ATTORNEY BUCKET NC}.:
`
`PA ENT
`(}S629I~50E}4-£}1
`
`LN‘ THE UNITE!) STATES PAIFENT s‘&N'£l TRADEMARK UFFICIE
`
`In re?i’ATENT .@%.PPL1CAT§ON of:
`E‘s«-’;'&NS at at
`!{}i8’?2,784
`Applicatian Nu:
`Filed:
`June: 22, 21104
`FOR:
`F€}R§6‘lULAT’I£)N
`
`) Cenfinnaiiun Na.
`
`3393
`
`161?
`{irstmp Art Unit:
`3 Examiner: Hui, San-Ming R
`g
`
`QECLARATIQTR‘ UNDER 35 U.S.£L 1.132
`{I}? PAUIV, Ri.€T}¥ARf} GIi1IA{..ER"I‘
`
`PAUL RICHARI} GELLERT cf xkstrazcnaca, 1-‘kiderley Park? Macciaatiéid, Cheshire, UK
`
`deciares:
`
`i.
`
`igraduatttd frmrt“: the: University Clf Cfixfbrcl in CI'2ex’r:is;try in 1934. E undertmk pastgraduate
`
`research with Prufesgar Brian Haowarci in the Physiczai Chemistry Laburatczry at {he-
`
`University :3f€T)xfi::rd icatiing til} the award «:::fVa E.}.Phi¥ in 1988. Frcml Fairxuary 1983 uniif
`
`the presearat I have been empkayeé by Astrafiezxeca, {ferxnerly Zeznezza and ICE") initiaily as a
`
`Sesnior Rergearcla Sciezxxtist and subsequwtiy as a Team. Lcaaziesrmianagszr, Principal
`
`Scienzist and, since EH04}, 3 Serum‘ Principal. Scientist.
`
`2.
`
`I have warked £3’: {ha fannuiatian and cimg deeziivrary ama thrcsughaui my Gamer wizh
`
`Astraxenieca, whmri: my research and dCf‘€é‘€§1C§§3f2‘i€fI1t wsark has ctwemd 3 range“: wf
`
`formuiatizxn types imtiuding sustained mieazsad injeciimls, includfaig fuivfistrazat.
`
`3. During tha ittfkliféifi tit‘ H132 atudy :11?‘ thus: subject applicatémz {hereinafter “the: Evans:
`
`A;ppiicatirat1""} anti the underiying data, Ehzwe:»13t:c:3me awam czf swveral transaripVtionL :3:
`
`{rather errata Eaetxmazan certain discI.:3sur::*:.~; af the suiaject agxplicatimx ~’e3;l:1f3 the underlying
`
`fabfiratmry iiéitethcficxk céata. (me pumme: c:ft¥*:i:.; Baciaratiaxa is m paint. (‘mt {he ezzszistemse
`
`DE isfiitivilfaflfx. 1.
`
`Astrazeneca Ex. 2135 p. 1
`Mylan Pharms. Inc. V. Astrazeneca AB IPR2016-01324
`
`

`
`and manure ::af1:hesearr»::+rs anvzi :0 repnn further tszzgtizxg Iha: has bazen carried out umier my
`
`g,uixi:1r:s;:e:<;> iamain aciziitécmai data (paragaphs 4- H} beimv and A{t3{‘:§‘1§I‘i£*ntS A-D), A
`
`furthaar purpnse cf this Declaratian is In mt mt: and dmumnzm the manner in which an
`
`£’:3:p£3i‘i£:§3C{:£i frmmniamr ssmtsld Iikciy haw: agzprc:-armerd the task :1-f devaloping 2; snstainnztai
`
`rirlflase izzjcctabie fmmlulaiian suitable flu: human use far a fiifllfibiiifii s::a::am;1s:)u:1<:i such as
`
`5:‘-u1w:stra21t in aimut Early EEZBIBO, which i mid:-;rs£a11d is wimn the priarity appiications
`
`smpprrartitzg the: Evans Appiication were flied (paragraphs 1 I
`
`- 3:3 bcziazrew and Attachmant
`
`E3). Cit3:}1:i{}flS to literamm and patent razfrsnfrncm in {Ilia Declaration will be in {hrs fauna’:
`
`Lead Aut:1mr({)atr;a), and the full cimtim1:~:. are giwsrn in the Tabla cf References at the end
`
`uni’ this Dmzlaratican. A copy af each cited referexzce {m aimed pctartiszzns :3? the hanger
`
`rtrmmamgi is irsclucieci in Attachment F under the Tab mamismr mated in the Tabb: of
`
`Rt’: fcrexzces.
`
`In T3513 2 {If the Evans Agsplicaiion, the saolubiiity sf fulvestram in casts: {xii apgmars tn:
`
`haste man transcribed inmrrectly from the originai murce, the iabnratarj; nmmbook. The
`
`Wilma in the latter is 24.5 mgfmi and nm 29 nzgfmi. in ether exprsriments to detennins the
`
`snlubiiity Elf fuixrestrant in casmr oi} anti 3130 in banzyi §)é§'1?gK.'§ai€,, some vazriabiiity was
`
`obszzwmi.
