We recently wrote about the continued, costly failure of Alzheimer’s disease drugs, often occurring in Phase 3 clinical development. Of 244 compounds tested from 2002 to 2012, only one gained approval. Today there are four compounds on the market for Alzheimer’s disease (of six total approved since research began), all aimed at treating symptoms rather than the disease process itself. None are very effective, their effects are variable and limited in duration, and not a single treatment has shown to slow disease progression.
Alzheimer’s disease is a huge — and growing — health problem, both in the United States and worldwide, where it is the major cause of dementia. Health care costs in the United States from dementia were about $818 billion in 2010, and experts estimate this figure will reach at least $2 trillion per year by 2030. So given the size of the problem and its economic impact, why has there been so little progress to date? And what will it take to see a change?
Alzheimer’s is a complex disease that affects people in different ways based on their individual biology. It is considered primarily a disease of the elderly — with the average age of onset in a person’s 80s, except in the rare case of early onset Alzheimer’s, which is usually inherited. Manifestations of the disease can be heavily influenced by other health issues, including other causes of dementia, further adding to its complexity. But in fact, Alzheimer’s dementia may represent the end-stage in a long disease process that begins decades before cognitive and memory changes first manifest. Thus, the failure of drug trials to date — conducted mostly in patients who are already showing signs of problems, however mild — may be a result of brain neurons being too far damaged to be helped much by the treatments under study.
But how can one effectively identify non-symptomatic patients at greatest risk who might benefit from very early intervention — other than those rare individuals whose high familial risk of Alzheimer’s is related to known genetic mutations? Good surrogate biomarkers are needed in order to effectively pinpoint those at risk of Alzheimer’s dementia who could benefit from preventive therapies and to track their response to treatment. Until we have well-validated, predictive biomarkers for the disease that can be easily and cost-effectively measured, as one today measures cholesterol levels as an indicator of coronary risk, wide-spread screening of non-symptomatic individuals is unlikely.
Until recently, definitive diagnosis of Alzheimer’s could be achieved only by brain autopsy after the patient’s death. Thus, part of the problem with drug development efforts for Alzheimer’s was that many of those who took part in trials of potential treatments were never going to benefit, because they did not have the disease. While there still exists no single test that can diagnose the disease, PET scans that look for the presence of toxic amyloid plaques in a person’s brain, spinal taps to detect the presence and changes in beta-amyloid and tau proteins in cerebrospinal fluid (CSF), or measurements of particular groups of proteins in blood are being studied as potential biomarkers for identifying and monitoring the disease from an early stage. Researchers are evaluating such potential markers over the long term in older adults who have an Alzheimer’s-affected parent as part of the National Alzheimer’s Project. Moreover, they are already employing PET measurements of amyloid in clinical trials of some compounds, including Biogen’s aducanumab — where Phase 1 results showed high doses of the antibody could reduce or even eliminate detectable amyloid in the brains of patients treated for the full 54 months of the study. But much more work will be needed to identify and validate biomarkers that can detect early signs of disease, monitor disease progression and predict response to specific therapies.
In light of this, pharma and biotech companies should focus more resources towards the identification and validation of candidate biomarkers for Alzheimer’s disease. Without appropriate biomarkers to stratify the patient population, to select patients for clinical trials, and to monitor (and potentially predict) response to treatment, the possibility of additional drug candidate failures is high.
Furthermore, while PET imaging provides important real-time information, access can be limited to institutions with the appropriate technology and equipment, and the utility for routine monitoring is limited. For this reason, we believe that biomarkers relying on analytes like blood will be highly valuable in this space. As we increase our understanding of the brain and related disorders, through initiatives such as the Brain Map Initiative (which we wrote about here), it is hoped that useful biomarkers will emerge, which can be used to better detect, monitor, and predict outcomes for Alzheimer’s disease.