An Analysis of US Food and Drug Administration Clinical Hold Orders for Drugs and Biologics: A Prospective Study Between 2008 and 2014

Abstract

Introduction: The US Food and Drug Administration (FDA) may issue a clinical hold order during drug development, notably for an unreasonable risk for participating subjects. Holds are important events; however, to date, there has not been a comprehensive study of this topic. The aim of this research was to prospectively analyze clinical hold orders of drugs and biologics between 2008 and 2014. Methods: A prospective analysis of Yahoo Finance: industry center—biotechnology and Yahoo Finance: drug manufacturers—major was conducted from August 2008 to August 2014. The literature was searched from 2000 to August 2014. After identifying a clinical hold for a drug or biologic, additional searches for information were conducted on the sponsor’s and the FDA’s websites. Finally, an additional targeted literature search with the name of the drug or biologic was performed. For any drug/biologic on clinical hold the following was collected: name, sponsor(s), indication, mechanism of action, stage of development or commercialization, partial or full hold, hold reason, hold duration, actions taken, and outcome. Results: Twenty-nine drugs were ordered on hold, 18 small molecules and 11 biologics, covering 37 indications. Oncology drugs were subjected to eight holds and hepatitis C drugs to seven. Twenty-four drugs were in development and all phases were represented. Five drugs were on the market. The hold was complete in 17 cases, partial in ten and undetermined in two. In most cases, hold orders were motivated by a clinical safety issue. For 23 holds, the safety issue was documented with data, but in three cases, it was only a potential risk. In five cases, the safety concern was preclinical. The most frequent safety issue (seven holds) was unexpected death(s). The second most frequent was that an organ for which the drug was prescribed was actually further damaged (liver toxicity five, osteoarthritis three, myopathy one). The median duration of holds was 8 months (range 1–48). Overall, 12 programs were discontinued and six were probably discontinued. After a successful lift of the hold, 11 drugs were still in development. Conclusions: An FDA hold order is a significant risk for a subsequent clinical development failure. There was no association between the stage of development at which the order was issued and subsequent failure. The longer the hold lasted the higher the risk of discontinuation. Hold orders had significant impacts on drug development timelines. Hold orders were not always synonymous with development failure, as for 11 of 29 drugs, the order was lifted.