`
`In Table 3 cf {hit Evans Application, the given rmiultzility vaiues were generated at 4"{Zf
`
`and mt at 2§"C3 as is mated in the title of Tabia 3. For “ftlivrsstrant fmmxulatiana, it is
`
`praferabhz that the fulvestrant rmnains mmplemly in s<:¥u£i{m at bat}: 40C and 259C. Tht:
`
`4"(Z temperature: mrrespnncis is the storage txtmparaturszt {iZ"C‘. ta: 8‘‘‘{? in the FBA apprcxsed
`
`Iflhfi} fur Fasimtlaax}, and the 25%’. temperature mrrttspimziis to the administrati-an
`
`temperature: {ambient temperature; In axiditiatm, {ha specified saiubiiity values on this
`
`"faiirie: 3 are mszam vaiuvzzs. ::a{c:u¥amd fmm E-3i‘:£i1}f$ii-3 ~;3frep1i::am samplzes {mm em: or mere
`
`trials. The i:1di1:i:iuaivaI,uesLare Bhawn in handwriting in 112:: amenéezi watrsimn ::fTaI:a£e 3
`
`in Attachment A. In additimz, it appears that {I33 mean vaiuefi far that East three
`
`cmrzptznsiiizzarzs have been inxtm-rectiy calculated. The cmrrcctcd mean values, together with
`
`the cmrection {zf {he iempsraturc: fmm “‘25“C" £63 mad "*4"‘C’fi are 3350 shown in
`
`handwriting in rim axnerzded versaimz az::fTahl::: 3 in Attaciwncnt A.
`
`138 E :‘t}l£':43fi{}{; L 5
`
`Wt
`
`Astrazeneca Ex. 2135 p. 2
`
`

`
`fan
`
`I have evaiaaied :11: trzmscrigntion and ether t3:lT{3I':£ againzst the miginai appliszatimt:
`
`disclasures and mncinde that these dc: mm: nhangtz ihe uitimairs cnnchzsinns made from fin:-.
`
`data as migjnallzs repnned. The aciciitiezm {if 15% W51; bmzy} henziizaate to ccampositiqns
`
`iwmzing txgrmi aimhol ccsnztantratisang in xzagtmr :::i§ 4:31‘ 1{)%, 35%, 23% and 3{}% w;"v
`
`unaexpatcttstlly prmédes a posiiive affirm ma fixlvca-;tram1; mlubiiity, significantly incmasing
`
`me scsiubility :31’ fuivesrrant in the c<m1p0sizimns¢;ies:pitz=; fhivestrant having 3 Inwer
`
`smiubiiity in hxanzyi benzoata than in either akznhctat £31“ caster mi.
`
`7. An aticiitietmai set cfiexperimenm has been caruiiucteéi at ?.S“’Ci mxdaszr my guidance in
`
`abtaén ccrnsisreni data with reduceci variability frurn 21 zaingie ism 1:2? rigmmusiy rzenircsilad
`
`:ss;:+iu§:xiEiLy €:x;:¢.riVr:1e::ts ami {£3 citamnnairaiti ihat; fiat: un::x;>::<:t¢:i mzmtasa: af 3€}iu§ilit§,f {sf
`
`i’i1§w::$trant by aéciing be.:1::y1 bztnzxzaatrzr intact c::1'r:;::c::$iti-zrms mntraining mhansi, heazyl
`
`alcmhmi am} caster czii, ia prr-maxi: across the izarxmcicr rmxge c~f;:»::>;np<;si:i0n encampassaed by
`
`{he cziaima being presenzeii with this Baciarmimn. ‘Fm: zmlubility 1::-f fulvesstrani in iimnsijyi
`
`§a::n:a:<>ai<3 mad in casmr ml was 313:; mrrasumci in tha: same: 3133 :::f£::-apszrizfxa-E:11t:s usmg the
`
`1-zarrzi‘: batch £3fIJm1zyI henzsoate ant} the same batch nf casmr mil as were used ts make up
`
`the -::<:«m;msitiom*. The Experimantal T6221 Procedure is ciescribnati in Attachment B
`
`3. Tbs‘: ntzsultss fmm Iihersc soiubiiity mvaperimxztnts are ahuwn. in rim mbic in Attachment C.
`
`’Ti”he:~:e results Shaw that the snlubiiity Of fnIv£:su'Ani in t;:a:~1tor nil amine (21-4 mgfml} is
`
`significantly greater than me mlubility of fulvegtrani in béxazyl benmate aicxne: (3.8
`
`mgfmi) and ::ier:m:msirate the mxexpecicd ixiereasas in fuiwstratxt aaiuhility on the adxiitisrm
`
`:3? ii}, 15 am} 25% um: benzyi bsznzasatc, in place ufsm equivalent amaunt ufca$t<:::: tail, to
`
`mxnpnsitimns having Imai aicohcxi csncentraiinrzs in casts: mil 0‘? lf}%, 15%, 23%, 25%
`
`and 3{}‘3f£a wins,
`
`9. Thug, .’l‘.§1t3: resuits that were nbtaimeé fmm ex pttrimennes rzonziucimi undaar rigaazarcmsiy
`
`:1:-Lmirxrmeci mnziitions and with an gxpanded range: <3? campasiiions, as Shawn in
`
`Atnmhzmstzt {L eozrzfizrm {kc ultimate conciusinns drawn fmm the rcsnitg iahawn in Tahh: 3
`
`3f the mriginefl appfécatian £iisc§05u.re, nameijg that ‘(hit a:3diii::n1 inf .EQ"fe‘5n ta 2§% wfiv btmzyl
`
`E33 1.-‘ia2{}«12:3{3¥x. I
`
`3
`
`Astrazeneca Ex. 2135 p. 3
`
`

`
`benzuate ta conxpaaiticms having imai alisalml c:cxz1s;::mtIati1ms in cazsmr {xii nf bemeen
`
`10% to 30% wfiv uxiaxpectedly pruvides a pasitive effect an fuivestraxxt salubility,
`
`significantly incraasiszg the s:.::£ubi}it:3s cf fhhsestraxxt in the campositimxs deapite
`
`fixivesstrant having 3 lower sciubifity in berszyi betnzoate than in eithezttr ziicnhai €31" castcyr
`
`oil.
`
`H}.
