deep calf veins arterialization for inferior limb preservation in diabetic patients with extended...

edicine 12 (2011) 10–19
Cardiovascular Revascularization M
Deep calf veins arterialization for inferior limb preservation in diabeticpatients with extended ischaemic wounds, unfit for direct arterialreconstruction: preliminary results according to an angiosome

model of perfusion

Vlad Alexandrescua,⁎, Christian Ngonganga, Gaetane Vincenta,Gilles Ledentb, Gérard Hubermontc

aDepartment of Surgery, Princesse Paola Hospital, Marche-en-Famenne, BelgiumbDepartment of Radiology, Princesse Paola Hospital, Marche-en-Famenne, Belgium

cDepartment of Diabetology, Princesse Paola Hospital, Marche-en-Famenne, Belgium

Received 17 November 2009; accepted 16 December 2009

Abstract Aim of the study: The study proposes to evaluate the limb salvage in diabetic ischemic limbs with

⁎ CorrespondingPaola Hospital, Rue dTel.: +32 84 219111;

E-mail address: v.

1553-8389/09/$ – seedoi:10.1016/j.carrev.2

foot wounds, where all types of common arterial reconstructions have previously failed or wereimpracticable, by using the “SAVES” technique (selective arterio-venous endoluminal switch) fordeep calf veins arterialization upon an angiosomes model of distribution.Materials and methods: Since January 2001 until September 2009, a series of 26 limbs withthreatening ischemic wounds in 25 diabetic patients at high risk for major amputation and no feasibleconventional revascularizations, were treated by the hybrid (surgical and endovascular) SAVEStechnique and were retrospectively reviewed. The method consists in selective arterialization of thedeep calf veins with synchronous endoluminal exclusion of the collaterals, guided by an angiosomes-model of vascularization. There were 12 limbs treated by preferential anterior tibial veinsarterialization, 11 with revascularizations in the posterior tibial and three others targeting theperoneal-related territorial wound distribution.Results: The initial technical success was achieved in 21 of 26 limbs (80%) with 0% 30-dayperioperative mortality rate. The cumulative primary and secondary patency were: 66%, 60% and48%, at 12, 24 and 36 months, respectively. Limb salvage revealed 73% at one year and steady 73%afterwards, while the clinical success was: 68%, 60% and 60% at identical time intervals.Conclusion: Selective deep calf venous arterialization oriented by an angiosomemodel for reperfusionmay represent a complementary alternative for limb salvage in extreme situations, inoperable by directarterial methods. Larger groups of study are needed to ascertain these preliminary observations.© 2011 Elsevier Inc. All rights reserved.

Keywords: Diabetes Mellitus; Critical-limb ischemia; Venous arterialization; Angiosomes; Embolization

1. Introduction

Advanced atherosclerosis with extended tibial arterieslesions is a common concern in diabetic patients having

author. Department of Vascular Surgery, Princessu Vivier, 21; 6900 Marche-en-Famenne, Belgium.fax: +32 84 [email protected] (V. Alexandrescu).

front matter © 2011 Elsevier Inc. All rights reserved.009.12.002

critical-limb ischemic (CLI) wounds [1] and often challengesthe classical methods for revascularization. Despite pro-gresses in both surgical and endovascular approaches, in 14–20% of these patients exhibiting advanced below the kneeocclusive disease, neither of the available procedures canafford appropriate distal arterial supply, turning to amputa-tion [2–5]. This observation was documented particularly indiabetic ischemic limbs, that currently add a neuropathic

mailto:[email protected]

http://dx.doi.org/10.1016/j.carrev.2009.12.002

11V. Alexandrescu et al. / Cardiovascular Revascularization Medicine 12 (2011) 10–19

background, local sepsis and variable amounts of tissueloss [1,2,4,5]. Although the concept of “venous arterializa-tions” as extreme alternative for delivering the oxygenatedblood to the distal tissues is not new, increasingcontemporary experience seems to reconsider this farthestalternative for limb revascularization in certain desperatecases at high risk for amputation [3,6]. While some reportsare claiming encouraging limb rescue results [3,6,7], othersremain reserved [8]. Parallel efforts for CLI limb salvagewere equally unfolded in the plastic surgery field, byintroducing an “angiosomes model for revascularization”(AMR), that proposes targeted tissue reconstructions uponspecific bundles for arterial and venous blood supply[9,10]. We examined in this study a hybrid technique(surgical and endovascular) for deep calf veins selectivearterializations, in conventionally inoperable diabetic CLIfoot ulcers, appending an intentional AMR policy tosupport tissue regeneration.

2. Material and methods

2.1. Patients

Since January 2001 until September 2009, a series of 26limbs in 25 diabetic patients with imminent amputation forthreatening inferior limb ischemic wounds and no feasiblearterial revascularizations, were treated by selective arterio-venous endoluminal switch (SAVES method) in the deepcalf veins following an AMR policy [9,10]. One patientreceived a staged bilateral intervention, at 1-year interval.These files were retrospectively reviewed having an ethicalcommittee approval. Patient’s selection was discussed in amultidisciplinary staff gathering vascular surgeons, inter-ventional radiologists, diabetologists, orthopedic and plasticsurgeons, infectiologists and general practitioners (Table 1).

