Early and Five-year Amputation and Survival Rate of Diabetic Patients with Critical Limb Ischemia: Data of a Cohort Study of 564 Patients

Transcription

1 Eur J Vasc Endovasc Surg 32, 484e490 (2006) doi: /j.ejvs , available online at htt:// on with Critical Limb Ischemia: Data of a Cohort Study of 564 Patients E. Faglia, 1 * G. Clerici, 1 J. Clerissi, 2 L. Gabrielli, 3 S. Losa, 3 M. Mantero, 1 M. Caminiti, 1 V. Curci, 1 T. Luattelli 2 and A. Morabito 4 1 Diabetology Centre-Diabetic Foot Centre, 2 Interventional Radiology Laboratory, Policlinico Multimedica, Sesto San Giovanni (Milano), Italy, 3 Vascular Surgery, and 4 Medical Statistics Unit, University of Milan, Italy Objective. To evaluate the early and late major amutation and survival rates and related risk factors in diabetic atients with critical limb ischemia (CLI). Design. Retrosective study. Methods. Revascularization feasibility, major amutation, survival rate and related risk factors were recorded in 564 diabetic atients consecutively hositalized for CLI from 1999 to 2003 and followed until June Results. Periheral angiolasty (PTA) was carried out in 420 (74.5%), byass graft (BPG) in 117 (20.7%) atients. In 27 (4.8%) atients both PTA and BPG were not ossible. Twenty-three above-the-ankle amutations (4.1%) were erformed at 30 days: 6 in PTA atients, 3 in BPG atients, 14 in non revascularized atients. In the follow-u of 558 atients (98.9%), 62 reeated PTAs and 9 new BPGs, 32 new major amutations (16 in PTA atients, 14 in BPG atients and 2 in nonrevascularized atients) were erformed. Major amutation was associated with absence of revascularization (OR 35.9, < 0.001, CI 12.9e99.7), occlusion of each of the three crural arteries (OR 8.20, ¼ 0.022, CI 1.35e49.6), wound infection (OR 2.1, ¼ CI 1.3e3.6), dialysis (OR 4.7, ¼ CI 1.9e11.7) increase in TcPO 2 after revascularization (OR 0.80, < CI 0.74e0.87). One hundred seventy three atients died during follow-u and this was associated with age (HR 1.05, < CI 1.03e1.07), history of cardiac disease (HR 2.16, < CI 1.53e3.06), dialysis (HR 3.52, < CI 2.08e 5.97), absence of revascularization (HR 1.68, < 0.001, CI 1.29e2.19) and imaired ejection fraction (HR 1.08, < 0.001, CI 1.05e1.09). Conclusions. In diabetic atients with CLI the revascularization is feasible in most cases and allows a low rate of early major amutation. This rate is higher in the follow-u eriod. Major amutation is very high in atients where revascularization is not feasible while the high mortality rate is due to the serious comorbidities observed in these atients. Keywords: Diabetic foot; Critical limb ischemia; Periheral angiolasty; Periheral byass graft; Above-the-ankle amutation; Survival. Studies on diabetic atients with eriheral arterial disease suort the effectiveness of revascularization, 1e3 however very few studies reort the outcomes of atients unsuitable for revascularization. 4 We do not know any study in diabetic oulation that used the diagnostic criteria of critical limb ischemia (CLI) roosed in 2000 by TransAtlantic Inter- Society Consensus (TASC). 5 The urose of this study is to evaluate the outcomes of diabetic atients with TASC criteria of CLI. *Corresonding author. E. Faglia, Diabetology Centre-Diabetic Foot Centre, Policlinico MultiMedica, Via Milanese 300, Sesto San Giovanni (Milan), Italy. address: ezio.faglia@multimedica.it URL: htt:// We reort the feasibility of revascularization with eriheral angiolasty (PTA) or byass graft (BPG), the early major amutation rate, the incidence of redo vascular rocedures and of new major amutations, the survival rate and related risk factors for amutation and survival in diabetic atients consecutively hositalized for CLI. Materials and Methods Protocol All diabetic atients referred to our Diabetic Foot Centre were assessed for the resence of sensory-motor neuroathy and eriheral arterial occlusive disease / $35.00/0 Ó 2006 Elsevier Ltd. All rights reserved.

