr. Lisa Dunkle presented preclinical results of this cytosine drug that has activity against HIV and hepatitis B virus (HBV). Dr. Dunkle suggested it might be of particular benefit for HIV-HBV coinfected persons, since it is 10-20 times more potent against HBV than is 3TC. The long intracellular half-life of the triple-phosphorylated component may allow for once-daily dosing. In the laboratory, L-Fd4C has activity against several NRTI resistance mutations, including 3TC's M184V mutation and the multidrug resistance mutations 151M and 69S. Dr. Dunkle said that NRTI drugs with the "unnatural L-configuration have emerged as potentially safer drugs for long-term dosing" due to decreased mitochondrial DNA toxicity that occurs with many of the Food and Drug Administration (FDA)-approved NRTI drugs (mitochondria produce energy within cells). In laboratory tests, L-Fd4C partially reversed mitochondrial toxicity due to d4T or ddI. This is because L-Fd4C "may inhibit transport of D-nucleosides into mitochondria through competitive inhibition of [a] carrier protein." Phase I studies began in 2001.
Dunkle LM and others. ACH-126,443: a new nucleoside analog with potent activity against wild-type and resistant HIV-1 and a promising pharmacokinetic and mitochondrial safety profile. Abstract and poster 303.
FTC is an experimental, once-daily NRTI cytosine drug under development by Triangle with activity against HIV and HBV. Three reports about FTC were presented at CROI. In the first report, Charles van der Hoorst, M.D., of the University of North Carolina at Chapel Hill presented the results of two Phase III studies that show quite similar efficacy when comparing FTC with 3TC in combination therapy.
In Study FTC-303, 440 subjects in North America (14% women, 36% non-Caucasian) with an undetectable viral load (limit 400 copies/mL) while taking 3TC-containing HAART were randomized to continue the same regimen (3TC twice daily) or to switch the 3TC component to open-label (not blinded) FTC 200mg once daily. The results after 48 weeks showed that the viral load became undetectable (limit 50 copies/mL) in 68% of the FTC arm and 75% of the 3TC arm.
In Study FTC-302, 468 treatment-naive subjects from South Africa (59% women, 88% non-Caucasian) were randomized to take either FTC or 3TC (standard twice-daily dosing), plus either efavirenz or nevirapine (Viramune) plus d4T. The study was double-blind (i.e., the drugs being taken were unknown to participants and treating physicians). Those with a baseline viral load of at least 100,000 copies/mL took efavirenz (18%), while those with a baseline level less than that took nevirapine (82%). The median baseline CD4 cell count was 373 cells/mm3, with a viral load of 4.6 log (40,000) copies/mL. Using a strict intent-to-treat analysis, the results after 48 weeks showed that an undetectable viral load (limit 50 copies/mL) was achieved by 61% of the FTC-containing arms and 65% of the 3TC-containing arms. Efficacy by baseline viral load or efavirenz versus nevirapine regimens was not presented.
Dr. van der Hoorst reported adverse events when comparing FTC- with 3TC-containing arms for both studies. Deaths were uncommon and similar when comparing the two drugs (see below). In Study FTC-303, there was significantly more nausea, vomiting, and gastroenteritis in the FTC arm (26%) than in the 3TC arm (17%), as well as significantly more headache (13% in FTC vs 6% in 3TC). However, all of those side effects occurred with near equal frequency in the FTC- and 3TC-containing arms (32-35%) of Study FTC-302. Conversely, dizziness was significantly less common among the FTC-containing arm (7%) than the 3TC-containing arm (13%) in Study FTC-302, yet with an equal rate in the FTC- and 3TC-containing arms of FTC-303 (4-5%). Grade 3-4 (severe to life-threatening) laboratory adverse events were similar when comparing the two arms of both studies. Discontinuation rates were not reported for the two studies.
Study 302 was terminated prematurely due to two deaths associated with liver failure (see report below) not due to FTC. Interestingly, Dr. van der Hoorst reported that the "higher rate of virologic failures not associated with resistance in the FTC arm of study FTC-302 (60%) [than in the 3TC arm, 23%] may imply adherence differences within [the] study." Measurement of FTC drug levels to confirm that were not performed or not reported. He concluded that there was comparable safety and antiviral efficacy in the FTC- and 3TC-containing arms of both studies, and that together with a low rate of resistance to the former makes FTC "an attractive [once-daily] component to a potent triple-therapy drug regimen."
An entire oral presentation was devoted to severe liver toxicity in Study FTC-302, which was terminated prematurely due to two deaths in women due to hepatitis. John A. Bartlett, M.D., of Duke University presented the report on behalf of the FTC-302 clinical investigators. One of the women who died had been randomized to take FTC and the other to take 3TC; both were taking nevirapine and d4T. The woman randomized to 3TC did have a positive test for HBV surface antigen, but with a negative HBV DNA test. Neither developed acute hepatitis A (HAV); hepatitis E status was not known or not reported.
Severe hepatotoxicity (liver toxicity) was evaluated in the study in some detail: it was defined as grade 3 or 4 increases in liver enzymes (ALT, AST, and alkaline phosphatase) and total bilirubin. Overall, severe liver toxicity occurred in 17% of the nevirapine arm and in none of the efavirenz arm. Within the nevirapine arm, severe liver toxicity was present in 14% of the FTC arm and 19% of the 3TC arm. When evaluating several potential predictive factors for severe toxicity among those randomized to nevirapine, the only significant one was female sex (72% of those with toxicity vs 57% of those without). The majority (77%) of those cases of severe liver toxicity occurred within the first four weeks, although in the nevirapine arm, both FTC and 3TC subgroups continued to accumulate more affected subjects up to week 48. One-third of those with severe liver toxicity had symptoms in their gastrointestinal tract, while a different third had a rash that commonly occurs with NNRTI drugs. Only 6% of those with severe liver toxicity had chronic hepatitis B at baseline (a positive surface antigen test), with 4% overall having "active replication" of HBV (a positive HBV DNA test). Baseline infection with HCV was present in only 3% of those with toxicity. There was no significant association between developing severe liver toxicity with either FTC or 3TC, iron levels, alcohol use, parasite (worm) disease, diet, traditional (Western) medication usage, age, Black ethnicity, body mass index (BMI, weight divided by height, squared), CD4 cell count, HIV viral load, or baseline liver tests. (Note that baseline information about this study is presented above.) Of those with severe liver toxicity, 39% permanently discontinued, 3% died, 3% unsuccessfully reintroduced study drugs, 28% successfully reintroduced study drugs, and 30% were successfully treated through the toxicity (their drugs were not stopped).
