`
`David D. Lawrence, ]r., MD 0 Chusilp Charnsangavej, MD
`- Kenneth C. Wright, PhD 0 Cesar Gianturco, MD - Sidney Wallace, MD
`
`
`
`Percutaneous Endovascular Graft:
`
`Experimental Evaluation‘
`
`An arterial endovascular graft was
`constructed by wrapping a Dacron
`cylinder around the Gianturco ex-
`pandable metallic stent. The device
`was passed through an 11-F Teflon
`catheter into the normal abdominal
`or thoracic aortas of nine dogs. At
`follow-up of 7-35 weeks, all but one
`graft remained patent. At necropsy,
`the grafts were almost completely
`covered by neo-intimal prolifera-
`tion. Similar proliferation was ob-
`served between the graft and the
`wall of the vessel.
`
`lndex tenns: Aorta, grafts and prostheses, 56.45
`- Arteries, grafts and prostheses, 94.456
`
`Radiology 1987; 163:357-360
`
`' From the Division of Diagnostic Imaging,
`Department of Diagnostic Radiology, Universi-
`ty of Texas System Cancer Center. M.D. Ander-
`son Hospital and Tumor Institute, Houston.
`From the 1986 RSNA annual meeting. Received
`November 10. 1986: accepted January 3, 1987.
`This work was supported in part by the I. S.
`Dunn Research Foundation and the George A.
`Cook Memorial Fund. Address reprint re-
`quests to D.D.L., Department of Radiology,
`University of Texas Medical School at Houston.
`6341 Fannin, Suite 2130 MSMB, Houston, TX
`77030.
`° RSNA, 1987
`
`THE Gianturco expandable metallic
`stent has been placed successful-
`ly within the venous system and tra-
`cheobronchial tree in humans, indic-
`ative of its utility in expanding
`narrowed lumens (1, 2). Endovascu-
`lar stents have also been placed in
`the arteries of animals, with eventual
`incorporation of the stents into the
`vessel walls (3, 4).
`Another use of the Gianturco stent
`is as a vehicle for the intravascular
`placement of other materials. We de-
`veloped such a modification to allow
`intravascular placement of a Dacron
`graft, using multiple Gianturco stents
`as a superstructure by which to an-
`chor and support the graft. Such a
`device may find use in the treatment
`of aneurysms, as it allows creation of
`a new lumen within the aneurysm
`while sealing off the aneurysm out-
`side of the device in a manner similar
`to a surgically placed graft. To evalu-
`ate the feasibility of the placement
`and short-term patency of this de-
`vice. we studied it in the normal aor-
`tas of nine dogs.
`
`MATERIALS AND METHODS
`
`Stent Construction
`
`The endovascular graft consisted of
`multiple stents in tandem connected to
`each other by metallic struts. The first
`stent was designated the lead stent; the
`last, as the trail stent (Fig. la). The Dacron
`tubing was wrapped around the outside
`of the middle group of stents, internaliz-
`ing them within the Dacron graft. The
`lead and trail stcnts acted as anchors for
`the graft, while the internal stents served
`to open the Dacron tubing when the de-
`vice was released from the catheter (Fig.
`1c).
`The lead and trail stents were con-
`structed, as previously described (3), of
`stainless steel wire with a diameter of
`0.016 inch. Each stent had six bends and
`no side barbs. The stents were approxi-
`mately 2.5 cm long, with a diameter of ap-
`proximately 2.5 cm when fully expanded.
`The internal stents were similarly con-
`
`structed, but wire of thinner gauge was
`used.
`The graft tubes were made from thin,
`woven Dacron sheets. The graft was ap-
`proximately 1.0 cm in diameter and usu-
`ally 5.0 cm in length (range. 2.5-9.5 cm
`long). The ends of the graft were attached
`to the limbs of the lead and trail stents
`with 6-0 suture material. Once construct-
`ed, the devices were loaded into 12-F Tef-
`lon cartridges and gas sterilized with eth-
`ylene oxide.
