Subintimal angioplasty as a treatment of femoropopliteal artery occlusions

Eur J Vasc Endovasc Surg 25, 578±582 (2003) doi:10.1053/ejvs.2002.1899, available online at http://www.sciencedirect.com on

Subintimal Angioplasty as a Treatment of Femoropopliteal Artery Occlusions E. Laxdal1, G. L. Jenssen2, G. Pedersen1 and S. Aune1 Departments of 1Vascular Surgery and 2Radiology, Haukeland University Hospital, Bergen, Norway Objectives: to report the results of subintimal PTA of femoropopliteal occlusions above the knee. Design: a retrospective study. Patients: in the period from January 1997 to January 2002, 109 patients were submitted to 124 interventions. The indication for treatment was intermittent claudication in 78 cases and critical ischaemia in 46. Methods: all cases of subintimal angioplasty were prospectively registered. A review of all cases treated with subintimal PTA for above-knee femoropopliteal occlusions were done. Primary assisted haemodynamic patency rate was calculated on intention to treat basis and for successfully treated cases. Comparison of patency with respect to comorbidities, indication, runoff and occlusion length was done with univariate and multivariate analysis (Cox' regression). Results: technical success rate was 90%. Primary assisted patency rates at 6, 12 and 18 months were 43, 37 and 31% calculated on basis of intention to treat and 48, 42 and 35% for successfully treated cases. Diabetes mellitus and critical ischaemia were found to be independent risk factors for re-occlusion. Conclusion: subintimal angioplasty is an alternative to open surgery for patients with femoropopliteal occlusions and intermittent claudication. The treatment is relatively atraumatic, complications are rare and in most cases treated with endovascular techniques. Patency rates are low. In cases of critical ischaemia, time can be important for outcome with respect to limb salvage. We therefore find that the poor patency rates of subintimal angioplasty of femoropopliteal occlusions contraindicate its use in the treatment of critical ischaemia with exception of cases unsuitable for surgical treatment. Key Words: Subintimal angioplasty; Femoropopliteal occlusions.

Introduction The technique of intentional subintimal recanalisation of femoropopliteal occlusions was first described by Bolia and his associates in 1990.1 Their publication of long-term results in 1994, suggested that this new approach was a useful alternative to bypass surgery.2 Reports from other centres have been scarce. Only one centre has published results as promising as those of Bolia and his group.3 Subintimal angioplasty has been used as a treatment of infrainguinal arterial occlusions at our institution since 1997. The aim of this study is to report our results of this treatment of above-knee femoropopliteal occlusions. Patients and Methods From January 1997 to January 2002, 109 patients underwent a total of 124 subintimal recanalisations  Please address all correspondence to: Dr E. Laxdal, Department of surgery, Haukeland University Hospital, 5052 Bergen, Norway.

of above-knee femoropopliteal occlusions. There were 69 men and 40 women (37%) aged 35±92 years (mean 72 years). The indication was disabling intermittent claudication (IC) in 81 cases (65%) and critical ischaemia (CI) as defined by the second European Consensus Document4 in 43 (35%). Other risk factors are listed in Table 1. The mean occlusion length was 13 cm (range 3±35 cm). The quality of the run off was defined on basis of the number of patent crural vessels, as good (41) in 80 limbs (65%) and poor (0±1) in 41 limbs (33%). It was not possible to retrieve Table 1. Associated conditions and comorbidity in 109 patients treated with subintimal angioplasty of femoropopliteal occlusions. The information was incomplete in some cases. Condition

Proportion

%

Heart disease Hypertension Diabetes mellitus Stroke COPD Smoker Creatinine 4 125 mmol/l

50/108 47/108 19/107 13/108 12/108 55/100 16/107

46 43 18 12 11 55 15

1078±5884/03/060578 ‡ 05 $35.00/0 # 2003 Elsevier Science Ltd. All rights reserved.

