Polyarylates containing derivatives of the natural amino acid l-tyrosine

~ 2~0~5~ --1- DF~cE2TpTToN POLYARYLATES CONTAINING n~rvATIv~ OF T~F NATuR~r AMrNO ACID n-TYROSINE TT'~I~A~. FT~.n The present invention relates to amino acidderived bioerodible polymers and to methods of synthe~izing such polymer~.

R~l ~KOUN u ~T Polyesters in qeneral are highly useful and intensively ~tudied mat-rial~.

Polyesters have been cla~ified according to the nature of the polyhydric comFo-~n~ and dicarboxylic acids from which the polymers are derived.

Polyesters derived from aromatic polyhydric com~o~ln~ are classified as polyarylates, with polyarylate~ of aliphatic diacids classified as aliphatic polyarylates and polyarylates of aromatic diacid~ cla~sified as aromatic polyarylates.

Polyesters derived from aliphatic polyhydric compounds and aliphatic diacids are clas~ified a~ aliphatic polyesters, and polyesters derived from aliphatic polyhydric compo~A~ and aromatic diacids are classified as aliphatic-aromatic polyesters.

See Imai et al., ~Synthesis and Characterization of Aromatic Polyesters,~ Ha~hook of Polv~er Science And T~rhnoloav, Vol. 1, N.

Cheremisinoff, ed., Marcel ~c~er, Inc., New York and Basel, 1989), 177-208.

A wide range of aliphatic polyesters (e.g., poly(lactic acid), poly(glycolic acid), polydioxanone, poly(hyd.G~yL~y ~te), polycapro-lactone and the like) have been explored as biomaterials, e~pecially in the context of degradable medical implants because of their hydrolytic instability.

See, Huang, ~Biodegradable Polyme~s,~ ~ncYclor~ of poly~er Science An~ TP~hnnlooy~ (F.H.

Mark et al., ed., John Wiley and Sons, New York 1985), vol. 2, 220-43.

Becau~e of their hydrolytic instability, h~ ~ver, aliphatic polyesters are not used as industrial engineering plastics.

21402~S _ '_ -2 Aliphatic-aromatic polyester~ include poly(ethylene terephthalate) (PET) and poly(butylene terephthalate) (PBT). Some aliphatic-aromatic polyesters are very important commercial plastics, but S (with the exception of DACRON ) have not been widely investigated in the biomaterials area. DACRON- i8 a PET u~ed ag an artificial VAr~ r graft. See, Silver et al., Biocom~tihilitY: Tnt~r~ct1on~ of Rioloaical and ~ n~hle M~t~ri~ls, vol. 1, (VCH Publishers, New 10York, 1989) 240-3.

Aliphatic-aromatic polyesters are hydrolytically much more stable than aliphatic polyesters and are degraded under physiological conditions at such a ~low rate that they are regarded as ~nondegradable.' 15Aromatic polyarylateQ have recently gained importance as liquid crystalline polymers. These polymers are derived from aromatic dicarboxylic acids and diphenols (bisphenols). Many aromatic polyarylates ~c~se-- highly desirable ~ e.~ie~ for use as engineering plastics, such as formation of ~L~GIIY films and fibers which are heat, solvent and flammability re6i~tant.

In addition, some aromatic polyarylates are electric in~ulators and exhibit liquid crystalline behavior. For ex_mple, Ca~G~-dum marketed poly(phydroxybenzoic acid) under the trade name of ~K~NOL in 1970. Shortly thereafter, ICI, T~ sF~e Eastman, Celanese and DuPont marketed polyarylates derived from iso- and t~ h~lic acid~ and bieph~nols such as hy~ none, sub~tituted h~dL~inone~ _nd Bisphenol A.

See, ~ AiAnA~ 'Polyesters, Main Chain Aromatic,' ~nCyc70p~r~;~ of po7~r Sc;~nce ~nA ~ ;n~r;rlg~ (Mark et al., ed., John Wiley and Sons, New York, 1985) vol. 2, 264-6. The major di~advantage of the~e early polyarylates was that they were difficult to ~L~e_G becau~e of their high melting po_nts. More Le_a..Lly, n-w polyarylates that are ~ore ea~ily ~ hle have 2140~5 -3 been obtained through modification of the monomers, copolymerization and ble~inq.

Aliphatic polyarylates are a fourth class of polyester and are derived from aromatic diols and aliphatic diacids.

Examples of aliphatic polyarylates include poly(hyd~G~inone -~ipate) and poly(Bi~p~Pnol Ar~~ac~te).

These polymer~ have le~8 commercial appeal bQcause they do not gen-rally pos~es6 the high thermal stabili~y of the aromatic polyarylates.

The synthesis of aliphatic polyarylates from bi~pbe~ols and aliphatic dicarboxylic acid~ wa~ first ~G~ Led by British Patent Nos. 621,102 and 636,429.

In the early work, hydroquinone and aliphatic dicarboxylic acids were used to develop aliphatic polyarylates of high melting points - 15 and hydrolytic ~ta~ility.

See, Earec~-~n, J. po Sci., 40, 399-406 (1959), E~OLO~ et al., VYcoko~olekul.

Soedin., ~, 1475 (1960) and Mikhailov et al., ~him Volokna, 1963~2~, 19-22 (1963).

Particularly important are the studies of Morgan, J. poly~.

Sci.~ pArt A2, 437-59 (1964) who synthesized a ~eries of ~liphatic polyarylates using either phenolphthalein or Bi-r~nQl A ~c the diphenol components.

By varying the length of the diacid, Morgan established the fir-ct 6ystematic stru~L~-f ~roperty correlations for this clas~ of polymers.

Morgan noted that all the polymers prepared were amorphous, soluble in a variety of organic solvents, and had relatively low soft~nin~ temperatures.

Thus, these materials were not the high t-mperature engineering plastics sought by industry.

Many other aliphatic polyarylates were later ~I.Lh~sized, for example, fro~ resorcinol, dihyd,o~yl.apthalenes, bis(4-hyd~oxy~l-enyl) alkanes and phthaleins.

See, Morgan, Co~encation poly~ers: ~y ~nterf~ial ~n~ Solution Me~h~c, (Interscience Publisher~, N-w York, 1965) 334-41.

Commonly owned U.S.

Patent No. 5,099,060 discloses amino acid-derived ~ir~enol compo~n~C, the chemical struc~u~a 5 of which were designed to be ~ ~ 2140255 _ -4 particularly useful in the polymerization of polycarbonates and polyimiro~Arbonates.

The resulting polymers are useful a~ degradable polymers in general, and as tissue compatible bioerodible materials -for biomedical u~es, in particular.

The suitability of these polymers for this end-use application is the result of their derivation from naturally occurring amino acids.

In part~ Ar~ the polyiminocArbonates and polycarbonates are polymerized from L-tyrosine derived ~irhsrl~ls.

m e L-tyrosine derived polyiminoc~rbonates are rapidly b;oAo~radable, while their polycarbonate counterparts degrade very elowly.

This patent suggests that when a material having a moderate degree of bioerodibility i~ required, that the two polymers be blended to achieve the required ratc of biodegradation.

While thi~ adeguately add~ ~c this requirement, a need remains for biodegradable polymers having a moderate rate of hydrolytic degradation.

