Meeting Report:
AAPS/FDA Workshop on Biopharmaceutics Classification System

Vivian A. Gray

V. A. Gray Consulting, Hockessin, DE

email correspondence to: vagray@rcn.com

The FDA and AAPS sponsored a workshop on September 25-27, 2002 on the implementation, challenges and extension opportunities for the biopharmaceutics classification system (BCS). In August 2000, the FDA published a Guidance to Industry titled; Waiver of In Vivo Bioavailability and Bioequivalence Studies for Immediate-Release Solid Oral Dosage Forms Based on a Biopharmaceutics Classification System. The major aspects of the criteria are solubility, dissolution, and permeability.

The goals of the workshop were to discuss issues and science related to the BCS guidance, review regulatory applications, influence future FDA guidance direction, and debate issues related to the BCS extension.

Gordon Amidon of the University of Michigan and Robert Lipper of Bristol Myers Squibb began the conference by defining the challenges ahead. Amidon reviewed the present guidance and outlined the possible extensions that would be considered during the workshop, especially during the breakout sessions. Lipper reiterated the importance of the BCS as a useful tool in many areas, but stressed the need to achieve international harmonization and uniformity of the BCS application. He also repeated a challenge from industry to clearly define where the extensions are headed and when they will be accomplished.

Ajaz Hussain, Deputy Director of the Office of Pharmaceutical Science in the Centerfor Drug Evaluation and Research at the FDA began his message with the statement that the agency realized that there were many biostudies that were not adding value to the assessment of bioequivalence. This led to the development of the biowaivers in the present guidance. He mentioned that a more recent draft guidance justifies biowaivers of certain food-effect studies. He focused on the need to understand the mechanisms of drug-excipient interaction before the extensions can safely be extended for Class III and possibly Class II drugs. Ajaz talked about how the European and US regulatory biowaiver approaches appear to be moving closer together. A discussion of the Process Analytical Technology (PAT) initiative was presented.

Mehul Mehta of the FDA discussed the preliminary results of a retrospective (spanning the years 1995 to 2001) FDA evaluation of in-house NDA data from reviews and original submissions. The objective was to learn about the appropriateness of the methods and criteria recommended in the BCS guidance and to gather data to evaluate whether the BCS guidance would need to be revised in the future. The methodology was to identify important biopharmaceutical variables such as the chemical properties of pKa, log P, acid or base, the pH solubility profiles, permeability assessment, formulation composition and dissolution profiles. The pharmacokinetic data such as absolute bioavailability, bioavailability relative to a solution, linearity, bioequivalence studies, metabolism, food effects and urinary recovery were studied. The survey found that about 20% of the compounds did not have solubility profiles and most of those that did were performed at 25°C.

The permeability data consisted of mainly mass balance. Absolute bioavailability data was available for 65 drugs. Intestinal permeability data was rare. Dissolution profiles for the three media were available for about 50% of the NDAs. There was not much information on GI track stability. The survey found that 46 out of 100 compounds were BCS Class 4, of these 15 showed BE failures. The definition of a BE failure was that either AUC or Cmax failed the BE criteria of 90% CI. He went on to say that the preliminary observations noted that no BE failures were observed for Class 1 drugs, 6 out of 12 BE failuresfor Class 2 drugs, and 3 out of 28 BE failures for Class 3 drugs. Some conclusions, although he cautioned that the database has not been audited for 100% accuracy, from the survey are that the BCS Class 1 criteria are stringent but can be met, there are no BE problems with Class 1 drugs, and a significant food effect seems to be associated with only the extensively metabolized drugs. He went on to add that the database has valuable information that can be used to evaluate possible extensions of waivers. The extensions being considered are: biowaivers for Class 3 drugs with rapidly dissolving formulations, reduction of the high permeability requirement from 90% to 85%, and defining a new intermediate permeability class. He suggested that perhaps a new intermediate solubility class could be defined and posed the question of biowaivers for food effect studies on Class 1 drugs.

The next FDA speaker was Barbara Davit, who gave some of the experiences of FDA with BCS based biowaiver applications. She covered three applications in detail and reported on their status. One compound met the criteria for BCS Class 1 and the applicant has been asked to conduct a fed BE study. The second applicant was asked to further evaluate dissolution using the SGF and SIF without enzymes, to verify that drug loss in the GI tract was from intestinal permeation rather that degradation, and verify that the drug substance binds to surfaces of plates used in the permeability studies.
The third application is still under review in the area of permeability, especially the role of presystemic metabolism. Her analysis was that the permeability aspects of the biowaiver are usually the most problematic for the applicants.

Donna Volpe of the FDA further discussed the permeability aspects. She reviewed the permeability classification in the BCS guidance, the concept of method suitability, and FDA's evaluation of a traditional 21-day Caco-2-cell permeability assay for 20 model drugs. She described in detail the method suitability criteria for the cell culture models and presented the rank order correlation between in vitro Caco-2 permeability and in vivo human extent of absorption of the model drugs. There was a clear segregation between high and low permeability drug substances. The study confirmed that the concept of method suitability outlined in the BCS guidance is a practical and feasible regulatory research tool for the evaluation of drug substance permeability for biowaiver applications. She went on to add that each laboratory must use method suitability to demonstrate that its own cell culture assay is capable of classifying drug substances as high or low permeability.

