In Vivo Predictive Dissolution (IPD) and Biopharmaceutical Modeling

Subscriber access provided by Fudan University

Communication

In Vivo predictive dissolution (IPD) and biopharmaceutical modeling and simulation: future use of modern approaches and methodologies in a regulatory context H. Lennernas, A. Lindahl, A. Van Peer, C. Ollier, T. Flanagan, R. Lionberger, A. Nordmark, S. Yamashita, L. Yu, G. L. Amidon, V. Fischer, E. Sjögren, P. Zane, M. McAllister, and B. Abrahamsson Mol. Pharmaceutics, Just Accepted Manuscript • DOI: 10.1021/acs.molpharmaceut.6b00824 • Publication Date (Web): 14 Feb 2017 Downloaded from http://pubs.acs.org on February 15, 2017

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Molecular Pharmaceutics is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 27

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Molecular Pharmaceutics

In Vivo predictive dissolution (IPD) and biopharmaceutical modeling and simulation: future use of modern approaches and methodologies in a regulatory context. Lennernäs H1; Lindahl A2: Van Peer A3; Ollier C4; Flanagan T5; Lionberger R6; Nordmark A2; Yamashita S7; Yu L8, Amidon GL9; Fischer V10; Sjögren E1, Zane P4; McAllister M11; Abrahamsson B12 Affiliations: 1 Department of Pharmacy, Uppsala University, Uppsala, Sweden 2 Medical Products Agency, Uppsala, Sweden 3 Janssen Research and Development, Janssen Pharmaceutica NV, Turnhoutseweg 30, 2340, Beerse, Belgium. 4 Sanofi US, 55 Corporate Drive, Bridgewater, New Jersey, USA 08807 5 AstraZeneca R&D, Macclesfield, UK 6 Office of Research and Standards, Office of Generic Drugs, Center for Drug Evaluation and Research, Food and Drug Administration, 10903 New Hampshire Ave., Silver Spring, Maryland, 20993, USA. 7 Faculty of Pharmaceutical Sciences, Setsunan University, Nagaotoge-cho 45-1, Hirakata, Osaka 5730101, Japan 8 Food and Drug Administration, Center for Drug Evaluation and Research, Silver Spring, MD, 20993, USA 9 College of Pharmacy, University of Michigan, Ann Arbor, Michigan 48109-1065 10 Drug Metabolism and Pharmacokinetics, Research & Development, AbbVie, North Chicago, Illinois 60064 11 GlaxoSmithKline, Pharmaceutical Development, New Frontiers Science Park, Harlow, Essex CM19 5AW, United Kingdom. 12 AstraZeneca R&D, Gothenburg, Sweden

*Address correspondence to: Hans Lennernäs, PhD Professor in Biopharmaceutics Department of Pharmacy Uppsala University Box 580 SE-751 23 Uppsala, Sweden Email: [email protected] Phone: +46 – 18 471 4317 Fax: +46 – 18 471 4223

ACS Paragon Plus Environment


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Table of Contents


Page 2 of 27

Page 3 of 27

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Abstract The overall objective of OrBiTo, a project within Innovative Medicines Initiative (IMI), is to streamline and optimize the development of orally administered drug products through the creation and efficient application of biopharmaceutics tools. This toolkit will include both experimental and computational models developed on improved understanding of the highly dynamic gastrointestinal (GI) physiology relevant to the GI absorption of drug products in both fasted and fed states. A part of the annual OrBiTomeeting in 2015 was dedicated to the presentation of the most recent progress in the development of the regulatory use of PBPK in silico modeling, in vivo predictive dissolution (IPD) tests and their application to biowaivers. There are still several areas for improvement of in vitro dissolution testing by means of generating results relevant for the intraluminal conditions in the GI tract. The major opportunity is probably in combining IPD testing and physiologically based in silico models where the in vitro data provide input to the absorption predictions. The OrBiTo project and other current research projects includes definition of test media representative for the more distal parts of the GI tract, models capturing supersaturation and precipitation phenomena, influence of motility waves on shear and other forces of hydrodynamic origin, addressing the interindividual variability in composition and characteristics of GI fluids, food effects, definition of biorelevant buffer systems and intestinal water volumes. In conclusion, there is currently a mismatch between the extensive industrial usage of modern in vivo predictive tools and very limited inclusion of such data in regulatory files. However, there is a great interest among all stakeholders to introduce recent progresses in prediction of in vivo GI drug absorption into regulatory context. Keywords: Biopharmaceutics classification system, in vivo predictive dissolution, biorelevant dissolution, theoretical modelling, PBPK, bioequivalence, intestinal drug absorption, generic drugs