Table 1Patient selection

Type of the lastconventional Therapythat failed before selectingfor the SAVES technique

n=26 Background history of the late

No feasibleaterial revasc.

Primaryprocedur

Failed femoro-infra-popliteal bypass

(n=2) 1

Failed femoro-tibial bypass (n=6) 1Failed femoro-pedal bypass (n=4) 1Failed btk. angioplasties

(intra- or extraluminal)(n=6) 1

Failed btk. stenting (n=2)Failed laser therapy (n=1)Unsuccessful iloprost infusion (n=2) 1Unsuccessful

vasodilatator+anticoagulanttreatment

(n=2) 1

btk, below-the-knee.

There were 18 men and the mean age was 71.9 years(in the range 56–84). Seven (27%) cases were insulindependent. Patient characteristics and risk factors aresummarized in Table 2. The main inclusion criteriaexpressed the presence of distal foot ulcerations and tissuenecrosis (Rutherford categories 5–6.) [11] and were assignedin all the treated limbs. Severe ischemic wounds strictlyconfined to the foot (Wagner Grade 3–4) [12] were noted in18 limbs (69%), whereas 8 (31%) others, associated complexfoot and below the knee trophic lesions (Table 2). Otherinclusion criteria gathered critical levels of trans-cutaneousoxymetry (tcPO2 b30 mmHg) [1,4,12] in previously failedrevascularizations without feasible arterial reconstructions(Tables 1 and 2), joining the lack of autologous venousmaterial (Table 3) and the presence of extended tissueinflammation (n=20 cases) (Table 2).

The exclusion criteria assembled: extended and irrecov-erable foot gangrene (Wagner Grade 5 lesions) [12], severecardiac insufficiency [left ventricular ejection fraction(LVEF) b30%, appraised by systematic cardiologic evalu-ation] and eventual disagreement expressed by the patient.

The technical features of the 23 hybrid (surgical andendovascular) interventions, are resumed in Table 4. Thepreoperative evaluation included the initial clinical codifi-cation [11–13], systematic tcPO2 measurements, joined to aduplex scan assessment of the arterial and venous (superfi-cial and deep) network, in all the affected limbs. The “ankle-brachial index” (ABI), added to the toe pressure measure-ments or the “toe-brachial index” when technically feasible,were used as complementary preoperative markers inlabeling the critical ischemic presentations. These datawere correlated either to DS arteriography (our earlyexperience), or to angio-computed tomography (CT) ormagnetic resonance imaging (MRI) followed by peropera-tive angiography (in the recent years). Multilevel arterial

st arterial revascularization

eSecondary procedure after: Tertiary procedure after:

Endovascular Surgery Endovascular Surgery

1

3 1 11 1 13 2

1 11

11

Table 2Patients characteristics/risk factors

No. of Limbs (n=26) (n%)

Age N70 years n=9 34%Coronary disease n=23 88%Chronic renal insufficiency n=15 57%End stage renal disease/dialysis n=11 42%Associated venous insufficiency n=7 26%Cerebrovascular disease n=9 34%Inferior limb neuropathy n=23 88%

Clinical features No. of limbs %

Failed previous primary oriterative femoro-distal bypasses

n=12 46%

Failed previous primary oriterative endovascular attempts

n=9 34%

Sole medical treatment and nofeasible previous arterialrevascularization

n=4 15%

Foot and calf wounds orinflammation impairing distalexposure of vessels

n=20 77%

Wagner Grade 3 foot lesions. n=7 26%Wagner Grade 4 foot lesions. n=19 73%Wagner Grade 3 and 4 foot lesionsadded to calf or ankle trophic defects

n=8 31%

Angiographic features No. of limbs %

2 or 3 crural arteries occluded n=15 57%Incomplete or occluded pedal arches n=11 42%Spread of calcifications N5 cm n=13 50%Infra-popliteal occlusions N10-cm length n=16 61%

Table 3Concomitant superficial vein conditions

No. of limbs (n=26) (n%)

Saphenous veins previouslyharvested or stripped

n=14 54%

Thrombosed saphenoussegments bordering the wounds

n=9 34%

Unusable and unavailable armveins or AV fistulas for dialysis

n=12 46%

Veins surrounded by tissuesepsis, ulcers or inflammation

n=11 42%

Varicose segments orpost-phlebitis degeneration

n=8 31%

Angiokeratomas in Fabry's disease n=1 3%Tissue calciphylaxis n=1 3%Contraindication for arm

vein harvesting (hemochromatosis)n=1 3%

12 V. Alexandrescu et al. / Cardiovascular Revascularization Medicine 12 (2011) 10–19

disease with preferential tibial location was a dominant item.Following the TASC II stratification [1] of the maininfrainguinal lesions, there were two Type B (7%), sixdominant Type C (23%) and 18 limbs (69%) showing severeType D lesions. Huge calf vessels calcifications (N5 cm),were present in 13 cases (50%) (Table 2). Adjuvant woundtherapy was applied following a multidisciplinary consensus,according to each presentation and the period-relatedexperience of our team (Table 5).

