UNITED STATES PATENT AND TRADEMARK OFFICE
`
`___________________
`
`BEFORE THE PATENT TRIAL AND APPEAL BOARD
`
`__
`
`APOTEX INC.,
`Petitioner
`
`v.
`
`CELGENE CORPORATION,
`Patent Owner
`__
`
`Case IPR2023-00512
`U.S. Patent No. 8,846,628
`Issued: September 30, 2014
`
`Title:
`ORAL FORMULATIONS OF CYTIDINE ANALOGS AND METHODS OF USE THEREOF
`
`DECLARATION OF HANNAH K. BATCHELOR, Ph.D.
`
`By:
`
`February 10, 2023
`
`Apotex v. Cellgene - IPR2023-00512
`Petitioner Apotex Exhibit 1003-0001
`
`

`

`
`
`TABLE OF CONTENTS
`
` 
`
` 
`
`Page
`INTRODUCTION ......................................................................................... 1 
`I. 
`II.  QUALIFICATIONS ...................................................................................... 2 
`III.  PERSON OF ORDINARY SKILL IN THE ART ...................................... 5 
`IV.  DOCUMENTS ............................................................................................... 6 
`V. 
`SUMMARY OF OPINIONS ......................................................................... 8 
`VI.  BACKGROUND AND STATE OF THE ART .......................................... 9 
`A. 
`Pharmacokinetics ................................................................................... 9 
`B. 
`In Silico Modeling of PK Properties of Oral 5-Azacytidine ............... 10 
`VII.  IN SILICO MODELING OF PK PROPERTIES FOLLOWING
`THE ORAL ADMINISTRATION OF 200 MG AND 400 MG
`NON-ENTERIC COATED TABLETS OF 5-AZACYTIDINE .............. 13 
`A.  GastroPlus Inputs ................................................................................ 14 
`Generation of Support Files ...................................................... 15 
`Compound Parameters .............................................................. 18 
`Physiology Parameters .............................................................. 19 
`PK Parameters ........................................................................... 20 
`Simulation Parameters .............................................................. 20 
` 
`Simulation Results ............................................................................... 20 
`B. 
`VIII.  CONCLUSION ............................................................................................ 21 
`
`
` 
`
` 
`
`i
`
`Apotex v. Cellgene - IPR2023-00512
`Petitioner Apotex Exhibit 1003-0002
`
`

`

`
`
`I, Hannah K. Batchelor, Ph.D., of Glasgow, UK, declare as follows:
`
`I.
`
`INTRODUCTION
`
`1.
`
`I have been asked by counsel for Apotex Inc. (“Petitioner”) to
`
`investigate and provide opinions relating to the in silico modeling of
`
`pharmacokinetic (“PK”) properties following the oral administration of non-enteric
`
`formulations of 5-azacytidine for the above-captioned inter partes review (“IPR”).
`
`2.
`
`I understand that Petitioner petitions for IPR of certain claims of U.S.
`
`Patent No. 8,846,628 (“ʼ628 patent”) and request that the United States Patent and
`
`Trademark Office (“USPTO”) cancel the challenged claims. I further understand
`
`that Celgene Corporation (“Patent Owner”) purports to own the ʼ628 patent.
`
`3.
`
`In preparing this Declaration, I have reviewed and considered the
`
`documents identified in Section IV in light of the general knowledge in the
`
`relevant art. In forming my opinions, I relied upon my education, knowledge, and
`
`experience, including in the fields of biopharmaceutics and pharmaceutical
`
`science, and considered the level of ordinary skill in the art as discussed below.
`
`4.
`
`I expect to be called to provide expert testimony, if necessary,
`
`regarding the opinions and issues considered in this Declaration. This Declaration
`
`identifies my opinions to date. I reserve the right to amend or supplement this
`
`Declaration, if allowed under the relevant rules, to address any issues raised by
`
`Patent Owner’s expert(s) or resulting from further discovery relating to any of the
`
`1
`
`Apotex v. Cellgene - IPR2023-00512
`Petitioner Apotex Exhibit 1003-0003
`
`

