Which Factors Increase the Risk of Conversion to Open Surgery Following Endovascular Abdominal Aortic Aneurysm Repair?

Transcription

1 Eur J Vasc Endovasc Surg 20, (2000) doi: /ejvs , available online at on Which Factors Increase the Risk of Conversion to Open Surgery Following Endovascular Abdominal Aortic Aneurysm Repair? P. W. M. Cuypers, R. J. F. Laheij and J. Buth, on behalf of the EUROSTAR Collaborators EUROSTAR Data Registry Center, Catharina Hospital, Eindhoven, The Netherlands Objective: to identify factors that increase the risk of conversion to open surgery following endovascular repair of abdominal aortic aneurysms (AAAs) and to assess their outcome. Design: analysis of 1871 patients enrolled in the EUROSTAR collaborators registry. Materials and Methods: patient characteristics, anatomic features of the aneurysm, type of endovascular device, institutional experience and the year in which the procedure was performed were related to risk of conversion. Results: forty-nine patients (2.6%) required conversion. In 38 patients conversion was performed during the first postoperative month (primary conversions) and in 11 patients during follow-up (secondary conversions). Primary conversion was mostly due to access problems and device migration. Secondary conversions were performed for rupture in six and for a persistent endoleak, with or without aneurysmal growth, in five patients. Patients who were converted were significantly older, had a lower body weight, and had a higher prevalence of chronic obstructive pulmonary disease. Conversion was associated with shorter, wider infrarenal necks and larger aneurysms. The conversion rate was lower when a team had performed more than 30 procedures, and in procedures performed during the last two years of the study period. The conversion rate was higher with EVT or Talent devices. Patients who required primary conversion had an 18% mortality rate, compared to 2.5% mortality in patients without conversion (p<0.01). Secondary conversion was associated with a perioperative mortality of 27%, and when performed for rupture 50%. Conclusion: both primary conversion and secondary conversion for rupture carry a high operative mortality. Awareness of the risk factors may reduce conversion rate as well as early and medium term mortality. Key Words: Abdominal aortic aneurysm; Endovascular; Conversion; Risk factor; Outcome. Introduction The reported incidence of complications following endovascular repair of abdominal aortic aneurysms (AAA) varies from 20 70% and appears to decrease with experience. 1 5 Many can be resolved by additional endovascular or surgical interventions, with preservation of the endograft. 6 7 Others, like severe arterial The aim of this study was to identify pre- and post-operative factors that increase the risk of open conversion and subsequent mortality. Materials and Methods injury or complete graft migration, usually require conversion to open surgery. Primary (i.e. at the time Data collection of the stent-graft placement) conversion rates of 2 20% have been reported. 4,8,9 The associated mortality rates Data from 1871 patients, treated between January 1994 vary from 0 to 17%, but are usually higher than those and July 1999 in 65 European centres, and enrolled in following uncomplicated endovascular or straightpatients, operated upon before July 1996, when the re- the EUROSTAR registry were analysed. Data from 383 forward open AAA repair. 7,8 Primary conversion is responsible for % of the overall perioperative gistry was established, were retrieved retrospectively mortality following endovascular AAA repair. 4,5,10 13 from hospital notes and outpatient records. Thereafter, pre-treatment registration by fax to the Data Registry Presented at the Thirteenth Annual Meeting of the European Center was compulsory, and provides prospective data Society for Vascular Surgery, 3 5 September on 1488 patients. Mean follow-up for the entire study Please address all correspondence to: P. Cuypers, Department of group was 6 (interquartile range 1 12) months. Surgery, PO Box 1350, 5602 ZA Eindhoven, The Netherlands. The EUROSTAR Collaborative Centres are listed in the Appendix Patient characteristics, aneurysm morphology, opto this article. erative data, postoperative outcome and follow-up /00/ $35.00/ Harcourt Publishers Ltd.