Subintimal Angioplasty

information on runoff in three cases (2%). Recanalisation was performed as described by Bolia1,5,6 and Reekers7 by four different interventional radiologists. The common femoral artery was punctured in an antegrade fashion on the ipsilateral side. Subintimal entry was achieved by pushing a 5 French straight angiography catheter into the lesion. The occluded segment was then traversed with a 0.03500 hydrophilic guidewire (Terumo). The tip of the guidewire was looped during the passage and followed by a 5 French straight catheter. After re-entry at the distal end of the lesion, the subintimal passage was dilated with either a 5 or 6 mm balloon angioplasty catheter (Schneider/Boston Scientific), depending on the diameter of the adjacent normal artery. Technical success was defined as an open canal with good antegrade flow at the completion of the procedure. The flow was not measured but visually evaluated by the radiologist. Procedures in which it was impossible to gain entry, re-entry or obtain an acceptable flow through the canal were regarded as technical failures. During the procedure the patient was given 5000 IU of heparin intravenously and, if tolerated, 160 mg aspirin daily thereafter. Follow-up was done at 1, 3, 6, 9, 12 and 18 months, with ankle/arm index measurement and duplex ultrasound of the reconstruction. Mean follow-up time was 7 months (median 3 months). Patients were submitted to follow-up until they reached an endpoint defined as re-occlusion or amputation. If a significant (450%) stenosis was found by duplex/ultrasound, the patient was referred back for angiography and PTA. If a 450% stenosis could not be confirmed by angiography, no further intervention was done and the patient continued routine follow up. If PTA was performed on an open reconstruction, the patency was regarded as primary assisted. An occluded reconstruction treated with a new subintimal PTA was regarded as a new and independent procedure. Operative mortality was defined as death within 30 days. Arterial perforations during the procedure were not regarded as complications as they were never a cause of haematoma, bleeding or worsening in the condition of the treated limb. All data were prospectively entered into a computerised registry and analysed with SPSS 10.0.7 for Windows. Patency rates were calculated with the product limit method and illustrated as Kaplan± Meier curves. Comparison of patency rates was done with the log rank test. Multivariate analyses were performed with Cox proportional hazard model. Means of continuous data were compared with Student's t-test. Statistical significance was accepted at p 5 0.05. All patencies are haemodynamic, primary assisted and based on intention to treat unless

579

specified otherwise. Patency rates and limb salvage rates are based on the number of treated legs. For cases not reaching a defined end point (occlusion or amputation), data were censored to last known follow-up.

Results Ninety percent of the procedures were considered technically successful. The operative mortality rate was 1.8%. Nine procedures were complicated by distal embolisation treated successfully with aspiration in six cases and thrombolysis in three cases. Distalisation of the arterial lesion past the knee joint was found in one case only. Primary assisted patency rates calculated on intention to treat basis was 43, 37 and 31% at 6, 12 and 18 months (Fig. 1). Primary assisted patency of successfully treated cases was 48, 42 and 35%, respectively. Critical ischaemia was an independent risk factor for re-occlusion both when calculated on basis of intention to treat (Fig. 2) and for successfully treated cases. Diabetes mellitus was also a independent risk factor for re-occlusion, but only for successfully treated cases (Table 2). Runoff was found to be a significant, but dependent risk factor for re-occlusion. The length of the recanalised segment was not found to have a significant effect on patency. The difference of patency rates when comparing the patency rates for the procedures on annual basis was not significant. Endovascular re-interventions because of significant restenosis were done once in 9 cases and twice in one case. Twelve patients required amputation, all of whom were initially treated for gangrene. Four of these had undergone a technically unsuccessful procedure and 10.0

45

0.9

30

18

0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 00.0 0

6

12

18

Fig. 1. Cumulated primary assisted patency of 124 above-knee femoropopliteal subintimal recanalisations calculated on basis of intention to treat. The numbers above the curve indicate patients at risk. Eur J Vasc Endovasc Surg Vol 25, June 2003

E. Laxdal et al.

580

10.0 0.9

I.C.: C.I.:

0.8

38 24

28 6

16 2

0.7 0.6 0.5 0.4 0.3 0.2 0.1 00.0

0

6

12

Months

18

p < 0.05

Fig. 2. Cumulated primary assisted patency of 124 above-knee femoropopliteal subintimal recanalisations: comparison of patency rates of treatment for intermittent claudication (unbroken line) and critical ischaemia (broken line) based on intention to treat. The numbers above the curve indicate patients at risk.

Table 2. The influence of risk factors on patency. Risk factor

Significance Significance intention to treat successfully treated

Heart disease Hypertension Diabetes Stroke COPD Smoking Renal insufficiency Indication IC/CI (63/37%)