2 485 Sensory-motor neuroathy was detected by means of vibration ercetion threshold > 25 V at biothesiometer, insensitivity in >5/9 foot oints at Semmes- Weinstein 10 g filament, and absence of Achilles tendon reflex. Periheral arterial disease was susected if one foot ulse was reduced or absent, ankle-ressure was <70 mmhg when assessable, transcutaneous oxygen tension (TcPO 2 ) at the dorsum of the foot was <50 mmhg, and significant obstructions were resent at dulex scanning. All atients with these arameters were referred to an angiograhic study and, if obstruction > 50% of vessel diameter was resent, PTA was erformed in the same session. 6 In atients unsuitable for PTA a by-ass graft (BPG) was considered. The atients in whom PTA or BPG were not ossible, received a theray with rostanoids (alrostadil-a-ciclodextrine 60e120 mgr/day) and ercutaneous symathectomy with henol injection under CT guidance. In atients in whom theraies did not relieve the rest ain or the gangrene was extended above the Choart joint, major amutation was erformed. For all atients, antalgic medication, TcPO 2 and ankle ressure were reassessed 5 days after the PTA, BPG, symathectomy or rostanoids infusion. Limb salvage In atients comlaining of rest ain without foot ulcer, the disaearance of ain with discontinuation of antalgic theray was considered to be a successful limb salvage. In atients with foot ulcer we considered limb salvage successful when the ulcer healed and lantar stand was maintained, even when achieved by tarsal-metatarsal amutation 7,8 and the atient was able to walk without crutches or artificial leg. Conversely, any above-the-ankle amutation was considered a failure (major amutation). Comlications Any event that required secific medical or surgical treatment or rolonged hosital stay following PTA or BPG was recorded and considered a comlication. Haematoma formation at the uncture site was considered not significant unless necessitating either surgery and/or acked red blood cell transfusion. Follow-u After hosital discharge all atients with foot ulcer were examined weekly until ulcer healing. All atients were rovided with extra-dee rocker shoes with soft thermoformable leather and customized insoles. Restenosis after PTA was susected when ain or an ulcer recurred. In these situations ankleressure and TcPO 2 were reassessed and dulex scanning was erformed. 9 If ankle ressure and TcPO 2 were significantly worse (<15% of the ost-pta value) and Dulex scanning was ositive, the atient underwent a new angiograhic evaluation and a further PTA, if ossible, or a BPG, if PTA imossible, was erformed. Morhological restenosis was not investigated since in absence of rest ain or ulcer reaearance we do not erform any revascularization, therefore we consider it clinically irrelevant. 10 BPGs were followed according to the vascular surgery rotocol. 11 The graft atency was assessed with clinical examination and ultrasound study at 30 days, 3,6,12 months and thereafter every 6 months. For every invasive treatment (PTA, BPG, symathectomy, surgical oeration on the foot) the written informed consent form was obtained from the atient. PTA clinical restenoses, BPG closures, reeat vascular rocedures, new above-the-ankle amutations, vital status as well as date and cause of deaths were recorded. Statistical methods Descritive statistics were reorted as average values and 1 standard deviation (SD) for continuous variables and as ercentages for discrete variables. The relationshi among the considered variables and the risk of above-the-ankle amutation was evaluated by multile logistic regression and data were reorted with odds ratio (OR) and 95% CI. The Cox regression model was adoted and Hazard Ratio (HR) for the tested variables was reorted in order to check the association of atients characteristics and the life time lasted from the entry in the study until the death or study closure. The time to the major amutation and death was studied by alying the Kalan Meier aroach and roduct/limit curves were built u. The Stata 7.0 software ackage (Statistics/Data Analysis, Stata Cororation, 4905 Lakeway Drive, College Station, Texas USA, 800-STATA-PC) was used. Results Patient oulation and treatment From January 1, 1999 to December 31, 2003, 902 diabetic atients were consecutively hositalized for foot ulcer and/or rest ain.

