Risk Management Strategies for
`Safety Qualification of Extractable
`and Leachable Substances in
`Therapeutic Biologic Protein Products
`
`Ingrid Markovic, Ph.D.
`U.S. Food and Drug Administration
`Center for prug Evaluation and Research
`/
`
`Comments expressed in this review represent the opinions of the author
`and are not intended to endorse any product or to reflect FDA policy.
`Abstract
`
`E xtractables and leachables (E&L) are chemical entities, which
`
`can be released into intermediate mat erial or final therapeutic
`biologic protein product at various times during upstream and/
`or downstream manufacturing steps, packaging operations and/or
`storage. These substances may pose a safety risk to the patient by
`causing toxicity, carcinogenicity, fmmunogenicity and/or endocrine
`dysregulatfon. They may also a lter product physico•chemical
`properties via direct interaction with the active pharmaceutical
`Ingredient or, indirectly, by interacting with the excipients In
`product vehicle, thereby adversely affecting the product quality.
`Current paper wlll address a risk-based approach to concept ualizing,
`evaluating and executing identification and chara<terization of E&L
`along with regulatory considerations regarding the impact of these
`impurities on product quality, patient safety and clinical efficacy.
`Selected case studies are presented and discussed.
`Introduction
`Extractables (E) are defined as chemical entities that can
`be extracted from components of a material by exertion of an
`exaggerated force (e.g., organic solvent, extreme temperature,
`ionic strength, pH, contact time, etc.). Leachables (L) are defined
`as chemical entities that can migrate from product-contact and/
`or non-product contact surfaces into a p rocess stream, bulk
`drug substance, product intermediate and/or fina l drug product
`under specified conditions of production, storage and use. While
`leachables are considered a subset ot extractab!es, they can also
`be derived by chemical modification of the original extractable
`component. In addition, not all substances identified as extractables
`
`96 APR MAY !JUNE 2009
`
`Novartis Exhibit 2178.001
`Regeneron v. Novartis, IPR2021-00816
`
`

`

`will be detected as leachables given that widely opposing extraction
`conditions are utilized for isolation of each set of compounds.
`~valuation of extractables is usually considered an essential step
`in the accurate prediction of leachables as well as in selection of
`adequate in-process equipment and/or final container/closure system
`employed in ~roduction of a given biologic product (Figure 1). In
`general, leaching <an occur at any of the multiple steps comprising a
`manufacturing process. Such Include but are not limited to upstream
`operations (e.g., media preparation, fermentation); downstream
`operations (e.g., concentration/buffer exchange, purification);
`formulation/fill; packaging operations and long-term storage of the
`product throughout its expiry period.
`Evaluating Extractables and Leachables
`Potential sources of E&L include materials used in the
`manuf_acturing, packaging, storage, filtration and transfer systems.
`These include but are not limited to components which are in d irect
`contact with the process fluid or product such as single use/stainless
`steel bloreactors, bags for intermediate and long-term storage,
`contain~rs, filters, transfer tubing, e lastomeric closures, ampoules,
`vials, syringes, bottles, etc. In addition to materials ln direct contact
`with the product, secondary packaging components, which are non(cid:173)
`product contact (e.g., cardboard containers, overwraps, overseals,
`~ontainer labels) can also be the source of leaching. For example,
`ink, epoxy adhesives and organic solvents originating from container
`labels have been detected in products packaged in prefilled syringes.
`Rega~ding their chemical nature, E&L are diverse compounds which
`~ay include but are not ilmited to phthalates (i.e., plasticizers),
`ni~rosamines, vulcanizing agents, accelerators, silicone, organic
`acids, hydrocarbons, cyclic esters derived from urethane adhesives
`anti-oxidants, residual solvents, antistatic agents, deaning agents'.
`residual metals, inorganic compounds (e.g., metal oxides), etc.
`Quality Risk Management Approach
`The presence of leachables during any stage of the production
`process or storage may pose a safety risk due to their potential to
`cause toxicity, carcinogenicity, immunogenicity and/or endocrine
`dysregulation [1, 2, 3, 4, 13). In addition, these substances may
`adversely_ impact the physico-chemical characteristics (e.g., via
`aggregation, oxidation, degradation, formation of particulates, etc.)
