Author + information
- Received November 13, 2013
- Revision received January 3, 2014
- Accepted January 16, 2014
- Published online July 1, 2014.
- Osamu Iida, MD∗∗ (, )
- Mitsuyoshi Takahara, MD, PhD†,
- Yoshimitsu Soga, MD‡,
- Kenji Suzuki, MD§,
- Keisuke Hirano, MD‖,
- Daizo Kawasaki, MD¶,
- Yoshiaki Shintani, MD#,
- Nobuhiro Suematsu, MD∗∗,
- Terutoshi Yamaoka, MD††,
- Shinsuke Nanto, MD, PhD‡‡ and
- Masaaki Uematsu, MD, PhD∗
- ∗Cardiovascular Center, Kansai Rosai Hospital, Amagasaki, Japan
- †Department of Metabolic Medicine, Osaka University Graduate School of Medicine, Osaka, Japan
- ‡Department of Cardiology, Kokura Memorial Hospital, Kitakyushu, Japan
- §Department of Cardiology, Sendai Kosei Hospital, Sendai, Japan
- ‖Department of Cardiology, Yokohama-city Eastern Hospital, Yokohama, Japan
- ¶Cardiovascular Division, Department of Internal Medicine, Hyogo College of Medicine, Nishinomiya, Japan
- #Department of Cardiology, Shin-Koga Hospital, Kurume, Japan
- ∗∗Department of Cardiology, Fukuoka Red Cross Hospital, Fukuoka, Japan
- ††Department of Vascular Surgeon, Matsuyama Red Cross Hospital, Matsuyama, Japan
- ‡‡Department of Advanced Cardiovascular Therapeutics, Osaka University Graduate School of Medicine, Osaka, Japan
- ↵∗Reprint requests and correspondence:
Dr. Osamu Iida, Kansai Rosai Hospital, Cardiovascular Center, 3-1-69 Inabaso, Amagasaki, Hyogo 660-8511, Japan.
Objectives This study sought to investigate factors associated with restenosis after endovascular therapy comparing TASC (Trans-Atlantic Inter-Society Consensus) II classes A to C with class D femoropopliteal (FP) lesions.
Background It is unclear whether the determinants of restenosis for TASC II class D lesions are the same as those for TASC II classes A to C FP lesions.
Methods We studied 2,400 limbs from 1,889 consecutive patients (73 ± 17 years of age; 31% women; 30% critical limb ischemia) who underwent successful endovascular therapy for de novo FP lesions. Predictors for restenosis in TASC II classes A to C and class D lesions were assessed using a Cox proportional hazards model.
Results The 5-year primary patency rate was 50% in TASC II classes A to C and 34% in TASC II class D lesions, respectively (p < 0.001). Overall, restenosis had a significant interaction with sex and renal failure (both p < 0.01). Female sex was a significant risk factor for restenosis in TASC II class D lesions (adjusted hazard ratio [HR]: 1.80, p < 0.001) but not TASC II classes A to C lesions (adjusted HR: 1.10, p = 0.352). Conversely, renal insufficiency was a significant risk factor for restenosis in TASC II classes A to C lesions (adjusted HR: 1.43, p < 0.001) but not TASC II class D lesions (adjusted HR: 0.79, p = 0.129). Diabetes mellitus, no stent use, chronic total occlusion, and poor below-the-knee runoff were shared risk factors for restenosis between TASC II classes A to C and class D lesions (all p < 0.05).
Conclusions For de novo FP lesions, diabetes, no stent use, chronic total occlusion, and poor below-the-knee runoff were shared restenosis predictors for TASC II classes A to C and class D lesions, whereas renal failure was a predictor for TASC II classes A to C lesions and female sex for TASC II class D lesions.
- endovascular therapy
- femoropopliteal lesions
- nitinol stent
- Trans-Atlantic Inter-Society Consensus (TASC)
Although 50% to 70% of patients with symptomatic peripheral arterial disease present with femoropopliteal (FP) lesions (1–3), the preferred revascularization strategy, particularly endovascular therapy (EVT), remains controversial (4). Based on recently reported mid-term outcomes of nitinol stent use, the 2011 European Society of Cardiology guidelines have recommended an EVT-first strategy for TASC (Trans-Atlantic Inter-Society Consensus) II classes A to C FP lesions and surgical bypass for TASC II class D lesions, with EVT indicated under certain conditions for the latter (5).
