This week there were two news items of Alzheimer Disease (AD) drug failures, noted on AlzForum and Endpoints News (not affiliated with either one). The first was developed in Japan, edonerpic, and evidently improved amyloid problems in rodents, even improved rodent memory and brain growth, but was no better than placebo in a human trial, and all human patients got worse. These results were announced in 2017, but the peer-reviewed writeup is now available in JAMA Neurology. The other agent, umibecestat, which works on the BACE-amyloid pathway, may be associated with brain problems, which is why the companies involved (Amgen and Novartis) are said to be considering limiting their AD research.
The interesting thing, at least to me, about edonerpic, is that it might have a different mechanism of action and target than the other AD drug failures, and might be protective against the neurotoxic effects of beta-amyloid in transgenic mice. It might also help mice who have brain injury. Sad that it didn’t work out.
Amyloid beta is a protein found in AD human brains at autopsy; BACE is a naturally occurring enzyme in the brain that works along that pertinent molecular pathway. There are many successful examples of such drugs that work on pathway enzymes, like tyrosine kinase inhibitors in cancer. Sadly, several BACE drugs have failed; I had commented about verubecestat on this blog earlier this year. That drug was used in both very early “prodromal” groups of patients, and in those with an established diagnosis of “mild-to-moderate” AD, but failed in both groups. Adverse events were noted.
At the same time, neurologists who do headache (HA) research are doing better, with two articles in NEJM, one about cluster HA and the other about migraine HA The treatments are related to the theory that certain HA problems involve the “calcitonin gene-related peptide receptor”. The one for cluster HA was a monoclonal antibody, a shot given under the skin; the migraine agent was a pill. Both did better than placebo, but not much. Still, it’s a step, it’s incremental progress, and cancer survival improvements have a history of small steps.
Here’s the conundrum from my lay perspective: if rodents, or fruit flies, or roundworms (good ole C. elegans) or dogs aren’t great models for drug screening or predicting the response of drugs in humans, it leaves humans as the primary model, with trials to be done ethically with safety in mind. Cognitive problems of subjects must be considered for informed consent. Since the “amyloid cascade hypothesis” seems stalled, without even a small step forward, it seems time for “lateral thinking.” But I wonder if anyone would be brave enough to take all the things that work on the brain, all FDA-approved neurology drugs, with known safety issues, and try them systematically and empirically on AD patients.
Disappointed that the drugs from the rodent trials in Japan didnt show any significant hope for AD patients. I agree, if all FDA- approved neurology drugs, including the HA drugs that did better than placebo, were administered carefully, we could be closer to making AD treatable and curable.
Thanks for your comment! Some call this idea “repurposing”; Dr. J. Cummings and others have written several articles about it in AD. For safety, this whole concept of empiric testing would be best done on an animal model (like tiny C. elegans or the fruit fly Drosophila) but even the transgenic mouse models don’t seem predictive enough. In cancer, there is a step wise system of human testing, from Phase “Zero” to Phase I, II and III trials, usually starting with the sickest patients. In my own career, I have administered agents in phases I-III. Yes, those trials are happening in Alzheimer and dementia research, but as far as I can tell, there’s no overall coordination or strategy, especially in repurposing. Yes there are proprietary and liability issues, but a coordinated strategy, perhaps using a designated cohort of volunteer subjects for a multiple crossover study (for example) would help deal with some of those barriers.