`of the final protein product (5). Furthermore, t here are concerns
`that leachables may pose a risk to cell viability during storage of
`live cells or during cell culture fermentation (e.g., in single use
`bion~actors) likely negatively affecting product yie ld and product
`quality characteristics. It is of note that biologic therapeutics may
`be espedally susceptible to impact of chemical leachables due to
`their (1) large size (e.g., in the KDa range) and complex structure
`(e.g., secondary, tertiary, quaternary); (2) extensive surface area
`and high frequency of potential sites of interaction; (3) route of
`~dministrat ion (i.e., most are sterile injectables) and dosing/volume
`(1.e., may be dosed at mg/ml and at re latively high volumes); and (4)
`because proteins may be more efficient in solubiiizing feachables
`compared to small molecules due to abundance of both hydrophilic
`and hydrophobic sites.
`Understanding of the system suitability criteria that are
`capable of defining and controlling E&L as critical quality attributes
`built_ Into the design space is of paramount Importance to ensure
`continuous production of high quality therapeutic biologic products
`with desired efficacy and minimal safety adverse events. For these
`reasons, it is recommended that drug product manufacturers
`perf<;>rm a .risk-based analysis as part of E&L evaluation taking into
`cons1derat1on product quality parameters as they relate to product
`safety and efficacy. The following factors may provide predictive
`
`parameters for identifying, evaluating and mitigating risks to critical
`quality and safety attributes (note that the factors are not ranked in
`the order of importance):
`
`• Toxic potential of studied E&L Including synergistic and/or
`additive acute affects as well as chronic toxicity
`• Drug d.ose, mode and frequency of administration (e.g.,
`many biologic therapies are presented as sterile injectables
`likely administered frequently at relatively high volumes
`and doses whereby higher (i.e., unacceptable) levels of
`leachable impurities may be delivered)
`• Prior clinical exposure to a particular leachable
`• Level of risk for adverse impact on product quality
`(e.g., may need to be assessed on a case-by-case basis as .
`biologic products and respective formulations likely have
`different susceptibilities to changes in the product due to
`interaction with leachables)
`• Surface area of tonfact and d uration of contact
`between material component and process fluid, product
`intermediate orfinal drug product
`• Process fluid storage temperature (i.e., leaching will be
`exacerb?ited at elevated temperatures; e.g., 37°C vs. - 196°C)
`• Type of t he processed/stored material (e.g., purification
`buffer vs. formu lated Drug Product)
`·· ·
`• Posi~lon in the process stream (e.g., upstream vs.
`downstream operations; typically the risks are greater
`as production moves closer to the finished product as
`oppo rtunity to clear potential impurities is diminished)
`• Type of construction material in use (e.g., PVC containers,
`bags or tubing are at high risk for leachlng dl(2·
`ethyihexyf)phthaiate, which has been shown to exert
`~arlous types of toxicities t o liver, testis, mammary, nerve,
`immune system, blood and fat tissue)
`• Formulation type where by a number of factors may be
`used to predict the risks for leaching. For example, liquid
`formulations are in continuous drug product-contact
`with elastomeric closure and/or container material
`compared to lyophllized ones and therefore at higher risk;
`formu lation excipients due to inte raction with leachables
`can jeopardize product quality {note case study #5]; pH of
`the formulation buffer may be important where alkaline
`solutions are thought to exacerbate leaching, etc.).
`• Therapeutic necessity of the d rug where higher levels
`may be tolerated if a given drug Is considered part of
`essential therapy (1 I
`Analytical Characterization of E&L
`Detection, identification, characterization and quantitation of
`leachables could be challenging since these substances represent
`diverse chemical classes of organic and/or inorganic compounds
`that co-exist in complex mixtures at trace amounts [6). Well
`designed extractables studies are important to provide an analytical
`roadma p and to identify early warnings signs regarding leachables
`released during up-stream and/or downstream operations and/or in
`storage (Figure 1). In situations where extractables are anticipated
`to adversely impact the physico-chemical/bioioglcal properties of a
`protein, characterization studies spiking the extractables into the
`product may be of value in assessing risks to product quality.