Aliphatic polyarylate type polyesters would be expected to have a moderate rate of biodegradation.

Such com~o~nA- could be prepared from the L-tyrosine derived ~ip~n~l~ used in the preparation of the polycarbonate~ and polyiminocarbonates disclosed by U.S.

P~tent No. 5,099,060.

However, unlike the polycarbonates and polyiminoc~rbonates~ which are polymerized by linkin~ biocompatible amino acid-derived diphenols by way of non-amide linkages to provide a nontoxic polymer that form~ nontoxic degradation products, the polymerization of polye~ters requires the reaction of a diacid with the A i phenol .

Diacids are therefore also required that will produce a nontoxic polyme-r that for~ nontoxic-degradation products.

There is no disclosure in the prior art that tyrosine-derived ~irh~nQls specifically designed for the polymerization of meçhAnically ~L~on~ polyiminocarbonates and polycarbonates would also form polyarylates with favorable engineering properties when 21~02~5 _ -5 reacted with variou~ aliphatic and aromatic dicarboxylic acids. Furthermore, the synthe~is of polyarylates using the amino acid-derived diphenols of U.S. Patent No.

5,099,060 po~e~ a challenge becau e of the relative S inertn-~s of the~e ~ ~ rh~ol 6 in polye~terifications and beca~-e of the pro~q -e of a chemically ~ensitive aliphatic e~ter bond in the monomer ~tructure. Contrary to the formation of aliphatic polyesters in which the thermodynamic equilibrium favors e~terification, the synthesis of polyarylate~ requires the activation of the diacid. Traditional methods of synthecis utilize diacid chloride derivatives as the active intermediates in either ~olution or interfacial polymerization reactions.

Tran~esterificationc at elevated temperature~ and direct lS polycon~?P-~tion~ using a variety of ~onA~nQing agents have al~o been explored. See, I~ai et al., ~Synthesis and Characterization of Aromatic Polyesters,~ ~AnAhnnk of PolY~er Science and T~rhnnlo~Y.

Aliphatic polyarylate~ of intermediate molecular weight (inherent viscosity about 0.52) were prepared by ~organ, J. polY~. Sci.. P~-t A~, 437-59 t 1964) under mild condition~ of low temperature and ~ re ~ through the diacid chloride and direct polyco~ene~tion routes. In the diacid chloride 25 interfacial reaction a ~-,v--~ly ba~ic aT~ solution of the bt~h-nQlate anion i~ reacted with a solution of a diacid chloride in a water-immiscible organic solvent such a~ dichloromethane. The reaction oc~, D at the interface and i~ accelerated by the u~e of a quaternary 30 ammonium or ~h~lphQnium ~alt or by a pha~e transfer catalyfit such a~ a crown eth r. S-e, Conix, Tn~. ~n~.

Chem., ~1, 147 (1959).

The diacid chloride reaction can also ~e done in an organic solvent and a ba~e ~uch a6 triethylamine i~ often used to neutralize liberated 35 h~dL~.. chloride. See, Morgan, J. PolYm. Sci..

P~rt ~7, 437~59 (1964).

The direct pol~. ~-r--tion route has significant advantages over the AiAci~ chloride 21~0255 '_ -6 t~chnigue in the laboratory preparation of aliphatic polyarylate6. The need for the unstable diacid chloride~ i~ eliminated, and the ~iphensl can have functional ~-ou~_ that would be incompatible with the S reactive AjAC~A chloride6 or the DLLv~ly basic agueous ~olutions ~ red in the interfacial ~h~igue. Direct polyconden6ation thereby facilitate~ the use of basésensitive ~ip~enols as monomers, such as the ~mino acidderived A, i phenQl6 diaclo6ed in U.S. Patent No.

5,099,060.

A variety of condensing agent~ have been used in the polymerization of aliphatic polyarylates, including triphenylphoEphin? (see, Ogata et al., Polym.

J., 13 (10), 989-91 (1981) and Yasuda et al., J.

Poly~ sci.: poly~. rhem.

Ed., ~, 2609-16 (1983) ), picryl chloride (~ee, Tanaka et al., polym. J., 14 ~8), 643-8 (1982) ) ~ phonrhorus oxychloride with lithium chloride monohyd,ate as a catalyst (6ee, Higashi et al., J.

Polym.

Sci.: Po~Y~. ~he~. F~ , 589 - 94 (1986) ), arylsulfonyl chlorides (see, Higa6hi et al., J. pol Sc-.: polY. ~h~. ~A., ~1, 3233 - 9 (1983) ) ~ A;F~?nY1 chloropho~phate (~ee, ui~ i et al., J. polY~.

Sci.:

Polym. ~hem.

Ed., ~, 3241-7 (1983) ), thionyl chloride with pyridine (see, Higashi, J.

PolY~.

Sci.: pol,Ym.

~. FA.~ ~ , 97-102 (1986) ), triethylamine a~ a base (see Elia~ et al., ~k~omol.

Chcm., 1~2, 681-6 (1981) ) and dii~v~v~ylcarbodiimide with the 6pecially designed cataly~t 4-(dimethylamino) pyridinium 4-toluenesulfonate (DPTS) (~ee, Moore et al., Macromol., 73 (1~, 65-70 (1990) ) .

Alt h~-~h numerou~ aliphatic polyarylate6 have thus been ~L~p~ed, they have ~ lly been derived from bioincompatible diacid or bisr~-nol components and would therefore not be expQcted to be ~uitable a8 degradable medical implant material~.

~ There remains a need for nontoxic polyarylates having a moderate rate of bioero~ion, ~uitable for u6e a~ ti~ e _omp~tible material~ for bio~edical u~e6.

--7- SUMMARY OF THE INVENTION It has now been discovered that amino acid-derived diphenols can be copolymerized with dicarboxylic acids by way of a carbodiimidemediated direct polycondensation to form nontoxic bioerodible polyarylates having numerous biomedical uses. In addition to being nontoxic in polymeric form, the polyarylates of the present invention are expected to form nontoxic degradation products. Furthermore, the polyarylates of the present invention degrade by hydrolytic chain cleavage under physiological conditions, which makes it possible to use the polymers as degradable, medical implant materials.

Therefore, according to one aspect of the present invention, there is provided an amino acid-derived polyarylate characterized by one or more recurring structural units represented by the formula:

~ CH2-CH-C-NH-CH-cH2 ~ O-C-R-Ce-o_ (I) wherein X is selected from the group consisting of -H, -NHL1, -NL1L2,:

O O -NH-C-Ll and -NH-C-OLl;

Y is hydrogen or an ester group having the structure -C- -L3 wherein L1, L2 and L3 are independently selected from the group consisting of straight or branched alkyl and alkylaryl groups containing up to 18 carbon atoms and a biologically or pharmacologically active ligand; and R ls selected from the group consisting of saturated and unsaturated, substituted and unsubstituted alkyl, aryl and alkylaryl groups containing up to 18 carbon atoms.

The polyarylates of the present invention meet the need for a moldable biodegradable biocompatible polymer derived from nontoxic monomeric starting materials.

The polymer is expected to hydrolyze to release the original diphenol and diacid, thus forming nontoxic degradation products, provided the monomeric starting materials are nontoxic.