Dr. Bartlett described in detail how the researchers concluded that neither FTC nor 3TC were the primary causative factors of the two deaths or of severe toxicity, and that the two study arms (FTC vs 3TC) were equivalent, including: number of enrollees (234 in each arm), mean ages of participants (31-32 years), proportion of Black subjects (89% FTC, 84% 3TC), proportion of female subjects (58% FTC, 60% 3TC), baseline median CD4 cell count (367 cells/mm3 for FTC, 355 cells/mm3 for 3TC), baseline viral load (4.5-4.6 log copies/mL), and the proportion of subjects with at least grade 1 (mild) liver abnormalities at baseline (21% in FTC, 23% in 3TC).
There are some limitations to Dr. Bartlett's presentation. Other possible factors or cofactors for severe liver toxicity that were not discussed include: intervening acute viral hepatitis infection (A, B, C, and/or E) or superinfection (hepatitis D), possible pregnancy, possible exposure to liver toxins including aflatoxin and/or acetaminophen (Tylenol), lactic acidosis associated with NRTI drugs, genetic markers in liver (P450), and possible pancreatitis associated with d4T usage. Also, although quite unlikely given the baseline criteria for starting nevirapine vs efavirenz, perhaps having a high baseline HIV viral load (greater than 100,000 copies/mL) was somehow protective against developing severe toxicity. Also, it is certainly possible that the NRTIs in the regimen were cofactors, even if they were not the primary factor. Another report at CROI that addressed liver toxicity, NNRTI drugs, and hepatitis virus coinfection is discussed below. The benefits of a once-daily regimen with FTC in a small, nonrandomized study were updated at CROI; see the "Once-Daily" section.
Bartlett JA and others. Severe liver toxicity in patients receiving two nucleoside analogues and a non-nucleoside reverse transcriptase inhibitor. Abstract and oral presentation 19.
Rousseau F and others. Emtricitabine (FTC): HBV DNA viral load assessments over 36 weeks in patients with chronic HBV infection. Abstract and poster 559.
Van der Hoorst C and others. Two randomized, controlled, equivalence trials of emtricitabine (FTC) to lamivudine (3TC). Abstract and oral presentation 18.
DAPD and its active metabolite DXG constitute a guanosine NRTI drug in Phase I/II clinical studies. Results from those studies had been presented at previous conferences or workshops in the past year. At CROI, Laurene H. Wang, Ph.D., of Triangle presented pharmacokinetic results of DAPD in 22 treatment-naive subjects (32% women) and 19 treatment-experienced subjects (5% women). Naive participants received twice-daily dosing of 100-500mg, while experienced participants received twice-daily dosing of 200-500mg, all for 14 days as monotherapy. The results showed extensive conversion of DAPD to DXG, with high plasma concentrations of the latter that were dose-dependent. However, there was wide variability among subjects, as has been reported for most anti-HIV drugs. The active component of DXG was well above the plasma IC50 (concentration inhibiting 50% of HIV in the laboratory) throughout the dosing interval for all doses. It may have been preferable to see active drug concentrations reported within immune cells, since the triple phosphate components of NRTI drugs within cells correlate best with anti-HIV efficacy.
Joy Y. Feng, Ph.D., of Triangle presented data indicating that the active triple phosphate form of DXG (DXG-TP) displays activity against selected HIV isolates that are resistant to AZT (Retrovir) or 3TC. Also reported was the fact that DXG-TP is poorly incorporated by human mitochondrial DNA, suggesting a lower likelihood of mitochondrial toxicity. Many of the currently approved NRTI drugs have toxicity that is associated with mitochondrial DNA abnormalities.
Feng J and others. Mechanistic studies of dioxolane guanosine 5'-triphosphate: implications for efficacy, lack of cross-resistance and selectivity of DAPD. Abstract and poster 306.
Wang LH and others. The disposition of DAPD and its active metabolite DXG in therapy-naive and -experienced HIV-infected subjects. Abstract and poster 752.
BCH-10618 is an experimental, preclinical NRTI under development by BioChem Pharma. Dr. Z. Gu reported that BCH-10618 "selected mutant [HIV] viruses that had a low level of resistance to [the drug] and the currently approved antiretrovirals." The so-called E-nucleoside drugs were found in laboratory testing to be quite potent against several multidrug-resistant HIV isolates, according to Dr. E. Kodama of Kyoto University in Japan.
Gu Z and others. In vitro selection and characterization of HIV-1 resistance to BCH-10618. Abstract 472.
Kodama E and others. 4'-ethynyl nucleoside analogues: potent inhibitors active against multi-drug resistant HIV variants in vitro. Abstract 305.
Tenofovir DF (TDF, PMPA)
Tenofovir DF resistance results after 48 weeks from Study 902 were presented by Michael Miller, Ph.D., of Gilead Sciences. TDF is a nucleotide reverse transcriptase inhibitor (NtRTI) that also has activity against HBV. A total of 189 highly treatment-experienced subjects with a mean 4.5-year duration of previous anti-HIV therapy added tenofovir or placebo to their existing regimen. With an overall mean baseline CD4 cell count of 375 cells/mm3 and a viral load of 4.7 log (5,011) copies/mL, those that added 300mg of TDF had a -0.6 log (four-fold) reduction in HIV RNA at weeks 24 and 48. Those with the signature 3TC mutation (M184V) at baseline without AZT mutations had a -0.8 log (six-fold) reduction in viral load. Other baseline NRTI mutations had little or no effect on viral response. Only 2% developed genotypic resistance to tenofovir (K65R mutation), which could have been due to other NRTI drugs and did not affect viral load reduction. Gilead opened its expanded access program for TDF just prior to CROI; physicians may call 1-800-GILEAD-5 for more information.
Miller M and others. Baseline and week 48 final phenotypic analysis of HIV-1 from patients adding tenofovir disoproxil fumarate (TDF) therapy to background ART. Abstract and poster 441.
Interleukin-2 (IL-2, Proleukin)
IL-2, also called T cell growth factor, has been shown to increase the number of T (CD4) lymphocytes in several smaller studies. Anti-HIV therapy that decreases the viral load also usually leads to increases in CD4 cell counts. Whether optimal therapy would include HAART plus IL-2 is currently under investigation. A few reports at CROI advanced current knowledge about IL-2.
The final, 84-week results of a Phase II, randomized study of adding IL-2 to HAART were presented by Ronald Mitsuyasu, M.D., of the University of California at Los Angeles (UCLA). A total of 204 subjects (9% women, 44% non-European-American) participated in the ACTG 328 study. Participants had never taken PI drug therapy and were started on HAART with indinavir plus two NRTI drugs. The required CD4 cell count range for study entry was 50-350 cells/mm3. For those who achieved a viral load below 5,000 copies/mL at 12 weeks, randomization to one of three arms took place. Group 1 continued taking HAART alone, Group 2 took HAART plus IL-2 by SC injection (subcutaneously, or under the skin) for five-day cycles every eight weeks as 7.5 MIU (million international units) twice daily, and Group 3 took HAART plus IL-2 by CIV (continuous intravenous injection) for five days every eight weeks as 9 MIU daily. Those in the CIV arm were allowed to switch to SC IL-2 for specified CD4 cell count increases. IL-2 dose reduction was allowed for toxicity. The median CD4 cell count at study entry was 199 cells /mm3, with a viral load of 4.5 log (31,622) copies/mL.
