Causality Assessment in Premarketing Drug Clinical Trials: Regulatory Evolution in the USA and Ongoing Concerns

Abstract

Since 1993, how to assess the causality of serious adverse events in premarketing drug clinical trials has undergone sustained regulatory evolution in the USA. In that year, an investigational drug study for chronic hepatitis B virus infection was emergently stopped after a patient suddenly exhibited hepatic failure and lactic acidosis, which later developed, along with pancreatitis and peripheral neuropathy, in several others after drug discontinuation. Five patients eventually died, including three despite emergency liver transplantation. The drug's multisystem toxicity was not predicted by preclinical animal studies, with grave injury to human mitochondria subsequently implicated. A concerned US Food and Drug Administration (FDA) created a task force whose findings would have a lasting impact on the agency's thinking. In 1994, the FDA proposed to amend its investigational new drug reporting requirements largely based on task force recommendations for ways to enhance the likelihood that sponsors and investigators would consider investigational agents as a possible cause of serious adverse events mimicking the underlying disease or concomitant drug toxicity. Then, in its 1997 final rule for expedited safety reporting requirements for drugs and biologics, the FDA advised sponsors that such reporting of serious, unexpected clinical trial cases would be expected when "there is a reasonable suspected causal relationship between the investigational product and the adverse event (i.e., the causal relationship cannot be ruled out)." This last clause was codified into the suspected adverse drug reaction definition in the FDA's 2003 safety reporting requirements for drugs and biologics proposed rule. The negatively received suspected adverse drug reaction and proposed causality standard were not adopted in the FDA's 2010 finalized investigational new drug safety reporting regulations, the agency stating that "'reasonable possibility' means there is evidence to suggest a causal relationship between the drug and the adverse event." However, such new requirements as aggregate analysis of specific events and expedited reporting of animal or in vitro data suggesting significant harm to humans, and subsequent guidance that sponsors develop "a systematic approach" to premarketing safety assessment, are among the components of the FDA's efforts to enhance determination of a "reasonable possibility" of causality. They are also philosophically consistent with the 1993 task force recommendations, and a reminder of the inherent hazards associated with the use of investigational drugs, particularly in the early stages of human study.