3 486 E. Faglia et al. 567 atients in whom our diagnostic rotocol indicated the resence of CLI were referred to an angiograhic study. Three atients had no angiograhic evidence of stenoses > 50% of vessel diameter, and were excluded from the analysis. Thus the study oulation included 564 (62.5%) atients. Revascularization was erformed in 537 (95.2%) atients. PTA was erformed in 420 (74.5%) atients. Table 1 reorts the number of treated stenoses and occlusions, the number of stents laced in every artery of the ischemic limb and the rate of successful PTA rocedures. A BPG was erformed in 117 (20.7%) atients. The BPG was axillo-femoral in one atient, femoraloliteal (16 PTFE, 45 vein graft) in 61 atients, femoral-infraoliteal (17 PTFE, 40 vein graft) in 57 atients. In 22 of these 117 atients a combined rocedure, PTA lus BPG, was erformed: in 10 atients an iliac PTA and femoral-oliteal BPG, in 12 atients femoral-oliteal BPG and an infraoliteal PTA. In 27 (4.8%) atients neither a PTA nor a BPG was ossible due to high surgical risk or lack of outflow. In all these atients a treatment with rostanoids was started, but was immediately stoed in 5 atients for hyotension and in 2 for angina. In 5 atients a ercutaneous chemical symathectomy was also erformed, while in the remaining atients this rocedure was not considered aroriate due to the resence of a serious cardioathy and other comorbidities (Table 2). The ankle-ressure could not be measured in 297 (52.7%) atients, because of the absence of both tibial arteries in 105 atients or due to the resence of arterial calcifications in 192 atients. In-hosital mortality Four atients died during their hosital stay. Three BPG atients died: 1 for neumonia, 1 for acute myocardial infarction and 1 for congestive heart failure. One non-revascularized atient died suddenly the night following the angiograhic study. Early limb salvage Of the 537 revascularized atients, in 85 atients with rest ain without foot ulcer the ain disaeared and the antalgic medication was discontinued. In 443 atients with foot ulcers a comlete healing of the lesions occurred with dressing or minor amutation. Nine (1.7%) major amutations were carried out at 30 days. Two above- and 4 below-the-knee amutations (1.4%) were erformed in PTA atients, in 5 for extensive infection of surgical wound of a Choart amutation, and in 1 for acute distal thrombosis after PTA not suitable for surgical revascularization. Of these six major amutations, three were erformed in atients treated without stent and three in atients treated with stent, without significant difference ( ¼ 0.358). Three (2.6%) above-the-knee amutations were erformed in BPG atients with crural by-ass, 2 with PTFE and 1 with venous conduit following acute non treatable graft closure. Of the 27 non-revascularized atients, in 2 atients the foot lesion healed with dressing whereas in 14 atients major amutation (10 above- and 4 belowthe-knee) was erformed at 30 days for rest ain and extensive gangrene. In the other 11 atients the foot lesion did not heal but there was not worsening, the ain decreased but did not disaear and the antalgic medication was reduced but not discontinued. PTA and BPG comlications Table 3 reorts the non fatal comlications in PTA and BPG atients. Follow-u 558 (98.9%) atients, 415 of the PTA grou, 116 of the BPG grou and 27 of non revascularized grou, were Table 1. Number of treated stenoses and occlusions, their ercent of TASC non eligible, number of stents laced in every artery and ercent of success of PTA rocedures Artery Stenoses Length > 4 cm Successful ta Stents Occlusions Length > 2 cm Successful ta Stents Iliac trunk (32.3%) 31 (100%) (50.0%) 4 (100%) 4 Femoral (66:4) 146 (100%) (65.4%) 78 (100%) 40 Politeal (36.6%) 112 (100%) (44.4%) 36 (100%) 18 Anterior tibial (87.9%) 76 (83.5%) e (90.3%) 63 (24.5%) 6 Posterior tibial (78.9%) 46 (80.7%) e (90.0%) 40 (13.7%) 7 Peroneal (71.2%) 100 (80.0%) e (62.1%) 32 (25.8%) 4 A concomitant femoral-oliteal and crural PTA was erformed in 151 atients: in 116 with angiolasty of one crural artery, more than one in 35 atients.

