1. Objective risk assessment via asking the right question at the right time: Knowledge based decisions for dosage form design/development/control Ken Morris, Ph.D. University Professor and Director Lachman Institute for Pharmaceutical Analysis Long Island University – Brooklyn Campus Arnold & Marie Schwartz College of Pharmacy and Health Sciences 1

2. Patient related failure modes as the basis for development, risk assessment and equivalence Examples Patient Failure mode Efficacy Toxicity Side effects Compliance Dosing regime Route of administration Examples of Pharmaceutical Source(s) of variability Stability (phys-chem) C p vs. time t ½ Dissolution Processing changes Company changes From where does the knowledge come? 2

3. Implementation of the Generic Drug User Fee Amendments of 2012 (GDUFA) Testimony of Janet Woodcock, M.D. Before the Committee on Oversight and Government Reform United States House of Representatives, February 4, 2016 Ongoing Challenges …”Second, there is a need for more research in the generics space. Some drugs lack generic competition because there is no convincing bioequivalence test method available. … Similarly, methods for showing chemical sameness for certain complex drugs are not available…” 3

4. Pharmaceutical Science for Generic Drugs: The Science of Equivalence Robert Lionberger, Ph.D., 4/30/15 (slide 26) Three for the Future of NIPTE Ensure modern manufacturing for the 5 out of 6 products that are generics Advance the science of equivalence: – Prioritize tools and science that make equivalence evaluation clear and development easier – Product design for human use Engage with GDUFA regulatory science 4

5. The Science of Equivalence: Therapeutic and Bio-Equivalence are only as good as the Pharmaceutical Equivalence BE is limited in scope (low variability normal healthy subjects) and relies on a clear definition and understanding of the PE to be of use Relying on BE in the absence of PE is a no-starter, TE assumes PE/BE/GMP Hence the need to set meaningful specifications for a product before it goes to the clinic (incl. est. shelf life) PE must be strong in prod/process understanding and inform SOP design Pharmaceutical Equivalence needs to move from the 19 th - century classification (tablet, elixir, etc.) to 21 st Century Classification – New drugs get the right CQA’s – considering patient needs and related failure modes. Same CQA’s and lot-lot variability for Generics. 5

6. RQRT #1: Levothyroxine Sodium Pentahydrate 6

7. Symptoms vs. problems Adherence vs. compliance – How do I know what I know? Critique the methodology 7 Why ?

8. Eric Duffy – CDER-FDA 8

9. Impact of the Solid State on Chemical Stability: Dehydration (Solid State Chemistry of Drugs, Byrn 1999) 9 Impact of the Solid State: Crystalline Hydrate Classification System ( Morris and Rodriguez 1993, Morris and Brittain1998) Dehydration can lead to amorphous material, new crystal structures, dehydrated hydrates which maintain the packing motif of the original structure, or mixed structures

10. 10 Dehydration of l-TH can produce a dehydrated hydrate which maintain the packing motif of the original structure as shown Katrusiak and Katrusiak, 93(12), JPS 2004

11. Does Dehydration Precede Degradation? 11 Pharm Dev and Tech Vol. 20(3), pgs.314-319 May 2015 Does O 2 Cause Degradation?

12. So a lot was known or discoverable about levo’s properties and issues – NTI (narrow therapeutic index drug) – Very low dose, 25-300ug – Chemically labile Stable if hydrated but oxidized if dehydrated/disordered Processing effected stability (ala structure) – Cincinati thesis Excipient interaction (often via pH)-thesis, Mansoor’s papers – Long PK t ½ of approximately 7 days Yet products were developed and approved apparently oblivious of the prior knowledge and/or logical concerns This illustrates the point that for generic products, the dosage form quality specs need to be developed at the preclinical or pre-bio- study stage (QTPP – Q8 -ala Q6A) Otherwise, you can’t know what variability is due to the patient variability or PD issues instead of the variability in the dosage form? 12 RECALL - this pre- dates the “Q’s” QbD, QbR, Q8 etc…so no Dev. Report in US