`

`
`
`opinions stated herein. I may also testify as to the matters set forth in any
`
`additional reports, declarations, witness statements, and/or testimony submitted by
`
`Patent Owner.
`
`II. QUALIFICATIONS
`
`5.
`
`I am a Professor in Pharmaceutics at the University of Strathclyde in
`
`the Strathclyde Institute of Pharmacy and Biomedical Science in Glasgow, UK.
`
`From August 2015 to July 2020 I served as a Senior Lecturer in Pharmaceutics,
`
`Formulation and Drug Delivery at the University of Birmingham in the School of
`
`Pharmacy in Birmingham, UK. From 2015 to 2018 I served as a Director of
`
`Research, and from April 2012 to August 2015, I served as a Pediatric
`
`Formulations Research Fellow, also at the University of Birmingham.
`
`6.
`
`Prior to my time at the University of Birmingham, I was a Research
`
`Portfolio Manager at the Heart of England NHS Foundation Trust (HEFT) between
`
`2011 and 2012; a Senior Scientist in Biopharmaceutics at AstraZeneca between
`
`2008 and 2011; a Lecturer in Pharmaceutics at Aston University in Birmingham,
`
`UK between 2000 and 2007; and a Formulation Scientist at Reckitt and Colman
`
`(now ReckittBenckister) between 1996 and 1997.
`
`7.
`
`I earned a Ph.D. in Drug Delivery from the University of London,
`
`School of Pharmacy in London, UK. My Ph.D. research was investigating
`
`bioadhesive potential of alginate as a means of enhancing therapy for gastric reflux
`
`2
`
`Apotex v. Cellgene - IPR2023-00512
`Petitioner Apotex Exhibit 1003-0004
`
`

`

`
`
`by forming a coat on the esophagus. It was funded by Reckitt Benckiser and the
`
`EPSRC and resulted in four publications and influenced the advertising campaign
`
`for Gaviscon® showing the product coating the esophagus. I earned my Bachelors
`
`of Science in Pharmacology and Chemistry from the University of Sheffield in
`
`Sheffield, UK. I also have a Graduate Certificate in Statistics with Medical
`
`Applications from the University of Sheffield, and a Postgraduate Certificate in
`
`Learning and Teaching (SEDA accreditation as a Teacher in Higher Education)
`
`from the University of Aston.
`
`8. My research focuses on pediatric biopharmaceutics and development
`
`of age-appropriate medicines for children. Specifically, I develop testing strategies
`
`to predict in vivo performance; undertake pediatric in silico modeling to optimize
`
`pharmacokinetic study design and work to understand the impact of drug-food
`
`interactions within pediatric populations. I am actively involved in involvement of
`
`children and young people in research, particularly clinical research that impacts
`
`upon the treatment of this population.
`
`9.
`
`To date, I have received in excess of £2.5m in research funding and
`
`have supervised more than 15 PhD students to completion. My current live
`
`funding (>£600k) supports 6 PhD students and a research technician. I have been
`
`invited to present my research at several relevant international conferences
`
`including: European Paediatric Formulation Initiative (EuPFI); American
`
`3
`
`Apotex v. Cellgene - IPR2023-00512
`Petitioner Apotex Exhibit 1003-0005
`
`

`

`
`
`Association of Pharmaceutical Scientists (AAPS); Controlled Release Society and
`
`the Royal College of Paediatrics and Child Health annual meeting.
`
`10.
`
`I have authored or co-authored chapters of several books in my field.
`
`I am a member of the Editorial Board for two journals in my field,
`
`Biopharmaceutics and Drug Disposition and Nature Scientific Reports, and am an
`
`editor of two books, including Biopharmaceutics: From Fundamentals to
`
`Industrial Practice, John Wiley & Sons Ltd. I also peer review for the following
`
`publications: AAPS PharmSciTech; Acta Biomaterialia; Acta Paediatrica;
`
`Advanced Drug Delivery Reviews; Archives of Disease in Childhood; BMJ
`
`Paediatrics Open; British Journal of Clinical Pharmacology; European Journal of
`
`Hospital Pharmacy; European Journal of Pharmaceutical and Biopharmaceutics;
`
`European Journal of Pharmaceutical Science; Expert Opinion on Drug Delivery;
`
`International Journal of Pharmaceutics; Journal of Asthma; Journal of Controlled
`
`Release; Journal of Drug Targeting; Journal of Pharmacy and Pharmacology;
`
`Molecular Pharmaceutics; Nanomedicine; Pharmaceutica Analytica Acta;
`
`Pharmaceutical Research; Pharmaceutical Sciences; PLOSOne; and The Journal
`
`of Pediatrics.
`
`11.
`
`12.
`
`I am listed as an inventor on at least one patent or patent application.
`
`I am the current Chair of the Academy of Pharmaceutical Scientists
`
`(APS), and have served as a committee member and Chair on the
`
`4
`
`Apotex v. Cellgene - IPR2023-00512
`Petitioner Apotex Exhibit 1003-0006
`
`