4 487 Table 2. Demograhic and clinical characteristics of study oulation (N [ 564) in revascularized atients with PTA and BPG, and in revascularized and non revascularized atients Variables PTA n ¼ 420 BPG n ¼ 117 Revascularized atients n ¼ 537 Non revascularized atients n ¼ 27 Age (years) Females (n) 148 (35.2%) 37 (31.6%) (34.9%) 13 (48.1%) Insulin theray (n) 255 (60.7%) 73 (62.4%) (61.1%) 14 (51.9%) Diabetes duration (years) Sensory-motor neuroathy (n) 343 (81.7%) 98 (83.8%) (82.1%) 24 (88.9%) Creatinine (mg/dl) (n ¼ 532) Dialysis (n) 24 (5.7%) 8 (7.0%) (5.9%) e e Antihyertensive theray (n) 304 (72.4%) 72 (61.5%) (70.0%) 20 (74.1%) Cardiac disease (n) 225 (53.6%) 64 (54.7%) (53.8%) 24 (88.9%) Ejection fraction (%, n ¼ 323) History of stroke (n) 55 (13.1%) 19 (16.2%) (13.8%) 9 (33.3%) Wagner grade (n) 0 66 (15.7%) 19 (16.2%) (15.8%) 3 (11.1%) (15.2%) 14 (12.0%) 78 (14.5%) 5 (18.5%) 2 59 (14.0%) 14 (12.0%) 73 (13.6%) 5 (18.5%) 3 41 (9.8%) 11 (9.4%) 52 (9.7%) 3 (11.1%) (45.2%) 59 (50.4%) 249 (46.4%) 11 (40.7%) Infected ulcer (n) 270 (64.3%) 75 (64.1%) (64.2%) 17 (63.0%) TcPO2 before the treatment (mmhg) TcPO2 after the treatment (mmhg) followed from January 1st 1999 until June 30th The mean follow-u was years. CLI recurrence during follow-u A first eisode of clinical restenosis in the PTA grou occurred in 76 (18.3%) atients. The cumulative atency rate at 5 years was 78% (CI 71e83%). PTA rocedures were successfully reeated in 62 (81.6%) atients and 1 atient died suddenly the night after PTA. In the remaining 14 atients a further PTA was not feasible, and 8 atients underwent a BPG. Non-treatable graft closure occurred in 22 (18.8%) atients of the BPG grou. The cumulative atency rate at 5 years was 77% (CI 67e85%). Table 3. Non fatal comlications and their treatment in revascularized atients (N [ 537) PTA atients (N [ 420) N Treatment Myocardial infarction 1 Intensive care unit Angina 1 Medical treatment Cardiac arrhythmia 1 Intensive care unit Left ventricular failure 1 Medical treatment Chest ain 1 Investigation, no treatment Acute renal failure 1 Medical treatment without dialysis Haematoma at the access site 2 Transfusion 1 Investigation without theray Access site seudoaneurysms 3 Surgical treatment Thrombosis 5 Thrombolysis effectiveness 1 Above-the-knee amutation BPG atients (N [ 117) N Treatment Thrombosis 8 Thrombectomy 3 Above the knee amutation Venous fistula 1 Surgical ligature Major amutation during follow-u Thirty-two major amutations were erformed. Three above- and 13 below- the-knee amutations were erformed in the PTA grou, 14 because of recurrence of rest ain and extensive gangrene due to non-treatable restenosis and 2 because of osteomyelitis of the heel. Ten above- and 4 below-the-knee amutations were erformed in the BPG grou because of non-treatable occlusion of the graft and 1 atient died during hosital stay because of setic shock. Among the major amutations in BPG grou, 3 were erformed in atients with femoral-oliteal graft (2 PTFE, 1 venous conduit), and 11 in infraoliteal graft (8 PTFE, 3 venous conduit). Two above-the-knee amutations were erformed in non revascularized atients because of a foot lesion worsening. Table 4 summarize the number of major amutations erformed in the early and follow-u eriod. Fig. 1 shows the Kalan-Meier curves of major amutation in the three grous. The results of multile logistic regression analysis erformed for the association between the recorded variables and the abovethe-ankle amutation are showed in Table (30.7%) atients died during follow-u. 120 (28.9%) atients died in the PTA grou, 31 (26.7%) in the BPG grou and 22 (81.4%)in the nonrevascularized grou. Fig. 2 shows the Kalan-Meier survival curves of the three grous. The causes of death are reorted in Table 6. Table 7 reorts the Hazard Ratio of the Cox model erformed for the association between the recorded variables and mortality.