`
`During the cflums: {if my study ct:»f the Evans; Sxgzpiicaticjn and the undtsrfying source:
`
`materisfis it was drawn m m}; attezntion that same of {he compositicm éata given for
`
`Ueiestrngen an-ii Deizziutin sumehnw had bean Silifiéiéi aria miumxa re: the right. Thus, fixer
`
`I}e:1est:¢*agen,t?1e 173% and 58% figures 3hcrwn under the B2133 cnlurnn shcmid have been
`
`under the 011; cziaiumxx; the 2€1"/éy and 4£Z}% fig;tu’esL aheewn under that BEZQH: cu-lumn shaufd
`
`have beam under the 332:8: <:cx1umr:;.:m~.:i the 2% figureg Shawn unéstr 1:§:C}E~I.:ahou¥d have
`
`balm ilfiiitfltf the“: 33:33 coiumn. Sirzaikariy ftiit‘ Iikalaiutirzg £3122’: “up ti} 2‘V:'” shcawn unefier that
`
`EROH wlumzz aheiulci have been umier the B:§()l-—l coiuznn. This table: rteparts that the
`
`scarce: saf this ciata wag .1 .PhaIm.Sci (1964) 53(8) 391, which is Rifflcin (I964) elsewhere:
`
`referred ta in this Ue::¥.arat.it:rn, ané I have: sssisu verified the cur:'wte& data fitzm the entrias
`
`far Ilalaiutiu and Delastrcegen in PER (I*'3'}‘3). A cup}? ::sfTabie 1 fmm the Evans.
`
`Appiicatican ii; regxruduced as fi5L1f¥3G}1i’I1€3I1t D, an which thaae Cflflfifliififlii have been made
`
`in h§£tldW'I'it§1'tg, and I’ have additiofialiy mars: cmrectly noted that Bclaiutin is I7-hydrUxy
`
`pmgcsteruns: ci‘£2f_H*‘{}£‘H‘€*, and that the “CC).M1”‘ designatisan fur Dc=:!.a!u1;ir; shnuld be “EMS”
`
`(Brist<3l—Mysz::rs Squibb). Attachment I} also includes a cane page expianatian of the
`
`ccm'ecticm::: to this Table I.
`
`H.
`
`In abauz early fiilifllé), 3 pa,-rsszsn r.=2sp::ms:;ih¥:: Fm ciexreiaping a $uSf$iit1f:€§ relassé injectaizak:
`
`fazxrmuiaiicaiz suitable fear aciministraticm {E} Ixumans far a new steruidal mmpound 313:}: as
`
`fulvesrrant, weukf have had apecializmfi training and sxparicnce in xiezvalaping
`
`phamlaszeuziczal fbxmuiatiens and methtixda Far their afiministratimi, In devetkxping Séuilh a
`
`fssnnulaticm far fulssastrantg the czbjective wuuid have been to fcxrrnulate an intramuazzzum
`
`{VIM} injectian that would prmzidwz fur the satisfactory sustaillcd miease tkffulwssrrant mm‘
`
`a perictnsi (If at Leasa £W‘£Z) weaks and praferahly (wit: a period of at ieasa: flzzrur weeks in
`
`meitzces the fretztgtaenuji erzf‘ a&miniatrati<:m. and wmifd have a target fu1\rc:sm*anz: cczznterzxt {If at
`
`I')I*3¥!€1?.*342€:-{Btu}
`
`at
`
`AstraZeneca Ex. 2135 p. 4
`
`

`
`least 45 mg;"mL 30 as to providz: a fulwzrstrant dose Bf at least 250 mg in a single 56 mL
`
`injection. Frmiii my personal ex perience and knmwledge of the: literature at abaut that
`
`timei Iheliexse that such an experienced formulator wmilcl likely have approached the task.
`
`of cievclopingla fonnuiatimn for fiilwstrant in abeutitha follcifwing manner.
`
`12. (Sivan the fbregeing objective, this experiencecl ftmnulamr would have appreciated that
`
`the traditimiai adrnisiigtration optimxis to explore. were intramuricuiar gm} injection {if a
`
`rsusiaimad release aqueous or oil suspansian or an ail-based smlntinn (depot) mntaining at
`
`least 250 mg of fillvsstrant in a vmlixmt-: of vehicle that is tolerable fm injcction, i.e., mi
`
`more than 5 or 6 ml;
`
`l3. Becauseiof the extremely low smlubility of fiilvestraiit in water, a tetasmnabic starting point
`
`would have: bean ms invegiigate intramuscular injection (if an aqursous or all suspensimn sf
`
`ifiilvestmnt, However, the {emulator would have found that injection of an aqmzmzs
`
`suspenaiszim of fulw-zstrant resulted in extensive local tisaue irritation at the injectimn aim as
`
`well as a poor release prcifiltt, such as reported in paragraph [0042] {if the Evans
`
`Application. Since suspensions thus were net 311 a(3CE:ptal3l€:.Q}}iiflI1—f0f fulvestrant, the
`
`experienced forrnulamr would have muveri on to» further explore whether 250 mg of
`
`fuivestrant cnuid ha solubilised in no more tlian 5~—6 mL mfan c3il~base:él V’6.l‘.liCl,~‘5£, z'..«2., to
`
`a€.3l‘li£:V£.’: the target flilwsstrant mncentratioii sf at least 45 mgz’mL.
