207
`
`3. Masked anti-inflammatory agents with reduced gastric irritation
`
`Despite the enormous amount of work that has been carried out on the development
`of anti-inflammatory agents, theirclinical usefulness is still restricted by their side-
`effects. The non—steroidal anti-inflammatory drugs (NSAID) are limited mainly by
`
`the steroidal agents, even when applied
`their gastric irritant properties whilst
`topically or used inappropriately, can produce an atrophy of the epidermis, steroid
`acne, stria, supra infections, impaired wound healing, purpura and occasionally
`systemic side—effects. The need for a safe anti-inflammatory agent remains, and con-
`sequently the search for a prodrug with reduced toxicity has continued during recent
`years.
`
`The gastric lesions produced by the acidic non—steroidal anti-inflammatory agents
`are produced by two different mechanisms: a direct contact mechanism on the
`mucosa [29, 30] and a generalised systemic action which appears after adsorption
`and can be demonstrated following intravenous dosing. The intensity of the mucosal
`Contact effect varies from drug to drug and from drug to prodrug and is not related
`to the anti-inflammatory activity.
`
`Aspirin, probably the most widely used drug in the world, is a potent, effective,
`and low—cost analgesic anti-inflammatory agent. However,
`it produces occult
`gastrointestinal blood loss in a large percentage of patients [31]. The administration
`of aspirin as part of a triglyceride has been investigated by a number of groups
`[32— 34]. The rationale for this approach was based on the well-established prin-
`ciples concerning the intestinal absorption of natural triglycerides. The actively ab-
`sorbed species from natural
`triglycerides are micellar particles composed of
`2—monoglycerides, free fatty acids and bile salts. Thus, synthetic triglyceride pro-
`drugs, with the anti-inflammatory acid attached at the 2-position, would be absorb-
`ed intact, without the intervention of significant quantities of free gastric irritant
`anti-inflammatory acid. This hypothesis has now been generally confirmed.
`Biological evaluation of the aspirin triglycerides (Table 2) having aspirin in the 2-po-
`sition and fatty acids of intermediate chain length (C4 — C18) in the l- and 3-positions
`(l5a,b) do not cause gastric lesions and have essentially all the systemic activity
`associated with aspirin [35]. The plasma salicylate levels following the oral ad-
`ministration of aspirin and an emulsion of compound 15a were compared in rats.
`The peak salicylate level produced by the aspirin triglyceride (15a) occurs later and
`is maintained longer than for aspirin itself. A comparison of the mean areas under
`the blood level curves showed that essentially 100% of the available salicylate in
`(15a) was absorbed.
`
`During the purification of aspirin triglyceride (15a), a by-product was isolated
`and characterised [32] as 1,3—didecanoyl—2-(2—methyl—4-oxo—l,3-benzodioxan—2-yl)
`glyceride (16) (‘cyclic aspirin triglyceride’). This compound was equipotent
`to
`aspirin on a molar basis as a systemic anti-inflammatory agent and did not cause
`gastric lesions in rat stomachs at comparable doses to aspirin.
`
`Petitioner Amerigen Pharmaceuticals Ltd.
`
`- Exhibit 1012 - Page 110
`
`
`
`
`
`
`
`
`
`206
`
`rate or pulse amplitude were noted and there was no difference in the incidence of
`subjective side—effects, both ibuterol and terbutaline being relatively free of side~
`effects at the bronchodilating dose. Human serum esterase hydrolysis [27] of a series
`of differently acyl—substituted mono— and diesters of terbutaline have been studied.
`The esterase—catalysed hydrolysis (Table 1) in this series was shown to be sensitive
`to steric hindrance in the acyl group of the ester. Increased branching in the acyl
`moiety lowered the rate of hydrolysis.
`Two different
`types of terbutaline ester prodrugs have been evaluated by
`Svensson and co—workers [28] in their continuing quest for B—stimulants having a
`prolonged duration of action, reduced side—effects and increased bioavailability. D
`2438 (11) was designed to function in the following way: the pivaloyl—phenol ester
`linkage, A, was less stable than the benzoyl—phenol ester linkage, B. Type A ester
`linkages, as seen in ibuterol (lOb), have a half—life of 3 seconds in human serum,
`whereas type B ester linkages, as measured for D 2435 (12), have a half—life of 1.5
`hours in human serum. Thus, first—pass hydrolysis and conjugation may operate
`preferentially on the outer ester linkage type A, still leaving terbutaline protected.
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`B ‘coo
`
`Bu*coo— ~co—o
`
`(11) D 2438
`
`Me2NCO0
`
`OH
`
`""°2Ncoo
`(*3) W0 2183
`
`I
`
`NHBU
`
`NHBu"
`
`NH Bu‘
`
`H0 co,o
`
`(12) D 2435
`
`“O
`
`on
`
`i /Mm”!
