In 1992, FDA established accelerated approval (AA) regulations to allow patients with HIV and other life-threatening diseases rapid access to therapeutics.
Accelerated approval allows for FDA approval based on surrogate clinical outcomes that are expected to correlate with clinically meaningful outcomes, like survival. These studies with surrogate outcomes are generally small single-arm phase II trials. Subpart H regulations require confirmatory phase III studies to verify benefit.
Between 1995 and 2005, about one-third of all cancer drug indications approved in this time-period received approval via the AA mechanism. Since then, only four cancer drug indications have received AA.
At a 2003 meeting of FDA’s Oncologic Drugs Advisory Committee (ODAC), obstacles to timely completion of confirmatory phase III studies for eight cancer indications were identified.
In 2005, ODAC again met with six sponsors who were having difficulties fulfilling Subpart H commitments, five of whom had attended the 2003 meeting. The consensus was that Subpart H commitments remain difficult to complete, despite focused and committed efforts from drug sponsors.
In an attempt to remedy this situation, FDA proposed that accrual to confirmatory trials be initiated at the time of AA and that interim analyses of phase III studies should preferentially serve as the basis for AA in the future. Following AA, these phase III trials would continue and hopefully provide evidence of improvements in clinically relevant outcomes.
In 2007, legislation passed by Congress allowed FDA to impose fines on drug sponsors and limit drug distribution if post-marketing commitments are not completed in a timely fashion.
Subpart H commitments remain incomplete for five of eight AA cancer indications discussed at the 2003 ODAC meeting and for five of six discussed at the 2005 ODAC meeting. These findings raise concern that, 15 years after initiating the AA program, reassessment of its benchmarks is warranted.
The RADAR project
At ASH 2007 and a March 2008 symposium convened by the American Enterprise Institute, Veena Shankaran, MD, and Kenneth Carson, MD, of the Research on Adverse Drug Events and Reports (RADAR) project of the Robert H. Lurie Comprehensive Cancer Center of Northwestern University, explored reasons behind the continuing difficulties with completing confirmatory trials for AA cancer indications.[1,2]
They noted that cancer patients are reluctant to enroll in randomized phase III Subpart H trials, since there is a high probability that they will not receive the novel drug. Also, after a novel drug is commercially available, physicians often view confirmatory trials as violations of the principle of equipoise, and are reluctant to advise patients to enroll.
These issues are particularly acute among cancer indications that affect small numbers of patients, where treatment options are generally very limited.
The RADAR team supported improvements proposed by some ODAC members at the 2005 meeting. Specifically, they believe that sponsors should be allowed the option of fulfilling Subpart H commitments by conducting long-term single-arm phase II studies and establishing comprehensive safety registries rather than completing a confirmatory phase III trial.
Consideration of this option is important for drugs used to treat cancers affecting small numbers of patients. For example, even for imatinib (Gleevec) for chronic-phase CML, the Subpart H phase III commitment is still pending.
Safety registries allow proactive assessment of serious adverse events and long-term assessment of clinical benefit. They represent an efficient use of finances for small companies with limited capital. A structured pharmacovigilance program in numerous clinical settings with a broad range of cancer patients in the post-AA setting can uncover safety and efficacy issues in a timely fashion.
Raising the bar
FDA has recently raised the bar for sponsors seeking AA, including increased emphasis on interim results of phase III trials and even requiring that companies report accrual to phase III trials before granting AA.
Overall, the number of new molecular entities that receive initial FDA approval via the AA route has slowed, and for entities that receive AA, required Subpart H commitments are difficult to complete.
Consideration should be given to the recommendation made by RADAR at the American Enterprise Institute symposium that after AA is granted, sponsors of these drugs should not uniformly be mandated to complete randomized phase III clinical trials to fulfill their Subpart H requirements. Rather, drug sponsors who receive AA for these types of cancer indications should be encouraged to maintain rigorous drug safety registries.
This policy change would facilitate early access to novel drugs for patients with rare cancers (and potentially increase the use of AA as an FDA approval option for these drugs); allow AA to be granted for these indications primarily on the basis of phase II studies (as was the original intent); and facilitate earlier detection of adverse effects in the post-approval setting.
1. Shankaran V et al: Waiting for Godot: Thirty-Six month follow-up on accelerated FDA approval of drugs to treat hematologic malignancies. ASH 2007, abstract 3340.
2. Carson KR et al: Accelerated approval in 2008. Presented at the American Enterprise Institute and University of Chicago Law School conference: “Oncology Drug Development: Rethinking FDA Oversight.” March 13-14, 2008, Washington, DC.