At 12 weeks, 85% of subjects met the criteria for randomization, with a median CD4 cell count of 249 cells/mm3 and a viral load of 1.7 log (50) copies/mL. The two IL-2 arms achieved significantly greater increases in CD4 cell counts than the arm randomized to HAART alone. Using a strict intent-to-treat analysis, the median CD4 cell count increase after 84 weeks in cells/mm3 was 121 (HAART alone), 293 (HAART plus SC IL-2), and 480 (HAART plus CIV IL-2). In the same three arms respectively, the percentages that achieved greater than a 50% increase in CD4 cell count over baseline were 41%, 73%, and 86%. There was a trend towards significantly increased naive CD4 cells (which are able to respond to new infections) in the IL-2 arms. After 84 weeks, the same three arms had similar percentages of subjects who maintained viral undetectability (limit 50 copies/mL): 84%, 83%, and 71%, respectively. Even though the study was not statistically powered to detect such differences, the HAART-alone arm had five AIDS events while the other two arms had none. Interestingly, at 52 weeks, the SC IL-2 arm had a significantly greater improvement in quality-of-life (QOL) measurements than either of the other two arms, which had decreases in QOL.
For those randomized to IL-2, 65% of the SC arm and 72% of the CIV arm completed at least six cycles. Eleven percent of the SC and 15% of the CIV arm did not complete three cycles. As a result of meeting required CD4 cell count increases, 76% of the CIV arm switched to SC. As a result of toxicity and dose reduction, the mean daily dose of IL-2 was 8.6 MIU in the SC arm and 8.2 in the CIV arm. Toxicities were similar to those reported for study drugs in the past, including moderate to severe influenza (flu)-like symptoms in the IL-2 arms. Grade 3 or 4 adverse events occurred in 10% of the HAART-only arm, 50% of the HAART plus SC IL-2 arm, and 63% of the HAART plus CIV IL-2 arm.
This study is the largest randomized trial of HAART plus IL-2 to date in moderately advanced subjects that demonstrates some benefits. While the toxicities of IL-2 cannot be discounted, the majority of participants completed six cycles of this outpatient (not requiring hospitalization) therapy. The true overall benefits of IL-2 will be known after the very large SILCAAT (baseline CD4 cell count below 300 cells/mm3) and ESPRIT (baseline CD4 cell count of at least 300 cells/mm3) trials are completed. For more information about ESPRIT, call 1-800-772-5464 ext. 58008.
In a meta-analysis (combined analysis) of three Vanguard studies of IL-2 added to antiretroviral therapy, Dr. R.C. Arduino of the University of Texas in Houston evaluated the best dose of SC IL-2. A total of 218 subjects with a baseline CD4 cell count of at least 350 cells/mm3 participated in the studies that had SC IL-2 added to their anti-HIV regimen. The IL-2 doses were 1.5 MIU, 4.5 MIU, or 7.5 MIU twice daily. The results "support[ed the] initial use of [the] 7.5 MIU dose" twice daily, with a "dose-dependent CD4 cell count increase" that also was associated with the baseline CD4 cell count and BMI. Starting at the 7.5 MIU dose decreased the number of cycles needed and increased the proportion of subjects who achieved a CD4 cell count increase at least twice the baseline level or to at least 1,000 cells/mm3. However, the 7.5 MIU dose increased the percentage of subjects who needed a dose reduction or who did not complete three cycles. Note that one-third of participants were from Thailand and included 67% women; almost all of them took only NRTI combination anti-HIV therapy, not HAART.
In an attempt to decrease toxicities from IL-2, Dr. J.A. Tavel and colleagues from the National Institutes of Health (NIH) tested adding prednisone (an immune blocker, 0.5mg/kg divided into twice-daily dosing) to the regimen. Unfortunately, the results showed that while adding prednisone to IL-2 significantly decreased the number of grade 3-4 toxicities, the CD4 cell increase was significantly blunted. The authors concluded that symptom management and dose reduction of IL-2 represent the best ways to address toxicities, while still maintaining CD4 cell count increases.
Researchers from the NIH also presented an interesting meta-analysis (combining results of three studies) of long-term follow-up of participants given anti-HIV therapy plus IL-2 to maintain CD4 cell counts. Required baseline CD4 cell counts were at least 200 cells/mm3 for 45%, and at least 500 cells/mm3 for the remaining 55%. Of 97 subjects initially enrolled in the IL-2 dosing studies, 77 of them (1% women) entered an extension phase. The mean duration of follow-up was 4.1 years. The mean baseline CD4 cell count was 543 cells/mm3, which had increased to a mean of 1,132 cells/mm3. The mean number of IL-2 cycles (five days every eight weeks) to maintain the CD4 cell counts was 10 (a range of 3-28). During the most recent cycle, the mean tolerated total daily dose of IL-2 was 11.6 MIU. Interestingly, the mean interval since the last cycle of IL-2 was 26 months (a range of 2-60). The mean HIV RNA viral load did not increase over time and tended towards continued small decreases over time. Only one AIDS illness developed: non-Hodgkin's lymphoma in one subject. The authors concluded that the "remarkably low frequency of intermittent SC IL-2 therapy" cycles needed to maintain CD4 cell counts "may contribute to patients' acceptance of SC IL-2 as a favorable long-term strategy." The lead author was Doreen Chaitt, R.N., M.P.H.
Arduino RC and others. Meta-analysis of the CD4 cell response to 3 doses of subcutaneous interleukin-2 (scIL-2) across 3 Vanguard studies. Abstract and poster 346.
Chaitt D and others. Extended therapy with subcutaneous interleukin-2 (scIL-2) in HIV infection: long-term follow-up of 3 trials. Abstract and poster 347.
Mitsuyasu R and others. Prospective, randomized, controlled phase II study of highly active antiretroviral therapy (HAART) with continuous IV (CIV) or subcutaneous (SC) interleukin-2 (IL-2) in HIV-infected patients with CD4+ counts of 50-350 cells/mm3: ACTG 328 final results at 84 weeks. Abstract and oral presentation 17.
Tavel JA and others. Prednisone decreases rIL-2 related toxicities but also blunts the rIL-2-related CD4+ cell response in HIV+ patients. Abstract and poster 348.