The toxicological concerns associatedwith polyarylates are met by using diphenols derived from L-tyrosine dipeptides and dicarboxylic acids that are either important metabolites or highly biocompatible compounds. Therefore, another aspect of the present , - 1 -8- 7 ~ ~ o ~ 5 ~ invention provides molded articles prepared from the polyarylates of the present invention.

The polyarylates of the present invention include polymers having pendant side chains on each repeating unit. This structural feature represents a further degree of freedom in the design of polyarylates and can be used to modify the overall physicomechanical properties of the polymer without changing the polymer backbone structure. Detailed studies correlating the structure of pendant chains with the physicomechanical properties of the polymer without changing the polymerbackbone structure.

Detailedstudies correlating the structure of pendant chains with the physicomechanical properties of polyarylates have apparently never been performed.

In addition, a biologically or pharmacologically active agent can be chemically incorporated into the pendant X and Y side chains of the polymer. Alternatively, the pendant side chains can be used to crosslink the polymer chains to form a polymeric matrix into which a biologically or pharmacologically active material can be physically -imbedded or dispersed.

Therefore, in a preferred aspect, the present invention includes polyarylates having repeating structural units according to Formula I in which R is the same as described with respect to Formula I and L1, L2 or L3 of X and Y include pharmacologically or biologically active ligands. Still yet another aspect of the present invention provides polyarylates having repeating structural units according to Formula I in which X, Y and R are the same as described with respect to Formula I, wherein the polyarylate is physically admixed with a biologically or pharmacologically active material.

Another preferred aspect of the present invention provides amino acid-derived monomers that are capable of being polymerized to form polyarylates. The monomers provided by this aspect of the present invention are diphenol compounds that are tyrosine derivatives having the structure of Formula II:

HO ~ CH2-lCH-C-NH-CH-CH2 ~ OH (II~ X Y in which X and Y are the same as described above with respect to Formula I.

Among the monomers of Formula II are the desaminotyrosyltyrosine esters of parent U.S.

Patent No. 5,099,060 used in the preparation o~ polycarbonates and polyiminocarbonates in which X is hydrogen and Y is:

9 -C-O-L3 wherein L3 is an alkyl group containing up to 18 carbon atoms.

This aspect of the present invention incorporates the discovery that useful polyarylates can be prepared from tyrosinederived diphenol compounds with the N-terminus present and suitably protected, as well as C-terminus protecting groups.

A further aspect of the present invention provides an improved carbodiimide-mediated direct polyesterification process for thepolymerizationofpolyarylatesusing4-(dimethylamino)pyridiniump-toluene sulfonate (DPTS) as a catalyst. A polymerization process according to this aspect of the invention includes the steps of contacting a diphenol with a dicarboxylic acid in a common solvent for the diphenol and the dicarboxylic acid in the presence of a catalytic amount of DPTS and an excess of carbodiimide, and recovering the resulting polyarylate.

According to this aspect of the present invention, there is provided a method for the polymerization of polyarylates comprising, mixing equimolar quan~ities of:

(a) a ~i~enol ha~in,g a S~uC~ ,Q~9t~'d by the formlll9~ o HO~,CH2-CH-C-~I~H-CH2~0H wherein X is ~d ~om t~e gro~p con~tin~ of H, -NL,~, O O Il 11 z5 -NH-C-L~ C-O~; and Y is s~l~t~d f~om the group a~r~i~ng of hy~ogen and ester grOUps having the structure:

O Il .

-C-OL~ where~n L,l, L2 and L3 ~re ir~d~P-lfly ~l~ om ~e group . 30 CQ~ ti~ of s~a~gbt or ~l~ncL~ al~grl and all~ylsrryl g~oups c~t~inin~ up to 18 ca~oon atoms; and !

,~ -9a- ~ ~ 4 ~ - (b) ~ dic4lbo,-rlic acid ha~ing a ~tructure ~ed by the fnrrnu~ O O Il HO-C-R-C-OH - wherein R is ~~ d from the g~up c~nqst~ of ~ d and ur~ t~ subs~ and ~ b~ h~t~d alkyl, aryl and alky~asyl groups c4~t.~ ing up to 18 carbon atoms;

toge~er w~th a c~y~c amount of ~(diu.~ u~o) py~dinium-~toluene sulfonate (DPrS) h a ~lvent for sa~t polyaryl~te, and 10adding at least one molar eqw~alent per mole of ~ carbo~ylic acid ~roups present of a ~ ;m;~e, ~o that a polysrylate ~ is formed cornr-i~d of ~IJ~struchuaI units le.~-Act~nt~ by the forrnula:

O O O Il .

Il ~ ~ C~-CH-C~ CH-CH2 ~ ~C-R-C-O _ 21402S5 ~_.

--10- This aspect of the present invention represents a modification of the PL o~edure of Moore et ~1., to u~e an ex~-~r of the carbodiimide coupling reagent. This modification unexpectedly S increa~es the molecular weight of the polymers obtained to over 50,000 daltons. It is not~o.~hy that this modification to incre~se the amount of coupling reagent is CG..~ ary to the conventional wi~dom employed when selecting the amount of coupling reagent to be used to mediate direct polyesterifications.

The unders~a~ g of the ordinarily ckilled artisan is th~t the molar equivalentc of the coupling reagent utilized should not eYc~e~ the molar equivalents of the functional ~o.~ to be coupled.

Thus, Moore et al. used a ratio of from 1.3 to 2.0 equivalents of coupling reagent to 2.0 equivalents of functional groups to be coupled.

By going against such conventional wisdom, the inventive ~ocr_s significantly increases the molecular weights obtained for the poly~rylates to levels heretofore unknown in the art.

R~rFF DFc~RTpTIoN OF ~ DRAWT~GS FIG. 1 depict~ the molecular weight degradation for adipic acid-derived aliphatic polyarylate embodiments of the present invention.

FIG. 2 depict~ a model drug release profile for the aliphatic polyarylates of FIG. 1 using p-nitroaniline a~ the agent released.

FIG. 3 depicts the model drug release profile of FIG. 2 as a function of the square-root of time.

RF-~T MOnFC OF ~ YT~G OUT T~v~ ON The polyaryl~tes of the ~~ent invention employ ~ir~enol compounds derived from dimers of L-tyro8~ne as a starting ~aterial. L-tyrosine derived diphenol compo~nAC suitable for the polymerization of polyarylates have the ~tructure depicted by Formula II in which X and Y are the same a~ described above with respect to For~ula I.

._ 2i~0255 L-tyrosine is a naturally G '~L~ ing amino acid. Among the preferred L-tyrosine derivatives of Formula II are derivative~ in which X is hydrogen.

The~e preferred compound6 are tyrosine dipeptide analogs known a6 de~aminoLy~osyl-tyrosine. Desaminotyrosine ~Dat) itself v~o~l. naturally in plants. In this preferred group, the ~irhPnol can be regarded as a A9~i~eptide.

The preferred desamino-Ly~o~yl-tyrosine compo~ln~- of ~ormula II are those compounds in which Y :

-~-0-L3 wherein L3 i~ an alkyl or alkylaryl group containing up to 18 carbon atoms. In this particularly preferred group, the Aiphenol starting ~aterial is properly referred to a~ a desamin~r.o~yl-tyrosine alkyl or alkylaryl ester. The preferred monomers of the group of desaminotyrosyl-tyrosine esters are the ethyl, butyl, hexyl, octyl and benzyl esters. The most pre~erred ester is the hexyl e~ter, referred to as desamh-o-yrG_yl-tyro ine hexyl ester, or DTH.

