PHARMACOLOGICAL THROMBOLYSIS: ONE MORE WEAPON FORFREE-FLAP SALVAGE

SALVATORE D’ARPA, M.D.,* ADRIANA CORDOVA, M.D., and FRANCESCO MOSCHELLA, M.D.

Despite the high success rate of free-tissue transfer, thrombosis still complicates 5�30% of cases. Meticoulous technique,careful vessel selection, and pharmacological prophylaxis are not always enough to avoid thrombosis. Early diagnosis andreintervention provide the only way to salvage a thrombosed free flap, in case of either arterial or venous thrombosis. Whenkinking, torsion, or external compression of the pedicle are ruled out, and thrombectomy and redo of the anastomosis areunsuccessful, the last resort to save the flap is thrombolytic therapy. The authors present their experience with the salvage oftwo otherwise lost flaps by means of urokinase thrombolysis through direct intra-arterial injection with the vein left open toavoid systemic diffusion of the drug, and give technical tips to improve drug delivery to the flap. Pharmacological thrombolysisis an additional and effective weapon to resolve thrombosis, if properly used, to be considered by every reconstructivemicrosurgeon. ª 2005 Wiley-Liss, Inc. Microsurgery 25:477�480, 2005.

Nonostante la percentuale di successi dei lembi liberi abbia raggiunto il 97%, la trombosi continua a rappresentare unatemibile minaccia. Profilassi farmacologica, attenta selezione dei pazienti e tecnica meticolosa non sono ancora sufficienti adevitare la trombosi, per la quale diagnosi precoce e intervento immediato rimangono l’unica possibile soluzione. Se larevisione dell’anastomosi e la trombectomia si rivelano inefficaci, l’ultima risorsa per salvare il lembo e la trombolisi farma-cologica. Gli autori presentano la propria esperienza in due casi in cui la trombolisi farmacologica, praticata attraversol’iniezione intra-arteriosa di urochinasi con la vena aperta per evitarne la diffusione sistemica, ha permesso di salvare duelembi altrimenti considerati persi. Danno inoltre alcuni dettagli tecnici per migliorare la diffusione del farmaco nel microcircolodel lembo. Usata correttamente, la trombolisi farmacologica rappresenta un’efficacie arma aggiuntiva per il salvataggio deilembi liberi e deve pertanto fare parte del bagaglio culturale di ogni microchirurgo ricostruttore. ª 2005 Wiley-Liss, Inc.Microsurgery 25:477�480, 2005.

Pharmacological thrombolysis is a procedure of well-known efficacy and safety. Since its first use in the 1960sfor intravenous treatment of pulmonary embolism,thrombolytic therapy has become the treatment of choicefor thrombi in coronary, pulmonary, peripheral, andvisceral arteries, dialysis grafts, veins, and intravenouscatheters (Samett E.J., Thrombolysis, peripheral,www.emedicine.com). Pharmacological thrombolysiswas first used inmicrovascular free-tissue transfer in 1987,when Schubert et al.1 andLipton and Jupiter2 reported onthe salvage of two free flaps with streptokinase.

Since then, not much has been written about phar-macological thrombolysis in microvascular reconstruc-tive surgery. Besides many case reports,2�8 only threelarger series (8, 5, and 20 flaps, respectively) were re-ported by Yii et al.,9 Serletti et al.,10 and Panchapakesanet al.11 Of these three, only Panchapakesan et al.11 triedto give indications for pharmacological thrombolysis,and reported their protocol for flap salvage in whichpharmacological thrombolysis was the last step.

Even if there are no large series regarding throm-bolytic free-flap salvage, large numbers are not requiredto assess the efficacy and safety of a procedure that hasbecome the gold standard for the treatment of acutethrombosis.

Pharmacological thrombolysis is mediated by agentsthat directly or indirectly activate plasminogen intoplasmin, its active form. Plasmin then cleaves fibrin inclots as well as fibrinogen and other coagulation factors(V, VIII, IX, XI, and XII).1

Four thrombolytic agents are currently available:streptokinase, urokinase, recombinant tissue plasmino-gen activator (rt-PA), and acylated plasminogen-strep-tokinase-activator complex (APSAC).

