Newly Approved Does Not Always Mean New and Improved

Pharmaceutical manufacturers are in business to make money, and they make money by bringing new products to market. This statement is not a value judgment; it is a simple fact. Profits are a stimulus for innovation in the pharmaceutical industry, and innovation is the basis of progress and improvement in health care.

The most profitable products for these companies are “blockbuster” drugs, which are commonly defined as drugs that generate more than $1 billion in revenue annually. Selective inhibitors of cyclooxygenase type 2 (“coxibs”) and atypical antipsychotics are 2 examples of drug classes that produced blockbuster sales. Recent controversies around both drug classes highlight some of the difficulties inherent in balancing the promise of innovation and improved care against real-world safety and effectiveness. Much has been made about the need for reform of regulatory agencies to ensure that the public is protected from unsafe drugs while still being afforded timely access to efficacious new ones.^{1 - 2} In this Commentary, we argue that regulatory reform alone is an insufficient step toward improving drug safety.

Initial licensing will, and should, continue to be guided by experimental evidence gathered from selected patient populations with specific diseases. However, randomized trials are an artificial environment, and information about the benefits and harms of new drugs will be lacking for patients who account for a substantial proportion of sales: those whose disease severity or comorbidities would have excluded them from clinical trials and those with “off-label” indications. Rarely is such usage actively discouraged by the manufacturer; indeed, it is sometimes quietly or even overtly promoted.³ In such instances, all regulatory agencies can do is cross their proverbial fingers and watch for trouble.

The history of coxibs and atypical antipsychotics highlights how the road to blockbuster status often follows a different route from the one envisaged by regulators. Atypical antipsychotics were initially licensed for use in the treatment of schizophrenia. Very quickly, however, they became popular for treatment of the behavioral symptoms of dementia, although this indication has never been approved by regulators in the United States. Indeed, use of these drugs in patients with dementia accounts for a large proportion of total sales.⁴

Coxibs were initially licensed for treatment of patients with rheumatoid and osteoarthritis and subsequently for the treatment of acute pain. These indications overlap with those of nonselective nonsteroidal anti-inflammatory drugs (NSAIDs). Once coxibs were on the market, their prescribing increased rapidly, and although the use of NSAIDs decreased slightly, the combined use of older NSAIDs and coxibs far exceeded the original market for NSAIDs alone.⁵ These new agents did not just replace NSAIDs; rather, they seemed to create an entirely new population of patients treated with anti-inflammatory drugs.

The increase in use of coxibs and atypical antipsychotics share 2 common elements: concerns about existing therapies (their toxicities in particular) and aggressive marketing that capitalized on intuitive leaps in logic. If the analgesic effect of coxibs is equal to that of traditional NSAIDs and if coxibs have less gastrointestinal toxicity, then they are likely safer and this safety advantage should persist in patients who otherwise might not have been prescribed an NSAID. Similarly, if atypical antipsychotic drugs are more effective or better tolerated than conventional antipsychotic drugs in patients with schizophrenia, then presumably they are also preferable for the treatment of other conditions for which conventional antipsychotic drugs traditionally have been used, such as behavioral problems in patients with dementia.

The perception that new drugs are better than old ones is reinforced by the natural instinct of consumers that price is directly correlated with quality. New drugs—especially ones from novel classes—are priced much higher than older products, which often have generic formulations. The higher price is not a barrier to use by most consumers because third-party payers often cover the cost of these new products, providing the illusion that no one pays.

The combination of aggressive marketing, the apparent endorsement of the scientific community through publications in prestigious journals, the sense that price reflects the value of the innovation, and the tacit endorsement by insurers (“The new product must be better; why else would my drug plan pay for it?”) collectively fuel the perception of patients and prescribers that a newly approved drug is a new and improved drug.

Although the sales of coxibs increased very rapidly, a small but vociferous group argued that these drugs, rather than being safer than NSAIDs, might actually be more dangerous.⁶ In 2004 Merck voluntarily withdrew Vioxx from the market following the observation of an increased risk of cardiovascular events in a long-term placebo-controlled trial. Journals that had published articles describing the benefits of these drugs became concerned that they had an incomplete picture of their safety profile. Quite predictably, the media, legislators, and lawyers all subsequently got involved; fingers were pointed and accusations were made. It was argued that the serious cardiac adverse effects from at least 1 drug, given its widespread use, had caused thousands of deaths.⁷ Regulators were asked to step in and review the entire class.

