Percutaneous coronary revascularisation has revolutionised the treatment of ischaemic heart disease during the past two decades. Despite technical refinements, however, long term results after using standard techniques remain limited by the phenomenon of restenosis—a process whereby elastic recoil and neointimal hyperplasia occur at the site of endothelial injury, often resulting in recurrent symptoms within six months of the procedure. Although the use of coronary stents is associated with lower rates of restenosis than balloon angioplasty alone,1 rates of up to 40% have been reported in some series, and treatment options are often unsatisfactory, with high recurrence rates after further intervention.2 Neointimal hyperplasia begins soon after coronary intervention as a result of platelet activation, inflammation, and proliferation of smooth muscle cells. Pharmacological inhibition of these processes by using drugs administered systemically has had little success in preventing restenosis. A platelet IIb/IIIa receptor antagonist, abciximab, has shown a modest benefit for patients with diabetes mellitus undergoing stent implantation,3 but trials of other drugs have often failed spectacularly despite promising preliminary work in animal models. Intravascular radiation (brachytherapy) using sources emitting γ rays or β rays is an effective way of treating established restenosis, 4 5 although its use for the prevention …