FDA Gives Salvage Therapy Testing a Push Forward.
In January, the FDA Antiviral Drugs Advisory Committee held a meeting on salvage therapy trial design. The meeting addressed thorny issues, including who to include in these trials; how to design them; and how to put together a successful New Drug Application (NDA) for a salvage therapy agent. Many of the new recommendations will appear in a "guidance document" on the use of HIV RNA measurements to support accelerated and traditional approval for phase III trials. Released for comment over a year ago, the finalized version should be available by the end of 2001, said Jeff Murray, an FDA antiviral medical officer.
These events are a major step forward, treatment advocates respond. Michael Marco, of the Treatment Action Group (TAG), commented after the FDA advisory committee meeting, "Today is the day when industry can stop saying the FDA is not sending a clear message."
Major Uncertainties on How to Test and Approve
Five years ago, protease inhibitors uncapped a revolution for people living with HIV. Today, many of the people who swallowed the first of these pills are facing a new dilemma. Although the FDA has "routinely" asked for anti-HIV drug developers to look at the effects of their drugs in treatment-experienced populations, little such information has been gathered, said Heidi Jolson, outgoing head of the Antiviral Division, at the January meeting.
Although doctors and patients have no choice other than to tackle difficult decisions about stopping or switching therapies, drug developers have been able to sidestep the issue by testing new drugs in treatment-naive individuals, where the drugs have the most striking activity. Trial results affect the drug's approval and marketing, and can also have an immediate influence on the company's stock price.
Any trial has to include a "background regimen" of active drugs that support the company's experimental candidate in attacking trial participants' HIV. Then there has to be an agreed-on method to isolate and evaluate the benefit of the particular drug in development. These requirements become difficult to fulfill in salvage therapy trials. Persons with long treatment histories and rebounding HIV likely have many strains of drug-resistant virus, and constructing an acceptably active supporting regimen may not be feasible. The experimental drug may have only a partial or transient effect as a result.
Persons eligible for salvage therapy trials also tend to have more advanced disease and experience HIV-associated opportunistic conditions. The symptoms and treatment of such infections or malignancies may complicate use of potentially toxic antiretrovirals.
Researchers considering salvage trials as well as many patients have changed their treatment goals. An undetectable viral load may be an unrealistic goal for salvage therapy, while a 0.5 log (two-thirds) drop in viral load is enough to discernibly improve health and boost CD4 counts. But will the prospect of a minimal, perhaps temporary drop in viral load be enough to enroll sufficient participants in a trial?
Companies have to address all these complexities while simultaneously finding a trial design that makes their new drug look good: effective, nontoxic and easy to take. The updated FDA guidance documents will provide the clearest road map yet for traversing this wild and woolly zone of trial design in treatment-experienced patients. Among other things, the new guidelines will urge drug developers to cosponsor trials, and depart from tried-and-true drug approval packages consisting of several trials in treatment-naive patients. Specific statements will be made about how to use baseline resistance testing and other diagnostics to identify the types of patients that are most suitable for participation in particular salvage therapy trials. The document will also advise on "superiority" designs that use the maximum number of drugs likely to have activity, whether experimental or approved, in an "optimized background therapy" regimen.
Guidance documents are the foundation for negotiations between trial sponsors and the FDA. In order to gain FDA approval for a drug, a company or researcher submits a package of data from a series of clinical trials. The guidance documents give recommendations on how best to design these trials to gain approval. Whether these documents will be enough to spur pharmaceutical companies into relatively unfamiliar territory remains to be seen. The recommendations are only suggestions, not requirements. "There's no teeth beyond encouraging [new trial designs]," acknowledged Murray.
