We all want safer drugs, but we normally settle for much less. Aside from medication errors, poor compliance, and manufacturing problems, the very mechanisms that medicines employ to heal patients frequently harm them instead.
In the end, safety problems account for most pharmaceutical development failures and product withdrawals, even when the drugs prove efficacious for the majority of patients taking them. Moreover, the high number of safety-related failures points to deficiencies in the methods, assumptions, and protocols companies use in drug development itself.
Now, for all of those reasons and more, companies are facing, and in some cases implementing, major changes in the status quo of drug safety.
Could the safety level of new drugs in general be significantly raised above current standards and expectations? What are some ways in which drug safety could be improved? What are obstacles to improvement? How great a change would a higher safety standard cause to the quality, impact, and innovation of new drugs? The following offers some answers — and proposes a collaboration of industry, academia, and regulation to achieve a “great leap forward” in the general level of new drug safety by quantifying, measuring, and rating drugs’ risks versus their benefits.
A New Space
Although a drug’s adverse effects may be well-known, doctors and patients typically have no firm basis for deciding whether its benefits warrant the risks of using it — or for comparing those risks to other everyday dangers from driving, eating, bicycling, and so on. Though beyond the main focus here, the problem extends to other therapies and procedures, and even over-the-counter medicines, vitamins, and supplements. Most often, as with radiotherapy, safety is only weakly defined as doing less or no more damage than a previous or comparable therapy.
Meanwhile, Big Pharma companies are seeking ways to “de-risk” their R&D investments, and more than anything else, such economic concerns may push them toward placing a higher priority on new-drug safety. Many, perhaps most life science companies, are facing new safety hurdles. The change is also evident in the emergence of a safety space — a sector of small companies focused on drug safety. Examples include Hepregen and the Drug Safety Alliance. Other vendors, such as CROs, conduct comprehensive safety programs, and niche players contribute pieces of regulator-required risk-management packages. (Examples include Apogenics, ParagonRX, United BioSource Corporation, BioTrak, and Gigamoto Technology Partners.)
Risk depends on knowledge. “Without accurate and predictive information about a drug, risk of ‘false negatives’ may be high,” says Bernadette (Bonnie) Fendrock, president and CEO of Hepregen. Her company is developing a highly sensitive in vitro cellular model for preclinical testing to detect liver-toxicity signals in drug compounds. But Fendrock also sees the need for addressing safety on a wider front. “Although meaningful to each, the safety issue is addressed differently by various constituents: regulatory agencies and internally by scientists. In addition, top levels of management need to provide incentives for increasing safety just as they do for improving the drug pipeline.”
Major drug-safety improvement must simultaneously progress along two major pathways: drug development and regulation. In development, companies will need to establish a “safety culture” that begins with an early focus on preclinical models and better early-stage trial design. Challenges include sifting safety signals from disease “noise” in patients, the cost of early-safety testing, and replacing the traditional sales-target approach with assessments of “therapeutic quality” in the selection of development candidates.
Dr. Ashraf Youssef, associate medical director, pharmacovigilance, at Takeda Global Research & Development Center, says, “Models of benefit/risk used in the economy can be applied to development of drugs or just making decisions in how you address the healthcare issue with the clues and the treatments you have available.”
Some models for safety-focused development exist. Wyeth introduced the Medical Differentiation Index (MDI) for its development products, which Pfizer has evidently adopted from its acquisition. MDI gives a high priority to safety as a key criterion, along with efficacy, dosing, tolerance, and convenience. MDI’s assessment remains relative to the existing standard of care and competitive products. Still, it represents a promising trend inside companies that may parallel new thinking on the regulatory side.
