By Judy Hanover
June 14, 2005 | The recent withdrawals of Vioxx, Tysabri, and other drugs have raised the public consciousness on the drug safety problem. Pharmaceutical companies, regulators, and doctors are actively seeking better ways to learn about safety problems earlier in the drug’s life. There is increased awareness of the need for improved detection efforts and disclosure of safety issues to protect patients from problem drugs.
Information technology, specifically safety databases and concomitant data-mining efforts, holds great promise for providing a potential solution to these daunting problems, despite lingering uncertainty. The opportunity to improve drug safety by applying technological solutions has made pharmacovigilance risk management one of the fastest-growing application areas in the drug development technology arena. In a recent report, Life Science Insights reviewed the available technology and found that despite a plethora of available technology solutions, true innovation remains elusive.
The ability to predict drug safety problems in clinical trials is the goal of adverse event reporting and pharmacovigilance risk management activities. Identifying the problem before the drug is marketed could prevent a catastrophic failure. However, premarketing studies are limited in their ability to predict adverse events. This is as a result of the small patient population, the narrow set of indications, and the short duration of clinical trial drug use. Statistically, only adverse events with an incidence of 1/1,000 or greater will be found during the course of a clinical trial with 3,000 participants.
It is clear, therefore, that current models for conducting premarketing studies will fail to detect all of the important adverse events for a drug before it goes on the market. Missed events can include those with a long latency, a small incidence, or an effect on a “real-world” segment of the population that was not studied. In the Vioxx saga, all three of these issues played a role. Once a product is on the market, the number and variety of patients exposed to a drug grow, increasing our ability to detect adverse events. The goal is to find problems yet keep the number of patients exposed to a minimum.
It is easy to blame the FDA’s reporting requirements, which may indeed be inadequate. The FDA has a limited ability to monitor the safety of marketed drugs; postmarketing reporting consists of spontaneous reports submitted voluntarily to the agency by doctors. Toughening up this reporting requirement would help, but there is a case to be made for simply working harder to get more information from the data available — here is where pharmacovigilance technology comes in.
Pharmacovigilance systems consist of a drug safety case information database with a varying set of tools and techniques for mining to arrive at the key conclusions or “safety signals” buried inside the massive data set. Drug safety data has many sources: preclinical studies, clinical trials, modeling and simulation technologies, post-launch adverse event reporting, and comparative drug class information. Risk management programs and Phase IV studies are increasingly providing useful data across real-life populations. Measurement of safety signals is complicated by reporting biases, the complex statistical models in use, the reliability of the providers reporting event data, and data quality and quantity issues.
Drug companies must carefully evaluate their safety investments and policies while leveraging technologies that allow them to make better decisions about drug safety. This can be enabled by technology such as that provided by companies including Lincoln Technologies, Galt Associates, Relsys, Aris Global, and Oracle. A recent Life Science Insights publication (“Adverse Event Reporting and Pharmacovigilance: Making Drug Safety Happen,” April 2005) reviews the existing application areas and their role in a pharmacovigilance strategy.
Both the FDA and pharma companies have a long way to go before they are effectively putting the tools into use to provide actionable, accurate, and easily understood information to the market. A detection strategy and procedures for handling safety signals are essential components to the strategic approach to pharmacovigilance.
Judy Hanover is a senior research analyst with Life Science Insights. E-mail: firstname.lastname@example.org.