Patterns of in-stent restenosis after carotid artery stenting ...


Published on

  • Be the first to comment

  • Be the first to like this

No Downloads
Total views
On SlideShare
From Embeds
Number of Embeds
Embeds 0
No embeds

No notes for slide

Patterns of in-stent restenosis after carotid artery stenting ...

  1. 1. CLINICAL RESEARCH STUDIES From the Society for Vascular Surgery Patterns of in-stent restenosis after carotid artery stenting: Classification and implications for long- term outcome Brajesh K. Lal, MD,a,b,c,d Elias A. Kaperonis, MD,e Salvador Cuadra, MD,a,d Indravadan Kapadia, PA,d and Robert W. Hobson II, MD,a,b,d Newark, NJ, and Athens, Greece Objectives: Factors predicting in-stent restenosis (ISR) and future need for target lesion revascularization (TLR) after carotid artery stenting (CAS) remain undetermined. We hypothesized that the patterns of restenotic lesions may provide prognostic information. In this study, we developed an ultrasound classification scheme for ISR based on lesion length and distribution and assessed factors that may predict the need for TLR. Methods: Patients were followed up after CAS with B-mode ultrasound imaging, and ISR lesions (>40% stenosis) were classified into type I (focal <10 mm end-stent lesions), II (focal <10 mm, intrastent), III (diffuse >10 mm, intrastent), IV (diffuse >10 mm proliferative, extending outside the stent), and V (total occlusion). The frequency of lesion types was assessed. Accuracy of the ultrasound classification was confirmed with angiography. We recorded patient (age, gender, comorbidities), lesion (severity, etiology, symptomatic status) and procedural features (type, number, length of stents), and the need for TLR. Results: Eighty-five ISR lesions developed after 255 CAS procedures. Their percentage distribution was type I, 40; type II, 25.9; type III, 12.9; for type IV, 20; and type V, 1.2. Accuracy of the ultrasound classification was confirmed by angiography (r2 0.82). Inter-rater agreement for the assignment of lesion type based on ultrasound was 0.88 (very good). TLR was performed in 13 that were >80% diameter reducing. On univariate analysis, the need for TLR was highest in type IV lesions (0%, 0%, 27.3%, and 58.8% [types I to IV, respectively]; P .001). History of ISR (2.9%, 0%, 0%, and 41.2% [types I to IV]; P .003) and diabetes mellitus (20.6%, 22.7%, 45.5%, and 52.9% [types I to IV]; P .02) occurred more frequently with type IV ISR lesions. On multivariate analysis of all patient, lesion, and procedural characteristics, only the type of ISR (odds ratio, 5.1) and a history of diabetes (odds ratio, 9.7) were independent predictors of TLR. Conclusions: The proposed classification accurately grades the magnitude of intimal hyperplasia after CAS and provides important prognostic information. Diffuse proliferative (type IV) ISR lesions and diabetes are important determinants of long-term outcome after CAS. This classification will facilitate a standardized description of recurrence after CAS and enable early identification of high-risk patients for additional monitoring, treatment, and investigation. ( J Vasc Surg 2007;46:833-40.) Carotid artery stenting (CAS) has emerged as a less long-term follow-up, we have observed in-stent restenosisinvasive alternative to carotid endarterectomy (CEA) for (ISR) of 40% diameter reduction in 42.7% of our patients,revascularization of extracranial carotid occlusive disease. and of 60% diameter reduction in 16.4% at 5 years ofOur institution1-5 and others6-10 have reported that CAS follow-up.5 Similarly, the Stenting and Angioplasty withcan be performed with low periprocedural morbidity. On Protection in Patients at High Risk for Endarterectomy (SAPPHIRE) investigators reported ISR in 19.7% of pa-From the Division of Vascular Surgerya, Department of Physiologyb, and tients at 1 year of follow-up.11 Therefore, ISR will become Department of Biomedical Engineeringc, University of Medicine and increasingly prevalent due to the exponential increase in the Dentistry, New Jersey-New Jersey Medical School; the Division of Vas- cular Surgery, St. Michaels Medical Centerd; and the Division of Vascular use of carotid stents. Surgery, University of Athense. Post-CAS ISR is currently treated at a threshold ofCompetition of interest: none. 80% diameter reduction (6.4% incidence at 5 years5).Supported by grants from the American Heart Association (RA5883, BKL) However, factors that predict target vessel failure remain and the National Institutes of Health (NS38384, RWH).Presented at VASCULAR 2007, Annual Meeting of the Society for Vascular undetermined. Primary stenting prevents carotid artery Surgery, Baltimore, Md, June 7-10, 2007. recoil and constrictive remodeling.12 Post-CAS ISR canReprint requests: Brajesh K Lal, MD, UMDNJ-New Jersey Medical School, threfore be primarily attributed to neointimal hyperpla- 185 S Orange Ave, MSB-H570, Newark, NJ 07103 (e-mail: lalbk@ sia,12,13 and studies of coronary ISR indicate that long$32.00 neointimal hyperplasia lesions diffusely involving the stentCopyright © 2007 by The Society for Vascular Surgery. surface correlate with the highest recurrence and reinter-doi:10.1016/j.jvs.2007.07.022 vention rates.14 833