2 184 P. W. M. Cuypers et al. data were collected by means of standardised EURO- nificantly associated with conversion in the univariate STAR Case Record Forms. 4,14,15 Follow-up clinical examinations analysis were entered in a multivariate regression and imaging studies were performed at 1, model. Statistical significance was reached when the 3, 6, 12, 18 and 24 months, and annually thereafter. p-value was less than The analysis of data was Data capture was for 95 to 98% complete depending performed with SAS statistical software (version 6.12; on the item. SAS Institutes, Cary, NC, U.S.A.). Study groups Results In 38 (2%) patients a primary conversion was con- Devices and procedures sidered. These patients comprise three subgroups: (i) immediate primary conversion, which was performed The configuration and label of the devices are preduring the initial procedure (28 patients), (ii) delayed sented in Table 1. The common cause of primary primary conversion, which was postponed until a later conversion was access failure (16 patients: main comdate, but within the first postoperative month (seven ponent 14, contralateral limb 2) (Table 2). Access was patients), and (iii) abandoned repair, which relates to complicated by rupture of an iliac artery with suba failed procedure without further attempt to obtain sequent urgent open repair in four cases. Migration success because of high operative risk (three patients). occurred in 10 patients which, in all cases, involved This third group, although not actually being concomplete device dislodgment from the proximal atversion, has been combined with the other two groups. tachment site. In two patients, the decision to convert Secondary conversion defined as explantation of the was taken before device insertion was attempted. as endovascular device and conversion to open repair after the first post-operative month was performed in the intraoperative assessment of the aortic neck was 11 patients. not in accord with the preoperative findings and was considered unfavourable. Secondary conversion was performed at a mean of 8 (range 3 18) months. Four patients had a persisting Study design endoleak with aneurysm growth. Six conversions were performed because of aneurysm rupture (Table 3). The following variables were analysed: patient char- Two (patients number 1 and 2) had a type III endoleak 17 acteristics (age, gender, ASA-classification, weight, as a result of disconnection of the separate iliac limb hypertension, smoking, diabetes and pulmonary status from the main component of a modular device. One according to the SVS/ISCVS scoring system 16 ), an- patient (patient number 4) had an aortoenteric fistula eurysm morphology (angulation of the aortic neck, following device failure. 18 Two patients (patients numthe aneurysm and iliac arteries, aortic neck diameter ber 3 and 6) had a type III endoleak following perand length, maximum aneurysm diameter, common foration of the Dacron fabric of the device. 19 All but iliac artery diameter and aortic diameter at the level one of the patients had an identified endoleak prior of the bifurcation), operating team related factors (level to rupture. of experience), study period (calendar year 1994 to 1999) and type of device. Statistical analysis Data were expressed as means and standard deviation (SD). The correlation of variables with the occurrence of conversion was studied by Chi-square analysis for categorical variables. T-tests were used for continuous variables with approximate normal distribution. The Mann Whitney test was used for the correlation of other continuous variables. This was performed by categorising the variables into patient-, anatomic-, and procedure-related groups and calculating the correlations for each group. Variables that were sig- Risk factor analysis Converted patients who were significantly older had a lower body weight and were more likely to have chronic obstructive pulmonary disease (COPD) (Table 4). Conversion was related to shorter (p<0.01) and wider (p=0.02) aortic necks (Table 5) and a greater aneurysm diameter (p<0.01). Common iliac arteries diameter, angulation of the neck, the body of the aneurysm and of the iliac arteries, did not influence the incidence of conversion. There was a greater than 50% reduction in conversion

