Background: We recently developed a CD8 T cell-induced mouse model that allowed development of a biomarker for Alzheimer’s disease based on quantitation of homeostatically-induced Amyloid Precursor Protein [APP] peptide-reactive CD8 T ("hiT") cells in blood (PMID: 38995966) using APP_471-479/Human Leukocyte Antigen(HLA)-A2 multimers (dextramers) as staining reagents in flow cytometry. Levels of hiT cells distinguished AD and related MCI from normal aging controls with high accuracyThe current version of the test requires patients to carry a common genetic marker, HLA-A2, which approximately half the population carries, and there is a need to develop assay solutions that increase the accessibility of the test.

Methods: To extend the test to patients without the HLA-A2 genetic marker, we identified an alternative non-self target peptide (ALIAPVHAV) pHLA-A2 dextramer (“Universal Dextramer”) that detects hiT cells regardless of a patient's HLA background. T cell binding to the Universal Dextramer correlated closely with binding to the original APP-based target in HLA-A2-positive patients (0.42-0.58x relative to APP/HLA-A2; r = 0.81; P<0.001), establishing it as a reliable substitute. Using this approach, we quantified hiT Cells in frozen whole blood and frozen buffy coat samples from 137 patients total: 84 with the HLA-A2 marker (79 in established Univ. of Antwerp cohort, 5 in new validation cohort from USC ADRC) and 53 without (44 UA, 9 USC).  Patients were drawn from four groups: Alzheimer’s disease (AD), mild cognitively impaired (MCI) with or without CSF biomarkers consistent with AD (MCI-AD and MCI-norm/other, respectively), and normal aging cohorts. We then assessed correspondence of dextramer+ hiT cells to AD and MCI status. As a second approach, we also tested a whole blood antigen-induced marker (“WAIM”) assay developed by Beckman Coulter. This method exposes blood samples to APP protein and measures the immune response (T cell activation) it triggers. The WAIM method can detect T cell reactivity independent of a patient's genetic background, offering another potential path to an HLA-independent assay.

Summary of Results: As with the APP pHLA dextramer in HLA-A2+ patients, levels of hiT cells binding to the “universal” pHLA dextramer were significantly diminished in AD and MCI-AD, and independent of HLA-A2 status. In the HLA-A2- cohort, AUC for hiT cells binding to the universal dextramer distinguished AD and MCI-AD from controls (AUC 0.912 for AD; 0.921 for MCI-AD) at least as well as the APP dextramer in HLA-A2+ patients. hiT cell reduction was also seen in HLA-A2-negative AD patients from USC ADRC in interim testing, supporting further validation studies using ADRC specimens. Reactivity to APP protein by both CD4 and CD8 T cells in the WAIM analysis of normal human blood reveals additional possibilities for HLA-independent assay formats.

Conclusions: Levels of hiT cells binding and/or responding to specific antigens strongly corresponds to AD and related MCI, and existing assays can be adapted for use in both HLA-A2+ and HLA-A2- patients. Ongoing work will focus on developing additional options for assay expansion to all patients regardless of HLA subtype status.