Streptokinase, derived from b-hemolytic strepto-cocci, forms a complex with one molecule of plasmino-gen, a complex that is able to activate other plasminogenmolecules. Streptokinase is antigenic, however, and cancause anaphylactic reactions (0.1% of cases) and canlose efficacy because of inactivation by circulatingantibodies whose titers rise within 5�6 days and fallwithin 6 months from the first administration.

Urokinase, derived from human kidneys or re-combinant DNA technology, directly activates plas-minogen without antigenicity.

Rt-PA binds specific sites of fibrin in clots, subse-quently activating plasminogen. As it only acts in clots,

Dipartimento di Discipline Chirurgiche ed Oncologiche, Cattedra di ChirurgiaPlastica e Ricostruttiva, Universita degli Studi di Palermo, Palermo, Italy

*Correspondence to: Salvatore D’Arpa, Via Giovanni Pacini 12, 90138 Pa-lermo, Italy. E-mail: turidarpa@hotmail.com

Received 30 December 2004; Accepted 28 February 2005

Published online 2 September 2005 in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/micr.20147

ª 2005 Wiley-Liss, Inc.

it theoretically has a higher fibrin specificity and sub-sequently lower systemic effects. The fact that it has ahigher fibrin specificity, and could then reduce systemiccomplications, has only been proven in vitro. A pro-spective randomized trial comparing urokinase and rt-PA was published in 1994 that found no differences inboth efficacy and safety between the two drugs in theclinical setting.12

APSAC is a preformed plasminogen-streptokinasecomplex inactivated by acylation. It is deacylated insolution and is theoretically highly selective, but noadvantages and high rates of systemic complicationshave been shown in the clinical setting.13

What is still lacking is agreement about a standard-ized means of safe and effective delivery to a free flap.

In this paper, the authors present their own experi-ence with the thrombolytic salvage of two flaps otherwiseconsidered dead, together with technical tips to improvedrug delivery and reduce systemic complications.

CASE REPORTS

Case 1

In February 2004, an 83-year-old man came to ourattention with a high-grade squamous-cell carcinoma ofthe scalp, with calvarial and dural invasion. After tumorexcision and dural reconstruction by a galeal graft, alatissimus dorsi free muscle flap was harvested andanastomosed in an end-to-end fashion to the superficialtemporal vessels and skin-grafted. On postoperative day2, the patient was brought back to the operating roombecause of black-blood spillage after pinprick, which weconsider an indication for immediate reexploration. Onsurgical reexploration, venous thrombosis was evident.The venous anastomosis was resected, and thrombecto-my was performed, followed by irrigation with heparin.After a few minutes of observation, no venous outflowwas noticed, even if the arterial anastomosis was patent.The decision was taken to perform pharmacologicalthrombolysis. The thoraco-dorsal vein was left open todirectly observe eventual venous outflow and to avoidsystemic drug spread, and 100,000 units of urokinasewere slowly injected into the recipient artery by directpuncture with a 26-gauge needle. After 4 min, venousoutflow was reestablished, and a vein graft from the leftsural vein was harvested while observing venous outflowthat was ultimately judged to be satisfactory. The veingraft was used to bridge the thoraco-dorsal vein to theexternal jugular vein in an end-to-end fashion.

The postoperative course was uncomplicated.

