Prospects for site-specific delivery of pharmacologic and molecular therapies

a



JACC Vol . 23. No. 5 Apd11994.1234-n,9

affSSEN AND ISNER SITE,WINCIFIC THERAPY

1235

designed to accomplish infusion of short-lived aqueous solutions into the vessel wall. Alternatively . polymers de. signed as miceuparticles or stoats have been modified to provide timed, including longer-lived, release of the thempenile agents . Devices for Intravascular Drug Delivery

Figure 1. lmnnmohietochemieal double staining of an atheredomy apeeimm froma revraoite human femoral artery retrieved 6 months after the laidal angioplasry. The and rattler stains; labels cells positive for ptalAYattng cell nuclear antigen, suggesting a very high paliferadaa activity in this specimen . Additional staining with the mastic scrim-specific antibody HHF-35 . indicated by beau. eytopeaak staining, moutons that moat of the proliferating cells arc in fact smooth muscle Cons. tart shows proliferating cell nuclear antigen-positive smooth muscle calls at a higher magnification .

of the agent deposited in the arterial wall. The latter depends principally on how the agent is delivered to the artery . Most catheter devices developed for local delivery have been

Figure 2. Schematic of several catheter systems used for inum asev r drug delivery. M Doublahmloon catheter used to inlne dug mission into an antonial compartment isolated by two balloons . II, Paeaaa balloon catheter. Agents are Itdeeted into the wascular wall tlveeah total inclinations in the siuglalunen balloon catheter . C. Mkreporoia baleen. A panes balloon is covered by a permeable polycarbmmae membrane (pore size 0.8 Non) to reduce jet-induced igjtay. D, Channel balloon. Drug are inflated through small parfomind channels its" an the :canes of an agioplosty balloo . Pressures fee balloon Inflation and drug infusion thus can be sepntad. F, Stem in a balloon . A different system far the low patoe inawlon of drags by ate of a expandable atom within the heinous ap~~ the balloon agulntt the arterial wall . P. Bydregel catheter. A hydsaphife polymer eating acts as a dugabashieg sponge. Agate au tsenrfmred ro the areral wall daing balloon inflation.

DonMtibagsou catheter . The first device used for the localized delivery of drugs or genes was the double-balloon catheter (7-9) (Fig. 2A) . By inflating the two balloons, a scaled Compartment is created in which the solubilized therapeutic want can be introduced after blood has been evacuated from the compartment. The injected mixture reaches the arterial wall by diffusion or as a result of applied hydrostatic pressure. It has been recently shown that infusion pressures of 150 mm Hg (0 .2 arm) are both nontraumatie and sufficient forthe uptake of deoxysibese nuclricocld )NA) plasmids into the arterial wall ; pressures X500 min Hg typically create additional injury (10) . The double-balloon catheter may be particularly well sealed for nonangiopesty applications in which the endothelium is the delivery target . Clearly, one of the potential advantages not shared by alternative delivery devices is that the double-balloon catheter provides a design option for evacuation, in addition to instillation, of the compartment isolated between two inflated balloons. This issue is particularly relevant for agents such as vital vectors in which systemic circulation may be undesirable . However, the clinical use of the double-balloon catheter is limited by certain practical problems, such as leakage of the solution through side branches and relatively long incubation times of 15 to 30 min . Foully, the inflation pressure required to accomplish a satisfactory seal by the two balloons of the delivery compartment can lead to additional vessel injury proximal and distal to the target site . Use of ultracompliant latex in lieu of conventional polyethylene balloon material may help to prevent this occurrence . Form balmw off. The porous balloon (Fig. 28) was first used for intramusndar drug delivery by Wolimky and Thong (11) and has recently been used for intravasmdar gone therapy by several investigators (12,13) . Drug or DNA solutions can be delivered into the balloon under pressure 1a1+~

