Author + information
- Received August 27, 2014
- Revision received October 19, 2014
- Accepted October 24, 2014
- Published online March 1, 2015.
- Sung-Jin Hong, MD,
- Young-Guk Ko, MD∗ (, )
- Dong-Ho Shin, MD, MPH,
- Jung-Sun Kim, MD,
- Byeong-Keuk Kim, MD,
- Donghoon Choi, MD,
- Myeong-Ki Hong, MD and
- Yangsoo Jang, MD
- Division of Cardiology, Severance Cardiovascular Hospital, Yonsei University Health System, Seoul, Republic of Korea
- ↵∗Reprint requests and correspondence:
Dr. Young-Guk Ko, Division of Cardiology, Severance Cardiovascular Hospital, Yonsei University Health System, 50 Yonsei-ro, Seodaemun-gu, Seoul 120-752, Republic of Korea.
Objectives This study sought to compare the outcomes of spot stenting versus long stenting after intentional subintimal approach for long femoropopliteal chronic total occlusions (CTO).
Background The optimal stenting strategy following the subintimal recanalization of long femoropopliteal chronic total occlusions has not been investigated.
Methods A total of 196 limbs in 163 patients, implanted with bare nitinol stents after subintimal approach in long femoropopliteal occlusions (lesion length 25 ± 8 cm), were retrospectively analyzed. The primary patency was compared between spot stenting (n = 129) and long stenting (n = 67).
Results Baseline characteristics and immediate procedural results were similar between groups. Adjusted-primary patency (47% vs. 77%, p < 0.001) and adjusted-freedom from target lesion revascularization (52% vs. 84%, p < 0.001) at 2 years were significantly lower in the long stenting group than in the spot stenting group. The incidence of stent fracture, fracture type, and restenosis pattern did not differ between groups. Long stenting was an independent predictor of restenosis (hazard ratio [HR]: 2.0) along with other risk factors such as nonuse of clopidogrel (HR: 3.3) or cilostazol (HR: 2.2), small stent diameter (HR: 0.6), poor run-off (HR: 1.9), and post-procedural ankle-brachial index (HR: 0.1). Compared with spot stenting after adjustment using inverse probability of treatment weighting, long stenting, especially involving the P2 or P3 segment of the popliteal artery, was independently associated with 7.5-fold increases in restenosis risk (p < 0.001).
Conclusions The primary patency was significantly higher with spot stenting than with long stenting following subintimal approach for long femoropopliteal chronic total occlusions. The risk of restenosis was especially higher when long stenting was extended to the distal popliteal artery.
Recent randomized studies have revealed that stent placement is associated with improved patency and clinical improvement compared with balloon angioplasty for the endovascular treatment of superficial femoral artery (SFA) lesions of increasing lengths (1–3). However, most of these patients had stenotic lesions rather than occlusions, and long lesions were rarely included. Therefore, treatment strategies for long occlusions of the SFA have not been standardized, although these occlusions are relatively common in patients with lower extremity artery disease (4,5). Subintimal angioplasty is a widely accepted treatment approach for the recanalization of long chronic total occlusions (CTO) of the SFA with favorable immediate and late outcomes (6–10). Despite the frequent use of subintimal angioplasty for CTO lesions of the SFA, the role of stenting in the subintimal tract is unclear. Whether spot stenting, which only covers segments with flow limitations or residual stenosis >30%, or long stenting, which covers the whole subintimal tract, is superior remains unknown, and an optimal stenting strategy has yet to be determined. Therefore, the purpose of this study was to compare the outcomes of spot stenting versus long stenting after subintimal approach for long CTO of the femoropopliteal artery.
Between 2003 and 2013, a total of 196 limbs in 163 patients who underwent successful stenting after intentional subintimal approach for long femoral CTO (lesion length ≥8 cm) were retrospectively analyzed. The treated limbs were classified into 2 groups, according to the stenting strategies: 129 limbs (66%) in the spot stenting group and 67 limbs (34%) in the long stenting group.
Before the angioplasty procedure, all patients underwent physical examinations, ankle-brachial index (ABI) assessments, and imaging tests, including computed tomography (CT), magnetic resonance angiography, or color duplex ultrasound. The Institutional Review Board at the Severance Hospital of the Yonsei University Health System approved this study and waived the requirements for informed consent for this retrospective analysis.