Some parts of the FDA's advice may be more appealing than others. In particular, drug companies may be more willing to assemble nontraditional NDAs. At the January meeting, Jolson bestowed the FDA's blessing on NDAs made up of one traditional trial that shows how well a drug works in treatment-naive populations, paired with rational, well-designed "supportive" trials that offer evidence that the drug also works in treatment-experienced individuals. This statement reflects the 1997 FDA Modernization Act, which revised FDA drug approval requirements by stating, "data from one adequate and well-controlled clinical investigation and confirmatory evidence" may be sufficient and constitute "substantial evidence" of a drug's merits. Previously, at least two large comparison trials, with a placebo or accepted treatment as a control arm, were required to support an NDA.
Abbott Pharmaceuticals was one of the first companies to use the new simpler package. The approval of its protease inhibitor, Kaletra (lopinavir) was based, in part, on a 57-person uncontrolled trial in protease-inhibitor-experienced individuals. The trial, which correlated response to a Kaletra-containing regimen with baseline tests of protease-inhibitor resistance, provided valuable information about the drug in treatment-experienced patients. Still, it was not without its limitations. (See "Doctors Hesitate on Kaletra," Treatment Insider, January 2001, http://199.105.9l.6/treatment/hiv+/january.pdf). In order to enroll, patients had to be NNRTI-naive. What the results showed, therefore, is that treatment-experienced patients responded well to a regimen containing Kaletra and another drug, the NNRTI Sustiva (efavirenz). Abbott maintains that the improvement in viral control was due to Kaletra, but it can support this claim only with the indirect evidence linking the extent of resistance to Kaletra with lac k of response.
Drug companies may be less enthusiastic about a planned FDA recommendation for co-sponsorship of experimental agents. Rather than having each company compete separately, the FDA will support drug companies testing two or more new agents at once. This would provide information about how drugs worked as part of a combination, rather than individually. It could, theoretically, lead to approval of both drugs with the requirement that they be used together. Jim Rooney of Gilead Sciences testified at the FDA meeting as a representative of the Inter-Company Collaboration (ICC), a trade group that seeks to promote joint trials that combine drugs from several companies. In an interview, he argued that this type of cooperation could work for companies, "as long as the indication could be modified later based on data from additional studies in different patient populations with different combinations, etc."
It may take more than a polite request from the FDA to get companies to break the mold of competitive research. "I can't say what kind of incentive could be provided [to companies who collaborate] beyond saving on patient resources," said the FDA's Jeff Murray. To date, there have been no significant jointly sponsored trials of new agents, even to determine how the drugs affect each other's metabolic elimination. Gilead and Trimeris have taken steps to include each other's experimental drugs, tenofovir and T-20, respectively, in ongoing efficacy trials. Trial participants are allowed to include the experimental drugs in their background regimens. This could be important for individuals deciding whether to take a single new agent or wait until a better combination becomes available. But having the option to take one experimental drug while in trial for another will not provide valuable data about how the two work as a strategy.
Activists and treatment advocates are ready to flex their muscles to see relevant salvage therapy trials become a reality. "If the FDA is enthusiastic, the community is enthusiastic, and the researchers' only concern is that we may not be able to persuade the drug companies [to collaborate], I think we may have a chance," says Yvette Delph, Antiviral Project Director of TAG. "I think the FDA has far more persuasive power than it has been willing to wield."
Traditional trial designs work like simple addition problems: experimental drug W is added to an approved combination X+Y+Z. At the end of the trial, there is data on whether X+Y+Z plus W works better than X+Y+Z alone. The next, hotly debated generation of trials will be more akin to multiplication. One model, which received considerable support at the FDA meeting, was the "modified factorial design." This type of trial is conceived in manner similar to a multiplication table, with particular drugs rather than numbers heading individual rows and columns. Each multiple -- in this case a combination of drugs -- occupies a different "cell," whose composition is determined by the intersecting rows and columns.
The advantage of this design is that it allows several new drugs to be tested in a single trial. Such a trial might place all participants on optimized background therapy and then randomly assign them to different cells, each of which contains at least one, and possibly more experimental drugs. The results from each of the cells could then be compared to identify the most effective combinations.
