Inclusion/exclusion criteria for drug studies can be overly restrictive

Key Points

Problems have arisen on the research horizon, however, because some potent systemic anti-inflammatories have caused elevations in serum triglyceride and cholesterol levels, as well as an increase in cardiovascular adverse events in testing. Some amazing drugs have been removed from further research based on these problems. This recent development has me reflecting on how drugs are developed and tested.

Safety (re)defined

In other words, drugs are studied in people who, outside of the one skin disease that is being addressed by the study drug, are considered healthy. But how many physicians use drugs under these conditions? Very few. It's no wonder that problems arise post-approval when the drug is used in a more representative population.

Perhaps one of the biggest impediments to accurate drug assessment is the rigorous inclusion/exclusion criteria. Is it possible to include the wrong subjects and exclude the right ones? Is it possible to bias a study for or against pharmaceutical performance based on poorly characterized inclusion/exclusion criteria? The answer to these questions is "yes." These are challenges that can only be overcome by mastering the art of clinical study design.

Additional limitations

Psoriasis studies seem to always specify subjects with a certain Psoriasis Area and Severity Index (PASI) score, percent of body surface area involvement and only plaque-type psoriasis. This means that the more challenging pustular, guttate and erythrodermic forms of psoriasis have no approved drugs and must be treated off-label. And insurance companies can deny treatment for these conditions.

There also are very few topical acne treatments approved for children age 9 or younger. What does the dermatologist do when an 8-year-old walks in with moderately severe inflammatory acne? How does malpractice look at the use of medications in situations for which they are not approved?

There is no doubt that inclusion/exclusion criteria are key to instructing investigators how to select the proper population for study. Research studies are more likely to yield positive results when the subjects are clustered with consistent disease. Questions arise when the inclusion/exclusion criteria restrict drug use in certain disease variants and in study populations not representative of those encountered in clinical practice, and allow insurance denial of off-label use in unstudied populations.

Maybe the constructs of research study design need to be reconsidered to meet the new undefined definition of "safe."