3 Which Factors Increase the Risk of Conversion to Open Surgery? 185 Table 1. Device label by configuration. Tube Aorto-uniiliac Bifurcated Total AneuRx Zenith EVT Stentor Talent Exluder Vanguard Others Total Figures indicate patient numbers. Other labels and home-made devices are combined together as others. AneuRx: manufactured by Medtronic, Zenith: manufactured by Cook, EVT: manufactured by Endovascular Technologies, Stentor: manufactured by Mintec, Talent: manufactured by World Medical, Excluder: manufactured by Gore, Vanguard: manufactured by Boston Scientific. Table 2. Cause of failure necessitating conversion to open surgery. Category of conversion Cause of failure of the endovascular procedure No. of patients Primary conversion or Access problems 16 abandoning the procedure Migration 10 Failed assembly contralateral limb 4 Device malfunction during deployment 3 Misjudgement aortic neck 2 Aortic rupture 1 Postoperative suspicion of rupture 1 Secondary conversion Rupture 6 Persistent endoleak with aneurysm growth 4 Symptomatic without rupture 1 Table 3. Conversion for rupture. Patient no. Time to rupture Cause of rupture Last diameter Death (month) measured (mm) 1 12 Graft limb disconnection Graft limb disconnection Fabric tear Aortoenteric fistula Proximal endoleak Fabric tear 50 rate when the team had performed 30 or more procedures (Table 6). Conversions were also significantly less frequent when performed during the last 2 years of the study period. There were more conversions in patients with EVT and with Talent devices but the configuration (straight, monoiliac or bifurcated) did not influence the incidence of conversion. In multivariate analysis (Table 7), body weight and chronic obstructive pulmonary disease (COPD), but not age, proved to be independent variables, as were width and shortness of the neck experience with <30 operations, procedures performed in the early period of the study, and the use of EVT and Talent devices. In a separate regression model, the use of straighttube grafts, neck diameter and length were entered together with the type of device. The EVT but not Talent graft maintained its independent correlation with conversion. Operative outcome The 30-day mortality for primary conversion during the initial procedure (28 patients) was 18% (five patients). In patients without conversion the mortality was 2.5% (p<0.01). Primary conversions during the initial procedure were also associated with major complications in 29% (five patients with multi-organ failure, two patients with cardiac complications, and one

4 186 P. W. M. Cuypers et al. Table 4. Patient characteristics by group. Conversion Without conversion n=49 n=1822 Age (years) 72.6 (7.0) 69.6 (8.3) Female gender 16% 8% (8) ASA physical status classification ASA I 7% 9% ASA II 40% 38% ASA III 44% 46% ASA IV 9% 7% Weight (kilograms) 73.2 (12.9) 79.1 (13) Hypertension 50% 58% Smoking 61% 57% Diabetes 7% 11% COPD 56% 36% Included are three patients with a failed procedure, in whom repair was abandoned because of high operative risk. Mean (SD). In 2% of patients ASA class was not available. SVS-ISCVS 16 risk score >0. Relative to patients without conversion, intergroup differences were statistically significant in: age, p=0.01; weight, p<0.01; sex, p=0.03; occurrence of COPD, p<0.01. Table 5. Anatomic features by group. Conversion Without conversion n=49 n=1822 Important angulation Neck 9 (18%) 344 (20%) Aneurysm 6 (12%) 155 (9%) Iliac 16 (35%) 568 (33%) Diameter (mm) Neck 23.4 (3.0) 22.4 (2.8) Aneurysm 60.9 (12.1) 55.9 (11.5) Aortic bifurcation 31.6 (12.8) 28.4 (11.8) Common iliac R 13.1 (5.9) 13.3 (6.5) Common iliac L 13.3 (5.8) 12.8 (5.9) Length proximal neck (mm) 22.9 (7.5) 28.1 (11.8) Figures indicate patient numbers with percentages between parentheses, unless indicated otherwise. Mean (SD). Included are three patients with a failed procedure, in whom repair was abandoned because of high operative risk. Relative to patients without conversion, intergroup differences were statistically significant in: diameter of neck, p=0.02; diameter of the aneurysm, p<0.01; and length of the neck, p<0.01. patient with respiratory insufficiency, because of pneumonia). In contrast, in primary delayed conversion (seven patients) and abandoning the procedure after primary failure (three patients) no mortality was encountered (p<0.01 compared to immediate primary conversion mortality). In the group with late conversion, no mortality was encountered in the five patients without rupture. In the six patients with rupture three died, resulting in an overall perioperative mortality for secondary conversion of 27%. Discussion Primary conversion Primary conversion occurred in 2% of the overall series (38 patients), similar to 2% reported by Stelter et al., 7 and slightly lower than the 3% and 4% reported by others. 6,10 In series that include the early experience with endovascular AAA treatment the primary conversion rate was considerably higher, ranging from 7 18%. 2,20,21