`Product manufacturers may choose to rely on the vendor of
`particular materia l component as a starting point for g eneral
`Information on the extractable substances. Such data are generated
`using model extraction solvents and exaggerated and/or exhaustive
`conditions of vendor's choice. In general, extraction studies are
`initiated by selecting appropriate exaggerated and/or exhaustive
`
`98 APR M4 Y !JUNE 2009
`
`Novartis Exhibit 2178.002
`Regeneron v. Novartis, IPR2021-00816
`
`

`

`t EXTRACTABLES & LEACHABLES .
`
`~
`
`.
`
`·
`
`}aical
`Led in
`f
`
`\d!or
`
`.
`
`Figure 1
`
`FJgul'e. l. Multi-tiered app1·oaclt to analyzing E&L
`
`Jncilrporale E&L as crilical quality atiribules when defiili111i
`a design space
`
`Initiate evaluation of
`extractables
`
`Identify, characterize
`and quantitate
`individual extractables
`
`... ---►
`
`Perform supportive toxicology studies
`
`cardnogenic and/or toxic potential.
`Correlate to toxic tl,reshold levels, if
`available
`
`r. ., ___ an_d_i_d-e1-1t_if_y_e_xtr-ac-t-ab_l_e_s,_~,_·t_h_-1
`
`Identified extractable.~ are
`known caninogens and/or
`toxins
`
`l
`
`No toxicity detected
`
`Proceed with ~he
`·1each~bles study
`
`Identify, characterbe and
`· quantify indiv(dual
`leacbables in stbrage
`
`I
`I
`I
`I
`I
`J
`
`----------
`,---------.
`
`t
`
`I
`Establish the limit of
`I quantitation (LOQ) end I
`I limit of detection (LOD) I
`..... _,,,. ........... --.1
`of the instrumentation
`
`,--------- ....
`
`1
`I
`
`I Assess the foUowiug:
`•Volatile rn1npolll)ds
`I •Semi-volatile
`I •Non-volatile
`I •Residuai metals
`I •Inorganic compounds
`I epH
`I •TOC
`I • Conductivity
`
`Process fluid representative of the
`actual process (e.g., cell culture media,
`Drug Product
`formulation buffer,
`etc.) should be used as a representative
`extraction solution in addition to others
`listed above for accurate prediction
`of extractables. In the event that
`representative process fluid
`is not
`applicable (e.g., due to interference
`with the analytical method), a well
`justified surrogate solution closely
`resembling the original is considered an
`acceptable alternative. Extraction study
`may be performed using a soaking or
`a recycling method for a given contact
`time (e.g., 180 days), under accelerated
`temperature conditions (e.g., 40°C) and
`preselected
`surface
`area-to-volume
`ratio, while exhaustive conditions may
`Involve more stringent temperatures
`(e.g., autoclaving a materlal component
`for a given time period, [7)). It is of
`note that small volume components
`(e.g., bags, containers, etc.) have a
`greater
`solution-contacting
`surface
`compared to large volume components.
`In general, they can be expected to
`generate higher level of extractables
`and therefore represent the worst-case
`scenario with regards to surface area-to(cid:173)
`volume ratio. Consideration should also
`be given to selecting the appropriate
`sample size (i.e., number of bags, filters,
`containers, elastomeric closures, etc.),
`which is greater than one in order to
`achieve accurate representation of the
`sample population.
`Testing of extractable and leach able
`substances typically includes the analysis
`of non-volatile, semi-volatile and highly
`volatile organic compounds as well
`as ana lysis of trace light and heavy
`metals. In addition, some less specific
`tests such as pH, conductivity and
`analysis of total organic carbon (TOC)
`may be performed. It is of note that
`TOC analysis can not be used on extract
`solutions that contain carbon. Highly
`selective analytical techniques should be
`employed for detection, characterization
`and quantification of these chemicals.
