[=ete
`
`FRESENIUS EXHIBIT 1026
`Page 1 of 54
`
`a)=ane—aeee
`
`tam
`
`c=,
`
`ee
`
`FRESENIUS EXHIBIT 1026
`Page 1 of 54
`
`

`

`'Ubtiif! ~ ·Ooligltffl Ca,~~-:PDV._imll ·hla
`
`ptm.~~ ~'1nm l edimd'IJ:)•
`
`~~m -of biopl . ··c1rac .~
`.icmnA. ~
`.
`i,. ~m.. ..,. <l}<.ngs:* -~~ ~ V- S;i_}
`!nd~inda,
`.
`?-!,,":SN o.,.;24 •m 1 ·X (~
`ra: ·dk. ~-)
`t ~~it21 ~~oo~, 2 h~t~ ~u1imu ..
`l ~eanpQ; Iutm A. ll . _ : ' .
`RS . 'J 048 199;
`6 tS\6---dt'l I
`
`wfk WW~ oal~~ d i~ 03. ·tiiit -~
`·¥ffim ~
`lfli,"rile: ·kt ~ ·~
`lf'~r ffit;~ inf~ioo..
`,qi~titil;ffii.
`'· ·
`Ma:tt~ di~ 3&1..~ below_
`
`··in _ bulk
`· -_ .. 1uf~,ooal
`
`9'1:···• 5956
`~
`
`~lbe:,r 1M ~ iIIQll' :i!UJ JFW:I. llffl{f be repn:id~ m lmumaH~ in my
`fu m ,or lb~
`iel ecit,ii)J}te Of tmd~i me{~ Jlbo•~i_. , .·
`,/ ~
`.
`iuy i n~ n •~ e Uid·.- ll!§riaw
`mkrofi!l4nil'I&~ ~ ~rdi~. orb,.
`s~~ ~i lhtmt ~ f t m ~'ridng ft~~ p ;~.
`·
`
`~ IIAal.liL DE-KJKER. ilf~"e.
`2.10 M':ooi'.ilim At-.i , · ·
`•. -
`tmp.~liVh iW.ddf&n.
`
`'Yolk" · . • Yett $00 6
`
`,( l uwni Firntin ~ •·" digit ~
`109·" 165-4l.12 l
`
`FRESENIUS EXHIBIT 1026
`Page 2 of 54
`
`

`

`---
`
`4
`
`Preformulation Development of
`Parenteral Biopharmaceuticals
`
`JOHN A. BONTEMPO
`Biopharmaceulical Product Development, East Brunswick, New Jersey
`
`I.
`
`II.
`
`INTRODUCTION
`A. Considerations of Domestic and I ntemational
`Distribution of the Product
`B. Points to Con5idc:r for Constituted Versus
`Lyophil1zed Fonnulations
`C. Unit Dose or Multidose
`D. Physicochemical Factors to Be Considered for
`Protein Drug Formulations
`
`IN fTIAL PREFORMULATION STUDIES:
`PARAMETERS AND VARIABLES TO BE T E STED
`A.
`Initial Variables to Be Tested
`B. Preliminary Analytical Development
`C. Experimental Conditions for the Initial
`Prcfonnulation Studies
`
`Ill. MECHANICAL AND PHYSICAL STRESSES
`A. Shaking E ffect on Protein Solution al the
`Preformulation Level
`B. Freeze- Thaw Experiments
`
`92
`
`93
`
`93
`94
`
`94
`
`95
`95
`96
`
`97
`99
`
`99
`100
`
`"f
`
`FRESENIUS EXHIBIT 1026
`Page 3 of 54
`
`