`
`Me2NC0O
`
`(14) KWD 2439
`
`The nitrogen-isostere of ibuterol, KWD 2183 (13), was also synthesised and
`evaluated. This prodrug (13) produced sustained blood levels (z 24 hours) of the
`parent drug in dog following a single oral dose. The biological persistence of this
`compound was attributed to the stability of the amide function and to the fact that
`
`inhibited their own
`KWD 2183, as well as the monourethane KWD 2439 (14),
`hydrolysis by reversibly binding to non-specific plasma esterases. Both D 2438 (1 1)
`and KWD 2183 (13) showed a dose-dependent inhibition of histamine-induced "bron-
`chospasm in unanaesthetised guinea—pigs by both the oral and local aerosol route.
`The ester prodrugs ll and 13 showed less cardiovascular side—effects than the parent
`drug terbutaline when administered orally to the dog. Thus, the prodrugs showed
`improved bioavailability, prolonged duration of action and reduced cardiovascular
`toxicity relative to the parent terbutaline.
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`Petitioner Mylan Pharmaceuticals Inc. - Exhibit 1012 - Page 110
`
`

`
`
`
`
`
`
`
`
`
`
`
`Although decreased gastric irritation has been demonstrated for both compounds
`15a and 16, there is no proof that these derivatives are true prodrugs of aspirin
`(acetyl salicylic acid) rather than of salicylic acid. Only blood salicylate levels and
`anti—inflammatory assay data were presented and, as salicylic acid is known to be
`a potent anti—inflammatory agent [36], anti-inflammatory data cannot be taken as
`evidence for the systemic presence of acetyl salicylic acid (aspirin). A series of ac-
`tivated methylthiomethyl, methylsulphinylmethyl and methylsulphonylmethyl
`(l7a~c) activated esters were synthesised and reported to be true prodrugs of
`aspirin (acetyl salicylic acid). As aspirin is a far more potent analgesic than salicylic
`acid, the detection of systemic acetyl salicylic, even though this was a transient
`species in in vivo biological systems, was considered to be a significant advance. No
`analgesic data nor gastric irritancy data were presented to support such claims.
`
`
`
`
`
`
`
`
`
`
`0
`
`Ac
`
`°
`
`o
`
`0CO(CH2)nCH3
`
`0CO(CH2)nCH3
`
`15 (a) n=8
`(h) n=16
`
`.
`
`16
`
`CH3
`
`o
`
`o
`
`0
`
`0
`
`0CO(CH2)8CH3
`
`0C0(CH2)3CH3
`
`coocH2xcH3
`
`OAc
`
`Ho
`
`cH2oH
`
`o
`
`OH R
`
`0II
`°-C
`
`208
`
`209
`
`The synthesis of a crystalline oz-D—deoxy glucopyranose derivative (18a) has been
`described recently [31]. This putative prodrug cleaves to aspirin in vitro in a reaction
`which is independent of pH, and has a half-life of 7 minutes at 37°C. The oz-D-
`
`
`
`
`glucopyranose derivative (l8b) slowly generates aspirin in solution. Although a
`number of simpler ester prodrugs of aspirin, including the 1 ’-ethoxy ethyl ester (19)
`[37], N-hydroxyethyl nicotinamide ester (20) [38], acyloxymethyl esters (21a,b) [39],
`and salicylamide ester (22) [40], have been described, little specific information
`regarding their biological properties is available. Decreased gastrointestinal irrita-
`tion was claimed for both benorylate (23) [41] and the lysine salt of aspirin (24) [42];
`
`
`
`
`
`
`
` Petitioner Amerigen Pharmaceuticals Ltd.
`
`- Exhibit 1012 - Page 111
`
`the latter producing three times less gastrointestinal bleeding than its parent acetyl
`salicylic acid. The blood loss was measured for human subjects using 51Cr-labelled
`red blood cells. Animal studies have shown that benorylate has anti—inflammatory,
`
`analgesic and antipyretic activities comparable to those of acetyl salicylic acid, its
`presumed but not proven parent. Clinically, the drug was reported to be as effective
`as aspirin as an anti—rheumatic drug.
`
`0
`
`OAc
`
`ocH2oc0R
`
`o
`
`0
`
`0A0
`
`CONH2
`
`21
`
`la) R=Bu‘
`(b) R=PhCH2
`
`22
`
`(30.0
`
`OAc
`23
`
`e
`CO2 E‘l3N(CH2)4('3H CO2ifl
`NH2
`
`+
`
`NHCOCH3
`
`OAc
`
`24
`
`A number of anhydride derivatives of aspirin have also been considered. Aspirin
`anhydride (25) has-low water solubility, and its further development was precluded
`on account of immunogenic problems [43,44]. Mixed anhydrides cannot be syn-
`thesised in pure form since some disproportionation always occurs and aspirin
`anhydride was formed as an impurity. Attempts to prepare a phosphoric acid—mixed
`anhydride (26) have been unsuccessful, even when stable dibenzyl esters (27) were
`used as intermediates, due to facile intramolecular reactions leading to the stable
`
`V
`cyclic salicylphosphate (28)..