Several presentations at CROI focused on once-daily regimens, which may be associated with easier adherence for many people. An update of ANRS 091 (the Montana Study) was presented by Dr. J.M. Molina of Hôpital St-Louis in Paris. In that study, the seven-pill (or capsule), once-daily regimen of FTC 200mg, efavirenz, and ddI was taken by 40 treatment-naive subjects (12% women). The median baseline CD4 cell count was 373 cells/mm3, with a viral load of approximately 4.8 log (58,400) copies/mL. Using a strict intent-to-treat analysis, the results showed that after 64 weeks, 90% had an undetectable viral load (limit 400 copies/mL) with a median CD4 cell count increase of 219 cells/mm3. Viral load results with an ultrasensitive test were not reported. Only one participant (2.5%) experienced virologic failure, or viral breakthrough. Two participants (5%) dropped out of the study due to adverse events related to study drugs (gastrointestinal symptoms from ddI or "vertigo" [spinning sensation] from efavirenz). Another 5% were lost to follow-up or decided to stop participation. The rate of grade 3-4 adverse events in the first 24 weeks was 15%, yet only half of those were related to study drugs (increases in liver enzymes, muscle enzyme CPK, or triglycerides). The results indicate moderate potency and tolerance of the once-daily regimen of FTC, efavirenz, and ddI.
The final 48-week results of a once-daily regimen of ddI (300mg), efavirenz, and 3TC (300mg) were presented by Dr. F. Maggiolo of Bergamo General Hospital in Italy. (Note that studies have shown that the active triphosphate component of 3TC in immune cells reveal a half-life that would allow for once-daily dosing of that drug.) A total of 75 treatment-naive subjects (21% women) participated, with a median baseline CD4 cell count of 251 cells/mm3 and a viral load of 5.0 log (123,000) copies/mL. HCV antibodies were present in 41%. Using a strict intent-to-treat analysis, results after 48 weeks showed that 78% had an undetectable viral load (limit 50 copies/mL), with nearly identical results for those with a baseline viral load of at least 100,000 copies/mL. For those with a baseline CD4 cell count greater than 200 cells/mm3, the undetectability rate was 88%, while those with a baseline level less than that had a 66% rate of undetectability. The median CD4 cell count increase was approximately 200 cells/mm3. Discontinuation rates were reported as being due to known adverse effects of study drugs (9%), virologic failure (5%), and nonadherence (4%). The authors concluded that the once-daily regimen was successful with easier adherence for subjects.
A once-daily regimen of amprenavir 1,200mg, ritonavir 200mg, plus standard (twice-daily) abacavir and 3TC led to 100% viral undetectability (limit 400 copies/mL) in a strict intent-to-treat analysis of 15 subjects (67% women, 73% non-Caucasian) after 12 weeks, with a CD4 cell count increase of 123 cells/mm3. For the 11 subjects who completed 20 weeks (in an as-treated analysis), the undetectability rate had fallen to 91%. Previously, subjects had been taking standard dosing of amprenavir with abacavir and 3TC for up to 24 weeks, with more than 85% achieving an undetectable viral load before switching to the once-daily double-PI drug regimen. Prior to that, subjects were treatment-naive, with a median baseline CD4 cell count of 182 cells/mm3 and a viral load of 4.9 log (70,794) copies/mL. Drug concentration studies showed that when compared with standard amprenavir dosing without ritonavir, the once-daily double-PI regimen led to a similar maximal concentration of amprenavir, a five-fold increase in the minimum concentration, and a nearly four-fold increase in the total drug exposure. A few participants experienced new-onset nausea, vomiting, and/or mouth tingling. One subject developed a grade 4 (life-threatening) increase in the liver enzyme GGT, while a few developed grade 2 (moderate) increases in liver enzymes or cholesterol. Since one other report at CROI indicated that once-daily dosing of abacavir might also be possible, this four-drug regimen might be possible to take entirely as once daily. The lead author was Dr. R. Wood of Somerset Hospital in Cape Town, South Africa.
A similar study (Merck 103/104) indicated that once-daily indinavir 1,200mg plus ritonavir 200mg combined with standard twice-daily dosing of d4T and 3TC might be possible. A total of 40 treatment-naive subjects (47% women) were enrolled with a median baseline CD4 cell count of 329 cells/mm3 and a viral load of 4.9 log (81,283) copies/mL. Using a strict intent-to-treat analysis, after 24 weeks, 72% achieved an undetectable viral load (limit 400 copies/mL), with 54% using an ultrasensitive test (limit 50 copies/mL). The CD4 cell count increase was approximately 120 cells/mm3. The discontinuation rate was 20%, but no subjects left the study due to drug-related adverse events. There was one possible kidney stone episode. Grade 2-3 adverse events related to study drugs occurred among 15%, including increases in total cholesterol and triglycerides. Previously, drug concentration studies have shown very favorable results with this ritonavir-boosted once-daily indinavir dosing. Note that since Bristol-Myers is developing a once-daily d4T "extended-release" formulation (100mg), this entire four-drug regimen might also be able to be dosed once daily in the future. The lead author was Dr. Jamal Suleiman of São Paolo, Brazil.
The safety and efficacy of switching to once-daily 3TC 300mg from standard twice-daily dosing of 150mg in combination was reported by Dr. M. Sension of North Broward Hospital District in Ft. Lauderdale, FL. Final 24-week results of the COLA4005 study included 81 subjects (13% women, 47% non-European-American). All participants had an undetectable viral load at baseline (limit 50 copies/mL) with a median baseline CD4 cell count of 514-532 cells/mm3. Study participants had been taking a stable (unchanged) HAART regimen for at least six months prior to study entry, with an undetectable viral load for at least three months. The remaining drugs in the regimen at baseline (other than 3TC) were d4T and either nelfinavir or indinavir. Study subjects were randomized to switch their 3TC to once daily or to maintain their standard twice-daily dosing. After 24 weeks, using a strict intent-to-treat analysis, 81-82% of each arm maintained an undetectable viral load (limit 50 copies/mL). The median increase in CD4 cell counts was 22 cells/mm3 in the twice-daily 3TC arm and 42 cells/mm3 in the once-daily arm. However, drug-related adverse events were slightly more frequent in the switch arm (28%) than in the no-switch arm (12%), with most of the increase in the gastrointestinal tract. No serious adverse events were due to study drugs; most were mild or moderate, and there were no discontinuations due to drug-related adverse events. The results provide additional evidence for the efficacy and safety of once-daily 3TC, in the combinations used.
Gulick RM. New antiretroviral drugs. Abstract and oral presentation S25 at session 95.
Harris M and others. Intracellular carbovir triphosphate levels among patients taking abacavir once daily. Abstract 746.
Maggiolo F and others. Once-a-day treatment for HIV infection: final 48-week results. Abstract and poster 320.
Molina JM and others. Once-daily combination therapy with emtricitabine, didanosine and efavirenz in treatment-naive HIV-infected adults: 64-week follow-up of the ANRS 091 trial. Abstract and poster 321.