`
`Graft Placement and Follow-up
`Study
`After intravenous induction of sodium
`pentobarbital into the dogs for general
`anesthesia and after systemic hepariniza-
`tion, an 11-F stent introduction assembly
`(Cook, Bloomington, lnd.) was intro-
`duced into the normal abdominal aortas
`of six dogs and the normal thoracic aortas
`of three dogs by a femoral arteriotomy
`procedure. Aortography was performed
`to document the size of the vessel. The
`grafts were then introduced through the
`cartridge and released under fluoroscopic
`guidance with the technique previously
`described for placement of a Gianturco
`stent (3). A second arteriogram was then
`obtained to enable evaluation of the im-
`mediate postplacement results. The ani-
`mals were not given anticoagulants rou-
`tinely following graft placement except
`for heparinization during follow-up stud-
`ies.
`
`Aortography was performed 1 week af-
`ter graft placement in the first six dogs
`and prior to death in all dogs. No imme-
`diate or delayed complications related to
`the graft placement were observed clini-
`cally.
`The dogs were killed by exsanguina-
`tion under deep pentobarbital anesthesia
`at 7 weeks (two dogs), 8 weeks (four
`dogs), and at 10, 16, and 35 weeks (one
`dog each) after the procedure. At necrop-
`sy, the section of the aorta containing the
`graft was removed, together with the kid-
`neys, if any portion of the device had
`bridged the renal arteries.
`
`RESULTS
`
`All grafts, with one exception, re-
`mained patent during the periods of
`observation. There was no evidence
`
`Edwards Lifesciences Corporation, et al. Exhibit 1029, p
`
`
`
`Edwards Lifesciences Corporation, et al. Exhibit 1029, p. 1 of 4
`
`
`
`Edwards Lifesciences Corporation, et al. Exhibit 1029, p. 2 of 4
`
`
`
`.1.
`
`b.
`
`C.
`
`(.1, ti) l(.2dioy,r.'ipl1s from lateral ;ibdomin:xl.1orto;;raint of dog 8 lv\'L‘L‘i4:ii1itt?t' placunicxit of ciidova.sml.:ir graft. Dacmu portion is be-
`Figure 2.
`twt-on two long .'trrn\\‘s.Tl1crv: is mild lumiiml l\M’1‘t')\\’it\g in the region of thegratt as. :1 result of m.-o-intimai }~rolit't-ration. Note occlixaion
`(short .'ll‘l‘i':\\') of nrigiii.» of t..~S lumbar nrzvricss by the g,r.it‘t, with reconstitution by mllnterals. (C) Anterupostcrior nortograni of sanie animal
`~lm\w,~ lL‘J\l 1-lvnl tiridgixig let’: renal arter)-‘ but not occluding,‘ it.
`
`sin was Ul).St‘.I‘VL‘ti in the nc~o-intimzi of
`one gm ft. Tlw degree of organimtion
`did not appear to be t't.‘lOit?L‘l to length
`01’ time that the graft had been in
`place, as tlcrisc organization could be
`SUCH at 7 wocks, and loussc organiza-
`tion was Lltflltttltrilrillttcl at 35 weeks,
`:1 ml tho various tir:grccs could be ob-
`S-t.‘!'\‘£.’(l in lllt.‘ smnc i;mt't. l‘-Iviclencc of
`I't‘\.‘.'lll(illZ£)tl0ll in the neo-intimn was
`})I't.‘.\t.‘I‘tl in $t.‘\'t.‘r£ll spccinwns (Fig. 4).