5 488 E. Faglia et al. Table 4. Number of above-the-ankle amutations erformed in the early and follow-u eriod in PTA, BPG and no revascularized atients Patients treatment Above-the-ankle amutation Total At 30 days Follow-u Angiolasty (N ¼ 420) 6 (1.4%) 16 (3.8%) 22 (5.2%) Byass graft (N ¼ 117) 3 (2.6%) 14 (12%) 17 (14.5%) No revascularization (N ¼ 27) 14 (51.9%) 2 (7.4%) 16 (59.3%) Table 5. Multile logistic regression analysis between recorded variables and above-the-ankle amutation Variables Odds ratio Confidence interval Absence of 35.9 < e99.7 revascularization (n) Occlusion of each of the crural e49.6 arteries (n) Wound infection (n) e3.6 Dialysis (n) e11.7 TcPO 2 increase (1 mmhg) 0.80 < e0.87 Discussion In this study atients were enrolled consecutively. No diabetic atient with TASC arameters of CLI were excluded. The ercentage of atients undergoing revascularization was very high considering that PTA and BPG were erformed in a oulation enrolled consecutively and not in selected atients with a resumed high success of revascularization. 12 The high revascularization rate of our cases deends on the use of both PTA and BPG. Many atients who underwent PTA were not suitable for BPG because their three leg arteries or edal arteries were not atent or because of the high surgical risk. A high ercentage of atients in whom PTA was not feasible were revascularized with BPG. In our rotocol PTA was the first-choice rocedure of revascularization allowing outcomes similar to BPG. 13 In our ractice PTA is effective also in long obstructions of the infraoliteal arteries. 14,15 The in-hosital mortality was very low in PTA atients, and was also low in BPG atients in comarison with the literature data. 16 Revascularization is the best theray to eliminate ain and heal foot lesion in atients with CLI. 17,18 Revascularization is esecially necessary when a foot surgery is required. 19,20 The extensive use of revascularization, in association with a good surgical aroach to the foot lesion, resulted in the short eriod in a very high rate of limb salvage. 2,21,22 However, major amutation were noted during follow-u. In the PTA grou most of the major amutations were erformed for clinical restenosis without further ossibility of endoluminal or surgical revascularisation. All the major amutations of BPG grou were erformed because of graft closure. Although some studies reorted a higher rate of edal by-ass failure in the early eriod than in the follow-u, 23,24 our data seem to be consistent with the fast rogression of atherosclerosis and the redominantly crural localization of the atherosclerotic obstructions, which are tyical features in diabetic atients. 25,26 The major amutation rate of the non-revascularized atients is very different. The early major amutation rate of these atients is very high. The low ercentage during the follow-u can be exlained by the very low survival rate of these atients due the associated serious comorbid conditions, cardiac esecially. The Number of atients (%) Years PTA BPG no revascularization Exosed atients at risk of death PTA BPG No revascularization Fig. 1. Kalan-Meier above-the-ankle amutation estimates in PTA, BPG and non revascularized atients. Number of atients (%) Years PTA BPG no revascularization Exosed atients at risk of death PTA BPG No revascularization Fig. 2. Kalan-Meier survival estimates of PTA, BPG and non revascularized atients

6 489 Table 6. Causes of death in followed oulation (N [ 558) Cause of death PTA N ¼ 415 BPG N ¼ 116 No revascularization N ¼ 27 Cardiac disease Sudden death Stroke Cancer Pulmonary e 1 e embolism Abdominal aneurysm 1 e e Renal insufficiency 4 e e Peritonitis 1 e e Gastric hemorrhage 1 e e Cirrhosis 2 e e Pneumonia 2 2 e Geromarasmus 2 e 1 Setic shock e 1 e Suicide 1 e e Total 120 (28.9%) 31 (26.7%) 22 (81.5%) mortality rate is high also in revascularized atients, although it is lower than in the non-revascularized atients. 