All procedures were performed under local anesthesia supplemented with intravenous sedation and analgesia when required. Either ipsilateral or contralateral femoral puncture was performed, depending on the distance to the target lesion. A 7-F introducer sheath (Terumo, Tokyo, Japan) was used for the ipsilateral approach, whereas a contralateral sheath (6-F to 8-F, Balkin; Cook Inc., Bloomington, Indiana) was employed for the crossover approach. A 0.035-inch hydrophilic guidewire (Radifocus, Terumo) and a supporting 5-F multipurpose catheter (Torkon NB, Cook Inc.) were used to cross the totally occluded lesion. In the intentional subintimal approach, a straight hydrophilic 0.035-inch wire was introduced at the level of the proximal stump eccentrically either into the medial or lateral wall of the occluded femoral artery using a 4-F or 5-F angled catheter. The wire was advanced distally to form a loop of wire, which was then pushed into the distal lumen supported by the catheter. We considered wire passage to be subintimal when linear or spiral dissections were visible at the proximal and distal stump. The Outback LTD re-entry catheter (Cordis, Bridgewater, New Jersey) was used when the guidewire failed to enter into the true lumen. After the subintimal passage of the guidewire through the CTO, the target artery was dilated with a balloon (5 to 6 mm in diameter).
The stenting strategy was chosen by the operators’ discretion. In the spot stenting group, the entire lesion length was not covered with stents after balloon dilation. At least 1 stent was routinely deployed into the proximal stump of the subintimal tract (10), and the segments with flow-limiting dissection, significant residual stenosis (>30%), or a pressure gradient >20 mm Hg were covered with additional stents. In the long stenting group, the entire lesion length was primarily covered with overlapping stents. Self-expanding nitinol stents (S.M.A.R.T. [Cordis]; Zilver [Cook]; Absolute Pro [Abbott Vascular, Redwood City, California]; Complete SE, [Medtronic, Santa Rosa, California]; or Protégé Everflex [Covidien, Plymouth, Minnesota]) of 6 to 8 mm in diameter were deployed into the subintimal channel. Deployed stents were routinely dilated with balloons for better apposition. After stenting, we routinely administered the combination of aspirin (100 mg/day) and either clopidogrel (75 mg/day) or cilostazol (200 mg/day) for at least 1 year. Thereafter, lifelong aspirin with or without cilostazol was given.
Follow-ups and evaluations of stent-related factors for restenosis
All patients underwent noninvasive hemodynamic evaluations before discharge, including ABI measurements, segmental pressures, and pulse volume recordings. Patients were observed at 1 month after the procedure and then were physically examined at 3-month intervals. Noninvasive hemodynamic evaluations were repeated at 1-year intervals or if the symptom status deteriorated. At least 1 imaging study, such as CT angiography, duplex ultrasound, or intra-arterial angiography, was performed in the event of either a >0.15 decrement in the ABI or worsening symptoms that were reflected by changes in the Rutherford category.
To evaluate the stent-related factors for restenosis, the extent of popliteal artery coverage of the distal stents, a stent-to-artery ratio, a stent fracture, and the presence of lesion calcification were analyzed. The extent of popliteal artery coverage was classified by segments (P1, P2, and P3) according to the location of the distal stent margin (11). The femoral artery size was calculated from the average diameter of the proximal and distal stump, where each stump diameter was measured from the long- and short-axis diameter on the pre-procedural CT images. Stent fractures were defined as the clear interruption (1 to 2 mm) of stent struts with kinks or misalignments along the axial length of the stent, and they were classified as types 1 to 5 (12). For evaluation of stent fractures, biplane radiographs (anteroposterior and oblique 30° view) of the implanted stents were obtained at 12 months or at repeat interventions using fluoroscopy or plain x-ray machine. Calcified lesions were defined as obvious densities observed within the apparent vascular wall in the angiogram.
Study endpoints and definitions
Technical success was defined as recanalization of the target lesion in the absence of residual stenosis >30% or flow-limiting dissection. A major complication was defined as any event that was either fatal or required surgical management or rehospitalization within 30 days of the procedure. The primary endpoint was primary patency, which was assessed by intra-arterial angiography, CT angiography, or duplex ultrasound. Peak velocity ≥180 cm/s or a lesion/adjacent segment velocity ratio ≥2.4 by duplex was considered to indicate significant (≥50%) restenosis. Additionally, the restenotic patterns were classified as type 1 (focal restenosis as ≤50 mm in length), type 2 (diffuse restenosis as >50 mm in length), or type 3 (total occlusions) (13). The secondary endpoint was the freedom of target lesion revascularization (TLR). All TLR were performed for restenotic lesions with both worsening symptoms and a >0.15 decrement in the ABI.
