Elements of a drug safety disaster

Previous posts have taken on narrower topics such as adverse events, medical research and the early history of finasteride. Each needs its own treatment, because each has its own language and key actors. But the landscape remains fragmented. This post identifies common themes across the divides. Unusually, several themes concern what has been left out. The final theme steps back even farther, offering a grand unified view: that disinformation has been used as a weapon to maintain institutional power.

On to the themes:

Physicians are entangled with the pharmaceutical industry. Merck enlisted prominent physicians to support the development of Propecia as investigators, consultants and speakers. In exchange for their academic endorsement, these physicians received research grants, publication credits in top journals, and payments. More broadly, pharma companies have long sponsored physician groups such as the American Academy of Dermatology, advertised in journals and bought booths at conferences. Pharmaceutical sales reps cultivate relationships with physicians, pushing them to prescribe the latest drugs and leaving behind medication starter kits, branded pens and pads. These entanglements make physicians dependent on the industry, and discourage them from challenging a drug maker’s position.

Almost all clinical trial data on finasteride was sponsored by Merck. This set up a conflict of interest: the company knew safety concerns could jeopardize drug approval, while they were also in control of safety reporting. Today, physicians still recite the finding from 25 years ago, that fewer than 4% of men using finasteride had a sexual adverse event—even though this appears wildly out of step with patient experiences.

Safety reporting in studies by dermatologists is inadequate. Papers reveal that investigators focused on benefits, while giving lackluster attention to safety concerns. Dermatologists lack expertise in sexual function or neuropsychiatric status, so they may not be competent to assess adverse events in these categories. Because finasteride is prescribed by many dermatologists and attracts patients, new safety concerns would be unwelcome. This acts as a disincentive to do rigorous safety reporting.

Taboos and embarrassment surrounding sexual dysfunction prevent the airing of concerns in studies and in real-world use. Patients and physicians may both avoid the topic—particularly since dermatologists lack expertise in sexual medicine. These social forces are also in play in clinical trials, leading to understatement of sexual adverse events. The medical debate has generally ignored the role of social factors, instead attributing men’s concerns to psychological problems.

There is little funding or prestige in drug safety research. Research dollars tend to go towards new drugs and devices. Once a drug has been approved and used for some years, it is considered safe. Raising questions about safety might disconcert patients and physicians (for different reasons). Lack of funding leads to a lopsided evidence base, with an abundance of commercially funded studies and few independent studies on safety. Studies that do exist tend to be small, so they may be criticized as “low-quality evidence.”

The mechanism of finasteride is more disruptive than Merck described. Prescribing Information only mentions changes to two steroid levels (testosterone and dihydrotestosterone). In fact, the drug directly suppresses five other steroids, and has indirect effects on other steroids. These substances support the reproductive system and brain, which underlie the most common deficits in men with lasting harms of finasteride.

FDA has a conflict of interest. Since the agency approved finasteride in the first place, it may be reluctant to acknowledge safety concerns after approval. David Graham’s testimony about Vioxx described how a safety group within FDA was discouraged from communicating serious adverse events, including deaths, within the agency.

The post-drug disorder is unfamiliar. Lasting dysfunctions after stopping finasteride or dutasteride affect multiple systems and vary across individuals. The notion of a post-drug condition is unfamiliar (although Merck observed this outcome in its own trials). Perhaps due to paternalism and bias against young patients, physicians have interpreted lasting dysfunctions as a low testosterone condition, ordinary depression or psychogenic erectile dysfunction. Physicians concerned about liability may be reluctant to attribute the condition to a prescribed drug.

Medicine enjoys prestige and authority, but doctors also have a peculiar culture. Medical knowledge is managed conservatively, such that updating or changing consensus can take decades. There is often low tolerance for uncertainty, and a tendency to disbelieve symptoms without any diagnostic signs. Physicians may consider their own knowledge and judgment superior to the patient’s reported experience. There is a history of physicians diagnosing unfamiliar presentations as psychological conditions. Physicians have a history as a professional guild, which may lead to favoring professional interests over those of patients. Finally, alliances with the pharmaceutical industry can prevent the recognition of drug-related harms.

Elastic uncertainty is used to defend the status quo. This means playing uncertainty up or down to support the institution’s interests. For example, Merck cast doubt on reports of lasting harms after stopping finasteride by pointing to many other factors: pre-existing conditions, the lack of a plausible biological mechanism, lack of lab findings and more. Counterarguments were not considered. Meanwhile, they judged their own studies to be certain, despite obvious weaknesses. A true scientific approach considers all arguments and chooses those best supported by evidence and reasoning. In contrast, Merck, FDA and physicians have played up uncertainty selectively to support their respective interests.

We end with a grand unified view, focused on institutions and power. Permanent drug harms threaten the authority of drug makers, physician groups and regulators, for they are collectively responsible for drug safety. Faced with claims of severe and lasting drug harms, they have defended the system, drawing on standardized tactics such as ignoring the information, doubting it, undermining it, or attacking the messenger. Disinformation is thus used as a weapon to defend a position of power.

Merck, physician guilds and FDA share common interests. If one organization deflects a claim about drug harms, it preserves the alliance of three. However, if one of them takes drug harms seriously, it threatens the whole alliance. Training, culture and procedures ensure that individuals within each organization uphold the status quo. They are thus self-preserving bureaucracies. The playbook of defense tactics—ignore, deny, deflect, attack—becomes embedded within the culture of all three institutions. Nevertheless, 25 years of delay, disinformation and patient-blaming have failed to quash concerns over lasting harms of 5-alpha reductase inhibitors.


Further reading