27e30 In order to achieve the best health-related quality of life we aim to avoid major amutation in all atients, excet in atients unable to walk and mentally incaable. In conclusion revascularization with PTA or BPG allows a very high ercentage of limb salvage in diabetic atients with CLI. In the early eriod the revascularization, in association with a good medical and surgical aroach to foot lesion, results in a very high ercentage of limb salvage, with very low in hosital mortality rate. During the follow-u the risk of above-the-ankle amutation is higher, but still lower comared to the literature data in non-revascularized atients with CLI. 31,32 Imossibility of revascularization because of the extent of arterial occlusive disease or surgical risk is a reliable marker both of a very high risk of major amutation and of a very low life exectancy, because of the severity of the comorbid conditions associated with the severity of CLI. Table 7. Cox model for association between recorded variables and mortality Variables Hazard ratio Confidence interval Age (1 year) 1.05 < e1.07 History of cardiac 2.16 < e3.06 disease (n) Dialysis (n) 3.52 < e5.97 Absence of 1.68 < e2.19 revascularization (n) Imaired ejection fraction (1%) 1.08 < e1.09 References 1 LOGERFO FW, GIBBONS GW, POMPOSELLI Jr FB, CAMPBELL DR, MILLER A, FREEMAN DV et al. Trends in the care of the diabetic foot. Exanded role of arterial reconstruction. Arch Surg 1992; 127:617e HOLSTEIN P, ELLITSGAARD N, OLSEN BB, ELLITSGAARD V. Decreasing incidence of major amutations in eole with diabetes. Diabetologia 2000;43:844e POMPOSELLI Jr FB, MARCACCIO EJ, GIBBONS GW, CAMPBELL DR, FREEMAN DV, BURGESS AM et al. Dorsalis edis arterial byass: durable limb salvage for foot ischemia in atients with diabetes mellitus. J Vasc Surg 1995;21:375e HOLSTEIN P, SORENSEN S. Limb salvage exerience in a multidiscilinary diabetic foot unit. Diabetes Care 1999;22(Sul 2):B97e B TransAtlantic Inter-Society (TASC): management of eriheral arterial disease (PAD). Eur J Vasc Endovasc Surg 2000;19(Sul A): 208e FAGLIA E, DALLA PAOLA L, CLERICI G, CLERISSI J, GRAZIANI L, FUSARO M et al. Periheral angiolasty as the first-choice revascularization rocedure in diabetic atients with critical limb ischemia: rosective study of 993 consecutive atients hositalized and followed between 1999 and Eur J Vasc Endovasc Surg 2005;29:620e GARBALOSA JC, CAVANAGH PR, WU G, ULBRECHT JS, BECKER MB, ALEXANDER IJ et al. Foot function in diabetic atients after artial amutation. Foot Ankle Int 1996;17:43e48. 8 REYZELMAN AM, HADI S, AMSTRONG DG. Limb salvage with Choart s amutation and tendon balancing. J Am Podiatr Med Assoc 1999;89:100e GROLLMAN J, LEVIN DC, BETTMANN MA, GOMES AS, HENKIN RE, HESSEL SJ et al. Recurrent symtoms following lower extremity angiolasty: claudication and threatened limb. American College of Radiology. ACR Aroriateness Criteria. Radiology 2000;215(Sul):95e ISNER JM, ROSENFIELD K. Redefining the treatment of eriheral artery disease. Role of ercutaneous revascularization. Circulation 1993;88:1534e ACR-aroriateness criteriaô for recurrent symtoms following lower extremities arterial byass surgery. Radiology 2000; 215:89e RAYMAN G, KRISHAN TM, BAKER NR, WAREHAM AM, RAYMAN A. Are we underestimating diabetes-related lower-extremity amutation rates? Diabetes Care 2004;27:1892e Byass versus angiolasty in severe ischemia of the leg (BASIL): multicentre, randomised controlled trial. Lancet 2005;366:1925e FAGLIA E, MANTERO M, CAMINITI M, CARAVAGGI C, DE GIGLIO R, PRITELLI C et al. Extensive use of eriheral angiolasty, esecially infraoliteal, in the treatment of ischemic foot ulcer: clinical results of a multicentric study of 221 consecutive diabetic subjects. J Intern Med 2002;252:225e HIRSCH AT, HASKAL ZJ, HERTZER NR, BAKAL CW, CREAGER MA, HALPERIN JL et al. ACC/AHC Guidelines for the Management of Patients with Periheral Arterial Disease (Lower Extremity, Renal, Mesenteric, and Abdominal Aortic). A Collaborative Reort from the American Association for Vascular Surgery/ Society of Vascular Surgery, Society for Vascular Medicine and Biology, Society of Interventional Radiology, and the ACC/ AHC Task Force on érectice Guidelines (Writing Committee to Develo Guidelines for the management of Patients With Periheral Arterial Disease). American College of Cardiology Web Site, Endovascular treatments for CLI, 2005,. 65e VIRKKUNEN J, HEIKKINEN M, LEPANTALO M, METSANOJA R, SALENIUS JP, The Finvasc Study Grou. Diabetes as an indeendent risk factor for early ostoerative comlications in critical limb ischemia. J Vasc Surg 2004;40:761e LEVIN ME. Preventing amutation in the atient with diabetes. Diabetes Care 1995;18:1383e1394.