`
`14. In the prefcmtnulation phase, the experienclad foimulamr would have Lzundnctzgci a
`
`‘literature review car mthemrise wmilci have become familiar with mminemially marketed
`
`injesztablts fcmnulations, particularly inj eszstabla sustained release fmmmlations of stemicls
`
`or either relatively insoluble compounds such as thmze listed in Table 1 of the Evans
`
`Application, with the objective of identifying potential oil vehicleii, co»-scilvents and either
`
`cxcipiems that already had been found to be mlsssratexi audio: to have passed thmugh
`
`rcgu1at0z'y1'£-viisew, and which might be candidateg far fi.1l‘lCl‘lf21’COI1S1lilBTé1ti£}n and testing for
`
`the fulvestrant fcinnulaticsn. This raeview also would have pmvicled guiriante witli respect:
`
`to cmncmitratimn lcvels of such C0*7${)l‘v’€3I1l3 and anther e:-mipients that generally had been
`
`foimci accaptablc in sustainmi Ifillfiafifif oil~baser:1 intramuscu1arinjectian.s administemci to
`
`DB li(i2U-'1-26{3f).l
`
`Ur
`
`Astrazeneca Ex. 2135 p. 5
`
`

`
`humans. This mbjactive is cmmnned, for example, in Name (199?) at page 166:
`
`Generaiiy, a knowledge ofwhich excipients have been deemeé
`safe by the FQA or are aiready present in 2: anarkrzmd product
`pnznziéses increasad assurance to the fonmslamr that these axcipients
`will probably be safe for their new drug product.
`Regulatory
`bociiszs may View an crxcipicmi pmvieusly apprnvcd in an injectahla
`dosage form fawmbly, and will fre::;uem1y require £633 safety data.
`
`The purpeseof this Nema paper was thus “to present the various excipients that have
`
`been included in the fmmulation of inj xsctabie products marketed in the USA.“ Similar
`
`objectives were intended to be served by the campilations of‘ commercial fonnulations in
`
`Strickirsy I (1999), Strickiey Ii (2000) and Strickley III (2000):
`
`This compilation will also be useful {hr those interested in
`knowing What additives are currentiy used in injectable products
`and at what eoflcentrations they are aciministered in p1‘ac-‘ace. This
`ccampiiation only focuses on marketed fonnulations. and does not
`delve into the subject of preclinical or drug disccrvery fmnnulatians
`assmciated with ear1y~stages phannacokinetics or pr00f«of:«cczncsept
`phannacodynamics, where the fannulation scientiai is not bmmd
`by regulatory constraints.
`
`(smckzcy I (1999) at 324).
`
`Pm-veil (1998) similarly statc$ at page .238 with respect ta its mmpilation Qfcmnmercialiy
`
`used excipients:
`
`Thus, the fonnulatien scientist is often faced with a diiemma —~
`which excipicnts are truly availabie fur use (baseri can what has
`been lifiitii previm1s1y},Lam:i which are nm‘? And at What
`mncentratinns, anti by what mute‘?
`Herein are listed the excipients found in most cf the approved
`and marketed parenteral fennulatinns, given syst-amatically by
`excipiem name. in this format it is easy to cletermine what
`cmncentratinxm were usad, the route. Bf administration, the main
`ratienale fer additicm of that exazipientg the drug that was
`formulated, the manufacmrer, brami name, etc.
`
`15. Fmm the Literature; review, the formulator wauld have muted reference to 3 number cf
`
`intramuscular injectabie sustained release oilwhasefi gtemidal fnznnulations thaé had been
`
`‘ 1*-Eizma (3997) ii(;4'l':$‘!:€il1fi£)I‘l. huwisver. that there“ is 11:: guarantee that the new drug pmfluci will 13:: Siiffl as cxgcipicuts
`are ccsmlirineci with mhcr additives arxcifor wflh a new drug, creating unfmcsficn patcniiaticzn m synergistic toxic
`szffizcis,
`
`DB E:'62{}=%26£}6.1
`
`{1}
`
`Astrazeneca Ex. 2135 p. 6
`
`

`
`ccsmmercialiy marketed:
`
`Strickiey I (1999), Table VII:
`
`Hakzperidnl Decanoateffialdol decanoate (30-100 mgfmi, in ssesame oii, benzyi
`
`aicohczl 1.2%‘);
`
`Testzz-5st:':r0:1r:: Etlanthaiefficlatestryi (200 mgz‘n1L in sesame ail, chimobutanol 5
`
`mgfmm;
`
`FDR (1973) at pages 1277-1278
`
`Pruiutcztzfprcagegterane (SQ m‘-§z’mL in ssesarmz oil, 150 mgfml banzyi hxznzoate, S
`
`mgfml benzyl alcmhoi, I mg/‘ml propylpagrah-en);
`
`FDR (1973) at pagss 13494354
`
`Dcsladuinonezffestcsterone Enanihatr: 8:, Estradiol ‘Valeratc: (90 & 4 mgfmi, in
`
`sesaxne oil, CL§% chiarcbutanoi);
`
`Delaahimonc Qfiffestosterane Enanthata 85 Estradini Valaraie (180 S2: 8 mgfmL in
`
`sesame oil, 2% benzyl alcohol);
`
`DelaIutinfhyéroxyprogesterone capmate {25(} mgfml. in 52% came: oil, 46%
`
`bm1z3.r1 benznate-3, 2% benzyl alcohoi);
`
`Defestrogenfestraciiol valrzratez £20 mgf:mL in ?8% Castor ail} 20% bcnzyl benzcata,
`
`2% balmy} alzruhol and 40 mgfmL in 58% zzasmr oil, 40% bmzyl be-xnzoate, 2%
`
`benzyi aI~::mhc»1);,
`
`Elelatestrylffestosterone: Enanthate (200 mg,:’mL in sesame 011, (3.5%
`
`chlerohntanol};
`
`Deialuteval 2}{fhydroxyprogze-fisteroms capmate: Sc esfradical valagrata (250 mgfm}; 8:;
`
`5 mg/mL in caster Q31: 45% benzyl benzmate, 1.6% benzyl alcohol);
`
`FDR (191734) at pages 1391-1392
`
`Prnlixin EinanthatefFIuphr:na;:inciinanthate (25 mgs'mL ‘in sesame: oil, 1.5% benzyi
`
`alceahol);
`
`Wang (1980):
`
`Drrpo-Testogteroneftestmstemm: cypiarmte (100 mgfmh in 8?.4% cottonseed oii,
`
`(3.1 mL. benzyl benmate, 9.45 mg benzyi alcohol a3 3 preservaiive);
`
`Macksy (1995):
`
`'I”cstc:.~vi1*0:1 fieputftcgtosierune enzmthate (250 mg*’mL in rszmmr ml and benzyl
`
`1:»m;e2042mns.1
`
`7
`
`Astrazeneca Ex. 2135 p. 7
`
`

`
`bntnzmttcjg
`
`as; wail as as numbaiar cf mhezr cxanmméarcialikaed 03,} based Iczsng-ac:%:ing IM irxjectable
`
`fiizrznulatimis a‘:;*;::.~<:srtcd on Table ‘I of {he Evang Appiicatim-11.
