As Alzheimer’s drug trials push into ever-earlier stages of the disease, researchers face new challenges in trial design. Chief among these is how to show clinical improvement when people are still normal on most of the cognitive and functional measures typically used for drug approval. The U.S. Food and Drug Administration provided some direction by issuing a new draft guidance for preclinical trials on February 16. The guidance recognizes three stages of pre-dementia sporadic AD and suggests different outcome measures for each. Notably, the agency says it will consider approving drugs based solely on neuropsychological test scores in preclinical disease. The FDA released other draft guidances at the same time as well, including the first one for amyotrophic lateral sclerosis. Researchers may submit comments online on the AD guidance until May 17, and the ALS one until April 17.
- An updated FDA draft guidance for early Alzheimer’s disease trials defines three pre-dementia stages.
- Drugs for the two preclinical stages could potentially be approved based on biomarker or cognitive outcomes alone.
- In addition, the FDA released its first draft guidance for ALS trials.
Researchers in the AD field welcomed the update. “This is great news,” Suzanne Hendrix at Pentara Corporation, Salt Lake City, wrote to Alzforum. “[The guidance] reflects acceptance of new approaches that are supported by research. … The acknowledgement that large effects in cognition may be important without a co-primary was refreshing,” she wrote (see comment below). Paul Aisen at the University of Southern California Alzheimer's Therapeutic Research Institute (ATRI) in La Jolla, California, agreed. “This is a valuable guide to current regulatory positions on early trial design that will facilitate progress in the field,” he wrote (see comment below).
The FDA released its first guidance for early AD trials in 2013. That document recognized prodromal and preclinical disease stages, and recommended assessing outcomes at these stages using measures that combine cognitive and functional tests, such as the CDR sum of boxes or newer cognitive composites (Mar 2013 news; Feb 2013 webinar).
The 2018 guidance refines these ideas. Instead of using the labels “preclinical” and “prodromal,” the FDA defines three stages of early AD in terms of their functional, cognitive, and biomarker change. In stage 1, biomarkers are abnormal, but people have no cognitive complaints or detectable decline even on sensitive tests. In stage 2, subtle cognitive effects crop up, but no functional deficits. In stage 3, people begin to have problems with some daily tasks. Lon Schneider at the University of Southern California, Los Angeles, said this categorization provides useful structure for trialists, and noted that stage 3 would correspond to mild cognitive impairment due to AD, while the first two are preclinical.
Recommended outcome measures differ at each stage. In stage 1, the agency notes that improvement on biomarkers alone could form the basis for an accelerated drug approval, as long as research suggests such biomarker approval is “reasonably likely to predict clinical benefit.” In an accelerated approval, clinical benefit needs to be confirmed by longer studies after the drug is on the market. The catch for this stage is that no biomarkers have yet demonstrated a clear relationship to clinical improvement, the FDA noted. The guidance suggests a second option, namely running a long enough trial that stage 1 participants progress to stage 2 and develop detectable cognitive problems. It also endorses using the time to first clinical symptoms as an outcome measure.
For stage 2, the agency will look for consistent improvement on several neuropsychological tests or a large magnitude of effect on one, ideally bolstered by positive biomarker change. Drug approvals could be either accelerated or full, depending on the strength of the evidence. This loosens requirements from the 2013 guidance, which said cognitive outcomes would be considered only for accelerated approval.
At stage 3, by contrast, functional improvement must be present. As in the 2013 guidance, the FDA will accept a single measure that combines cognitive and functional aspects, for example scoring participants’ ability to handle financial transactions or social conversations.
The new guidance does not use the term “disease-modifying,” but instead suggests looking for an effect on disease progression by using a crossover design, in which the placebo group switches to active drug at a given point in time. If the placebo group fails to catch up to the benefit seen in the active group, this indicates a lasting effect of the treatment on the course of disease. The FDA emphasized that this benefit must be shown on clinical tests, not just on biomarkers.
The guidance discusses only late-onset disease and does not address familial AD, but the agency notes that some of the same principles may apply to trials in this population.
Schneider liked the specificity of the new guidance. “The 2018 edition of the FDA draft guidance better conceptualizes the 2013 draft guidelines,” he wrote. However, he added that the changes overall neither lower nor raise the barriers for drug approval (see comment below).
Some in industry called the new guidance a positive development as well. “We welcome the agency’s willingness to progress the thinking around disease stage-appropriate endpoints at earlier stages of Alzheimer’s disease,” David Caouette at Biogen wrote to Alzforum.
Unlike the AD field, ALS researchers have lacked clear trial guidelines until now (Jul 2015 news). The new guidance seeks to rectify this. It applies only to drugs that are intended to treat neuromuscular problems. For approval, such a drug needs to demonstrate a clinically meaningful effect on function, symptoms, or survival, the guidance states. The FDA recommends combining function and survival into one measure to account for patients who die during the trial. One way to do this is to use the joint rank test, which first ranks pairs of participants on survival, then on clinical outcomes (Finkelstein and Schoenfeld, 1999). The guidance notes that survival analyses should include data on whether participants need help breathing, as this is a common symptom of end-stage disease.
The agency recommends using a placebo control group, rather than historical data on progression, due to variability in how fast different populations progress. Researchers are also encouraged to assess function at frequent, regular intervals to minimize the effects of day-to-day fluctuations. The FDA cautions researchers to watch for any worsening of disease progression in the treatment group. As with AD, researchers can also use the time to clinical worsening as an outcome. The agency notes that because of the lethal, untreatable nature of ALS, companies can be sparing with preclinical and pharmacological testing to move drugs into trials faster, but they do need to include a dose-response curve.
Jonathan Glass at Emory University in Atlanta said the new guidance will aid researchers. “It clearly identifies what the FDA is looking for in clinical trials for ALS, and also what they are likely to accept for approval. [It] provides structure for current and future drug development for ALS,” he wrote to Alzforum.—Madolyn Bowman Rogers
- FDA Invites Comment on Drug Testing Guidance for Early AD
- Divvying Up the Ice-Bucket Dollars: Looking Long-Term
- Finkelstein DM, Schoenfeld DA. Combining mortality and longitudinal measures in clinical trials. Stat Med. 1999 Jun 15;18(11):1341-54. PubMed.