`Several ester derivatives of the newer non—steroidal anti-inflammatory drugs have
`
`[29] have investigated the toxicological and phar-
`been described. Cioli et al.
`macological profile of ibuprofen guiacol ester (AF, 2259) (29) (Table 2). The gastro-
`intestinal toxicity, behavioural disorders and acute toxicity of AF, 2259 are reduced in
`comparison to ibuprofen. Both products, at equimolar doses, are equally active on
`oedema and fever. Ibuprofen guiacol ester (29) is better tolerated than its parent
`drug because of its peculiar pharmacokinetics; the ester releases ibuprofen slowly,
`thus reducing its local and general toxicity.
`Triglyceride derivatives of naproxen (30a,b) [34, 45] and indomethacin (3la,b)
`[34, 45] together with those of a number of putative anti—inflammatory heterocyclic
`acids [34] have been prepared. The l—glyceride (30a) and the 2—glyceride (30b) of
`naproxen have been evaluated for gastric irritancy in the rat (Table 2). The gastro-
`
`
`
`
`
`
`
`CH3 c o
`3
`
`18 (a) R=H
`
`(b) R=0H
`
`0
`
`0/\/NHC0
`
`\\N
`
`20
`
`OAc
`
`
`
`
`17 (a) x=s
`(b) x=sg
`
`(c) x=so2
`
`0 CH3
`CO CH 0Et
`
`II
` :
`
`19
`
`I
`
`
`
`
`Petitioner Mylan Pharmaceuticals Inc. - Exhibit 1012 - Page 111
`
`

`
`210
`
`25
`
`O
`
`O
`
`0
`
`OAc
`
`AC0
`
`0
`1); _
`0
`
`O
`
`OAc
`
`6
`
`2
`
`
`
`211
`
`
`
`
`
`in therapeutic index ,
`domethacin (31a,b) showed a 2.5 — 3-fold improvement
`(ulcerogenic dose causing lesions in 50% of the animals/anti—inflammatory dose to
`give 50% inhibition in rats) (Table 2). This result may be contrasted to the more than
`80-fold improvement in therapeutic index observed for the aspirin triglycerides
`(15a,b). The authors [45] conclude that these results indicate that gastrointestinal
`damage due to indomethacin is produced mainly by systemically circulating drug,
`whereas damage caused by aspirin is due primarily to local action of the drug on the
`gastric mucosa.
`
`The pharmacology and metabolism of a novel kind of anti—inflammatory agent,
`fenbufen (32a), was described recently [47]. In man, fenbufen was metabolised
`mainly into 4-hydroxy~(l,1—biphenyl)-4-butanoic aid (32b) and biphenyl acetic acid
`(32c), the active metabolite. Fenbufen was shown to have anti—inflammatory and
`analgesic potency in animal tests comparable with aspirin and phenylbutazone. The
`metabolite biphenyl acetic acid (32c) was considerably more active as an anti-
`inflammatory and analgesic agent. When fenbufen and its metabolites were given
`subcutaneously or orally to rats in equimolar doses, it was found that biphenyl
`acetic acid was considerably more ulcerogenic than the parent (32a) or its other
`metabolite, 4—hydroxy-(1,l—biphenyl)-4-butanoic acid (32b). When fenbufen (32a)
`was administered orally, and since biphenyl—4-acetic acid is formed by metabolism
`
`of the prodrug in the liver, local high concentrations of biphenyl—4-acetic acid in the
`stomach would not be anticipated, and thus ulceration was minimal. In terms of
`safety, as measured by the ratio of efficacy to the dose causing ulceration in the
`gastrointestinal
`tract,
`fenbufen was claimed to have the widest safety margin
`amongst the non—steroidal anti—inflammatory agents investigated.
`Sulindac (34) may be considered an isostere of indomethacin. From extensive
`structure—activity relationships among large numbers of indene and indole acetic
`acids, Shen [48] concluded that replacement of the indole nucleus with indene
`reduces CNS side—effects. The only biotransformations undergone by sulindac are
`
`changes in the oxidation state of the sulphinyl group, namely, oxidation of the
`parent sulphoxide group to the sulphone (35) or its reduction to sulphide (33) [49].
`The formation of the sulphone metabolite (35) was irreversible, and this compound
`elicited no significant biological response. The formation of the sulphide metabolite
`(33), on the other hand, is readily reversible in five animal species. The reversible
`interconversion of the prodrug sulindac to the active anti—inflammatory species, the
`sulphide (33), provides a theoretical basis for the long plasma half—life of the active
`drug and, in animal species, for the favourable gastrointestinal tolerance observed.
`The biological profile (Table 2) observed for the prodrug sulindac can be rationalised
`by the proposal that the sulphide (33), which is oxidised to sulindac in the course of
`scavenging oxidising radicals generated in inflammatory lesions, may be efficiently
`regenerated, for further scavenging, in situ.
