USC Study Identifies Three Unique Cognitive Trajectories in Preclinical Alzheimer’s Disease to Improve Clinical Trial Accuracy

USC researchers identify stable, slow, and fast decline paths in Alzheimer's patients using P-tau217 markers to improve clinical trial accuracy and prognosis.

By: AXL Media

Published: Apr 28, 2026, 5:57 AM EDT

Source: Information for this report was sourced from Keck School of Medicine of USC

A Departure From Conventional Linear Alzheimer’s Models

Recent findings from the Keck School of Medicine of USC challenge the long held assumption that Alzheimer’s disease progresses at a uniform, slow pace for every patient. By analyzing data from asymptomatic individuals in the A4 and LEARN studies, researchers discovered that cognitive decline is highly variable rather than a collective slide. According to Michael Donohue, PhD, a professor of neurology at USC, focusing on the average decline of a group often masks the significant differences between individuals, which suggests that the disease is far more heterogeneous than previously documented in medical literature.

Biomarker Precision in Identifying Early Disease Paths

The investigation marks a pivotal moment in neurology by successfully linking specific biological markers to cognitive outcomes before symptoms even appear. Using brain scans and specialized blood tests, specifically measuring the protein phosphorylated tau or P-tau217, the team could distinguish between those who would remain cognitively resilient and those destined for rapid decline. The data revealed that individuals entering the fast decline category typically exhibited higher baseline P-tau217 levels and possessed a smaller hippocampus, the brain region most critical for memory formation and the primary target of early Alzheimer’s pathology.

Statistical Significance of Patient Stability in Trials

One of the most striking revelations of the six year study was that approximately 70% of participants remained cognitively stable throughout the observation period. This high rate of stability among preclinical patients presents a significant hurdle for researchers attempting to measure the efficacy of new pharmaceutical interventions. When a vast majority of a trial group shows no decline regardless of treatment, it becomes statistically difficult to determine if a drug is truly working, potentially leading to the premature dismissal of effective therapies due to diluted data pools.