Widespread use of EVT in more complex FP lesions has been fueled by recent advances, including the newer generation of nitinol stents with associated lower stent fracture rates and improved mid-term outcomes (6,7). Current evidence indicates that for FP lesions ≥5 cm, mid-term patency is better for nitinol stents than for balloon angioplasty (6–8), and in clinical practice, EVT is often selected for TASC II class D lesions because of a patient's poor general condition (9–11).
Despite EVT's common use, risk of restenosis in TASC II class D lesions has not been systematically studied. It is also unclear whether the risk of restenosis for TASC II class D lesions is the same as those for TASC II classes A to C lesions. This study aims to clarify whether factors associated with restenosis after EVT in TASC II class D lesions overlap with those in TASC II classes A to C lesions.
This study is a subanalysis of REAL-FP (Retrospective Multicenter Analysis for Femoropopliteal Stenting), a nonrandomized multicenter registry study approved by the institutional review boards at all 13 participating cardiovascular and vascular institutions in Japan (University Hospital Medical Information Network—Clinical Trials Registry no. UMIN000010986).
During the study period, 14,682 patients underwent EVT, of whom 4,697 had FP lesions treated with a provisional stenting strategy. The following exclusion criteria were then applied: 1) restenotic lesions (n = 1,487); 2) history of endovascular or surgical revascularization (n = 164); 3) acute onset limb ischemia (n = 128); 4) persistent sciatic artery (n = 2); 5) cystic adventitial disease (n = 2); 6) popliteal artery entrapment syndrome (n = 1); 7) femoral access closure device-related occlusion (n = 1); or 8) post-amputation (n = 1). Of the remaining 2,911 patients who underwent EVT of an FP lesion, 5.2% (152) failed because of: 1) failure of guidewire passage (n = 124); 2) failure of balloon catheter passage (n = 9); or 3) procedure discontinuation secondary to complication (n = 19). Finally, 3,471 limbs in 2,759 consecutive patients who underwent EVT for de novo atherosclerotic FP lesions were enrolled from January 1, 2004 to December 31, 2011. In all participating centers in this study, FP disease was defined as obstruction of the femoral and popliteal arteries by atherosclerosis. In the current study, we excluded cases with missing data on baseline characteristics of interest. Consequently, a total of 2,400 limbs in 1,889 patients were finally included.
The indication and strategy for the EVT procedure were decided by consensus among consulting vascular specialists, including vascular surgeons and interventional radiologists, based on findings on a computed tomography or duplex ultrasound prior to diagnostic angiography of the lower limb. The indication for revascularization included symptomatic patients, with >50% diameter FP stenosis determined by angiography and mean pressure gradient >10 mm Hg determined by a 4-F diagnostic catheter. The EVT approach and stent selection were decided by the operators based on anatomical features. The contralateral approach was the most commonly used for access. After a 6-F sheath insertion, either a 0.035- or a 0.014-inch wire was advanced across the lesion; unfractionated heparin (5,000 units) was administered. Intravascular ultrasound (IVUS) was often used, albeit not on a regular basis and at the operator's discretion, to confirm wire passage and to assess vessel diameter to decide optimal balloon and stent size. A 5-mm-diameter balloon was most frequently used at nominal pressure for 1 min in almost every case. A nitinol stent was implanted in patients presenting with a residual systolic pressure gradient >10 mm Hg, residual stenosis >30%, and/or flow-limiting dissection following balloon dilation in accordance with the latest guidelines. Self-expanding stents with diameters 1 to 2 mm larger than the reference diameter were used. When multiple stents were implanted, the overlap was 10 mm or longer. Provisional stenting was conducted with self-expanding nitinol stents chosen at the physicians' discretion: Smart (Cordis, Miami Lakes, Florida); Luminexx (BARD, Murray Hill, New Jersey); Zilver (Cook Medical, Bloomington, Indiana); or Misago (Terumo, Tokyo, Japan). Dual antiplatelet therapy (aspirin 100 mg/day with cilostazol 200 mg/day or ticlopidine 200 mg/day or clopidogrel 75 mg/day) was started at least 1 week prior to EVT and continued until the end of follow-up.