`Suxibuzone (36a), a prodrug of phenylbutazone (36c), exhibits extremely low
`ulcerogenicity [50], although its anti—inflammatory, analgesic and antipyretic pro-
`
`Petitioner Amerigen Pharmaceuticals Ltd.
`
`_ Exhibit 1012 _ page 112
`
`
`
`
`
`
`0 I
`
`%0
`
`/P
`\O_
`
`,9/ocH2Ph
`P
`0/ \ocH2PH
`
`OAc
`
`27
`
`.
`
`28
`
`0
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`intestinal damage was assessed by measuring the rise in an acute—phase protein, Oll-
`acid glycoprotein, by an immunological assay of the serum. The concentration of
`oz1—acid glycoprotein has been shown to correlate well with the incidence and severi-
`ty of gastric lesions, as determined by the usual visual method [46]. At equimolar
`doses, the 2-glyceride (30b) produced no significant change in the basal concentra-
`
`tion of ozl-acid glycoprotein, while the 1—glyceride (30a) effected a small rise in the
`concentration of
`this protein. Comparison of
`the 2-glyceride (30b) with
`2—(6—methoxy-2-naphthyl)-propionic acid for gastric irritation, as determined by the
`minimum chronic dose producing occult blood in either the faeces or urine in dog,
`gave a dose ratio in favour of the 2-glyceride of at least 3. The 2—triglycerides of in-
`
`
`
`
`
`
`
`
`
`
`
`
`OCOR3
`
`R2000-[ocon1
`
`(MR1:
`
`M90
`
`R =R =c H
`
`3
`
`1735
`
`2
`
`(b)R-1= R3 = CWH35"
`
`3°
`
`R2:
`
`Meo
`
`
`
`"
`
`
`
`R 3
`
`2
`
`(a) R = cocH2cH2co2H
`(b) R = CHOHCHZCHZCOZH
`(c) R = CH2CO2H
`
`C‘
`
`
`
`
`
`31 (a) R = CH3
`(b) R = cgnmn
`
`
`
`Petitioner Mylan Pharmaceuticals Inc. - Exhibit 1012 - Page 112
`
`

`
`
`
`E;
`
`
`
`
`
`212
`
`perties remain intact when compared at equimolar doses with the parent drug.
`Metabolic studies [51] indicate that suxibuzone was hydrolysed by esterases to give
`4-hydroxymethylphenylbutazone (36b), which is considered to be labile [52] and to
`decompose spontaneously to phenylbutazone (36c) unless otherwise conjugated with
`glucuronic acid at the oxymethyl side-chain. Suxibuzone is considered to be less
`
`ulcerogenic than phenylbutazone (Table 2) because any suxibuzone remaining in the
`
`gastrointestinal tract is less harmful than phenylbutazone, since unchanged sux-
`ibuzone does not inhibit the function of either mitochondria or prostaglandin syn-
`thetase.
`
`F
`
`cH2co2H
`
`F
`
`cH2co2H
`
`. CH3
`
`? CH3
`
`33
`
`Mes
`
`F
`
`CH3S
`g
`
`34
`
`H co H
`C 2
`2
`
`Ph
`
`Ph
`
`
`
`
`
`
`0
`\\
`3
`\\
`0
`
`CH
`
`3
`
`U
`1
`
`35
`
`.
`
`\N
`
`Bu"
`
`N’
`
`R
`
`36 (a) R = CH2OC0(CH2)CO2H
`(b) R = CHZOH
`(
`) R = H
`
`°
`
`L
`
`37
`
`(a) R = P032"
`(b) R = H
`
`
`
`C”
`
`3
`
`
`
`
`
`
`
`
`
`
`Piroxafos (37a) [53], the phosphoric ester of piroxicam (37b), was shown to have
`anti—inflammatory, antigranuloma and analgesic activity and was considered to be
`a prodrug of the parent anti-inflammatory agent.
`In comparison with in-
`domethacin, aspirin and phenylbutazone, piroxafos was less active as an anti~
`inflammatory agent than indomethacin, but equiactive with the other two agents.
`Piroxafos was tolerated significantly better by the gastrointestinal tract than were
`the comparison drugs.