`Slt‘l1(_3f§l5}1l§‘;l“tli‘lCiitltt‘I10llt;l1 to pro-
`(iucm l1_vp:.~rtrupl\it-ti collnts-r.il.-; dovet-
`.opt.~ti in two tgrgiftzc. Ono stcmisis was
`related to usv of an ovcrsiz.<.'d graft,
`which liL‘i)l‘l)' 0L‘Cltt(iL‘Ll the aorta. The
`other was mused by fziiliirc of the
`czimlsit mid of tho l.);u:r0n to open
`l'ull_\', d.\ it \\'.'lS not tiricig;L*dt\y;ii1 in-
`ternal stcnl. Small, urg;1rii7.eul throm-
`lwlit‘ \’t“;:t*t;tti(>n_s \\’t'rc pros-.'i1t 011
`tins of llw tiv\'iu-5 (at ,7 \~.:'-.-wigs and at
`33 \\’L*t"l\I~':l but \\'t‘l't' not .‘~'t,‘t,'l1 in .'iny of
`thv other y,ml't_-.
`l'l1t.‘ l.).it‘rtm i;r.‘il't that did not re-
`
`\.’,.l..ns.» I-.’:
`
`l\l..n~|....-"2
`
`main patent had been disrupted by
`heat from inadvertent autoclaving
`before insertion. At necropsy. 16
`weeks after placement. the graft was
`found to be laiyered against the anter-
`olntcrat wall of the vessel and, inter-
`estingly, was completely covered
`with nco-intima.
`
`DISCUSSION
`
`The goal in the cievclopmcnt of the
`I;‘X1\i(J\’£l>‘t.‘Ltli3f graft was to produce: :1
`device that could be placed by it
`trzirisczitlxoter approach to treat an .m—
`eurysm. The device should pass
`through a relatively small catheter
`and expand to fit the lumen of the
`vesaeel. When in place, the device
`should then act in the saiiiuz way as .2
`surgically placed graft, providing‘ a
`new conduit for blood flow and sup-
`porting the \\’t;‘£ll<L'{l\('L"l v;i:;cul.1r wall.
`Our prelimin;i.r_v ¢'.'\'&'tltl.'itit)n of the
`u:icl(wa.<ctilar device in normal aortas
`
`of clogs duiriciiistrates that our design
`was easily placed through an 11-17
`catheter into the vessel. It was shown
`that the grafts occluded the side
`br;mcl'n.-2; tliey tiridgecl as well as pro-
`ctucing, a nco-intima similar to that
`seen in surgically placard gr:\t't.~:; (5-
`11). The vessels bridged by stunt
`wires alone dcnionstratud potency
`without cvicleiice of distal emboii.
`This feature will allow pl:iceinont' of
`the ltratl stem across the renal arter-
`ies, pcrmitting placcn1cr1toftl‘ic Cc-
`phntic portion of the graft just below
`the ltrvul of the rt,-nal ai'tt:r_v. Space
`between the ggraft wall and the native
`intima was t'ill:'.-cl, stxggc-stiiig that the
`anenrysm nmy fill in with .1 t'ibropr0-
`till-rativc tissiic respoiisu. Our future
`0.\PL‘I'llI1(‘l’l ts will ¢;‘\’alLlOiC this.
`l\-lost of the coniplicatiomz (perform
`lion, stciitvsis) that occurred were re-
`latml to technical factors and are pn-
`tontial ly zwoitlutila by cnreftil
`attention to graft plaiccmcnt and to
`
`R 241 inlnuu o ’1§()
`
`Edwards Lifesciences Corporation, et al. Exhibit 1029,&
`
`Edwards Lifesciences Corporation, et al. Exhibit 1029, p. 3 of 4
`
`
`
`tailoring the graft to the diameter of
`the vascular lumen. The presence of
`vegetation on two of the stents is
`bothersome and needs further evalu-
`ation.
`
`While this endovascular graft was
`
`designed primarily for evaluation in
`aneurysms, other potential experi-
`ments include its use as a postangio-
`plasty adjunct and as stents in the ve-
`nous system and other organ
`systems. I
`
`a.
`
`b.
`
`(a) Cross specimen of device in abdominal aorta. There is nearly complete cover-
`Figure 3.
`ing of the device by neo-intima. (b) Cross specimen shows lead stent wire bridging left renal
`artery. The vessel remains patent, and there is no neo-intimal covering over the portion of
`the wire bridging the artery. The Dacron graft and its junction with the native vessel (arrow)
`are completelycovered with neo-intima.