`

`
`
`Biopharmaceutics focus group within APS and as a committee member of the New
`
`Scientists Focus Group within APS. I am a Workstream Leader of the
`
`biopharmaceutics theme within the European Paediatric Formulation Initiative
`
`(EuPFI). I am also a member of UNGAP (The European Network on
`
`Understanding Gastrointestinal Absorption-related Processes), Expert Advisory
`
`Group on Medicinal Chemistry to the British Pharmacopoeia, Standards
`
`Committee for IPEC (International Pharmaceutical Excipients Committee), Drug
`
`Delivery Research Network (DDRN) (founding member), and Young Academic
`
`Network for Chemical Engineers (YANCE). I also serve as a formulations expert
`
`within the Connect 4 Children Collaborative Network for European Clinical Trials
`
`for Children (c4c).
`
`13. A summary of my education, experience, publications, awards and
`
`honors, patents, publications, and presentations is provided in my CV, a copy of
`
`which is attached as Exhibit A to this Declaration.
`
`14.
`
`I am being compensated for my time in connection with this IPR at
`
`my standard consulting rate, which is £250.00 per hour. My compensation is not
`
`dependent in any way upon the outcome of this matter.
`
`III. PERSON OF ORDINARY SKILL IN THE ART
`
`15.
`
`I have been asked to investigate and provide opinions from the
`
`perspective of a person of ordinary skill in the art (“person of ordinary skill” or
`
`5
`
`Apotex v. Cellgene - IPR2023-00512
`Petitioner Apotex Exhibit 1003-0007
`
`

`

`
`
`“POSA”) at the time of the priority date of the challenged claims of the ʼ628
`
`patent, which I understand is December 5, 2008.
`
`16.
`
`I understand that a POSA relating to the subject matter of the ʼ628
`
`patent would have had (1) a Pharm.D., or a Ph.D. in pharmaceutical sciences,
`
`chemical engineering, chemistry, or related discipline; and (2) at least two to four
`
`years of experience with pharmaceutical design, formulation, development, and/or
`
`manufacturing of oral dosage forms. A POSA may also work as part of a
`
`multidisciplinary team and draw upon not only his or her own skills, but also take
`
`advantage of certain specialized skills of others in the team to solve a given
`
`problem. To the extent necessary, this person would have worked in collaboration
`
`with others with the requisite education and experience in candidate drug selection,
`
`clinical use, clinical testing, design, formulation, development, and/or
`
`manufacturing of pharmaceutical oral dosage forms.
`
`17.
`
`I have applied this definition of a POSA herein.
`
`IV. DOCUMENTS
`
`18.
`
`In preparing this Declaration, I reviewed and considered the
`
`documents identified in the below table:
`
`Exhibit No.
`
`1014
`
`Description
`
`Chan et al., “5-Azacytidine Hydrolysis Kinetics Measured by High-
`Pressure Liquid Chromatography and 13C-NMR Spectroscopy,”68
`(7) J Pharma. Scis. 807 (1979) (“Chan”)
`
`6
`
`Apotex v. Cellgene - IPR2023-00512
`Petitioner Apotex Exhibit 1003-0008
`
`

`

`
`
`Exhibit No.
`
`Description
`
`1017
`
`1018
`
`1027
`
`1039
`
`1040
`
`1041
`
`1042
`
`1043
`
`1044
`
`1045
`
`1046
`
`1047
`
`Marcucci, G., et al., “Bioavailability of Azacitidine Subcutaneous
`Versus Intravenous in Patients With the Myelodysplastic
`Syndromes,” 45 J. Clin. Pharmacology 597 (2005) (“Marcucci”)
`
`Thomson, A., “Back to basics: pharmacokinetics,” 272 Pharma. J
`769 (2004) (“Thomson”)
`
`GastroPlus version 5.2 Manual (“GP5.2 manual”)
`Buck et al., Prediction of Human Pharmacokinetics Using
`Physiologically Based Modeling: A Retrospective Analysis of 26
`Clinically Tested Drugs, Drug Metabolism & Disposition,
`35(10):17661780 (2007) (“Buck”)
`Dannenfelser et al., Development of Clinical Dosage Forms for a
`Poorly Water Soluble Drug I: Application of Polyethylene Glycol–
`Polysorbate 80 Solid Dispersion Carrier System, J Pharma. Sci.,
`93(5):11651175 (2004) (“Dannenfelser”)
`Press Release, Simulations Plus, Inc., “Simulations Plus Releases
`GastroPlus(™) 5.2” (Nov. 30, 2006) (“GP5.2 PR”)
`IARC Monographs on the Evaluation of Carcinogenic Risks to
`Humans, Pharmaceutical Drugs, Vol. 50, 1990 (“IARC”)
`Israili et al., The Disposition and Pharmacokinetics in Humans of 5-
`Azacytidine Administered Intravenously as a Bolus or by Continuous
`Infusion, Cancer Res. 36:14531461 (1976) (“Israili”)
`Kuentz et al., A strategy for preclinical formulation development
`using GastroPlusTM as pharmacokinetic simulation tool and a
`statistical screening design applied to a dog study, Euro. J. Pharma.
`Sci., 27:9199 (2006) (“Kuentz”)
`Li et al., IV-IVC Considerations in the Development of Immediate-
`Release Oral Dosage Form, J Pharma. Sci. 94(7):13961417 (2005)
`(“Li”)
`Obach et al., The Prediction of Human Pharmacokinetic Parameters
`from Preclinical and In Vitro Metabolism Data, J. Pharma. & Exp.
`Thera., 283(1):4658 (1997) (“Obach”)
`Parrott & Lavé, Prediction of intestinal absorption: comparative
`assessment of gastroplus™ and idea™, Euro. J. Pharma. Sci.,
`17:5161 (2002) (“Parrott”)
`
`7
`
`Apotex v. Cellgene - IPR2023-00512
`Petitioner Apotex Exhibit 1003-0009
`
`