`
`
`
`
`
`TABLE 2
`
`Biological Activity of Prodrugs of the Non-steroidal Anti—inflammatory Agents
`
`213
`
`Anti-inflammatory
`activity, carrageenin
`test EDX (pmol/kg)
`
`Ulcerogenic
`(UD50/;tmol/ kg)
`
`Therapeutic index
`(UD50/EDX)
`
`Reference
`
`
`
`
`
`Compound
`number
`
`Aspirin
`
`15a
`15b
`16
`Ibuprofen
`
`29
`30a
`30b
`Indomethacin
`31a
`31b
`32a
`
`32c
`
`33
`
`34
`
`36a
`36c
`37a
`37b
`
`498
`1016
`368
`190
`402
`459
`12
`12
`21
`121
`5
`14
`13
`11.7
`
`ED25 =
`ED” =
`ED25 =
`ED25 =
`ED25 =
`ED30 =
`ED” =
`ED30 =
`ED50 =
`ED5o =
`ED50 =
`ED25 =
`ED25 =
`ED30 = >
`(umol/paw)
`ED30 =
`(umol/paw)
`ED50 =
`
`,ED50 =
`0'
`
`ED30 =
`ED3o =
`ED50 =
`EDSO =
`
`84
`1278
`
`>2562
`>l690
`194
`12
`30
`
`—
`—
`
`10
`67
`78
`17.7
`
`0.17
`0.795
`
`13.5
`4.20
`0.42
`1
`2.5
`~
`4-5 X Naproxen
`2
`5
`6
`1.51
`
`0.46
`
`~
`
`6
`
`15
`
`228
`162
`8.5
`9.1
`
`20 (gastric)
`103 (intestinal)
`76 (gastric)
`194 (intestinal)
`1826
`272
`29
`30.2
`
`'
`
`~
`
`3.3
`17.2
`12.9
`
`8.0
`1.7
`3.4
`3.3
`
`35
`47
`35
`35
`32
`47
`29
`29
`34
`34
`47
`45, 47
`45, 47
`47
`
`47
`
`47
`
`47
`
`51a
`51a
`53
`53
`
`4. Steroidal anti-inflammatory agents with reduced local and systemic side-effects
`
`in corticoid prodrugs is evident, undoubtedly
`A recent resurgence of interest
`stimulated by the adverse side—,effects encountered through excessive application of
`topical fluorinated steroid formulations. The main therapy-limiting side—effects
`following topical administration have been noted by Schopf [54], and include der-
`mal atrophy, thymus involution and suppression of the adrenal—pituitary functions.
`In an attempt to reduce the level of systemic side—effects, Bodor and Sloan [55] have
`investigated a series of hydrocortisone derivatives (Table 3) incorporating biphasic
`solubilising groups such as 21-dialkylsuccinamate. The use of amidocarboxylic acid
`esters of steroids was shown to enhance the rate of delivery of hydrocortisone
`
`through the skin for compound 39a and resulted in reduced local toxicity but com-
`parable systemic toxicity to the parent drug hydrocortisone (38a). A mixed ester
`
`
`
`Petitioner Amerigen Pharmaceuticals Ltd.
`
`— Exhibit 1012 — Page 113
`
`Petitioner Mylan Pharmaceuticals Inc. - Exhibit 1012 - Page 113
`
`

`
`215
`
`3-spirothiazolidines [56] of hydrocortisone (40a,b) which readily revert to their
`
`parent steroidal ketone have been investigated (Table 3). The anti—inflammatory ac-
`tivity of the thiazolidines (41a,b) were determined in the mouse croton oil irritation
`model and in the human blanching model. All the thiazolidines of hydrocortisone-
`2l—acetate, especially compounds 41a and 41b, were more active than their parent
`compound, hydrocortisone—21-acetate, and there was no clear correlation between
`activity and stability of the thiazolidine derivative. In general, thiazolidines derived
`from hydrocortisone—l7—esters were less active than the parent ester. The prodrugs
`41a and 41b were less toxic systemically and this appears, at least in the case of com-
`pound 41a, to be due to the fact that less hydrocortisone was delivered through the
`skin from this masked form than from hydrocortisone—21—acetate. The therapeutic
`
`efficacy of these prodrugs, when measured in terms of their anti—inflammatory ac-
`tivity opposite their local propensity to atrophy the skin, was not enhanced.
`
`41
`
`(a) x = H R = COCH3
`(b) x = C02Et R = C0C4H9"
`
`"0
`
`on
`O
`o“°“
`
`Sun”
`
`X
`
`OR
`
`0
`\\\OH
`
`42
`
`ta) R = COCH3
`(b) R = c°C15”31
`
`n
`
`HO
`
`0
`
`.
`.
`.
`.
`A series of 21-esters of hydrocortisone (42a,b) were synthesised and incorporated
`into liposomes [34] prepared from dipalmitoyl lecithin and egg phosphatidic acid.
`The degree of incorporation of the steroid esters into liposomes was measured using
`radioactively labelled steroid samples, and almost complete incorporation (98%)
`was observed for hydrocortisone-21~palmitate (42b). Fortuitously, also the rate of
`release of steroid from the liposome into phosphate—buffered saline at neutral pH
`was least for the hydrocortisone—2l—palmitate liposome. The release of drug from
`prodrug in the presence of sonicated rat macrophages occurred at a significant rate,
`with a half—life of the order of 50 hours. The liposome—encapsulated steroid prodrug
`
`Petitioner Amerigen Pharmaceuticals Ltd.