2.2. Definitions

Clinical grading of the inferior limb ischemia wasevaluated according to the current standardized reportingpractices [11–13] and the TASC (II) recommendations [1].Other specific clinical findings concerning the “diabeticfoot” were guided by the Wagner revised classification [12].The technical success was consented if, at the completionangiograms, a direct arterial axis starting the iliac levelthrough the selected deep calf veins until the peri-malleolarvenous network was present after devalvulation and selectiveembolizations of side branches, without N25% narrowing ormajor steal by collaterals. Beyond duplex notification of thegraft patency, a pulsatile Doppler spectrum (arterial flowvelocity) in the distal deep calf veins was needed to consentthis statement before discharge.

Clinical success was recorded if marked healing trends(improvement with at least two Rutherford categories [1]with or without minor amputations) added to superior tcPO2measurements above the critical level (b30 mmHg) [1,4],were noted. Limb salvage implied no request of majoramputation and it was consented if the functional autonomyof the patient was recovered (walking or standing), despiteeventually minor amputations. Since the revascularizationswere not conducted on the foot arterial network and the distalarterial flow was not directly stressed by the SAVES method,the ABI, the toe pressure and the toe-brachial indexmeasurements were not included as postoperative reportingitems in this cohort of patients.

2.2.1. The angiosome model for revascularization and deepveins selection

In each case, the deep venous territory for arterializationwas selected upon the location of the main foot wounds, byfollowing the angiosomes model of perfusion in the foot andankle described in the dedicated literature [9,10,14]. Inchoosing the appropriate deep calf veins, specific duplexnotifications focusing the patency, the valvular competence,eventual anatomical variants and the level of duplicationwere appended. Complementary information referring thesurrounding tissues in the calf was added, in order toestablish the optimal location of the distal arteriovenousanastomosis. A succinct illustration of our schematicangiosomes working model is appended in Fig. 1, focusingon the main venous distribution.

2.3. The SAVES technique

All the combined interventions were performed in theoperating room using a mobile C-arm roentgenographicunit. In each case, the preoperative medication includedanti-platelet therapy (160 mg aspirin daily or clopidogrel75 mg/day) started at least 72 h before the procedure. In thefirst stage of the procedures, a classical femoro-below-the-knee polytetrafluoroethlylene (PTFE) bypass protocol was

Table 4The technical features of the 26 combined surgical /endovascular procedures

Type of prosthetic bypass Localisation of the targeted venosome Sites of collateral embolization

Femoro-popliteo/distal (n=17). Anterior tibial and forefoot (n=7) TPt+collat. ATa (n=3).TPa+collat. ATa (n=4).

Posterior tibial and plantar/ heel (n=8). ATa+collat. PTa (n=8).Peroneal and lateral ankle (n=2). ATa+PTa+collat. peroneal (n=1).

TPa+collat. Ata (n=1).Popliteo-popliteo/distal (n=2). Anterior tibial and forefoot (n=2) PTa+peroneal (n=2).Femoro-tibial (n=5). Anterior tibial and forefoot (n=2). TPt+collat. Ata (n=1).

PTa+collat. Ata (n=1).Posterior tibial and heel (n=2). Peroneal+collat. PTa (n=1).

Collat. PTa (n=1).Peroneal and lateral ankle (n=1). PTa+collat. peroneal (n=1).

Popliteo-tibial (n=2). Anterior tibial and forefoot (n=1) PTa+peroneal+collat. Ata (n=1).Posterior tibial and plantar (n=1). ATa+peroneal+collat. PTa (n=1).

ATa, anterior tibial artery; PTa, posterior tibial artery; TPt, tibio-peroneal trunk.


underwent. The common femoral artery was usually thechosen in-flow vessel (n=20) albeit, the superficial femoralartery (SFA) (n=2) and the popliteal artery (n=4) wereseldom employed. The locations of the distal anastomosis arespecified in Table 4 and were realised on one of the paireddeep veins in a termino-lateral (n=19) or ter`mino-terminal(n=7) manner. In all the cases, 8 mm PTFE armed prosthesis(n=21/Distaflow BARD USA, and n=5/GORE MedicalProducts USA) were placed. Before clamping the femoralartery, 2500 IU heparin (adjusted to the individual weightand clearance in the 42% cases of severe renal insufficiency)were given. A moderate anticoagulant therapy was preferred,in order to avoid major interferences with the secondprocedural stage, the collateral embolizations. The proximalsegment of the dedublated vein was ligated, when the lateralsided anastomosis was chosen. Although the distal poplitealvein was found more often paired than single according tosimilar reported observations [15], in seven cases (26%), thedistal anastomosis was performed directly in the venoustibial trunks (Table 4) and oriented toward the woundcorresponding angiosome.