`

`
`
`Exhibit No.
`
`Description
`Simulations Plus, Inc., Form 10-QSB (January 16, 2007) (“SLP 10-
`QSB”)
`The SDF file for the PubChem compound entry for Azacitidine
`(PubChem CID 9444) (September 16, 2004)
`(“Conformer3D_CID_9444 (1).sdf”), printed to PDF
`ADMET Predictor Output file (“Azacitidine ADMET.txt”), printed
`to PDF
`CDD file for 200 mg dose (“Azacitidine 200 mg.cdd”), printed to
`PDF
`IV plasma concentration-time data file for 200 mg dose
`(“Azacitidine 200 mg.ipd”), printed to PDF
`Result data file for 200 mg dose (“Azacitidine 200 mg All Data.txt”),
`printed to PDF
`CDD file for 400 mg dose (“Azacitidine 400 mg.cdd”), printed to
`PDF
`IV plasma concentration-time data file for 400 mg dose
`(“Azacitidine 400 mg.ipd”), printed to PDF
`Result data file for 400 mg dose (“Azacitidine 400 mg All Data.txt”),
`printed to PDF
`V. SUMMARY OF OPINIONS
`
`1048
`
`1052
`
`1053
`
`1054
`
`1055
`
`1056
`
`1057
`
`1058
`
`1059
`
`19.
`
`In silico modeling of PK properties (AUC, Cmax, and Tmax) following
`
`the oral administration of 200 mg and 400 mg non-enteric coated (“non-EC”)
`
`tablets of 5-azacytidine to a human test subject results in the following PK
`
`properties:
`
`PK Property
`
`AUC (area under the curve)
`
`Cmax (maximum plasma concentration achieved
`during a period of time)
`
`8
`
`Simulation Result
`264.72 ng*h/mL
`(200 mg dose)
`529.46 ng*h/mL
`(400 mg dose)
`105.3 ng/mL
`(200 mg dose)
`
`Apotex v. Cellgene - IPR2023-00512
`Petitioner Apotex Exhibit 1003-0010
`
`

`

`
`
`Tmax (time necessary to reach Cmax)
`
`210.6 ng/mL
`(400 mg dose)
`67.2 min
`(200 and 400 mg dose)
`
`
`(Appx II.D, Appx II.E, Appx III.D, Appx III.E; EX-1056; EX-1059.)
`
`VI. BACKGROUND AND STATE OF THE ART
`
`A. Pharmacokinetics
`
`20. Pharmacokinetics (“PK”) is the study of the relationship between drug
`
`administration and the resulting levels of that drug in the body. (EX-1018,
`
`Thomson at 769 (summarizing principles of pharmacokinetics and important
`
`parameters).) One important aspect of PK is drug concentration in plasma, or
`
`“plasma concentration.” A profile that reports plasma concentration of a drug over
`
`time (i.e., a plasma concentration-time profile) can be used to calculate PK
`
`properties such as: (1) Area Under Curve (“AUC”) (total amount of drug present in
`
`the plasma over a period of time, measured by the area under the curve of the
`
`plasma concentration-time profile); (2) Cmax (maximum plasma concentration
`
`achieved during a period of time); and (3) Tmax (time necessary to reach Cmax).
`
`(EX-1017, Marcucci at 599-600 (reporting pharmacokinetic properties of 5-
`
`azacytidine following subcutaneous or intravenous administration); EX-1027,
`
`GP5.2 manual at 39 (user manual for GastroPlus software version 5.2).)
`
`21. The PK of a drug may inform how to optimize its dose, dosing
`
`schedule, or even dosage form. In particular, this information may be used to
`
`9
`
`Apotex v. Cellgene - IPR2023-00512
`Petitioner Apotex Exhibit 1003-0011
`
`