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Introduction The Innovative Medicines Initiative (IMI) is Europe's largest public-private initiative in the life science sector between the European Union (EU) and the European pharmaceutical industry association (EFPIA). The overall aim of IMI is to shorten development time of better and safer medicines for patients by creating multidisciplinary networks of industrial and academic experts 1. The five year OrBiTo project (www.orbitoproject.eu) passed mid-term evaluation in April 2015 and has been one of the more successful IMI programs based on its high impact scientific output 2. The overall objective of OrBiTo is to streamline and optimize the development of orally administered drug products through the creation and efficient application of biopharmaceutics tools. This toolkit will include both experimental and computational models developed based on improved understanding of highly dynamic gastrointestinal (GI) physiology relevant to the oral absorption of drug products in both fasted and fed states. . The OrBiTo project has identified, described and reported knowledge gaps in several key areas related to GI drug absorption and has started to deliver novel framework for rational application of in vivo predictive dissolution (IPD) as well as other biopharmaceutics tools for oral drug delivery 1, 3-6. Thus, it is expected that the outcome from the OrBiTo project will shift industrial product development from an essentially empirical approach to a more rational model-based approach. In addition, OrBiTo aims to collaborate with regulatory agencies to provide the basis for revised regulatory guidelines in the context of Quality by Design (QbD). This will be achieved by providing reliable biopharmaceutics links between the combination of active pharmaceutical ingredient (API) and its formulation, and to clinical performance enabling more clinically relevant specifications and extension of possibilities to waive in vivo bioequivalence studies based on in vitro/in silico methods that are predictive. The OrBiTo research program is based on four separate work packages (WP1-4) with different focus: (WP1) development of suitable biopharmaceutics tools for API characterization, (WP2) designing and characterization of the formulation, (WP3) increased understanding of physiological conditions in the GI tract and (WP4) optimization of absorption related processes included in physiologically-based pharmacokinetic (PBPK)


Page 4 of 27

Page 5 of 27

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


models, respectively. A key objective for OrBiTo is to evaluate the predictive accuracy of available in silico and in vitro biopharmaceutical models and to provide validation data in different dosing scenarios to increase their use in regulatory submissions and decision making throughout drug development process. During the 2015 annual OrBiTo meeting hosted at Sanofi, Chilly-Mazarin, Paris, one afternoon was dedicated to the presentation and discussion of the most recent progress in the development of the regulatory use of PBPK in silico modeling, in vivo predictive dissolution tests and their application to biowaivers. The primary objective of this regulatory session was to publicly review opportunities and progress with representatives from regulatory agencies from US, EU and Japan as basis for continued work in OrBiTo. These research performed in OrBiTo are envisaged to have a future regulatory impact globally and to stimulate similar multidisciplinary research activities. This scientific report is based on the presentation at the meeting and the discussions generated during the session and scientific program is presented elsewhere (http://www.orbitoproject.eu/node/113).

Review of the presentations. Dr. Anders Lindahl (Swedish Medical Products Agency, a member of the OrBiTo consortium) started the regulatory part of the annual OrbiTo meeting by focusing on the future trends on regulatory sciences and raised timely issues such as: i) should in vitro drug product dissolution rate profiles in biorelevant media be considered more trustful, ii) or on physiologically-based biopharmaceutics (PBBP) and pharmacokinetics (PBPK) modelling and simulation as future approaches in regulatory risk assessment and decision making? It was emphasized that regulatory discussions and interactions will facilitate the development and application of biopharmaceutics tools with regulatory implications. In particular there is a consensus that regulatory discussions and interactions are very important for the progress of the OrBiTo project and accordingly the OrBiTo project has an associated external regulatory stakeholder group.