After declamping the PTFE arterio-venous bypass, an 80cm /6F introducer sheath (Cook Inc. Bloomington, IN USA(commercialized by Cook Ireland LTD.)) was antegradelymounted via the common femoral artery and throughout theprosthetic graft and distal anastomosis, allowing road-mapping of the infra-popliteal deep venous network

Table 5Adjuvant wound treatment

No. of limbs (n=26). (%)

Vacuum-assisted wound closure n=6 23%Biosynthetic skin substitutes n=2 7%Maggot therapy n=4 15%Rotational skin flaps n=1 3%PACE technology n=2 7%Sole epithelialization-stimulating dressings n=10 38%Inability for adjuvant local treatment n=1 3%

(Fig. 2). The competence of the valvular system was nextdocumented (often already intrinsically affected in theneuropathic limbs) [16] and the level of collaterals. Ifneeded, devalvulation of the deep veins was initiated by thecrossing of a 0.035 hydrophilic guidewire (Terumo, Japan)and accomplished by the passage of the same 6F 80-cm-long introducer sheath (occasionally necessitating few“back and forth” co-axially over the wire movements).After advancing the long sheath until the ankle region inthe selected venous pathway, by injecting small amounts ofcontrast, the principal venous collaterals and perforatorswere located. Using a floppy tip 5F vertebral or Cobra 2-3catheters (Cook), or seemingly with a 4F curved hydro-

Fig. 1. A simplified illustration of the angiosomes (i.e., venosomes) of thefoot and ankle: 1, (medial calcaneal), 2, (medial plantar), 3, (lateral plantar)dependent from the posterior tibial territory, 4, (dorsalis pedis) from theanterior tibial and 5, (lateral calcaneal) from the peroneal tract.

Fig. 2. Selective arterialization of a left posterior tibial vein: A, B, C: staged embolizations in the anterior tibial and peroneal veins. (C, D, and E) collateral andperforator complementary embolizations in the posterior tibial vein territory.


philic micro-catheter (“Slip-cath” selective catheter, Cook),the main venous collaterals were selectively catheterizedand occluded with metal coils (5F “Tornado,” “Nester” orstainless steel 5−8 mm coils/Cook Medical). The emboli-zation coils were delivered mainly by the “coaxial” andoccasionally the “anchor” techniques [17] in the principaldeep venous ramifications (Fig. 2). There was an averageof 6.8 (minimal, 5 and maximal, 16) coils released perprocedure. Not all the collaterals were approachable duringthis stage of the procedure, particularly the medium-sizedto small diameter veins, usually located in the perimalleolar

deep network. To avoid prolonged time and contrast con-suming interventions, the selective branch embolizationswere restricted below the ankle level, if the distal bloodpressure measured at the edge of the long 6F introducersheath were equal or superior to 50 mmHg (to excluderesidual critical ischemic perfusion) [1]. Hemostasis wasdirectly performed at the arterial puncture level after thewithdrawal of sheaths. All patients were prescribed aspirin(160 mg/day) added to clopidogrel (75 mg/day) during thefirst 3 months after the procedure followed by aspirin atlifetime, unless intolerance or previous ticlid or clopidogrel

Fig. 3. Clinical outcome after preferential posterior tibial arterialization targeting the No. 1 (medial calcanean) and No. 2 and 3 plantar (medial plantar and lateralplantar) angiosomes. (A) Toes necrosis and extended sepsis in the forefoot and plantar zones (angiosomes 2 and 3/posterior tibial dependency). (B) Forefootevolution after trans-metatarsal amputation, at 1 year. (C) Concomitant heel ulceration in the initial clinical setting (angiosome 1/ posterior tibial dependency).(D) Heel evolution at 1 month, at 3 months (E) and at 5 months (F).


medication, in which case the former was prolonged. Oralanticoagulation, if present, was continued without addingantiplatelet therapy.

2.4. Follow-up

In the postoperative period, patients were assessed byclinical examination (Fig. 3), tcPO2 measurements andduplex scan evaluations before discharge and at one monthafterward, and by clinical and duplex assessment every 6months, thereafter. For those who have expressed clinicaldeterioration as primary end-point sustained by correspon-dent cutaneous oximetry and duplex findings, an angio-CTor MRI was undertaken. Mean follow-up was 21.5 months(in the range 1–62 months).

2.5. Statistical analysis

All results were reported in an intention to treat analysis.The Kaplan-Meier life-table method was utilized todetermine the outcome of patency, the clinical success,limb salvage and patients survival rates at sequentialintervals. All data were incorporated in a “Graph Pad InStat” statistics software (Graph Pad Software, San Diego,CA, USA).