Continuous data are presented as mean ± SD, and categorical data are presented as counts (percentages). Baseline clinical and lesion characteristics were compared between the 2 groups using a Student t test for continuous data or Fisher exact test for categorical data. Primary and secondary endpoints were estimated using the Kaplan-Meier survival analysis and compared by log-rank test. We performed a univariate analysis using the Cox proportional hazards regression with all variables listed in Tables 1 and 2⇓. The variables achieving p < 0.1 in the univariate analysis were entered into the multivariate analysis model. Two models of multivariate analysis were used to separately evaluate the impact of long stenting and the degree of the popliteal coverage of stenting on the risk of restenosis. In the first model, the long stenting group was compared with the spot stenting group. In the second model, the long stenting group was further classified into the long stentings limited to femoral artery, those extended to P1, and those extended to the P2 or P3 segment, respectively. All variables met the proportional hazards assumption when we evaluated them by including an interaction between covariates and the logarithm of time and by examination of log (-log [survival]) curves. Considering that some patients contributed data from >1 limb, generalized estimating equations and clustered Cox regression analysis were performed.
To reduce the impact of selection bias and potential confounding in an observational study, adjusted survival curves and weighted Cox proportional-hazards regression models were also constructed using the inverse probability of treatment weighting after stabilization and trimming (14). Weights for patients receiving long stenting were the inverse of (1 – propensity score), and weights for patients receiving spot stenting were the inverse of the propensity score. Propensity scores were estimated using multiple logistic regression analysis. A full nonparsimonious model was developed that included all variables in Tables 1 and 2, except number of stents, stented length, mean femoral artery diameter, stent-to-artery ratio, and extent of popliteal artery stent coverage. For stabilization, we multiplied the inverse probability of treatment weighting weights by the marginal prevalence of the treatment actually received. For trimming, weights were set equal to 0.10 if the stabilized weight was <0.10 and set equal to 10 if it was >10. Findings were considered significant at p < 0.05. All statistical analyses were performed with SAS (version 9.2, SAS Inc., Cary, North Carolina).
Baseline and procedural data
Baseline characteristics are summarized in Table 1. The lesion and procedural characteristics are presented in Table 2. Most of the treated lesions were TransAtlantic Inter-Societal Consensus (TASC) II C or D lesions with a mean length of 25 ± 8 cm. There were more patients with TASC II C/D lesions in the spot stenting group. However, the lesion length, the lesions involving the popliteal artery including P2 or P3 segment, and pre-procedural ABI did not differ between groups.
All of the 129 limbs in the spot stenting group achieved technical success. In 5 limbs of the long stenting group, spot stenting was initially attempted; however, remaining flow-limiting dissection after the first stenting required conversion to the long stenting strategy. The other 62 limbs treated with long stenting showed technical success. Lower numbers of stents were implanted, and the stented length was shorter in the spot stenting group. More stents with larger diameters were implanted in the spot stenting group. However, the stent-to-artery ratio was similar between groups. Although only 9 limbs (7%) in the spot stenting group had distal popliteal artery coverage, 38 limbs (56%) in the long stenting group were covered in the popliteal artery segment with stents. In particular, 15 limbs (22%) of the long stenting group were distally covered up to the P2 or P3 segment with stents.
The immediate procedural results are shown in Table 3. Post-procedural ABI were similarly increased in both groups (p = 0.262). There were no procedure-related deaths. Distal embolization to the distal SFA or arterial perforation was observed at similar frequencies in both groups.
Primary patency and clinical outcomes, according to stenting strategies
During the median follow-up period of 1.7 years, loss of patency was found in 67 limbs. In the spot stenting group, 37 limbs (29%) showed restenosis: type 1 in 4 limbs (11%); type 2 in 10 limbs (27%); and type 3 in 23 limbs (62%) (Figure 1). In the long stenting group, 30 limbs (45%) developed restenosis: type 1 in 2 limbs (7%); type 2 in 9 limbs (30%); and type 3 in 19 limbs (63%) (Figure 1). The frequencies of different restenotic patterns were similar between groups (p = 0.830). Fluoroscopic or x-ray images for the evaluation of stent fracture at the 12-month follow-up were available for 76 limbs (39%). There were no significant differences in the incidence and type of stent fractures between groups (Table 3).
The Kaplan-Meier survival curves and adjusted survival curves demonstrated that the spot stenting group had significantly higher primary patency rates than the long stenting group did (Figures 2A and 2D). Adjusted-patency rates at 1 and 2 years were 87% and 77% for the spot stenting group and 56% and 47% for the long stenting group, respectively.