5 Which Factors Increase the Risk of Conversion to Open Surgery? 187 Table 6. Conversion rate for procedure-related variables. Total no. of procedures Conversion n=1871 n=49 Experience team Ζ10 procedures (3.6%) procedures (3.5%) >30 procedures (1.5%) Year of procedure (6.3%) (0.7%) (3.9%) (4.3%) (1.7%) (1.0%) Type of device AneuRx (1.8%) Zenith 69 1 (1.5%) EVT 55 9 (16%) Stentor (2.4%) Talent (4%) Vanguard (1.9%) Excluder 51 0 Other 23 1 (4.2%) p=0.02 p<0.01 p<0.01 Included are three patients with a failed procedure, in whom repair was abandoned because of high operative risk. Table 7. Results of three multivariate analyses relating risk factors to the probability of conversion (all conversions) for patients, anatomic and procedure related variables. Variable p-value OR 95% CI Patient related Age 0.08 Weight 0.02 COPD absent 1 COPD present Anatomic features Diameter neck 0.04 Diameter aneurysm 0.14 Length proximal neck <0.01 Procedure related Experience Ζ30 procedures 1 Experience >30 procedures Procedures performed Procedures performed Vanguard 1 AneuRx Zenith EVT < Stentor Talent < Excluder 1 Other OR: odds ratio (the OR is relative to the variable reference class, which is indicated by OR=1); 95% CI= 95% confidence interval. Continuous parameter. The low conversion rate in this registry population may reflect the large proportion of patients treated in more recent years, and the associated benefits of advanced team experience, 1 5 improvements in interventional techniques and technology refinements. 2,5 Our observation that primary conversion during the initial procedure was associated with a perioperative mortality as high as 18% is in agreement with the 11 22% reported in previous series. 10,22,23 Thus, when discussing with the patient the possibility of a con-

6 188 P. W. M. Cuypers et al. version to an open procedure, the associated risks endotension 31 and aneurysmal growth requiring should not be minimised by a comparison with the conversion. 32 risks of a primary elective surgical AAA-repair. It has been suggested that rupture of the AAA in Immediate primary conversion was the most common the presence of a stent graft may be associated with a approach. The policy of delayed primary con- lower mortality. 33 This is not the case in the present version was followed in seven patients. Earlier series, in which the perioperative mortality rate of publications by May et al. 2,23 emphasising the high secondary conversion for rupture was 50%. operative mortality of immediate primary conversion, The use of EVT-endografts had an independent as- may have influenced surgeons in their decision to opt sociation with adverse outcome. The latest modification for delayed primary conversion. The fact that none of of EVT-devices make them currently easier to the patients treated in this manner died underline the use than earlier models, and low complication rates appropriateness of their advice. As a second alternative have recently been reported by others. 34 for immediate primary conversion, a wait and see policy in the way of abandoning repair was adopted in three patients because of high operative risk, again without early mortality. In May s series renal failure as Conclusions a result from contrast load, and supracoeliac clamping was a major complication. 8 This complication was Anatomic characteristics, doctor s experience, as well encountered in the EUROSTAR patients in com- as the type of device, correlated with adverse outcome. bination with other symptoms of multi-organ failure. Identification of these factors is important to reduce There are few circumstances in which abandoning the the need for conversion to open surgery. Primary procedure is not a realistic option. Iatrogenic aortic conversion, if performed during the initial operation, perforation usually requires open surgery, although and secondary conversion for aneurysm rupture, were quick completion under balloon-blocking of the inflow associated with the highest mortalities. has been described. 