Data Driven Taxonomy for Antipsychotic Medication: A New Classification System

Abstract

BackgroundThere are over 25 licensed antipsychotic medications with diverse pharmacological and clinical profiles. Antipsychotics are commonly described as either typical or atypical, but this does not accurately reflect pharmacological profiles. There is thus a need for a data driven antipsychotic classification scheme suitable for clinicians and researchers which maps onto both pharmacological and clinical effects. MethodWe analysed affinities of 27 antipsychotics for 42 receptors from 3,325 receptor binding studies. We used a clustering algorithm to group antipsychotics based on their pattern of receptor affinity. Using a machine learning model, we examined the ability of this grouping to predict antipsychotic-induced side effects quantified according to an umbrella review of clinical trial and treatment guideline data. ResultsClustering resulted in four groups of antipsychotics. The predominant receptor affinity and effect/side effect fingerprints of these four groups were defined, as follows: Group 1 - Muscarinic (M3-M5) receptor antagonism; Cholinergic and metabolic side effects. Group 2 - Dopamine (D2) partial agonism and adrenergic antagonism; Globally low side effect burden. Group 3 - Serotonergic and dopaminergic antagonism; Globally moderate side effect burden. Group 4 - Dopaminergic antagonism; Extrapyramidal and motor side effects. Groups 1 and 4 were more efficacious than clusters 2 and 3. The novel classification was superior to existing approaches when predicting side effects. ConclusionsA receptor affinity-based grouping not only reflects compound pharmacology but also detects meaningful clinical differences to a greater extent than existing approaches. The approach has the potential to benefit both patients and researchers by guiding treatment and informing drug development.