`
`_ Exhibit 1012 _ page 114
`
`
`
`214
`
`
`
`
`
`from the same series exhibited less systemic toxicity than its parent,
`(39b)
`hydrocortisone—17—valerate (39a), whilst retaining a portion of the local activity of
`the parent drug. The systemic toxicity was assessed by measuring the thymus weight
`of weanling rats, whilst the local toxicity was determined using hairless mice and
`measuring the double-fold skin thickness. The topical anti—inflammatory effect was
`assessed using the croton oil mouse ear
`irritation test. A further series of
`
`
`
`OH
`\\%R
`\\
`
`°°°‘°”2l2°°NE‘2
`\\\\°0R
`
`O
`
`38
`
`(a)R=H
`=
`lb) R
`COC4H9
`
`n
`
`O
`CA6
`0
`
`39 <a>R=H
`b R = coc H “
`l
`)
`4 9
`
`
`
`
`
`
`(b) X = H
`
`40
`
`(a) X = COZEI
`
`TABLE 3
`
`Toxicity and Anti—inflammatory Potencies of Steroidal Prodrugs
`M
`
`-
`
`
`
`
`
`
`Compound
`
`"/0 Reduction in
`thymus weight
`(systemic toxicity)
`
`Skin thickness in
`XLO [4]
`«_L S.E.
`(local toxicity)
`
`Croton oil irritation
`ED50 (M)
`
`
`
`
`
`
`
`
`
`
`
`
`
`38a
`39a
`38b
`3%
`40a
`40b
`Hydr0C0rtiS0n€-
`21-acetate
`
`36
`39
`21
`16,,
`22a
`11“
`
`79 i 2
`94 i 3'3
`88 i 5°
`_
`88 : 6
`37 i 5
`
`0.0084
`—
`“
`0.0019
`0.0033
`0,0055
`
`36
`
`: 7
`
`90
`
`0.016
`
`
`
`
`a. P < 0.01 compared to la.
`b. P < 0.05 compared to la.
`c. Data from Ref. 56.
`
`d. P < 0.001 compared to hydrocortisone-21—acetate.
`
`
`
`
`
`
`
`
`1
`
`
`
`
`
`Petitioner Mylan Pharmaceuticals Inc. - Exhibit 1012 - Page 114
`
`

`
`
`
`216
`
`217
`
`was evaluated in an acute experimental arthritis in the rabbit. Intra—articular treat-
`ment of this experimental arthritis with a liposome preparation containing cortisol
`palmitate resulted in a significant and sustained fall in the temperature and the
`diameter of the affected joint. This reduction in both these parameters occurred
`within 24 hours and was maintained for at least 3 days. Clinical evaluation of this
`liposome therapy in patients with classical reheumatoid arthritis was carried out.
`The liposome preparation of the prodrug realised an improvement in both the sub-
`jective and objective indices of joint inflammation. Post-injection pain was not ex-
`perienced by any patient, nor were any adverse systemic side—effects noted. This
`combination of liposome and prodrug technology significantly improved the ef-
`ficacy of intra-articular steroid therapy. This was achieved by a steroid—sparing ef-
`fect, by direct injection of the prodrug complex locally to the inflamed joint, and
`by targeting the liposomes through phagocytosis to the appropriate cell line.
`
`5. Prodrugs of anticancer agents
`
`5.1.
`
`PEPTIDYL-LATENTIATED FORMS OF THE ANTHRACYCLINES AND RELATED
`ANTINEOPLASTIC AGENTS WITH REDUCED CARDIAC TOXICITY
`
`The practical therapeutic use of daunorubicin (43a) and doxorubicin (43b), two an-
`tineoplastic drugs of the anthracycline family [57], is limited by the cardiotoxic ef-
`fects produced by these agents. The cardiotoxic effects [58] can be subdividedlinto
`acute, subacute and chronic, depending upon their temporal relationship to the ad-
`ministration of the drugs. The acute effects consist of hypotension, tachycardia and
`arrhythmias, and develop within minutes after intravenous infusion of the drug.
`
`0
`
`0
`
`43 (a) R = H daunorubicin
`(b) R = OH doxorubicin
`
`NH2
`
`(a) X = Leu
`(b) X = Ala
`(c) X = A|a—Leu
`(d) X = Leu-Leu
`(e) X = Lys
`(f) X = Arg
`(9) X = 0m
`(h) X = 2,4 diaminobutyric acid
`(i) X = D-Lys
`
`The subacute effects are characterised by fibrinous pericarditis or myocardial
`dysfunction, and occur within 4 weeks of the first or second dose of the drug. The
`chronic effects become evident only after several weeks or months of treatment and
`are manifested by the insidious onset of severe, often fatal, congestive heart failure.
