A blood-based screening algorithm developed by researchers at the Keck School of Medicine of USC could make recruitment for Alzheimer’s clinical trials more efficient. In an international phase 3 trial, the approach substantially reduced the number of participants undergoing PET scans who were subsequently found to be ineligible.
The research evaluated the algorithm within AHEAD 3-45, a trial investigating whether earlier treatment with lecanemab can improve outcomes in people who have amyloid-beta accumulation but do not yet show cognitive impairment.
Finding participants
Lecanemab removes amyloid-beta aggregates associated with Alzheimer’s disease and can slow clinical progression by around 30 percent. However, it is currently prescribed to patients who already have symptoms of cognitive decline. AHEAD 3-45 is investigating treatment at an earlier stage, when amyloid has accumulated in the brain but cognitive symptoms have not yet emerged.
Finding suitable participants is challenging. Only around 30 percent of cognitively healthy adults over 65 have amyloid levels high enough to qualify for trials such as AHEAD 3-45. Before blood-based screening was introduced, potential participants underwent a process that could take up to three months. PET imaging was used to establish whether their amyloid levels met the trial criteria. More than 70 percent of candidates who received these relatively costly scans ultimately proved ineligible.
With the algorithm, the proportion of participants who underwent PET imaging but did not qualify fell from more than 70 percent to 31 percent. According to the researchers, this reduced the burden on both participants and clinical trial sites.
Combining biomarkers and risk factors
The USC algorithm uses blood biomarkers associated with Alzheimer’s pathology. Its first version incorporated the amyloid-beta ratio, which can provide an early indication of amyloid accumulation. A later version added p-tau217, a biomarker that more reliably reflects amyloid burden in the brain.
The model also considers age and whether a person carries the APOE4 variant, both associated with an increased likelihood of amyloid accumulation. Researchers developed the algorithm using data from 1,080 AHEAD participants and validated it independently with data from the Wisconsin Registry for Alzheimer’s Prevention. Three versions were developed during recruitment between 2020 and 2024.
When the first version was introduced in February 2022, the proportion of people receiving PET scans but failing to meet the trial criteria declined from 71 to 50 percent. After p-tau217 was incorporated in May 2023, that percentage fell further to 31 percent.
The algorithm uses a statistical method known as Mixture of Experts. Instead of simply categorising people as amyloid-positive or negative, it estimates their position along a continuous spectrum of amyloid accumulation. This is relevant because some people have intermediate amyloid levels that cannot be adequately characterised using a single plasma marker.
Blood test as a first filter
The findings add to evidence supporting blood biomarkers as an initial screening step before more intensive testing. In this case, blood testing did not replace PET imaging. Instead, the algorithm helped determine which candidates were more likely to benefit from undergoing a PET scan.
That distinction is important because the study evaluated the technology specifically as a recruitment tool for a clinical trial. Its immediate value lies in reducing unnecessary imaging and making it easier to identify people who meet predefined research criteria. More efficient screening could also support trials investigating Alzheimer’s treatments at increasingly early stages of the disease. Amyloid accumulation can begin decades before cognitive symptoms become apparent, creating a potential window for studying preventive interventions.
According to corresponding author Oliver Langford of the USC Epstein Family Alzheimer’s Therapeutic Research Institute, earlier and more accessible testing could eventually help identify Alzheimer’s pathology sooner. For clinical research, the current results demonstrate how blood biomarkers combined with algorithmic analysis can narrow a large group of potential participants before costly imaging is performed.
The AHEAD 3-45 study will ultimately help determine whether treating amyloid accumulation before cognitive decline develops can translate into better clinical outcomes.
Alzheimer blood test
Last year, researchers developed a new blood test that showed strong potential for early and accessible diagnosis of Alzheimer’s disease. The test measures key biomarkers in blood plasma, including Aβ42/40 and p-tau217, which are associated with amyloid plaque formation. Notably, p-tau217 proved highly reliable in identifying Alzheimer’s-related memory problems.
Tested in over 500 patients, the approach offers a simpler, less invasive and more affordable alternative to PET scans and spinal taps. Researchers highlight its potential not only in clinical care but also in selecting and monitoring patients in clinical trials. Further validation is needed across diverse populations, but the findings underscore the growing role of blood-based biomarkers in advancing more accessible and patient-friendly dementia diagnostics.
References
Alzheimer’s & Dementia (research)
Add ICT&health on Google
Show more content from ICT&health in Google Search.
