The target: an over-active molecular switch
A kinase is an enzyme that adds a phosphate group to other proteins, flipping them on or off — the cell's main wiring for growth signals. In many cancers a kinase is stuck in the 'on' position and keeps screaming "grow," so the cell divides without stop. A kinase inhibitor blocks that one over-active switch. This was the dawn of targeted therapy: instead of poisoning every dividing cell the way old chemotherapy did, hit the specific molecule the cancer depends on.
The breakthrough drug, imatinib, treats a leukaemia driven by an abnormal kinase called BCR-ABL. It is a tyrosine kinase inhibitor that sits in the kinase's active site where ATP would normally bind, a competitive inhibitor of the cell's own energy molecule. With the ATP site occupied, the kinase can't add its phosphate, the growth signal goes silent, and the cancer cells die.
The hard part: selectivity among 500 cousins
Here is the catch. The human body has over 500 kinases, and the ATP site — the pocket the drug aims for — looks almost the same in all of them. Hit the wrong ones and you get toxicity. So the entire craft of this class is selectivity: finding the tiny differences between your target kinase and its 500 cousins, and shaping the drug to exploit them. The famous trick is to catch the kinase in an inactive shape, reaching into a small extra pocket next to the ATP site that only the target opens up.
Resistance, and the covalent answer
Even a perfect inhibitor faces a moving target. Cancer cells mutate, and a single changed amino acid in the active site can stop the drug fitting — a kinase resistance mutation. The drug still works everywhere else in the body, but the tumour has slipped its grip and starts growing again. The class answers this by designing successive generations, each shaped to bind the mutated pocket the last one lost.
The boldest modern move is the covalent inhibitor: drugs like osimertinib carry a mild reactive group that forms an actual chemical bond to a specific cysteine in the target kinase. Because the bond is permanent, the drug stays put even when reversible binders are pushed off, and the bond can be aimed at a residue unique to the target for extra selectivity. It is a striking bookend to this track — the very first class we met won by refusing to react, and the newest one wins by reacting on purpose, with surgical precision.