Sension M and others. Efficacy and safety of switch to 3TC 300mg QD vs. continued 3TC 150mg BID in subjects with virologic suppression on stable 3TC/d4T/PI therapy (COLA4005): final 24-week results. Abstract and poster 317.
Suleiman J and others. Preliminary results from indinavir and ritonavir in a once-daily regimen (Merck 103/104). Abstract and poster 336.
Wood R and others. Amprenavir (APV) 600mg/ ritonavir (RTV) 100mg BID or APV 1200mg/RTV 200mg QD given in combination with abacavir and lamivudine maintains efficacy in ART-naive HIV-1 infected adults over 12 weeks (APV20001). Abstract and poster 332.
Liver and Kidney Transplants for Persons with or without HCV, HBV Coinfection
Two presentations focused on organ transplantation in persons with HIV. In the past, liver or kidney transplants were considered too risky for people with HIV infection due to the requirement for immune suppressive therapy to prevent graft rejection. Another consideration has been the overall shortage of available organs. Yet HAART use in the developed world has led to relatively stable HIV disease for many people, which in turn has led to liver complications for HIV positive persons who are coinfected with HCV or HBV.
Dr. A.E. Boyd of King's College Hospital in London presented a case series of eight HIV positive persons, each of whom had a liver transplant. Among the four who had chronic hepatitis C (25% women, 100% Caucasian), death occurred at 3-25 months after transplantation, including in two subjects who were treated post-transplant with interferon alpha plus ribavirin; three of four maintained anti-HIV therapy post-transplant. The three (including one Black woman) who had liver failure from HBV or chronic hepatitis B were still alive 1-35 months post-transplant, had maintained HAART, and took hepatitis B immune globulin post-transplant. The eighth subject was a Black man with "non-A, non-B fulminant hepatic failure" who was still alive 15 months after transplant, had taken hepatitis B immune globulin, and continued his anti-HIV therapy. All eight subjects received standard immune suppressive therapy for transplants including tacrolimus and prednisone; one also received cyclosporine and azathioprine. For the three subjects still alive 5-15 months after transplantation, HIV viral load was very low or undetectable (limit 50 copies/mL) with stable CD4 cell counts (244-552 cells/mm3). Dr. Boyd concluded, "HIV infection, per se, should not be a contraindication to 'orthotopic' liver transplantation. . . ."
Michelle Roland, M.D., of the University of California at San Francisco (UCSF) presented six case reports of HIV positive persons who had solid organ transplantation at UCSF, including two liver transplants and six kidney transplants (one person received two sequential liver transplants and a kidney transplant). The Latino teenaged boy who received two livers and a kidney was still alive ten months after the first transplant, had chronic hepatitis C, and required HAART interruptions post-transplant. The other five HIV positive persons each had a kidney transplant (20% women, 60% non-European-American) due to HIV-associated kidney disease, high blood pressure, and/or diabetes (note that one did have chronic hepatitis C). All five were alive from 1-10 (mean 6) months after transplant, with three episodes of transplant rejection in two of them. All five took cyclosporine immune suppression therapy and continued HAART post-transplantation (except one person who stopped HAART during a period of altered kidney function). In all cases, Dr. Roland reported that there has been "no significant HIV clinical, virological, or immunologic disease progression." More solid organ transplants among persons with HIV infection will take place at UCSF. For more information, visit http://hivinsite.ucsf.edu or call Laurie Carlson at 415-502-8322 (e-mail: CarlsonL@surgery.ucsf.edu). Note that there are two reports elsewhere in the medical literature of HIV positive persons who underwent a kidney or liver transplant who survived with normal organ function up to eight years.
Ahuja TS and others. Long-term survival in an HIV-infected renal transplant recipient. American Journal of Nephrology 1997;17:480-482.
Boyd AE and others. Liver transplantation and HIV -- a case series of 7 patients. Abstract and poster 578.
Roland M and others. Solid organ transplantation in HIV disease. Abstract and poster 579.
Other Data on Hepatitis Virus Coinfection
Neither chronic hepatitis B nor C significantly increased the risk of severe liver toxicity (grade 3-4) among HIV coinfected persons who received either nevirapine- or efavirenz-containing HAART, according to Mark Sulkowski, M.D., of Johns Hopkins University. While the overall incidence (rate of new cases) of severe liver toxicity was 17% among persons taking nevirapine and 8% among persons taking efavirenz, the incidence per 100 patient-months was nearly identical for the two drugs: 1.3 for nevirapine and 1.2 for efavirenz. (Patient-months are determined by multiplying the number of persons by the number of months. For example, one person followed for ten months equals ten patient-months, and ten persons followed for one month also equals ten patient-months.) The large Johns Hopkins database of HIV positive persons was used for the study in which 203 people started nevirapine-based HAART and 97 started efavirenz-based HAART. HCV coinfection was present in 49% of those taking nevirapine and 44% of those taking efavirenz. A significantly higher percentage of persons taking nevirapine had chronic hepatitis B (9%, by positive surface antigen test) compared with those taking efavirenz (2%). Dr. Sulkowski concluded, "NNRTI therapy should not be withheld from patients with chronic viral hepatitis."
HCV coinfection with HIV appears to be associated with a blunted CD4 cell response to HAART, according to Dr. Sulkowski. However, HCV coinfection did "not appear to adversely affect HIV disease progression or survival after controlling for the use and effectiveness of HAART." Forty-five percent of 1,742 HIV positive subjects (30% women, 76% African-American) were coinfected with HCV. The topic of HCV infection and HIV-HCV coinfection at CROI was of great interest to many attendees. A State-of-the-Art Lecture on HCV Coinfection and Summary was presented by Kenneth Sherman, M.D., Ph.D., of the University of Cincinnati. His lecture is available by audio (with slides) at www.retroconference.org.
In another study, in 35 HIV positive subjects (3% women) coinfected with HBV, the addition of low-dose adefovir (a nucleotide analog, at 10mg daily) to HAART with 3TC led to a significant reduction of HBV DNA after 36 weeks. Adefovir was well tolerated without kidney toxicity, and there was no HIV rebound or significant change in CD4 cell counts in persons who had previously responded to HAART, according to Yves Benhamou, M.D., of Hôpital Pitié-Salpêtrière in Paris. Prior to starting adefovir, 100% of the participants had developed HBV resistance to 3TC. HBV DNA decreased by a mean -3.8 log (6,309) copies/mL, while three subjects lost the hepatitis B "e" antigen.