`
`16:. A3 3 further gar: sf“ this prefammlatircm phage, {ha cxpcriaxxcefl fm’Iam.1§atm' would have
`
`aonductaeci 3 prfiafannuiatien sniuiziiiiy screen, separately measuring the zmlubiiity mi”
`
`fulvestrant in .3; range :3!’ pure soivexum, i:3cIudi:1gI“h«:: pmteniial oil anci ::+::--mlvent
`
`canciidatcxg that had been identifimi in thri: abmm: literature review mg being Suitable fm
`
`inciuaion in intramuscular injrsctinxx fuamuiatium, Sea, for %X£i’!I‘lp§¥f:, Gupta (1999),
`
`Chapter 1“? at pagr-3 462, mader the heading “Fmm:_1iation Dfl'¥£%1€}pm£3t‘:f"i
`
`The activities _nect:ssa1;v 2:3 iixwciap 3 parenteral preduct atarz be
`plarmd imam the ftmewing flame broad areas: prefonnulation.
`ftprrnulmiam. and scale.-2~up‘ Wham tlaerae are alternative develapment
`persgractives, 3&1 development uitimamly needs ta acnamgaiiah the
`same activiéiefi. Prefgnnnulatisgm incluées the characierizatics nf the
`
`bulk drug 1331.13 initial scrréelaitng {hr éxcipient cmmpatibility with {ha
`drug.
`
`“Prcfm'mniaiinn a;mx'iiAe:-:” are said in "’;:rrz:wi<ie fnndamentai dam anci szxpariexxce nccassmy
`
`In §t3ViiIiIl}3'f7iZIff{3t.11aii{]I32§ for :3 sprzcifixz cnnaptzund” inxzlmiing, :13; item 3.1 in the izufkinc <3?
`
`areas nf specific interest. a aimennizmiiim :::f“‘$mi12I:~i}ity"’ in ”‘$z33z*::::t¢.:z§ 1-zuiveanmf’ {at 433 }.
`
`“Signific.m21 fimnulatim astiviiies begin with iniiiai prefonmaiatimz data and knmvledge
`
`cf £113 spccificz :mm£: 0f’administrati(:m" {£1485}, which “fonnuIa¥:iL:m mrtiviriczs include that}
`
`identificatinn and seiectinn af 2: suitable vehicle fiaqueuua, nanaqumus or to-solvcnt
`
`system} H." {at 2:104}. it is. further mateii that “irzj:~2mirm wrsiumaz ire; m3e?u::sf‘fi1:=; most
`
`irnpertant :::{?11‘Lfi3§{i&1’iiii{3I‘t$ in the finnnuiatmn dmreiegxmrrrai {}f a mmmczrcfial pmziuct” {at
`
`405“). when cartryiazg nut such 3 prefunnulatism snl::¢i::«i,lity scrmn with fnlvestrant, the:
`
`Vfannuiamr weuid have feund that fuiwaszram had exiremeiy ixm-* seiubility in water, haw
`
`soiuhiiity in mass: 1:615 {but highest in cassmr mii), kow snlubility in henzyl bcnzeate, anti
`
`the highest :~*A;>¥uhi§ity in ethzmoi and 1:m:1:+:y1 alcchnifi such as retpcm-::d in Table 2 of the
`
`Elvans Appiisgzatixm.
`
`1?,
`
`With £1223: ;i:1%1:m1a£§:.1n an prim: c<::r::r;1crciaEi2:a:c3 Emrznixiatitamt and mi: fuivesirani saiuiaiiity
`
`data frv;m1 the pmfcsmlulazian :~:r;:§:2:.=:n {mach 21:; rcpertrzti in Table I’. mfthc: Evans
`
`C3333 £3,121‘?-«*»l?.f:*Z5l:. 1
`
`3
`
`Astrazeneca Ex. 2135 p. 8
`
`

`
`Application), the expericxacazd fonnulamr wxtmlcl lzawz selected castnr oil as §.l‘lf.’3 all vehicle
`
`because ml’ the higher solubility cxf fixlvestrant in Castor oil relative to the other ails tested.
`
`Nevertheleszs, he would have appreciated that the target fixlvestrant concentration of at
`
`15333145 nagfml, coulai mt be achieved with caster ail aims, and that a <:0—soI‘w.=:nt wntzulci
`
`bi: required.
`
`18.