`

`92
`
`Bontempo
`
`C. Filling Systems
`D. Stability Evaluation
`IV, DEGRADATION MECHANISMS
`A. Oxidalion
`B. Deamidation
`C. Hydrolysis
`D. Racemizalion
`E.
`lsomerization
`F. Disulfide Exchange
`G. Beta-Elimination
`V. PHYSICAL DEGRADATLONS
`A. Covalent Aggregation
`B. Noncovalenl
`C. Aggregation
`D. Denaturation
`E. Precipitation
`F. Adsorption
`VI. SUMMARY
`
`REFERENCES
`
`JOO
`101
`
`102
`!02
`103
`l03
`!03
`103
`103
`104
`
`104
`104
`104
`104
`105
`105
`106
`
`I06
`
`107
`
`I. INTRODUCTION
`Preformulation research studies of protein therapeutics encompass biophar(cid:173)
`maceutical, physicochemical, and analytical investigations fo support of
`subsequent stable formulations for preclinical, clinical, and market usage.
`In this highly competitive protein therapeutics field, it is very impor(cid:173)
`tant to obtain significant, measurable progress with preformulations studies
`in a timely manner. How extensive these studies are will depend on the
`availability of the crude, active drug substance and the intended route
`of administration. Most often, these studies begin with exfremely small
`amounts of crude bulk active substance and, as more material becomes
`available with greater purity, more studies are initiated.
`From an industrial point of view, the preform1Jlation studies are de(cid:173)
`signed to cover a wide range or properties in a shon time to learn as much
`as possible, but not in great depth. The pharmaceutical fonnulation scien(cid:173)
`tist is very much interested in identifying potential problems early enough
`to evaluate potential alternatives to stabilize future form.ulation(s).
`
`FRESENIUS EXHIBIT 1026
`Page 4 of 54
`
`

`

`Preformuletion Development
`
`93
`
`As previously stated, there must be a strong interdisciplinai)' collabo(cid:173)
`ration team .to review, identify, and maximize the most productive leads
`toward fonnulation development. Preformulation studies are short in du(cid:173)
`ration, two to three months, and some of these are performed under varying
`stress conditions which will be described later in this chapter, It is important
`to remember that no two proteins are alike and studies designs will vary
`case by case.
`Prior to the onset of prefonnulations, the pharmaceutical team must
`review some very important factors which will have an impact on the pre(cid:173)
`fonnulation and formulation development.
`
`A. Considerations of Domestic and lnternatlonat
`Distribution of the Product
`Many global joint ventures and partnerships today in the biophannaceuti(cid:173)
`cal industry dictate various pharmaceutical, clinical, and marketing strate(cid:173)
`gies. The regulatory requirements and acceptance of formulation excipi(cid:173)
`ents, packaging components, unit dose versus multidose product, and
`stability conditions vary from continent to continent. Constituted and/or
`lyophilized dosage forms must also be considered. The development of for(cid:173)
`mulation considerations should be on a worldwide acceptance basis.
`
`8 . Points to Consider for Constituted Versus Lyophillzed
`Formulations
`Some of the key points to be -considered for a constituted formulation arc:
`• A constituted formulation may be less stable than a lyophilized one
`• Effect of agitation during manufacturing and shipping
`• Interaction of the liquid with the inner wall of the glass vial and
`with the elastomeric closure
`• Aggregation problems
`• Head space within the vial
`• Preservative effectiveness
`Some of the key points to be considered for a lyophilized formulation
`
`are:
`
`• Better stability than a constituted product
`- Determination of an optimal lyophilization cycle
`• Effects of residual moisture on the activity and stability of the product
`• Ease of reconstitutability. Clinicians, nurses and trained home us•
`ers, prefer reconstitutability of the product within two minutes.
`
`FRESENIUS EXHIBIT 1026
`Page 5 of 54
`
`

