Canadian Society for Vascular SurgerySociété canadienne de chirurgie vasculaire

EXPERIMENTAL LAPAROSCOPIC AORTOBIFEMORAL BYPASSFOR OCCLUSIVE AORTOILIAC DISEASE

Yves-Marie Dion, MD, MSc, FACS, FRCSC;*† Félix Gaillard, MD;* Jean-Claude Demalsy, MD, CSPQ;‡ Carlos R. Gracia, MD, FACS§

From the *Department of Surgery, Hôpital Saint-François d’Assise, the †Institut Biomateriaux de Québec, Université Laval, Quebec, Que., the ‡Centre Hospitalier Régionalde Rimouski, Rimouski, Que. and the §California Laparoscopic Institute, San Ramon, Calif.

Modified from a paper presented at the annual meeting of the Canadian Society for Vascular Surgery, Montreal, Que., Sept. 15, 1995

Accepted for publication May 9, 1996

Correspondence to: Dr. Yves-Marie Dion, Department of Surgery, Hôpital Saint-François d’Assise, 10 de l’Espinay, Quebec QC G1L 3L5

© 1996 Canadian Medical Association (text and abstract/résumé)

OBJECTIVE: To describe a totally laparoscopic technique for aortobifemoral bypass to treat aortoiliac athero-matous occlusive disease.DESIGN: A feasibility study.SETTING: A university teaching hospital.SUBJECTS: Six piglets weighing between 70 and 80 kg were submitted to a totally laparoscopic retroperi-toneal aortobifemoral bypass, performed through six trocar sites, with abdominal suspension and a gaslesstechnique. No minilaparotomy was performed. After systemic heparinization, the infrarenal aorta wascross-clamped and the aortic bifurcation stapled. An end-to-end aorto–prosthetic anastomosis was per-formed. Retroperitoneal tunnels were created to allow each limb of the graft to join its correspondingfemoral artery by a conventional anastomosis.INTERVENTION: Totally laparoscopic aortobifemoral bypass.MAIN OUTCOME MEASURES: Duration of the procedure, intraoperative blood loss and operative complica-tions, bleeding in the immediate postoperative period. Evaluation of the aortic anastomosis at autopsy.RESULTS: All aortobifemoral bypasses were completed in less than 4 hours. Intraoperative blood loss didnot exceed 250 mL. No intraoperative complication was encountered except occasional bleeding at theaortic anastomosis upon releasing the arterial clamp. This was controlled with a collagen sponge (threecases) or extra stitches (two cases). The animals were observed for 15 minutes before sacrifice. Autopsy re-vealed a normal aortic anastomosis in all cases and a normal progression of the limbs of the graft under theureters in the retroperitoneal tunnels.CONCLUSIONS: This animal model demonstrates the feasibility of the aortobifemoral bypass through a la-paroscopic approach. The retroperitoneal anatomy of the piglet is similar to that of man. Aortic surgery canbe conducted as for the standard technique. We used a similar approach to perform the first human, totallylaparoscopic aortobifemoral bypass with an end-to-end anastomosis.

OBJECTIF : Décrire une technique totalement laparoscopique permettant d’effectuer un pontage aortobifé-moral comme traitement de la maladie athéromateuse occlusive aorto-iliaque.CONCEPTION : Une étude de faisabilité.CONTEXTE : Un hôpital d’enseignement universitaire.SUJETS : Six porcelets pesant de 70 à 80 kg furent soumis à un pontage aortobifémoral effectué de façontotalement laparoscopique par voie rétropéritonéale. Un appareil de suspension abdominale et une tech-nique ne nécessitant pas d’insufflation de gaz furent utilisés. Aucune mini-laparotomie ne fut nécessaire. Lachirurgie fut effectuée au moyen d’une instrumentation laparoscopique insérée à travers six sites de trocar.Après héparinisation systémique, l’aorte abdominale fut clampée et la bifurcation aortique clipée. Uneanastomose aorto-prosthétique termino-terminale fut effectuée. Les tunnels rétropéritoneaux furent créés

14459 December/96 CJS /Page 451

CJS, Vol. 39, No. 6, December 1996 451

Afew years ago, we described atechnique for laparoscopy-assisted aortic surgery in hu-

mans.1 Further investigation led to adifferent approach to the abdominalaorta for totally laparoscopic aorto -bifemoral bypass.2,3 The purpose ofthe present article is to demonstrate atechnique, developed in piglets, thatallowed the first totally laparoscopicaortobifemoral bypass with an end-to-end aortic anastomosis in a humanwith occlusive aortic disease.

