FDA Approves Second Protease Inhibitor in Record Time – As Abbott Chief Warns Many to “Wait for the data”
And Indinavir Makes Three
Treatment activists met recently with representatives from Abbott Labs at their Abbott Park, IL headquarters in advance of the February 29th FDA hearing which reviewed Abbott’s application for approval of its protease inhibitor ritonavir (Norvir). Ritonavir received full FDA approval the following day; Merck’s indinavir (Crixivan) received conditional approval the following week. In attendance at the Chicago meeting were Mark Harrington (TAG), David Barr (GMHC), Ben Cheng and Martin Delaney (Project Inform), Jules Levin and Gordon Nary (Physicians Association for AIDS Care). Representing Abbott Labs were Andre Pernet, John Leonard, a PR guy and a PK guy.
Andre Pernet reviewed the data from the three studies of ritonavir. These included the advanced study, which was presented with much fanfare at the IDSA meeting in Washington late January; the French uncontrolled combination study; as well as data from the AZT-naïve study, which was to be presented for the first time at the FDA hearing on 2/29.
The Advanced study
1,070 HIV-infected individuals with CD4 cell counts < 100 were randomized to standard-of-care with or without ritonavir, followed for a median of six months (range, 0.2-7 months). Ritonavir lowered the risk of death by 43% and the risk of progression by 58%. 17% of the study participants discontinued ritonavir due to side-effects, versus 6% on placebo. Study participants have since been offered open-label ritonavir, and long-term follow-up continues.
The French study
29 HIV-infected volunteers with CD4 cell counts 100-250 were given open-label ritonavir+AZT +ddc. Median CD4 cell counts at entry were 156 cells; and median plasma RNA, 70,000 copies/ml. 21 (72%) of the study participants tolerated the treatment (it was the vile-tasting liquid formula of ritonavir, no longer in use) and were followed for six months. Virus became undetectable in the blood cells in 6 of 21 (28.6%) and undetectable in the plasma in 5 of 21 (23.8%) at six months. 12 of the 21 (57%) recorded greater than a 99.99% reduction in infectious cells at six months.
The Naïve study
356, supposedly AZT-naïve individuals (No check was made to ensure they were actually naïve–and 10% of those in the study had no expected AZT “T-cell bounce.”) with CD4 cell counts > 200 and viral load > 15,000 copies/ml were randomized to ritonavir alone, AZT alone, or ritonavir + AZT. The ritonavir was again formulated in an unpalatable liquid form for most of the 16-week surrogate marker trial. CD4 rises and viral load drops were better in the ritonavir monotherapy arm than in the ritonavir + AZT combination arm, with the AZT monotherapy arm doing the worst: this was not expected. Viral load dropped by a maximum of 1.6 logs on ritonavir and combination, and stabilized (or rose slightly). CD4 counts rose by 35 cells in the combination are and 80 cells in the ritonavir alone arm at 16 weeks. In the AZT monotherapy arm, viral load dropped by 0.6 log and CD4 cell counts rose by 20 cells. Drop-out on ritonavir arm was 26% versus 39% on combination and 19% on AZT. Nausea and vomiting rates were 100% higher on combination than on ritonavir monotherapy.
“Do you have any lymph node data from the study?” asked Jules.
“I have one lymph node,” replied John Leonard. “We got it after 18 months of treatment. The patient is culture negative and PCR negative, but we cultured virus out of the node.”
“Does ritonavir cross the blood-brain barrier? asked Jules.
“At about 1% of the levels it reaches in the blood,” replied John.
Marty asked why not use viral load, rather than CD4 “immunological decline,” for the labeling indication. John Leonard explained that, exciting as it is, the plasma viral load test is still too new and ever-changing to use for indication purposes. He said many doctors who treat AIDS shouldn’t; it should be a specialty, like oncology. Treatment decisions are more patient-driven in AIDS than in any other field. Marty then asked if any follow-up studies were planned.
