Introduction: . Alzheimer's disease models (ADMs) are currently used for drug development (DD). More than 20,000 molecules were screened for AD treatment over decades, with only one drug (Aducanumab)FDA-approved over the past 18 years. A revision of pathogenic concepts and ADMs are needed.

Areas covered: The authors discuss herein preclinical models including: (i) in vitro models (cell lines, primary neuron cell cultures, iPSC-derived brain cells), (ii) ex vivo models, and (iii) in vivo models (artificial, transgenic, non-transgenic and induced).

Expert opinion: The following types of ADMs have been reported: Mouse models (45.08%), Rat models (15.04%), Non-human Primate models (0.76%), Rabbit models (0.46%), Cat models (0.53%), Pig models (0.30%), Guinea pig models (0.15%), Octodon degu models (0.02%), Dog models (0.54%), Drosophila melanogaster models (1.79%), Zebrafish models (0.50%), Caenorhabditis elegans (1.21%), Cell culture models (3.31%), Cholinergic models (8.26%), Neurotoxic models (6.79%), Neuroinflammation models (6.92%), Neurovascular models (7.88%), and Microbiome models (0.45%).No single ADM faithfully reproduces all the pathogenic events in the human AD phenotype spectrum. ADMs should be different for (i) pathogenic studies vs basic DD, and (ii) preventive interventions vs symptomatic treatments. There cannot be an ideal ADM for DD, because AD is a spectrum of syndromes. DD can integrate pathogenic, mechanistic, metabolic, transporter and pleiotropic genes in a multisystem model.