3. Results

Duplex assessment documented deep venous reflux in 12(46%) limbs, whereas the level of veins duplicationsexpressed a distal popliteal location in 18 (69%), a proximalpopliteal in 6 (23%) and superficial femoral distribution in 2

(8%) cases. Anatomical variations in the origin anddistribution of the deep venous trunks were assigned in 6(23%) instances (mostly regarding the anterior tibialterritory). The postoperative tcPO2 evaluation showed amean increase in the cutaneous oxygen tension with 17mmHg (12–32 mmHg) in the early post-procedural period in23 limbs (88%), practically unchanged values in two (8%)and decline of the initial level in one other case (4%) (listed inthe initial technical failures). The more relevant raises of thetcPO2 were detected in the angiosomes dependent to theposterior tibial versus anterior tibial or peroneal irrigation.Primary technical success was assented in 80% (21/26) cases.In five instances, the procedure could not be completedfollowing the common protocol. In one case, the distalvenous anastomosis of the PTFE graft was unachievable in acontext of extensive fibrosis (multiple previous interven-tions) and inflammation. In one other, we noted a walllaceration of the venous conduit during the devalvulationstage, requesting hemostasis by direct embolization of themain selected trunk. An alternative deep vein was arterializedoutside the angiosomes orientation, yet the case was rankedas technical failure. There were three other situations whenthe distal embolizations of the collaterals could not becompleted: in two limbs, a rich network of fine branches withsharp angled retrograde emergence and difficult endovas-cular manoeuvres were noted, while in the remaining onecase, the embolization stage was discontinued following anintolerance to contrast agents. For the whole five technicalfailures, two compulsory arterializations (non angiosomes-oriented), one minor and one major amputations wereneeded. The cumulative primary and secondary patency,including the initial five technical failures in an intention to

Fig. 5. Survival curve of the patients.


treat analysis was (±S.E.M.): 66±9%, 60±10% and 48±14%,at 12, 24 and 36 months, respectively. Limb salvage,applying the same approach was: 73±10% at 1 year andsteady 73±10% afterwards, while the clinical successexpressed: 68±10%, 60±11% and 60±11% at identical timeintervals (Fig. 4). The overall 30-day perioperative compli-cation rate was 26% (7/26 cases). Among the six majorcomplications noted, in four cases, diligent medical treatmentprovided a favorable clinical outcome: two transient cardiacdeficiencies after the initiation of the arterio-venousv(a-v)flow, compensated by inotropic and diuretic agents and tworeversible contrast induced renal failures, controlled bytemporary dialysis. The two other unavoidably complica-tions assembled a compartmental syndrome and venousgangrene in the first and graft infection followed by itssuppression in the second, requesting in both cases majoramputations. One minor complication was documented,referring a venous branch perforation during the selectivecatheterization of the collaterals, with self-limiting effectsand no clinical implications.

The angiosome model for arterializations [9,10] wasintended in the following orientations: 12 limbs with trophiclesions of the anterior leg and ankle, the forefoot, toes anddorsum of the foot were treated by preferential anterior tibialdeep veins arterializations (anterior tibial and dorsalis pedisdependent angiosomes), 11 other cases owned revascular-izations via the posterior tibial veins for ulcers in the medialcalcaneal, plantar, hallux, medial malleolar and Achilleanlocalizations (angiosomes relying the posterior tibial andplantar vascular bundles) (Figs. 2 and 3), while in theremnant three others, targeted arterializations of the peronealdeep veins were focused referring to wounds located in thelateral calcaneal and lateral malleolar territory (Table 4).

Clinical surveillance in a multidisciplinary program: forthe whole group, complete healing with or without minoramputations (Fig. 3) were noted in 14 treated limbs (54%)

Fig. 4. Cumulative patency, clinical success and limb salvage data.

(six limbs/Wagner Grade 3 and eight limbs/Wagner Grade4). In six other instances (23%), conspicuous recovery trendswith improvements in the quality of life were noted;however, those limbs have never become totally ulcer-free(one /Wagner Grade 3 and five/Wagner Grade 4). Recurrentulcerations were observed in 15 cases (57%), mainlybetween the 10th–15th months of follow-up. Globally, asum of six (23%) major amputations (all in Wagner Grade 4and severe neuropathic limbs) were documented at differentstages of the follow-up (one technical failure, two incomplications, two after iterative bypass thrombosis andone requested by a systemic uncontrolled sepsis). Neverthe-less, it should be noted that for all the successfully treatedlimbs in this series, either minor amputations (phalanges orforefoot) or intentional mummification of the affected toeswere requested. The 30-day perioperative mortality rate was0%. The survival rates (Fig. 5) were 93%, 67% and 54% at12, 24 and 36 months, respectively.

4. Discussion

Venous arterialization to treat advanced CLI presenta-tions has emerged as extreme therapeutic alternative in casesfeaturing impracticable conventional arterial reconstructionssince many decades ago [6]. In the recent years, someauthors seem to provide rather encouraging [3,6,7,18] thanpessimistic results [8] of this evolving technique in studiesfocusing essentially on arterializations of the superficialsaphenous system.