Follow-up protocol and study outcomes
The clinical evaluation, including symptom changes, ankle-brachial index , and lesion patency assessed by duplex ultrasound, was performed at baseline, 24 h, 1 month, and at every 3 months after the procedure. Reintervention was performed only when indicated clinically by symptom recurrence. The primary outcome measure was primary patency defined as peak systolic velocity ratio <2.4 assessed by duplex ultrasound (12).
Data are shown as mean ± SD for continuous variables or as percentage for dichotomous variables, unless otherwise mentioned. Between-group differences were evaluated by the unpaired t test and the Fisher exact test for continuous and dichotomous variables, respectively. Primary patency was estimated by the Kaplan-Meier method, and the difference among groups was assessed by the log-rank test. The association between baseline characteristics with the outcome was assessed by the Cox proportional regression model. Hazard ratios (HRs) are reported with their 95% confidence intervals (CIs). The influence of the TASC II classification on the association between baseline characteristics and outcome was evaluated by interaction effects. A p value less than 0.05 was considered statistically significant. Statistical analyses were performed by use of IBM SPSS Statistics (version 19, IBM, Armonk, New York).
Baseline patient characteristics in TASC II classes A to C versus class D lesions
Baseline characteristics are shown in Table 1. The mean age was 73 ± 9 years and 69% of patients were male. The prevalence of diabetes mellitus and renal insufficiency was 62% (n = 1,170) and 38% (n = 709), respectively. Thirty-five percent of limbs presented with critical limb ischemia (CLI). TASC II class D lesions were observed in 28% (n = 680) of the cases. Compared with the TASC II classes A to C group, the TASC II class D group had a higher prevalence of advanced age (>80 years) (29% vs. 23%, p = 0.025), nonambulatory status (24% vs. 16%, p < 0.001), elevated C-reactive protein levels (>3 mg/dl) (14% vs. 10%, p = 0.015), arterial calcification (65% vs. 57%, p < 0.001), chronic total occlusion (CTO) (79% vs. 34%, p < 0.001), stent use (69% vs. 93%, p < 0.001), IVUS use (29% vs. 23%, p = 0.004), aorto-iliac lesion (20% vs. 17%, p = 0.024), poor below-the-knee (BTK) runoff (49% vs. 44%, p = 0.033), and cilostazol use (56% vs. 43%, p < 0.001). Renal insufficiency and a small reference diameter (<5 mm) was more frequently observed in TASC II classes A to C lesions (p = 0.020 and 0.021, respectively).
Primary patency in TASC II classes A to C versus class D lesions
Figure 1 shows the comparison of primary patency between TASC II classes A to C lesions and class D lesions after EVT based on a provisional stenting strategy. The primary patency was 80% in the TASC II classes A to C group versus 69% in TASC II class D group at 1 year, 62% versus 48% at 3 years, and 49% versus 34% at 5 years, consistently demonstrating higher rates of primary patency in the TASC II classes A to C group up to 5 years after provisional stenting-based EVT.
Association of baseline characteristics with loss of primary patency in TASC II classes A to C versus class D
Table 2 shows the association between baseline characteristics and restenosis in TASC II classes A to C and class D. Interaction analysis demonstrated that female sex, nonambulatory status, renal insufficiency, and IVUS use had a significantly different unadjusted hazard ratio for restenosis between TASC II classes A to C and class D (all p for interaction <0.05). Figures 2 to 5⇓⇓⇓⇓ show the primary patency rate among subgroups stratified by these baseline characteristics as well as the TASC II classification. Female sex was associated with an increased restenosis risk in the TASC II class D but not the classes A to C group (Fig. 2). On the other hand, nonambulatory status, renal insufficiency, and nonuse of IVUS was associated with an increased restenosis risk in the TASC II classes A to C group, but not in the TASC II class D group (Figs. 3 to 5).
