Genetic Determinants of Tau-PET Imaging in Alzheimer's Disease: A Narrative Review of Genome-Wide and Candidate-Gene Studies.

Alzheimer's disease (AD) is characterized by extracellular β-amyloid (Aβ) plaques and intracellular hyperphosphorylated tau neurofibrillary tangles, with tau pathology correlating more strongly than amyloid pathology with cognitive decline and neuronal loss. While traditional genome-wide association studies (GWAS) have identified loci associated with tau-related phenotypes in cerebrospinal fluid, these approaches cannot capture the regional brain patterns of tau deposition that may underlie disease heterogeneity. Recent GWAS incorporating tau-PET imaging have begun to address this gap, identifying novel genetic variants associated with tau-PET signal through chromatin modification, microtubule dynamics, and immune regulation. Some of these mechanisms may operate at least partly independently of established amyloid-related pathways. This narrative review synthesizes the peer-reviewed evidence from genome-wide and candidate-gene association studies of tau-PET in AD, classifying findings by biological mechanism to characterize the reported genetic determinants of regional tau-PET signal, to grade the strength of the evidence behind each, and to consider their clinical applications. Throughout, the phenotype under analysis is tracer uptake and not directly measured neuropathological tau, and each locus is interpreted with that constraint in view. Transcriptional regulators such as ZBTB20 and EYA4 have been associated with tau-PET signals, although both findings rest on a single, unreplicated analysis. Common MAPT variants showed only nominal associations, whereas the CYP1B1-RMDN2 locus, which influences microtubule dynamics and oxidative stress, reached genome-wide significance and is the only locus to date that has been formally replicated in independent cohorts. CYP1B1 is notable because it offers a plausible mechanism linking oxidative stress to tau aggregation. Genes controlling tau phosphorylation, such as PPP2R2B and FYN, have also been implicated, although the PPP2R2B association was reported without a replication sample and the FYN association was nominal (p = 0.048) in a candidate-gene analysis of 146 participants. Immune- and trafficking-related loci, including CR1, CLU, and BIN1, suggest contributions of neuroinflammation and endosomal transport to regional tau accumulation. Lastly, APOE ε4 is the most frequently reported predictor of tau-PET burden across brain regions, but its effect attenuates in population-based, predominantly cognitively unimpaired cohorts, indicating that it may be substantially amyloid-conditional, not amyloid-independent. Across these studies the genetic architecture of tau-PET signal appears at least partly separable from that of clinically diagnosed AD, although this remains preliminary: no formal genetic correlation has been estimated, the contributing cohorts overlap substantially, and the samples are small enough that a null result carries little weight. The evidence supports continued study of tau-PET as an endophenotype, and it also indicates that the therapeutic implications of these loci are less direct than these studies generally claim.