Vinod Shah of the FDA discussed the utility of BCS in establishing dissolution specifications. He pointed out that the BCS guidance is only for immediate release, non-narrow therapeutic index drugs. The dissolution media and conditions were reviewed and the rapidly dissolving criteria of 85% in less than or equal to 30 minutes in each media (using one apparatus and speed) discussed. Vinod examined the difference in setting specifications for new drugs versus generic drugs, especially compendial and non-compendial drugs. The drug product should comply with the specification throughout its shelf life, so the final specification should be based on experience gained during the development process and include stability data. He emphasized that profile data is needed for tests that have longer than 15 minute specifications (80% Q). The BE dissolution test will include the QC test, which may be a single point specification, and additional dissolution tests at the three pH's and profile points. Based on BCS, dissolution specifications for Class 1 drugs can be set at 80% Q in 30 minutes.

James Cook of Pfizer presented the industry experience with the BCS guidance. He discussed its value and limitations. The obvious benefits are the decreased expense of biostudies but he showed that only 24% of drugs under the current requirements would be candidates. He advised that there needed to be an effective strategy to identify waiver candidates to maximize the savings. There are some costs however in performing the initial solubility and permeability studies. Also convincing the responsible parties to apply for the biowaivers and deciding who will lead the initiative is troublesome. He closed on a positive note that the implementation has saved Pfizer one million dollars so far.

Charles DiLiberti of Barr Laboratories gave the generic industry perspective on the BCS guidance applied to ANDAs. His assessment of BCS biowaivers was that the current rules severely limit the savings in cost and subject numbers. He went on to explain that because the drugs qualifying forBCS waivers have inherently low intrasubject variability, it is unlikely that, even with proposed extensions, BCS waivers will ever save a large fraction of R &D dollars or reduce subject numbers. Major concerns are the uncertainties in acceptability of data and the potential use of data by competitors. If the FDA permits referencing data from another firm's permeability study, this may severely limit business opportunities for the few firms that conduct these studies. DiLiberti did add that the industry would gladly welcome any BCS extensions and stated that the greatest potential benefit of BCS waivers are for the following: drugs where human exposure is highly undesirable, to avoid the need for bio-INDs, long half-life drugs with high inter-subject variability, difficult to recruit subject populations, difficult bioanalytical analysis, and endogenous compounds.

Robert Lipper from Bristol Myers Squibb reported on the application of BCS in candidate selection and in guiding early clinical development strategy. He examined BCS as an aspect of determining "developability" of a compound, considering permeability, solubility and even dose. Once a lead compound is identified then information relevant to probable BCS classification is critical to selection of Phase 1 dosing strategy and can significantly influence the preclinical development timeline. He also asserted that feedback from early clinical BA/PK evaluation of lead compounds can assist discovery in improving biopharmaceutical characteristics of subsequent candidates. Lipper stated there are caveats to BCS, such as, first pass metabolism and dealing with transporters and efflux.

William Charman of Monash University discussed the practical approaches to delivery and assessment of BCS Class 2 drugs. With poorly soluble drugs one must consider the formulation type and pathway to the in vivo solubility. With conventional immediate release formulations the in vivo dissolution can be rate limiting to absorption therefore there is a direct effect of the formulation and in vivo conditions on absorption. The expectation of an in vivo/in vitro correlation with Class 2 is an area of active research. In formulation design Charman listed the areas of improved technology that enhance solubility: lipid-based systems, solid dispersions, particle size manipulation, crystal engineering, and complexation approaches. The biopharmaceutical factors of greatest concern for Class 2 drugs are the absorption pathway, excipient effects on GI function, and food effects. His focus shifted to the enhanced absorption of poorly soluble drugs in the post-prandial state. The post-prandial state, midway between fed and fasted, has yet to be addressed. He elaborated further in describing how small amounts of lipid can alter the luminal environment and ultimately effect the absorption of the Class 2 drug.

Joseph Polli of GlaxoSmithKline, Inc. spoke on the usefulness of in vitro P-glycoprotein (Pgp) assays in the context of BCS. Pgp effects the absorption, distribution and clearance of compounds. He emphasized that it is important in drug development to identify the key drug transporters and their genetic variants and elucidate their role in the disposition of a compound. This establishes the necessity for robust transporter assays. He also highlighted the influence of other membrane transporters. He stated that monolayer efflux assays were his method of choice to identify Pgp substrates.

Future directions of BCS in drug discovery were examined by Vincent Lee of the University of California. His main focus was that a comprehensive computational framework might be developed that incorporates the three drug properties, solubility, dissolution rate, and intestinal permeability. Drug affinity for its receptor would also be a component. BCS will possibly accelerate the pace of a movement towards a cellular and subcellular focus as the basis of optimizing drug bioavailability.