  2. 2. JOURNAL OF VASCULAR SURGERY834 Lal et al November 2007 The patterns of ISR developing after carotid artery to standard techniques used by our group previously4,5,16stenting have not been described, and their prognostic to estimate the degree of stenosis. Velocity criteria forutility has not been studied. We therefore developed an native carotid arteries (peak systolic velocity 130 cm/s)ultrasound classification scheme for post-CAS ISR based on were used to identify potential patients with ISR. Thesethe length and distribution of the lesion with respect to the underwent detailed B-mode imaging with power Dopplerstent and verified its accuracy with carotid angiography. We to select arteries with any ISR 40%. These lesions werethen assessed long-term clinical follow-up to determine finally included in the present analysis; therefore, onlywhether the classification system predicted the need for patients with confirmed visible restenotic lesions were in-future therapeutic reintervention in the form of target cluded in the study. This strategy ensured that we did notlesion-site revascularization (TLR). miss any patients with ISR. B-mode imaging studies during follow-up were furtherMETHODS used to define the morphology of the ISR lesions according Patient population and treatment. We performed to the length and distribution of the lesion with respect to255 CAS procedures from January 1, 1996, through De- the stent. Standard imaging techniques used previously bycember 31, 2006. Lesions were treated with a WallStent our group were applied to obtain grayscale images of the(Boston Scientific Corp, Natick, Mass) or ACCULINK ipsilateral cervical carotid artery.5 The transducer wasstent (Guidant Corp, St. Paul, Minn). Procedural details placed directly over the stented carotid artery segment tofor CAS at our institution have been published in detail by obtain a longitudinal image. It was then swept from theour group previously.1-5 base of the neck to the angle of the mandible to obtain All patients received aspirin (325 mg daily) and clopi- multiple cross-sectional images of the entire common anddogrel (75 mg twice daily) for at least 48 hours before the internal carotid arteries. The images were recorded onprocedure. Clopidogrel (75 mg daily) was continued for 30 magneto-optical (MO) disks and analyzed off-line with adays after the procedure, and aspirin was continued indefi- computer-assisted image-analysis program (Metamorphnitely. 6.1, Universal Imaging Corp, Downingtown, Pa) by inde- Patients early in our experience underwent CAS with- pendent observers blinded to clinical findings. The longi-out embolic protection. The ACCUNET (Abbot, Menlo tudinal image was used to measure the length of the lesion,Park, Calif) antiembolic device was used in all subsequent defined as the distance from the proximal shoulder to thepatients. At last follow-up, 85 arteries developed ISR of distal shoulder of the lesion, and to determine its location 40% and constitute the cohort for the current study. with respect to the stent. The cross-sectional views were Patients underwent endovascular retreatment when used to determine the luminal diameters.their ISR reached a threshold of 80%, regardless of neu- Classification of in-stent restenosis. B-mode imagesrologic symptoms. Devices used to treat ISR included in longitudinal and serial cross sections were reviewed byballoon angioplasty, cutting balloon angioplasty, and re- two independent observers who classified the lesions onstenting and were selected at the discretion of the treating separate occasions (Fig 1). Disagreements were resolved byphysician. No patient required surgical revascularization for consensus.ISR. Lesions after the first recurrence were excluded from ● Type I (focal end-stent group): Lesions are 10 mmthe primary analysis to avoid introducing any statistical bias long and are positioned at the proximal or distal mar-in the multivariate model. gin (but not both) of the