`

`
`
`maintain the drug’s plasma concentration within its therapeutic window for a
`
`desired period of time, which may inform drug product design or how it is
`
`administered to make it more effective, safe, and/or convenient. (EX-1018,
`
`Thomson at 769.)
`
`B. In Silico Modeling of PK Properties of Oral 5-Azacytidine
`
`22. Drug development, particularly clinical testing of drug product
`
`candidates, is expensive and time consuming. (EX-1046, Obach at 46 (study
`
`validating human PK prediction methods); EX-1047, Parrott at 51-52 (evaluation
`
`of in silico prediction software, GastroPlus, as a tool for drug discovery and
`
`development); EX-1045, Li at 1413 (predictive tools are important to minimize
`
`costly animal and human experiments during drug development).) One method of
`
`mitigating the time and expense associated with clinical testing was to utilize in
`
`silico methods. In silico physiologically-based pharmacokinetic methods were
`
`developed for modeling PK of oral dosage forms, enabling scientists to more
`
`readily identify the best candidates for drug product development. (EX-1047,
`
`Parrott at 51-52; EX-1045, Li at 1413; EX-1039, Buck at 1766 (confirming
`
`successful human PK predictions by GastroPlus for 26 clinically tested drugs); EX-
`
`1044, Kuentz at 99 (GastroPlus successfully modeled drug absorption based on
`
`preformulation data).) These models were routinely used during early drug
`
`discovery, preclinical development, and clinical trials. (EX-1039, Buck at 1766;
`
`10
`
`Apotex v. Cellgene - IPR2023-00512
`Petitioner Apotex Exhibit 1003-0012
`
`

`

`
`
`EX-1044, Kuentz at 92-93; EX-1045, Li at 1409-1410; EX-1047, Parrott at
`
`Abstract; EX-1048, SLP 10-QSB at 21 (“GastroPlus is the ‘gold standard’ in the
`
`industry for its class of simulation software” and used by “virtually every major
`
`pharmaceutical company”).)
`
`23. Physiologically-based in silico models have been routinely and widely
`
`used since the early 2000s to determine the expected PK of orally administered
`
`pharmaceutical compositions. (EX-1039, Buck at 1766; EX-1044, Kuentz at 93;
`
`EX-1045, Li at 1409-1410; EX-1047, Parrott at 52; EX-1041, GP5.2 PR (“This
`
`latest release of GastroPlus adds several important improvements … new features
`
`[of version 5.2] provide greater accuracy for certain types of simulations . . . we’re
`
`now getting reports from the field that GastroPlus simulation results are becoming
`
`required in many of their internal project reports”); EX-1027, GP5.2 manual.) In
`
`silico modeling had been recognized as being able to accurately predict in vivo PK
`
`for orally administered drugs. (EX-1047, Parrott at 51; EX-1039, Buck at 1766.)
`
`24. The clinical bioavailability and in vivo PK for 5-azacytidine dosage
`
`forms were well known for SC and IV administration. (EX-1017, Marcucci.)
`
`Based on the foregoing and known physical and chemical properties of 5-
`
`azacytidine, in silico models of the oral dosage forms of 5-azacytidine could have
`
`readily been constructed.
`
`11
`
`Apotex v. Cellgene - IPR2023-00512
`Petitioner Apotex Exhibit 1003-0013
`
`

`

`
`
`25. One example of a routinely-used and widely-known PK modeling
`
`software is GastroPlus (version 5.2) (“GP5.2”).1 (EX-1041, GP5.2 PR; EX-1027,
`
`GP5.2 manual at 4-9.) GP5.2 and its earlier versions were available before 2007,
`
`before the priority date of the challenged claims of the ʼ628 patent, which I
`
`understand is December 5, 2008. (EX-1027, GP5.2 manual (dated November
`
`2006).)
`
`26. Major pharmaceutical companies such as Roche, Johnson & Johnson,
`
`and Novartis employed GastroPlus (EX-1044, Kuentz at 91 (reporting use of
`
`GastroPlus at Roche in clinical formulation development); EX-1039, Buck at 1766
`
`(reporting validation of GastroPlus predictions by Johnson & Johnson
`
`Pharmaceutical Research and Development); EX-1040, Dannenfelser at 1165
`
`(Novartis using GastroPlus to predict the oral bioavailability of a compound)),
`
`demonstrating that this software was “the tool of choice” for modeling PK when
`
`integrating in vitro and in vivo data with in silico prediction. (EX-1047, Parrott at
`
`60; EX-1044, Kuentz at 93; EX-1045, Li at 1409-1410). GP5.2 and its earlier
`
`versions were capable of generating in silico models for oral dosage forms based
`
`on in vivo intravenous PK results. (EX-1041, GP5.2 PR; EX-1027, GP5.2 manual;
`
`
`1 GP5.2 was commercially available from Simulations Plus
`
`(https://www.simulations-plus.com/).
`
`12
`
`Apotex v. Cellgene - IPR2023-00512
`Petitioner Apotex Exhibit 1003-0014
`
`

