Pushing back against pressure to shorten the path to the U.S. market for new drugs and medical devices, the FDA has released an analysis of nearly two dozen case studies in which phase III trials didn't bear out the promises of phase II testing.
The case studies – running the gamut from biologics and small-molecule drugs to devices and vaccines – are just the tip of the iceberg, the agency said as it dug in its heels on the importance of phase III randomized, controlled trials (RCTs) in advancing the scientific understanding of the safety and efficacy of medical products.
"As we continue to explore alternatives to requiring phase III testing, it is important to keep in mind the benefits they provide to both patients and to the medical research enterprise," the agency said in the paper released last week in the final days of the Obama administration.
The case studies follow in the wake of the 21st Century Cures Act, which Congress passed last month with the express intent of streamlining development of promising new treatments. Key provisions of the $6.3 billion package are intended to bring drugs and devices to market more quickly and at less cost by expediting the review of breakthrough devices, increasing patient involvement in the drug approval process, streamlining the review process for drug/device combination products, and fostering the development and use of biomarkers as clinical trial endpoints. (See BioWorld Today, Nov. 29, 2016, and Dec. 8, 2016.)
Noting that about nine in 10 drugs and biologics that are tested in people are never submitted for approval, the FDA acknowledged the growing interest in exploring alternatives to requiring phase III testing, such as relying on different types of data and unvalidated surrogate endpoints. The agency points to the case studies in the analysis as support for its insistence that phase III RCTs remain the gold standard for drug and device development and approval.
The analysis looks at 22 instances across the FDA centers in which a promising drug or device didn't demonstrate efficacy compared with placebo or other treatments in phase III or the trial revealed new safety concerns.
The FDA conducted the analysis to better understand the nature of the evidence obtained from many phase II trials and the contributions of phase III trials, FDA press officer Sarah Peddicord told Medical Device Daily. In many ways, the paper reads like a tutorial for laypeople on the purpose of clinical trials in drug and device development.
The cases were "chosen from a large pool of similar examples in which phase II results were not confirmed in phase III," Peddicord said. The analysis, which is consistent with other research, "indicates that medical products that look promising at early stages of development or clinical testing may turn out not to provide any clinical benefit or cause harms when evaluated in larger, more rigorous clinical trials," she added.
The analysis revealed both safety and efficacy failures even when the phase II trials were relatively large or when the product already had been approved for another condition. In some instances, short-term results found in phase II didn't translate into long-term benefit in phase III. In other cases, the phase III trial uncovered a toxicity not apparent in the phase II study.
The FDA pointed out that unexpected evidence from a later trial doesn't always result in nonapproval, though. In the case of Eli Lilly and Co.'s Zyprexa Relprevv, a long-acting formulation of olanzapine pamoate, the phase III trials revealed a risk of post-injection delirium sedation syndrome. Given the efficacy of the schizophrenia drug, especially in patients with a history of nonadherence, the FDA approved the long-acting formulation with a risk evaluation and mitigation strategy that requires monitoring of all patients for three hours following the injection.
The analysis also shows that a medical product's apparent mechanism of action doesn't always predict clinical outcomes. Most of the products in the case studies had a plausible mechanism of action, but the phase III results showed no benefit. And in some cases, the experimental product actually increased the frequency of the problem it was expected to prevent based on mechanism of action and biomarker data.
Pfizer Inc.'s torcetrapib, for example, was intended to reduce heart attacks by increasing "good" cholesterol (HDL) and lowering "bad" cholesterol (LDL). Phase II trials showed the drug did increase HDL and lower LDL. However, in the phase III trial that examined whether the drug actually reduced heart attacks, patients taking the drug were 25 percent more likely to suffer a major cardiac event than those in the control group.
The FDA's conclusion? Its analysis demonstrates "the public health implications of undue reliance on phase II studies," the agency said. And the case studies support the agency's stance that phase III RCTs are golden for showing safety and efficacy.
The late-stage trials also "help researchers understand when a purported mechanism of action is credible and merits further development, allowing researchers to avoid investing substantial time and resources going in the wrong direction, resources that could be deployed to identify a truly effective product," the FDA said.