13. What are some of the right question(s) at the right time: during development and ANDA review for levothyroxine sodium pentahydrate? Development What are the patient failure modes used in product design What does the literature say about the science and API What are the CQAs Is the API chemically stable What is the solid form of the API and is it physically stable over the shelf life Does the process compromise API stability Specs and Risk Assessment Review Is the DH complete (including problems observed) What is the API characterization process Are the analytical methods really targeting CQAs Are the specifications relevant Are the process controls sufficient to insure integrity of the API Is the dosage form PE to the RLD 13

14. NTI quality classification: knowledge base development (generating “new” prior knowledge) Drug Oral Dose (mg) Stability C p t 1/2 (hr) Solubility (mg/mL) Product Recall Basis (2012 - 2016) Mol. Wt (g/mol) Prazosin1Light Sensitive2.50.5 (HCl salt)-383.4 Warfarin1 Temperature sensitive 400.017 Super potetnt, CU, Stability 308.33 Clonidine0.1+1450-230.09 Valproic Acid125 Light and Temperature Sensitive 12.550 Failed Dissolution 144.21 Digoxin0.05Light Sensitive420.0648-580.94 Levothyroxine0.025 Light Sensitive, Oxidation 168 0.000105 0.15 (salt) Subpotent, stability, (45) 776.87 Phenytoin30 Temperature sensitive 170.032 Failed dissolution 252.27 Isoetharine Mesylate 0.35 Temperature sensitive 3.18-335.42 Disopyramide100+6.70.0449Failed dissolution339.47 14 Many more descriptors are relatively straight forward to obtain or generate

15. RQRT #2: Methylphenidate 15 [11-13-2014] FDA concerns about therapeutic equivalence with two generic versions of Concerta tablets (methylphenidate hydrochloride extended- release) Based on an analysis of data, FDA has concerns about whether or not two approved generic versions of Concerta tablets (methylphenidate hydrochloride extended-release tablets), used to treat attention-deficit hyperactivity disorder in adults and children, are therapeutically equivalent tothe brand-name drug. In some individuals, the Mallinckrodt and Kudco products may deliver drug in the body at a slower rate during the 7- to 12-hour range. The diminished release rate may result in patients not having the desired effect. As a result, the FDA has changed the therapeutic equivalence (TE) rating for the Mallinckrodt and Kudco products from AB to BX. This means the Mallinckrodt and Kudco products are still approved and can be prescribed, but are no longer recommended as automatically substitutable at the pharmacy (or by a pharmacist) for Concerta.

16. Methylphenidate (Clinical Therapeutics, 34(5), p1170, 2012) 16 Pharmaceutically equivalent? Not really Therapeutically equivalent? Not always

17. What are some of the right question(s) at the right time: during development and ANDA review for methylphenidate? Development What are the patient failure modes used in product design What does the literature say about the API/dosage forms What is the mechanism of modifying the release profile What are the controls for the consistency of the modified release coating What constitutes adherence to PE principles What are the specs and risk assessment Review Is the DH complete Were the patient failure modes used appropriate What’s the BE vs. the target demographic Does the level of PE cover all marketed dosage forms What is the appropriate RLD and has PE been established for it 17

18. RQRT #3: Mesalamine, RLD variable, generics? Asacol—Average dissolution curves for three different lots measured at (a) pH 6.5, (b) pH 6.8 (c) pH 7.2, JPS, 97(4), 4/08 Mean dissolution time as a function of the overall mean tablet coating thickness measured by terahertz pulsed imaging 18

19. Draft Guidance on Mesalamine 19 It’s known that the dissolution is dependent on coating thickness, what good is BE w/o PE?

20. What are some of the right question(s) at the right time: during development and NDA or ANDA review for mesalamine? Development What are the patient failure modes used in product design What does the literature say about the API/dosage forms What is the mechanism of modifying the release profile What are the CQAs and the controls for consistency of the modified release coating (deign space) What is the risk assessment Review Is the DH complete What is the scale-up strategy Is the CQA monitoring technology sufficient to assess process capability Is the use test (dissolution) sufficiently powered to detect variability at the appropriate level Should a bio-waiver be possible for the RLD and/or a generic 20

21. RQRT #4: Fosinopril Sodium: NCE ACE inhibitor NCE successfully developed in the 80’s NDAs are more complicated and there’s less prior knowledge However, using scientific principles and experience several challenges (all public) were addressed A Chemical stability and excipient interaction illustrates RQRT principles 21