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Dr. Talia Flanagan (Astra Zeneca) and Dr. Achiel Van Peer (Janssen) presented the results from recent survey performed among OrBiTo’s EFPIA members7. The survey focused on the current use of in vivo predictive in vitro and in silico biopharmaceutics tools during drug development and in regulatory interactions and decision making. The objectives of the survey were threefold: i) to capture todays practice and experience in applying in vitro and in silico biopharmaceutics tools in a regulatory setting at different stages of pharmaceutical development, ii) to identify what tools from OrBiTo will provide particular value in regulatory setting, and finally iii) to provide a benchmark to determine if there is any increased use of these biopharmaceutics tools at the end of the OrBiTo research period. The survey was sent out to all 13 EFPIA partners in May 2015 and the response frequency was 85%. At the Paris OrBiTo meeting preliminary results from 10 responders were presented. The full results of the survey have since been published 7. The survey revealed that biorelevant dissolution testing methods in simulated dissolution media and PBPK are widely used in drug and formulation development in the EFPIA companies of OrBiTo to further improve the understanding of the interplay between various biopharmaceutics factors. In Figure 1 the frequency for the various rationales for using biorelevant dissolution methods is displayed (Figure 1); this highlights that biorelevant dissolution testing is used to address a variety of biopharmaceutical issues during development, including to support key decisions such as formulation selection and optimization. When dissolution testing is first initiated, 80 % of the companies use biorelevant media (SGF, FaSSIF, FeSSIF) as a first step prior to any dissolution testing in simplified buffers for biopharmaceutics classification system (BCS) class II and IV APIs. The survey also highlighted that these data are rarely submitted to regulators (about 70% of companies seldom or never submit this data at the IND stage, and 60% at NDA/MAA). The potential usefulness of biorelevant dissolution media as a part of the quality control (QC) product release testing in manufacturing was also considered. This has the potential to address known issues with current QC dissolution methods that can reduce their biorelevance, such as the need to add non-physiological amounts of surfactant to QC media for poorly soluble compounds (BCS II and IV). However, even if the increased use of biorelevant methods in manufacturing can be supported from the scientific principles, there is a need to resolve


Page 6 of 27

Page 7 of 27

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


practical challenges such as variability and reproducibility, which are related to the instability of the colloidal structures in biorelevant media. The use of biorelevant dissolution-absorption PBPK/PBBP modeling and simulation (M&S) at different stages of drug development was also investigated in this survey. In early development it was most commonly used to examine the impact on particle size on GI drug absorption and to bridge between different formulations (88%). It is fairly common, with around half of the survey respondents using this approach during BA/BE study optimization and to propose an in vitro dissolution strategy. At an early IND stage, it was also clear that a large proportion (83%) of the companies who responded choose to adopt a PBPK/PBBP-based modelling and simulation strategy to aid in BE waiver discussions and relaxing the ranges for clinically relevant specifications. It was also clear that in silico biorelevant dissolution-based absorption M&S was rarely used in marketing applications (such as MAA and NDA). When the approach was used it was to demonstrate BE across formulation concepts, establish API and formulation design space and manufacturing controls, anticipate post-approval manufacturing changes and to obtain biowaivers. In this EFPIA investigation the challenges in regulatory biopharmaceutics with highest priority were identified. There is a need for industry to share more examples of these in vitro and in silico predictive approaches with regulators such as clinically relevant specifications; IVIVC/R for IR products; ability to estimate human intestinal permeability. Accordinglythis will lead to improved feedback from regulators for the applicants on the acceptability of these approaches and how they can be used in regulatory decision making. Other issues of common interest are the regulatory policy regarding arguments on clinical relevance for dissolution acceptance criteria outside the traditional f2 similarity test, and model acceptance criteria for in silico models (PBPK modelling guideline) and how to use these models together with biorelevant tools to establish biowaivers.