stent. Lesions 10 mm long Demographics and follow-up. Clinical demograph- at both ends of the stent are defined as type I, multi-ics and laboratory results were collected in a prospective focal end-stent.registry. Risk factors that were tabulated included coronary ● Type II (focal intrastent group): Lesions are 10 mmartery disease (currently or previously symptomatic, requir- long and are confined to within the stent(s) withouting intervention), medically treated diabetes mellitus, med- extending outside the margins. Two or more discreteically treated hypertension, medically treated hypercholes- lesions 10 mm in length located within the stent areterolemia (or if serum cholesterol was 180 mg/dL), and defined as type II, multifocal intrastent.smoking (current or prior smoker). Clinical follow-up was ● Type III (diffuse intrastent group): Lesions are 10performed with office visits and duplex ultrasound exami- mm long and are confined to within the stent(s) with-nations. The occurrence of TLR events was recorded. out extending outside the margins. Ultrasound examination. Duplex ultrasound exami- ● Type IV (diffuse proliferative group): Lesions are 10nations, including Doppler velocity measurements and B- mm long and extend beyond the margin(s) of themode imaging studies, were performed in our noninvasive stent(s).vascular laboratory, which is approved by the Inter-societal ● Type V (occlusion group). Lesions have no progradeCommission on Accreditation of Vascular Laboratories,15 flow and no lumen is identified.before and after CAS within 3 days of the procedure andduring each annual follow-up visit. The studies were per- Angiographic analysis. The angiographic percentageformed with a 7-13 MHz linear array transducer (Sequoia of stenosis was measured before and after stent deployment in512, Acuson, Mountain View, Calif). Doppler velocities all patients undergoing CAS and was based on standard tech-were obtained with appropriate angle correction according niques using multiple projections. All angiograms were ana-
  3. 3. JOURNAL OF VASCULAR SURGERYVolume 46, Number 5 Lal et al 835 Fig 1. A, Schematic images show the five patterns of carotid in-stent restenosis based on the introduced classification. The shaded area represents the stent. B, Representative B-mode ultrasound images of in-stent restenosis correspond to the patterns I through IV.lyzed off-line with a computer-assisted quantitative edge- (NASCET) criteria.17 The in-stent least luminal diameterdetection algorithm (MDQM; MEDCON Telemedicine was compared with the distal nontapering portion of theTechnology, Inc, Livingston, NJ) by an independent observer internal carotid artery, which served as the reference seg-who was blinded to the ultrasound and clinical findings. In 13 ment. Lesion length was measured as the distance from thearteries, ISR 80% was found and they underwent TLR. In proximal shoulder to the distal shoulder of the lesion. The15 additional ISR lesions, confirmatory angiography was done angiographic lesion length was compared with that ob-for progressively increasing velocity measurements on ultra- tained by B-mode imaging.sound imaging and these were found to be 80% diameter- Statistical analysis. Statistical analysis was performedreducing lesions. Preprocedural B-mode imaging was com- using GraphPad Prism 3.00 software (GraphPad Softwarepared with angiography in these patients to test the accuracy Inc, San Diego, Calif) and SPSS software (SPSS Inc, Chi-of our classification scheme. cago, Ill). Categoric data are presented as percentages and Angiographic stenosis was determined using North continuous data as mean SD. Categoric data were com-American Symptomatic Carotid Endarterectomy Trial pared using the 2 test, and continuous data were com-
  4. 4. JOURNAL OF VASCULAR SURGERY836 Lal et al November 2007Table I. Baseline patient characteristics Patterns of in-stent restenosis by type I, focal end-stent II, focal intrastent III, diffuse intrastent IV, diffuse proliferativeCharacteristic* (n 34) (n 22) (n 11) (n 17)Distribution 40.0 25.9 12.9 20.0Age, y 73.9 9.4 74.6 6.7 70.5 6.7 69.2 15.9Male sex 61.8 63.6 63.6 70.6Diabetes mellitus† 20.6 22.7 45.5 52.9Hypertension 85.3 86.4 81.8 70.6Coronary artery disease 67.6 45.5 45.5 47.1Hypercholesterolemia 64.7 72.7 63.6 64.7Smoking 41.2 45.5 44.2 29.4*Categoric variables expressed as percentage; continuous variables as mean SD.