MATERIALS AND METHODS

The animal experiments were ap-proved by the institutional AnimalCare Committee and were conductedaccording to the guidelines of theCanadian Council for Animal Care.Six female piglets (average age

6 months and weight range from 75to 80 kg) were premedicated with at-ropine sulfate (Astra Pharma, Missis-sauga, Ont.), 1.2 mg intramuscularly.They were anesthetized with pento-barbital sodium (25 to 35 mg/kg).Anesthesia was maintained with a con-tinuous intravenous infusion of pen-tobarbital sodium (5 to 10 mg/kg perhour).4 Through an endotrachealtube, they were artificially ventilatedby means of a volumetric Bird respira-

tor at a tidal volume of 7 to 8 mL/kg,a frequency of 10 to 12 breaths/minand an air mixture containing 25%oxygen. The ventilator was adjustedto maintain the PaO2 between 95 and135 mm Hg and the PCO2 less than 40 mm Hg.5 An arterial line was in-serted into the right carotid artery.The right internal jugular vein was dis-sected to allow insertion of a catheterfor infusion of crystalloids or with-drawal of blood samples.

Surgical technique

Our technique of laparoscopicretroperitoneal exposure of the aortahas been described elsewhere.3 Briefly,six trocars were used to insert variouslaparoscopic instruments and the graft.The retroperitoneal cavity was main-tained by the use of an abdominal walllifting device (Laparolift; Origin Med -systems, Menlo Park, Calif.).Our technique of laparoscopic aor-

tobifemoral bypass for aortoiliac occlu-sive disease has been modified as fol-lows: Laparoscopic scissors are used todissect the infrarenal aorta and bothexternal iliac arteries. The inferiormesenteric artery is clipped and sev-ered. The lumbar arteries near the in-ferior mesenteric artery are dissected,clipped and divided (Fig. 1). Then, the

aorta, at a point slightly cephalad tothe inferior mesenteric artery, is care-fully dissected circumferentially withthe use of an Endo Maxi Retract (AutoSuture Company, Norwalk, Conn.)(Fig. 2). Prior ligation of the lumbararteries at this level facilitates the manoeuvre. The vascular graft (Hema -shield; Meadox Medicals, Oakland,NJ) is introduced through one trocarport and fashioned in such a way(6 mm × 6 mm) that the prosthesiswould be of the same calibre as thepiglet aorta, which averages 6 mm. In-cision in the femoral regions allows ex-posure of the common femoral arter-ies. The retroperitoneal tunnels arecreated by dissecting the proximal iliacarteries laparoscopically, then insertinga large Crawford clamp from thefemoral region. Its proximal progres-sion is followed under laparoscopiccontrol. Each limb of the graft is intro-duced into its respective retroperi-toneal tunnel and pulled into thefemoral region. Care is taken to ensureproper placement of the ureters abovethe limbs of the graft. After systemicheparinization (100 IU/kg), the aortadistal to the left renal artery is cross-clamped with a laparoscopic vascularclamp (Laborie Surgical, Brossard,Que.). A laparoscopic vascular TA-30stapler (Auto Suture Company) is used

DION ET AL

14459 December/96 CJS /Page

452 JCC, Vol. 39, No 6, décembre 1996

de façon à permettre à chacun des membres de la prothèse d’être acheminé vers sa région fémorale respec-tive. Une anastomose conventionnelle fut effectuée au niveau des artères fémorales.INTERVENTION : Pontage aortobifémoral effectué totalement par laparoscopie.PRINCIPALES MESURES DE RÉSULTATS : Durée de l’intervention, pertes sanguines et complications per-opéra-toires, saignement durant la période post-opératoire immédiate et évaluation post-mortem de l’anastomoseaortique.RÉSULTATS : Tous les pontages aortobifémoraux furent effectués en moins de 4 heures chacun. La pertesanguine per-opératoire fut inférieure à 250 mL. Aucune complication per-opératoire ne fut notée saufpour la survenue d’épisodes de saignement lors du déclampage aortique. Ce type de saignement fut con-trôlé par application d’une éponge de collagène (trois cas) ou de sutures supplémentaires (deux cas). Lesanimaux furent observés pendant 15 minutes avant sacrifice. L’autopsie révéla à ce moment une anasto-mose aortique normale dans tous les cas et un cheminement normal des membres de la prothèse à la foispar rapport aux uretères et dans leur trajets rétropéritonéaux.CONCLUSIONS : Ce modèle animal démontre la faisabilité du pontage aortobifémoral utilisant une tech-nique laparoscopique. L’anatomie rétropéritonéale du porcelet est similaire à celle de l’humain. La chirurgieaortique est effectuée selon les principes de la technique standard. Nous avons utilisé une approche simi-laire pour effectuer, chez l’humain, le premier pontage totalement laparoscopique au moyen d’une anasto-mose termino-terminale.

to staple the distal aorta, where theEndo Maxi Retract instrument is posi-tioned and protects the inferior venacava (Fig. 3). The aorta is transectedwith laparoscopic scissors distal to theaortic cross-clamp (Fig. 4). The graftis then sutured end-to-end to theproximal aortic stump with running 4-0 Prolene (Fig. 5). After the anasto-mosis has been tested for leaks, theaorta is again cross-clamped. The limbsof the graft are placed under appropri-ate tension and the femoral anasto-moses are performed with 5-0 Prolene

(Fig. 6). The animals are observed for15 minutes after completion of theprocedure and are then killed.