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

There are several significant barriers to use of biopharmaceutics predictive tools (in vitro, in silico) in regulatory submissions. It is a new scientific field where expertise and validation to integrate all the elements is still limited. There is no clear regulatory framework and there is a need to increase the acceptability of using the models in submissions. Determining how to incorporate in silico modeling and biorelevant dissolution within regulatory risk assessments is of major importance for advancements in this field. It is also clear that the accuracy of prediction of many of the biopharmaceutical tools needs to be improved and that further work is needed to understand the criticality of input parameters or there is a lack of data supporting certain processes (such as human intestinal permeability (Peff). Further achievements and increased regulatory utility will be based on an integration of in silico PBPK/PBBP modeling in biopharmaceutics risk assessment and biowaivers. Dr Anna Nordmark (Medical Products Agency, Sweden) stated that the regulatory agency had observed an increased use of PBPK analysis in applications for marketing authorisation of new chemical entities 8. Today, an EMA concept paper is presenting the principles and the general approach to the use of PBPK analysis at various stages during drug development and regulatory interactions 9. Currently, a CHMP guideline is under development to provide guidance on what should be included in PBPK reports including model qualification of the intended use 10. A guideline is proposed which will provide general advices applicable to the content of PBPK reports and qualification of PBPK software in order to harmonize the quality and the extent of submitted data. In addition, the accuracy of any library files used in the analysis needs to be supported. It is also clear that demands on model evaluation increases with the increased regulatory importance of the analyses. The PBPK drug model should be able to predict existing in vivo data and all the model parameters. Dr Nordmark also emphasized the importance of that consequences of over- and under-predictions of in vivo data are reported, analyzed and discussed as well as that a sensitivity analysis is performed on the model assumptions and uncertain parameters. The current EMA regulatory position is not to accept M&S as the only source of information for granting biowaivers or for approval decisions on food effects or formulation changes; further supporting evidence of the


Page 8 of 27

Page 9 of 27

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


ability of the models to predict these situations including drug product examples are still needed. In this respect, it is encouraged to publish studies linking GI absorption M&S with clinical pharmacokinetic performance. In that way a knowledge base can be set-up that is similar to that already existing for modelling the clinical impact of drug-drug interactions (DDI). According to Dr. Robert Lionberger (FDA, Office of Generic Drugs, US) more than 85% of all patients in the US receive their drug product treatment by the use of generic products (Figure 2). To secure relevant innovation, research and development in the area of product equivalence will be of paramount benefit to a major part of the patients11. Understanding the science of equivalence, or product bridging, is crucial to reduce the risk that a drug product difference may impact therapeutic substitutability 12. Whilst bioequivalence studies to show the equivalence (substitutable) of pharmaceutical products are performed under rigorous controlled clinical testing conditions, it is important to reflect of how a drug product may be used (or mis-used) by the patients themselves. The role of M&S during generic drug evaluation may be to risk assess differences between a test and reference product and evaluate how it may impact therapeutic substitutability. Dr Lionberger emphasized that the context for biopharmaceutics M&S in the generic drug framework is different than from M&S in drug discovery and development, where usually much more knowledge about drug, formulation properties and human PK are often available. Interestingly, research activities in this large area are being conducted by FDA rather than by individual generic pharmaceutical companies. There is a need for best practices to ensure regulatory acceptability of a bottom (in vitro)-up (in vivo) PBPK model for drug products. Identified knowledge gaps that need attention are the impact of GI motility, GI lumen hydrodynamics and gaining more confidence on the colon physiology, particularly important for M&S of oral modified release (MR) formulations. Information on intrasubject variability of GI parameters is of shortage, which means that in higher risk cases FDA may request fully replicate bioequivalence study designs for generic MR drug product during fasted and postprandial conditions. The interactions between formulation properties with GI physiological factors are crucial for the design of in vivo bioequivalence tests and especially for complex formulations such as solid dispersion and nano-