`

`--.....,,--
`
`1
`
`I
`
`Bontempo
`
`• Stability of the reconstituted product
`• Preservative effectiveness (ifth1s is a multidose product)
`• Cost effectiveness, Lyophilization technology is expensive along
`with cost of utilities
`At the onset of preformulations studies, it is difficult to predict with
`certainty which of the two types of fonnulations will have a marketable
`advantage for an extended shelf life. Al this early stage of developmenr,
`there are usually very small amounts of the bulk active drug substance avail(cid:173)
`able. The formulator must make very efficient use of the active drug sub(cid:173)
`stance. Neverlheless, both formulations should be considered and started
`at the same time, Stability results should be the deciding factor as to which
`form will be selected for further development.
`
`C. Unit Dose or Multidose
`
`The decision to select unit dose versus multidose should be based upon
`input from clinical investigators, focus groups, marketing surveys, and com(cid:173)
`petitors' products. A multidose formulation will require significantly more
`time for development.
`The multidose will require the screening and incorporation of com(cid:173)
`patible preservativc(s) with the protein formulation. This formulation will
`be tested to determine if it is efficacious enough to meet the United States
`Pharmacopeia (USP) requirements. Meeting these requirements, it can
`qualify as a "multidose" for the U.S. market. However, if the formulation
`is also designated for international marlc:et, there are three additional fac(cid:173)
`tors that must be taken into account. Thefi'3t is that for the "antimicrobial
`effectiveness test, .. a particular country may or may not accept the preserv(cid:173)
`ative selected. Secondly. the concentrations of the preservative present in
`the formulation may be different fro111 the USP requiremefltS. Thirdly, the
`time periods required for the inhibition of the bacteria and fungi strains
`tested may also differ. Consequently, I strongly suggest that the interna(cid:173)
`tional regulatory requirements for compliance should be well researched
`and understood by the scientific aod management staff. Other excipients
`should also be thoroughly reviewed for international acceptance.
`
`D. Physlcochemlcal Factors to Be Considered for
`Protein Drug formulations
`
`Some of the most important physioochemical properties of protein drugs
`required for the development of parenteral preformulations and formula(cid:173)
`tions are found in Table I.
`
`FRESENIUS EXHIBIT 1026
`Page 6 of 54
`
`

`

`Preformulatlon Development
`
`95
`
`T A.BLE 1 Physicochemlcal Factors 10 Be Considered ror Protein Drug
`Formulations
`
`Structure of the protein drug
`lsoelectric point
`Molecular weight
`
`Amino acid composition
`
`Disulfide bonds
`
`Spectra! properties
`Agents affecting solubility:
`Detergent
`Salts
`Metal ions
`pH
`
`Agents affecting stability
`pH
`Temperature
`Light
`Oxygen
`Metal ions
`Freeze-thaw
`Mechanical stress
`
`Polymorphism
`Stereoisomers
`FIitration media compatibility
`Shear
`Surface denaturation
`
`Since this may be an early stage of process development, some of the
`properties listed in Table 1 may not be available initially, simply because
`there was not enough time or personnel to perform the work.
`
`II. INITIAL PREFORMULATION STUDIES: PARAMETERS
`AND VARIABLES TO BE TESTED
`The pharmaceutical formulation scientist will consider several factors in the
`preformulation designs. The data received from the Process/Purification
`section are reviewed for structure, pH and purity of the substance, prelimi(cid:173)
`nary bioassay, and an immune assay used in terms of semiquantitative
`measuremems,
`Other imponant infonnation 1ha1 may or may not be available are
`product solubility, preliminary stability, potential degradation routes. From
`personal experience, there is only minimal crude bulk active substance at
`this early .stage.
`
`A. Initial Variables to Be Tested
`
`Perhaps 10 or more initial preformulation combinations should be consid(cid:173)
`ered. The initial variables to be tested with various protein concentrations
`are the effects of buffer species. ionic strength, pH range, temperature,
`
`FRESENIUS EXHIBIT 1026
`Page 7 of 54
`
`

`

`96
`
`somempo
`
`initial shear, surface denaturation, agitation, and aggregation. Since it has
`been well documented that protein solutions are unstable, some selective
`excipients from various classes of stabilizers should also be included in or(cid:173)
`der to evaluate stability requirements. Stabilizers will be discussed later in
`this volume.
`
`B. Prellmlnary Analytical Development
`
`In order to determine the initial stability re.suits, it is necessary to have
`developed, or to have under development, analytical methods 10 measure
`the potency of the specific formulation under various experimental condi(cid:173)
`tions. Ultimately some of these analytical methods will be needed lo moni(cid:173)
`tor stability to detect physical and chemical degradation, Regulatory com(cid:173)
`pliance for the beginning of Phase I Clinical Studies may requite at least
`two different methods that are "stability indicators," rnost often fully vali(cid:173)
`dated. Dr. Sharma. in Chapter 6, will cover the bioanalytical development.
`
`TABLE 2 Bioanalytlcal Methods to Evaluate Initial Preformulalion Development
`
`Method
`
`Bioassay
`
`Immunoassay
`
`pH
`SOS-PAGE (Reduced &
`non reduced)
`RP-HPLC
`
`IEF
`
`SE-HPLC
`
`N-terminal sequencing
`UV
`
`CD (circulardichroism)
`In tile UV region
`
`Function
`Measure of activity throughout shelr fife ol a
`formulation
`Purity assessment and measures concentra(cid:173)
`tion of a particular molecular species
`Chemical stability
`Separation by molecular weight, characterlza(cid:173)
`llon of proteins and purity
`Estimation of purity, identity, and stability of
`proteins. Separation and anaJysiS of pro(cid:173)
`tein dige.sts .
`Oetermtnes the isoeleciric point of lhe protein
`and detects modfficatlons of the protein
`Method of separating molecules according
`to their molecular size and purity
`determination
`Elucidation of the C-terminus, identity
`Detection of individual component, concentra(cid:173)
`tion, a,nd aggregation
`Detects secondary and tertiary conformation
`and quantltates various structures
`
`FRESENIUS EXHIBIT 1026
`Page 8 of 54
`
`