Case 2

In October 2004, a 41-year-old man came to ourattention for osteomyelitis of the distal third of the tibia

that developed after an open fracture treated by externalfixation. Extensive debridement was performed, and afree gracilis muscle flap was transferred and anastomo-sed in an end-to-side fashion to the anterior tibial artery.Because of the poor condition of the recipient artery,probably due to trauma, 2 h were lost before completionof the arterial anastomosis. The flap was dark, andneither color improvement nor venous outflow wasnoted after completion of the anastomosis and assess-ment of its patency by milking and Doppler test. In-traoperative diagnosis of microthrombi in the flapmicrocirculation was done, and the decision to performthrombolytic therapy was made after finding no evi-dence of thrombi in the flap’s vein. As the arterialanastomosis was undoubtedly patent, a decision not toredo the anastomosis was made. Two stitches from theanastomosis were released, and 50,000 units of uroki-nase were slowly injected into the artery. The anasto-mosis was then closed, and the flap was observed. After5 min of observation, no improvement was noted,probably because with the anastomosis open, not en-ough drug was delivered into the flap’s microcirculation.To directly inject the drug into the flap’s artery wasconsidered dangerous because of the small diameter ofthe artery (1.1 mm), which would have exposed it to anunacceptably high risk of spasm. To inject the drug intothe anterior tibial artery, even if it was clamped down-stream to the anastomosis, was not considered becauseof the risk of systemic spread of the drug after clamprelease and of poor delivery to the flap’s microcircula-tion. The drug was then directly injected into the flap’sartery through cannulation by means of a puncture witha 26-gauge needle, perpendicular to the anterior tibialartery. Another 50,000 units of urokinase were theninjected slowly. After 3 min, venous outflow andimmediate improvement of the flap’s color were noted.Venous outflow was judged to be satisfactory, and ve-nous anastomosis was performed in an end-to-endfashion to one anterior tibial vein. The postoperativecourse was uncomplicated. One month postoperatively,the bone had healed perfectly.

DISCUSSION

Failure of a free flap is due to either arterial or ve-nous thrombosis, the latter being by far the more com-mon because of the slower flow rate on the venous side.Measures to prevent these phenomena include carefulpatient selection, careful recipient vessel selection,pharmacological prophylaxis, meticulous surgical tech-nique in flap harvesting, vessel preparation and micro-surgical anastomosis, and careful postoperativemanagement to prevent all maneveurs and events (e.g.,external vessel compression or avulsion, hematoma, or

478 D’Arpa et al.

seroma formation) that could compromise flow throughthe anastomosed vessel.

Despite all these measures, the threat of thrombosisis still present (5�30% of all free-tissue transfers),14 andclose monitoring of the flap’s vitality is mandatory topromptly intervene in case of thrombus formation.Prompt intervention has always been advocated toavoid irreversible tissue damage by promptly reestab-lishing vascularization of the flap. But this is not theonly reason. Another reason to promptly intervene isthat, if pharmacological thrombolysis has to be per-formed, it would only be effective within 6 h fromthrombus formation. After 6 h, irreversible tissue dam-age ensues, causing cellular and microvascular damageand the so-called no-reflow phenomenon.

Major contraindications to pharmacological throm-bolysis, to be considered only when systemic circulationof the drug is allowed, are recent stroke or malignancy(particularly if brain metastases are likely), renal insuf-ficiency, allergy, cardiac thrombus, diabetic retinopathy,and coagulopathy. A recent arterial puncture or surgeryare relative contraindications (Samett E.J., Thrombo-lysis, peripheral. www.emedicine.com). These contrain-dications do not exist if the vein of a free flap is releasedand the drug is injected directly into the flap’s circula-tion,15 as in the cases described here. Furthermore,urokinase is inactivated after 15 min, and recombinanttissue plasminogen activator (rt-PA) after 5 min (SamettE.J., Thrombolysis, peripheral. www.emedicine.com).Hence if the venous anastomosis is redone after at least15 min, risks of systemic complications are irrelevant.As it is highly antigenic, it is preferable not to usestreptokinase. The choice of urokinase over rt-PA wasmade because of the greater availability of the former.

Pharmacological thrombolysis must be consideredthe last resort, when all other attempts at flap salvagehave failed. Panchapakesan et al.11 proposed a protocolin which pharmacological thrombolysis is used after allother measures (thrombectomy)16 fail and arterial in-flow is present.