Although not new, the deep vein arterialization wasseldom reported as farthest alternative for limb salvage,gathering either experimental [19] or staged interventions inthe posterior tibial veins [20], often without systematicvalvular avulsion [6,20].

Though rational, the concern for destruction of the venousvalves has given rise to miscellaneous techniques [3,6,7,18]concerning either metallic olives or retrograde balloons and0.035-mm guidewire-enhancing valves disruption, sustainedor not by Fogarty catheters [6]. The sole hydraulic effect of


the new pulsatile flow has been equally evoked to be strengthenough to neutralize the valvular barrier, especially in thedeep venous circulation [6,20]. This latest concept remainshowever controversial [6]. In our practice, we stressed theeffort to restore the distal-calf and foot perfusion, by tibialveins arterializations inducing devalvulation and endovas-cular collateral occlusion, in inoperable diabetic arteriallesions [1,20], when all types of conventional reconstruc-tions [22] have previously failed or were impracticable(Tables 1 and 3).

In this series, the valvular incompetence was primaryinduced by the first crossing of the 0.035 hydrophilicguidewire and completed by the passage of a 6F 80-cm-longintroducer sheath through the distal deep vein. The lack ofvalvular tensile stress was documented by injecting a smallamount of contrast before proceeding the embolization stage.Interestingly, deep venous system reflux was assessed in 12(46%) limbs, whereas only 7 (26%) exhibit previous clinicalnotifications for venous insufficiency (Table 2). This findingseem to correlate to the presence of 88% neuropathic limbsin this cohort of patients (Table 2), withholding a possible“idiosyncratic deep venous distension” context by autonom-ic denervation [16].

Peripheral neuropathy has also been evoked to enhancecomplementary capillary shunting and local redistributionof the dermal, subdermal and muscular blood flow[1,12,16,21]. This process that equally involves ascendantvenous collectors [23] making possible the blood masstransfer from deep to the surface [16,23], may provide atheoretically benefic effect in remote tissue oxygenationaccompanying arterializations [6]. Similar conditions couldbe assumed to be casted by the SAVES technique thatinduces a rise in the venous pressure in these open shunts andredistribution of the oxygenated blood flow.

In comparing the superficial saphenous veins [3,6–8,18]versus the deep veins arterializations, these latest mayprovide the advantages of a larger availability and theavoidance of manipulations in surrounding infected skinareas that delineate ulcers, commonly inducing inflamma-tion, sclerosis and thrombosis of the neighboring vessels.Additionally, the duplicated calf veins may allow adeliberate orientation in arterializations by targeting precisetissular and cutaneous territories (angiosomes) upon thelocation of the main wounds. The “angiosomes model” forperfusion was originally described by Taylor [9,24] anddeveloped by Attinger et al. [10,14] in the plastic surgicalfield. Although without unanimously proved clinicallyevidence, it seems to gain more interventional adeptsparticularly in the diabetic limb salvage domain [14,25].The concept delineates the human body several “block-zones” for tissues perfusion, the “ angiosomes,” encompass-ing a topographical dependency upon specific arterial(arteriosomes) and subsequent venous (venosomes) bundlesfor irrigation. This topic was adopted also by ourmultidisciplinary diabetic “foot-team” (Fig. 1) in approach-ing the ischemic diabetic foot ulcers [26]. In the present

study, by adding the AMR-policy in the SAVES technique,we documented a correct healing with or without minoramputations in 54% of cases and avoidance of the previouslystated high risk for major amputation in 20 limbs (73% limbsalvage and 68% technical success rates at 12 months offollow-up).

A global increase in the tcPO2 (mean 17 mm Hg), wasnoted in 88% of cases, particularly following arterializationsin the posterior tibial dependent angiosomes. This observa-tion may designate the more dense network of fine collateralsand ascendant perforators described in this territory [23,24].Although the ischemic presentations are reputed to share a“patchy distribution” in diabetic neuroischemic limbs[12,21], Faglia et al. showed in a recent study that improve-ments in the pcPO2 above 40 mm Hg after revascularizationcorrelates to a N90% likelihood for wound healing [4].

From the AMR viewpoint, the venous tributaries couldbe more often than the arterial axes subject to anatomicalvariations [17,23,24]. We believe that systematic preopera-tive duplex assessment added to the road-mapping imagesmay provide important information in detecting theseparticularities (23% limbs in our cohort). We also observedthat staged embolizations of the major collaterals weretechnically more demanding in the posterior tibial venousterritory, according to the steady presence of largerperforators or communicants, gaping in the frequentneuropathic context, sometimes appointing a sharp retro-grade emergence and more arduous cannulation.