`

`Preformulalion Developm11nl
`
`!J7
`
`However, some oft he following bioanalytical method!i listed in Table 2 can
`be applied to begin initial evaluation of the preformulations degradation
`(if any) under test.
`
`C. Experimental Conditions for the Initial
`Preformulatlon Studies
`
`Protein Concentration
`
`Protein drugs are extremely potent: therefore, very low concentrations are
`required for their respective therapeutic levels. Dosage fonns development
`need to be tested at varying ranges of activity. The respective concentra(cid:173)
`tions may range from nanograms to micrograms to milligrams and the con-
`centration will vary from protein to protein.
`·
`
`pH Range
`
`Initially, a range of pHs should be selected, for example, 3, 5, 7 and 9. Spe(cid:173)
`cific pH units will be determined during the formulation studies. The pH
`changes may have varying impacts on the solubility and stability of the for(cid:173)
`mulation. pH control in pharmaceutical dosage forms hs very critical (1).
`The proper pH selection is one of the key factors in developing a stable
`product.
`
`Buffers
`
`The buffer(s) selection should be made from the USP physiological burfers
`list and should be selected based upon their optimal pH range. Some of
`these buffers are acetate pH 3.8-5.8, succinate pH 3.2-o.6, citrate ptt 2.1-
`6.2, phosphate pH 6,2-8.2, and triethanolamine pH 7.0-9.0. These pH
`ranges will differ from protein to protein,
`Buffer concentrations should be in 1he range of O.ot to 0, 1 molar con(cid:173)
`centration. As buffer concentration goes up, so does the pain upon injec•
`lion. In selecting the proper buffer, phosphate should be the last in one's
`choice. Phoi;phate buffer reacts with calcium from the glass vial and zinc
`from the rubber stopper to cause glass laminates ;i.nd eventually haziness
`of the solution during stability periods.
`
`Other Excipients to Be Considered
`
`As it was stated previously, the objective of a preformulation study is to
`select potentiallyoompatibleexcipients in order to hasten the development
`of stable formulations. Based upon protein chemical and physical instabil-
`
`FRESENIUS EXHIBIT 1026
`Page 9 of 54
`
`

