Dementia Diagnosis: A tool for healthcare and pharma

There are currently no drugs for Alzheimer’s Disease, AD, available that stop or cure the disease. Results from recent clinical trials, where the drug has been shown not to work, suggest that many of those enrolled in the trial did not have AD at the start. While patients with AD are expected to deteriorate over the course of a trial in terms of their memory recall and ability to look after themselves, many enrolled subjects did not show this deterioration. The inclusion of a large portion of patients who were not declining confounds the results and possibly hides any real affect of the drug. The difficulty in diagnosing patients correctly is not just a problem for drug trials. Patients visiting their GP are even less likely to get an accurate diagnosis as the patient visit in trials is more thorough than in primary care, or even than in most memory clinics. A correct diagnosis can have significant positive impact on patients even in the absence of drugs but will be critical to ensure patients get early access to drugs as they become available.
In this project, we will develop a dementia prognosis tool that uses data collected in clinics (MRI scans, cognitive tests, demographic data), to help diagnose patients by predicting their outcome and expected rate of decline.
We have demonstrated proof-of-principle on a relatively small amount of data and in this project we will work closely with some large pharmaceutical companies and Imperial College to develop better methods and test it on many 1000s of patients from a global population including 100s of patients from each of N America, Europe, China, Australia, Japan.