`Peptide prodrugs of numerous anticancer drugs have been investigated in an at-
`tempt to increase the selective toxicity towards the malignant cells. Baurain and co-
`workers [59]
`(see Table 4) have prepared and reported some promising in vivo
`
`TABLE 4
`
`Biological Activity and Toxicity Data for Peptidyl Derivatives of Daunorubicin
`
`Compound
`number
`
`Percentage increase Medium survival
`in lifespan (dose)
`time in days (dosing
`on subcutaneously
`schedule) against EL4
`implanted L1210
`lymphoma
`leukaemia
`
`No. of toxic
`deaths in EL4
`test
`
`Weight change
`(0/0) on day 8 in
`L1210 test
`
`43a, dauno-
`rubicin (DNR)
`
`44a, Leu-DNR
`
`65 (10 mg/kg/day)
`67 (11 mg/kg/day)
`83 (12 mg/kg/day)
`166 (40 mg/kg/day)
`215 (44 mg/kg/day)
`225 (46 mg/kg/day)
`
`17 ( 80 pg/mouse, X 1)
`
`8
`
`6
`15 ( 20 [Lg/IHOUSC, X6)
`19 (100 p.g/mouse, X6) 0
`
`15 (100 p.g/mouse, X8) 4
`
`-4.3
`-5.9
`-6.3
`-1.3
`-3.3
`-8.5
`
`- 5.9
`-8.5
`
`-5.0
`-8.2
`-23.4
`
`+0.4
`-2.8
`-7.4
`
`37](60 mg/kg/day)
`61 (66 mg/kg/day) —
`
`168 (53 mg/kg/day) —
`193 (60 mg/kg/day)
`73 (66 mg/kg/day)
`
`65 (53 mg/kg/day) —
`210 (60 mg/kg/day)
`153 (66 mg/kg/day)
`
`44b, Ala-DNR
`
`44c, Ala-Leu
`DNR
`
`44d, Leu—Leu—
`DNR
`
`44e, Lys-DNR
`
`44f, Arg-DNR
`
`44g, Orn~DNR
`
`-
`
`—
`
`—
`
`26 (250 pg/mouse, X8)
`
`2
`
`35 (250 pig/mouse, X7)’ 0
`19 (400 pg/mouse, X 6)
`1
`
`30 (300 pg/mouse, X 8) O
`11 (600 pg/mouse, X4)
`4
`
`30 (650 ug/mouse, X8)
`
`0
`
`14 (300 pg/mouse, X 8) 0
`12 (600 fig/mouse, X4)
`4
`
`44h, Azbu-DNR -
`
`441, D-Lys-DNR -
`
`45, D-Va1—Leu-
`Lys-DNR
`
`13 (60 mg/kg, ><3) —
`21 (75 mg/kg, ><3)
`against B16
`melanoma
`
`Petitioner Amerigen Pharmaceuticals Ltd.
`
`- Exhibit 1012 - Page 115
`
`Petitioner Mylan Pharmaceuticals Inc. - Exhibit 1012 - Page 115
`
`

`
`218
`
`results [60] on amino acid and dipeptide derivatives of daunorubicin (43a), and
`singled out the L—1eucyl daunorubicin (44a) as being four times less toxic than
`daunorubicin, whilst being equipotent with the parent when tested against Ll210
`leukaemia in mice. An improved therapeutic index was thus claimed for the prodrug
`(44a). Sela and Levin [61] (see Table 4) have also reported on over 30 amino acid
`and peptide derivatives of daunorubicin. These authors also confirm that the amino
`acid and di— or tripeptide derivatives of daunorubicin (44a~ i) were uniformly less
`toxic than the parent. In general, the hydrophobic amino acids, such as alanine and
`leucine, produced more toxic derivatives (44a,b) than did the more hydrophilic,
`basic and acidic amino acids. The reduced toxicity observed for a given derivative
`could not always be correlated with improved therapeutic efficacy. The aspartyl and
`prolyl daunorubicin were found to be nontoxic and poorly effective as antitumour
`agents. The most promising therapeutic effect, in terms of both antitumour activity
`and toxicity, was found in a group consisting of the basic amino acids lysine (44e),
`arginine (44f), ornithine (44g), and, in particular, the 2,4—diaminobutyric acid (44h)
`derivative; all of these derivatives (44a— h) being superior to the parent molecule.
`These variations in toxicity and therapeutic index of the amino acid derivatives were
`attributed to several parameters, namely, differences in tissue distribution, different
`rates of cellular uptake and different patterns of intracellular localisation. Evidence
`was presented demonstrating that these derivatives acted as ‘slow release’ forms of
`the parent drug, where the activation procedure was enzymically mediated. Thus,
`D—lysyldaunorubicin (44i) exhibits a very poor curative effect on L121O leukaemic
`mice,
`in marked contrast to L—lysyldaunorubicin (44e),
`its enantiomer. In most
`cases, these drugs do not have blocked amino termini so, once again, liberation of
`the parent drug by non—tumour—specific aminopeptidases seems most likely. The im-
`provement seen in terms of selectivity with these prodrugs may well be due to
`changes in the overall pharmacokinetics of the drugs rather than to any particular
`tumour—specific drug targeting.