D

1 1 1 , h

a

E

C

F

A

IACC Vol . 23, No. s April 1994 :1274-44

eatheterleads to high local tissue concentrauom :of the drug . which. however, decrease rapidly within a few hours (26). Show. Scents have been proposed as useful carriers for sustained local drug delivery of antithrombutic or amiprotif . crative drugs, or both, and genes . Several approaches have heap investigated to achieve continuous drug release from a stets. First. metallic steers have been seeded with genelically modeled endothelial cells to date agents such as tissue plasminogen activator (t-PA) (27). Second. stets can be coated directly with dregs or with drug-eliding biodegradable polymers. Third . stems can be maoulietoed completely ..f.~d.1rug-eluting biodcgadable polyntas (25.29) . Several biodeo Bradnble polymers, such as potylecdde, polyglycotide, polynethoster, polycepmMctone. pdyhydroaybutyrate and polyethylene oxiddpolybutylene terephthalate. have been investigated for this purpose (28,29) . Preliminary experience with these materials may favor stems made of potytactide because ibis material canbe used to release drugs and causes only a mild tissue reaction (30), in contrast to the other materials that have been shown to evoke an extensive foreign body reaction (31). Fibrin has also been used to coat metallic stents and has been reported to reduce the incidence of stet occlusion withoa causing a significant inflammatory response (32). Finally, removable nitinal stems, developed with a drugShtting polyurethane coating (331, represent a new approach in which therapeutic agents might be delivered for a definite period, after which the device . and any stent-related liabilities, could be removed . Potential Therapeutic Agents for Local Drug Delivery CowerWrl drugs. In the past few years a number of d''erent, An still paeliminary, studies have investigated the efech of variou agents for local drug delivery . Because of its known aarithrcmbotir and antiproliferstive properties. heparin was aim of the first drugs tested . Delivered through a porous balloon, heparin has been demonstrated to persist in the media of the arterial wall for up to 48 h (11). A hydruguFcaated balloon has also been shown to achieve elficiam delivery of heparin into the vaucuhr wall in vitro (34) . By use of a Dacron graft shunt model, local heparin delivered in vivo from a hydrogel balloon was fouad to inhibit plaleltdepeodeat thranhosis (35) . However. experimental studies Involving local heparin delivery in a rabbit model of atherosckrosis failed to show any effect on restettoils when hepatic was delivered through a porous balloon (36). Heparin-coated scents likewise failed to reduce intimal prolileaation in several experimental studies (37,38), but, a preliminary study, testing local delivery of low molecular weight heparin through a porous balloon showed a favorable effect of this heparin preparation on thrombus and utrnimimal formation in an atherosclerotic rabbit model (39). Local delivery of angiopeptin (40). 0tiol protease inhibitors (41), Bole hicine (42). doxondlicin (16) or methotrexate through a porous balloon (43) or a scent (37) all filed to inhibit the

RIFSSEN AND I5NER SITE-SPECIFIC THERAPY

1737

development or intimat hyperplesia . One possible explanalion for the failure of these latter experimental studies may be the multifactorial nature of restenosis that resists any treatment strategy designed to address only a single mechanism. It is possible to consider combination chemotherapy, that is, delivery of several agents that synergistically act to retard vessel narrowing through multiple mechanisms . Agents that block protein synthesis . extracellular matrix production and thrombus formation could thus be applied in combination . Because drug solutions can be cleared rapidly from the arterial wail by both convection and diffusion, another potential explanation for the failure of locally applied singledose drug delivery to thus far achieve dramatic benefit is that the residence time of the agents used has been insufficient . Mare than 9055 of deposited colchicine . for example. is cleared within a few hours after local intraluminal delivery from a porous balloon (42) . In contrast . sustained release of heparin from a periadventitial depot effectively reduced neointimal proliferation and local thrombus formation in the balloon-injured rat carotid artery made] (44-46). Heparin in these cases was administered in an ethylene-vinyl acetate copolymer matrix (44) or a polyvinyl alcohol gel surrounded by a Silastic shell (45,46). Similarly, dexamethasone has been shown to inhibit neointimal proliferation in the ballooninjured rat carotid artery model when delivered locally from a silicone polymer implanted in the adventitia (47) . Although surgical placement of drageluting matrices is impractical for most clinical applications, one approach to minimize drug washout try use of percutaneous, transarterial delivery is the use of nondittitnble. drug-eluting auicroparlicles . Similar to biodegradable sterns, these nticroparticks can be manufaclured from a variety of synthetic polymers (e.g.. polylactide) or natural substances such as proteins or polysaccharides . variables such as total dose of drug and kinetics of release ran he strategically manipulated . Again . biocompatibilitY of the material or materials used is a major issue to be resolved because of the potential for a foreign body reaction precipitated by the polymer. Microparticles can be injected of lciently into the arterial wall through a porous balloon (48) or a balloon over asrenr (20) and are retained in the arterial wall as well as in the periadventitial tissue for at least 2 weeks. However, although biodegradable micropartieles that release cotchiciae have been shown to inhibit smooth muscle cell proliferation in vitro (49). no such effect could be demonstrated in vivo (50). In addition to inhibiting resterosis. local delivery of therapeutic agents might also prove useful in other, more acute situations . In a preliminary study, Mitchel et al . (51) reported that acute thrombotic occlusions of coronary arteries after angioplasty could be successfully reopened by balloon dilation with urokinasecoated hydrogel balloons . This approach may well represent an attractive alternative for patients with contraindications to systemic thtmnbolysis or patients with systemic therapy that failed to lyse occlusive thrombus .



