The role of clinical trials in the Food and Drug Administration approval process for cardiovascular devices.

T he Food, Drug and Cosmetic Act of 1938 (the Act) required manufacturers to provide evidence of safety to the Food and Drug Administration (FDA) before drugs could be marketed. Amendments to the Act in 1962 added the requirement of demonstration of effectiveness to the condition for approval (21 CFR 314).1 The Medical Device Amendments of 1976 extended these controls to devices, establishing a tiered system of classification based on the level of controls required to ensure safety and effectiveness before marketing a device (43 FR 32988).2 Devices with performance characteristics that are life sustaining, the failure of which presents unreasonable risk of injury or illness, are required to undergo premarket approval, with sponsors providing valid scientific evidence to support claims for safety, effectiveness, and utility. The most recent amendments to the Act were introduced with the Safe Medical Devices Act (SMDA) of 1990 and Medical Device Amendments of 1992. Among other issues, the SMDA addresses the need to monitor "chronic" function and safety through discretionary or required postmarket surveillance of devices using scientifically designed studies (21 CFR 821). The SMDA thus recognizes that the premarketing approval process may need to be supplemented by assessment of long-term performance of devices. The need for such programs was underscored by the late failure of prosthetic heart valves.3 The FDA's role, thus, is to ensure that marketed devices are safe and effective and have demonstrable clinical utility. The approval process for "significantrisk" devices may take one of two routes. A premarket notification [510(k)] requests marketing clearance for devices demonstrated to be substantially equivalent in technology and function to a device marketed before the 1976 amendments. A premarket approval application (PMA) is required for Class III (significant risk) devices or devices marketed subsequent to this amendment. Although device approval under 510(k) procedures may not require clinical trials, devices requiring a PMA must undergo more stringent documentation of safety and efficacy with data that are scientifically derived, using acceptable laboratory (in vitro and in vivo) and clinical studies. The regulations (21 CFR 860.7) spell out in hierarchical fashion what constitutes "valid scientific clinical evidence." The controlled clinical trial, academically acknowledged as the experimental gold standard for clinical studies, is afforded preeminence.45