RESULTS

All animals survived the procedure.Intraoperative blood loss did not ex-ceed 250 mL and was principally dueto the flushing techniques used be-fore definitive unclamping. Extra su-tures were applied to the aortic anas-tomosis in two cases because of bloodleakage. A collagen sponge sufficed to

control oozing in three cases. Nobleeding occurred at the site of aorticstapling. No procedure took longerthan 4 hours.At autopsy, a small amount of blood

was found in the retroperitoneum inmost animals, as would be encounteredin any surgery of this kind. No majorbleeding source was noted. In all cases,the ureters were in an appropriate posi-tion where they crossed the limbs ofthe graft. The proximal anastomoseswere transected and found to be ade-quately sewn.

LAPAROSCOPIC AORTOBIFEMORAL BYPASS

14459 December/96 CJS /Page 453

CJS, Vol. 39, No. 6, December 1996 453

FIG. 1. Blades of laparoscopic scissors surround one lumbar artery,which has been clipped (arrow). Aorta (large arrowheads) is seen. Por-tion of inferior vena cava (small arrowheads) is visualized.

FIG. 2. Aorta is surrounded by Endo Maxi Retract (Auto Suture Company,Norwalk, Conn.) (arrowheads), which is placed cephalad to incised infe-rior mesenteric artery on which clip has been applied (arrow).

FIG. 3. Laparoscopic TA-30 stapler (Auto Suture Company, Norwalk,Conn.) (large arrowheads) has been fired. Aortic staple line is seen(small arrowheads). Endo Maxi Retract (arrow) protects inferior venacava from injury during stapling.

FIG. 4. Aortic clamp (arrow) has been applied and aorta transected.Proximal aortic stump is visible (arrowheads).

DISCUSSION

We previously described a techniqueof laparoscopy-assisted aortobifemoralbypass1 that has since been performedsuccessfully by others.6,7 All agree thatthe potential benefits afforded by thelaparoscopic approach in other fieldsof surgery could alter positively thepostoperative course of patients whoundergo aortoiliac vascular proce-dures.Because of difficulties we encoun-

tered with the transabdominal ap-proach, such as inefficient bowel re-traction and the need to close theretroperitoneum at the end of the pro-cedure, we prefer the retroperitonealroute for the safety of laparoscopicaortoiliac surgery. The use of the peri-toneum as a barrier against intrusionof intra-abdominal organs into thesurgical field eases the procedure.2,3

The large piglet is a reproducibleand excellent model because its ab-dominal wall and retroperitoneum aresimilar to those of man8 and becausethe same instrumentation can be usedin humans.We believe that an abdominal wall

lifting device is essential for this pro-cedure. It maintains a permanent cav-

ity, which does not threaten to col-lapse when prolonged suctioning isnecessary. We showed in the labora-tory that the procedure can be per-formed solely under pneumoperi-toneum, but the lifting device addssafety if bleeding occurs. We are as-sessing the value of combining aretropneumoperitoneum with the lift-ing device. We recently demonstratedthat maintenance of euvolemia coulddecrease appreciably the danger of car-bon dioxide pulmonary embolism,which might occur after laceration ofa large vein like the vena cava (or alumbar vein).9

The actual technique of laparo-scopic aortobifemoral bypass used inthis model resembles the conventionalopen retroperitoneal technique withminor modifications. Ligation of theinferior mesenteric artery allows forthe creation of a larger retroperitonealcavity. Ligation of a pair of lumbar ar-teries makes for easy dissection of thedistal aorta and the passage of the la-paroscopic vascular TA-30 stapler. Weelected to secure the graft in place be-fore clamping the aorta for two rea-sons: first, it shortens cross-clampingtime without compromising the abil-ity to apply extra stitches at the back

of the anastomosis, because no ten-sion is applied on the limbs of thegraft; second, it allows the graft to sitin a good position for the perfor-mance of the aortic anastomosis. Thelevel of anticoagulation obtained issimilar to that encountered in hu-mans.10,11 We have now performed 22aortobifemoral bypasses in animalsand have found that the ureters arenot endangered. The left ureter is dis-sected from the retroperitoneum andpositioned on the left psoas muscle atthe beginning of the procedure.This animal model does not allow

performance of the anastomosis in anatherosclerotic aorta. Our initial expe-rience in man confirms that an athero-matous aorta does not pose a problemto a laparoscopic anastomosis, whichis performed under the magnificationafforded by a videocamera. A calcifiedaorta would necessitate endarterec-tomy as in open surgery.In summary, this animal model

demonstrates the feasibility of a totallylaparoscopic aortobifemoral bypass foraortoiliac occlusive disease. A similartechnique allowed us to perform thefirst human totally laparoscopic aortob-ifemoral bypass for aortoiliac disease us-ing an end-to-end aortic anastomosis.