`Such include but are not limited to High
`Performance
`liquid Chromatography
`coupled with Mass Spectroscopy (HPLC(cid:173)
`MS), Gas Chromatography coupled with Mass Spectroscopy (GC-MS),
`Inductively Coupled Plasma with Mass Spectroscopy (KP-MS}, Proton
`Nuclear Magnetic Resonance (1H·NMR), Fourier Transform Infrared
`(FTIR) Spectroscopy and Atomic Spectroscopy (e.g., atomic absorption,
`atomic emission spectroscopy). For examp!e, non-volatile compounds
`can be analyzed with HPLC-MS; whereas, highly volatile and semi(cid:173)
`volatile organic compounds can be resolved using GC-MS. Alternative
`methods, such as ICP-MS, can be used for detection of residual metals.
`Emerging analytical technologies with appropriate sensitivity and
`specificity should be considered in addition to the currently avallable
`methods in design of E&L characterization studies,
`
`,. _ _ _ _ _ _ _ _
`I Correlate leachables
`I levels to the thresllold
`l
`levels published in
`I
`toxicological
`1
`I __ !a'!b~e:a __ I
`
`- -►
`
`/
`
`No cha11gein
`product quulity
`and no safety
`concerns
`
`Assess tlui impact on product quality
`(e_.g., oxidation, aggregation,
`degradation, etc.) in storage
`
`Cha11 ges in pro duet
`quality identified
`
`Figure 1. MultJ.1iered app,oach to &nalyiin9 E&L
`conditions not stipulated for manufacture, storage and/or use
`in ?rder to isolate chemicals from relevant material components.
`It 1s recommended that the extraction studies employ worst·
`case conditions with regards to pH, ionic strength, contact time,
`temperature, surface area•to-votume ratio and, if applicable, with
`organic solvents of varying polarity (i.e., from highly polar to non(cid:173)
`polar) as extraction media. in cases where material component
`under examination
`is not compatible with organic solvents,
`aqueous media may be used. Use of detergent (I.e., polysorbate
`20)-containing extraction media is also recommended.
`
`APR MAY !JUNE 2009 99
`
`Novartis Exhibit 2178.003
`Regeneron v. Novartis, IPR2021-00816
`
`

`

`Evaluation of extractables should be comp lemented with
`assessment of leac;hables. In addition to monitoring the substances
`that are leaching under recommended conditions of use and storage,
`leachables studies may be designed to Identify interactions and the
`resulting effect of such Interactions on the in-process material and/
`or on the product under accelerated conditions. In many cases,
`Drug Substance and Drug Product stability studies should be used
`to support conclusions regarding the impact of these substances on
`product quality over time. This is particularly important in situations
`where there is no downstream purification step that could eliminate
`the impurities such as in the final formulated Drug Product.
`In addition to determining the chemical identity, quantity and
`composition of E&L and the impact on process fluid/product quality,
`E&L should be assessed for their cytotoxidty (e.g., USP chapter <87>
`(81), acute toxicity in animals (e.g., USP chapter <88> [9)) as well as
`chronic toxicity. The chronic toxicity data may be especially helpful in
`ensuring that safety and clinical efficacy are not adversely affected
`in patient population which is treated for extensive time-periods
`(e.g., lifetime treatment) and thereby subjected to chronic exposure
`of impurities. For acute threshold levels, publically available
`literature sources such as ICH Q3C(R3) [1 0) or 'product Quality
`Research Institute Leachables and Extractables Working Group (11 ),
`which stipulate the safety threshold levels for such impurities may
`prove especially useful.
`
`Case studies
`Case study # 1
`Please note that this case study was previously published in
`(12) with the aim of the current paper to provide an update on
`the corrective actions taken. A therapeutic protein product was
`changed from a iyophllized to a liquid presentation. Due to this
`change, a divalent metal cation migrated from the rubber stopper
`into the Drug Product vehicle. The released metal cation activated a
`metalloprotease (a process-related impurity that co-eluted with the
`active pharmaceutical ingredient) causing N-terminal degradation
`of the product. The problem was uncovered during stability studies
`under inverted conditions and was resolved by adding a chelator
`(i.e., ethylenediaminetetraacetic acid, EDTA) to the Drug Product
`formulation. Unfortunate ly, the new formulation was associated with
`adverse safety outcomes recognized by an increase In cardiovascular
`events as well as changes in the pharmacokinetic properties of
`the drug. This formulation was consequently withdrawn from the
`market and replaced with the original one. The leaching of divalent
`metal cations was mitigated by implement ing a modification in the
`elastomeric closure, which is now coated with Teflon.