The preparation of the diphenol starting material~ can be depicted by the following scheme:

Tyr Tyr IJ ~L~Lected Tyr C-protected Tyr ~ i rhon pure A~p~nol -- polyarylate C-terminus protected alkyl and alkylaryl ester~- of tyrosine con~aining up to 8 carbon atom~ are prepared according to the ~.occ~ e disclosed in J.P. GL~e~atein and N.

Winitz, ~h~;Q~ry of ~hP Amino Acid~, (John Wiley ~ Sons, New York 1961), p. 929.

21~025~ '_ -12 C-terminus protected alkyl and alkylaryl esters of tyrosine cont~ini~g more than 8 carbon atoms are prepared according to the ~LO- e~ ,e disclosed in Overell, U.S.

Patent No. 4,428,932.

N-terminus protected tyrosines are prepared following s~an~ard p.~_e~l~rQs of peptide chemistry such as di~closed in R~A-n~ky, Practice of Pe~tide Synthesis (Springer-Verl_g, New York, 1984).

The protection of either terminuC is omitted if X or Y of Formula II is hydrogen.

The crude tyrosine derivatives are sometimes obt_ined as oils and can be purified by simple Le_~y-~allization.

Crystallization of the pure product i8 accelerated by crystal 3ee ~in~.

The ~irb~nQls are then prepared by carbodiimide mcdiated coupling reactions in the presence of hydroxybenzotriazide following sta~A~rd p,o~e~ res of peptide chemistry such as disclosed in ~o~Anszky, Pr~ctice of PeDtide Synthesis (Springer-Verlag, New York, lg84) at page 145.

The crude diphenols can be recrystallized twice, first from 50% acetic acid and water and then from a 20:20:1 ratio of ethyl acetate, hexane and meth_nol, or, alternatively, flash chromatography on silica gel is used, employing a 100:2 mixture of methylene chloride:methanol as the mobile phase. DTH is prepared by the carbodiimide mediated coupling of desami.-~yLo~ine and tyro~ine hexyl ester in the ~r ~~n~e of ~ydLox~benzotriazole.

The ~iF~~~ol compo~ndc are then reacted with aliphatic or aromatic dicarboxylic acids in a carbodiimidc-mediated direct polyesterification using DPTS aC a catalyst to form aliphatic or aromatic polyarylates.

DiCArhoyylic acids suitable for the polymerization of polyarylates have the structure of Formula III:

O O HO-C-R-C-OH (III) 21~025~ -13 in which, for the aliphatic polyarylates, R is selected from saturated and unsaturated, substituted and un~ub6tituted alkyl or alkylaryl ~,ou~ containing up to 18 carbon atom~, and preferably from 2 to 12 carbon atom6. For the aromatic polyarylate~, R is selected fro~ aryl group~ containing up to 18 carbon atoms and preferably from 6 to 12 r~r~on atoms.

R iB preferably ~elected so that the dicarboxylic acids employed as starting materials are either important naturally-occurring metabolites or highly bioco~patible compo~ . Preferred aliphatic dicarboxylic acid starting materials therefore include the intermediate dicarboxylic acid~ of the cellular respiration pathway known a~ the ~reb6 Cycle. These lS dicarboxylic acids include alpha-ketoglutaric acid, ~uccinic acid, fumaric acid, malic acid and oxaloacetic acid, for which R i~ -CH2-CH2-C(-O)-, -CH2-CH2-, -CH=CH-, -CH2-CH(-OH)- and -CH2-C(SO)-, L e~pe~~tively.

Another naturally-occurring, preferred, aliphatic dicarboxylic acid starting material is adipic acid (R = (-CH2-)4), found in beet juice. Still yet another preferred biocompatible aliphatic dicarboxylic acid i6 sebacic acid (R - (-CH2-)8), which ha~ been ~tudied exten~ivQly and has been found to be nontoxic as part of the clinical evaluation of polytbis(p-carboxyphCI~o y) propane-co ~hAcic acid anhydride) by Lsurencin et al., J. Rio~ed. M~ter. Res., ~, 1463-81 (1990).

Other preferred biocompatible aliphatic dicarboxylic acids include oxalic acid (no R), malonic acid (R-l-CH2-)), glutaric acid (R-(-CH2-)3), pimelic acid (R-(-CH2-)5), ~uberic acid (R~(-CH2-)6) and azelaic acid (R~(~CH2~)7)- In other word~, among the aliphatic dicarb~xylic acids ~uitable for use in the pre ent invention are compounds in which R ~c~ ents (-CH2-)z, wherein Z iB an integer b~r~ -n zero and 12, inclu~ive.

Among the preferred aromatic dicArhQ~ylic acid~ ~uitable for use in the present invention are 21~0~5S -14 terephthalic acid, i~ophthalic acid and bis(p-carboxyphenoxy) Al~A~ec such as bis(p-carboxyphenoxy) propane.

The biocompatibility of bi~(p-ca bo~yl~h~nQ~y) propane is di6clo~ed by the above-cited Laurencin et al., J. ~iomeA. MAter.

Res.

r~ca~ the A j ph~ol~ of the present invention are hag~ citive, the polyarylates of the present invention are prepared by direct polyesterification, rather than by diacid chloride ~echniques.

Polyesterification ~o~ncing agents and reaction condition~ ~hould be rh~-~n that ~re compatible with the hA-e ~-enaitive Aip~ol starting materials.

Thus, the polyarylates can be prepared by the prc-ess disclosed by Ogata et al., PolY~. J., l3rl0), 989-91 (1981) and Yasuda et al., J.

Polym.

Sci.

PolyE. ~h~ A. ~ 2609-16 (1983) ufiing triphenylpho~rhine as the condensing agent, the ~ e-~ of Tanaka et al., PolYm.

J., 14(8~, 643-8 (1982) using picryl chloride as the c~nA~ncing agent, or by the process of Higashi et al., J. poly~.

Sci.: PolYm. ~h~. FA. ~ 589-94 (1986) using pho~phQru~ oxychloride a6 the condensing agent with lithium chloride monohydrate a~ a cataly~t.

The polyarylatec can al~o be prepared by the method disclosed by Higashi et al., J. polvm~.

Sci.:

Polym. rh~ A. ~ 3233-9 (1983) using arylsulfonyl chloride as the conden~ing agent, by the process of Higashi et al., J. polY~.

Sci.: PolYm.

Chem.

Ed., ~, 3241-7 (1983) using diphenyl chlorophocphAte as the conA~n~ing agent, by the pl~_~S~ of Uig~hi et al., J.

PolY~L Sci.: poly~. ~hem. ~A. ~ 97-102 (1986) using thionyl chloride with pyridine as the conA~n-ing agent, or by the ~ e-- of Eliac, et al., ~romol.

Chem.,-~L~2, 681-6 (1981) using thionyl chloride with triethylamine. A preferred polye~terification ~ re i8 the-method di~closed by Moore et al., Macromol., 23, 65-70 (1990) utilizing carbodiimide coupling reagent~ a~ the condens~ng agents with the specially decigned cataly~t DPTS.

21~025~ -15 A particularly preferred polyesterification technigue modifies the method of Moore to utilize an _x_a~C of the c~rh~i imide coupling reagent.

