Beruflich Dokumente
Kultur Dokumente
AAPS workshop on BE, BCS and Beyond, May 2007, co-sponsored with FDA
Bertil S. Abrahamsson Ph.D. Paul A. Dickinson Ph.D.
bertil.abrahamsson@astrazeneca.com paul.dickinson@astrazeneca.com
Outline:
Introduction to Quality by Design (QbD)
A driver to go beyond BE and BCS
Review the current accepted practices for demonstrating clinical quality Focus on approaches for BCS2 compounds
Case study FDA pilot programme Establishment of the Clinical Boundary of the Design Space
Evaluation of clinical impact of key product and process variables Potential Regulatory Flexibility and Continuous Improvement Generic overview
Aim:
Propose to demonstrate how BCS and dissolution testing can be applied to QbD to assure clinical quality
Acknowledgments
PAR&D Paul Stott Sheena Behn Andy Townsend Ryan Gibb Andy Sheridan John Smart Chris Potter
Clin Pharm Peter McCormack Parvis Ghahramani Tracey Hammett Reg CMC Linda Billett Bob Timko Carol Stinson
AstraZeneca have Adopted a Science and Risk Based Approach to Clinical Quality
ICH Q8
Design Space
Multi-dimensional space that encompasses combinations of product design, manufacturing process design, critical manufacturing process parameters and component attributes that provide assurance of suitable product quality and performance
ICH Q9
4
What are the expectations for demonstrating clinical quality (safety and efficacy) today?
High Biowaiver:
Permeability Low High
Solubility
Bioequivalence Study.
Dissolution for some SUPAC changes
1 2 3 4
6
Bioequivalence Study.
Dissolution for some SUPAC (smaller) changes
No No No
>300mg
100
10
BCS Class II
1
0 1 2 3 4 5 6 7 8 9
pH
Applying a scientific and risk based approach an High Level Quality Risk Assessment (QRA) was performed predicated on prior knowledge from other projects and specific knowledge about the drug substance. This identified two high level risks for the drug substance 1. Impact of product / process variables on in vivo performance (BCS Class II) 2. Compression properties leading to poor physical quality
Wet granulated IR tablet
10
11
Tablet variants, encompassing the extremes of these variables, were manufactured with the intention of producing tablets with retarded dissolution rates.
12
Process
The four tablet variants were: Standard Clinical Manufacture; Large Particle Size Variant; Process Variant (Over granulation); Formulation Variant (less disintegrant, more binder) Tested in vitro to develop dissolution method capable of identifying changes in product quality due to changes in these most relevant process parameters and quality attributes Tested in vivo evaluate impact of these variables on in vivo performance and if appropriate develop a mathematical relationship (IVIVC)
14
% Dissolution
Variant A (Standard tablet) Variant B (Larger particle size) Variant C (Process variant) Variant D (Formulation variant)
12 Time (minutes)
15
Geometric Mean and Individual Plasma Concentrations from Tablet Variants (n = 10 for Variants A and B, n = 11 for Variant C and n = 9 for Variant D) versus Oral Solution (n = 15) after dosing to healthy volunteers
No difference in PK performance (Cmax and AUC) after dosing the tablet variants to volunteers 16 i.e. not appropriate to try to develop an IVIVC
Study Conclusions The in vivo evaluation demonstrated that changes in the most relevant Process Parameters and Quality Attributes did not affect PK in volunteers
i.e. will have no impact on safety and efficacy in patients
Cannot develop an IVIVC because the PK profiles overlap and there are no differences in vivo However, the impact of these variables on dissolution when tested by the most appropriate method, could be detected
i.e. over discriminatory method
As the variants dosed encompassed three different mechanisms to alter drug release from the tablet the overly discriminatory dissolution test is an appropriate surrogate to assess clinical quality of all outputs from further processing studies
17
Proposed Design Space for Clinical Quality If product has a dissolution profile faster than that of slowest profile dosed to volunteers (Variant D) then clinical efficacy and safety will be comparable to clinical trials material (Clinical Quality Boundary)
18
Future changes such as site, scale, equipment, method of manufacture can be qualified using this dissolution method and limit Question posed to the Agency at recent face to face meeting: If we present the clinical boundary of Design Space in this way and include the supporting data in the NDA, does this meet the FDA expectation of Drug Product design space?
19
Developing Process Understanding and the Impact on Clinical Performance - How this might work in practice.
10
% Dissolved
% Dissol ved
P/4156/7A P/4156/7B P/4156/8A P/4156/8B P/4156/9A P/4156/9B P/4156/10A P/4156/10B P/4156/11A P/4156/11B Variant C Variant D Time (minutes)
P/4156/1A P/4156/1B P/4156/2A P/4156/2B P/4156/3A P/4156/3B P/4156/4A P/4156/5A P/4156/5B P/4156/6A P/4156/6B Variant C Variant D Time (minutes)
% Dissolved
P/4156/12A P/4156/12B P/4156/13A P/4156/13B P/4156/14A P/4156/14B P/4156/15A P/4156/15B P/4156/16A P/4156/16B Variant C Variant D Time (minutes)
We have demonstrated that Product manufactured at the extremes of the manufacturing parameters studied exhibits a dissolution profile significantly faster than variant D Therefore, all variants within the manufacturing parameters studied will be bioequivalent
21
22
11
Generic Summary/Overview
3.
4.
5.
12
Understanding clinical importance of relevant manufacturing variables role of dissolution and BCS
High
Dissolution accepted
Solubility
Low
1 2 3 4
25
Bioequivalence Study Or Follow principles of BCS2 or BCS3 if can demonstrate that compound behaves more like BCS2 or BCS3 in vivo
Why dissolution testing acceptable surrogate for clinical for class III drugs in QbD
Modelling indicates that class III drug products with rapid dissolution is insensitive to changes in dissolution, eg
Difference Cmax <10% for solid with complete dissolution after 1 hour vs solution for wide range of elimination rates and permeabilities, Kortejrvi, Eur J Pharm Sci 2007, 30, 155 Metformin, Cheng Eur J Pharm Sci 2004, 22, 297 Ranitidine, Polli Adv Expr Med Biol 1997, 423, 191-198
Published biowaiver monographs for class III drugs based on literature data, show no BA differences for different IR formulations eg
Atenolol Ranitidine
QbD implies higher level of understanding of manufacturing process and product compared to historical position
26
13
High
Solubility
Low
Limit set based on clinical Complete dissolution within bioavailability data 30 minutes in most discriminating simple media (physiological pH range). If slower: bioavailability data or additional mechanistic information Complete dissolution within Limit set on case by case 15 minutes in most basis discriminating simple media (physiological pH range). If slower: bioavailability data or additional mechanistic information
1 2 3 4
27
Setting a BCS 2 dissolution limit that assures clinical quality based on BA data two acceptable approaches
Classical IVIVC
Cmax or AUC
Safe space*
Cmax or AUC
*A
10% *B limit *C
*A
*B
*C
limit
in vitro dissolution
in vitro dissolution
A = biobatch, B and C = side batches based on most relevant manufacturing variables *For this type of approach to be acceptable the most relevant risks to clinical quality need to have been assessed (i.e. in a QbD setting)
28
14
Conclusions
Dissolution testing should be applicable to assure desired clinical performance for wide range of drugs in QbD based on BCS considerations and specific product knowledge Failure to establish classical IVIVC, could be succesful outcome of in vitro/in vivo study in context of QbD, if all variants produce the same exposure
29
15