In addition to their relatively unexpected cardiac toxicity, the coxibs, it turns out, were not devoid of gastrointestinal toxicity after all. Studies showed that the introduction of coxibs was associated with population-wide increases in hospitalization for gastrointestinal bleeding.⁸ Some of this unexpected surge in harm undoubtedly reflected the use of coxibs among patients who, by virtue of comorbidity, were poor candidates for traditional NSAIDs in the first place.

The story of atypical antipsychotic drugs continues to evolve but highlights some of the same issues and also raises new concerns. Several years following the increase of use of these agents for patients with dementia, regulators in Canada and the United States, in conjunction with drug manufacturers, released a series of warnings regarding serious adverse events—stroke and mortality—associated with the use of 3 most commonly used atypical agents in patients with dementia.⁹ Regulators were put in the decidedly awkward position of not only having to issue warnings about the drugs, but also having to point out to prescribers that these drugs were not even approved for the indication that was the subject of the warning.

Prior to these warnings, a systematic review of studies measuring the effectiveness of atypical antipsychotic drugs in patients with behavioral problems of dementia found little evidence of benefit.¹⁰ Others have shown that randomized trials in general may have failed to provide information about serious adverse events and mortality.¹¹ To date, most third-party payers have not changed their coverage for these drugs,¹² even though they cannot help but be aware that they are paying for drugs that are commonly being used for off-label indications and that carry significant risks.

Prescribers, patients, and their families now face a dilemma regarding the use of antipsychotic drugs. They have become increasingly comfortable with the use of atypical antipsychotic drugs: in Ontario, more than 30% of nursing home residents receive these drugs.¹³ How many of these patients truly need these medications is unknown, but that many of them do not is a virtual certainty.

In the face of evolving safety concerns over new drugs, what should physicians do? A huge new market is not easily undone. One option might be to replace the newer agents with older agents, but this might not solve the problem. For atypical agents, evidence suggests that the older conventional agents are at least as risky, if not more risky, than newer agents in terms of mortality.^{14 - 15} With coxibs, regulatory agencies have concluded that traditional NSAIDs also impart significant cardiac risks.^{16 - 17}

There are relatively few speed bumps on the road to blockbuster drug status. Many different groups drive the development of these new markets, and no single group really stands to lose. Physicians are happy because they feel up-to-date with the latest pharmaceutical innovations. Patients are happy because they feel they are getting the newest and best drugs. Pharmaceutical companies are undoubtedly happy because happy physicians and happy patients are good for their bottom line. Private third-party payers, while they strive to ensure value for money, simply diffuse the incremental costs among policyholders. Public drug insurance plans might resist covering new drugs but would do so at the risk of public outrage and negative publicity, so often they simply go along.

Given the “win-win-win” of an emerging blockbuster drug class, who is held accountable for the safety-related blame that can only become apparent with the benefit of hindsight? It is easy to lay blame at the feet of regulators, but they cannot do it all. They must make a dichotomous decision about approval for every new product on the basis of a limited amount of information at a fixed point in time. They should not be expected to provide guarantees of safety when drugs are used in different populations or for unapproved indications after licensing. However, requiring postmarketing studies could provide essential information about safety.

But if regulators are not solely responsible for real-world drug safety issues, then who is? Some suggest that authors and journal editors should do a better job of including data on risks of harm in published clinical trials.¹¹ Some blame manufacturers for overly aggressive marketing.³ Others say that third-party payers need to focus more on safety issues.¹²

There is another important perspective. Maybe part of the problem rests with physicians and patients. Patients must become more skeptical and better informed about new drugs before they take them. Prescriptions are written by physicians, who need to be more discriminating in how they prescribe new drugs. Specifically, physicians need a better understanding of the possible risks they expose their patients to when they prescribe drugs for unapproved indications or to patients different from those who helped generate early information about safety.

Changing the mindset of physicians and patients is no easy task, but one possible solution may lie in renewed investment in educational outreach programs (sometimes called “academic detailing” or “counter-detailing”): the delivery of impartial prescribing advice, so that physicians (and subsequently, their patients) can have a more balanced view of where a new drug really should fit into prescribing practices and what the relevant safety concerns might be. Outreach education is not a new idea¹⁸ ; it was previously proposed to help physicians understand that new drugs may simply be more costly without a commensurate clinical benefit. Today, it seems useful to expand the concept to safety concerns, including the use of case studies (such as the ones described in the commentary) to make the point.

The message for physicians and patients is simple: new and approved does not always mean new and improved. With every new drug, the safety profile is incomplete, and there is always more to come. At present, this message—as important and as irrefutable as it is—is seldom communicated, because there is little genuine interest in saying it until it is too late. This needs to change.