In an article published in Nature Reviews Neurology, Dr Steven Kiddle from the MRC Biostatistics Unit, together with international colleagues Harald Hampel, Sid O’Bryant, José Molinuevo, Henrik Zetterberg, Colin Masters, Simone Lista, Richard Batrla & Kaj Blennow review the current state of research into a blood test for Alzheimer’s disease.

Alzheimer’s disease, the most common form of dementia, leads to worsening memory and cognitive ability and eventually to death. A blood test would be especially useful for helping to diagnose the disease within general practice, where many patients seek help for memory problems. Existing tests rely on brain scans, or extracting fluid from the spine, which can be expensive or invasive. A blood test could also one day help doctors to spot the disease early, in the hope that it could be treated, although currently available medicines are only able to provide temporary relief of symptoms.

Many challenges exist in developing a blood test for Alzheimer’s disease, including the limited ability for proteins and molecules to pass between the blood and the brain. This has not stopped researchers looking for blood markers of Alzheimer’s, as to date 1,404 studies looking for such markers have been performed. We found that of the 19 most promising blood tests, only five had been replicated, i.e. found to perform well when applied to a new set of patients. Where replication had been successful, it had rarely been performed by independent researchers or based on tests published beforehand, two steps necessary to demonstrate that they are ready to be given to patients. (Figure below taken from published paper in Nature Reviews Neurology)

We recommend a validation process for a blood test for Alzheimer’s disease that will address this concern: (1) asking researchers to predict the disease status of a set of patients based on blood samples, but providing no information in advance on which patients have Alzheimer’s disease, (2) setting up the blood test in an independent laboratory, (3) validation in a large independent set of patients. For the final step we highlight that no existing sets of patient samples represent the population who would most benefit from a blood test, those within primary care, and therefore recommend that a study is set up to provide that.

We conclude that close cooperation is needed among academic institutions, industry partners and regulatory bodies to develop a blood test for Alzheimer’s disease that can make a difference to patients.

---

To read the full published paper in Nature Reviews Neurology, please click on this link: https://www.nature.com/articles/s41582-018-0079-7