Ameeta Parekh of the FDA reviewed the evolution of the BA/BE regulatory standards from the 1977 BA/BE regulations (Bio Regs) to today's BCS biowaiver guidance. Some of the key questions over the years have been- can in vitro dissolution specifications be physiologically meaningful for immediate release products, what types of manufacturing/formulation variables are minor, how can in vitro testing be used to approve major formulation changes, what mechanistic evidence could be used in lieu of in vivo BE establishment, and can such evidence be extrapolated to the fed condition? She reiterated the extramural regulatory research on the effects of manufacturing variables, permeability studies, and variables influencing in vivo BE studies. The ultimate product of all this work has been confidence that the BCS is based on sound scientific rationale. She suggests that it now seems reasonable to extend the BCS to the fed state.

Gordon Amidon of the University of Michigan described in detail the basis for extending BE dissolution standards. He stated that there could be a new paradigm of reducing unnecessary in vivo studies based on scientific principles, suggesting that there could be biowaivers for over 50% of drug products. The extensions were numerous and the topic of the breakout sessions. He reasoned that Class 3 could have a dissolution rate specification if the excipients would not interfere with absorption. Other extensions were to change the solubility range from pH 1 to 7.4 to pH 1.0 to 6.8 and Class 2 drugs that were highly soluble in pH 6.8 should be considered for biowaivers. He also commented on applicability of BCS to the PAT initiative.

Brian Houston of the University of Manchester presented on the role of metabolism in predicting BA/BE in the context of BCS. His focus was that first pass metabolism is important in dictating the systemic concentrations of a drug following oral administration. The metabolic clearance is presently considered an important additional drug property in BCS. The impact of intestinal metabolism causing a marked change in therapeutic outcome is well known. These issues are not addressed by the in vitro test but this is the subject of recent research. Ronald Borchardt of the University of Kansas continued on the subject of transporters, discussing whether the cell culture models could be predictive of transporter mediated drug permeation in the context of BCS.

The excipient effect on permeability in vivo was examined by Jim Polli at the University of Maryland. He discussed the methodology for permeability classification, including considerations for excipient effects. He higlighted the excipients sodium lauryl sulfate, Tween 80, HPMC, Vitamin E TPOS, and cremophor.

Lawrence Yu of the FDA spoke on the challenges and opportunities for BCS-based biowaiver extensions. He observed that using literature data was not found to provide adequate support for regulatory requirements. Yu stated that there were many possible ways to extend the BCS biowaiver. The first was to change the current solubility range from pH 1.2-7.4 to pH 1.2 ­ 6.8. The second was to look at 500 mL as the average volume for the small intestine; therefore the 250 mL may be replaced by 500mL for solubility determinations. The third extension may be to drop the permeability level from 90% absorbed to 85%. The fourth possibility was to use the acceptable Class 1 dissolution rates for Class 2 with the use of surfactants or 6.8 pH buffer. The fifth extension was to look at Class 3 and apply the dissolution rate of 85% in 15 minutes for all three media. With Class 3 the issue of transporters was discussed along with excipient effects.

Hans Junginger of Leiden University spoke about the pilot project of the FIP BCS Special Interest Group Database called the International Biopharmaceutical Classification Data Base (IBCDAB). This project was undertaken because there is a need for BCS data of active substances and their finished products to be available in the public domain.
Arthur Straughn of the University of Tennessee presented the results of a FDA study performed to evaluate the potential for bioinequivalence between AB rated BCS Class 1 generic formulations under fed conditions. The conclusions were that food did not effect BE of the formulations tested, that a biowaiver would seem rationale for Class 1 drugs, and there was preliminary evidence for Class 3 biowaivers.

Dale Conner of the FDA reported that the draft guidance for food effects was available for comment. The guidance allows for BCS biowaivers for fed studies. The guidance will be official in the near future.

The breakout sessions with the summaries can be found on the AAPS website (www. aaps.org) under meetings ­ past meetings and summaries. The key break out session on the subject of dissolution was the session entitled Solubility Classification and Dissolution Classification. One major issue raised in this session was that applying the F2 calculation was difficult. The suggestions were to raise or eliminate the CV requirement, put in a two point spec-15 and 30 minutes, or that F2 only applies if the time goes out to 45 or 60 minutes. The extension of the time for the Class 1 requirement to 45 or 60 minutes for the pH 6.8 buffer was suggested. Another suggestion was to raise the volume in the solubility studies from 250 mL to 500 mL for pH 4.5 and pH 6.8. An issue with much support was to allow the paddle speed to be increased to 75 rpm when coning was present.

Lawrence Yu ended the meeting summarizing the proceedings and the next steps. There were extensions to be considered for Class 3 and Class 2, the 90 % absorption level may be lowered, biowaivers may be in place for fed studies for Class 1, concerns on the method suitability of the permeability classification will be addressed, mechanistic causes for a BE failures in Class 3 will be studied, and changes to the solubility range and volumes will be considered.

Some of the presentations of this workshop can also be found at the AAPS website. This information will be available for only three months. There is also another way to view the proceedings of the workshop at the AAPS website. Go to Education, then Distance learning and under this workshop title the presentations with overlay of the speakers voices are available for all the talks of the workshop. This is an excellent way to get the full information of the meeting. You must be an AAPS member to get to this information.