This blog post comes courtesy of Aysha Akhtar, MD MPH. Aysha is double Board-Certified in both Neurology and Preventive Medicine and has a Master’s Degree in Public Health. She is the Deputy Director of the Army’s Traumatic Brain Injury Program. She is a Lieutenant Commander in the U.S. Public Health Service, in which she deploys to assist with national public health emergencies. Aysha is also a Fellow of the Oxford Centre for Animal Ethics and is a Consultant Editor for the Journal of Animal Ethics.
In March of this year, clinical trials on aducanumab, a drug developed by Biogen as a possible therapy for Alzheimer’s disease (AD) was prematurely halted. The drug targeted the build-up in the brain of beta-amyloid, one of the two proteins found to accumulate in the brains of those with AD, leading to brain cell death. Biogen was undergoing two large phase three clinical trials (the last stage of human trials), when an independent data monitoring committee determined there was insufficient evidence supporting the efficacy of aducanumab in treating the disease. The trials were stopped and the drug abandoned.
The research community called this failure of aducanumab a “devastating blow.” The drug was considered the most promising candidate in a field that has seen failure after failure. Hopes were high with aducanumab because studies on mice showed that removing the buildup of the amyloid plaques “worked wonders.”
A failing model
And here lies a critical problem. Despite the fact that animal testing remains the default “gold standard” in medical research, it’s use in understanding human diseases and predicting human outcomes has become increasingly questioned. In general, it has been noted that more than nine out of ten drugs that are shown to be safe and effective in animals are later shown to be unsafe and/or ineffective in humans. The unreliability of animal testing is especially evident in the field of AD research. After successes in animals, more than 172 drug development failures have occurred in the treatment of human AD.
One of the major problems in using other animals to model human diseases is the inability to reproduce the complexity of a naturally occurring human disease. Mice, rats and other animals used in research don’t naturally develop AD. So, experimenters will try to artificially induce AD in animals. But for an “animal model” to accurately portray a human disease, it must reproduce the full the pathology (the causes and effects) of the disease and the symptoms and signs of the disease.
Experimenters have altered genes in mice to create models of AD. In humans, AD is characterized by the presence of several key features in the brain, but each mouse model is different and no single mouse model shows all the pathologic features of human AD. Instead, each model displays bits and pieces of Alzheimer’s and many display features not present in human AD. Consequently, these models often give conflicting results because they differ in regard to the signs that manifest and the causes behind these signs.
Back in 2007, one of the key messages of the Inaugural Alzheimer’s Drug Discovery Foundation Meeting was that the patient is currently the only true model of AD. Existing animal models replicate various aspects of the disease but do not fully mimic the human condition, resulting in a low predictive value. The conference further concluded that using models with low predictive value provides little understanding of the pathophysiology (the physiology and functional changes) of a disease.
One investigator commented that “in reality, disease models usually model only certain aspects of clinical symptomatology, and because only rarely is the etiology of diseases well understood, the induction of the disease state in the model can differ from the clinical condition.”
In other words, because we rarely fully understand how and why a disease occurs in humans, when we try to replicate that disease in animals we are usually falling well short off the mark. We take a few observations from humans then try to recreate those observations in animals, and we end up relying on the animal models in place of under- standing the full disease in humans.
David Horrobin, founder and editor of the journal Medical Hypotheses, once commented on the obstacles the pharmaceutical industry faces in delivering new therapies and criticized assumptions made about the congruence of animal models of disease to human diseases. For an animal model of disease to be congruent with the human disease, he argued, three conditions must be met:
We must fully understand the animal model;
We must fully understand the human disease; and
We must have examined the two cases and found them to be substantially congruent in all important respects.
Horrobin contended that these three conditions have not been fulfilled for any human disease. He asked: “Does the use of animal models of disease take us any closer to understanding human disease? With rare exceptions, the answer to this question is likely to be no.”
Despite the warnings from Horrobin, the 2007 Inaugural Alzheimer’s Drug Discovery Foundation Meeting, and countless others, animal experimentation continues as the foundation of AD research. Substantial effort has been made to improve the relevancy of AD mouse models. Despite these attempts, these new mouse models still fail to appropriately mimic what occurs in humans. To date, only two types of medications have been FDA approved for the treatment of AD over more than thirty years and none have been approved since 2000. And the two medications that are approved? The impact they have on patients with AD is minimal. Given the repeated failures in AD research using animals, isn’t it time to focus on developing more effective, human-based testing methods?