24 Another indication for immediate conversion is inadvertent covering of renal artery orifices by the endograft, not correctable by traction. This complication was not encountered in this series. References 1Naslund TC, Edwards WH Jr, Neuzil DF et al. Technical complications of endovascular abdominal aortic aneurysm repair. J Vasc Surg 1997; 26: May J, White GH, Yu W et al. Endovascular grafting for abdominal aortic aneurysms: changing incidence and indication Secondary conversions for conversion to open operation. Cardiovasc Surg 1998; 6: Cuypers Ph, Nevelsteen A, Buth J et al. Complications in the The incidence of secondary conversions in this series endovascular repair of abdominal aortic aneurysms: a risk factor analysis. Eur J Vasc Endovasc Surg 1999; 18: was 0.6% (11/1871). However, the mean follow-up 4Buth J, Laheij RJF, on behalf of the EUROSTAR Collaborators. Early complications and endoleaks after enperiod in the study was only 6 months. The secondary dovascular abdominal aortic aneurysm repair: report of a conversion rate in the series of Jacobowitz et al. with multicenter study. J Vasc Surg 2000; 31: a long follow-up period, up to 40 months, was 4%. 10 5Moore WS, Kashyap VS, Vescera CL et al. Abdominal aortic aneurysm. A 6-year comparison of endovascular versus trans- In a recent EUROSTAR analysis of 1023 patients fol- abdominal repair. Ann Surg 1999; 230: lowed for more than 12 months (over 20 months 6 Holzenbein TJ, Kretschmer G, Dorffner R et al. Endovascular average) the late conversion rate was 2%. 25 At this management of endoleaks after transluminal infrarenal ab- dominal aneurysm repair. Eur J Vasc Endovasc Surg 1998; 16: moment it is difficult to make an estimate of future secondary conversion rates, as these will be influenced 7 Stelter W, Umscheid Th, Ziegler P. Three year experience with modular stent-graft devices for endovascular AAA treatment. J by both the duration of the follow-up and the im- Endovasc Surg 1997; 4: proving quality of newer endografts May J, White GH, Harris JP. Techniques for surgical conversion of aortic endoprosthesis. Eur J Vasc Endovasc Surg 1999; 18: Secondary conversions performed for actual or im pending aneurysm rupture, or for endoleak with anrepair of abdominal aortic aneurysms: where do we stand? Mayo 9Seelig MH, Oldenburg WA, Hakaim AG et al. Endovascular eurysmal growth. In three cases aneurysmal rupture Clin Proc 1999; 74: was due to fabric holes or separation of the metal stent 10 Jacobowitz GR, Lee AM, Riles TS, for the EVT Investigators. frame or limb dislocation. 18,19 Immediate and late explantation of endovascular aortic grafts: the Endovascular Technologies experience. J Vasc Surg 1999; 29: Proximal migration or endoleak, as cause of sec ondary abdominal procedures, may be due to device 11 May J, White GH, Waugh R et al. Adverse events after endo- luminal repair of abdominal aortic aneurysms: a comparison failure 27,28 or dilatation of the infrarenal neck. 29,30 during two successive periods of time. J Vasc Surg 1999; 29: Thrombosed endoleaks, although not identifiable at Lawrence-Brown M, Sieunarine K, Hartley D et al. The Perth routine CT or angiographic examination, may be as- HLB bifurcated endoluminal graft: a review of the experience sociated with increased intrasac pressure known as and intermediate results. Cardiovasc Surg 1998; 6:

7 Which Factors Increase the Risk of Conversion to Open Surgery? May J, White GH, Yu W et al. Results of endoluminal grafting Appendix are dependent on aneurysm morphology. Ann Vasc Surg 1996; 10: Harris PL, Buth J, Mialhe C et al. The need for clinical trials for endovascular abdominal aortic aneurysm stent-graft repair: The EUROSTAR Collaborative Centres are: the EUROSTAR project. J Endovasc Surg 1997; 4: (1) Amsterdam, The Netherlands University Hospital; 15 Cuypers Ph, Buth J, Harris PL et al. on behalf of the EUROS- TAR Collaborators. Realistic expectations for patients with (2) Arnhem, The Netherlands; (3) Athens, Greece Unistent-graft treatment of abdominal aortic aneurysms. Report of versity Hospital; (4) Barcelona, Spain University Hosa multicenter registry. Eur J Vasc Endovasc Surg 1999; 17: Rutherford RB, Flanigan DP, Gupta SK et al. Suggested pital; (5) Bonheiden, Belgium Imelda Hospital; (6) standards for reports dealing with lower limb ischemia. J Vasc Bournemouth, UK Royal Bournemouth Hospital; (7) Surg 1986; 4: White GH, May J, Waugh R et al. Type III and Type IV endoleak: Bristol, UK Royal Infirmary; (8) Draguignan, France toward a complete definition of blood flow in the sac after Clinique Notre-Dame; (9) Dublin, Ireland St. James endoluminal AAA repair. J Endovasc Surg 1998; 5: Norgren L, Jernby B, Engellau L. Aortoenteric fistula caused Hospital; (10) Düsseldorf, Germany Augusta Kranby a ruptured stent-graft: a case report. J Endovasc Surg 1998; 5: kenhaus; (11) Eindhoven, The Netherlands Catharina Breek JC, Hamming JF, Lohle PNM et al. Spontaneous perforation of an aortic endoprosthesis. Eur J Vasc Endovasc Surg Spectrum Twente; (13) Frankfurt, Germany Städ- Hospital; (12) Enschede, The Netherlands Medisch 1999; 18: Chuter TAM, Risberg B, Hopkinson BR et al. Clinical experience tischen