“We have big plans,” replied Pernet (who’d conveniently ducked out during the presentation of the naïve study). “We spent $65 million on this drug last year, and we’re going to spend $65 million more this year. We’re going to look at combinations of protease inhibitors, nucleosides plus protease inhibitors, and David Ho’s quatro (four-drug) trial. We’ll have drug for 100,000 patients by May and 400,000 by mid-summer.”
“How do you expect third-party payors to reimburse expensive multi-drug combination regimens without a compelling plan to show that, especially in asymptomatics, they prolong health or life?” I asked.
“Well, first of all, the combinations won’t be as expensive as you think,” replied Leonard, “and secondly, we do have plans to address that.”
“What plans are those?” asked Martin Delaney.
“Well, it might take a very large, long trial,” explained Leonard, “perhaps 20,000 patient years of follow-up. The event rate would be low; that would mean 20,000 patients followed for one year–or 2,000 followed for ten.” The long-forgotten, dreaded TAG large, simple trial reared its ugly head. Maybe our original sample size calculations weren’t all that off the mark. After all, we wanted to answer the question for all populations: early, middle and late. Abbott did the easy part, the late population, and has no plans for clinical endpoint studies in earlier populations.
“We can’t do it by ourselves,” explained Pernet. “We had lunch last week with Bill Paul, Donna Shalala and Phil Lee. We think government and industry need to work together to answer these post-marketing questions. A new body has to study the ideal combination. The ACTG, CPCRA or industry cannot do it by themselves. We need a new mechanism–a business plan–concerted for multiple questions and to avoid duplication of efforts. They were receptive. We have someone meeting with Al Gore as we speak.”
“We couldn’t agree with you more,” said Delaney, “because there is a real danger that, in a rush to use these drugs, people will use them sub-optimally.”
John Leonard asked, “Will you help us restrain patients who shouldn’t be treated yet?”
Andre Pernet went on further, “Don’t use ritonavir up too early, because there’s no data. We have good [sic] triple combination data from France on 100-250 CD4 cells, and survival data on <100 CD4s, but nothing on people with CD4 cell levels above 250. So wait for the data.”
The only true efficacy study is the advanced study. It is a good, six-month clinical endpoint study. But since everyone has been switched to open-label ritonavir, we’ll never know how long the duration of the beneficial effect is. The naïve study, due to the vile tasting ritonavir formulation, was something of a fiasco; yet it still shows that ritonavir is superior to AZT as initial monotherapy in this population (CD4> 200), when measured by surrogate markers. There are no clinical events, no long-term follow-up, and no further studies in this population. The French study is a provocative phase I/II study, but since it was open-label and uncontrolled, it’s nothing more.
We should request that Abbott conduct both a surrogate marker and a clinical endpoint study in earlier populations. The surrogate marker study could be a 16-week, 4-arm study testing AZT+ddI vs. AZT+3TC vs. those arms+ritonavir, using the gel-coated capsules. The clinical endpoint study should be powered to get an answer in no longer than two years.
John Leonard asked, “Why is AZT+3TC being presumptively ratified as the standard of care when it may not be the optimal underlying nucleoside regimen? Eventually we’ll need to pursue novel and cost-effective regimens: why not look at two protease inhibitors? Saquinavir [with ritonavir] is attractive to us.” [NB: They were playing us like a violin to try to get us excited about the saquinavir+ ritonavir combination, and it hasn’t even been in HIV-infected people yet.] “Roche now shares our interest. We’ve also offered ritonavir to Vertex-Glaxo and Agouron. It’s possible to have highly effective two protease inhibitor regimens, but when?”
Andre Pernet then peered into his crystal ball and pronounced, “In a year from now, we’ll be sitting here and be seeing that ritonavir+saquinavir when used together can suppress the virus indefinitely.” Marty asked, “Is it appropriate, given how little we now know, to say that patients should be on combination [protease+protease] therapy–or even on monotherapy?” There was no answer. [Multiple-dosing phase one studies are currently being conducted in HIV-negative volunteers. Studies in HIV-infected adults may begin by the end of April with results presented at the international meeting in Vancouver in July.]