6-3 ~

K

.8

rev

-g

a

g

p

8

ill ~3 ~

i°

=

$!t

V

A

VI

e

as ~~ 1111 jjq4!IjIllIh1I1Iih1HIII!Lidui

9p

HIM " TM I



JAM VoL 23. No. 3 April IW1:1731-l4

lovirns, Runs sarcoma virus or simian virus 40 promoter (10 .13,22) . Although the use of these promoters has been associated with varying levels of gene expression, none are tissue specific . hsur-specific endogenous promoters (e .g . . an actin promoter if the target cells are smooth muscle cells) may further refine arterial gene tmosfen by allowing one to resttict gene expcss(on to certain cell types within the arterial segment at the site of transfecxion . The third component affecting tran*ction efficiency is The trammtaial delivery device. The first device used for iMravaseular gene delivery of marker genes such as betagelecwsidase and luciferase was the double-balloon catheter (9,64) . In our experience, transkctinas performed with Ieclismae plasmids . mixed with liposotttes and applied through a double-balloon system in a rabbit model resulted in low 6egneney and low levels of gene expression (64). In contrast, arterial gene transfer with luciferase DNA only (no 8posomes. viruses or other vectors) through a hydrogel balloon was consistently successlhl (22). and mean lucdaase expression was increased by more Yore two orders of magniNde compared with previous experiments using the double-balloon system . Gene expression was achieved with infstion times as short as 1 min . Auturadiographic studies demonstrated that DNA was delivered to the entire media, with the highest concentration in the aubimimal cell layers (Pig . SA), Additional transfeetiaas with the marker gene for nocleonapeelfic belallilacmsidase (Pig . 58) showed that nanhction was restricted to only a few cells, predomi nastiy located in the subintimal layer, indicating that the higher DNA concentration and mechanical faces achieved directly at the haloon-artery iderihce optimize cellular uptake ofDNA. Others (12,13) have used purcus balloons to peefam lam transfer with retrovleuses or planmid DNA . Cutretltly. ail of these devices are also being evaluated for adenoviml gene wanner 172) . hidhect one transfer co st(nnes an alternative, albeit sews wmhersoma, to direct modes of on transfer previallnly described . Cells targeted for gene transfer are removed from the potential recipient and are transtected with the temMbisaM gene . after which the successfully transfected units am selected far relmpla mom at a speeme recipient site in vivo. Both andado ial cells and smooth muscle cells We ban reimplanted according to this paradigm at previenuly denuded sites by at of a double-balloon catheter (81,82?. Similarly. stems seeded with genetically erglneered M dusts ial cells have been used to achieve tranavescular Introduction (27). AMieame ullpauchetidm . Antisense oligonuckotides are short segments of synthetic DNA that are in general chemically modified to resist enzymatic degradation by intracellular or exnaeellular nucleases. Alter cellular uptake, antiseese ofgonucleotides can bind in a relatively specific manner to a complementary messenger ribonucleic acid (mRNA). The action of RNase H on the resulting DNAIRNA hybrid readers this segment of mRNA unavailable for translation. The advantage of aMisenne oligonude-