`

`98
`
`Bontempo
`
`ity, it is highly probable that some excipients may be included in the prcfor(cid:173)
`mulation. In so doing, the designs of fonnulations to follow can be more
`specific in selecting the proper excipient(s) to control specific degradation
`pathways.
`Chelating Agents
`The crude bulk protein drug during fennentation and purification steps has
`passed through and contacted surfaces such as metal, plastic, and gla~. If
`metal ions are present in the liquid bulk active. it is highly recommended
`to use a chelating agent such as ethyrenediamine retraacidic acid (EDTA)
`to effectively bind trace metals such as copper, iron, calcium, manganese
`and others. A recommended dose of (EOTA) would be about 0.01 to
`0.05%.
`Antioxidants
`Since oxidation -is one ofthe major factors in protein degradation, it is highly
`recommended, should the use of a specific antioxidant be required, to in(cid:173)
`clude into the preformulation an antiol(idant such as ascorbic acid, sodium
`disulfide, monothio-g)ycerol, or alpha tocopherol. The role of an antioxi(cid:173)
`dant is to deplete or block a specific chain reaction. Antioxidants will be the
`preferential target and eventually be depleted, or may block a specific chain
`reac1ion. Argon and/or nitrogen gas can also be used to flood lhe head space
`of a vial or ampule during sterile filling to prevent or retard oxidation. A
`recommended antio~idant dose would be about 0.05 to 0.1 % .
`Preservatives
`If a multidose fonnulation is required, an antimicrobial agent, called pre(cid:173)
`servative, is required to be incorporated into the formulation. The preserv(cid:173)
`ative effectiveness must comply with the USP requirements to be qualified
`as multidose. The most often used pre.,r;ervatives and respective concentra(cid:173)
`tions are pheno1 (0.3 to 0.5% ), chloro'outanol (0.3 to 0.5%) and bcnzyl al(cid:173)
`cohol (1.0 to 3.0%). Additional details are provided in Chapter S.
`Surfactants
`Judicious selection of surfactants can result in the prevention of aggrega(cid:173)
`tion and stabilization of proteins (2). Polysorbate 80, poloxamer 188, and
`pluronic 68 have been used in injectable fonnulation. The purity of the
`surfactant may have an impact on the chemical stability of the prefonnula(cid:173)
`tion. Peroxide residues in the surfactant have been implicated in oxidations
`of protein.
`
`FRESENIUS EXHIBIT 1026
`Page 10 of 54
`
`

`

`Preformu/atlon Development
`
`99
`
`Glass Vial Selection
`
`Type J glase,, as classified in the USP, should be used. The solection of 11
`glass vial must also be taken into consideration when dealing with adsorp(cid:173)
`tive properties of the respective protein. Adsorption of proteins will be
`treated later in this volµme.
`
`Rubber Stopper Selection
`hi studymg both tile liquid and reconstituted protein drugs, the selection of
`a rubber stopper is also of major concem considering the potential reactiv(cid:173)
`ity of a protein solution with a rubber stopper, as well as the reactivity of
`the reconstituted tyophilizcd solution during storage conditions prior to
`use. For parcnteraf fom1Ulations, the biopharmaceutical ,industry has been
`using rubber stoppers with a very thin film of various inert polymers in order
`to a~hicvc greater compatibility, fleiubility, low levels of particulates, and
`machinability. In addition, adsorption, absorption, and penneation through
`the stopper are essentially eliminated. Extcnsi~e details may be found in
`Chapter 8.
`
`Membrane Filter Selection
`
`Membrane filtration is the most often used technique to sterilize protein
`solutions. The chemical nature of the fitter and the pH of the protein soJu•
`tion are the two most imponant factors affecting the protein adsorption (3).
`However, there are other issues that require consideration, The fonnula(cid:173)
`tion scientist must be aware of particles or fibef"S release(! during the filtra·
`lion, the poiential extractables that may occur, the potential tox.icicy of the
`tilter media and the product compatibility wlth the membrane. Of all the
`filters tested (unpublished data) polyvinylidcne difluoride, polycarbonate,
`polysulfone, and regenerated cellulose were found to be the most compat(cid:173)
`ible with various proteins and with minimal amounts of protein binding and
`deactivation.
`
`Ill. MECHANICAL AND PHYSICAL STRESSES
`A. Shaking Effect on Protein Solution at the
`Pre-formulation Level
`
`Some of the various physical modes ofvialed protein solution can undergo
`begin with the bulk active fonnulation, filling of formulated solution into
`vials or ampules, visual inspection, labeling, packaging, shipping, and re•
`ceiving. Simulation of some of the functions described above need to be
`
`FRESENIUS EXHIBIT 1026
`Page 11 of 54
`
`