Regulators must also create a new culture that shifts organization and resources to the evaluation of benefit/risk and also emphasizes scientific review over enforcement. Challenges include overcoming tradition, political conflict, and general disharmony among major industry players. Both the U.S. and European regulatory authorities have drug-safety programs under way: The FDA launched RiskMAPs for pharmaceuticals in 2005, and, as directed by the FDAAA (Food and Drug Administration Amendments Act of 2007), it followed with REMS (risk evaluation and mitigation strategies). European Medicines Agency (EMA) previously created a similar program called EU-RMP (EU Risk Management Plan). REMS and EU-RMP consist of company-supplied information and procedures for professionals and patients designed to minimize typical therapeutic side effects. REMS are required, not optional, and they have spawned a new, REMS-dedicated discipline in some companies.
Dr. John Balian, senior VP of worldwide safety and regulatory operations at Pfizer, and two coauthors from BMS advise a full-development approach to REMS. (See “Further Reading.”) “The drug approval process is increasingly dependent upon the development of risk-management plans that ensure appropriate prescription behaviors and balanced risk levels,” they say. “Although these requirements place new responsibilities on pharmaceutical companies, innovative and effective designs of regulatory applications and commercialization plans that ensure optimal usage can help mitigate the associated burdens and even help secure benefits for affected products and patients.”
REMS and similar programs mainly address safety in the usage of drugs, however; they do not necessarily encourage the development of safer drugs. Still, the REMS-imposed discipline can draw companies closer to the cross-functional collaboration and coordination needed to detect safety “signaling” from the earliest development stages on, using data sources that include but also augment pharmacovigilance. Now, some propose that regulators step beyond the “use-intervention” approach of REMS by reforming the way they review and authorize all drugs (and more). One group of industry leaders, collaborating with Dr. Robert Klitzman, director of the master of bioethics program at Columbia University, is exploring ways to start a broad-based conversation on reforming the FDA that includes new and better ways to manage benefit/risk assessment. “We need to look at every aspect of how the FDA can regulate drugs in ways that encourage innovation,” says Klitzman. “But safety is a huge issue, one of the biggest issues, in FDA reform.”
For both development and regulation, the key to improving safety is first to define it. Safety standards are shaped by benefit/risk; safety is qualified by level of efficacy, so to define one requires defining the other. A sufficiently broad-based consortium could identify the risks that should be ameliorated — along with the risks of amelioration itself. It could also determine, based on historical data and experience, the reduction in risk now possible to attain, as well as realistic but aggressive targets for the future. A key goal might be the quantification of benefit/risk, a numerical array similar in some respects to the Oregon Health Plan calculations for reimbursed procedures in Medicaid, though focused entirely on patients’ response to treatment.
A member of Klitzman’s group, Dr. Llew Keltner, president of Novici, believes the FDA should move toward a formal benefit/risk index: numerical ratings for drugs and other treatments and procedures the agency reviews. A public panel would assign, periodically review, and when justified by new data, revise the rating numbers, which would be included in labeling and published for all approved uses. “The benefit/risk numbers would define the safety endpoints for drug development,” Keltner says. “The task then would be to define the in vitro, preclinical, and early-stage clinical surrogates for approximating the real-world targets. A company could decide if it wanted to take the sales and reimbursement risk with a drug, given its assigned ratio, and then the consumer would need to accept the responsibility for deciding how much risk to take in treatment.”
Quantifying medical benefit and risk may not be easy. Currently, the concepts are largely subjective. Risk in particular means different things to different players: patients, physicians, pharma companies, payers, and of course regulators. Even within the FDA, benefit/risk assessments from two different auditors differ considerably, notes Klitzman. But Keltner maintains that a more objective benefit/risk rating system is at least technically feasible — and necessary for objective assessment.
“Define the risks, define the benefits, use some reasonable method to measure and rank or prioritize them, systematize it, and then go,” Keltner says. “There is a very big field of research and literature on benefit/risk analysis — plenty of very well-tested models for how to quantify benefit/risk.”
Safety in Practice
Adoption of a B/R (benefit/risk) index or comparable reforms in drug development and regulation would have tremendous effects on clinical practice. Prescribing decisions and reimbursement would inevitably depend largely on each product’s rating, perhaps as the most important component of its labeling. That means life sciences companies, from large to small, would be speaking the benefit/risk language and thinking about reimbursement from the earliest development stages on.