DION ET AL

14459 December/96 CJS /Page

454 JCC, Vol. 39, No 6, décembre 1996

FIG. 5. Close-up view of aortic anastomosis after unclamping. Imprintsof running 4-0 Prolene suture are visible (arrowheads) uniting graft toaorta. Knot can be seen (arrow).

FIG. 6. Modified bifurcated graft measuring 6 × 6 mm is shown fromanastomosis (arrow) to point where each graft limb enters its ownretroperitoneal tunnel (arrowheads).

References

1. Dion YM, Katkhouda N, Aucoin A.Laparoscopy-assisted aortobifemoralbypass. Surg Laparosc Endosc 1993;3(5):425-9.

2. Dion YM, Chin AK, Thompson TA.Experimental laparoscopic aorto -bifemoral bypass. Surg Endosc 1995;9:894-7.

3. Dion YM, Gracia CR. Experimentallaparoscopic aortic aneurysm resec-tion and aortobifemoral bypass. SurgLaparosc Endosc 1996;6(3):184-90.

4. Vik A, Jenssen BM, Brubakk AO.Paradoxical air embolism in pigs witha patent foramen ovale. Undersee Bio-med Res 1992;19(5):361-74.

5. Vik A, Brubakk AO, Hennessy TR,Jenssen BM, Ekker M, Slordahl SA.Venous air embolism in swine: trans-port of gas bubbles through the pul-monary circulation. J Appl Physiol1990;69(1):237-44.

6. Berens ES, Herde JR. Laparoscopicvascular surgery: four case reports. JVasc Surg 1995;22(1):73-9.

7. Chen MH, Murphy EA, Halpern V,Faust GR, Cosgrove GM, Cohen JR.Laparoscopic-assisted abdominal aor-tic aneurysm repair. Surg Endosc 1995;9:905-7.

8. Odlaug TO. Laboratory anatomy ofthe fetal pig. 3rd ed. Dubuque (IA):WMC Brown, 1966:8-19.

9. Dion YM, Levesque C, Doillon CJ.Experimental carbon dioxide pul-monary embolization after vena cavalaceration under pneumoperitoneum.Surg Endosc 1995;9:1065-9.

10. Heras M, Chesebro JH, Penny WJ,Baily KR, Badimon L, Fuster V. Laboratory investigation: effects ofthrombin inhibition on the develop-ment of acute platelet-thrombus de-position during angioplasty in pigs:heparin versus recombinant hirudin,a specific thrombin inhibitor. Circu-lation 1989;79(3):657-65.

11. Lam JY, Chesebro JH, Steele PM,Dewanjee MK, Badimon L, Fuster V.Deep arterial injury during experi-mental angioplasty: relationship to apositive indium-111-labeled plateletscintigram, quantitative platelet de-position and mural thrombus. J AmColl Cardiol 1986;8:1380-6.

Addendum

As of August 1996, two of the authors(Y.-M.D. and C.R.G.) have successfullyperformed four more totally laparoscopicaortobifemoral bypasses and two totallylaparoscopic iliofemoral bypasses in hu-mans. The abdominal wall lifting devicewas judged unnecessary, and the proce-dures were done with the use of carbondioxide pneumoperitoneum. The retro -peritoneal approach has been modified tofacilitate the procedure.

LAPAROSCOPIC AORTOBIFEMORAL BYPASS

14459 December/96 CJS /Page 455

CJS, Vol. 39, No. 6, December 1996 455

CORRECTION

There was an error in abstract number 707 on page A30 of the abstract supplement tothe August 1996 issue of the Journal: the name of Dr. D.C. Taylor was omitted fromthe list of authors. The information on the title, authors and centre for abstract 707should appear as follows: PROGRESS IN ABDOMINAL AORTIC ANEURYSMSURGERY: FOUR DECADES OF EXPERIENCE AT A TEACHING CEN-TRE. J.C. Chen, H.D. Hildebrand, A.J. Salvian, D.C. Taylor, Y.N. Hsiang. Department of Surgery, Vancouver Hospital and Health Sciences Centre, Univer-sity of British Columbia, Vancouver, BC. We apologize to Dr. Taylor and his coau-thors for this error.