FDA Advisory Committee Recommends PMPA Approval For Treatment-Experienced

When It Rains It pours: PMPA FDA hearing

Written by Jules Levin
This is a first report and further reports may be forthcoming.

As expected, PMPA (Tenofovir) will receive FDA approval in treatment experienced patients. There was no surprise data presented at the FDA hearing to negatively impact on that. The Advisory Committee voted 9-7 to approve PMPA only for treatment-experienced patients, not for treatment-naīve. When you consider only committee panelists with a right to vote I think the vote was 8-5. But the FDA did not request a formal vote saying they had heard enough to make a decision. Perhaps they did not want to take a formal vote in the interest of time since the meeting had a delay due to "rain". If the WTC tragedies havenšt been enough, the ceiling in the auditorium where this FDA hearing was held started leaking on top of a few panelists. They had to halt the hearing to repair the situation and the hearing ended up with about a 15 minute delay due to the leak. Everyone on the FDA Antiviral Drug Advisory Committee recommended that the FDA approve PMPA for treatment experienced patients. The committee roundly felt strongly that PMPA is a very important drug for patients with resistance to the currently available HIV medications, and needs to be quickly made available. There was agreement that PMPA has an impressive resistance profile.

A main controversy at the hearing was whether PMPA should receive approval for use in treatment-naīve patients. Committee members wanted to wait until they had more data in treatment-naīve patients before deciding. A large study in treatment-naīve is ongoing. Another key discussion was the question whether PMPA presents bone problems for patients. Early data in animals found bone problems when PMPA was administered at much higher doses in animals than used in humans. And when doses were reduced in animals to doses equivalent to that used humans the toxicities went away. This subject received much discussion at the hearing. Bone experts were on the panel and another bone expert was brought in by Gilead.

The bone experts on the panel agreed that we donšt know the mechanism causing the bone problems seen in animals. And not knowing the reason why this occurs limits our ability to predict outcomes following long-term use in humans. The experts said we need to follow patients for a longer term to evaluate this concern. Having said that, so for in human studies going out about 1 year bone problems have not been seen. This concern was also raised as a consideration in wanting to see more long-term data before granting approval for use in treament-naīve. The FDA agreed that no bone problems are likely to be seen for 48 weeks in humans. Gilead suggests that the bone toxicity appears likely due to an imbalance in phosphate and not a direct effect of PMPA. The bone experts on the panel felt this was a fair assessment. But they would like to see more long-term follow-up, as they concluded that we donšt know the cause and until we do know the cause it makes predicting longer-term outcomes difficult. One of the bone experts on the panel said that until we know answers to this question he recommends a battery of blood work and a Bone Mineral Density test before patients start on PMPA therapy. He listed a number of tests he recommends. For emphasis, Roger Pomerantz, a panelist, asked the bone expert to repeat the list of tests he recommends. In particular, the bone experts said before starting PMPA itšs important to see if there is Vitamin D deficiency. An expert said that 10% of normal people can have Vitamin D deficiency and so in sick patients, as in HIV, deficiency could be more prevalent.

In women, the bone concerns may be more pronounced since post-menopausal women have a greater potential for developing osteopenia, and so this needs further research attention. Panelists also raised concerns about potential bone problems for kids since they have different situations. Kids are still developing as opposed to adults. In kids with resistance to currently available HIV meds, the risk-benefit ratio appears to favor PMPA use. But in treatment-naīve kids waiting until there is more data may be preferable. Panelists generally agreed that in kids and adults the risk-benefit ratio for treatment-experienced patients who need a new drug favors approval for and access to PMPA. It was also suggested we may be able to get answers to bone question more quickly in kids.

Key HIV experts on the panel agreed that it is premature to approve PMPA for treatment-naīve patients. Some other panelists were willing to approve for naīve. But after the bone concerns were discussed in more depth, several key HIV experts felt all the more strongly that we need more data before approving for naives.

Another key point of discussion was whether the 3TC classic mutation M184V was relevant to response to PMPA therapy. Resistance data reported at the 4th Resistance Workshop in 2000 showed that when the 184 mutation was present patients had a better viral load reduction than if the 184 was not present. Data was presented here showing otherwise. When Gilead was asked today about this, they said they are not recommending patients continue on 3TC to maintain the presence of the 184 mutation. Therefore, it appears as if Gilead does not believe the 184 improves response. I think resistance experts on the panel agree.

There is more interesting resistance data not covered in this report. For the sake of brevity in this report Išm limiting the amount of info. Essentially, the more NRTI mutations a patient has the less likely they are to respond well to PMPA. But PMPA appears to be effective for many patients with NRTI mutations. Patients with 3 or less mutations tend to respond better and patients with more mutations tend to respond less well. When the 41 or 210 AZT mutations are present with other NRTI mutations response may be reduced. In study 902, phenotype resistance data shows when PMPA resistance is 4 fold the viral load reduction is .17 log. This compares to a .71 log reduction when phenotypic resistance is 1 fold or less. When resistance is 3-4 fold viral load reduction was .55 log.