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

based drug delivery systems. In order to reduce these knowledge gaps, FDA and the generic pharmaceutical industry agreed that a small fraction of generic drug user fees are invested in research relevant to generic drug product development and approval process. Research issues related to biopharmaceutics include: GI lumen hydrodynamics, effect of product release rate on first-pass metabolism in different enzymatic phenotypes, and the effect of different manufacturing and critical pharmaceutical process parameters. This provides a benefit to the entire generic industry in the form of better biopharmaceutical tools for drug product development and approval processes. Volker Fischer (ABBeVie) demonstrated an increasing number of applications and higher impact of PBPK M&S as part of the pharmaceutical development and internal decision making process within industry in the area of drug pharmacokinetics, metabolism and DDIs. The progress in this area might provide some guidance in best practice by means of applying PBPK modeling in the biopharmaceutics context. Dr. Shinji Yamashita (Setsunan University, Japan) started his presentation by informing the audience that there is no BCS-based biowaiver regulatory guidance in Japan, but that biowaivers for drug products can be granted based on in vitro drug release tests in multiple dissolution media in the case of minor changes in the formulation and dose strength. The rationale to why the BCS is not used in Japanese bioequivalence guidelines was discussed, and the main reason are being attributed to the fundamental use of API biopharmaceutical properties such as solubility and intestinal permeability, and not fully accounting for the impact of pharmaceutical formulation and manufacturing process on in vivo performance. A concern in Japan is that the in vivo human intestinal permeability values for many APIs are not determined. However, bioequivalence tests for most oral IR products can be demonstrated by the multimedia in vitro dissolution test, and accordingly there is no direct need for BCS in Japan 13. Examples of discriminative in vitro dissolution tests are displayed in Table 1. A pH range is necessary to apply to be able to ensure bioequivalence in normal and achlorhydric subjects, since some oral products demonstrate similar dissolution rate at one pH value, do not always show the same dissolution at another pH. In Japan,


Page 10 of 27

Page 11 of 27

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


more than 60 % of elderly people, defined as >60 years, have an achlorhydric stomach with almost neutral gastric pH. Ongoing pharmaceutical research in Japan focuses on aspects that differ across different global regions, such as the ingested fluid volume with a drug product (250 mL for FDA and 150 mL for PMDA). In a survey (n=512 subjects) of the water volume taken orally together with the drug products, it was shown that the mean ingested volume was about 88 mL (Figure 3). The impact of ingested fluid volume on GI absorption during in vivo bioequivalence studies showed that pitavastatin could dissolve completely in the stomach irrespective of the water intake, while glimepride and lansoprazole could not. Without water intake, un-dissolved drug particles might be retained in the stomach longer and delay the absorption of drug from the small intestine 13. Recently a broad discussion for an ICH initiative on bioequivalence was started. It is an ongoing process to define regional global differences on IVIVC and bioequivalence, which has created an expectation of a future harmonization across regulatory agencies on this important task. Changes in the way in which BCS guidelines are used in Japan are likely to be driven by the increasing approval and clinical use of generic pharmaceutical products, which historically have been at a lower level of clinical use relative to marketed pharmaceutical products in the EU and US. Dr. Lawrence Yu (Office of Pharmaceutical Quality, FDA, US) explained the new organizational structure at FDA and the regulatory path forward on Pharmaceutical Quality that promotes a continuous manufacturing process quality as well as a pharmaceutical quality that are primarily related to the medical need in each patient. FDA’s adapted quality vision may reduce the tremendous number of perfomed human in vivo bioequivalence studies in the recent past 14. The introduction of BCS in the regulatory context is calculated to have saved several hundred million dollars in avoidance of in vivo bioequivalence study costs. In this context, FDA expanded the biowaiver policy in their revised bioequivalence guide line to also include BCS III APIs when appropriate in vitro dissolution testing methods are available. Granting IVIVC-based biowaivers for BCS IV compounds seems difficult so far, but granting a biowaiver for BCS II weak acids seems not impossible in the future 15. Based on earlier experience by regulatory reviewers, it



