A dilemma for the 1990s. Choosing appropriate experimental animal model for the prevention of restenosis.

Pe ercutaneous transluminal coronary angioplasty (PTCA) has become a successful and widely used treatment for patients with coronary disease since its first clinical application byAndreas Gruentzig in 1977. Despite the increase in procedure and case complexity, primary success rates have improved. However, late restenosis, which constitutes the most important problem after successful angioplasty, continues to occur in 30-40% of patients within 3-6 months. Experimental, pathological (autopsy and atherectomy specimens), angioscopic, and angiographic observations, coupled with recent observations in cell culture and in situ hybridization, indicate that restenosis after PTCA involves a fibroproliferative response to vascular injury in the setting of mural thrombosis with platelet activation, thrombin generation, and the release of mitogens.' An alternative or associated mechanism of restenosis after PTCA is recoil at the site of PTCA injury.2 Attempts to modify the fibroproliferative response to PTCA in humans by pharmacological interventions have met with very limited success. Therefore, despite more than a decade of intensive research, PTCA interventions that have shown promise in limiting restenosis in laboratory animal models have not led to similar findings in patients. Difficulty applying the results of animal models to humans is not surprising, considering the complexity of the response of the arterial wall to injury. Therefore, differences in techniques used to initiate injury, degree of stretch, and the type of injury (deep versus mild) that results may modify the outcome of pharmacological manipulation in a given animal model; complex mechanisms regulating the initial deposition of platelets and fibrin on and within the damaged vessel wall and the subsequent fibroproliferative response are not fully understood and may vary among species; the influence of atherosclerotic plaque composition on the response of the vessel wall to mechanical insult is not considered by most animal models of acute balloon injury and confounds the problem of extrapolating experimental re-