There are drawbacks to this design. FDA advisory committee member John Mellors (University of Pittsburgh) pointed out that such trials will have to be quite large, to ensure that there are enough people in each cell for meaningful statistical comparisons. In addition, not all cells are created equal. While each participant gets at least one new drug, he or she may be deprived of the most promising combinations. Modified factorial design could also lead to a tangle of data on side effects, in which it might be difficult to clearly assign blame.
These and other criticisms notwithstanding, the FDA used the meeting to re-emphasize its support of modified factorial trials as part of approval packages. Dr. Jolson reminded attendees of a 1999 letter to industry indicating support for this approach.
Fast Approval, Plus Long-Term Followup
For some patients with no treatment options, time is of the essence. Although the FDA guidance document will not include a clear statement about the appropriate length of salvage therapy trials, Murray said that the FDA could accept as little as 16 weeks of data on evaluating salvage therapy safety and efficacy. At present, 24 weeks of data is the standard minimum. Researchers support this slimmed-down timeframe. "I'd vote for it one hundred percent," said Dr. Michael Saag (University of Alabama at Birmingham), who spoke at the FDA meeting. "Yes, it's taking a risk to release a potentially toxic medication for which we don't have a safety profile beforehand, but how does that toxicity look to a patient who is going to die [without the drug]."
In this scenario, drug approval will not hinge on 16-week data alone. However, 24- and 48-week safety and efficacy information might be gathered in a separate trial, perhaps in a less pretreated population. "We often let companies submit under earlier data and then update during the review [process] anyway," said Murray. A TAG discussion paper agrees that 16-week data could be used for "possible submission for accelerated approval" but underscores that "safety issues mandate the continuation of such studies up to 48 weeks and beyond (after approval)."
The move to accept shorter trials comes as the consequences from the first round of rapid approvals are becoming painfully clear. All of today's antiretrovirals earned approval with unprecedented speed, via the FDA's "accelerated" approval process. Today, long-term toxicities, such as lipodystrophy and bone disease, are emerging. These toxicities were not apparent in the trials conducted for approval of the drugs. Even one- to two-year studies did not begin to show how common, or severe, these problems would become.
To Saag and others, this long-term data is the crux of the issue, not the eight weeks lopped off of a 24-week trial. Here again, the FDA's power to change the industry is limited. Its actual tactics are limited to the extreme step of removing a drug from the market and the rather mild threat of changing the wording of a drug's official package insert. Accelerated approval is conditional, and can be revoked if a company does not fulfill its post-approval commitments to long-term follow-up and monitoring. Still, many companies have altogether failed to make good on promises of Phase IV post-marketing studies.
In the face of this patchy record of long-term HAART use, there is a growing call for new data collection systems to fill in the blanks. One possibility is to establish well-coordinated databases that track adverse events and side effects in all patients on approved therapies. "What may be useful is to have more prospective cohorts that are monitoring everyone [for toxic events]," said Yvette Delph. As the list of long-term toxicities grows, it may be increasingly difficult for the drug industry to sidestep its obligations to help create such collection systems.
The salvage therapy arena has become an important one for all parties concerned. Patients on failing therapy are looking for rapid rescue. And drug companies are looking for rapid testing and marketing of their products. Meanwhile, the case for accelerated review of first-line therapies is waning, now that the FDA has approved 14 anti-HIV agents. The agency now is looking at how to meet the unmet needs in HIV with all deliberate speed.
Salvage therapy is chief among such needs, and community, industry and government have some convergence of interest in demonstrating new drugs' efficacy in this setting. Still, problems with determining individual agents' contribution to combination regimens as well as the agents' Long-term effects may yet undermine the efforts at cooperation evinced at the January meeting.
|Printer friendly Cite/link Email Feedback|
|Publication:||AIDS Treatment News|
|Date:||Mar 23, 2001|
|Previous Article:||Building Alliances for World Health Funding.|
|Next Article:||Africa Access: AIDS Activists Organizing Burkina Faso Summit on Generics.|