These findings indicate that sex, ambulatory status, renal insufficiency, and IVUS use had different impacts on restenosis between TASC II classes A to C and class D. We therefore performed multivariate analysis, in which the impact of these 4 variables was stratified according to TASC II classification, whereas other baseline characteristics were treated as having a common impact across TASC II classes. The results are shown in Table 3. Female sex was an independent risk factor for restenosis in TASC II class D (adjusted HR: 1.80, p < 0.001), but not in TASC II classes A to C (adjusted HR: 1.11, p = 0.352). On the other hand, renal insufficiency was an independent risk factor for restenosis in TASC II classes A to C (adjusted HR: 1.43, p < 0.001) but not in TASC II class D (adjusted HR: 0.79, p = 0.129). Diabetes mellitus, no stent use, CTO, and poor BTK runoff were common risk factors for restenosis for TASC II classes A to C and class D (all p < 0.05). The adjusted HR of IVUS use was 0.80 (p = 0.114) in TASC II class D and 0.56 (p < 0.001) in TASC II classes A to C (p for interaction overall = 0.060). Similar findings were observed even after adjustment for the use of antiplatelet drugs (Online Table 1). In this model, cilostazol use was independently associated with a lower risk of restenosis, with its adjusted HR 0.67 (p < 0.002).
Association of baseline characteristics with loss of primary patency in claudicants versus CLI patients
We additionally investigated the determinants of restenosis in claudicants versus CLI patients. As shown in Online Table 2, CTO and IVUS use had a significantly different unadjusted hazard ratio for restenosis between claudicants and CLI patients (p for interaction = 0.017 and 0.001, respectively). The subsequent multivariate analysis demonstrated that CTO was an independent risk factor for restenosis in claudicants (adjusted HR: 1.54, p > 0.001) but not in CLI patients (adjusted HR: 1.21, p = 0.174). However, the adjusted interaction effect was not statistically significant (p = 0.0134). On the other hand, although IVUS use was independently associated with a lower risk of restenosis both in claudicants (adjusted HR: 0.79, p = 0.031) and in CLI patients (adjusted HR: 0.38, p < 0.001), the adjusted HR was significantly smaller in CLI patients than in claudicants (p for interaction = 0.002).
The findings were unchanged when the difference in the prognostic impact of risk factors between TASC II classes A to C and class D and between claudicants and CLI patients was simultaneously taken into consideration (Online Tables 3 and 4).
We investigated differences in restenosis determinants between TASC II classes A to C and class D lesions. The incidence of restenosis was higher in TASC II class D than in classes A to C lesions. Common restenosis determinants for TASC II classes A to C and class D lesions included diabetes mellitus, no stent usage, CTO, and poor BTK runoff. On the other hand, female sex was a restenosis determinant in TASC II class D lesions only, and renal insufficiency was a determinant in TASC II classes A to C lesions only (p for interaction = 0.001 and 0.003, respectively). As for IVUS use, the overall interaction effect on patency did not achieve statistical significance because of a lack of statistical significance for TASC II class D lesions (p for interaction overall = 0.060).
In FP disease, better mid- and long-term outcomes secondary to device improvement contributed to the guideline update favoring increased EVT use (5). Despite being a retrospective investigation, the positive results for stent treatment are consistent with its more recent use as mainstream therapeutic strategy (Table 3). Furthermore, the supplemental analysis (Online Table 1) showed that cilostazol use was associated with a lower risk of restenosis, which was consistent with previous studies reporting a favorable effect of cilostazol on restenosis (13). Interestingly, the current study revealed that restenosis risk factors differed between TASC II classes A to C and class D lesions.
The current study revealed that female sex was a risk factor for restenosis only in TASC II class D lesions, whereas renal failure was a risk factor in TASC II classes A to C but not class D lesions. This result appears relevant to deciding treatment strategy in the clinical setting. Accumulating evidence has demonstrated that female patients have poorer outcomes in terms of cardiovascular death, myocardial infarction, stroke, patency after bypass surgery, and wound infection (14–19). Although the true mechanisms remain unknown, it might be because female patients not only have a relatively small vessel diameter but also are more likely to suffer from a proinflammatory state. Severer arterial lesions, that is, TASC II class D lesions, might be more subject to these potential risks in female patients. The treatment strategy for TASC II class D FP lesions in female patients should be reconsidered.
In contrast, renal insufficiency was only a restenosis determinant in TASC II classes A to C lesions, and not in TASC II class D lesions. Renal insufficiency has been reported to affect the vascular lesion itself, by, for instance, promoting calcification and progression to BTK (20). In this study, TASC II classes A to C lesions had better primary patency than class D lesions did. One can posit that when disease progresses to the more severe TASC II class D, renal insufficiency has no affect as a restenosis factor because overall outcomes become poor; in contrast, renal insufficiency has an impact on classes A to C lesions because of their lower restenosis rate.