“Payers need ways of measuring value that are widely accepted as valid and appropriate,” says reimbursement expert Howard Tag of Tag & Associates. “Right now, there is no such metric. Cost per quality adjusted life year is the best we now have, but it is not widely accepted. If a B/R index could be converted into a value index, that could be a new, more valid, and appropriate measure.”
Tag reasons that health insurance coverage could then be structured to link premiums with the B/R index. A lower premium would cover only higher-index treatments. “In a simplified, less quantitative way, that’s what happens when the least-expensive Medicare Part D plans cover only generic drugs (high benefit/low risk).”
Future discussion must surely address how such a system would affect the decision-making dynamics among all players. But, if the system truly focuses on treatment risks and benefits to patients, it would logically shift power toward them and those who treat them.
The B/R index is a bold proposal and is guaranteed to face stiff opposition. The FDA’s response to our query on the proposal was less than encouraging: “As the benefit/risk framework develops, there may be aspects of benefit/risk decision making that might be appropriate for quantification to aid in the decision-making process. We do not think that all of the considerations that go into benefit/risk assessments can be reduced to something like a B/R index. We, CDER [Center for Drug Evaluation and Research], do note that we have been developing a more structured approach to benefit/risk assessment that approaches this topic mostly from a qualitative perspective.”
The FDA’s Critical Path Initiative (CPI) in some ways fits the pattern of a broad-based collaboration. CPI is an admirable and even essential outreach by the FDA, but it may not be the ideal forum for a balanced discussion among all key players. CPI highlights new technology and tools to improve evaluation of typical safety parameters, particularly if the tools can replace or speed human trials. Yet the new tools, however necessary, are only a part of the larger picture. A broad, independent coalition would be free to set anything and everything on the table, including proposals FDA may now consider impractical or unwise, to define safety in a more objective way.
Without changing anything else, making drugs safer would be a revolution in itself. But to accomplish that goal alone will require something like a constitutional change — a formal commitment to raising the safety standard dramatically. Perhaps it is time to envision a “Constitutional Convention on Drug Safety,” calling together all concerned players to explore, debate, caucus, and hammer out the new standards and structure for overturning the old status quo. Hopefully, a convention could emerge out of initiatives already under way in related areas such as risk mitigation. (See “Safety Programs, Initiatives & Collaborations.”) But new leadership may also arise, calling for a great leap forward.
Roadmap to Risk Evaluation and Mitigation Strategies (REMS) Success, John D. Balian, Janice C. Wherry, Rachpal Malhotra, and Valerie Perentesis
Pharmacogenomics of Adverse Drug Reactions: Practical Applications and Perspectives, L. Becquemont
Adaptive Trials and Bayesian Statistics in Drug Development, D.A. Berry
Extreme Measures: New Tactics Required in Drug Value Differentiation, M.D. Christel
Outcomes Research Collaborations Between Third-Party Payers, Academia, and Pharmaceutical Manufacturers. What Can We Learn From Clinical Research?, H.G. Eichler, S.X. Kong, and J.P. Grègoire
Safety Programs, Initiatives, Collaborations
REMS (Risk Evaluation and Mitigation Strategies)
FDA ETASU (Elements to Assure Safe Use)
FDA Safe Use Initiative
EU-RMP (EU Risk Management Plan)
FDA RiskMAPs (Risk Minimization Action Plans)
FDA Critical Path Initiative
FDA Sentinel Initiative
DAMOCLES Study Group (NHS Health Technology Assessment Programme)
PSTC (Predictive Safety Testing Consortium)
iSAEC (International Serious Adverse Event Consortium)
OMOP (Observational Medical Outcomes Partnership)
PhRMA Benefit-Risk Assessment Team
ASTER (ADE Spontaneous Triggered Event Reporting) — Brigham and Women’s Hospital, Partners Healthcare, CDISC, CRIX, Pfizer
The Erice Manifesto (Europe)