`
`A number cf tlm commrsrcialized fnrmulations that would have been identified in the
`
`literature review (including the Castor oil-based fcmnulatmns} have a substaniial henzyl
`
`benzmate component, which may be ‘present as a cn~30Iv:3nt. Sea, for Example, Delalutin
`
`noted in paragraph 15 above, which is reported in FDR (1973) and noted in Table: I of the
`
`Evans Application, and is one of the fonnulatinns discussed in Riffkin (l%}4), “Castor
`
`Gil as a Vshiclva for Paremeral Administration of §l;£‘:£‘€;)l(§ Hormom=:s” (3:-re Riffkin n. {is},
`
`Dreslaintin is 250 mgfmi. l7~hydmxyprz3ge$t£:rons3 zzaprceatc: diasolved in 52% Castor oil,
`
`:=t6% benzyl benzczate and 2% bcnzyl alcnlml. Hewever, Riffkin Table 321 reports that the
`
`solubility of 17-hycluxyprogesiernne capmate in master oil aims is only 55.6 mga’mL, but
`
`the scslutility :31’ 17~hyciroxypr0g::stmom: capmate in benzyl benzoate is substantially
`
`higher, being at lsaat 251} mgfml. (see exam;-ls 4 offluber {US ‘$20) anti Attachment E
`
`discussed below). Even if net needed as a cosolveni, Riffkin (1964) notes tha1“th<~:
`
`addition of ‘benzyl alcohol or betzzyl bemzcsatez to caster oil msulted in a lower and more
`
`flavorable viscosity, making it easier to injcs:£”‘ (paragraph bridging pages 893-894}.
`
`19.
`
`Hewevm“, the skilled faxmulamr would have appreciated fmm the fulvestrant solubility
`
`data gmxerated in tha prefmmulation screw that fiilvatrant had very cliffergnt solubility
`
`charactezisiicsz relative: m the steroids mf pravious cmmnercial formulatinmz. Attachment E
`
`is a cempilation showing the chemical stmemres and Lrelative salnbilities in Castor oil and
`
`sesame oil crf the wmpmuncis named in Riffkin {1964}Tab1e Ii mmpared to the smmture
`
`am} the gnlubility of fulvfistrant in these oils. It can be seen thai the solubility cf
`
`fixlvestrant in castor oil and in sesame all {20 mgs‘mL and (3.58 mgfmL, respectively, from
`
`Table 2 of the Evans Application} is appreciably lower than the salubility ofihe ether
`
`st:2:mid3 in these oils (taken fmm Tabla 11 cf Riffkin (1964)). The second page of
`
`Attachment E mlmlates the mncentration in benzyi benzgoate of five: namecl steroideg, takeai
`
`D33 l;fi2£lJ.-2666.!
`
`9
`
`Astrazeneca Ex. 2135 p. 9
`
`

`
`fmm Examples L5 c1f'Hubcr(L5S ‘SZ2i})., ranging from 200 to 40:1 mgfmi? By
`
`ccsmparist:-n, the scziubiiity of fillvesflamt in benzyl benzoata is reported in Tabie 2 of the
`
`Evans Appiicatimn as being 01113: 6.15 mg;‘mL., and {mi}; 3.3 mgfmh as cietennimzd in the
`
`recently cxtmduczterfi tests _re.=pm-ted inmztachment C.
`
`20. The flxpmienced fonnulator thug would have expected that bemzyi henzoate wcxuid :10! act;
`
`as a cowmhxent far fuivestrant in casmr oil because the soiubility a:»f‘fuivestran£ in benzyi
`
`benzczate was gignificmtly irzvwer than its salnhility in castor oil. The addition of benzyl
`
`henzoate in {3Ei$tQI‘ oil, for whatever reason, wcmld harm 133211 expected to decmase, rather
`
`man increases, the solubiiity mf fimrestrant in the resulting caster oilfbenzyl banzaate
`
`mixture. ‘This is canfinned in Tabla 4 of the Evans Application, which mpons a
`
`fixlsmstrant seiubiliiy ofczmiy 12.6 mg/mi. in the magma: mi} whicie containing cmiy 15%
`
`hcnzyl bmzoatc, compared to the 2{)1ngfm1.. mlubility of fixlvfistrant in Castor oii aione a$
`
`reparted in Table 2.3
`
`21. Baaerd on the solubility ciata determined in ma: preformulation rscrseza (stmh as reported in
`
`Table: 2 of the Evans Application), cthanol antidor benzyi alcszrhol wmflxzi have been scan as
`
`the best {:0-scziwznt candidates for raising the fiflvestrant soiuhility to the 45 mg,:’mL targat
`
`in the castm‘ oil ‘vehicie, and wcmid also funciicm to lower the viscosity of the resulting
`
`fermuiaticm and make it easiar to inject, Consistent with this soiubiiity data, Dukes (US
`
`‘$14) added 49% WW berazyl alcnhrsrf in ordtr to dissolve SE} mgflmi; ftxlvestrant in the
`
`Castor oi1~basa::I formulation used in the e:xpe:rimcnta1 rat studies of his Exampie 3. It thus
`
`would have berm: appamxxt that 40% WW benzyi alcohui could fmactiun as a co-solwzm in
`
`caster oi! 1:0 achieve: the: target fulvestmm mncentratmn. Nevertheless, this: skified
`
`fmmulamr would have been cmnczmed with using such a high alcrznhvgal content in
`
`imramuscuiar ifajactabla fxmnulatiéng far administration tn) a human.
`
`3 Data taken ‘fmm the Examples (if H1.zber{US ‘S2-Q); these are m;m::e21trations'us+2c1 in $112: examples and mm
`ncc:essa:'i}y tbs actual maximum solubility of gash steroid in bcnzyl benzaam, which may be higher. Huber was a co»
`auiimr an Riffldn (1964).
`7’ It shouizi he noted that in the fizrther tests that were recantly emanated under my g11idance{paragrap£1:~; 759 above
`amzzi Attaclznments B and C‘ herein), tin: sxtsiubility czffiglvsstrant in castor oil iflfilhh-£3 was: again Iesmd and found tn: 1::
`31.4 mgfmi, and thas-nlubility raf‘fi1lve::irantifi1 benzyi benzaate alum: was; again tested and fimatd In be Qnly 3,8
`mgjml... which fizrtixcr ::0nfin11;<.thz1£ §:.'rm1;:yl bcnzceaic would rm: be c3q7u=:r:tcd in act as 2: cmsnlvcsit for Fu!vesh'a_nt in
`czzstmr oil.