RIEsEN AND ISNER SITE-SPECIFIC THERAPY

1239

otides is that they can be specifically targeted against mRNAs for those proteins essential for cell proliferation. Some of these proteins are strategically bested near the final common pathway into which multiple . paetalty redundant signal transduction pathways (including growths factors and their receptors) converge . Antisease oligoucleutides targeted against a variety of mRNAs, including proliferating cell nuclear antigen (83), aonmusde myosin (B4) and the proto-oncogenes c-myb (84) and raryc (85.86), have been shown to partially inhibit smooth muscle cell proliferation in vitro. This effect has been shown so persist for several days in cultured cells exposed to aatiseuse far only 2 It . In experimental animal models . antisense digonucleotides administered in vivo have been found to reduce neointimat formation after denudation of art carotid (87,88) or porcine coronary arteries (T.alewshi A, personal communication, De cember 1993) . Simpns et a)- (87), for example, applied c-Mb aatisense oligonucleotides in combination with a pluronic gel to the surgically exposed adventiti of she art carotid artery . The polymer used in this case has the physical property of being liquid at 4°C but then forts a highly viscous gel at body temperature. Morishita et at. (88) mixed antisense oligonucleotides, prepared against cyclin cdc2 and prolifersting-cell nuclear antigen, with inactivated viral particles, tipoaomes and a nuclear protein and then injected the solution into the hnmenof a ligated rat carotid artery . Studies designed to lest The erect of a variety of antisense oligonuclcotides administered percutaneously to inhibit arterial narrowingare currently ongoing o a number of laboratories . In all cases, selecting of the particular antisense strategy is based in part on idontilyltg a targeted mRNA present in low abundance and a eoneaponding protein with a short half-life, Issues that remain to be addressed if this strategy is to be successfully applied clinically as therapy include the amount of oligoaudeotide that can be delivered, the stability of the oligonucleoudes in the vessel wad and, perhmost perha crital inthe ea se of restenosis,timing of administration compared with mechanical intervention.

Systemic Adminialrntbn of Inactive Agents Followed by Localized Activation Photodyamie therapy constitutes a form of local therapy in which the activity of light-excitable photosensitizera is exploited to produce Injury of targeted cells. Photosensitlz. em are relatively nontoxic substances unless they are activated by the appropriate wavelength of light . Hematopor phyrin derivatives, activated by red light with a wavelength of 635 am, have been used for the treatment of neoplastic disorders (89). These substances accumulate in the proli er ating tissues, such as tumors, and have also been ahowrr to he present in higher concentrations (n atherosclerotic plat ue than in normal arteries after systemic administration (SO) . Cultured smooth muscle cells from athetueclerotic plagtes also show a higher sensitivity to hematoporphytins than



1240

RIESSEN AND tSNFR SITE-SPECIFIC THERAPY

S. A, Automdiogmm of a mbbh carotid artery performed ngioplasty with a hydrogen balloon cooled with "Sradklebeled piasmld DNA . The highest density of silver grape indicating the presence oftediolabeled DNA Is Panad In the htyers of the arterial wall bordering the lumen . d, Histologic section of ∎ rabbit carotid artery 3 days after tmnsfcetiaa with the corker gene for nuclear-epeclftc beta-galacsosidase . A hydra el balloon was coated with Mpg plasmid DNA and Inflated in The artery for S min at 9-arm pressure. Three subimimd calls MUMS the gem for nmlear•apeeIftc bear-galaesosidase can beientified by the darkhim omlear staining (arrow). Additional homunoemiolao with antibodies against muscle acdo (HHF-33) shows the prCdnmimoce of amotth muscle cells in the arterial wall . Hematuxylin-coin.

::M-

smooth muscle cells from normal arteries (91) . Catheterbased light delivery systems are now available for intravascollar photodynamic therapy (92) . Photodynamie therapy involving administration of a hem atoporphyrin derivative, followed by flash-lamp irradiation delivered through a quartz fiber catheter with a didbsing tip was shown to inhibit necinlimal formation in a rabbit model (93). This effect, achieved when the treatment was administered immediately or delayed for I week after balloon injury . was associated with extensive medial smooth muscle cell necrosis, indicating that medial smooth muscle cells also take up photosensitizer directly after arterial injury . Treatment at later time points was shown to target mom selectively the neointimal tissue . It remains to be shown whether medial smooth

. 23, Nn 5 JACC Val Ape 1991:1276-44

Flgare 6 Tq, Uauastsd rat carotid smty 3 weeks alter bdkoa injury demonstrating a think aeoislina OW). lilies, Section era rat cmadd artery 3 weeks after beflom &cud lkat foliwed by radiation with low dates dWasvkdet light goomated atom a diode laser in the absence ofany photansiGum . The presence of only few cell euclel in the medic (M) Indicates smooth marcle ad necrosis. The artery dams not show any rm®timd proliferation . IBM - Internal shark memhmm,

muscle cell necrosis predisposes to vascular complications such m aneuryemal dilation or perforation, but clinical experience with laser balloon angioplasty (94) suggests this is unlikely to be a frequent concern . A second issue for which there is limited data (9S) is the possibility that proteins released from necrotic cells might represent potent mitagens capable of Initiating a secondary wave of cell proliferation . The clinical use of hemelopoephydne in the past has been limited by a cutaneous photosensitivity that can lest for several weeks. To reduce the systemic photdoxicity, lower doses of new hematoporphysin derivatives have been administered locally through a porous balloon (96). Likewise. chlorodominum-sulfoaated phthalocyanin inhibited neolmti mal formation in rats after balloon injury and also produced necrosis of the media (97) . This substance is nonphototosic and is also preferentially taken up by neointimal cells (9d). Pscealen represents an alternative photosensitizer currently under investigation as an agent to be administered in con-