NS NS NS NS NS NS NS p 5 0.05

NS NS p 5 0.05 NS NS NS NS p 5 0.05

were not candidates for further intervention for reasons unrelated to the procedure. In the remaining eight cases, the recannalisations were re-occluded at the time of amputation. The limb salvage rate of patients with CI was 70% at 3 months and 50% at 6 months. Six of the cases treated for CI that reoccluded were submitted to surgery: bypass in six cases and remote TEA in one. All operations were successful and still patent at last follow-up. Discussion The surgical as well as endovascular treatment of long (43 cm) femoropopliteal occlusions is controversial because of unsatisfactory long-term patency rates and complication profiles. The promising results of subintimal femoropopliteal recanalisations published by London et al. indicated a new approach to this Eur J Vasc Endovasc Surg Vol 25, June 2003

problem that seemed worth considering.2 Twelvemonth patency rates of 71% gave a reason to expect an adoption of the procedure at centres of high endovascular expertise. According to Bolia, the technique does not require extensive experience by the operator and should therefore have a potential for wide application.8 Today, 10 years later, only four papers have been published on the matter, reporting the treatment of 309 cases. This small number of reports does not necessarily signify that other vascular and endovascular centres have not attempted to adopt the method. Technical difficulties and disappointing results are possible reasons for abandoning the method at an early stage without reporting the experience. The technique as described by Bolia et al. and Reekers et al. implies traversing the subintimal space past the occlusion and squeezing its contents to the arterial wall. Thereby a twisted canal is created past the occlusion with an eccentric, oval lumen, resembling a ribbon spiral. On a single plane angiogram the ``spiral twists'' may resemble significant stenoses without necessarily being so. The antegrade flow rate through the canal, estimated visually by the interventionist in charge of the procedure, is therefore used as the main indicator of technical success.1,6,7 Elimination of all residual stenoses 430% is an additional criteria of technical success introduced in the reports of London et al.2 and Reekers et al.3 Their methodological description does not refer to whether they did a completion angiography in one or more planes. In our material, evaluation of technical success was primarily based on flow estimation as described by Bolia and Reekers and we did single plane completion angiography only. The irregular configuration of the subintimal canal may render it difficult to assess with duplex ultrasound as well. Significant stenoses (defined as peak systolic velocity increase of more than 150% of baseline velocity proximally to the entry zone) found on duplex ultrasound were frequently seen to be insignificant on single plane angiography and therefore not submitted to PTA as planned. Perhaps a two-plane angiography is a more precise method for estimation of such stenoses and we therefore intend to apply this as a routine in such cases in the future. The selection of patients suitable for this treatment is not yet clearly defined. Reekers et al. found the extent of calcification to have a significant predictive value for technical failure. Bolia et al. stated that in addition to extensive calcification, recent occlusions (3±6 months old) and arteries with extensive, distal atherosclerotic disease should not be treated with subintimal recanalisation.8 We adhered to these principles

Subintimal Angioplasty

in main, thus obtaining a high initial success rate. But these selection criteria do not account for the tendency to later re-occlusions. The high initial success rate of 90% may indicate that there were procedures accepted as satisfactory that should have been regarded as failures. By creating an intimal and subintimal injury one may expect a situation with increased thrombogenicity within the subintimal canal. We do not know the duration of this state nor the necessity for adjunctive anti-thrombotic therapy. London et al. administered 5000 IU of heparin during the procedure, but had otherwise no defined policy for post-procedural administration of aspirin. They found aspirin administration as a significant, but dependent factor for patency in their patients.2 Reekers et al. administrated i.v. heparin for 24 h after the procedure, followed by 80-mg aspirin twice daily for an undefined length of time.3 Bolia and Bell recommend anticoagulation with i.v. heparin 6000 IU four times during the first 24 h after the procedure followed by a prescription of 150- or 300-mg aspirin daily for three months. Furthermore they administer Tholazoline and Nitroglycerine during the procedure and local application of Nitroglycerine after the procedure to avoid distal spasm and thereby increase flow through the reconstruction.8 McCarthy et al.9 do not mention the antithrombotic therapy in their group of patients. In previous reports, symptomatic patency as well as haemodynamic patency has been used to evaluate the method, the symptomatic patencies being somewhat higher than the haemodynamic. A relief from symptoms in spite of re-occlusion might be caused by other factors unrelated to the procedure itself, i.e. placebo effect or training. We have also seen cases that had little relief of or even worsening of symptoms in spite of a patent reconstruction because of progression of atherosclerosis distally to the popliteal artery. We find that this new method should be evaluated on the basis of haemodynamic patencies only, preferentially con-