This pr~A~c~ aliphatic polyarylates having molecular weightc greater than those obtained by Moore.

Essentially any carbodiimide commonly used as a coupling reagent in peptide chemi_try can be used as a condensing agent in the preferred polye6terification pLG.e-~.

Such carbodiimide_ are well-known and disclosed in Ro~Anszky, PrActice of Pe~ti~ Svnth~c;c (Springer-Verlag, New York, 1984) and include dicyclohexylcarbodiimide, dii~opropylcarbodiimide, l-(3-dimethylamil.o~L~yl)-3-ethyl carbodiimide hydLo~l.loride, N-cyclohexyl-N'-(2'-morpholinoethyl)r~r~oAiimide-metho-p-toluene ~ulfonate, - 15 N-benzyl-N'-3'- dimethylaminopropyl-carbodiimide hydrochloride, l-ethyl-3-(3-dimethylami..o~ G~ 1 ) carbodiimide methiodide, N-ethylcarbodiimide hydrochloride, and the like.

The preferred carbodiimidec are dicyclohexyl carbodiimide and diisopropylcarbodiimide.

A reaction mixture i~ formed by contacting exactly eguimolar guantitie_ of the diphenol and the dicarboxylic acid in a solvent for the diphenol and the dicarboxylic acid.

Suitable solvent~ include methylene chloride, tetrahydrofuran, dimethylformamide, chloroform, carbon tetrachloride and N-methyl pyrroli~in~-~.

It i8 not ne_~ ry to bring all reagents into complete solution prior to initiating the polyesterification reaction, although the polymerization of slightly ~oluble monomer~ such as desami~.oLy~osyltyro~ine ethyl ester and succinic acid will yield higher molecular weight polymers when the amount of solvent i~ increased.

The reaction mixture can also be heated gently to aid in the partial dissolution of the reactants.

- In this particularly preferred method, the reaction mixture will also contain an ~YC~Ffi of the carbodiimide coupling reagent.

When carbodiimides are used in peptide synthesis as disclosed by ~o~n~zky, 21~025S _ '',_ -16 between 0.5 to 1.0 molar equivalents of carbodiimide reagent is used for each mole of carboxylic acid group pre~ent. In the ~L- -~nt invention, greater than 1.0 molar equivalents of carbodiimide per mole of carboxylic acid group present are used. This ic what is meant by de~cribing the reaction mixture as conta i n j ng an excess of carbodiimide.

The polymer molecular weight significantly increa~e~ as the amount of coupling reagent used is increa~ed. The degree of molecular weight increase only begin~ to level off around four molar equivalents of carbodiimide per mole of carboxylic acid group.

Increasing the amount of coupling reagent beyond four equivalent~ of carbodiimide ha~ no further beneficial effect. While quantities of c~rh~Aiimide greater than four eguivalent~ are not detrimental to the polye~terification reaction, ~uch qu~ntitie6 are not cost-effective and are thus not fa~oL~d for this reason.

The carbodiimide-mediated direct polyesterification is performed in the p ~~ence of the catalystDPTS. DPTS is prepared in accordance with the procedure of Moore et al., Macromol., 23~1), 65-70 (1990). The amount of DPTS i~ not critical hec~r? the material is a true cataly~t that i regenerated. The catalytically effective quantity i~ generally between about 0.1 and about 2.0 molar eguivalent~ per mole of c~rboxylic acid group, and preferably about 0.5 equivalents per mole of carboxylic acid group.

The reaction proceed~ at room temperature, or about 20-30-C. The reaction mixture can be heated ~lightly (~60-C) prior to carbodiimide addition to partially ~olubilize le~ soluble monomers. ~W_~eL~ the po~ymerization reaction it~elf should be cQnAl~cted ~ between 20-C and 30-C.

Within this temperature range, the reaction can be continued, with ~tirring, for at least 12 hours, and preferably for from one to four days. The polymer is L.c~ered by ~e~ching the reaction mixture in methanol, from which the polyarylate 21~025S _.

-17 ly precipitates while the residual reagents remain in solution.

The precipitate may be separated by me~h-nical separations such as filtration and purified by solvent wA~bing.

In a preferred p~ re, equimolar amounts of pure, dried tyrosine-derived ~irh~nol and dicarboxylic acid are weighed preciaely (+ O.ooOl g) and placed in a round bv~Lomed flask, ~e l.ied at 130-C. A suitable magnetic stir bar i8 placed into the flask.

Then 0.4 equivalents of DPTS are added.

The flask is fitted with a septu~ and flu~hed with nitrogen or argon to remove traces of moisture from the reaction mixture.

Next, a quantity of HPLC grade methylene chloride is added via a syringe and the reaction mixture i~ stirred - 1~ vigorously to suspend the reactants.

The amount of methylene chloride used will ~ p~nd upon the solubility of the ~irh?nol, or the dicarboxylic acid, or both monomers.

At this stage, the reaction mixture may be slightly heated to partially dissolve the monomers.

While it is not essential that the monomers be completely dis~olved, the quantity of solvent ahould be sufficient to disaolve the polymer as it forms and thus slowly bring the monomera into aolution.

4.0 equivalent~ of dii~v~v~ylcarbodiimide are then added to the reaction ~ixture via a ayringe.

After about 10 minutea, the reaction mixture becomes clear, followed by the formation of a cloudy precipitate of diiospropylurea.

After stirring between 20'C and 30 C for one to four days, the reaction is terminated by pouring the reaction mixture slowly and with vigorous ~tirring into ten volume~ of methanol.

The polymer precipitate~ whiie the re~idual reagents remain disaolved in methanol, resulting in the form~tion of the clear aupernatant.

3S - The polymeric product ia retrieved by filtration and washed with copius amounta of methanol to remove any impuritiea.

If desired, the polymeric products can be further purified by dissolving in 21~025~ '" -18 methylene chloride (10% or 20% w/w) and reprecipitating in methanol. The polymeric product is then dried to constant weight under high vacuum.

The polyarylates of the present invention thus include one or more .e_~LLing structural units ,~L~ ted by Formula I a~ de~cribed above, with X, Y and R being the ~ame a~ dQscribed above with L e_~oct to Formula I. The ~e_u, L ing structural units of preferred polyarylate are ,~L~--ented by ~ormula I as described _bove, with X being hydL~yc-- _nd Y being:

-C-0-L3 with L3 being an _lkyl or alkyl_ryl group containing up to 18 carbon atoms and preferably being an ethyl, butyl, hexyl, octyl or benzyl group. R is a saturated or uns_turated, ~ub~tituted or unsub~tituted, alkyl, aryl or alkylaryl group contAining up to 18 carbon atoms and i~ preferably A saturated or un~_tur_ted, sub tituted or unsubstituted _lkyl group containing from 2 to 12 carbon atoms.

Preferred polyarylates have weight-average molecular weights above 50,000 daltons, and preferably above 100,000 daltons. Molecular weights are calculated from gel permeation chromatG~ hy relative to poly~LyLene ~t~nA-rd~ without further correction.