JACC Vd. a . Na. s Ap 71954:1214-44

RIESSEN AND ISNER Srt0$PECIFie THERAPY

junction with ultraviolet irradiation . This photosensitizes . routinely used for the treatment of dermatolngic disorders . has limited systemic toxicity . A single exposure to psoralen ultraviolet A (PUVA) radiation was shown to result in complete stasis of cell proliferation over a 211-day period

without cell necrosis (99) . Fatally, preliminary studies from our laboratory have suggested that very low doses of ultraviolet radiation generated from a diode laser and delivered locally to an isolated segment of rat carotid artery in vivo can inhibit smooth muscle cell proliferation and cause medial necrosis even when used alone, with no phoeoacdve agent (fig . 6) (100).

Syetesnic Admiofstrailon of Agents With a Specific Affinity to Proliferating Smooth

1241

even in the case of a systemically administered agent designed to be activated locally, there may well be a role for local delivery. Preliminary experience in our laboratory has demonstrated that local administration of the fusion protein to a segment of balloon-injured rat carotid artery resulted in markedly reduced netJmtimal formation with clearly reduced systemic side effects (till) . Thus, although this treatment strategy was originally designed as targeted systemic therapy . its future clinical use might well be local delivery through an intravascular delivery system . Several questions, however, must first be resolved. These include foremost the optimal ti,ae point at which to apply the toxin . Application immediately after angioplasty might be most practical, but uncertainty with regard to the time interval between balloon injury and the onset of increased expression of receptors in proliferating cells remains a concern . Second, the long-term consequences of intimal and medial cell necrosis, or both, as discussed previously, are unclear .

Muscle Cells

Proliferating smooth muscle cells, compared with quiescent cells, often express higher numbers of cell surface receptors for a variety of growth factors . Differential expression of growth factor receptors may provide the opportunity to use cytotmdc agents that specifically target proliferating smooth muscle cells. Targeted delivery of cytotoxiae to specific cell surface receptors may be accomplished with recombinant Ibsion proteins . These chimeric agents combine a potent toxic with peptide ligands to cell surface receptors. After receptor binding and internalization of the prtltein. action of the toxin results he cell death . For the receptor

rargetirtg component, several growth factors have been used, Including misinforming graMh factor alpha (10l) . epidermal growth actor (102). basic dbroblast growth factor (103 .104) and oteleukio-2 (1031 . Toxins that have been used to complete 0u hybrid include a modified Pseadomonas exotoxin (101 .103), diphtheria toxin (102) and saporin (104), a toxin directed at ribosomal RNA . In cell culture, these fusion toxins preferentially kill rapidly proliferating smooth m scle colts, compared with quiescent cells (101,104) . A recombinant CYtomxmt specific for the epidermal growth factor receptor was also shown to prevent smooth muscle cell outgrowth from human atheroscle otic plaque in culture (102). We (106) and others (104) have used the ballooninjured rat carotid artery model to study the effects of recombinant fusion proteins on necintimal formation in vivo . Systemic intravenous delivery of afusion protein was shown by Casace0u et al . (104) and more recently by Picketing et al . (106) to inhibit iatimal proliferation but was also associated with systemic toxicity, manifested principally as weight loss . Such systemic toxicity is mot altogether unanticipated because of the fact that growth factor receptors are not exclusively expressed by proliferating smooth muscle cells but are also present an cells from other organs such as the liver and gut . Previous nonvaacular use of these fusion Dancing suggests that the doses used in the rat would be umiteely to exhibit similar toxicity in humans . Nevertheless,

Conclusions

Local drug delivery represents a potentially valuable adjunct to the treatment of restenosis and other vascular diseases . Several prototype devices have been developed and tested in vivo. Further modifications are expected to achieve more efficient delivery without concomitant tissue damage . Additional studies will be required to demonstrate the logical but yet unproved concept that the efficacy and safety of pharumcalogic substances conventionally administered systemically is superior when delivered locally and to clarify the efficacy and safety of alte native, locally targeted strategies, such as g m therapy, amtisense otigmtucleotides, pholodynamic therapy and recombinant fusion proteins . We apsssfuay ackeowtsdge the coambutimu of Guy Ledere. J. Geowey rtakesins. saemni Takeshlla. Censlephar Passers . Takayuht Mature. Lawrence Weir . Marinate Reactor and JadyNS JekaWwski to wart de . scribed in INS rc'dn9 .