HCV might be sexually transmitted by specific sexual practices among homosexual and bisexual men, according to an analysis of 662 men from the Vancouver Lymphadenopathy AIDS Study by Dr. Kevin J.P. Craib in British Columbia, Canada. Six percent of the men were HCV antibody positive, yet 49% of those denied a history of injection drug use, a major risk factor for HCV transmission. Significant behavioral risk factors included oral-anal contact and insertive fisting (finger-anal contact). However, such behaviors might have represented associations and not causal or actual means of transmission. Previous studies of homosexual and bisexual men have revealed a low (10% or less) rate of HCV infection when there is no history of injection drug use. The analysis is reminiscent of past behavioral studies about HIV transmission or those for Kaposi's sarcoma (KS), an AIDS-related cancer.
Benhamou Y and others. An open label pilot study of the safety and efficacy of adefovir dipivoxil in HIV/HBV co-infected patients with lamivudine resistant HBV. Abstract and oral presentation 36.
Craib KJB and others. Evidence of sexual transmission of hepatitis C virus in a cohort of homosexual men. Abstract and poster 561.
Sulkowski M and others. Hepatotoxicity associated with NNRTI use: role of drugs and chronic viral hepatitis. Abstract and poster 618.
Sulkowski M and others. Effect of HCV co-infection on HIV disease progression and survival in HIV-infected adults. Abstract and oral presentation 34.
Causes of Lipodystrophy (Fat Redistribution)
One of the more interesting posters about the topic of fat redistribution was authored by Nevena Christeff, M.D., of the Pasteur Institute in Paris. Dr. Christeff found that several blood markers were highly statistically associated with the clinical evolution of fat redistribution among 31 HIV positive men taking anti-HIV therapy (94% HAART) and 20 healthy controls. At baseline, 20 of the 31 HIV positive men had fat redistribution, measured by CT (computerized tomography x-ray scans) and physical examination. The other 11 of 31 men did not have fat redistribution. All of them were followed prospectively for up to 24 months. Anthropomorphic tests (waist-hip ratios and skin fold measurements) and physical examinations were carried out at regular intervals. Participants were then categorized as having lipodystrophy that improved, remained unchanged, or worsened, or (if starting without lipodystrophy) as developing or remaining without lipodystrophy. Several types of fasting blood tests were performed including lipids (fats), cytokines (intercellular messengers), and sex hormones.
Among those without lipodystrophy at baseline, two of 11 (18%) developed the condition, which was significantly associated with a decrease in DHEA (dehydroepiandrosterone, a natural steroid hormone precursor that normally decreases with age), an increase of the cortisol-to-DHEA ratio (cortisol is a natural "stress" hormone from the kidney), an increase in blood levels of interferon alpha, and an increase in blood lipids (predominantly VLDL, very low-density lipoprotein cholesterol). Those who did not have lipodystrophy at baseline and who did not develop it (9 of 11, or 82%) had a very high level of DHEA, high cortisol, a normal cortisol-to-DHEA ratio, a normal level of interferon alpha, and normal lipids. Among the 21 subjects with lipodystrophy at baseline, 30% had improvements in follow-up, which was significantly associated with a 95% increase in DHEA, a normalized cortisol-to-DHEA ratio, normalized interferon alpha, and normalized blood lipids, predominantly a 75% decrease of VLDL. Among those with baseline lipodystrophy, 25% had worsened over time, which was significantly associated with a 50% decrease in blood DHEA, a 95% increase in the cortisol-to-DHEA ratio, and persistently high levels of interferon alpha and lipids. The remaining 45% of those with baseline lipodystrophy were unchanged at follow-up; this was associated with a slight decrease in DHEA, a slight increase in the cortisol-to-DHEA ratio, a decrease in interferon alpha, and persistently elevated lipids.
These interesting interim results lead to more questions than answers. First, it must be emphasized that statistical "association" does not mean "cause." It is possible that whatever is causing the evolution of lipodystrophy might also be causing the changes in blood markers, rather than the blood markers themselves leading to lipodystrophy evolution. Second, what is causing the changes in these various blood markers over time? The authors say that no subjects were taking complementary medications that might have affected the results, including glucocorticoids or ketoconazole (an antifungal medication). There is an assumption that none were taking (or using) steroid hormones unknown to the researchers. However, in the U.S., DHEA is available without a prescription as a nutritional supplement. This reviewer has heard anecdotally that DHEA is not legally available in Europe, but it is possible that those with increased blood DHEA levels were taking it as a supplement, unknown to the researchers. Otherwise, what would have led to increased DHEA? Is it possible (although quite unlikely) that some of the anti-HIV regimens might have led to the blood marker improvements? Is it possible that some participants had started intensive stress-reduction interventions that might have lowered their cortisol levels? (The authors do not indicate whether or not this is the case.) Were some subjects taking prescription or over-the-counter substances that might have improved their blood lipid profiles? (If by prescription, the authors likely would have stated so.)
While these and other issues must be addressed before any conclusions can be drawn, the following potential implications immediately come to mind for those with HIV taking HAART: (1) would taking over-the-counter DHEA improve or prevent fat redistribution? If so, how much DHEA should be taken and how often should blood levels be measured? (note that it is advisable to talk with a physician before taking any so-called alternative or complementary medication), (2) would vigorous stress-reduction management improve or prevent fat redistribution?, and (3) would the treatment of abnormal blood lipids improve fat redistribution? Much additional research is needed that should help to answer these questions. One major limitation of the report is that the authors did not subdivide "lipodystrophy evolution" into lipoatrophy (fat loss under the skin) and fat accumulation, since there appears to be emerging information that the two phenomena are distinct with likely different or possible overlapping causes.
Christeff N and others. The clinical evolution of lipodystrophy syndrome is closely related to cortisol:DHEA ratio and serum interferon-a. Abstract and poster 659.
Causes of Increased Cholesterol and Triglycerides with PI Drugs
Researchers from the University of Basel in Switzerland have reported a genetic marker that predisposes HIV positive individuals to an increased blood cholesterol level after a PI drug is started. A study of 67 persons with HIV and 2,727 control subjects led to the discovery of a genetic variation in SREBP (sterol-regulatory element-binding protein)-1c. Those carriers of genotype 11/12 had significantly increased blood cholesterol after starting a PI drug regimen, while those with homozygous genotype 22 did not have a significant increase. Dr. André R. Miserez and colleagues concluded, "We report the first polymorphism [gene variant] of pharmacogenetic relevance in a gene encoding a key regulator of cholesterol . . . [and that] testing prior to antiretroviral treatment may predict PI-associated hyperlipoproteinemia." There are also implications for potential new treatments of increased cholesterol among both HIV positive and HIV negative persons. Moreover, the new genetic marker might have some predictive implications for HIV negative persons prone to increases in cholesterol that, in turn, increase the risk of artery disease (heart attacks and strokes).
In a related presentation by Dr. T.M. Riddle of the University of Cincinnati, ritonavir-treated mice showed a significant increase in blood triglycerides and VLDL cholesterol that was associated with increased activated SREBP in cell nuclei (centers). The authors concluded that ritonavir (in mice) "may interfere with the . . . degradation [breakdown] of activated nuclear SREBP-1, which may then explain, in part, the development of hyperlipidemia and lipodystrophy in patients treated with HIV protease inhibitor drugs."
