See article by Yasuda et al.  (pages 481–486) in this issue.
Restenosis remains the major challenge to percutaneous coronary intervention procedures, despite substantial improvements. Standard balloon angioplasty results in restenosis in 30–50% of patients within several months after successful revascularization , requiring repeat target vessel revascularization. Placement of stents effectively reduces early restenosis due to elastic recoil of the artery, but there is increased late restenosis due to excessive intimal hyperplasia, yielding an overall restenosis rate of about 20–30% . Although this restenosis rate is higher than desired, stenting has become the main interventional procedure for treatment of flow-limiting atherosclerotic lesions. In our clinic and numerous others almost 60–70% of critical lesions are stented. This dominant procedure has prompted the colloquial coining of the ‘oculo-stento reflex’—an interventionalist sees a lesion and reflexively stents it.
Rapamycin-coated stents have virtually eliminated restenosis in early trials [3–5]. After 8 months of follow-up there have been no adverse events , thus if rapamycin-coated stents have long-term efficacy beyond these initial trials this method shows great promise. Drug-coated stents offer the practical means of controlling the dose and duration of these new drugs. Unlike polymeric stents, which had a potential to cause inflammation and thrombosis, the latest technique of dip-coating provides immediate delivery to the target vessel wall with minimal effects on adjacent tissue . Cautious optimism is warranted regarding the long-term outcome of rapamycin-coated stents, however, as radiation-emitting stents were also touted as the ultimate cure when they were first introduced. Unfortunately, thrombosis, edge stenosis (‘candy wrapper’ effect), and other complications have limited the usefulness of radiation . A major point is that 30–40% of critical lesions cannot be stented, largely because they occur at branch sites or in small arteries of <3 mm i.d. Hence, it must be appreciated that other methods for prevention of restenosis beyond the ‘oculo-stento reflex’ need continued development.
Yasuda et al.  in this issue address this need with compelling evidence that arterial restenosis can be prevented by catheter delivery of an anti-mitotic agent, not by stenting. The results of the study are dramatic and convincing. Rabbits were subjected to classical injury of the iliac artery elicited by over-inflation of a balloon. Local delivery of the microtubule polymerizing agent docetaxel virtually eliminated restenosis measured 28 days post-balloon injury. Solid evidence for the anti-mitotic mechanism of docetaxel was provided by the 67% decrease in proliferation measured as the reduction in Ki-67-labeled cells. Taxoid microtubule polymerizing agents are widely used as anti-cancer drugs and the major side effect of systemic use, leukocytopenia, is well-documented . Yasuda et al. show that docetaxel did not reduce leukocyte number over the 28-day study, indicating that local application of cytotoxic drugs is without systemic side effects. Finally, blood flow was maintained adequately to the distal limb during the relatively long duration (20 min) local delivery of docetaxel using their specialized catheter. A potentially negative outcome is that long-term local tissue events may translate into arterial dilation, aneurysm promotion, and arterial rupture . Note that a common feature of rapamycin studies using stents [3–5] and Yasuda's work is the use of anti-proliferative agents, thus providing sound rationale for continued focus on development of anti-proliferative pharmacological agents.
An excellent study sometimes raises more questions than it answers and provides a stimulus for additional research. In particular, several needs emerge from the excellent study of Yasuda et al. . (1) Restenosis studies on therapeutic efficacy and, especially, basic cellular and molecular mechanisms, should be conducted on large animal models that better mimic the development of atherosclerotic lesions, e.g. primates and swine. The recent review of Johnson et al.  provides a wealth of information regarding limitations of rodent models of restenosis. A related issue is that mechanisms of restenosis must be studied in animal models having unique, accelerated coronary artery disease, such as in diabetes [12–14]. (2) Catheter delivery techniques that will enable presentation of pharmacological agents to the coronary circulation without limiting flow [15–18] must be refined. Although Yasuda's specially designed local delivery catheter minimized ischemic insult to the limb (skeletal muscle) , there is limited evidence that catheters can be used for coronary interventions where ischemia–reperfusion injury would be much more serious. Local delivery of the drug at the lesion site is fraught with a number of problems; namely, drug to target contact, duration of action, and remote side-effects due to immediate wash-out. Thus, coupling of efforts in pharmacology with interventional device technology seems essential. (3) Perhaps the most urgent need is the study of these pharmacological agents with interventional device technology in naturally occurring atherosclerotic lesions, rather than balloon injured healthy arteries. Application of the findings to restenosis in humans could be limited because restenosis is studied after injury of normally, healthy arteries, rather than after angioplasty of mature, natural atherosclerotic lesions.
In conclusion, promising results of Yasuda et al.  and other anti-proliferative therapy [3–5] strongly indicate that coupling of pharmacology with interventional device technology for prevention of restenosis could add another dimension beyond the ‘oculo-stento reflex’. Is it possible that utilization of all devices coupled with pharmacotherapy will cure restenosis?