Bio-relevant Dissolution and Drug

Absorption Using Predictive Modeling

Tools: Regulatory Perspective

and Experience

November 14, 2016

Kimberly Raines, Ph.D.

Division of Biopharmaceutics

FDA/CDER/OPQ/Office of New Drug Products

2

Disclaimer

The views expressed in this

presentation are those of the speaker

and not necessarily those of the Food

and Drug Administration (FDA).

3

Introduction

• Emerging use of bio-relevant dissolution andPBPK modeling as a biopharmaceutics toolstems from the ability to predict the impact ofthe many factors, including properties of theAPI, that can influence oral drug absorption.

• Biopharmaceutics tools, such as dissolutiontesting using physiologically-relevant in vitrosystems can thus enable a more efficient drugproduct development process by helping torelate drug product quality to the patient.

4

Bio-relevant Dissolution

Bio-relevant in vitro dissolution is widely usedwithin the pharmaceutical industry to address anumber of biopharmaceutics factors:• Dissolution method development

• Discriminate formulation factors (e.g. polymers, particlesurface area, or physical and chemical characteristics of thedrug)

• Reflect significant differences in bioavailability

• Input into PBPK-based modelling and simulation

5

Application of PBPK modeling

• Study effects of particle size, shape and particle density on dissolution and precipitation

• Examine effects of solubility on dissolution and absorption

• Deconvolute in vivo dissolution for mechanistic In Vitro – In Vivo Correlation (IVIVC)

• Development of “clinically relevant specifications”

Formulated Drug

Solubilized Drug

Absorbed Drugkd kp

6

Dissolution Model Links to

Mechanistic Absorption Model

• An extension of dissolution model for in vivo performance prediction

• Ensured product quality with known efficacy and safety

• Clinically relevant design space and potential CMC flexibility

Drug substance/formulation

properties(e.g., solubility, pKa, LogP,

particle size, permeability, etc.)

In vitro dissolution profile

Mechanistic absorption

and PK model

In vivo performance

Absorption, bioavailability, blood

concentration profiles

D A PV SC

PermeationCarrier-mediated transport

Dissolution model

Fa FDp F

Metabolism Metabolism

7

Challenges and Opportunities

CHALLENGES• In vitro in vivo correlations (IVIVC) success rate is low• In vivo link to quality is challenging (especially for poorly soluble

drugs (BCS class II and IV drug products))

OPPORTUNITIES• Leverage the use of bio-relevant media which closely mimic the

fluids of the human stomach and intestine allowing for bettersimulated conditions of the gastrointestinal tract.

• Use of in silico absorption modeling to assess the impact of invitro dissolution on in vivo performance

8

Regulatory Examples

CASE STUDY #1:PBPK modeling utilized as evidence to supportspecifications for dissolution and particle size wouldensure suitable clinical performance of batches

CASE STUDY #2:Application of PBPK modeling in setting clinically relevantdrug product in vitro dissolution acceptance criteria.

9

Case Study 1

• Lesinuard (ZURAMPIC)

• FDA Approved 2015

• URAT1 inhibitor indicated in combination with a xanthine oxidase inhibitor forthe treatment of hyperuricemia associated with gout in patients who have notachieved target serum uric acid levels with a xanthine oxidase inhibitor alone.

• pH dependent solubility

• In-vitro permeability of lesinurad was evaluated using Caco-2 monolayers and[14C]-lesinurad. In these studies, indicated that lesinurad was activelytransported across Caco-2 monolayers.

• Development of an acceptable QC dissolution testing method able todiscriminate between batches of acceptable bioavailability and batches with aslower rate and extent of absorption, Cmax and AUC.

http://www.accessdata.fda.gov/drugsatfda_docs/label/2015/207988lbl.pdf

10

PBPK Modeling Strategy

Pepin, X. J. H. et al. Mol. Pharmaceutics 2016, 13, 3256-3269

11

Parameter sensitivity analysis (PSA) for failed BE batch

Pepin, X. J. H. et al. Mol. Pharmaceutics 2016, 13, 3256-3269

12

Dissolution space for predicted BE to lesinurad tablets

Pepin, X. J. H. et al. Mol. Pharmaceutics 2016, 13, 3256-3269

13

Case Example 1 - Summary

• In silico absorption modeling has been performed, to assess theimpact of in vitro dissolution on in vivo performance for ZURAMPIC(lesinurad) tablets.

• Dissolution profiles of lesinurad tablets generated using the QCmethod were used as an input to estimate in vivo dissolution in thevarious parts of the GI tract and predict human exposure.

• The model accounts for differences of dosage form transit,dissolution, local pH in the GI tract, and fluid volumes. Thepredictability of the model was demonstrated by confirming theCmax observed in a clinical trial.

• The model indicated that drug product batches that pass theproposed dissolution acceptance criteria of Q = 80% in 30 min areanticipated to be bioequivalent to the clinical reference batch.

