(233ap) Prediction of Formulation Effects of Continuously Formulated Solid Oral Dosage Forms on Bioavailability Using PBPK Modeling | AIChE

(233ap) Prediction of Formulation Effects of Continuously Formulated Solid Oral Dosage Forms on Bioavailability Using PBPK Modeling


Hartmanshenn, C. - Presenter, Rutgers University
Androulakis, I. P., Rutgers, The State University of New Jersey
Ierapetritou, M. G., Rutgers, The State University of New Jersey
Sinko, P. J., Rutgers University
The Quality by Design (QbD) paradigm posits that in depth understanding of the relationship of a pharmaceutical productâ??s Critical Quality Attributes (CQAs), ingredient (including API) Critical Material Attributes (CMAs), and Critical manufacturing Process Parameters (CPPs), to the in vivo characteristics of the product as defined by the appropriate Quality Target Product Profile (QTPP) and Clinically Relevant Specifications (CRS), including absorption/bioavailability, will lead to better products more rapidly and economically for the benefit of the patient [1]. As stated, the QbD paradigm essentially requires establishing reliable quantitative relations between the â??inputâ? to the system (CMAs and CPPs) and the surrogate outputs (CQAs), which are presumed to be predictive of the â??real outputsâ?, (CRS). Ideally, the relationships between CMAs, ingredient ratios, and CPPs to highly predictive CQAs would be established a priori without extensive experimentation, especially in vivo, or substantial time/cost investments. For establishing such quantitative relations, we need to a) accurately, effectively and efficiently tune CQAs, ideally using predictive relationships to the ingredient proportions, CMAs and CPPs; and b) accurately, effectively and efficiently predict CRS from the CQAs (and perhaps, also using information from CMAs and CPPs).

Meeting this target has remained elusive. In recent years, two major advances have made a significant impact. First, Continuous Manufacturing (CM) has clearly demonstrated itself as a superior alternative to batch processing allowing for tight control of product specifications, thus making it possible to target specific CQA. Second, physiology-based mathematical models of dissolution and absorption are becoming increasingly more reliable and dependable for connecting in silico CQA and CRS. However, despite previous studies linking in silico CQA and CRS, we still lack a clear demonstration of an integrated framework that considers the overall â??supply chainâ? of drug manufacturing, starting from the API, manufacturing process parameters, ingredient proportions and CMAs, and that eventually leads to in vivo predictions of the biorelevant profile. The lack of such integrated formalism a) creates sub-optimal designs and delivery systems; and b) does not take full advantage of the opportunities offered by CM. We propose to develop, implement, and demonstrate a systematic computational framework for connecting CAQ with CRS in silico in the context of continuous solid oral dose manufacturing processes.

More specifically, we explore computational methods for linking formulation-dependent dissolution characteristics to in vivo bioavailability. We achieve this aim in two main steps: (1) using physicochemical and processing parameters we predict in vivo dissolution from formulation and process parameters; and (2) using the predicted dissolution profile along with physicochemical and physiological properties we predict in vivo bioavailability. These two steps are accomplished by making use of DDDPlusTM and GastroPlusTM respectively. DDDPlusTM simulates the dissolution curve of a drug product based on parameters such as diffusivity, solubility, particle size, and pKa, along with tablet properties such as composition, compression force, and tablet dimensions. The second step consists of feeding the output dissolution curve from DDDPlusTM into a GastroPlusTM model. GastroPlusTMis based on the ACAT (Advanced Compartmental Absorption and Transit model) which has the ability to simulate the expected systemic bioavailability of a drug in the form of a plasma versus time graph [2, 3].

We hypothesize that since formulation regulates the parameters inputted into the dissolution model, for which the output profile serves as an input into the absorption model, it should have a direct impact on simulated bioavailability. An important part of this work will thus consists of quantifying the extent to which formulation can affect bioavailability. To aid in this endeavor, a parameter sensitivity analysis (PSA) will be performed to establish the parameters within the manufacturing design space that have the highest impact [4]. Furthermore, the significance of process variability on bioavailability will be studied in order to determine optimal operation range and to map out a design space for a given target bioperformance.

The advantage of mathematical modeling is the ease with which useful information can be produced compared to more labor-intensive and costly experiments. We seek to demonstrate the tremendous potential of PBPK models in guiding pharmaceutical processing development for oral tablet manufacturing within a Quality by Design framework. Future work also involves utilizing this mechanistic system to advance personalized medicine. By varying physiological properties, the model can be made to account for age, sex, ethnicity, extent of disease, and circadian rhythms, an aspect which could significantly enhance the customizing of drug formulation for specific populations.

1. Kesisoglou, F. and A. Mitra, Application of Absorption Modeling in Rational Design of Drug Product Under Quality-by-Design Paradigm. AAPS J, 2015. 17(5): p. 1224-36.

2. Huang, W., S.L. Lee, and L.X. Yu, Mechanistic approaches to predicting oral drug absorption. AAPS J, 2009. 11(2): p. 217-24.

3. GastroPlus User Manual - Simulation Software for Drug Discovery and Development, Cognigen, Editor. 2015, Simulations Plus.

4. Zhang, X., et al., Utility of physiologically based absorption modeling in implementing Quality by Design in drug development. AAPS J, 2011. 13(1): p. 59-71.