All the patients presented discrete to moderate lower limbedemas after arterializations .This phenomenon showed aspontaneous resolution during the first month in 20 cases(77%), excluding the technical failures. As noticed, in onecase (4%) persistent swelling added to a compartmentalsyndrome and venous gangrene were documented endingwith limb amputation. Two patients (7%) developedprecocious and reversible postoperative symptoms of cardiacinsufficiency (both featuring LVEF=35%), and this subsetseems to agree with other similar observations suggestingthat the more distal the location of the outflow a-v shunt, thelower likelihood to enhance secondary cardiac intolerance[6,18]. There was one other (4%) more insidious andretarded presentation of left ventricular decline, in a patientwith severe inoperable coronary disease who died fourmonths after the intervention. Referring the 66% cumulativePTFE graft patency noted at 1 year, the rate should correlate42% end-stage-renal-disease (ESRD) patients receivingsystematic low-molecular-weight heparin during dialysis,18 (69%) having previous ticlopidin or clopidogrel medica-tion for synchronous cardiovascular reasons and 24 (92%)including statins in their treatment.

By assembling the SAVES method and the synchronousmultidisciplinary approach, our 73% limb preservationquota at 1 year seems to match with other publishedreports [3,6,7,18] although cautiously comparing theinclusion criteria in each of these cohorts. Seemingly, ininterpreting the relatively modest clinical success (68% at


12 months and 60% at 24 and 36 months) in our group, thepresence of 73% Wagner Grade 4 foot lesions, hidingmiscellaneous form of osteomyelitis and 88% peripheralneuropathic limbs (Table 2), may represent two explanatoryfactors for the observed delay in tissue recovery. From thepatient's point of view, the main benefit of the SAVESmethod at this stage of the technique, may probably be thiscomplementary attempt in preserving the ankle and theback- or forefoot, with major implications in ambulationand survival [1,2,4,12]. This goal may be of greatestconcern in threatened diabetic limbs without direct arterialsolutions, albeit at high jeopardy for major amputation[2,12], even though tissue loss by minor amputations ortoes mummifications was inevitable.

The present study may express some clear limitations:beyond its retrospective nature and the small number ofcases included, the results in terms of limb salvage andclinical recovery are difficult to be compared with otherseries focusing on conventional arterial reconstructions orsuperficial venous arterializations, since the technique itselfand the profile of the cases enrolled are different. Secondly,the “SAVES” method is probably an evolving procedure.Some technical difficulties and inherent complicationsencountered, particularly in the collateral embolizationstage, may be eluded by an unfolding experience and theuse of new and more sophisticated endovascular devices.Third, additional measures to assess the inflow rate in thearterialized calf veins, or improvements in the peripheralmicro-circulation (laser Doppler fluxometry, fluorescencevideomicroscopy, plethysmometry or stress echocardiogra-phy) could be added in the follow-up protocol and mayprovide useful information for patient selection and betterpostoperative surveillance. Finally, the evaluation of theclinical success and limb preservation in these “high-risk”diabetic patients was undoubtedly influenced by theparticipation of the multidisciplinary team of specialists[5,12,26], that represents an autonomous variable, with hisown changing efficacy during the different periods ofthe study.

5. Conclusion

Deep calf venous arterialization (the SAVES method)might be considered as extreme alternative to attempt tissuepreservation in limbs unfit for conventional arterial revascu-larizations. The guidance of the angiosome model forreperfusion seems, at this stage of our experience, a probablyuseful tool in enhancing distal ulcer healing. Larger groupsof study, providing comparative and randomized data areneeded to ascertain these preliminary observations.

Acknowledgments

We would like to acknowledge all the members of themultidisciplinary “Diabetic-Foot Group” also the radiology,

cardiology and nephrology departments of our institutionand their teams, for the assiduous work and punctual supportin managing these challenging cases.

References

[1] Norgreen L, Hiatt WR, Dormandy JA, Nehler MR, Harris KA, FowkesFGR, on behalf of the TASC II Working Group. Inter-societyconsensus for the management of peripheral arterial disease (TASCII.). Eur J Vasc Endovasc Surg 2007;33(Suppl1):S32–S55.

[2] Dormandy J, Heek L, Vig S. Major amputations: clinical patterns andpredictors. Semin Vasc Surg 1999;12:154–61.

[3] Jacop S, Nassef A, Belli AM, Dormandi JA, Taylor RS. Vascularsurgery society of Great Britain and Ireland: distal venous arterializa-tions for nonreconstructable arterial disease. Br J Surg 1999;86(5):694.

[4] Faglia E, Clerici G, Caminiti M, Quarantiello A, Curci V, Morabito A.Predictive values of transcutaneous oxygen tension for above-the-ankle amputation in diabetic patients with critical limb ischemia. Eur JVasc Endovasc Surg 2007(33):731–6.

[5] Apelquist J, Larsson J, Agardh CD. Long-term prognosis for diabeticpatients with foot ulcers. J Intern Med 1993;233(6):485–91.