The panelists recommended that more drug interactions studies need to be conducted in HIV+ individuals. In particular, Gilead reported on a drug interaction study in healthy volunteers receiving PMPA and Kaletra (lopinavir/ritonavir), ddI, EFV, indinavir, and 3TC. 3TC and ddI are renally excreted as is PMPA. The administration of PMPA in combination with Kaletra caused a statistically significant increase in the Cmax (peak levels) and AUC of PMPA (about 30%). Gilead said these results are difficult to interpret and could be due to a food effect rather than a drug interaction. Administration of PMPA with ddI caused a 28% increase in Cmax and 44% in AUC of ddI. Gilead also said that when looking at two studies they saw no evidence for an increased frequency in ddI-related adverse events or lab abnormalities (pancreatitis, peripheral neuropathy, elevated lipase or amylase). Gilead concluded this suggests the interaction is not likely to be clinically significant.

PMPA also had slight but statistically significant effects on the disposition of Kaletra causing about a 15% decrease in Cmax and AUC and about an 11% decrease in Cmin. Again, Gilead felt these small changes are not clinically significant. A panelist suggested an interaction study be conducted using PMPA with 600 mg of ritonavir. Gilead reported that overall these interaction studies demonstrate a lack of clinically significant effects on the PK of PMPA, and PMPA had no significant effect on the PK of 3TC, indinavir, or EFV. Panelists also recommended that since PMPA is excreted renally, interaction studies should be done with other drugs renally excreted.

Studies of PMPA have been conducted with patients having a meal. Use with food enhances oral bioavailability by 40%. The datasets were small but showed no differences in response to PMPA for men vs women and whites vs non-whites. Panelists recommended a further study be conducted in women of color to confirm these findings. The response to therapy for patients with high viral load (>50,000 copies/ml).

Pivotol clinical studies reported in the New Drug application from Gilead to the FDA focosed on the 902 and 907 (n=552) studies where treatment-experienced patients with detectable viral load intensified their regimens by adding PMPA. At the hearing data from monotherapy studies were also discussed.

In the small 901 monotherapy 28-day study, patients who completed the study taking 300 mg once daily PMPA had a viral load reduction of 1.22 log. Patients taking 600 mg perday had a viral load reduction of .80 log. 300 mg once per day is the chosen dose. For patients who were treatment-naīve the viral load reduction was 1.57 log and for treatment-experienced the reduction was .97 log.

In study 902 (n=186; randomized to 3 dose groups), the overall viral load reduction was about .60 log but the baseline viral load for the patients was low at about 14,000 copies/ml. In study 907 the viral load reduction was also about .60 log and the baseline viral load was even lower at about 4000 copies/ml. So perhaps in my opinion they could not evaluate the full viral load reduction in some patients because their viral load was so low to begin with.

There does not appear to be a concern about nephrotoxicity with the data so far accumulated. Serum creatinine abnormalities were uncommon in Gileadšs central lab database. Gilead reported that despite extended treatment duration no patent had a serum creatinine elevation above grade 1 and 5% in the PMPA group had grade 1 elevations. Of the 32 patients with grade 1 elevations 18 (3%) had serum creatininešs „ 1.5 mg/dL, with a maximum of 1.9 mg/dL.

15% of the patients receiving PMPA had grade 1 or 2 hypophosphatemia. There were 4 patients reported to have grade 3 or 4 hypophosphatemia. One patient with grade 3 toxicity has this at baseline, and the other 3 patients (2 grade 3, 1 grade 4) had values within normal ranges when the tests were repeated within 10 days.

Gilead reported that to evaluate the "sporadic and transient" appearance of hypophosphatemia, an analysis was performed on patients with grade 2 or more hypophosphatemia. Of the 61 patients in the PMPA group, 51 had an abnormal value at only one visit, and the abnormality resolved at the next visit. 10 patients had 2 consecutive grade 2 values; 1 patient had 3 consecutive grade 2 values. No patient was discontinued due to hypophosphatemia. Gilead summarized that serum phosphate decreases and serum creatinine elevations occur sporadically, do not worsen, and resolve upon treatment interruption. To date, there remains no evidence of a significant PMPA-related effect on these parameters.

Only 1% of patients discontinued treatment with PMPA due to GI side effects. PMPA displayed little concern in the way of side effects and lab abnormalities in studies 902 and 907. The incidence of patients experiencing side effects appeared no different for patients receiving PMPA or PMPA placebo. Interestingly, the incidence of ALT elevations appeared low and suggests use of PMPA in HCV+ patients may be helpful. Further studies should be conducted by Gilead including evaluation by biopsy.



to top

women & HIV>