The authors use the same protocol and point outthat the efficacy of pharmacological thrombolysis istime-related. Not too much time has to be lost in tryingto reestablish flow or observe the flap after thrombec-tomy has been unsuccessfull. Conversely, thrombolytictherapy must always be kept in mind and considered asa useful weapon in flaps otherwise considered lost.

Perhaps pharmacological thrombolysis has beenoverlooked by many reconstructive microsurgeons be-cause of a fear of systemic complications. With propertechnique, drug delivery can be maximized, and systemicspread and subsequent complications can be avoided.

The flap’s vein must be left open in order to directlyobserve the quality and quantity of venous outflow, and

avoid systemic spread of the drug. When an end-to-sideanastomosis is performed, drug injection must be donedirectly into the flap’s artery, as described in case 2, inorder to maximize drug delivery to the flap’s circulationwhile avoiding systemic spread that could occur withinjection into the recipient artery. When the anastomosisis performed in an end-to-end fashion, drug injectioncan be done in the recipient artery, which is alwayspreferred to the flap’s artery.

Serletti et al.10 redo a venous anastomosis beforeadministering thrombolytic therapy. The authors prefernot to redo the anastomosis, not only because this willavoid systemic drug spread, but also because directlyobserving venous outflow is the best method for evalu-ating its effectiveness.

Yii et al.9 performed pharmacological thrombolysisthrough the resected end of the vessel or via a sidebranch. Panchapakesan et al.11 used the same methodsor a small arteriotomy. It is better not to manipulate thepedicle, to avoid spasm or damage. To allow circulationof the drug into the flap with the bloodstream with anintact arterial anastomosis could warrant a better andmore homogeneous diffusion of the drug in the flap’smicrocirculation, while avoiding ischemia. These samebeliefs are shared by Bonde et al.17

Moreover, it could be risky to open a well-func-tioning anastomosis to allow introduction of a catheterfor drug administration.

Published outcomes of a flap’s salvage with throm-bolytic therapy vary from the 30% (6 out of 20) ofPanchapakesan et al.11 (the largest series ever) to the100% of Serletti et al.10 (5 out of 5). These differencescould be due to different timeliness in intervention. Thereason why thrombolytic therapy did not always workin Panchapakesan et al.11 series may be that it was notalways undertaken within the 6 recommended hours. Ifwe consider the time that is always lost, from whenthrombosis is diagnosed, to bringing the patient back tothe operating room, to reexploring the anatomosis, toperforming a thrombectomy before deciding to performa pharmacological thrombolysis, 6 h can easily pass.

Thrombolytic therapy, however, does neither resolvethe causes for thrombosis nor eliminate the risk of re-thrombosis. Thus eventual extrinsic or intrinsic causesfor thrombosis must be removed and anticoagulant andantiplatelet therapy must not be withdrawn. In the au-thors’ institution, the protocol for thrombosis prophy-laxis includes: 1) low molecular weight heparin given onthe day of the operation and for 15 days postopera-tively; 2) aspirin, 1.4 mg/kg orally on the preoperativeday and for 15 days postoperatively;18 3) lysin acetyl-salicilate 1 g, given as an i.v. bolus just before pedicledivision to prevent platelet aggregation; 4) lactatedRinger’s for 3 postoperative days, in order to keep

Thrombolysis in Free Flaps 479

hematocrit at around 30%;18 and 5) intraoperative vesselirrigation with heparinized saline, 100 U/ml. Bothplatelet aggregation with aspirin and anticoagulationwith heparin are warranted. Close monitoring everyhour for the first 72 h is also warranted, to allow earlyrecognition and treatment of complications.

CONCLUSIONS

Indications for pharmacological thrombolysis are allthrombosed flaps in which thrombectomy fails, or thoseflaps with normal arterial inflow and absent venousoutflow even after thrombectomy.

Uk and rt-PA, due to their short half-lives and theirabsence of antigenicity, are the drugs of first choice.

Pharmacological thrombolysis must be performed assoon as possible to avoid irreversible tissue damage andthe no-reflow phenomenon.