`
`Acknowledgments: We thank lrene Szwarc,
`R. T., and Raquel Collins. B.S., for their invalu-
`able technical assistance, and John Kuykendahl
`and Jay Johnson for photography. We are
`grateful to L. Clifton Stephens, D.V.M., i’h.D.,
`for reviewing and photographing the histolog-
`ic specimens and to Katie lsch for secretarial
`help.
`
`References
`1. Charnsangavej C, Carrasco CH, Wallace 5, et
`al. Stenosis of the vena cava: preliminary as-
`sessment of treatment with expandable me-
`tallic stents. Radiology 1986;161:295-298.
`2. Wallace M], Charnsangavej C, Ogawa K, et al.
`Tracheobronchial tree: expandable metallic
`stents used in experimental and clinical a -
`plications, work in progress. Radiology 1 86;
`lS8:309-312.
`3. Wright KC, Wallace 5, Charnsangavej C, Car-
`rasco CH, Cianturco C. Percutaneous endo-
`vascular stents: an experimental evaluation.
`Radiolog l985;l56:69-72.
`4. Palmaz]
`, Windeler SA, Garcia F, Tio F0,
`Sibbitt RR, Reuter SR. Atherosclerotic rabbit
`aortas: expandable intraluminal grafting. Ra-
`diology I986; 1601723-726.
`.
`5. Takebayashi J, Kamatani M, Namba S. Kata-
`gami Y, Hayashi K. Early reparative process
`of implanted arterial prosthesis: a morpho-
`logical and hematological study. J Surg Res
`I974; l7:l02-ll0.
`6. Sauvage LR. Externally supported, non-
`crimped. external-velour. weft-knitted Da-
`cron prosthesis for axillofemoral, femoropo-
`pliteal, and femorotibial bypass. ln: Wrig t
`CB, ed. Vascular grafting: clinical applica-
`PSG. 1983.168-l 6.
`tions and technicgues. Boston: John Wright-
`7. Herring MB, Dilley R. Jersild RA, lr., Boxer L.
`Gardner A, Clover I. Seedin arterial pros-
`theses with vascular endothe ium. Ann Surg
`1979; l90:84—90.
`'
`8. Mason RC. The interaction of blood hemo-
`static elements with artificial surfaces. Prog
`Hemost Thromb 1972; 1:141-I64.
`9. Warren BA. Brock LC. The electron micro-
`scopic features and fibrinolytic properties of
`"neo-intima." Br] Exp Palhol I964; 45:6l2-
`617.
`10. Herring MB. Endothelial seeding of blood
`flow. ln: Wright CB, ed. Vascular grafting:
`John Wrig t-PSC, 1983; 275-314.
`clinical applications and techniques. Boston:
`ll. Sauvage LR. Ber er KE. Wood 5), Yates SC,
`Smith )C, Mans ield PB.
`lnterspecies healing
`10 :698-705.
`ofgorous arterial prostheses. Arch Surg 1974;
`
`
`
`a.
`
`Figure 4.
`(a) Photomicrograph of cross section of abdominal aorta shows Dacron
`graft (short arrow) covered by thick neo-intima on its lu-
`minal surface. A layer of cells resembling endothelium is seen at the luminal surface 0
`fthe neo-intima (long arrow). The large smooth holes
`in the neo-intima represent a cross section through the internal stent wires. (b) "Pleated" graft has longitudinal peaks and valleys. Note rel-
`al surface. There is also recanalization within the
`atively more filling by the neo-intima in the valleys, in attempt to produce an even lumin
`ma.
`neo-intima and in the fibroproliferative response between the Dacron graft and the native vessel. Arrow points to region of the native inti-
`
`360 o Radinloev
`
`Mav
`
`1987
`
`Edwards Lifesciences Corporation, et al. Exhibit 1029, 1%
`
`Edwards Lifesciences Corporation, et al. Exhibit 1029, p. 4 of 4