`

`100
`
`Bontempo
`
`performed by doing some shaking experiments to determine their affect on
`aggregation induction.
`Some of these preformulation experiments should also contain vary(cid:173)
`ing concentrations of surfactant(s) with appropriate controls. These short
`and inexpensive experiments can beset up on reciprocal shakers for periods
`of time from 1 to 6 to 24 hours, shaking from 10, 30, and 60 reciprocal
`strokes per minute. Reciprocal strokes disrupt and break up the flow of the
`liquid, while rotary strokes move the liquid circularly without breakup.
`These studies are intended to determine precipitation and aggregation ef(cid:173)
`fects. Detailed aggregation experiments: and results will be described later
`in these chapters.
`
`B. Freeze-Thaw Experiments
`
`These experiments will also be described in later chapters and will be part
`of Chapter 5. These experiments require a fair amount of active drug sub(cid:173)
`stance as well as a fair amount of work. At this point of development there
`may not be enough active drug substance available.
`
`C. FIiiing Systems
`
`Of all the filling types employed to dispc;nsc:: liquid, such as time-pressure,
`piston, and rotary pump, the rolling diaphragm mc::tcring pump is the:: one
`of choice for filling biopharmaceutical solutions. The internal paru of the::
`pump do not come in contact with one another where the liquid solution
`flows. This is the "TL Systems Rolling Diaphragm Liquid Metering Pump"
`(4). One of the most imponant features of this pump is that it eliminates
`the principal cause of particulate generation which is most often induced
`by parts coming together creating shedding of microscopic particles.
`There are ttiree other important parameters to control while dispens(cid:173)
`ing protein solutions. (l) The speed at which liquid is filled into the vials.
`With protein solutions the maximum speed is between 25 to 30 vials per
`minute, delivering 0.5 to 2.0 mL volume per 5- or 10-mL vial per single
`filling head. If a large number of vials need to be filled, this filling system
`can accommodate variable numbers of filling heads, thus allowing it to fill
`a large number of vials. Filling at a faster rate will result in protein precipi•
`tation and aggregation. (2) The inner diameter of the filling cannula should
`not be so very small as to induce shearing and aggregation of the protein
`solution. (3) The tip of the cannula for the tilling head should be bent at
`such an angle as to deliver the fluid against the inner wall of the vial and
`not perpendicular to the bottom of the vial, TI1is will result in a gentle flow
`
`FRESENIUS EXHIBIT 1026
`Page 12 of 54
`
`

`

`Preformufatlon Development
`
`101
`
`touching the inner wall of the vial when the cannula enters the vial and
`delivers the required amount of fluid. The proper bend on the tip of the
`cannula may also eliminate aggregation and/or shearing or the protein
`solution.
`
`D. Stability Evaluation
`
`The development of acceptable analytical methods while isolation, charac(cid:173)
`terization, and purification of a bulk active drug substance are going on is
`very important. h can be an aid in generating semiquantitative and quanti(cid:173)
`tative measurements of the active bulk drug at various stages of the process.
`Significant marketing advantages in this competitive pharmaceµtical
`market would be to achieve a longer shelf .Jjfe of the product and storage
`temperature at room temperature. Today the lyophilized protein drug of(cid:173)
`fers refrigerated temperature storage between 2 and 8°C.
`The present storage conditions set up by the USP on storage require-
`ments are as follows:
`• Cold storage. Any temperature between 2 and 8°C
`• Cool. Any temperature between 8 and l5°C
`• Room temperature. Temperature prevailing in a working area
`• Controlled room temperature. Temperature controlled thermo(cid:173)
`statically between 15 and 30"C
`• Excessive heat. Temperature exceeding 40°C
`Table 3 summarizes the initial guideline time points and tempera(cid:173)
`tures that preformulation solutions ~hould be exposed to. The results from
`the preformulations will allow the review team to determine directions to
`manipulate the excipienls to obtain better stability.
`
`TABLE 3 Guideline for Preformulation Stability Studies
`Temperatures
`Frozen controls (--80 and -20•C)
`Refrigerated (2--8°C)
`
`Tlrnepolnt
`
`Intermediate (20, 30, 37"C)
`
`High temperature (40, 45, 60°C)
`
`Reference control sample as needed
`T = O, 6, 12, 24& 48 weeks
`Continue if stable
`T-= O, 4, 8, 12, 18, 24 weeks
`Continue if stable
`T -= o, 1, 2, 4, 8, 12 weeks
`Continue if stable
`
`FRESENIUS EXHIBIT 1026
`Page 13 of 54
`
`

