Revised PMPA Update: PMPA 48-Week Phase II Safety and Efficacy Results; Preliminary Data Presented at ICAR in Baltimore April 16, 2000

Recently, Gilead Sciences held a meeting for about 20 US community advocates and several from Europe to discuss PMPA data and development plans. They discussed the data reported by Melanie Thompson, MD, Director of the AIDS Research Consortium of Atlanta, at the International Conference for Antiviral Research (ICAR). In this study PMPA (tenofovir DF) was added as intensification to a stable regimen in 189

treatment-experienced HIV patients. First, a couple of caveats. The viral load drop at week 2 was about -0.75 log. for patients randomized to the 300 mg. At 4 weeks into the study participants were permitted to change the regimen they were on when they started the study. Preliminarily, it appears as though 36% of patients in placebo arm had at least one change in background therapy between week 4 and 24; 40% in the 75 mg arm; in the 150 mg arm it was 33%; 35% in the 300mg arm. Analysis is ongoing, and it's too soon to know what the actual changes in therapy were.

48 weeks of treatment with the highest dose of tenofovir DF (300 mg once daily) in 41 (54 received drug) anti-retroviral experienced patients was associated with a mean change in HIV RNA of -0.68 log10 copies/mL. 51 patients were randomized to 150 mg once daily dose, 54 to the 75 mg dose, and 28 to placebo.

Thompson reported viral load reductions seen previously with PMPA in various populations: 300 mg in ART-experienced, -1.06 log; in an early clinical study (Study 901), preliminary analysis of data from treatment-naive patients randomized to receive 300 mg (n=4) of PMPA for 28 days had mean reduction in plasma HIV RNA levels

from baseline of 1.5 log10 copies/mL.

The 48-week double-blind, dose-ranging study enrolled 189 treatment-experienced patients who were on a stable antiretroviral regimen of no more than 4 antiretroviral drugs

for at least 8 weeks prior to entering the study. Patients were randomized to receive one of three PMPA doses (300 mg, 150 mg or 75 mg) or placebo in addition to their existing treatment regimen. At week 24, all patients receiving placebo were switched in blinded fashion to the tenofovir DF 300 mg treatment arm.

Prior to enrollment, patients had received a mean of 4.6 years of antiretroviral therapy. At baseline, patients had a mean HIV RNA of 3.7 log10 copies/mL (5000 copies/ml) and a mean CD4 cell count of 375 cells/mm3. In the 300 mg PMPA arm the baseline HIV RNA was 5500 copies/ml. Baseline genotypic analyses of HIV isolates from 187 of the 189 patients enrolled in the study showed that 97% had evidence of nucleoside reverse transcriptase inhibitor (NRTI) resistance mutations, 59% had protease inhibitor resistance mutations and 34% had non-nucleoside reverse transcriptase inhibitor (NNRTI) resistance mutations.

Discontinuations and Adverse Events

In the 300 mg arm, 20% (11) patients discontinued through 48 weeks. While 24% (12), 26% (14), and 25% (7) discontinued in the 150 mg, 75 mg, and placebo arms, respectively. In the 300 mg arm 4% withdrew for adverse events and 11% "withdrew consent/lost to follow-up". The percentages were similar in the other arms. After a minimum of 32 weeks data (n=54), 9% (5) patients experienced a serious adverse event, compared to 4% (1) in the placebo arm. In the 300 mg arm, the serious adverse events included 1 depression and 1 pneumonia. In the 300 mg PMPA arm at week 48, 39% (21 of 54) experienced grade 3 or 4 lab abnormalities: 11% triglyceride elevation; 11% creatine kinase; 7% AST elevation; 7% neutropenia; 6% amylase elevation; 4% ALT elevation, 0% serum glucose elevation, and 2% each for serum lipase and bilirubin elevations. In the placebo arm, 29% reported experiencing lab abnormalities of grade 3 or 4.

Thompson reported that through 48 weeks, no patients randomized to tenofovir DF 300 mg had confirmed elevations (greater than or equal to 0.5 mg/dL) of serum creatinine (a marker of kidney function). There were 2 unconfirmed cases (4%) in the 300 mg dose arm by week 48, and 1 in the placebo (4%) arm by week 24. Gilead said that the pattern of phosphate declines <2.0 was different in this study than with adefovir. Gilead reported that 5 patients (10%) in the TDF 300mg dose group and 1 patient (4%) in the placebo group (censored at 24 weeks when they crossed over to active therapy) had a serum phosphate < 2.0 mg/dl through week 48 of the study. The decreases in serum phosphate occurred (by looking at Kaplan-Meier curves) between weeks 4 and 28. From week 28 to week 72 the rate remained stable in both arms. At week 76 there was one case of phosphate decline <2.0 mg/dL in the 300 mg PMPA arm. In the adefovir studies, 47% of patients treated with the 120 mg dose of adefovir experienced serum phosphate <2.0 mg/dL by week 48, but didn't occur until after 20 weeks on therapy. As well, phosphate <2.0 mg/dL was often associated with creatinine abnormalities.