581

firmed by duplex, to avoid confusion of confounding factors. The primary assisted patency rates after 6 and 12 months were 43 and 37%, which is considerably inferior to results reported by other centres (Table 3). Our dissatisfaction with these results is reflected in the decline of interventions during the last two years of the study. The 6-month patency rates for patients with IC was 54% and CI 24%. This difference was statistically significant. In our study the proportion of patients with CI was relatively high compared to the proportion reported by London et al. (Table 3) that may contribute to the high re-occlusion rates in our material. In our series the majority of re-occlusions occurred within the first 6 months after the procedure. This explains the difference in the mean (7 months) and the median (3 months) follow-up time and a mean follow-up time of only 7 months. This implies that the early controls are the most important in order to detect factors of significance for the patency and prevention of re-occlusion. Risk factors for re-occlusion vary in the different studies. Apart from CI, we found that diabetes mellitus was as independent risk factor for re-occlusion in successfully treated cases. Runoff was a significant but dependent risk factor for reocclusion. London et al. found that in addition to runoff, smoking and occlusion length were independent risk factors for re-occlusion. In our series, neither occlusion length nor smoking affected patency significantly. Angiographic results were the only factor of significance for patency rates in the group of patients Reekers et al. investigated. McCarthy et al. reported the runoff status as the only independent factor affecting patency rates. We conclude that subintimal angioplasty of femoro-popliteal occlusion is an alternative to surgical intervention in cases of disabling IC. The treatment is less traumatic than surgery. Complications are rare, not fatal, and may in most cases be treated with

Table 3. Comparison of patency calculated on basis of intention to treat. Author

n

Initial success (%)

Patency 6 months (%)

Patency 12 months (%)

Indication CI/IC

Significant factors for patency (p 5 0.05)

London et al.

200

80

64

57

178/22

Reekers et al. McCarthy et al. Haukeland

40 69 124

85 74 90

43

60 51 37

29/11 26/43 78/46

Smoking y Length y Runoff y Angio graphic result z Runoff  Indication y Diabetes y

 Pearsons w2 test. y Multivariate analysis. z Method of calculation not described in the report. Eur J Vasc Endovasc Surg Vol 25, June 2003

E. Laxdal et al.

582

endovascular techniques. Patency rates are low but may possibly be improved by a better patient selection, clearer differentiation between technical successes and failures, and more aggressive antithrombotic therapy prior to as well as after the procedure. In cases of CI, the patency rates are very low. Time can be important for outcome with respect to limb salvage in this group of patients. We therefore find that subintimal angioplasty of femoropopliteal occlusions cannot be recommended in the treatment of CI with exception of cases that are unfit for surgical treatment. Further research of factors affecting and defining technical success vs failures, the haemodynamic properties of the subintimal canal, criteria for secondary interventions and the significance of antithrombotic therapy are needed.

Acknowledgement The authors are grateful to Stein Atle Lie, PhD, Section for Medical Statistics, Institute of Surgical Sciences, University of Bergen for his advice in appliance of statistical methods.

Eur J Vasc Endovasc Surg Vol 25, June 2003

References 1 Bolia A, Miles KA, Brennan J, Bell PRF. Percutaneous transluminal angioplasty of occlusions of the femoral and popliteal arteries by subintimal dissection. Cardiovasc Intervent Radiol 1990; 13: 357±363. 2 London NJM, Srinivasan R, Sayers RD, Naylor AR, Hartshorne T, Ratliff DA, Bell PRF, Bolia A. Subintimal angioplasty of femoropopliteal artery occlusions: The long-term results. Eur J Vasc Surg 1994; 8: 148±155. 3 Reekers JA, Kromhout JG, Jacobs MJHM. Percutaneous intentional extraluminal recanalisation of the femoropopliteal artery. J Vasc Surg 1994; 8: 723±728. 4 Second European Consensus Document on chronic critical leg ischaemia. Eur J Vasc Surg 1992; 6: 1±32. 5 Bolia A, Brennan J, Bell PR. Recanalisation of femoro-popliteal occlusions: improving success rate by subintimal recanalization (letter). Clin Radiol 1989; 40: 325. 6 Bolia A. Percutaneous intentional extraluminal (subintimal) recanlisation of crural arteries. Eur J Radiol 1998; 28: 199±204. 7 Reekers JA, Bolia A. Percutaneous intentional extraluminal (subintimal) recanalisation: how to do it yourself. Eur J Radiol 1998; 28: 192±198. 8 Bolia A, Bell PRF. Femoropopliteal and crural artery recanalisation using subintimal angioplasty. Seminars in Vascular Surgery 1995; 8: 253±264. 9 McCarthy RJ, Neary W, Roobottom C, Tottle A, Ashley S. Short-term results of femoropopliteal subintimal angioplasty. Br J Surg 2000; 87: 1361±1365. Accepted 3 February 2003