References 1 . lp J. Foster V . Badiaea L . Bad®en J. taubxan M. Chastise 1 . Stmaimmss of acceknad ethe mchresh : rude of vascular Injury was smash muscle cell pmBarniae J Am Coll Consul 19%7:166M7 . 1. Cassedh W . Migaunn of somdk museh andeudahelial sets. CSilisal avenue o esrsoosis. Rreshnion 1992 :16:727-9 . 1 . Schlrane RS . Miens DR, Topd E3. The resemuds paradipo revisited : m ahemaive proposal for cekalsr s eelaotses . J Am Call Consul !9n;.-J : 1181-93, 4 . Hooke H. Slrdaehuetder T . Oberhoff M, Ban E . Karsch KR. Tire course of smooth muscle OR pmtteraioa in the lotion and media of arteries following experimental angiopla'ry . Circ Sec 1990:67:6SIA. . S . Fickrrng JG . Weir G Jdmsowski 1. glearmey MA, four 3M . Proffer alive sdivily in peopherah and mm-my atherosderotie plaque among patients uMmpsing percumeous nwaseuhdmion . I CRn Invest 1993: 91:1469-gI. G Mach KL, Wikmky RL . Hathaway DR . Novel drug ad device

randulauioas tee targeted prevendan of resuesods . Cardio Intervention 1912:1 :1 h16.

R

n

c

fi

Oil

r

ll ~,ag "R ~L ~ ~ ai'p'

Q xfk

PF V.

