Issue link: https://www.e-digitaleditions.com/i/492796
Tablets & Capsules April 2015 29 very tightly defined initial conditions, and that outside of these boundaries, the model no longer holds. Models, therefore, cannot tell us how nature works, only give us a glimpse into very small slices of approximate reality. That's why lab (in vitro) testing doesn't represent the con- ditions under which drug products treat patients (in vivo). That leaves us with two possible choices: Increase the allowable variance in IVIVC results or make dissolution and absorption kinetics part of the empirical evidence used for drug approval in Phase I and Phase II clinical trials. The outcomes If the hypothesis that models correlate with reality is flawed, we should expect to get exactly the results that we have: irregular, largely irreproducible, and not predic- tive over a very wide range. That raises a bigger philosophical question: If we are looking for causality—i.e., changes to the materials used to control release cause changes in dissolution in IVIVC test- ing—we're going the wrong way with our analyses. That's because, to date, the models have not accounted for very many variables, and in order to make the models perform better, we keep refining the initial conditions and control- ling more factors. In fact, we've reached the point where all we claim is a limited correlation; we have so severely lim- ited our data sets that divining causality is impossible. We can show that our model holds true in extraordi- narily limited circumstances. So what? That isn't helpful in generalizing the effect of different doses, controlled- release excipients, patient demographics (e.g., the effect of age on PK), etc. (Figure 2). Genetic and other differences between people and groups of people will always make a "standard" dose or con- trolled-release technology subject to wide variation in its effects. That's why the next paradigm shift within medicine will address how individual genetic profiles (genotypes) react to different environmental exposure. It's possible that, because of changes in our environment and exposures, what presented little risk a few decades ago may now pose signif- icant risk [4]. For example, perhaps the recent focus on high-protein, low-carbohydrate, and high saturated fat will reveal new genetic risks that were never anticipated prior to consumer fads that created a societal-scale experiment. It's one reason the term "personalized medicine" will continue to appear in the literature and the news. To answer the question of whether we'll ever attain truly personalized medicine will require us to conduct other thought experi- ments and test other falsifiable experimental designs. But I do know that if we could formulate better questions about IVIVC and other models, we'd get better answers. T&C References 1. Locwin, B. (2012). "Regulatory compliance: Patients talk," Tablets & Capsules. 10:7, p. 48. 2. Fisher, R. A. (1966). The design of experiments, 8th edition. New York, Hafner Publishing, p. 17. 3. Box, G. E. P., and Draper, N. R., (1987). Empirical model-building and response surfaces, John Wiley & Sons, New York, p. 74. 4 Locwin, B. (2015). "Is obesity rooted in your genes? Not exactly." Genetic Literacy Project: www.geneticliter- acyproject.org/2015/01/08/is-obesity-rooted-in-your- genes-not-exactly/ Ben Locwin, Ph.D., MBA, is president of Healthcare Science Advisors, 959 Maplewood Ave., Suite 2, Portsmouth, NH 03801. Tel. 603 397 7304. Website: www.healthcare- scienceadvisors.com. He writes and speaks on a variety of sci- entific topics, and he researches and consults for companies in the pharmaceutical, food and nutrition, and behavioral psy- chology industries, as well as academia. He last wrote for Tablets & Capsules' July 2014 issue. Figure 1 PK information that we wish dissolution testing could provide { MTC MEC Onset Duration of action Therapeutic range C max T max Concentration Time AUC Figure 2 Multilayered pellets for use in controlled-release tablets and capsules complicate IVIVC models Release-control polymer Protective layer or additional release control Drug layer Core granule or crystal