`Case study #2
`For another protein product, human serum albumin (HSA) was
`replaced with poiysorbate as a critical excipient while keeping the
`same container/closure system (I.e .. pre-filled syringe). Associated
`wlth this change, bromine from the coated bromobutyl plunger
`stopper and tungsten from the syringe needle were found by
`ICP-MS. Both impurities can be powerful oxidants but the impact
`on product oxidation and aggregation were
`inadequately
`monitored. It is of note that methionine is present as an exciplent
`in the formulation, which may be critical in mitigating the possible
`damage due to bromine and tungsten. In order to identify the risks
`to product quality the sponsor was asked to, evaluate the effects
`of bromine and tungsten on Drug Product quality both individually
`and in combination using robust analytical methods that included
`orthogonal methods for monitoring protein aggregation over the
`shelf life.
`
`Case study #3
`This case study pertains to a change In the material of
`construction of the closure system from latex to chlorobutyl
`rubber stopper for a lyophilized product. As a result, butylated
`hydroxytoluene (BHT), a common antioxidant and food additive,
`leached from the stopper and was uncovered at the 12-month
`stability time point using Reversed Phase High Performance Liquid
`Chromatography (RP-HPLC). The reachable (i.e., BHT) was quantified
`and measured levels proved to be extremely low and moreover
`significantly below the LOSO values established for BHT in animal
`models. This alleviated concerns associated with adverse effect on
`patient safety. In order to assess the impact to product quality in
`storage, the Sponsor performed additional stability-indicat ing
`assays and found no other anomalies in product physico-chemical
`parameters. Furthermore, additional studies evaluating E&L are
`being performed with re~ulttcurrently underway. Finally, in order to
`control and monitor the level of the leach able, it was recommended
`that an acceptance criterion for BHT be established.
`Case study__ #4
`This example involves a change from vials to staked needle
`prefilled syri~ge. Due to this change, organic solvent-from partially
`dried epoxy /glue used for needle attachment to syringe barrel
`leached into the product and caused an increase in protein oxidation
`followed by aggregation via d isulfide switching. The problem was
`resolved by allowing syringe barrels to dry for 6 months prior to use.
`Case study #5
`In this case study there was a change from molded to tubing glass
`vials, which resulted in the leaching of aluminum oxide' produced as
`a by-product of the new glass vial manufacturing process. Due to
`this change, phosphate in the formu lation buffer interacted with
`aluminum forming aluminum phosphate crystals. The problem was
`observed as an out of specification (OOS) result for visible particulates
`of up to 150 µm diameter ln size in samples that were allowed to age
`for more than 12 months with no changes in other physico-chemkal
`parameters. A variety of analytical methods was used to characterize
`the particulates including Scanning Electron Microscopy, Fourier
`Transform Infrared Spectroscopy and X-Ray Diffraction. The 005
`result led to a recall of the lo~ that failed the acceptance criterion.
`The issue with leaching was resolved by coating the glass vials with
`silicone using a baked-on siliconization process.
`Summary
`Biologic protein products can be very sensitive to seemingly
`minor impurities and changes in the container/closure system and/
`or formulatfon composition. Undetected differences in product
`impurity profile may have a significant impact on clinical safety and
`efficacy parameters such has been reported in the case of leachabtes
`acting as adjuvants triggering immune response [13). Presented case
`studies illustrate that corrective actions should employ a simplest
`approach to resolve a problem with a least potential to alter product
`quality as it relates to safety and efficacy.
`
`Acknowledgements
`The author gratefully acknowledges Drs. Barry Cherney and
`Steven Kozlowski from OBP, CDER, FDA for critical review of the
`manuscript. The author also thanks members of the Division of
`Therapeutic Proteins, OBP, CDER, FDA for helpful discussions and
`for providing the case studies.