7 490 E. Faglia et al. 18 OURIEL K. Periheral arterial disease. Lancet 2001;358:1257e CAMPBELL WB, PONETTE D, SUGIONO M. Long-term results following oeration for diabetic foot roblems: arterial disease confers a oor rognosis. Eur J Vasc Endovasc Surg 2000;19:174e LEPANTALO M, BIANCARI F, TUKIAINEN E. Never amutate without consultation of a vascular surgeon. Diabetes Metab Res Rev 2000;16(Sul 1):S27eS HUNTER GA. Results of minor foot amutations for ischemia of the lower extremity in diabetics and non diabetics. Can J Surg 1975;18:273e YEAGER RA, MONETA GL, EDWARDS JM, WILLIAMSON K, MCCONNELL DB, TAYLOR LM et al. Predictors of outcome of forefoot surgery for ulceration and gangrene. Am J Surg 1998;175: 388e PANNETON JM, GLOVICZKI P, BOWER TC, RHODES JM, CANTON LG, TOOMEY BJ. Pedal byass for limb salvage: imact of diabetes on long-term outcome. Ann Vasc Surg 2000;14:640e DOMENIG CM, HAMDAN AD, HOLZENBEIN TJ, KANSAL N, AULIVOLA B, SKILLMAN JJ et al. Timing of edal byass failure and its imact on the need for amutation. Ann Vasc Surg 2005;19:56e JUDE EB, OYIBO SO, CHALMERS N, BOULTON AJ. Periheral arterial disease in diabetic and nondiabetic atients. A comarison of severity and outcome. Diabetes Care 2001;24:1433e VAN DER FEEN C, NEIJENS FS, KANTERS SD, MALI WP, STOLK RP, BANGA JD. Angiograhic distribution of lower extremity atherosclerosis in atients with and without diabetes. Diabet Med 2002; 19:366e MOULIK PK, MTONGA R, GILL GV. Amutation and mortality in new-onset diabetic foot ulcers stratified by etiology. Diabetes Care 2003;26:491e CRIQUI MH, DENEMBERG JO. The generalized nature of atherosclerosis: how eriheral arterial disease may redict adverse events from coronary artery disease. Vasc Med 1998;3:241e BOYKO EJ, AHRONI JH, SMITH DG, DAVIGNON D. Increased mortality associated with diabetic foot ulcer. Diabet Med 1996;13:967e LEIBSON CL, RAMSOM JE, OLSON W, ZIMMERMAN BR, O FALLON WM, PALUMBO PJ. Periheral Arterial Disease, Diabetes, and Mortality. Diabetes Care 2004;27:2843e BERTELE V, RONCAGLIONI MC, PANGRAZZI J, TERXZIAN E, TOGNONI G, on the behalf of the i.c.a.i. (CLI) Grou. Clinical outcome and its redictors in 1560 atients with critical leg ischemia. Eur J Vasc Endovasc Surg 1999;18:401e PETRAKIS IE, SCIACCAV. Sinal cord stimulation in diabetic lower limb critical ischaemia: transcutaneous oxygen measurement as redictor for treatment success. Eur J Vasc Endovasc Surg 2000;19:587e592. Acceted 14 March 2006 Available online 26 May 2006