The drug must be administered without disruptingthe arterial anastomosis. When the anastomosis is end-to-end, the drug must be injected directly into therecipient artery. In the flap’s artery, through a puncurein the recipient artery when the anastomosis is end-to-side.

Even after pharmacological thrombolysis, prophy-lactic therapy must be administered to avoid furtherthrombotic events.

Pharmacological thrombolysis must always be keptin mind to further improve the success rate of micro-surgical free-tissue transfer by trying to salvage freeflaps that would otherwise be considered lost.

REFERENCES

1. Schubert W, Huter DW, Guzman-Stein G, Ahrenholz DH, SolemLD, Dressel TD, Cunningham BL. Use of streptokinase for the

salvage of a free flap: case report and review of the use ofthrombolytic therapy. Microsurgery 1987;8:117�121.

2. Lipton HA, Jupiter JB. Streptokinase salvage of free-tissue trans-fer: case report and review of the literature. Plast Reconstr Surg1987;79:977�981.

3. Wechselberger G, Schoeller T, Ohler K, et al. Flap salvage in a‘‘flow-through’’ flap by manual thrombectomy plus thrombolytictherapy. J Reconstr Microsurg 1998;14:127�129.

4. Tonks A. Streptokinase salvage of a rectus abdominis free flap.Correspondence. Plast Reconstr Surg 1995;95:933�934.

5. Fudem GM, Walton RL. Microvascular thrombolysis to salvage afree flap using human recombinant tissue plasminogen activator. JReconstr Microsurg 1989;5:231�234.

6. Atiyeh BS, Hashim HA, Hamdan AM, Musharafieh RS. Localrecombinant tissue plasminogen activator (rt-PA) thrombolytictherapy in microvascular surgery. Microsurgery 1999;19:265�271.

7. Goldberg JA, Pederson WC, Barwick WJ. Salvage of free tissuetransfer using thrombolytic agents. J Reconstr Microsurg1989;5:351�356.

8. Noordanus RP, Hage JJ. Late salvage of a ‘‘free-flap’’ phallopl-asty: a case report. Microsurgery 1993;14:599�600.

9. Yii NW, Evans GR, Miller MJ, Reece GP, Langstein H, Chang D,Kroll SS, Wang B, Robb GL. Thrombolytic therapy: what is itsrole in free flap salvage? Ann Plast Surg 2001;46:601�604.

10. Serletti JM, Moran SL, Orlando GS, O’Connor T, Herrera HR.Urokinase protocol for free flap salvage following prolonged ve-nous thrombosis. Plast Reconstr Surg 1998;102:1947�1953.

11. Panchapakesan V, Addison P, Beausang E, Lipa JE, Gilbert RW,Neligan PC. Role of thrombolysis in free-flap salvage. J ReconstrMicrosurg 2003;19: 523�529.

12. The STILE Investigators. Results of a prospective randomised trialevaluating surgery versus thrombolysis for ischemia of the lowerextremities (the STILE trial). Ann Surg 1994;220:251.

13. Marder VJ, Sherry S. Thrombolytic therapy: current status. NEngl J Med 1988;318:1512�1520.

14. Kubo T, Yano K, Hosokawa K. Management of flaps with com-promised venous outflow in head and neck micorsurgical recon-struction. Microsurgery 2002;22:391�395.

15. Esclamado RM, Carrol WR. The pathogenesis of vascularthrombosis and its impact in microvascular surgery. Head Neck1999;21:355�362.

16. Wheatley MJ, Meltzer TR. The role of vascular pedicle thromb-ectomy in the management of compromised free tissue transfers.Ann Plast Surg 1996;36:360�364.

17. Bonde CT, Heslet L, Jansen E, Elberg JJ. Salvage of free flaps aftervenous thrombosis: case report. Micorsurgery 2004;24:298�301.

18. Conrad MH, Adams WP Jr. Pharmacologic optimization ofmicrosurgery in the new millennium. Plast Reconstr Surg2001;108:2088�2096.

480 D’Arpa et al.