Accelerated Approval and the Tension Between Safety and Access

Mikkael A. Sekeres, M.D., University of Miami Health System, discusses his latest book Drugs and the FDA and the use of accelerated approvals based on surrogate endpoints.

The accelerated approval last year of Biogen’s Aduhelm (aducanumab) to treat patients with Alzheimer’s disease is just the most recent example that demonstrates the tension that exists regarding the FDA and access to safe and effective medications, said Mikkael A. Sekeres, M.D. “This is a desperate patient population with very few options, none of which worked particularly well. I get the sense they were feeling political pressure and patient advocacy pressure to get a drug approved.”

Aduhelm’s approval, he said, was similar to the controversy after the accelerated approval in 2008 of Genentech’s Avastin (bevacizumab) — after an advisory committee voted no — to treat women with HER2-negative metastatic breast cancer.

In his new book, Drugs and the FDA, Sekeres, who is professor, chief of the Division of Hematology, at the University of Miami Health System, gives a first-hand account of the 2011 hearings of the Oncology Drug Advisory Committee, of which he was a member for five years, to discuss removing Avastin’s breast cancer indication after a confirmatory trials showed no significant benefit.

“Avastin was the very first time that the FDA had approached a manufacturer of a drug that was approved under accelerated approval to remove this from the market and that company said no. That’s bold,” Sekeres said in an interview with Formulary Watch. “I always felt like this was this momentous point in time for the FDA, where their authority to remove for a drug was being questioned at that the highest levels, and very publicly.”

Sekeres explains in the book how patient deaths and drug-related safety issues led first to the 1906 Pure Food and Drug Act and the 1936 Food, Drug & Cosmetic Act. (See timeline below.)

The modern FDA was created by the 1962 Kefauver-Harris Amendments, which followed another tragedy. In the 1950s, severe birth defects were seen in children born to women who had taken thalomide while pregnant. The 1962 law created the structure for well-controlled clinical trials in people with a focus on safety and efficacy. Prior to this, manufacturers did not have to prove a new drug was effective, and chemistry studies were enough to show safety.

But it was different tragedy that forced the FDA to consider a faster approach for drug development. The HIV/AIDS epidemic of the 1980s saw activists protesting both at the FDA and pharmaceutical companies doing research in HIV, demanding earlier access to investigational medications. At the time, no medications were available to treat HIV or the opportunistic infection that resulted from AIDS.

The development and approval of zidovudine, also known as AZT, changed the way the agency viewed serious diseases. Zidovudine was approved in March 1987 without the manufacturer conducting a phase 3 study and less than two years after a first-in-human trial began. Close collaboration between the FDA and the manufacturer led to a faster assessment of the therapy’s effectiveness using a surrogate marker, in this case viral load.

It was Anthony Fauci, M.D., director of the National Institutes of Allergy and Infectious Disease — who later became America’s doctor during the COVID-19 pandemic — who suggested in 1989 a parallel track for promising therapies for serious diseases if it showed meaningful benefit over existing therapies.

For zidovudine and other HIV therapies that followed, phase 3 trials conducted after the accelerated approval confirmed that lowered viral load led to longer lives. Newer therapies and dosing regimens tested also led to therapies with fewer toxicities for patients. Throughout the 1990s, therapies for AIDS/HIV and related conditions dominated the accelerated approvals. By 1999, oncology therapy approvals starting to overtake AIDS/HIV therapies. Since the accelerated approval pathway began in 1992, 278 drug applications have been granted approval by the FDA’s Center for Drug Evaluation and Research, according to the Office of Inspector General.

But, as Sekeres explains, to the FDA, meaningful benefit always meant lives longer or lives better. The agency has always said that follow-up trials need to confirm a positive benefit-risk profile to maintain approval and, if significant benefit is not confirmed, the drug should be removed from the market. As of Dec. 31, 2021, 12% of accelerated approvals have been withdrawn, either voluntarily by the sponsor or involuntarily after FDA proceedings, according to a recent study.

In the case of Avastin, it was first approved to treat patients with metastatic colon cancer in 2004, and in 2008 the FDA granted accelerated approval to treat patients with HER2-negative breast cancer. A phase 3 trial showed a 52% reduction in the risk of disease progression or death and a doubling in progression-free survival, which is considered a surrogate endpoint. But Genentech’s confirmatory study showed minimal improvement in progression-free survival compared with the previous study and no improvement in overall survival. Avastin also demonstrated serious side effects, including bleeding, lowering of the immune system, high blood pressure, and fever, and Genentech did not have any data related to quality-of-life measurement.

“With progression-free survival, I could never figure out how I would communicate that to a patient as a benefit,” Sekeres said. “Imagine that conversation: ‘this drug showed that you’re not going to live longer, you’re not going to feel better. But you’re going to go a longer period of time without your tumor worsening. Is that something that you want?’”

Sekeres, however, said a mechanism to get medications to the patients faster is critical. “I do actually believe in surrogate markers. Endpoints like progression-free survival shouldn’t be an endpoint in and of themselves, unless they’re accompanied by data demonstrating the patients have improved quality of life.

“And they should be flagged [on the label] as just that, as surrogate markers. We should be adding to labels a big, fat asterisk that says, ‘this approval is based on a surrogate marker, and this approval may be temporary as we await more safety and efficacy data.’ I do believe in getting drugs that show a real strong signal of efficacy to patients who desperately need them quicker. But we in the healthcare community and the FDA need to do a much better job communicating that these drugs are approved based on an endpoint that isn’t yet clinically meaningful, and that more data are yet to come.”