22. Fosinopril Sodium (JPS 88(7), 1999, p696) Hypothesized reaction pathway(s) and excipient (MgSt) interaction was tested when observed in later development 22 Kinetics/risk analyses showed the effect was not significant to self- life

23. What are some of the right question(s) at the right time: for stability issues during development and NDA review for fosinopril sodium? Development What are the patient failure modes used in product design What does the literature and paper chemistry say about the API What excipient interactions might be expected What applicable ICH Q6A tenets apply What are the specs and risk assessment Review Is the DH complete Were the risks identified sufficiently explored Were the observed problems explained and supported by data/models Were corrective actions required Does the risk analysis and monitoring support the decision(s) 23

24. Right Question at the Right Time – Sponsors in practice The development history/report should be a log of the rationale and, therefore, the internal science and risk assessment questions are asked at the time Documenting unexpected or out of spec/trend results is essential in supporting decisions made and addressing related questions and “information requests” Every development history becomes part of the institutional knowledge base and an integral part of risk assessment, reduction and/or mitigation Adherence to the precepts of QbD and not just simple compliance with regulations and quidances is the most efficient method for quality product development Right question at the right time should lead to “Right the First Time” approvable applications on the first cycle 24

25. Right Question at the Right Time - FDA in practice FDA’s asking the right question – at the right time – critically contributes to effective and efficient development of pharmaceutical products and incorporates science and risk assessment To ask the right question at the right time FDA needs portions of development report available for review by reviewers with the necessary ‘peer’ expertise, the earlier the better and reviewed with a mind willing to understand Even with ICH Q8 – effective/congruent review of new drug development reports is yet to be achieved across the ICH regions, so the time is right to encourage this in the U.S. With the formation of OPQ and the efforts to align new and generic (including biotech) CMC review processes, the QbD-QbR approach holds promise but is in need of significant attention to ensure it is really using the knowledge gained to assess risk and incorporate the principles QbD 25

26. From where does the knowledge come? Knowledge Bases and Knowledge Management A Development History captured in the Development Report is the key to both guiding development and being ready for review. Essential aspects of the development history (DH) include: – The rationale for all phases of development Fundamental principles, prior knowledge, heuristics Adherence (not just compliance!) to Q8 and Q6 tenets are implicit in DH – The knowledge base created and effectively curated Makes the DH an electronic living document capturing all changes Usage of new and prior knowledge to make decisions Capturing of failure modes (patient/product) and sharing of knowledge between FDA review and inspection 26

27. What can you do with a knowledge base? Manufacturing Science in the 21 st Century Pharmaceutical: A Case Study on Quality by Design 27

28. What can you do with a knowledge base? “Ontological informatics infrastructure for pharmaceutical product development and manufacturing” Venkatasubramanian et al., Computers and Chemical Engineering 30 (2006) 1482 An ontology is a formal, explicit specification of a shared conceptualization Conceptualization refers to an abstract model of some phenomenon in the world by having identified the relevant concepts of that phenomenon. Explicit means that the type of concepts used, and the constraints on their use are explicitly defined. Formal refers to the fact that the ontology should be machine-readable. Shared reflects the notion that an ontology captures consensual knowledge i.e. it is not private to an individual, but accepted by a group 28 Formalizes asking the right question at the right time!

29. Summary: Needs and NIPTE/FDA Approaches Epistemology – Objective risk assessment Ontology – what is the “nature” of pharmaceutical knowledge and information Methodology – How are data generated and analyzed to create new knowledge EPISTOMOLOGY – Understanding of patient failure modes Determination of modes attributable to pharmaceutical properties/behavior Establishment of CQAs for design – Science of Equivalence – PE to TE, generics Basic research in materials DF design Establish PE for known CQAs Specifications and controls before clinicals ONTOLOGY – Knowledge curation and management Curation center for scientific and prior knowledge management – generalizable, reusable Bayesian – generate “new” prior knowledge Tipping point analysis and opportunities – Synthesis RQRT – Educational Initiatives METHODOLOGY – 21 st century analytical methodology and new technology – Science and Risk-based cGMPs and SOPs Adherence directed compliance 29