• To explore the dissolution space, simulations were performed toindicate that such a batch will also be bioequivalent to standardclinical batches despite having a dissolution profile, which would failthe proposed dissolution acceptance criteria of Q = 80% in 30 min.available for dissolution.

Pepin, X. J. H. et al. Mol. Pharmaceutics 2016, 13, 3256-3269

14

Mechanistic Modeling and Simulation for Drug Product B

• Immediate release tablet

• BCS class 4 drug substance

– (low solubility and low permeability)

• Three Strengths

– A (lower), B (middle), and C (higher)

– A and B studied in phase 3 clinical trials

– A pivotal BE study comparing

• Higher strength C (C1 and C2) to middle strength B

• Dissolution method: Atypical behavior and under-discriminating

– For changes in drug substance and product attributes

– Did not reject batch that was not BE to the phase 3 clinical batch

• Development of a new dissolution method

Case Study 2

15

Setting of Clinically RelevantDissolution Acceptance Criteria

• Dissolution acceptance criteria based on new dissolution method:

– Commercial and registration stability batches

– BE and Non-BE batches

• Proposed dissolution acceptance criteria

– Q = 75% at 30 min

Ο Higher strength C2 (BE Batch)

Δ Higher strength C1 (Non-BE Batch)

Other Profiles (Commercial batches of higher strength C2)

Q = 80% at 30 min?

Mechanistic absorption and PK modeling and simulation

Q = 75% at 30 min clinically relevant?

16

Mechanistic Modeling and Simulation Strategy for Drug Product B

MODEL BUILDINGo I.V. PK Data

o Oral PK Data of Strength B (CTF)

o Physicochemical properties, absorption parameters, and PK

parameters

o Dissolution Profiles (Discriminating dissolution method)

MODEL VALIDATIONo Internal Validation (Strength B)

o External Validation (Strength C – C1 and C2)

o Repeated Virtual BE Trials (C1 is not BE to B; C2 is BE to B)

o Accurate predictions for all of the above

MODEL APPLICATIONo Support bio-predictive nature of the new in vitro dissolution

method

o Set dissolution acceptance criteria

Strength B

Use of Virtual Dissolution Profiles

17

Dissolution Acceptance Criteriafor Higher Strength (C)

Bioequivalent

Q = 75% at 30 min Clinically relevant for higher

strength

Q = 75% at 30 min

Clinically relevant for other strengths (A and B) ?

Virtual ProfileQ = 75% at 30 min

18

Dissolution Acceptance Criteria for Lower (A) and Middle (B) Strengths

Lower (A) and Middle (B) StrengthsQ = 75% at 30 min

Not clinically relevant

Lower (A) and Middle (B) StrengthsQ = 80% at 30 min Clinically relevant(data not shown)

19

Case Example 2 - Summary

• Availability of in vivo and in vitro data enabled physiologicallybased (mechanistic) absorption and pharmacokinetic modelbuilding and validation

• Successful application of validated mechanistic model– Supported bio-predictive nature of the developed dissolution method

– Establishing clinically relevant dissolution acceptance criteria

• Application of biopharmaceutics principles and integration ofin silico tool, and in vitro and in vivo data for establishingclinically relevant dissolution acceptance criteria

20

Demonstration of Clinically Relevant Dissolution in a Regulatory Submission

• In vivo data (e.g., BA/BE, PK) to demonstrate that batcheswith dissolution profiles in line with the proposeddissolution acceptance criteria(on) have a similarsystemic exposure compared to that of the pivotal PKand clinical batches.

• In silico PBPK (physiologically-based pharmacokinetics)modeling and simulation predictions on the impact ofthe quality of a batch with an in vitro dissolution profilerepresenting the extreme of the proposed dissolutionacceptance criteria(on) on the systemic exposure of thedrug product.

21

Considerations for PBPK model submission supporting a Clinically Relevant Dissolution

i. Modeling summary report, including model development, modelverification/modification, and model application.

ii. Detailed information on the inputs used in the construction andvalidation of the model(s). When the parameters are optimized, thedata source selection, the estimation method, the justification forthe optimization algorithm, and the assumption used should beprovided.

iii. Name and version of the software, and (for custom modelingsoftware) schematics of model structure and differential equations.

iv. The methodological approach to model verification, modelverification results, and sensitivity analyses to evaluate therobustness of the model.

v. The results of using the verified model to address the studyquestion(s) should be presented.

22

Future Direction

• Introduction of regulatory framework on theuse of bio-relevant dissolution methodologiesand PBPK modeling in regulatory submissionsspecific for biopharmaceutics

• Encourage the establishment of a correlationbetween critical process parameters andclinically relevant release and stability in vitrodissolution acceptance criteria

23

Acknowledgments

• Division of Biopharmaceutics

– Dr. Paul Seo

– Dr. John Duan

– Dr. Sandra Suarez-Sharp

– Dr. Poonam Delvadia

– Dr. Min Li

– Dr. Ho-Pi Lin

– Dr. Meng Wang

– Dr. Fang Wu