Thc polyarylates can be worked up by known methods com~only employed in the field of ~ynthetic polymers to produce a variety of useful article~ with valuable physical and chemical propertie~, all derived from ti~F~le co~patible monomer~. The u~eful article~ can be ~haped by cu.,~e..~ional polymer-forming t~h~i~ue6 such a~ extru~ion, compre~sion molding, injection moldingr, and the like. Nolded articles prepared from the polyarylate~ are u~eful, inter alia, a6 degradable biomaterial~ for medical implant application Such applications include the u~e of the molded articles a~ va7~1Ar graft~ and stent~, bone plates, ~uLu~ implantable ~ensor~, barriers for ~urgical adhe~ion 21~025S ,....

~,_ --19- ~.-ve..Lion, implantable drug delivery devices and other therapeutic aids and articles which decompose harmlessly within a known period of time.

The polyarylatec can also be formed into drug delivery i~plants that degrade to release pharmacologically or biologically active agents within a predictable controlled release time. Such controlled drug delivery sy~tems can be prepared by incorporating the active agents into the polymer chAin~ as pen~t side chains or by cross lin~ing the ~en~lnt side chains to form a polymeric matrix into which the active Agents are physically emk~A~A or ~i~psrsed. Controlled drug delivery system implants can also be formed by physically admixing the polyarylates with a biologically lS or pharmacologically active agent. The foregoing ~oce~ L3 are es~entially c~ c--Lional and well-known to those of ordinary Ckill in the art.

Biologically or pharmacologically active agents can be chemically incorporated into pendant side rh~inC of polyarylate~ having repeating ~tructural units according to Formula I, wherein X is an unprotected amino group (N-terminus) or Y is an unprotected carboxylic acid group (C-terminus). The formation of polyarylates h~ving de ~.oLected N- or C-termini from the polyarylate~ of the ~re ~~t invention having protected N- or C-termini can be achieved by the use of temporary protecting ~LoU~ as di_closed by PoA~nczky.

The selective u~e of such temporary protecting ~LOU~6 is well-known to those of ordinary skill in the art of peptide synthesi~.

A variety of pharmacologically or biologically active ligand~ having functional ~LoU~5 capable of being coupled- to a free amino or carboxylic acid group can then be covalently incoL~oLated into the deprotected monomeric units of the polyarylate~ by c6~.~e.~Lional coupling ~e~hnique6. The resulting polyarylate has repeating ~tructural units according to Formula I in 21~02~ '_ -20 which X i8 -NHLl, -NL1L2:

O o U 11 -NH-C-Ll or -NH-C-O-L or Y is:

-C-0-L3 wherein Ll, L2 or L3 is a pharmacologically or - biologically active ligand.

Ll or L2 of X include pharmacologically or biologically active li~nA~ which have a free carbocylic acid group through which the ligand i8 covalently honA~A to X. Such ligand~ are ~elected from ibuprofen, methotrexate, N-acetyl-cis-hyd~y~oline, ce~haArine, ampicillin, and the like. L3 of Y includes pharmacologically or biologically active ligands which have a free ~mino or hydroxyl group, through which the ligand is covalently ~on~~A to Y. Such ligand~ are selected from dopamine, cis-hydroxyproline ethyl ester, doxorubicin, histamine, ampicillin, cephradine, and the like.

Eor ~o..L.olled drug delivery ~y6tem~ in which a biologically or pharmacologically active agent is phy~ically ~m~-~d~A or Ai~r-rsed into a polymeric matrix or phy~ic~lly admixed with a polyarylate, ~uitable biologically or phar~acologically active agents include in principle any active agent that has to be repeatedly admini~tered over prolonged periods of time.

The following non-limiting ex~mples set forth hereinbelow illustrate certain a~pects of the invention.

All part~ and percentages are by weight unle~6 otherwi~e noted and all temperatures are in de~L~-~ Cel~iu~.

Dicarboxylic acid~ and all other reagent~ were purchar~ in pure form and were used as received. Solvents were of ~HPLC grade.~ All ~;p~enQlic ~onomers (e.g., the ethyl,-butyl, hexyl and octyl e~ters-of desaminoLy~G_yl tyrosine) were prepared according to the ~LG~-e~ e provided in Example I of U.S. Patent No. 5,099,060.

Al~ho~h thi~ re refer~ ~pecifically to DTH, -21 monomers having esters other than the hexyl ester (including the benzyl ester) can be readily prepared by the same basic procedure.

The DPTS catalyst was prepared as described by Moore et al., Macromol., 23~1), 65-70 (1990).

F~AMPT.F~.~ The examples use the following polymer characterization pror~nre6.

SPFCTROSCOPY FT-IR ~pectra were recorded on a Matson Cygnus 100 spectrometer.

Polymer samples were dissolved in methylene chloride and films were cast directly onto NaCl plate6.

All spectra were collected after 60 scans at 4 cm 1 re~olution. W -vicible spectra were recorded on a Perkin-Elmer Lambda 3~ spectrophotometer.

NMR spectra were recorded on a Varian VXR-200 spectrometer.

MOLECUT~R WEIGHTS Molecular weights were determined by gel permeation chromatography (GPC) on a chromatographic system consi~ting of a Perkin-Elmer Model 410 pump, a Waters*Model 410 RI detector, and a PE-Nelson Model 2600 computerized data station.

For polymer an~lysis, 2 PL-gel GPC columns (300 nm x 7.7 nm, particle size 5 micron, pore ~ize 105 and 103 Angstrom, respectively) were placed in ~eries and were operated at a flow rate of 1 mL/min in DMF containing 0.1% (w/v) of LiBr or in methylene chloride.

Molecular weights were calculated relative to polystyrene stA~A~rds without further correction~.

~!A~. PRO~r ~ ~ The glass transition temperature (Tg) was mea~ured with a DuPont 910 Differential ScAnning Calorimeter (DSC) calibrated with indium.

The decomposition temperature (Td) wa~ measured with a DuPont 951 Thermogravimetric Analyzer at a heating rate of lO-C/min.

* - Trade-rnark - ~ '~ -22 S~TVENT CASTING The polymer (1.5 g) was dissolved in 25 mL of methylene chloride. The solution was filtered through a syringe filter (0.45 micron pore, 25 mm PTFE filter3 and poured into a 4~ X 4~ glas~ mold in ~ glove bag under nitrogen. The film was first dried for three days in the glove bag, followed by drying in high vacuum to constant weight.

~ NI~AT PR0~ The mechanical properties of thin, solventcast polymer films were measured on an Instron Tensile Tester according to ASTM stAnAard D882-83. In all cases, unoriented, non-crystalline films were used, yielding results that were representative of the bulk - 15 properties of the polymer~ at room temperature.

HYDROr YTTC ~Ft'.R~n~TTON STuDTT~ Samples of various ~Ap~fi and sizes were cut from solvent cast films. The samples were incubated at 37 C in phosphate buffer-saline solution (0.1 M, pH 7.4) contAi~inq 200 mg/L sodium azide to inhibit bacterial growth. The degradation process was followed weekly by recording the weight change of five individual samples, by measuring the residual polymer molecular weight, by solution and by surface analysis (FTIR, ESCA and SEM) and by me~A~ical property testing of partly degraded Famples after expo~ure to the bùffer solution.

