Raising a Red Flag: Deficiencies Found in the Reporting of Important Parameters in FDA Approved Drug Profiles

Physicians have an innate tendency to trust in governing medical entities like American Medical Association (AMA), Food and Drug Administration (FDA) and Centers for Disease Control (CDC). Maintaining a reproducible and sustainable standard for reporting clinically relevant pharmaceutical data for new and established drugs has always been a priority for the FDA.

At the very end of the 20th Century (in December 2000), the  FDA issued a call to improve the professional labeling that is found in the drug package inserts and other accredited drug information reporting agencies such as the Physician Drug Reference (PDR). In 2006, the standard format for this package insert was modifiedin an attempt to make it more user-friendly and to serve as an efficient resource tool for patients, physicians, and researchers. In today’s fast-changing challenging healthcare environment, physicians tend to accept this reported data, and the lack of it, as this is all there is available to them to treat their patients.

Being able to manipulate the dose for a patient with renal failure based on given elimination half life is an excellent adjunct to the management, yet, is not an option on many of the FDA approved drugs due to deficiency in the reporting of this parameter in the drug inserts. Furthermore for each drug profile, the new FDA reforms allow for major differences in the reporting methods for a set of properties provided by the different manufactures and there has not been much assessment of the pharmaceutical company’s compliance with FDA guidelines. Also, the Investigator’s Brochure (IB) supposedly provides guidelines that describe the necessary reporting requirements for clinical and non-clinical pharmacokinetic, pharmacological, metabolism, toxicological, drug efficacy, and safety data needed to provide guidance for a clear understanding of the risks and adverse drug reactions during drug use. Yet, as reported by us and a number of groups, deficiencies in pharmacokinetics reporting is evident in many of the FDA approved drug packaged inserts.

The United States Government Accountability Office (GAO) recently concluded that the FDA lacks the authority for requiring certain studies and is deficient in the necessary resources for obtaining post-market drug safety data. Nonetheless, it is always the challenging question of research scientists that makes a difference.

Four years ago using a multifaceted approach to train undergraduates in bioinformatics and chemometrics, we began using the web-accessible FDA approved drug files to extract a drugs chemical, pharmacokinetic, and toxicological data. Initially we evaluated 75 randomly selected common consumer drugs and created a FDA Common Consumer Drug Database© using the commercially available KnowItAll® informatics platform from Bio-Rad Laboratories. Recently using a similar impartial randomising method, we built a Cancer Drug Database© containing 85 chemotherapy drugs. The reporting of important biological properties such as bioavailability (BIO), plasma protein binding (PPB), elimination half-life (t1/2), volume of distribution (VD), and water solubility were found to be markedly deficient. The evidence is so compelling that the apparent laxity in enforcing reporting regulations for many of these pharmaceuticals is alarming.

The FDA and other agencies have repeatedly refined their regulations to ensure comprehensive reporting for the above mentioned as well as other parameters, yet, and as mentioned in this article, specific guidelines for reporting pharmaceutical data is kept lenient.

During the past decade, quantitative structure-activity relationships (QSAR) have proved to be very effective in finding the underlying relationship between molecular structure and pharmacokinetic or toxicological properties. In other words, variation in biological response can be expressed as a function of the variation in the chemical composition. Hence during the testing and evaluation of the two drug databases, we were able to use commercial models developed by ChemSilico, Syracuse Research, and Strand Genomics, all available within the KnowItAll® platform, to predict some of the missing data in drugs containing a common chemical structural motif. In trying to validate the models predicting ability, we found that possible reasons for the outliers could be sought by evaluating consistencies in physiological properties (such as BIO and PPB) among the drugs sharing the same functional group.

Perhaps the real impact of this research and similar studies is the red flags they raise and the ensuing prospects that emerge. The “given” taken for granted by physicians does not have to be accepted with unquestioned indifference by the FDA. The guidelines need to be revisited with emphasis on extreme vigilance in reporting all parameters that directly or indirectly affect patient care.

On the other hand and as clearly alluded to in this article, the concomitant advent in technology will make it easier for pharmaceutical companies as well as ancillary research groups to accurately report or at least estimate most of the deficient data. Additionally, undergraduates trained in multidisciplinary research programs to interpret drug data in such emerging environments and skilled in organic/medicinal chemistry, human physiology, data management, analytical problem solving, and communication skills, would greatly benefit as the demand for researchers with drug metabolism expertise increases in the fields of pharmacy, medicine, and medical technology.

Acknowledgements

The Wesley College research programis supported by grant number 2 P2O RR016472-10 from the National Center for Research Resources (NCRR), a component of the National Institutes of Health (NIH). This IDeA Network of Biomedical Research Excellence (INBRE) grant to the state of Delawarewas obtained under the leadership of the University of Delaware, and the authors sincerely appreciate their efforts.