`
`1.
`
`100 APR MAY !JUNE 2009
`
`Novartis Exhibit 2178.004
`Regeneron v. Novartis, IPR2021-00816
`
`

`

`13. Boven K, Stryker S, Knight J, Thomas A, van Regenmorte/ M,
`Kemeny DM, Power D, Rossert J, Casadeva/1 N. The increased
`Incidence of pure red cell aplasia with an Eprex formulation In
`uncoated rubber stopper syringes. Kidney Int. 2005,
`67:2346-2353.
`
`OSTENBERG RE: Potential toxicity of extractables and
`/eachables in Drug Product. American Pharmaceutical Review
`(2005) 8:64-67
`CAFMEYER NR and WOLFSON BB: The possible leaching of
`diethyl phthalate into levothyroxine sodium tablets. American
`Journal of Hospital Pharmacy (1991) 48:735-739.
`CDRHIFDA guidance: Safety assessment of di(2-ethylhexyl)
`phthalate (DEHP) re/eased from PVC medical devices.
`Available at: http:l/www.fda.govlcdrhlostldehp-pvc.pdf.
`GA YATHRI NS, DHANYA CR, INDU AR et al: Changes fn some
`hormones by low doses of di(2-ethyl hexyl) phthalate (DEHP),
`a commonly used plasticizer in PVC blood storage bags and
`medical tubing. Indian Journal of Medical Research (2004)
`119:139-144.
`MARKOVIC, I. Evaluation of safety and quality impact of
`extractable and /eachable substances in therapeutic biologic
`protein products: A risk-based perspective. Expert Opin Drug
`Saf. (2007) 6:487-491.
`WANG, Q. Selection of analytical techniques for
`pharmaceutical leachab/es studies. Am. Pharm. Rev.
`(2005) 8:38-44.
`Anon. U.S. Pharmacopeia, 28th ed. Rockville, MD: The U.S.
`Pharmacopeial Convention, Inc.; 2004. <381> Elastomeric
`Closures for Injections, Physicochemical Test Procedures.
`Anon. U.S. Pharmacopeia, 28th ed. Rockvifle, M D: The U.S.
`Pharmacopelal Convention, Inc.; 2004. <87> Biological
`Reactivity Tests, In Vitro.
`Anon. U.S. Pharmacopeia, 28th ed. Rockville, MD: The U.S.
`Pharmacopeial Convention, Inc.; 2004. <88> Biologfcal
`Reactivity Tests, In Vivo.
`/CH Q3C(R3). Impurities: Guideline for residual solvents.
`Available at http:l!www.ich.org/LOBlmedialMEDfA423.pdf.
`Product Quality Research Institute Leachables and Extractabfes
`Working Group, Safety thresholds and Best Practi,es for
`Extractab/es and Leachab/es in Orally Inhaled ancJiNasal Drug
`Products, 2006.
`MARKOVIC I: Challenges associated with extractable and/or
`teachable substances in herapeutic biologic erotein products.
`American Pharmaceutical Review (2006) 9:20-27.
`
`I 2009AAPS
`
`National
`Biotechnology
`Conference
`
`\,
`
`.. ---:,''f.'i}l~
`. . ..
`.. '"·~s·~-:.'}:.>
`Vi sit us at the 2009 AAPS
`•;j~\f.\f~:.:~~t
`.
`\'\L.~
`· ·?ff;'?,'S')\:1>~
`National Biotechnology Confere nee
`"iot:fl~ · "'';:i'~,,,, .. , ..... ·
`Booth #320
`f~
`;:1 - - - - - - - - - - - - - - -- - - - - -- - - - - -~-;,~:,,:..,: .. c,"· ·•'.·7.'~·:·C:•"-""~,·· - - - - - - - - !
`~'----- - --- --- --------- -------- --- - --- - --- --- ----
`
`v. ~
`
`'~ ''"
`
`APR MAY !JUNE 2009 101
`
`Novartis Exhibit 2178.005
`Regeneron v. Novartis, IPR2021-00816
`
`