MODT'T ~DRUG ~TT~ST~'~ STUDIES An exactly weighed amount of p-nitroaniline (pNA) was dissolved in methylene chloride. To this solution, an exactly weighed ~mount of polymer wa5 added, such that the required ratio of dye to polymer was obtained. Any undissolved residues were removed by filtrat~on through a 0.45 micron membrane filter. The ~ filtered solutions were then used for solvent casting as described above. Homogeneous and transparent films were obtained. Release devices (circular disks, diameter approximately 1.0 cm, thisknesq about 0.01 - 0.03 cm, * - Trade-mark f~ , 21~025~ l_ -23 - weight about 10 - 40 mg) were excised from the dried film~ with a cork borer.

The relea~e devices were placed into 10 mL of rh~rh-te buffer (0.1 M, pH 7.4) and kept at 37 C. The amount of dye relea~ed was determined spe_L-G~hotometrically. After each 6uch analysis, the device wa~ tran~ferred into fresh, sterile pho~rhAte buffer.

The following table define~ the abbreviations adopted for the ~irh~nols and polyarylates illustrated by the examples below:

S~ S T Desami.. oLyl~_yl-tyrosine hexyl ester DTH De~amh.~LyLG_yl-tyrosine ethyl ester DTE Desami--oLy~yl-tyrosine octyl ester DTO Poly(DTH succinate) DTH-suc Poly(DTH adipate) DTH-AA Poly(DTH ~O~ACAte) DTH-SA Poly(DTE adipate) DTE-AA Poly(DTO adipate) DTO-AA Poly(DTH terephthal~te) DTH-Tere F~'XI~MPt.F 1 DTH-8Uc was prepared by placing DTH (10.3380 g, 0.025 mole~) and ~uccinic acid (2.9523 g, 0.025 mole~), into a 500 mL dry fla~k. DPTS (2.9940 g, 0.4 equivalents) was added and after sealing the flask and fl~ ;ng with nitrogen, methylene chloride (180 mL) was added.

The reaction mixture was stirred vigorously to au~pend the reactants.

Dii7~ o~ylcarbodiimide (12.6230 g, 4.0 equivalent~) was added via syringe.

After four (4) days, _the reaction was terminated by pouring the reaction mixture into 1,800 mL of methanol.

The crude polymer was purified by di~solving in methylene chloride (20% w/w) and .e~e_ipitating from 110 mL of methanol.

8.18 g of a colorles6 polymer wa~ obtA i ncA (66% of ~h~ ical yield) having a weight ave~e "~ 21~0~S5 -24 molecular weight of 102,000 daltons, a number average molecular weight of 68,000 daltons and a polydispersity of 1.50.

F~AMPT-F~ 2 S DTH-AA was prepared by placing DTH (8.2712 g, 0.020 moles) and adipic acid (2.9228 g, 0.020 moles) into a 500 mL dry fla~k. DPTS (2.3552 g, 0.4 equivalent~) was added and after sealing the flask and fl~ ing with nitrogen, methylene chloride (150 mL) was added.

The reaction mixture wa6 stirred vigorously to ,e"l the reactant~.

Dii~ o~ylcarbodiimide (10.0984 g, 4.0 equivalents) wa~ added via syringe.

After one day, the reaction was terminated by pouring the reaction mixture into 1,500 mL of methanol.

The crude polymer was purified by di~solving in methylene chloride ~5% w/w) and ~ e_ipitating from 500 mL of methanol.

5.58 g of a colorle66 polymer wa6 obtained (53~ of theoretical yield).

The polymer had a weight average molecular weight of 232,000 daltons, a number average molecular weight of 132,000 dalton6 and a polydi~-r6ity of 1.76.

F~MPT~ 3 DTH-SA was prepared by reacting sebacic acid (5.0560 g, 0.025 moles) with 0.025 moles of DTH according to the ~oce~ e de~cribed in Example 1.

Following purification by dissolving in methylene chloride and ~e~e-ipitating from methanol, 11.32 g of a colorle~s polymer wa~ obtained (76~ of theoretical yield).

The polymer had a weight average molecular weight of 212,000 daltons, a number average molecular weight of 133,000 dalton~ and a polydi~per~ity of 1.59.

PT~4 DTE-AA was prepared by placing DTE (8.9100 g, 0.025 moles) and adipic acid (3.6535 g, 0.025 moles) in a 500 mL dry flask. DPTS (2.9440 g, 0.4 equivalents) wa added and after ~ealing the flask and f~ in7 with ni~l6ye.., methylene chloride (275 mL) wa~ added. The 2I ~ 025~ -25 reaction mixture was stirred vigorously to suspend the reactants. DiisG~ ylcarbodiimide (12.6230 g, 4.0 equivalents) was added via syringe. After one day, the reaction was terminated by pouring the reaction mixture into 1,800 mL of methanol. The crude polymer was purified by dissolving in methylene chloride (20% w/w) and .e~ecipitation from 80 mL of methanol.

7.66 g of a colorless polymer was obtained (66% of theoretical yield). The polymer had a weight average molecular weight of lSl,000 daltons, a number average molecular weight of 68,000 daltons and a polyAi ~-rsity of 2.2.

~AMP~.F 5 DT0-AA was prepared by placing DT0 (9.7148 g, 0.022 moles) and adipic acid (3. 2151 g, 0. 022 moles) into a 300 mL dry flask. DPTS (2.5907 g, 0.4 equivalents) was added and after sealing the flask and fl~ching with nitrogen, methylene chloride (140 mL) was added.

The reaction mixture was stirred vigorously to suspend the reactants.

DiisG~u~lcarbodiimide (11.1082 g, 4.0 equivalents) was added via syringe.

After one day, the reaction was terminated by pouring the reaction mixture into 1,400 mL of methanol.

The crude poly~er was purified by dissolving in methylene chloride (20~ W/W) and ~ ecipitation from 100 mL of methanol.

9.52 g of a colorless polymer was obtA i n~A (78~ of theoretical yield).

The polymer had a weight average molecular weight of 218,000, a number average mole~ weight of 128,000 and a poly~i~persity of 1.70.

TpF~MAT. PROrr ~ Several batches of polymers were prepared according to the examples given above.

The thermal 3 5 properties of these polymers were analyzed as described above and are listed in Table II.

The aliphatic polyarylates listed here have glass transition temperatures between 17-C and S6-C and become soft ~140~Sj '_, -26 enough for compression and injection molding about 40'C above their .e_~e_Live glass transitions. Because the decompo6ition temperatures are well above 300-C in all cases, the polymers can be readily processed by thermal polymer ~ ing tec~niques.

TT ~oly~ r ~ T Td o~ OD~ C) ~ 102,000 68,000 4S 371 DTH-AA 232,000 132,000 34 356 208,000 85,000 33 360 186,000 126,000 32 356 DTH-SA 212,000 133,000 17 368 DTE-AA 209,000 114,000 56 340 DT0-AA 218,000 128,000 28 357 DFr-l~nz~T~s:)N STUnT~.C A degradation study was started for the adipic acid serie~ (DTE-AA, DTH-AA and DT0-AA) to determine the time frame and me~h~nism for degradation.

During the first five weeks of the ~tudy, none of the polymers lost any weight.

Uou~v~r, a decrea~e in molecular weight was detected, confirming the degradability of the polymers under physiological conditions.

The decrease in molecular weight i~ shown in FIG. 1.

The rate of degradation will most probably be controlled by the overall l.yd~o~l.obicity of the polymer and its morphological state.