`
`II)133:{:22}=326{1t3.1
`
`1;}
`
`Astrazeneca EX. 2135 p. 10
`
`

`
`22. Firm ztrf 313, the E}‘.pfi3‘i£:3I1iL‘-(‘iii fannuiator would want in sninisntzfi the amzmnt {ref mt»
`
`mivcnts anti ztxsipicnts in any injectahisa f::m'nuiat:i<:an. F*”m* iixamgle, as statexfi in Gmpta
`
`£2999)‘ Chapter Ff, “Fonnuiation and A<i:ni:1iss£mzi<::an Tcchniques 1:1: Minimize hzjactinn
`
`Pain and Tifime Btamage Assuciaméi with Parental ?rm»:iuc:£3" at page. 4&4:
`
`Ceusamvents are czomtncsniy used in ctahance drug mlubiiiiy and
`fiabiiity. Cxasoivants may imziudn <:thzm»::>l_. gnrmpyisnc giycol,
`pnlyethyieixe giycals, and giyr:m*ix1rs:. Thesgc -i:£}m;§0!1B11IZShaV§3
`intrinsic effects on himimgic tissmc: and can alter the prmpcrties of
`omssr <“:x::;:ipiem3, thus infitmncing the tissue damage or pain caused
`by at pmdwst. There is. a dsanh cf” iiteratum cm {Em pain caused by
`msnlvcnts, but there £3 21150 a grvwing body of lcnowiedge an the
`tissue damage that {hey can mange. it is mist mrtain that tissue:
`damage is always ciirectigé cormlaitcd with the izzjectimn gain, but
`minimizaiirm csf both pain can injmrticm and pnmntiai tbs‘ tissue
`danaagra slmuid be inciucieti in the prxncfium cimseinprnent pian.
`
`See akin (Emma (1999)., Chapter 3 1, titlad C3o5:11ve;’x‘£ Use in izijectable Farmulations, page
`
`21?:
`
`Idszrzsliy, it is best is 331%: anti um 1-mlvrrixizs that wgmid maximize
`the solubility ofthe compaunfi. ,Mmcimi:zi11g the solubility 0f .23
`mmgmunci in a gzsarticniar camivent aystam wnuld msuh: in lower
`{ma} Iesveis Qf the non~aqne<:aus S()I\f€.i’Iii(S) being miminisiered ‘K0 the
`patient, thcrehy Iowering the chance far gmtxzntial aide: effects.
`
`'I’}:tis objactive Wonk} have appiied to aqurmuis anti x:2i1-based systems alike, in that the
`
`prexzrstdemt of ccxmmercializeci fonnulations idenigifixad in the iitzrmtunz review Wonk} have
`
`czmxfirrneci that fixed oiis, such as casts): aii, have icmg beam mmmerrciatty used and
`
`'us1r.:m:;::t::*:<3 as; the majar comgaozzent 31°" 02%-based ffiflfiifiiflfiii mieastz intramuscniar injectable
`
`szcmidal .f':mnuI:a1i0ns. fin {ha ether hand, £30-solvents such as £ifi13;3{}i £31“ benzyl aicehal
`
`have gtzzmraliy been used only in far lesser cancizntraiiong. 33 discussad in the fmilms.-‘ing
`
`paragraph.
`
`23. Thus, use afsuch a high cement nf either imnzyi aI<:<3§2<:I :12: ethane: waniti have been
`
`cmmrary to precedem as shcexvn fmm the rmirsw uf mmmerciaiimrci mi-based
`
`intramuscular ixzjatctabie sustained release f€}.fI'¥3L1}&IiiI.I!!1$, The Iimramre rmeiew as rcsf early
`
`2{.1{3{} xmzmid lmw: ahawn that any henzyi afmzhnz} in sun}: fammuiaimzis
`
`aimost aiways
`
`{}¥:3I§{12i}-i2fi’E}fa,I
`
`1 1
`
`Astrazeneca EX. 2135 p. 11
`
`

`
`presczm as a pmsexnrativc in 2: concentration of abcut 2% or less, oczcasionally at 3
`
`concztntration of up :0 5%, but rzmly rarely at higher concemtratinns. With respect to
`
`benzyl alcohol see, for example:
`
`Gupta (£999), Chapter ll at page 22?} stating that benzyl alcohol “is; typically ns~:*c1‘i‘n
`
`conceniratinns of up to 2 percent aa a pregervative and up in 5 prrzsfent as a solvent,”
`
`and then discussing reported toxicities.