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

is not common practice that an appropriate IVIVC is developed that would support a scientifically sound dossier for claiming a biowaiver. The lack of a coordinated ICH effort in the area of oral biopharmaceutics would explain the various regulatory guidances issued and in use around the globe. Dr Yu further discussed the agency’s focus on in vitro dissolution methods and their use. Especially, the view and intention to move these in vitro tests of in vivo pharmaceutical product performance from a traditional QC specification pass/fail test to methods that relate directly to clinical performance in the patient. Recently proposed changes to regulatory guidances include a reduction in the ingested fluid volume used to 500 mL to more closely resemble the situation in their clinical use. Also, it was recognized that improve the in vivo relevance further with even smaller fluid volumes will be challenging from a dissolution equipment perspective. However, there is likely to be a continued regulatory emphasis towards more biorelevant in vitro dissolution methodologies in the future. Dr. Gordon Amidon (University of Michigan, Ann Arbor, USA) supported the increased focus on in vitro dissolution and its in vivo predictiviness. He also supported the recent FDA funding initiative to increase the in vivo knowledge on luminal fluid mechanics, motility and transit along the human GI tract. Dr. Amidon also emphasized the potential of in vitro biorelevant dissolution media testing for subclasses of BCS class II and IV APis of acids and bases to further extend regulatory biowaiver guidances in the future 16. It was presented and confirmed from earlier studies that there exists discontinuous distribution of liquid pockets along the human small intestinal 17. The individual distribution pockets in these individual subjects are shown in Figure 4. This is considered to be important for the rates and extents of dissolution, luminal precipitation, and intestinal absorption of oral dosage forms and APIs in vivo in the human GI tract and can be used to develop more physiologically relevant in vitro dissolution methodologies, as well as M&S predictive computational tools. Dr. Amidon demonstrated that it is crucial for any in vitro dissolution method to consider the bicarbonate (HCO3−) secretion by the duodenal epithelial cells and from pancreas into the proximal small intestine. This proximal small intestinal secretion neutralizes the gastric acid emptied into the duodenum and


Page 12 of 27

Page 13 of 27

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


buffers the intestinal fluid to maintain an almost neutral small intestinal luminal pH. Conducting in vitro dissolution experiments in bicarbonate buffer would be more physiologically realistic, but the preparation and use of bicarbonate buffer in vitro methodologies is complex. For instance, CO2 gas must be constantly added to the media to maintain a constant bicarbonate buffer concentration, which is generally a slow process that may affect the hydrodynamics and in vitro dissolution of drug product/particles because of the potential presence of gas bubbles at solid–liquid interfaces. Consequently, applying a buffer solution that produces equivalent buffer effect on in vitro drug dissolution as bicarbonate buffer, would be a better choice. Matching the in vitro dissolution rate of weak acid and weak base drugs in phosphate and bicarbonate buffer systems is possible. However, Dr Amidon emphasized that it is crucial to consider the very complex interplay between buffer pH and pKa(s), drug pKa, solubility, and finally diffusion layer thickness 18, 19. Finally, Dr Amidon ended by concluding that the validation of the various dissolution methods and conditions, sharing information and interaction with the regulatory agencies are considered to be of utmost importance in order to facilitate an appropriate regulatory risk assessment by the regulators and to be a part of the development of future regulatory guidance on biowaivers beyond BCS I and III.

Discussion The progress of the oral biopharmaceutics area to novel regulatory practices and guidelines has been incremental since the introduction of BCS in the late 90’ies 20. There is now an excellent opportunity to take a new step based on in vivo predictive dissolution (IPD) and physiologically based GI absorption in silico tools developed during the last 15-20 years. These tools are now routinely used at different stages of pharmaceutical product development in the industry, but rarely included as a part of the regulatory submissions. The primary objective of the OrBiTo project is to further improve the utility and confidence of such in vivo predictive methods to further replace in vivo experimentation while achieving products