Of the 67 limbs that developed restenosis during the follow-up, 39 and 8 limbs were treated with repeat percutaneous revascularization and bypass surgery, respectively. Twenty limbs were treated medically because of patient refusal (n = 6), poor medical conditions (n = 5), or mild symptoms (n = 9). Percutaneous revascularization was successful in 37 of 39 limbs (95%). The revascularization attempt failed in 2 limbs, because the wire could not be passed through the totally reoccluded lesions. The Kaplan-Meier survival curves and adjusted survival curves showed that the spot stenting group showed significantly higher TLR-free survival rates at 1 and 2 years than the long stenting group did (91% and 84% vs. 61% and 52% after adjustment, p < 0.001) (Figures 2B and 2E).
Multivariate analysis of the predictors of restenosis
The body mass index, critical limb ischemia, nonuse of clopidogrel, distal run-off vessels ≤1, small stent diameter, lower post-procedural ABI, stent coverage of the popliteal artery extending to the P2 or P3 segment, and long stenting were found to be associated with an increased risk of restenosis as determined via univariate analysis (Figure 3). On the multivariate analysis, long stenting (hazard ratio [HR]: 1.97) was identified as an independent predictor of restenosis, along with the nonuse of clopidogrel (HR: 3.25) or nonuse of cilostazol (HR: 2.23), small stent diameter (HR: 0.64), distal run-off vessels ≤1 (HR: 1.85), and post-procedural ABI (HR: 0.06) (Table 4). In the second model, long stenting extending to the P2 or P3 popliteal artery segments, compared with spot stenting, was associated with 3.37-fold (p = 0.008) increases in the risk of restenosis. The Kaplan-Meier survival curve also showed a graded relationship between the primary patency and the extent of popliteal artery coverage (p < 0.001) (Figures 2C and 2F). When adjusted by inverse probability of treatment weighting, long stenting extending to the P2 or P3 popliteal artery segments, compared with spot stenting, was associated with 7.51-fold increases in the risk of restenosis (Table 4). Findings of clustered Cox regression were consistent (Online Table 1).
The principal findings of the present study are that the primary patency and the freedom of TLR were significantly lower with long stenting than with spot stenting following intentional subintimal approach for long femoropopliteal CTO. The risk of restenosis was significantly higher when long stenting was extended to cover the P2 or P3 segment of the popliteal artery.
Long stenting for femoropopliteal occlusions
Recent randomized trials have mostly focused on short-to-medium lesions, and results from these trials have demonstrated higher patency in primary self-expanding nitinol stenting than in provisional stenting, particularly for increasing lesions of femoropopliteal arteries (1–3). However, the majority of lesions in clinical practice are considerably longer, and whether the benefit of primary stenting could be extended to longer CTO with TASC II C/D lesions has not been proven. In the present study, which reflects clinical practice, we compared spot stenting with long stenting in long femoropopliteal CTO with mean lesion lengths of 25 ± 8 cm and 94% TASC II C/D lesions. Although there are limited data on the outcomes of stenting after subintimal angioplasty, especially for femoropopliteal CTO, registry data show that longer stenting is associated with poor outcomes. Boufi et al. (6) reported that the implantation of stents covering the whole subintimal tract does not benefit primary patency. Siablis et al. (9) found a >6-fold increase in restenosis with full-lesion stenting. Moreover, Treiman et al. (15) reported that the primary patency was 85% at 1 year but only 18% at 3 years when they routinely stented the entire length of the dissection after subintimal angioplasty. Here, we found that the primary patency and the freedom of TLR were significantly lower with long stenting than with spot stenting after subintimal angioplasty. Adjusted-patency rate was 47%, and the TLR-free survival rate was 52% at 2 years for the long stenting group, which is similar to the findings of a previous report (15).
Implication of long stenting as a risk factor of restenosis
The nonuse of cilostazol or clopidogrel, poor distal run-off vessels, and lower post-procedural ABI were found to be independent predictors for restenosis, and these results are similar to previous reports (9,10,16). Larger stent diameter may be more advantageous to obtain larger lumen of the subintimal tract and to maintain the femoral patency after intentional subintimal approach.