[6] Lu XW, Idu MM, Ubbink DT, Legemate DA. Meta-analysis of theclinical effectiveness of venous arterializations for salvage of criticallyischaemic limbs. Eur J Vasc Surg 2006;31:493–9.

[7] Lengua F, La Madrid A, Acosta C, Barriga H, Maliqui C, Arauco R,Lengua A. Arterialization of the distal veins of the foot for limb salvagein arteritis. Technique and results. Ann Chir 2001;126:629–38.

[8] Matzke S, Pitkanen J, Lepantalo M. Does saphenous vein arterializa-tions prevent major amputation in critical leg ischemia? A comparativestudy. J Cardiovasc Surg (Torino) 1999;40:845–7.

[9] Taylor GI, Pan WR. Angiosomes of the leg: anatomic study andclinical implications. Plast Reconstr Surg 1998;102(3):599–616.

[10] Attinger CE, Evans KK, Bulan E, Blume P, Cooper P. Angiosomes ofthe foot and ankle and clinical implications for limb salvage:reconstruction, incisions and revascularization. Plast Reconstr Surg2006;117(7 Suppl):261S–93S.

[11] Dayal R, Kent C. Standardized reporting practices. Rutherfordvascular surgery. 6th ed. Philadelphia: Elsevier Saunders Inc, 2005.p. 41–52.

[12] Driver V, Landowski MA, Madsen JL. Neuropathic wounds: thediabetic wound. Acute and chronic wounds-current managementconcepts. Philadelphia: Mosby Elsevier Inc, 2007. p. 307–36.

[13] Diehm N, Pattynama PM, Jaff MR, Cermonesi A, Becker GJ, HopkinsLN, Mahler F, Talen A, Cardella JF, Ramee S, van Sambeek M,Vermassen F, Biamino G. Clinical endpoints in peripheral endovas-cular revascularization trials: a case for standardized definitions. Eur JVasc Endovasc Surg 2008(36):409–19.

[14] Attinger CE, Evans KK, Mesbahi A. Angiosomes of the foot andangiosome-dependent healing. Diabetic foot, lower extremity arterialdisease and limb salvage. Philadelphia: Lippincott Williams &Wilkins, 2006. p. 75–107.

[15] Wind GG, Valentine RJ. Popliteal artery. Anatomic exposures invascular surgery. Baltimore: William&Wilkins, 1991. p. 373–410.

[16] Pureval TS, Goss DE, Watkins PJ, Edmonds ME. Lower limb venouspressure in diabetic neuropathy. Diabetes Care 1995;18(3):377–81.

[17] White RI, Pollak JS. Controlled delivery of pushable fibered coils forlarge vessel embolotherapy. Vascular embolotherapy a comprehensiveapproach, Volume 1, general principles, chest, abdomen and greatvessels. Springer, 2006. p. 35–43.

[18] Ozbek C, Kestelli M, Emrecan B, Ozsoyler I, Bayatli K, Yasa H, LafciB, Gurbuz A. A novel approach: ascending venous arterialization foratherosclerosis obliterans. Eur J Vasc Endovasc Surg 2005;29:47–51.

[19] Ozek C, Zhang F, Lineaweaver WC, Chin BT, Newlin L, Eiman T,Buncke HJ. Arterialization of the venous system in a rat lower limbmodel. Br. J Plast Surg 1997;50(6):402–6.


[20] Chen XS, Lin T, Chen DL, Guan YB. Venous arterialization ofposterior tibial vein channel for extensive arterial occlusion of lowerextremities. J Surg Concepts Pract 1998;3:219–21.

[21] Krisnan SM, Rayman G. Microcirculation and the diabetic foot.The foot in diabetes. 4th ed. London: Wiley & Sons Ltd, 2008.p. 41–51.

[22] Beard JD.Which is the best revascularization for critical limb ischemia:endovascular or open surgery? J Vasc Surg 2008(48):11S–6S.

[23] Calota F, Mogoanta S, Intorcaciu M, Pasalega M, Popescu CF, VasileI, Mesina C. The venous system of the lower limbs. Rom J MorpholEmbriol 2007;48(4):355–60.

[24] Taylor GI, Caddy CM,Watterson PA, Crock JG. The venous territories(venosomes) of the human body: experimental study and clinicalimplications. Plast Reconstr Surg 1990;86(2):185–213.

[25] Neville RF, Attinger CE, Bulan EJ, Ducic I, Thomassen M, SidawyAN. Revascularization of a specific angiosome for limb salvage: doesthe target artery matter? Ann Vasc Surg 2009;23(3):367–73.

[26] Alexandrescu V, Hubermont G, Philips Y, Guillaumie B, Ngongang C,Vandenbossche P, Azdad K, Ledent G, Horion J. Selective angioplastyfollowing an angiosome model of reperfusion in the treatment ofWagner 1–4 diabetic foot lesions: practice in a multidisciplinarydiabetic limb service. J Endovasc Ther 2008;15:580–93.

deep calf veins arterialization for inferior limb preservation in diabetic patients with extended...

Documents