In terms of overall hydrophobicity, the r_te of degradation should decrease with increasing length of the polymer side chain.

In term~ of morphological state, polymer~ that are in the rubbery ~tate _t 37-C ~hould degrade faster than polymer~ that are in the gla~y state at that temperature.

Thu~, the gla~s transition temperature (Tg) becomes the e~n~ rate determining parameter in the degradation p~G~e ~~ .

As shown in FIG. l, from the data obtained over the initi_l five week~, DT0-AA arp~rs to degrade ~ 2I~025S '_ -27 fastest. This polymer is in~ in the rubbery state at 37-C, while the polymers that App~ar to degrade the lea6t, DTE-AA and DTH-AA, have higher glass transition temperatures A Thus, the current invention indicates that there are at least two interrelated parameters (length of side chain and glase transition temperature) that can be used to co..~ol the rate of degradation of tyrosine-derived polyarylate~.

MOnF~T. nRt~G 12FT.~A.~:F STUnTF.C Preliminary model drug release studies were performed for the polymers in the adipic acid series in an attempt to establish the feasibility of using tyro~ine-derived polyarylates as implantable drug release formulations. In these studies, p-nitroaniline (pNA) served a~ a model ~drug~ becau~e its release can be cG..~c.~iently aonitored by W -visible spectL--~opy.

The release profiles of pNA from disk ~~re~ polymer devices loaded with 5% (w/w) of pNA are shown in FIGS. 2 and 3.

It is evident that the release rate is related to the morphological state of the polymer, with polymers in the rubbery ~tate releasing much faster than polymers in the glassy state.

Thus, the order of drug release increases with increasing side chain length (ethyl to hexyl to octyl) ~G~ L ~ ~ponding to the decrease in the polymer gla~s transition temperature.

This i8 illustrated in Table III, which correlates the release rate (slope of thè FIG. 3 release curves) with the glass transition temperatures of the Le_~ective polymers.

T~r.~ ~TS Poly~r ~o 810p - ~ - - G ~) Tg ~-C) DTE-AA 2.98 x 10 4 56 DTH-AA 8.56 x 10-4 33 DTO-AA 11.21 x 10 4 28 The difference between FIG. 2 and FIG. 3 is in the representation of the data. FIG. 2 shows the accumulated fraction of drug released as a function of 2140Z5~ ; ,~ -28 time, while FIG. 3 shows the accumulated fraction of drug released as a function of the sguare root of time.

The fact that a significant portion of the release profiles in FIG. 3 are straight lines is an indica~ion that drug relea~e in the sy6tem~ is predominantly ~diffu~ion controlled.~ The release rate~ (as defined by the slopes of the linear regions of the release profiles in FIG. 3) are related to the Tg of the polymer, indicating that it will be possible to formulate a wide range of drug delivery systems by selecting different polymers from among the tyrosinederived polyarylate~.

Obviously, drugs other than pNA will have release characteristics different from the data shown in FIGS. 2 and 3. The data shown here simply intend to demonstrate the potential uRefulne~ of aliphatic polyarylates in the are~ of druç delivery.

PHysTco~ T~AT~ PRorr~ Preliminary mech~nical property studies have been done for the polymers DTH-AA, DTH-SA, DTE-AA and DTo-AA and the results are tabulated in Table IV:

T~T.~' nr Ten~ 8trength % ~longation Poly~-r ~an- ~Pa) ~t br--~ DTE-AA 37.1 DTH-AA 37.2 498 DTO-AA 44.2 634 DTH-SA 46.4 727 B_sically, each polymer differs in ~pp~Arance~ ductility and vi-coelasticity Aepen~ing upon whether the polymer i8 in the glassy state or the rubbery state at room temperature. Because DTE-AA is in the gla~sy state at room temperature, it is much less ductile than the other three polymers. All polymer~ in the series can be oriented by heating above Tg and ~tretching. Oriented films are ~ignificantly DL~G..yer than unoriented films.

DTH-SA is the most ductile of the four polymers. It can ~ 2140~55 ._ -29 be ~tretched to long extensions and will return back to its original shape.

This polymer thus shows elastomeric behavior at room temperature.

F~MPT~ 6 ~ .re wa6 prepared by reacting terephthalic acid (1.1629 g, 0.007 moles) with 0.007 mole~ of DTH (2.8949 g) in a 250 mL dry flask according to the p~o~ re de~cribed in Example 1.

Unlike the reactions involving aliphatic acids, the reaction mixture never cleared up, reflecting the decrea~ed colubility of the reagent~.

Instead, a gelatinol~- polymeric mass formed.

After one day, the reaction was terminated by pouring the reaction mixture into 800 mL of methanol.

The resulting product was dried under high vacuum overnight.

3.54 g of a colorless polymer was obtained (93% of ~h-~.e~ical yield). r~c~ee of the in~olubility of the polymeric product in common organic ~olvents, the molecular weight could not be determined.

~ AT PRO~ C OF DTH-Tere Analysis of the polymer by DSC indicated that DTH-Tere has a glas~ transition at 96-C and an additional endothermal transition at 190-C.

The additional transition at 190-C may be indicative of polymer crystallinity.

The exact nature of this tran~ition ic currently not understood.

SOTJl~TT.TTy OF Lr~ Lre TN OI2eANTc SOT~ NTS 10 mg aliquots of polymer were placed into te~t tubes and 2 mL of the indicated solvent were added.

If the ~olution became clear, the polymer was classified a~ soluble.

If the polymer appeared to ~well or dis~olve partially, it wa~ classified a~ partially soluble.

If no ~h-nr~ occurred, the polymer wa~ cla~ci~ied a~ in~oluble.

Solvents in which the polymer waC cla~sified a~ partially soluble or insoluble were then heated to their ~ p~ctive boiling points and the polymer was observed again.

The solubility of the polymer in variou~ ~olvents with and without the application of heat is illu~trated in Table V:

214U2~5 .._ -30~r.~ V 80~ t R~ t~ ~o B-~t ~-~t ~pD~ Dichloromethane Insoluble Insoluble Partially Partially Chloroform Soluble Soluble Partially Tetrachloroethylene Soluble Soluble Partially Partially DHF/lithium bromide Soluble Soluble Tetr_~ydxofuran Insoluble Insoluble Partially Pyridine Soluble Soluble Acetonitrile Insoluble Insoluble Ethyl acetate Insoluble Insoluble Ethyl ether Insoluble Insoluble Partially N-methyl pyrrolidinone Soluble Soluble Accordingly, a family of polymers is provided having pendent chain~ and a bac~hone structure that can be readily varied to modify the thermal and physicomec~anic_l properties.

By modifying the polymer structure _nd therm~l properties, the hydrolytic stability and the drug release properties of the polymers can be controlled.

The foregoing exa~ples and description of the preferred embodiment should be taken as illustrating, rather than a~ limiting, the ~.~ invention as defined by the claims.

As will be readily appreciated, numerous variations and combinations of the features set forth ~ithin the foregoing description and examples can be utilized without departing from the present invention.

STA~ T OF ~~V~l~TAr. APptT~RTTTTy The polymers of the ~ ~nt invention are useful in the preparation of biodegradable and 219025S -31 biocompatible film~ and molded articles and are particularly useful in the preparation of molded medical ~rticle~, e~pecially medical implants, including controlled drug delivery systems.