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 14 of 27

with optimal clinical performance in patients (1). This could also include regulatory applications of these predictive tools like biowaivers of formulations bioequivalence studies, food studies or quality standards like dissolution or particle size specifications. The major opportunity is probably in combining in vivo predictive dissolution testing and physiologically based in silico models where the in vitro data provides input to the absorption predictions. The limitation of standalone use of in vitro dissolution data is thatthere is only a limited number of products where there is a simple, direct relation between in vitro dissolution data and in vivo biopharmaceutic variables exists. This is primarily for well-designed extended-release formulations. For immediate release formulations, the absorption is almost always modulated by other factors such as gastric emptying and intestinal permeability and level A IVIVCs for IR tablets are therefore very rare. This problem can be addressed by using physiologically based in silico prediction tools where the different factors influencing absorption are continuously integrated throughout the dynamic GI drug absorption process. There are still several areas for improvement of in vitro dissolution testing by means of generating results relevant for the intraluminal conditions in the GI tract. For example, the OrBiTo project and other current research projects include definitions of test media representative for the more distal parts of the GI tract, models capturing supersaturation and precipitation phenomena, influence of motility waves on shear and other forces of hydrodynamic origin, addressing the interindividual variability in composition and characteristics of GI fluids, food effects, definition of biorelevant buffer systems and intestinal water volumes. In OrBiTo, advances are in most areas built on novel mechanistic in vivo studies in humans studying processes of interest. Furthermore, the OrBiTo projects include extensive validations of different methods versus in vivo data based on a historical database populated by data from the industry partners in OrBiTo. This will be an important aspect in further defining the “space” by means of API and formulation properties where different methods provide reliable in vivo predictions. Thus, the next years will provide a significantly strengthened scientific basis for application of in vitro predictive testing in regulatory context, such as biorelevant diss olution tests.


Page 15 of 27

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


The use of PBPK prediction models has been introduced in a regulatory context primarily in the drugdrug interaction area but also to some extent regarding special patient populations, like paediatrics. A concept paper has also been published by EMA on Qualification and reporting of physiologically-based pharmacokinetic (PBPK) modelling and analyses” and a new guideline is under preparation 9, 10. This facilitates the introduction of these tools also in the regulatory biopharmaceutics area since the tools are already available at the leading regulatory agencies and some initial principles for regulatory usage are emerging. This opportunity will be further boosted by multiple activities in the OrBiTo project which aims to improve GI absorption predictions by physiologically based in silico models including gap analysis/validations versus historical in vivo data, improved GI physiology system understanding, standardisation and increased biorelevance of in vitro based input variables into predictions, meaningful integration of product in vitro dissolution data into models, improved and extended algorithms better describing the GI absorption process, identification of best practice in predictions and training of modelers. For example, as a major task within OrBiTo in this area, a first blinded, fully bottom-up gap analysis with existing predictive in silico absorption tools has been completed including a data set of over 600 human PK study arms performed by scientists from 15 different OrBiTo partner organizations. A renewed analysis will be performed at the end of the project where a multitude of different improvements, based on the increased knowledge generated from OrBiTo tasks, has been implemented. Thus, the combination of the established usage of physiologically based in silico models for oral absorption predictions in the industrial development process, spearheaded of regulatory applications in certain PK areas and developments in progress within the OrBiTo projects will provide an excellent platform for enhancing the use of this type of tool in a regulatory context. Another critical factor to ensure progress in the regulatory context is for the pharmaceutical industry to start include in silico simulations and predictions in the regulatory submissions and also publish examples, which was clearly encouraged by senior regulatory agency representatives at the meeting. This allows for an increased familiarization of such in silico tools and helps evolve standards and practices. Further



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

evolution of best practices including description and validation of model assumptions are critical success factors. Finally, the initial gap analysis in OrBiTo revealed significant variations in predictions between different researcher not only verifying the need in standardization but also the need for enhanced training of the model users. In the first instance, perhaps such data mainly will be of supportive nature in case of biowaivers or defining pharmaceutical quality criteria. A possible immediate application of more pivotal nature is to establish PBPK/PBBP based in vitro/in vivo correlations or relationships that can be used as basis for biowaivers in analogy with classical IVIVC approaches also using the same acceptance criteria (e.g. prediction error

In Vivo Predictive Dissolution (IPD) and Biopharmaceutical Modeling

Recommend Documents