`

`
`
`EX-1018, Thomson at 769-70; EX-1039, Buck at 1766; EX-1044, Kuentz at 93;
`
`EX-1045, Li at 1409-1410; EX-1047, Parrott at 52.)
`
`27. Another useful feature of GP5.2 is its ability to model the in vivo rate
`
`of chemical degradation based on in vitro degradation data. (EX-1027, GP5.2
`
`manual.) Such degradation data for 5-azacytidine was well known and reported in
`
`the literature. (EX-1014, Chan.) Thus, GP5.2 accounts for any degradation in the
`
`gastrointestinal tract when determining the PK of an oral dosage form of 5-
`
`azacytidine.
`
`28. Prior to testing known compositions such as 5-azacytidine in humans,
`
`a POSA would have been motivated to obtain data on in silico modeling, for
`
`example, from others of skill in the art that a POSA may have collaborated with,
`
`using GP5.2 based on at least reduced time and cost considerations.
`
`VII.
`
`IN SILICO MODELING OF PK PROPERTIES FOLLOWING THE
`ORAL ADMINISTRATION OF 200 MG AND 400 MG NON-
`ENTERIC COATED TABLETS OF 5-AZACYTIDINE
`
`29.
`
`I have been asked to investigate and provide opinions relating to the in
`
`silico modeling of PK properties following the oral administration of a 200 mg
`
`non-EC tablet of 5-azacytidine and a 400 mg non-EC tablet of 5-azacytidine.
`
`30.
`
`In silico modeling of PK properties following the administration of a
`
`200 mg non-EC formulation of 5-azacytidine to a human test subject resulted in an
`
`AUC of 267.72 ng*h/mL, a Cmax of 105.3 ng/mL, and a Tmax of 67.2 minutes. In
`
`13
`
`Apotex v. Cellgene - IPR2023-00512
`Petitioner Apotex Exhibit 1003-0015
`
`

`

`
`
`silico modeling of PK properties following the administration of a 400 mg non-EC
`
`formulation of 5-azacytidine to a human test subject resulted in an AUC of 529.46
`
`ng*h/mL, a Cmax of 210.6 ng/mL, and a Tmax of 67.2 minutes.
`
`31. GP5.2 inputs and results are discussed below for the simulation that I
`
`conducted.
`
`A. GastroPlus Inputs
`
`32. GastroPlus simulation (i.e., in silico modeling) involves input of
`
`compound, physiology, PK, and simulation parameters. All parameters for the
`
`simulation were known and available to a POSA including structural information,
`
`degradation characteristics, and known PK of other dosage forms. Such
`
`parameters were inputted, and the simulation was run, in line with how a POSA
`
`would have used the GastroPlus software.2 (See, e.g., EX-1041, GP5.2 PR; EX-
`
`1027, GP5.2 manual; EX-1045, Li at 1409-1410; EX-1047, Parrott at 52.)
`
`
`2 Certain relevant inputs were analyzed in the ADMET (Absorption, Distribution,
`
`Metabolism, Excretion, and Toxicity) or PKPlus modules to generate certain
`
`parameters used by GP5.2. The output from the ADMET analysis is provided
`
`herewith. (EX-1053, Azacitidine ADMET.TXT.) This is consistent with how a
`
`POSA would have used the GastroPlus software.
`
`14
`
`Apotex v. Cellgene - IPR2023-00512
`Petitioner Apotex Exhibit 1003-0016
`
`