`More recently, there has been a resurgence of interest in designing potential site-
`specific anticancer prodrugs based on the rationale that tumours that contain a high
`level of some specific enzyme might convert the prodrug to the pharmacologically
`active drug in the vicinity of the tumour, resulting in lower drug concentrations at
`sites of limiting toxicity, such as the heart in the case of the anthracyclines. Many
`transformed cells display a high level of proteolytic activity [62] due to the presence
`of membrane-bound serine proteases, plasminogen activators. Although plasmi-
`nogen activators are found in normal cells, substantially increased levels are found
`in both virally and chemically transformed cells. Katzenellenbogen et al. [63] sug-
`gested that the plasmin activity associated with tumours producing elevated levels
`of plasminogen activator presented an opportunity for designing such site—specific
`cytotoxic agents. Acivicin (46a) and N,N-bis(2-chloroethyl)—p—phenylenediamine
`(45a) were derivatised with the tripeptide D—Val-Leu—Lys to the prodrugs 46b and
`45b. The conversion of these agents to latentiated forms rendered them 5 ~— 10 times
`
`2l9
`
`more toxic for malignant chick embryo fibroblasts than for the corresponding nor-
`mal cells. This increase in cytoxicity of the latentiated forms for the malignant cells
`was not observed in the presence of a known plasmin inhibitor, p-nitrophenyl—p-
`guanidino benzoate. A prodrug derivative in which the L-lysine residue of the pep~
`tide moiety was replaced by D—lysine, a substitution known to block the activation
`of the prodrug by plasmin, yielded a compound (450) with no increase in selective
`toxicity towards transformed cells.
`In the absence of plasminogen,
`the pep-
`tidylacivicin (46b)
`showed greatly decreased potency and no selectivity‘ for
`transformed cells. The prodrugs were thus considered to be selectively toxic to
`transformed cells because they were specifically activated by the plasmin in the
`transformed cell cultures.
`
`
`
`x H N
`
`CHZCHZCI
`
`N<cH2cH_,_c|
`
`45
`
`(a) X = H
`(b) X = D—Va|—Leu—Lys
`(c) X = D——Val Leu—D—Lys
`
`HOZC CH NHX
`
`Cl
`
`46 (3) X = H
`(b) X = D—Va|-—Leu——Lys
`
`O
`
`47
`
`CH2
`
`H0
`
`D——Va|——Leu—Lys——HN
`
`The peptidylphenylenediamine mustard (45b) was tested in vivo against B16
`mouse melanoma, a transplantable tumour known to be rich in plasminogen ac-
`tivator activity and also sensitive to the parent phenylenediamine mustard (45a), and
`shown to be clearly active; however, it failed to produce an improved malignant cell
`selectivity when compared to the parent drug. The prodrugs were thus demonstrated
`to possess improved malignant cell selective toxicity in vitro but a similar selectivity
`was not observed in vivo, even though antitumour activity was seen for both pro-
`drug and drug in vivo. Two main reasons were considered for the comparative
`failure of the in vivo experiment, either the presence in plasma of inhibitors of pro-
`drug activation, diluted plasma (8070 activation in 1 hour) was a more effective ac-
`tivator of prodrug than was concentrated plasma (4% activation in 1 hour), or the
`
`comparative hydrolytic instability of the chloroethyl moiety in the prodrug (IV: =
`30 minutes at 37°C) giving noncytotoxic hydrolytic products. Further difficulties
`
`Petitioner Wan Pharma°ei"
`
`‘I
`
`Petitioner Mylan Pharmaceuticals Inc. - Exhibit 1012 - Page 116
`
`

`
`220
`
`221
`
`were noted concerning the problem of effectively testing a prodrug opposite a drug,
`particularly when both compounds are injected in the vicinity of the tumour, and
`the non-specific activation of prodrug at extra—tumour sites.
`A similar study was carried out on peptidyl prodrugs derived from the clinically
`important drug, doxorubicin (43b), and almost identical results were obtained. The
`peptide derivative of doxorubicin (47) was shown to be more malignant cell selective
`in vitro than the parent drug, and that in vivo it was far less toxic. The poor potency
`of the prodrug prevented any satisfactory exploration in vivo, aimed at determining
`if the in vitro therapeutic gains observed could be realised in whole animals.
`
`5.2.
`
`PRODRUGS OF 5-FLUOROURACIL WITH REDUCED HOST TOXICITY
`
`Novel prodrug derivatives of 5-fluorouracil (48) possessing a broader spectrum of
`antitumour activity, and in particular fewer side—effects, have been sought diligently
`in a number of laboratories in recent years. The clinical usefulness of 5-fluorouracil
`
`in the treatment of solid tumours has been restricted, and its efficacy limited, by
`the severe side-effects encountered such as disorders of the bone marrow or the
`
`epithelium of the gastrointestinal tract. The search for less toxic prodrugs of
`5-fluorouracil was stimulated further by the discovery in 1967 of l—(2—tetrahydrofu