Finally, the use of IVUS appears to correlate well with prognosis in at least TASC II classes A to C lesions. IVUS use has been associated with better prognosis in the coronary artery field by guiding implantation position and selection of an appropriate stent diameter to vessel diameter. IVUS findings regarding the presence or absence of an edge dissection affected long-term patency. In the coronary artery field, minimizing stent underexpansion and maximizing lesion area coverage with IVUS guidance has improved outcomes (21,22). We found in TASC II class D FP lesions, the association of IVUS use was not statistically significant (p for interaction: 0.060). Although the reason for the lack of statistical significance in TASC II class D lesions is unclear, one might posit that the longer calcified TASC II class D lesions produce a shadow that precludes deriving the full benefit from IVUS use; also, some of the cases could not be properly evaluated. On the other hand, TASC II classes A to C had better primary patency results than did TASC II class D lesions. Therefore, appropriate acute results with IVUS guidance may have more impact on loss of patency. Stent use reduced restenosis rates in TASC II classes A to C lesions, and IVUS guidance during stent implantation could further potentiate this advantage by providing an assessment of lesion length, thus ensuring complete coverage. This study suggests that using similar strategies in the FP territory to those for coronary arteries may be important, at least in TASC II classes A to C lesions.
These associations were still observed even after adjustment for antiplatelet drugs (Online Table 1), and after a simultaneous consideration of the difference in the prognostic affect between claudicants and CLI patients (Online Table 4).
Despite being large-scale, this study was not prospective. There could be bias concerning treatment strategy among institutions: particularly in those where EVT was selected or first-generation nitinol stents were used. Drug-eluting stents and drug-eluting balloons, which are regularly used in recent clinical settings, were not used. IVUS use was not regularly or consistently used to properly evaluate appropriateness of stent selection. However, there has not been a similar study thus far, and the results are relevant to the treatment strategy choices in the clinical setting.
For de novo FP lesions, diabetes, no stent use, CTO, and poor BTK runoff were shared restenosis predictors between TASC II classes A to C and class D FP lesions, whereas renal failure was a predictor in TASC II classes A to C lesions and female sex in TASC II class D lesions.
The authors would like to thank the cardiac catheterization laboratory medical staff and clinical research coordinators in the participating centers.
For supplemental tables, please see the online version of this paper.
The authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- Abbreviations and Acronyms
- below the knee
- confidence interval
- critical limb ischemia
- chronic total occlusion
- endovascular therapy
- hazard ratio
- intravascular ultrasound
- Trans-Atlantic Inter-Society Consensus
- Received November 13, 2013.
- Revision received January 3, 2014.
- Accepted January 16, 2014.
- American College of Cardiology Foundation
- Zeller T.
- Anderson J.L.,
- Halperin J.L.,
- Albert N.M.,
- et al.
- Tendera M.,
- Aboyans V.,
- Bartelink M.L.,
- et al.,
- for the ESC Committee for Practice Guidelines
- Laird J.R.,
- Katzen B.T.,
- Scheinert D.,
- et al.,
- for the RESILIENT Investigators
- Krankenberg H.,
- Schlüter M.,
- Steinkamp H.J.,
- et al.
- Davaine J.M.,
- Azéma L.,
- Guyomarch B.,
- et al.
- Iida O.,
- Yokoi H.,
- Soga Y.,
- et al.,
- for the STOP-IC Investigators
- Kullo I.J.,
- Bailey K.R.,
- Kardia S.L.,
- Mosley T.H. Jr..,
- Boerwinkle E.,
- Turner S.T.
- Selvin E.,
- Erlinger T.P.
- Zheng Z.J.,
- Rosamond W.D.,
- Chambless L.E.,
- et al.,
- for the ARIC Investigators
- Nguyen L.L.,
- Hevelone N.,
- Rogers S.O.,
- et al.
- Brosi P.,
- Baumgartner I.,
- Silvestro A.,
- et al.
- Roy P.,
- Steinberg D.H.,
- Sushinsky S.J.,
- et al.
- Fitzgerald P.J.,
- Oshima A.,
- Hayase M.,
- et al.