Could a simple blood test transform Alzheimer’s diagnosis?

Hi, I'm Ashton Harper. I'm the Global Medical Director for Neurosciences for Roche Diagnostics, and in my role, I'm responsible for generating clinical evidence to support our diagnostic innovations.

There are many challenges facing the diagnosis of Alzheimer's today. It's a very, very prevalent condition, so approximately 57 million people internationally are living with dementia. Alzheimer's is the major cause of this, in 70% of cases, and this is expected to rise dramatically to 2050, probably tripling in numbers. So, it’s a high-prevalence disease, and we really need to be trying to diagnose this condition early.

Some of the issues with that are the fact that there's a lot of stigma around this condition, which means that patients don't often present very early to a clinician. Then the first point of contact in the system is primary care, and primary care doctors don't typically have the expertise for diagnosing Alzheimer's. They use subjective pen-and-paper cognitive tests that lack sensitivity in early stages of the disease. And then, with referral on to specialist care, there's a shortage of specialists, which creates more delays. The existing gold standard tests, such as amyloid PET CT and cerebrospinal fluid, CSF, are costly, invasive, have minimal access, and they're inconvenient and uncomfortable for patients. So again, this leads to an issue with diagnosis. In fact, 75% of individuals internationally don't have a diagnosis currently. Those who do typically wait three and a half years to receive one.

So, as I was saying, there are issues with the current gold standard, so amyloid PET and CSF. There's limited access to these tests and they're very expensive. By bringing in a blood test, it enables us to do very quick, simple, routine blood tests that you would have done for any other diagnosis. And this could potentially be done in primary care or, essentially, out-of-hospital settings. This enables broad access to diagnostic testing, which we hope will enable early diagnosis, timely diagnosis of Alzheimer's, and enable patients to be started on treatment or management plans that are tailored to their condition at the earliest possible time.

Roche is presenting lots of interesting data at this conference. I think, by far and away, the data on pTau217 is the most interesting. So this is a plasma biomarker for early diagnosis of Alzheimer's. What we're presenting here is that this test can be utilized across the care pathway. A unique element of this test is that it can be used in both primary and secondary care, and we have validated this test on a very broad population across these different care settings and across different stages of the disease.

Quite a significant proportion of patients were in very early stages, so subjective cognitive decline or mild cognitive impairment, as well as mild dementia. But we really focus on this because we want to have the earliest possible diagnosis to enable any treatments or management plans to have the greatest impact possible, and we've shown that this test has significant accuracy at detecting disease versus non-diseased individuals. And so, showing excellent sensitivity and specificity well above 90%, which is in line with CEO guidance on diagnostic thresholds for blood-based biomarkers.

What will Alzheimer's diagnosis look like in five years? I've got lots of hopes for this. And I think the most important thing really will be to try to reduce the diagnostic wait times that are currently a significant issue for patients and to enable more patients to actually have a definitive diagnosis compared to what they're having today. So I hope we get there, and that's going to require broad adoption of tests such as pTau217. I also think we need to have tests for screening, so for accurately identifying patients that are in a pre-clinical situation, which would enable us to enrol patients into secondary prevention trials. I also think we need tests for prognostication and also for treatment monitoring as well. So, a lot of work to be done, but I'm hopeful that in the next five, maybe ten years, we'll get there.