In the present study, long stenting was identified as an independent predictor of restenosis. Theoretically, longer lesion length, increased fracture rate, small artery or stent diameter, or the stent-to-artery diameter mismatch may lead to a higher restenosis rate in the long stenting group than in the spot stenting group. However, we found no significant differences regarding these variables. Whether the stent fractures are generally associated with poor outcomes is still controversial (12,17). We did not routinely investigate for stent fractures. Therefore, the stent fracture rates in our study may have been inaccurate. Additionally, the small sample size may have led to insignificant difference in stent fracture rates between groups. Interestingly, we found that the longer extent of stent coverage into the popliteal artery, especially those that involved the P2 or P3 segment, was independently associated with a 3.37-fold increased restenosis rate. This is compatible with the previous observation that popliteal stenting in TASC II D lesions was associated with a 4.28-fold increased risk of restenosis (18). A randomized controlled trial demonstrated superior outcomes of primary stenting for the treatment of isolated popliteal artery lesions compared with provisional stenting (19). However, this study included only isolated popliteal artery disease and the mean target lesion length was 42 mm. Recent studies suggested that longer stents may cause more kinking in the adjacent vessel segments and also more increased wall shear stress due to the axial and radial rigidity of the stented segment (20,21). These biomechanical changes by long stents may be also responsible for the increased restenosis risk.
In order to avoid problems related to a metal scaffold left behind in the popliteal artery segment, a drug-coated balloon with or without atherectomy may be a more favorable treatment option for the distal segment of the femoropopliteal artery. Also, drug-coated balloons combined with spot stenting may be advantageous for the treatment of long femoropopliteal CTO after subintimal recanalization (22). The placement of a new self-expanding interwoven nitinol stent (SUPERA, IDEV Technologies, Webster, Texas) may be another favorable option for the treatment of popliteal lesions (11).
First, this was a retrospective study of single-center registry data from a relatively small group of patients. Second, noninvasive testing during the follow-ups was less rigorously carried out than would be achieved in a prospective study. Third, there was no pre-defined limit of stented length for spot stenting. Fourth, we did not perform intravascular ultrasound during the procedure to prove the subintimal passage of the wires. Therefore, we use the term “intentional subintimal approach” rather than “subintimal angioplasty.” However, the subintimal approach is a commonly performed procedure for long CTO of the femoropopliteal artery. Only a small number of operators use intravascular ultrasound in daily practice.
The primary patency was significantly higher with spot stenting than with long stenting following intentional subintimal approach for long femoropopliteal CTO. The risk of restenosis was especially higher when long stenting was extended to the distal popliteal artery.
This study was supported by the Healthcare Technology R&D Project, Ministry for Health, Welfare and Family Affairs, Republic of Korea (no. A085012, A102064, A120478, and HI08C2149), the Korea Health 21 R&D Project, Ministry of Health and Welfare, Republic of Korea (no. A085136), and the Cardiovascular Research Center, Seoul, Republic of Korea. The authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- Abbreviations and Acronyms
- ankle-brachial index
- computed tomography
- chronic total occlusion
- hazard ratio
- superficial femoral artery
- TransAtlantic Inter-Societal Consensus
- target lesion revascularization
- Received August 27, 2014.
- Revision received October 19, 2014.
- Accepted October 24, 2014.
- American College of Cardiology Foundation
- Krankenberg H.,
- Schluter M.,
- Steinkamp H.J.,
- et al.
- Schillinger M.,
- Sabeti S.,
- Dick P.,
- et al.
- Laird J.R.,
- Katzen B.T.,
- Scheinert D.,
- et al.
- Rooke T.W.,
- Hirsch A.T.,
- Misra S.,
- et al.,
- for the Society for Cardiovascular Angiography and Interventions; Society of Interventional Radiology, Society for Vascular Medicine, and Society for Vascular Surgery
- Tendera M.,
- Aboyans V.,
- Bartelink M.L.,
- et al.,
- for the European Stroke Organization
- Hong S.J.,
- Ko Y.G.,
- Kim J.S.,
- Hong M.K.,
- Jang Y.,
- Choi D.
- Scheinert D.,
- Werner M.,
- Scheinert S.,
- et al.
- Iida O.,
- Nanto S.,
- Uematsu M.,
- Ikeoka K.,
- Okamoto S.,
- Nagata S.
- Tosaka A.,
- Soga Y.,
- Iida O.,
- et al.
- Iida O.,
- Soga Y.,
- Hirano K.,
- et al.
- Baril D.T.,
- Chaer R.A.,
- Rhee R.Y.,
- Makaroun M.S.,
- Marone L.K.
- Rastan A.,
- Krankenberg H.,
- Baumgartner I.,
- et al.
- Zeller T.,
- Rastan A.,
- Macharzina R.,
- et al.