`

`
`
` Generation of Support Files
`
`33. GastroPlus utilizes well-known information, including structural
`
`information to generate values for physiochemical properties of a drug, for use in
`
`pharmacokinetic simulation. (EX-1041, GP5.2 PR; EX-1027, GP5.2 manual; EX-
`
`1045, Li at 1409-1410; EX-1047, Parrott at 52; see also Appx I.A.) These
`
`physiochemical properties include molecular weight, logP, solubility, diffusion
`
`coefficient, estimated effective permeability, plasma protein binding, and volume
`
`of distribution. (Appx I.A.2.) The 3D structure of 5-azacytidine was obtained
`
`from PubChem as an SDF file.3
`
`34. To the extent a compound degrades following oral administration,
`
`chemical stability data can be used to model the amount of the compound that
`
`remains available for absorption. The degradation rate constants for 5-azacytidine
`
`hydrolysis in aqueous buffer solutions at 37 °C were reported for various pH
`
`values between 4.5 and 8.0. (EX-1014, Chan at 810, Table II, 811, Fig. 8.) The
`
`percent degradation per hour can be calculated using the following equation:4
`
`
`3 The PubChem compound entry for Azacitidine (PubChem CID 9444) was
`
`created on September 16, 2004. The SDF file was obtained from
`
`https://pubchem.ncbi.nlm.nih.gov/compound/Azacitidine#section=3D-Conformer
`
`and is provided herewith. (EX-1052.)
`
`4 Adapted from equation 3 of Chan. (EX-1014, Chan at 809.)
`
`15
`
`Apotex v. Cellgene - IPR2023-00512
`Petitioner Apotex Exhibit 1003-0017
`
`

`

`
`
`% degradation/hr = 100 – (100 e (-k *60)),
`
`where k is the degradation rate constant
`
`35. Based on the reported degradation rate constants reported in Chan, the
`
`percent degradation per hour were calculated for each of the pH values as shown in
`
`the table below.
`
`pH
`
`Degradation rate constant
`
`% degradation/hr
`
`4.5
`15.6 x 10-3 M, min-1= 0.0156 min-1
`60.78
`5.6
`11.2 x 10-3 M, min-1= 0.0112 min-1
`48.93
`7.0
`7.33 x 10-3 M, min-1= 0.00733 min-1
`35.58
`7.4
`12.8 x 10-3 M, min-1 = 0.0128 min-1
`53.61
`8.0
`35 x 10-3 M, min-1 = 0.035 min-1
`87.75
`36. The above data was entered into GastroPlus to generate a chemical
`
`degradation data file (CDD file). (Appx I.B; EX-1054, Azacitidine 200mg.cdd;
`
`EX-1057, Azacitidine 400mg.cdd.)
`
`37. Predictions of distribution and clearance can also be generated based
`
`on in vivo human preclinical intravenous pharmacokinetic data or in vitro data.
`
`(EX-1046, Obach at 46-47.) In vivo human preclinical intravenous
`
`pharmacokinetic data for 5-azacytidine was reported. (EX-1017, Marcucci at 600,
`
`Table 1, Fig. 1.) The data reported in Figure 1 of Marcucci was extracted using
`
`image-to-data software to ensure its accuracy. Figure 1 of Marcucci and the data
`
`extracted therefrom are reproduced below:
`
`16
`
`Apotex v. Cellgene - IPR2023-00512
`Petitioner Apotex Exhibit 1003-0018
`
`

`

`
`
`
`
`Time (h)
`
`Plasma Concentration
`(ng/mL)
`0.0
`0
`0.08
`2620
`0.17
`2514
`0.5
`377
`1.0
`119
`2.0
`27
`38. The intravenous plasma concentration-time data from Marcucci was
`
`entered into the PKPlus module in GastroPlus to generate an intravenous plasma
`
`data file (IPD file). (Appx I.C; EX-1055, Azaciditine 200mg.ipd; EX-1058,
`
`Azaciditine 400mg.ipd.) PKPlus generated compartmental models and a non-
`
`compartmental model for calculating PK parameters from the plasma
`
`concentration-time profile. (Appx I.D.) The two-compartment model was selected
`
`based on the teachings in Israili that the analysis of in vivo data “fits a 2-
`
`compartment model (r > 0.95).” (EX-1043, Israili at 1455, 1457; Appx I.D.4.)
`
`17
`
`Apotex v. Cellgene - IPR2023-00512
`Petitioner Apotex Exhibit 1003-0019
`
`

`

`
`
`The R2 value of 0.9995 reported by PKPlus confirmed that the Marcucci data fit
`
`the two-compartment model. (Appx I.D.4.)
`
`
`
`Compound Parameters
`
`39. The parameters related to properties of the drug compound, 5-
`
`azacytidine, were entered in the “Compound” tab.
`
`40. First, inputs relating to the dosage form and initial dose were selected
`
`to reflect the non-EC tablets containing either 200 mg or 400 mg of 5-azacytidine.
`
`(Appx II.A; Appx III.A.) Specifically, an immediate release (IR) tablet was
`
`selected as an input for the dosage form to reflect the non-EC tablets containing
`
`either 200 mg or 400 mg of 5-azacytidine. Values for subsequent doses were not
`
`changed and kept at 0 mg as the model was for a single dose response.
`
`41. Next, well-known information related to the chemical properties of 5-
`
`azacytidine were available to a POSA, including from the International Agency for
`
`Research on Cancer (IARC). (EX-1042, IARC at 47-63.) Such information for 5-
`
`azacytidine included the molecular formula (C8H12N4O5), molecular weight (244
`
`g/mol), and solubility in various solvents at different pH values. (EX-1042, IARC
`
`at 47-48; EX-1043, Israili at 1458.) The pH for the reference solubility was set at
`
`pH 1.1 to reflect the pH of stomach acid and the solubility at that pH 1.1 was
`
`utilized. (EX-1042, IARC at 48 (28 mg/mL in 0.1 N hydrochloric acid).) The
`
`diffusion coefficient was calculated by GastroPlus based on molecular weight. In
`
`18
`
`Apotex v. Cellgene - IPR2023-00512
`Petitioner Apotex Exhibit 1003-0020
`
`