`

`102
`
`Bontempo
`
`IV. DEGRADATION MECHANISMS
`To predict degradation pathways of new biophannaceuticals is ve,y diffi(cid:173)
`cult Depending on the stress conditions, each protein may react differently
`than another protein. As stated previously, the objcctivcsofpreformulation
`are to evaluate ~tress conditions such as pH, temperatures, and buffers and
`b1,:gin evaluation of some ini1ial breakdown producls. At this particuhu
`stage of development, it is necessary to have some analytical method(s) with
`some reliability to detect initial degradation. It is difficult to begin evalu(cid:173)
`ation of degradation products without the reliability of these prclimimuy
`assay methods.
`The pulJ>Ose of initial preformul111ion studies is to begin understand(cid:173)
`ing of protein instability via chemical and physical stress conditions (5). In
`order to stabifo:e potential useful pharmaceutical products, it is important
`to undeBtand how proteins degrade, how they are affected by the compo(cid:173)
`sition of the fonnulation, and the effects of stability conditions. The major
`pathways of protein degradation are chemical and physical. Under chcmi•
`cal degradation, changes and modifications occur due 10 bond fonnation
`or cleavage, yielding new chemical entities. One or more of the following
`can occur: oxidation, deamidalion, hydrolysis. racemization, isomeriuuion,
`beta elimination, and disulfide eltchange. Physical instability can occur in
`the fonn of denaturation, aggregation, precipitation, and ads(lrption with(cid:173)
`out covalent changes.
`
`A. 01tldatlon
`
`Oxidation of protein is perhaps one of the most common degtadation
`mecha.nisms that can take place during various stages of the processing,
`such as fermentation, purification, filling, packaging, and storage of the
`biophannaceuticals. Under oxidative stress and in the presence of trace
`metals, amino acids such as methionine (Mei) can be oxidized Jo methion(cid:173)
`ine sulfoxide, cysteine (Cys) to cysleine disulfide, as well as tryptophane
`(Try) and histidine (His) via other modifications.
`Oxidation can be controlled or minimized by (1) the addition of an(cid:173)
`tioxidants, (2) having strict controls on the processing operations, (3) using
`nitrogen gas to flood head space of the container.
`Oxidized human growth hormone (hGH) retains only 25 percent the
`activity of the native molecule, recombinant interferon-beta loses consid(cid:173)
`erable antiviral activity due to oxidation (5). Oxidation can be detected by
`reversed phase HPLC (RP-HPLC). high-perfonnance isoelectric chroma(cid:173)
`tography (HP-IEq, ·peptide mapping, amino acids analysis, and mass spec-
`
`FRESENIUS EXHIBIT 1026
`Page 14 of 54
`
`

`

`Preformulallon Development
`
`103
`
`trometry (MS) (6). In terms of total p,rotein concentrdtion, ultraviolet spec(cid:173)
`trophotometry is the method most often used (7).
`
`B. Deamldation
`
`Deamidation is another more frequent degradation mechanism affecting
`pharmaceutical protein stability. Deamidatioa is the hydrolysis of the side
`chains amide on asparagine (Asp) and glutamine (Gin) to fonn Asp and/or
`Gln residues. Extensive reports have elucidated mechanisms of deamida(cid:173)
`tion reactions (8).
`Deamidation can be detected by isoelectric focusing, ion exchange
`chromatography, tryptic mapping and HPLC (9).
`
`c . Hydrolysis
`
`Hydrolysis is another most likely cause of degradation of prote ins. It in•
`volves a peptide (amide) bond in the protein backbone (5). The most influ(cid:173)
`ential factor affecting the hydrolytic rate is the solution pH.
`
`0 . Racemlzatlon
`
`Proteins may also degrade via other modifications (10) such as racemiza•
`tion. This mechanism invoives the removal of the alpha proton from an
`amino acid in a peptide to yield a negatively charged planar carbanion. The
`proton can then be replaced into this optically in,1ctive intermediate, thus
`producing a mix1ure of D and L enantiomers (2). Racemization can yield
`enantiomers in both acidic and alkaline conditions.
`
`E. lsomerlzatlon
`
`Protein degradation is also induced by isomerization. Hydrol}'Sis of cyclic
`amides of asparagine, glutamine, and aspartic acid will result in isomeriza(cid:173)
`tion. Low pH accelerates hydrolysis of asparagine and glutarninc. However,
`high pH ac.celerates hydrolysis of aspartic acid and glutamicacid (2,11 ,12).
`
`F. Disulfide Exchange
`
`Disulfide exchange may result from a degradation other than covalent
`modification. These reactions may include the disulfide exchange of cys(cid:173)
`teine. This reaction is base, catalyzed and promoted by thiol antioxidants
`(13).
`
`FRESENIUS EXHIBIT 1026
`Page 15 of 54
`
`

