Mechanism of Action (MOA)
A mechanism of action is the molecular account of what a drug binds and what that binding does, traced through to the physiological change that produces the clinical effect.
A complete mechanism names a target, states what the drug does to it, and connects that event through identified intermediate steps to the outcome measured in patients. Semaglutide provides a tidy example: agonism at the GLP-1 receptor, glucose-dependent insulin secretion and suppressed glucagon, slowed gastric emptying, and hypothalamic signalling that reduces energy intake. The related phrase mode of action is sometimes reserved for the physiological level rather than the molecular one, though in practice the two are used interchangeably.
Regulators approve drugs on demonstrated efficacy and acceptable safety, not on mechanistic understanding, and several long-established medicines are still argued over. Lithium's action in bipolar disorder has no settled account. Metformin's effects have been attributed variously to hepatic mitochondrial complex I, AMP-activated protein kinase, and the gut. None of these are in doubt clinically.
Mechanism earns its place by generating predictions rather than conclusions. It anticipates class effects, suggests where drug interactions and off-target liabilities will appear, indicates which populations may respond differently, and tells you what to monitor. What it cannot do is substitute for outcome data, because a mechanism that should work is an argument for running a trial, not a result from one.
The characteristic abuse in this field is the mechanism cascade assembled from fragments. A page will list receptor interactions, growth factor pathways and cytokine changes, each traceable to a cell culture or rodent experiment, and present the chain as an explanation of a human benefit that has never been measured. BPC-157 marketing is the clearest instance: an elaborate signalling narrative built on rat studies, offered in place of the controlled human trials that do not exist.