CPF

r~



$

g

II!1

a

s~

e .c

$3

Ik

m

-

~

-H

IRK

~~

8>7e ~_

OF F [

,~=

s ~r r

k

•, ~

-~--.{^

lp

E W

(;T

zNR

r

all

~g

m

o =

g~

~ B

PL



JACC VOL 23. No. 5 Aped 19Pk1234-44

en uhicine motion die not bl0 b reslenmis in amerasckradc rabbit kmoml -Modes [abstmctl. Circulation 1993:6 Soppl 11311 . 51. M1lU IF. Aen iMA .FtanDB.aaLL99Wtreatmentotinaacotolmry thmmbm and mumbo= amupt closure forming mikd comnmy mtpoplasty wilt urdduce-cwled hydrvgel nhl gum labstmall. Ciau114e019ylAS Sorel 1:1.569 52. Zhn N- 1.400 0 . Lin Y, De6e R. Syltenk goon enpoegeion die, imrame tuna DNA delivery teas hole mice . Sale- 199! :261 99-1 I . 33. ismmdeard P. Ions M. Vented . 1. Crystal RG. h viva 910, ums6r and eapocesiar in award uninjured blood -Is 1104 repluatioodegcka reeamHaut mlamvhus vectors . Ch Rev 1993:T2:113ra. 54 . Lynch MC. Closer MM . Otome WRA. Clowns AW. Miller .A0 . Lm"- expansion d turner Wanton. dentorsousc in -al. m call muiele cell d rats : a MOM for gene therapy . Ale Non Aced Sid USA 1992$41138-o,1 . 55- Tdeddta S . Lom000e DW. Ledea G. Rcberiog JG. tang JM . Tire motors of kamm growth haruuoe sardiuo fid1 wing arterial gene Ieurkr. dmodd foevmeular geec Ebenpl • Inowml . I Am Coil Camiu IRMd1:179A . Sir Nobel EG . Rzulo G. Nebd G . 7kamducdm of,foreign bistmonparibiliity gene ido the material well ionLoul -the. eroe Nod Aced Son LISA L992d9:5157-6i . 57. Mrhl Eli. Yang ZY . Pkmz 0. m al. Reeenbiuan RbtHast grmuth Deue-I proamhs trthow byperplais lord mgioneeesic in ner{es is vivo . Nakuc ;362m45-6, 193% 58. Miehdksitz D, Halsey 0. (ken H. Co0didored inbbhlen al uwofoogrdaaand dcrEpm]ifgntip .byoxmRrmmr-a edLisemwd rd p5!. Cell 199962:671-80. 59. Take" S . Timers LP. Amino T. at .L In vieu evidene otenhmad myarsta6 fdkwiog Sad umdalgene trunderdnheplaadd encod6g vucuhr eod0NeWl Feuds t :.caa Io mound CirenWlun 1993d Skppl 1 :1476 G0. Diebel DA. Gate ameda in the uemnen of thrombosis . T6romb Removal 1993.711:19"l. 61. Diehek DA. Nrwbrm 0, Depen SIR Anderson WF . Enlmmettem of thelbriedylic nct[diy, d deep eoddhe8d tens by reuoviru4mdiuled peg, weaker. Blood 891:7153141. 62, P9dmak EM, Whitehill TA, E9hteru D, Wil duos WD . Memhn LM. Franky JC 11111"'d eaprmaton d amoMnam hums EPA in cold . mi dernimw&tlwtinlrst6 seedorvmyingcoo lidens d re voaisi pace lmrkr. A N Sung 19925100446-.12. 63 . LoEW, KUOML . Dir1et DA. Expaeiandmmchomd uihi.aseia ft .pld emla4id all neabrme. J Sid Clam 1992267i13020-7. M. Ledoe G.G.l D,'hkeaSu& Ndd S . Wdr L.ImuJM .24aunndm .4.101 Joe binder in a edible PadeL Egeieaty to normal end .434-44. hdko&4hud ubmmelerede adeeim . I Ceo invest 119040 m,PAeftm.k .WILD.WeirLTalmshh0S.LosordoD,knrled . rLipme m- . .- ^,e0 pale harder ids, human vascular seam tnuele ed4 ChcobA n 19Mtp413-21 . 66• Hmoley lichron P, CUM= V. Gebey'ehu G, James 1, Fugue PL. Llpokmmhe° raepm: a mw. Idaho, ofoiwq pabsuwek bgP acme InmiMim moat . From 199J ;IY73-9. 47. MulISmBC.Theboskmimcedgeneliteracy. ScicnapJ13;2W :92632. . Gene lmufer by reumvm vecos 4L MMer DG, Adam MA, Miter A0 stems mfr in edit tin me actively rephrasing at do flow or adhatain . Md Cal Bid 1989:10:4239--4L al Tikabla S. ad 0 . Lecka 0. et W. lngeued por expression folowhi Eposmuinediew wind gone hander asmdnad Was imt . Mg ntemh nude all pedietbe . In view and in vivo Austria in a rrbbh mold o unende injury. J Clii Invert 109490142-I . ad-Perrkwdel LA Mabeh L Aaicmm A M, grand P. wide 70 . B rJNw md w crew sl. eM murder ad horn . . Ikera l~d In I:9 21. WSleed J& lronm ME. Grad ED. MeideD RA ResnmMwm adesovi. wa 6 m clinches vaster him in vivo June transfer and ran be pekme, 1611y dsecled at vmahreadothelion oranmath muscle alt [nhnhactl . Circulation 1992;86 Sup el 1 :1473. 7L Sew PG, Fd&mm L So--= JY. e t eL Local deivery ofadmmvinnfee

bd0010la~ hmnc ~ G. Circulation 19345 uppe 1 :1443.. 73 . Leo SW. Ttepul BC . Raft 33. Vivmni R. 1 ;X11011 DAX h vimo