`
`Mama (1997), Tablfi; ‘V’ at page £68, rcpnxting that lmnzyl alcohol was present as an
`
`antimicrobial preservative in ‘F4 injectable formulations (net limited :0 oil-based IM
`
`fonnulations} 3; cmncemirations af from 0.756% {nme ihat benzyl alcolml is net
`
`included at all in Nema Table 1, “Sc2l*s2<311ts and CO-S£)lVnBI‘l’ES”;
`
`Powell (1998), the benzyl alculml lisiing at pages 244-246», particularly those
`
`indicated as lming used in IM fcmnulatiomsg
`
`Strickley I (1999) at page 32'?) notes the inclusimt zzsf 2% benzyl alcohal in an lM
`
`lmrazapam fnnnulatimn in 2: pmpylem: glycol vehicle, but times not include henzyl
`
`alcohol at all in Table ‘VI listing “Ccpsolvems Uszad in Parenteral Fonnulati0ns;"
`
`Lopatin (1972) mating in Table 3 at page 727 apposite Benzyl alcohol, “Tame. User}
`
`in cmxcentration 31° not ever 3%. Has irritant aciion in cancentration of S%;’"''
`
`Cczmelima {US ‘3fi3), col. 1, lines 30-35 stating, “It is known that the selnbility of
`
`steroids in vcgrstable or animal oils can be increased by the addiiionl of excipirsnts such
`
`as be-nzyl alcohml and benzyl benzmatsz. An objection to the use of such excipients, and
`
`specifically’ benzyl almlml in somewhat higher cosxcentrations, is that these agents
`
`may irritate the tissues,”
`
`Thfi limratum revit;-W as of early 2000 alsa wmfld haw shcmm that, with few exceptions,
`
`ethanol was: not included in such fcrnnulations in excess of abmxt 10%. See, for example:
`D
`
`Gupta {l 999), Chapmr I I at page 225 mating that athannl has ‘been urged at 3:.-'-Nels up
`
`in 50 przrment, but these lntvels typically are associated with pain on injection;
`
`Strickley IT (1999), Table V}, “List sf Cosolvents Used in Parenteral I?<>rmulaticms”
`
`more specifically lists. the etlaanol content in [M formulations far specifically
`
`icicnitifixfi drugs, which cnncentraiinns range only from 2.5 to 10%; an {MEIV
`
`lctmazapam, fbrmu12:1:i(:m in a propylene glyccal vehicle is noteci at page 329 as having
`
`18% alcohol, but is not included with the {M fbnnulatians in Table VI;
`
`[EEK léé3£lal2i’;i)6 .1
`
`12
`
`Astrazeneca EX. 2135 p. 12
`
`

`
`«-
`
`Exfczma {i9',~}‘?}, Tame: I, “Scs‘i%.re;11s, and {3<:m~saaixe*:::1t5;°” at page SEQ ligts ethane? as i:u::-mg
`
`in E4 fmnnnlatimns with a mmzemratima range of Q.fa?~80‘E>’a {far ?r<::gra1}; mam that this
`
`is misieaéizzg, hawfivcr, since Pmgmf £55 :1 €?Q!:I£2a:'r1£r£I£r3
`
`fizz
`
`izltrawzzxans infusicxn xtanky,
`
`ami $5 in his diimtt-sci 250 to $988 timm br.:f:::::: aiiminixtratmx;
`
`an
`Fczwazli {£998}, Iistzé “a1m11<::§" at page 242 2121:} "‘et¥'1yi aicohofi ‘ at page 255, wherein
`
`4»
`
`that: ::t:han<:.:1 cmmentration far {M ibnnulatians ranges fr:::\ :1 £1.61-38%.
`
`24.
`
`'I‘hus, even Lhcmgh iZ1111=,:a::s(US ‘S 14) had dezxnorzstratcd that I a target 45 mg;'mL
`
`Fulwmtmnt mncentratian ccrnid 133 achieved lay aciciing 40% b<-may} alcohol tea the castmr
`
`{}i1'Vfi3hi£:1E:, the prececirzznt of commmcializecl ¥M <:>i1«basc:d $y$':=:rns wstruid have mmiivatesri
`
`1124:: experienced fnrmuiatur tn subsianiiatlfy rcduce the bertzgzi aim mi cxmtxznt cf the
`
`fmznutazian irziaxxderi far human use, anti {his cenzmerciai prxacecient weuié have mafia
`
`him very rraiurgztam to mplace benzyi aictzshnl with the fifibfitiifiliai a noun: {zef ethaaml that
`
`wrimid ht: neacicsd £9 Vmaintain if»: targszi fuiwacstmnt. cmazcntratiatz, Benzyi bcnmatc chatty
`
`wtgstaid am he ccmsgidercd tn sceiw: ihi3 tiilmnmza, but rather wouid bee: expzi-cteti in have a
`
`riegative :ef'f‘ec:t an fuivesmmzt seluhiliry since: fuivesazranz was even less sniuble in bensagsl
`
`benznam than 3:: came: 3&1, that is, me wcmifi have xzaapegzieci that adding benzyI‘i:>:::1?.0ate
`
`wmzld rcquim fitiil mare akzohni In maimaén 1112;: target fuivgsiraxxt »::{:n::entrati::rn.‘5‘
`
`25. Umier £hes;e cirsmmstances, the dégcaveny by Evazuza :2»?! a3‘., that ‘£1212 addition Qf benzyi
`
`bnsnzcmte to the caste: czsilfaicohcal mixtura actualiy increases thfit aolubiiity cf fulvcstrant
`
`such that mm-e fuiwssstrant squid has difismivsad in a given mzlumiz ztzf fbnnuiation, was
`
`uz1expc::;e3ci and tmiy surprising. This pmsétive ‘mmzrzyi tmxxzmatc ef:‘f:=:ct cm iixivestrant
`
`mlnbility in the resuiting fonnumtiata is Shmrm in Tzabla 3 (if the Spéfiification ifanré £5 11:31
`
`chaxzgcti by the ahcave-Imted mrrmximns}, anti is r:.::3:1firm::c§ am} xixzmmzstrated ever a
`
`brnader range 0?‘ formulaticm camposixiczn by tin: addiiianai set inf‘ experiments ctnmimtteti
`
`under my guieiance anti discussed in paragraphs ‘L9 s:¥:>»;:«w:, the results saf which are
`
`mgmrmd in Attachmexztfi C.
`
`awn fmm this sainhiiiiy issue, there wc-»u§r:i imam brmrs nu malivatiqm tcs add insnayi
`“ It siixmid be samad that
`banzaam {Em xsizxcmfiity :'e:3m:ti-:3n aim’.-L: aha saignificant quasxtity :3? 2:§::s;:1mi M3-aid 5<i3‘:“.-'6.‘ am chm} finzzciisxn x:-1‘ac:mg as :1
`.::;::~,~:c1m:n1 as; M25! as rg-réuscisng fiat: injection xrisc-ztmity zmd making it tzasiasr m irzjcct. whfimas tin: bzmzzjfl bcnzraaivs
`wstmlxi he mpcxtmi In hszm: :1 raxzgat