Miserez AR and others. Hyperlipoproteinemia in HIV patients is linked to sterol-regulatory element-binding protein (SREBP)-1c. Abstract 500.
Riddle TM and others. HIV protease inhibitor therapy increases hepatic lipoprotein production via stabilization of activated sterol regulatory element-binding protein-1 (SREBP-1) in the nucleus. Abstract 659.
Other Data on Adverse Effects of Anti-HIV Therapy
The importance of including an HIV negative, age-matched control group when evaluating the presence of fat redistribution was highlighted by a presentation by Dr. L. Kingsley of the University of Pittsburgh. At the 31st follow-up visit of the Multicenter AIDS Cohort Study (MACS) with 868 men (62% HIV positive), waistline fat accumulation ("paunch") was present in approximately 26% of HIV negative and 39% of HIV positive men, while breast enlargement was present in approximately 8% of HIV negative and 11% of HIV positive men. However, peripheral (arms and legs) fat wasting, or lipoatrophy, was more specific to HIV infection with treatment. In addition, the prevalence (cumulative rate) of lipodystrophy syndrome appeared to plateau after two years of HAART, without additional increases thereafter.
New additions to the black box (emphasized drug-label) warnings for d4T and ddI have been made by the FDA to reflect an increased risk for fatal hepatotoxicity (liver toxicity) when either or both are combined with hydroxyurea (Hydrea, an anticancer, anti-sickle cell drug). The FDA had received reports about 34 cases of liver toxicity associated with the use of any NRTI drug plus hydroxyurea. Dr. D. Boxwell of the FDA concluded, "Patients taking hydroxyurea and NRTI [drugs], in particular ddI/d4T, should be aggressively monitored for hepatotoxicity." The FDA encourages reporting of similar cases through the MedWatch program (1-800-FDA-1088).
Several presentations at CROI indicated that HIV itself may be a cause of osteopenia/osteoporosis (loss of bone mineral density) without anti-HIV treatment, in adults and children. However, HAART might worsen the loss. Osteoporosis (severe loss of bone mineral density) represents a risk for spontaneous bone fracture that can be painful and debilitating; however, there have been scant reports of this having occurred to date as a result of HIV. In addition, there were two reports about persons with HIV (adults and children) being at significantly increased risk for osteonecrosis (also known as avascular necrosis) or areas of spontaneous bone death, due to inadequate blood supply. These bone conditions may in future be known as part of the natural history of HIV disease rather than as adverse effects of anti-HIV drugs.
Despite much ongoing research about lactic acidemia (high lactic acid in blood) in HIV positive persons taking anti-HIV therapy and the potential association with several types of adverse events, "routine lactate measurements [in persons with HIV are] not justified at this time," according to Andrew Carr, M.D., of St. Vincent's Hospital in Sydney. Dr. Carr presented a summary of lactic acidemia. Similarly, Dr. S.M.E. Vrouenraets of Academic Medical Center in the Netherlands concluded in his report that "the value of lactate measurement for individual treatment monitoring remains obscure . . . [in part because] elevated lactate levels were not consistent in the same individuals."
Dr. Carr did report that in a statistical multivariate analysis of 221 HIV positive persons, lactic acidemia was significantly associated with osteopenia. However, Dr. S. Claxton and colleagues from Washington University in St. Louis found no such association in their 30 subjects, underscoring controversy and unknown factors in this area and possibly different types of tests used.
Gemfibrozil (Lopid) therapy did lead to very mild decreases of high triglyceride levels in 38 HIV positive subjects taking PI drug-based HAART, but not sufficiently into a normal range, according to Dr. J. Miller of the University of New South Wales in Sydney. Dr. Miller concluded that other lipid-lowering agents need to be investigated.
A similar rate of heart disease hospitalizations was seen among 4,541 HIV positive men both with and without exposure to PI drugs (5.8 and 5.2 events per 1,000 patient-years, respectively). However the overall rate among them (5.5) was higher than that for age-matched, HIV negative control men (3.4). The HIV positive men had a total follow-up of 14,703 patient-years and were derived from the Kaiser Northern California database. Rates of smoking and diabetes were similar for the HIV positive and negative men. Yet high blood pressure was significantly more common among the HIV negative men (21% vs 13% in HIV positive men), while high blood cholesterol tended towards being more common among those with (21%) rather than those without (15%) HIV infection. The lead author, Daniel Klein, M.D., of Oakland, CA, postulated that "HIV . . . may be implicated in the pathogenesis [cause of] coronary heart disease." However, PI drug exposure was "significantly associated with [the] occurrence of MI [myocardial infarction, or heart attack] . . . [and a] dose-effect relationship with higher MI incidence [rate] in patients exposed [for] 18 months or more," according to Dr. M. Mary-Krause of INSERM in Paris. A total of 42,787 HIV positive persons were included in that analysis.
Boxwell D and Toerner J. Fatal hepatotoxicity associated with combination hydroxyurea and nucleoside reverse transcriptase inhibitors: cases from the FDA adverse event reporting system. Abstract and poster 617.
Carr A. State-of-the-Art summary and discussion issues in metabolic complications: controversy or consensus. Session 64 (no abstract).
Carr A and others. Lactic acidemia is associated with spinal osteopenia in HIV-infected men. Abstract and poster 631.
Claxton S. Circulating leptin and lactate levels are not associated with osteopenia in HIV-infected men. Abstract and poster 634.
Keruly JC and others. Increasing incidence of avascular necrosis of the hip in HIV-infected patients. Abstract 637.
Klein D and others. Do protease inhibitors increase the risk for coronary heart disease among HIV positive patients? -- follow-up through June 2000. Abstract and poster 655.
Kingsley L and others. Prevalence of lipodystrophy and metabolic abnormalities in the multicenter AIDS cohort study (MACS). Abstract and oral presentation 539.
Knobel H and others. Osteopenia in HIV-infected patients. Is it the disease or is it the treatment? Abstract and poster 629.
Mary-Krause M and others. Impact of treatment with protease inhibitor on myocardial infarction occurrence in HIV-infected men. Abstract and poster 657.
McGowan I and others. Assessment of bone mineral density in HIV-infected antiretroviral treatment naive patients. Abstract and poster 628.
Miller J and others. A randomized, double-blind study of gemfibrozil for the treatment of protease inhibitor-associated hypertriglyceridemia. Abstract and oral presentation 540.
Vrouenraets SME and others. Hyperlactatemia in HIV-infected patients: the role of NRTI treatment. Abstract and poster 625.