51535 4 AND ISNER SrrE-SPE= THERAPY

1243

tdesnr•h al valor-mameredpue trader into baBoort-injured rote carotid onories . Cite iks 1993;77:797-907 . 74. Here J . Grid RD . Adenovirneunedise d hander of low dinghy Ii. poprolein reapmr gene acutely aaelaales cholesterol clearance in normal tdoo Pat: Neil Aced Sd USA 1993 :992512-6, 73. Yen. RL . Ademvhus -am igloo d r brain. Trends Neurord 1993 ; I6 257-3. 76. tlanda Y . Ived K, Gchids T. Increased expression of DNA coint,r dead ant indoor 9ogeht b aduh to liver. ScLnce 19W.243,37". 77 . Mwishim R, Gibbons OR, Rmeda Y .Ogiham T. Coon YJ. Novel and eEeeIivegone tmatfcr lee0nigro cur At* of vouches grog sniotendn system, J C1h Invest 1993:91 :258 5. 78. Cotten M. Warner E. Zmkrdd K Phillips S. Cur[el DT, Binmlel ML. Hiheidceacy readceouedaled delivery of small and horn t48 kilo. base I gene eaamme union the endmamedsmptioe activity of d*cdse or chemlolly [activated adeoovirus psrtidea Pro ; Had Acrd Sd USA 1997 d:fi09e-9. N cool .WoenaE.Cateo1,etd .Dueetievnuwgentmnsireumaay epidrelium empbyirm ulenovinrptdylyeiueDNA cumplexr . Hum Gone Titer 193W :17-31 . 90. RV J . One FRI. Gene t,anfer into coabbdced std primary fiibroblast an firms: compataan of trmsfection mahods toed proorrem . Rioteda;gues 1992 :13:995-603. 51. Nobel EG. Plants G, Boc PM. Stairs IC. Nobel GJ. Recombinant a . elprcasm m duo within eadodrllal calls of the asterial wall . Seirnoe 19bbI,3441342-4. 82 . Nmoz G, Nobel HG, W6d GJ. Intreducdmaf vmw6r wudh trued, alu eapresdngrecanbioml grids a vivo. CSrcdation 1991 :53:578-83. 83 . Speir E . 4stain SE labibrbondsmoah muscle call proliferation by an ar dxeme oEgodeoa5rwdeotidl Isolating The Meander RNA enwdhtp prdifermir9 can nuclear amigen. CSsculalioa 1992:M:J38-47. 4. lower t1. Resistant RD. Andsome nonwoude myeom heavy chain adoosyboigonocleoudussupports smooth muscle call prolfemlionin uitra. Cite Res 1992:70,835-43. 85. giro & Fu YM, Yu ZX, Einstein SE . Inhibitory effects of artherom dtiodeosyaaekdides tmgelipg c-eyc mRNA on smooth muck cell prodferadon end migration . Pmc Neil Aced Sol USA 1993;90:6544. m. SbiY .Hutgb40gn HG, Hill DI. Zeleneebi A. lbw,eergidutieieofo-myn espreukar by adense olgeourkonidea inhibits proliferation albumen smooth nmcleeoW. Circulation 19931:It90.9. S7. Siring M. Edelmn . ER, DeKebmo IL, Linger R. Rosaabotp RD. An iscost c-,gb diSeeoIck01dea inMbE 1181011 nskrial month musk cell accmulNm in vim. Motor. 1992,35947-70 . m. Magi hit . R, Gi66oeoCH .556-c ICE . ta W. Si might hiududd delivery of mluonu tdc2 hhirr+e and prdbfenl:Wrel nnlam entiJn 014000 .rplula Arc elomide remarks in chronic mhitidm d maludmd hyp :m Need Mod Sid USA 1903:99474-A 5% Prom OR. Lin CW. Bularr R. P6omtherapy with Mmetopnrphyein derivative 6 the mountain of .epukkl operational son caulmourat of flit bladder. N he 3 MW 1997;31'RH31-5. lexpetimm. 91 SparnJR. Savor J. ShmJh[re D, Addin S . Florescence a to atMmueloar pNgmn with kmdopleplyrW doe0lEve . J Clin inner 19U317I :193-9. 91 . Derives PC. Iech04er T. Belt E. Respmses of allured nNah muck eels he Banta nmder9ldemdc enmfm end primary sterrmim5 6sions oiler pho5madin an: implistatiew l phomdymnde therapy d vascular eenow.1 Am Call Cetdbl 1940 :15:1543-50. 92. -Fang G, flyepm 5, Schndper 1H . Gisrom SL . Application of phow , dynaff& thmupy m the trenbmel al lhumdemlic plaques . Neun:ur gory 1993 :30 .488-43. 93. America T. Used M, Ameniyo T, d d . Pathological e9ets d ba0am Xclg11as5lamp nomination u plolodytmmkthmapy fmtleprevemion of astenm6 [abstau1 . I Am Cod Cordm1 1951 ;21:IISA. 94 . Won GJ, Pamamm RM. Jentior RD, et al . Lam balloon mgiopbdy : clinical, mgiagrWhie and hi mkp ; rondts . J Am Cal Cordial 1991:18 : 193-202. 95. Gejdusek CM . Schwartz SM. Comparison of intmcella6r and eximalhda mgegenk activity . J Cell Physio1194 ;121J16-22 . 96 GmtschiorP.ErdemciA.GellmaF .ad-Selective hemmrparphyrin detivadve SHMDI application i arterial vessels usiru a porous ballon catheter,eadre in equivalent levels m cneopued to highdose systemic administration. Z Icerdk11991 :80:731-45.