Kliniken Frankfurt am Main-Hoechst; (14) Freiwith a bifurcated endovascular graft for abdominal aortic an- burg, Germany Universitätsklinik; (15) Gilly, Belgium; eurysm repair. J Vasc Surg 1996; 24: Moore WS, Rutherford RB. Transfemoral endovascular repair (16) Glasgow, UK Gartnavel Hospital; (17) Groningen of abdominal aortic aneurysms: results of the North-American The Netherlands University Hospital; (18) Groningen, EVT phase I trial. J Vasc Surg 1996; 23: Chuter TAM, Wendt G, Hopkinson BR et al. Transfemoral The Netherlands Martini Hospital; (19) Hamburg, Gerinsertion of a bifurcated endovascular graft for aortic aneurysm many Algemeines Krankenhaus Altona; (20) Hanrepair: the first 20 patients. Cardiovasc Surg 1995; 3: May J, White GH, Yu W et al. Conversion from endoluminal nover, Germany Henriettenstiftung; (21) Hull, UK to open repair of abdominal aortic aneurysms: a hazardous Royal Infirmary; (22) Leuven, Belgium University procedure. Eur J Vasc Endovasc Surg 1997; 14: May J, Schulze K Successful completion of endoluminal repair Hospital; (23) Lille, France Hôpital Cardiologue; (24) of an abdominal aortic aneurysm after intraoperative iatrogenic Liverpool, UK Royal Liverpool and University Hosrupture of the aneurysm. J Vasc Surg 1999; 30: Lahey RJF, Buth J, on behalf of the EUROSTAR Collaborators. Immediate and longterm results of endovascular AAA Poland L Académie de Médecine; (27) Lund, Sweden pital; (25) London, UK St. Mary s Hospital; (26) Lublin, repair in the EUROSTAR series. The need for secondary interventions. In: Branchereau A, Jacobs M, eds. Surgical and endovascular treatment of aortic aneurysms. New York: Futura pitalier; (29) Lyon, France Clinique Jeanne d Arc; (30) University Hospital; (28) Luxembourg Centre Hos- Publishing Company, 2000: Umscheid Th, Stelter WJ. Time-related alterations in shape, Maastricht, The Netherlands University Hospital; (31) position and structure of selfexpanding, modular aortic stent- Madrid, Spain Hospital de la Princesa; (32) Mangrafts. A 4-year single-center follow-up. J Endovasc Surg 1999; 6: chester, UK Withington Hospital; (33) Modena, Italy 27 Lambert AW, Williams DJ, Budd JS et al. Experimental assess- Ospedale e Agostin; (34) Munich, Germany University ment of proximal stent-graft (intervascular) fixation in human cadaveric infrarenal aortas. Eur J Vasc Endovasc Surg 1999; 17: Rechts der Isar; (35) Nancy, France Policlinique d Es sey; (36) Newcastle-upon-Tyne, UK Freeman Hospital; 28 Schurink GWH, Aarts NJM, van Baalen JM et al. Stent attachment site-related endoleakage after stent-graft treatment: (37) Nieuwegein, The Netherlands St. Antonius Hos- an in-vitro study of the effects of graft size, stent type and pital; (38) Oslo, Norway Aker Hospital; (39) Oslo, Noratherosclerotic wall changes. J Vasc Surg 1999; 30: Matsumara JS, Chaikof EL. Continued expansion of aortic way Ulleval Hospital; (40) Paris, France Henry necks after endovascular repair of abdominal aortic aneurysms. Mondor; (41) Paris, France Hôpital Broussais; (42) Rot- J Vasc Surg 1998; 28: Wever JJ, de Nie AJ, Blankensteijn JD et al. Dilatation of the terdam, The Netherlands St. Clara Hospital; (43) Rotproximal neck of infrarenal aortic aneurysms after endovascular terdam, The Netherlands University Hospital; (44) San AAA repair. Eur J Vasc Endovasc Surg 1999; 19: Gilling-Smith G, Brennan J, Harris P et al. Endotension after Sebastian, Spain Hospital de Gipuzkoa; (45) Stockendovascular aneurysm repair. Definition, classification, and holm, Sweden Karolinska University Hospital; (46) Tel strategies for surveillance and intervention. J Endovasc Surg 1999; 6: Aviv, Israel Sourasky Medical Center; (47) Tilburg, The 32 Torsello GB, Klenck E, Kasprzak B et al. Rupture of abdominal aortic aneurysm previously treated by endovascular stent-graft. Netherlands Elisabeth Hospital; (48) Tilburg, The J Vasc Surg 1998; 28: Netherlands Tweesteden Hospital; (49) Toulouse, 33 May J, White GH, Waugh R et al. Rupture of abdominal aortic aneurysms: a concurrent comparison of outcome of those France Hôpital de Rangueil; (50) St. Truiden, Belgium occurring after endoluminal repair versus those occurring de St. Trudo Hospital; (51) Trondheim, Norway Uninovo. Eur J Vasc Endovasc Surg 1999; 18: Deaton DH, Bogey WM, Chiang K et al. Bifurcated endovascular grafting for abdominal aortic aneurysm. Ann Vasc Surg pital; (53) Utrecht, The Netherlands University versity Hospital; (52) Ulm, Germany University Hos- 1999; 13: Hospital; (54) Veldhoven, The Netherlands St. Joseph Accepted 5 June 2000 Hospital; (55) Vienna, Austria University Hospital; (56) Zürich, Switzerland.