Genetic Risk for Low Blood Levels of Anti-HIV Drugs
The benefits of anti-HIV therapy require minimum blood concentrations of the drugs. Drug resistance has also been linked with low blood levels of anti-HIV drugs. Conversely, many toxicities of these drugs are related to concentrations that are too high. Yet several studies have shown a wide variation between adults who take standard doses of anti-HIV medications, even after correcting for body weight. As reported at CROI, Dr. J. Fellay of CHUV in Lausanne, Switzerland, has found two potential explanations for drug levels that are either high or low: genetic differences in the p-glycoprotein transporter molecule and/or the P450 CYP2D6 isoform in the liver. P-glycoprotein, which occurs naturally, is partly responsible for "pumping out" unwanted drugs from the inside of cells (and is therefore associated with resistance). The P450 enzyme system in the liver is partly responsible for the normal metabolism (breakdown) of many drugs, including the PI and NNRTI drugs.
A total of 63 Caucasian persons with HIV were tested who had an undetectable viral load (limit 400 copies/mL) while taking HAART with either a PI drug(s) or efavirenz. Among them, 43% had drug levels (of PIs or efavirenz) that were classified as high, while the remaining 57% had levels that were low. The results were analyzed in a multivariate regression that adjusted for age and sex. For the low drug level group, there was an eight-fold increased risk (or odds ratio) of having a homozygous (double dose) TT allele (gene marker) in p-glycoprotein and/or a UM (ultrarapid metabolizer) genotype in CYP2D6. For the high drug level group, there was a two-fold increased risk of having a homozygous CC allele in p-glycoprotein and/or a PM (poor metabolizer) genotype in CYP2D6.
While these results need to be confirmed in larger numbers of persons, they provide a newer understanding as to why drug levels vary so much between people who are adherent with dosing. The virologic significance in the current study was not readily apparent, since the entire group had undetectable virus at baseline. Perhaps those persons with lower drug levels might be more prone to virologic rebound in the future. The findings would have greater significance if applied to larger groups of persons who never achieve undetectability, experience viral rebound, and/or experience higher rates of drug toxicities. Nonetheless, it is possible to envision a time in the future when a baseline genetic profile might be obtained to determine who is more likely to have low blood levels of drug -- even with perfect adherence -- which might require higher initial dosing or perhaps frequent initial drug level measuring in the first days after starting (and if necessary, increasing drug doses). Similarly, a different genetic profile that might increase the risk of high drug levels associated with toxicity might suggest frequent blood level monitoring in the days after starting a drug, leading to possible dose reductions or a more risky approach of lower initial doses even before drug level testing. Also, it is likely that other genetic markers will, in part, account for varying drug levels in blood.
Therefore, additional research is needed to find these other markers, before a composite genetic profile could be known and routinely ordered before anti-HIV therapy is started. Also, it is quite possible that such markers would differ by race/ethnicity, further underscoring the need for additional research in this area.
Fellay J and others. Predictive power of p-glycoprotein and CYP2D6 polymorphism for plasma levels of antiretroviral agents. Abstract 260.
Garlic Supplements and Saquinavir (Fortovase)
Over-the-counter garlic capsule supplements taken twice daily significantly decreased the blood levels of saquinavir by approximately one-half in healthy volunteers, according to Stephen Piscitelli, Pharm.D, of Virco. Trough (lowest), peak (maximal), and total exposure drug concentrations were decreased by 49-54%. Even after a ten-day "washout" period (no garlic), saquinavir levels still had not returned to normal. The results suggest that people taking HAART should avoid garlic supplements. Whether dietary garlic would have similar effects is unknown.
Piscitelli SC and others. Garlic supplements decrease plasma saquinavir concentrations. Abstract 743.
Combination oral contraceptives (birth control pills) may increase the risk of HIV sexual transmission through the cervix due to significantly increased expression of the CCR5 coreceptor for HIV on cervical T (CD4) cells, according to Dr. Manyu Prakash of Imperial College in London. Also, the CXCR4 coreceptor for HIV was expressed more than CCR5 on cervical T cells. The percentage of cervical T cells expressing activation markers also was higher than that in blood. Dr. Prakash concluded that these findings suggest "one possible mechanism that may increase transmission rates . . . in women using oral contraceptives."
Prakash M and others. T lymphocytes in the cervix of normal women show high levels of activation markers and increased levels of CCR5 receptor expression in women taking the combined oral contraceptive. Abstract and poster 76.
The major focus of this year's CROI was on clinical and basic science issues. However, the program committee chose to include two speakers for plenary sessions that addressed access to treatments and advocacy for HIV positive persons in resource-poor nations. One of them was Jeffrey D. Sachs, Ph.D., of the Harvard University Center for International Development, who gave a similar presentation at the XIII International AIDS Conference in Durban, South Africa, this past summer. Dr. Sachs presented a rather convincing argument that the total amount of money that has been donated by Western (i.e., Northern) countries to the HIV/AIDS pandemic in sub-Saharan Africa is an exceedingly small percentage of their budgets and that increased funds are needed to finance anti-HIV treatments there. His estimate of $24 million annually to treat with HAART those HIV positive persons in sub-Saharan Africa would constitute less than 0.05% of the gross national product of the most highly developed, high-income countries, such as the U.S and Japan. Unfortunately, Dr. Sachs did not specifically address the issues of health-care access, lack of infrastructure, and certain forms of corruption that are prevalent in many of the countries in need.
A second plenary speaker on the final day was Anne-Valerie Kaninda, M.D., M.P.H., of Médecins Sans Frontières (Doctors Without Borders). She announced that Cipla, Ltd. of India would charge $350 for a one-year supply of triple-drug HAART per HIV positive person that would include generic nevirapine, d4T, and 3TC. The same regimen for governments of resource-poor countries would cost $650, while the general wholesale cost would be $1,250. Since then, several of the major pharmaceutical patent holders have announced steeply discounted prices of their anti-HIV drugs for sub-Saharan African countries. Unfortunately, even at a very low rate of approximately $1 per day for HAART, this is still out of reach for the majority who need therapy in these countries. And the cost of drugs only does not address costs of diagnostic testing, counseling, paraprofessionals, access to health care, and other critical factors (e.g., nutrition and access to clean running water).
Abate T. AIDS Drugs in the Third World: Drugmakers yield to pressure. San Francisco Chronicle, March 25, 2001, page A15.
Kaninda A-V. The access challenge: AIDS treatment in resource-limited settings. Abstract and oral presentation S26.
Sachs JD. From talk to action in fighting AIDS in developing countries. Keynote Lecture and abstract L3.
Harvey S. Bartnof, M.D., is a staff physician at the AIDS Virus Education and Research Institute (AVERI) in San Francisco, and the former Medical Editor of HIVandHepatitis.com.
This article was provided by San Francisco AIDS Foundation. It is a part of the publication Bulletin of Experimental Treatments for AIDS. Visit San Francisco AIDS Foundation's Web site to find out more about their activities, publications and services.