At the community meeting and in follow-up telephone conversations with Gilead the following information was gathered. In pre-clinical animal studies (rat, dogs, monkeys) decreased bone mineral density (BMD) has been previously observed at PMPA drug levels 10-30 times the amount used in humans. Gilead said that in experimenting with giving a dose of 5 times the human dose (10 mk/kg) to young animals (who Gilead says may be more susceptible to bone loss), they did not see bone loss out to two years of dosing. So far (48 weeks) in the human study described here (#902), Gilead reported a 1-2% BMD loss which Gilead reported was within the assay variability. Gilead said clinically significant, in terms of being at risk for bone fracture, occur when there is a 10-20% bone loss. Gilead also said there is data on women for osteoporosis that says if there is 1-2% per year bone loss that accumulates to 20% over 10 years, there is increased risk for bone fracture. Gilead reported there were no changes in serum markers of bone disease (serum calcium, serum phosphate, alkaline phosphatase, parathyroid hormone) and no clinical reports of bony abnormalities (spontaneous fractures or pain requiring hospitalization) through 48 weeks. To evaluate the risk of bone loss associated with PMPA, longer term studies appear to be required. Gilead will continue to evaluate bone density changes in their phase 3 program using standardized methodologies across study centers. Changes in bone mineral density have recently been reported in HIV patients taking HAART therapy including protease inhibitors.

At the EASL (European Association for the Study of the Liver) meeting held in Rotterdam April 12-14, there was a study reporting that BMD loss was associated with HCV and advancing HCV disease progression. See NATAP EASL Reports on NATAP web site.

Due to the BMD observations associated with PMPA, the FDA may be requesting 48 week safety before granting commercial availability. Accelerated approval is granted with 24 weeks data. Gilead and the FDA will be meeting for further discussions. PMPA has utility for salvage therapy, but requiring 48 weeks safety data will delay availability.

HIV-RNA and CD4s

The mean change in HIV RNA from baseline for patients receiving the 300 mg dose was

-0.68 log10 copies/mL after 48 weeks of treatment (n=41). Patients in the 150 mg and 75 mg PMPA arms had mean changes in viral load of -0.6 log10 copies/mL (n=34) and -0.43 log10 (n=40), respectively. 28% percent receiving PMPA had <400 copies/ml at week 24. In addition, patients who had received placebo for the first 24 weeks of the study and switched to the 300 mg dose for the second 24-weeks had a mean change in HIV RNA of -0.71 log10 at 48 weeks (n=19). At week 24, the CD4 count in the 300 mg dose arm was 14 lower than at baseline (n=50), while it was about 11 CD4s above baseline at week 48 (n=41). The same type of CD4 response was observed in the other arms.

PMPA 24-Week Resistance Analyses

Michael Miller, Ph.D. of Gilead Sciences reported on the PMPA resistance data from a virology sub-study of all patients enrolled in Study 902. In this sub-study, genotypic analyses of plasma samples from 124 patients were performed at week 24; their median HIV RNA was 4731 copies/ml; the remaining patients had insufficient viral load for genotypic analyses. 34 patients had HIV RNA too low (median 311 copies/ml) to test. 23 patients had <50 copies/ml.

Miller reported the antiretroviral drugs patients were taking at baseline-- 99% of patients were taking NRTIs, 66% 3TC, 59% d4T, 28% ddI, 24% AZT, 7% abacavir, and 2% ddC; 68% were taking protease inhibitors: 37% NFV, 18% SQV, 14% RTV , and 14% with IDV; 30% were taking NNRTIs: 22% NVP, and 4% each EFV and DLV.

Baseline genotypic analysis showed that 97% of patients had any NRTI mutation (n=181), 74% had an AZT mutation (n=139), 67% had a 3TC mutation (n=125), 48% had an AZT mutation with 184 (3TC) (n=90), 27% had another NRTI mutation (n=51), and 3% had no NRTI mutation (n=6). High level AZT resistance may preclude meaningful virologic response to PMPA, as it affected adefovir. The data isn't yet available from this study on how many people had such high level resistance and how they fared.

Patients were still blinded at this analysis so resistance developed could be in PMPA or placebo arms. Of those 124 patients evaluated, the percentage of patients who developed

new resistance mutations to background NRTIs, protease inhibitors and NNRTIs

was 33% (n=41), 6% (n=8) and 2% (n=3) respectively. 32 patients developed an AZT mutation (AZT, d4T, or abacavir). Three patients developed the T69D/N (AZT or d4T) mutation. 2 patients developed the L74V (ddI or abacavir) mutation. No one developed the Q151M muti-drug resistance mutation or a T69 double insertion mutation. The remaining 78 patients had no evidence of new resistance mutations. Only two patients developed the K65R reverse transcriptase mutation, which has been selected by PMPA

in vitro and by ddI, ddC and abacavir in vivo. Neither patient showed evidence of viral rebound at 24 weeks. One of the two patients showed no decline in viral load through week 32, but at week 12 when there was still no resistance detected at K65 viral load had not declined. Because the study was still blinded at the time of data analyses, it is not known whether the patients were on PMPA or placebo, although both patients were taking either abacavir or ddI.

Gilead has reported preclinical studies have shown that tenofovir is eliminated by the kidney, is not metabolized by the liver and is not associated with cytochrome p450 interactions.

Gilead initiated Phase III clinical testing with Study 907 in November 1999. Underway at the United States, Europe and Australia, Study 907 is enrolling patients who have been on stable antiretroviral therapy of no more than four agents for at least eight weeks. A second 48-week Phase III trial (Study 903) in treatment-naive patients is anticipated to begin shortly.