`

`....
`
`104
`
`Bontempo
`
`Disulfide exchange can occur in misfolded conformers due to incor(cid:173)
`rect intramolecular disulfide bonds (14 ).
`
`G. Beta-Elimination
`Another degradation residue can be the beta-elimination of ser, thr, cys, lys
`and phe residues. These reactions arc accelerated by basic pH, tempera(cid:173)
`ture, and the presence of metal ions (16).
`
`V. PHYSICAL DEGRADATIONS
`Aggregation
`Protein aggregation can be of a covalent or noncovalent nature (17,18).
`
`A. Covalent Aggregation
`This pathway involves modification of the chemical structures resulting in
`new chemical structures and may include reactions, such as oxidation, de(cid:173)
`amidation, proteolysis, disulfide interchanges, racemization, and others.
`
`B. Noncovalent
`This instability may be induced by agitation, shear, precipitation. and ad(cid:173)
`sorption to surfaces.
`
`C. Aggregation
`Protein aggregation derived rrom either physical or chemical inactivation,
`is presently a major biophannaccutical problem ( 17- 21 ). Aggregation can
`be either covalent or noncovalent, occurring during any phase of product
`development from purification to formulation. An early detection of aggre(cid:173)
`gation via biochemical or spectrophotometric methods, or both, can be of
`significant guidance to fonnulation scientists in selecting compatible excipi•
`ems to minimize and/or prevent its formation in the e,cpcrimental fonnu(cid:173)
`lation.
`Formation of aggregation can begin by the fonnation of initial parti(cid:173)
`cles from protein molecules via the Brownian movement. This is followed
`by collision of these molecules and aggregates of varying sizes can be formed.
`These aggregates can be generated by shear or collisional forces (22).
`Dct~tion and measurements of aggregations can be performed by a
`number of techniques. Visual observations, light scattering, polyacrylamide
`gel electrophoresis, UV, spectrophotometry, laser light diffraction panicu-
`
`FRESENIUS EXHIBIT 1026
`Page 16 of 54
`
`

`

`Preformulatlon Development
`
`105
`
`late analysis, fluorescence spectra and differential scanning colorimetry
`(DSC); RP-HPLC, and SE-HPLC (7,14). Conformational changes can also
`lead to aggrega1ions and can be measured by DSC (23),
`A formulation scientist should focus on some important obseivations
`that need to be made to answer some potential problems on aggregation.
`• Determination of initial approximate number of aggregates
`• Determination of approximate size and distribution of aggregates
`• Do the aggregates increase in size and number over time?
`• Do the aggregates affect the efficacy of the proteins?
`• What is the effect of aggregation on the long-term storage of the
`potential marketable product?
`
`0. Oenaturatlon
`Dena tu ration of proteins can be the result of 5everal processes and reported
`by several investigators (27).
`Factors which induce denaturation are heat or cold, extreme pHs,
`organic solvents, hydrophilic surfaces, shear, agitation, milling, filtering, shak(cid:173)
`ing, freeze-thaw cycles, ionic strength, and others. Thermal inactivation
`processes will induce conformational side reactions and destruction of amino
`acids (28). The loss of biological function may well be attributed to the
`effect of the temperature on the higher-ordered structure of the protein.
`Thermal denaturation of proteins is of great interest to the fonnulation
`scientist. Thermal probes offer tools to study protein structure and stability
`chat ultimately can be or significant use to· stabilize protein drug fonnula(cid:173)
`tions. Modifications of protein thermal effects have been reviewed (8).
`The ability of the protein to refold from a denatured state, a reven;ible
`heat denatu!"lltion, is also of considerable interest for the stability of a pro(cid:173)
`tein formulation. These processes of renaturation are veiy complex (29),
`and each protein does have Its own unique renaturation mechanisms.
`Since filtrations and volume reductions occur from the fermentation
`to process purification, there is very likely inactivation of the protein attrib(cid:173)
`utable to shearing effect.
`
`E. Precipitation
`Precipita1ion in formulations ca.n occur by a variety of mechanisms such as
`shaking, heating, filtration, pH. and chemical interactions. Aggregation is
`the initial ons