`

`
`
`addition, the octanol-water partition coefficient of the neutral drug molecule is
`
`disclosed as being <0.005. (EX-1043, Israili at 1458.) The logP of 5-azacytidine
`
`can be calculated from the partition coefficient using the equation below:
`
`log10 (Partition Coefficient) = LogP
`
`log10 (0.005) = - 2.3
`
`42. The parameters based on the known physical and chemical properties
`
`of 5-azacytidine were generated by the ADMET module and thereafter utilized.
`
`(Appx II.A.; Appx II.A.1; Appx III.A.; Appx III.A.1; EX-1053.) Additionally,
`
`previously generated chemical degradation data was utilized, as I described above
`
`in Section VII.A.1.
`
`43. Finally, the remaining inputs included dosage volume (250 mL),
`
`particle size (25 μm), mean precipitation time (900 seconds), and drug particle
`
`density (1.2 g/mL). (Appx II.A; Appx III.A.) Default values for these parameters
`
`were prepopulated and were left unchanged, which is consistent with how a POSA
`
`would have used the GastroPlus software.
`
`
`
`Physiology Parameters
`
`44. The parameters that define the type of clinical trial to be simulated
`
`and the model used for the simulation were entered in the “Physiology” tab. The
`
`prepopulated default values of “Human – Physiological – Fasted” and “Opt logD
`
`Model” for the ASF Model were unchanged. (Appx II.B; Appx III.B.)
`
`19
`
`Apotex v. Cellgene - IPR2023-00512
`Petitioner Apotex Exhibit 1003-0021
`
`

`

`
`
`
`
`PK Parameters
`
`45. The PK related parameters such as the subject’s body weight,
`
`blood/plasma concentration ratio, unbound percent in plasma, and distribution and
`
`clearance related values can be entered in the “Pharmacokinetics” tab. (Appx II.C;
`
`Appx III.C.) The blood/plasma concentration ratio (~0.8) and unbound percent in
`
`plasma (>99%) for 5-azacytidine was known. (EX-1043, Israili at 1453, 1458.)
`
`The distribution and clearance related values generated by PKPlus (EX-1055; EX-
`
`1058) from the PK data in Marcucci (EX-1017) were loaded and the body weight
`
`was entered based on the average body weight (76.8 kg) reported in Marcucci.
`
`(EX-1017, Marcucci at 599.) All other parameters were unchanged from their
`
`default values.
`
`
`
`Simulation Parameters
`
`46. All remaining simulation parameters are specified in the “Simulation”
`
`tab. The previously generated chemical degradation data was loaded. The default
`
`option of “Built-in pKa-based Solubility Model” and default simulation length or
`
`24 hours was used. (Appx II.D; Appx III.D.)
`
`B. Simulation Results
`
`47. The results of the simulation for the oral administration of 200 mg and
`
`400 mg non-EC tablets of 5-azacytidine to a human test subject are summarized in
`
`the table below.
`
`20
`
`Apotex v. Cellgene - IPR2023-00512
`Petitioner Apotex Exhibit 1003-0022
`
`

`

`
`
`PK Property
`
`AUC (area under the curve)
`
`Cmax (maximum plasma concentration achieved
`during a period of time)
`
`Tmax (time necessary to reach Cmax)
`
`Simulation Result
`264.72 ng*h/mL
`(200 mg dose)
`529.46 ng*h/mL
`(400 mg dose)
`105.3 ng/mL
`(200 mg dose)
`210.6 ng/mL
`(400 mg dose)
`67.2 min
`(200 and 400 mg dose)
`
`(Appx II.D, Appx II.E, Appx III.D, Appx III.E, EX-1056; EX-1059.)
`
`VIII. CONCLUSION
`
`48.
`
`I hereby declare that all statements made herein of my own
`
`knowledge are true and tha