† P .02.Table II. Characteristics of the original treated lesion Patterns of in-stent restenosis by type I, focal end-stent II, focal intrastent III, diffuse intrastent IV, diffuse proliferative Etiology of treated lesion* (n 34) (n 22) (n 11) (n 17)In-stent restenosis† 2.94 0 0 41.2Primary atherosclerosis‡ 38.2 72.7 36.4 23.5Post CEA-restenosis 58.8 27.3 63.6 35.3Severity of treated lesion, % stenosis 81.9 82.5 88.9 83.5Preprocedure neurologic symptoms 29.4 22.7 27.3 23.5Details of implanted stents Stents/lesion 1.1 0.4 1.1 0.3 1.2 0.4 1.3 0.5 Length, mm 28.8 10.3 31.4 8.3 30.0 6.3 28.2 11.7 Types of implanted stents§ Wallstent 50.0 36.4 27.3 47.1 Acculink 50.0 63.6 72.7 52.9Post-treatment result, % residual stenosis 6.4 7.6 7.1 6.8 6.9 7.3 7.7 8.3CEA, Carotid endarterectomy.*Values are expressed as percentages, continuous variables are expressed as mean SD† P .003.‡ P .02.§ P .03.pared using analysis of variance with the Tukey post-test. P according to our proposed scheme, 40% (n 34) were .05 was considered significant. focal end stent (type I), 25.9% (n 22) were focal intras- The primary end point of the analysis was the associa- tent (type II), 12.9% (n 11) were diffuse intrastent (typetion of lesion classification with TLR. Univariate variables III), 20% (n 17) were diffuse proliferative, and 1.2% (nwith P .2 were entered into the multivariate logistic 1) developed an occlusion. Inter-rater agreement for theregression model; forward stepping was used to determine assignment of lesion type based on ultrasound imaging wasthe independent predictors of TLR. Independent variables 0.88 (very good).were considered significant risk factors at P .05. To Baseline patient characteristics are presented in Table I.confirm the accuracy of the classification scheme, ISR le- None of the patients had neurologic symptoms in associa-sion length measurements derived from duplex ultrasound tion with the development of ISR. All groups were wellimaging were compared with angiographic measurements matched with respect to age and sex and for associatedusing linear regression. Inter-rater agreement on the ultra- comorbidities such as a history of hypertension, coronarysound classification was assessed by calculating the statis- artery disease, hypercholesterolemia, and smoking. How-tic; a score of 0.81 to 1.0 was defined as very good agree- ever, increasing levels of ISR classification were associatedment. with an increasing prevalence of diabetes mellitus (20.6%, 22.7%, 45.5%, and 52.9% for types I to IV, respectively; 2RESULTS trend, 5.4; P .02). Patient characteristics. Of the 255 CAS procedures Lesion characteristics. The severity (degree of steno-performed, ISR developed in 85 arteries during a mean sis) of the original lesion treated with CAS was comparablefollow-up of 19.3 months. Of these 85 ISR lesions classified in all ISR classes, as indicated in Table II. The proportion of
  5. 5. JOURNAL OF VASCULAR SURGERYVolume 46, Number 5 Lal et al 837Table III. Target lesion revascularization Patterns of in-stent restenosis by type I, focal end-stent II, focal intrastent III, diffuse intrastent IV, diffuse proliferativeVariables* (n 34) (n 22) (n 11) (n 17)Incidence of TLR, %* 0 0 27.3 58.8Devices used for ISR, No. Balloon angioplasty 0 0 1 3 Stent 0 0 1 5 Cutting balloon 0 0 0 1 Cutting balloon stent 0 0 1 1Post-treatment result, % residual stenosis N/A N/A 10.4 6.9 11.9 6.1TLR, Target lesion revascularization; ISR, in-stent-restenosis; N/A, not applicable.*P .001.lesions treated for neurologic symptoms, the number ofstents used per lesion, and the mean stent length used werealso similar across ISR types. Technical success wasachieved in all patients and no differences were noted in theangiographic residual stenosis after therapy across all ISRtypes. In this cohort, 37 patients had been treated with aWallStent and 48 with an Acculink stent. The incidence ofISR did not vary with the type of stent used or the type oflesion treated. Univariate analysis indicated a difference inthe patterns of ISR between stent types. Intrastent ISRpatterns (types II and III) occurred more frequently afterplacement of Acculink stents compared with WallStents(P .03). Univariate analysis also indicated that focalintrastent (type II) lesions occurred more frequently aftertreatment of primary atherosclerotic carotid stenosis com- Fig 2. Linear regression analysis shows comparison of lesion length measurement by quantitative angiography vs B-mode ultra-pared with treatment for post-CEA restenosis (P .02). sound imaging. Finally, 13 patients had recurrent ISR, of which threewere type III lesions and the remaining 10 were type IV.On univariate analysis, higher levels of ISR classification the accuracy of our proposed ultrasound classification ofwere associated with prior ISR (2.9%, 0%, 0%, and 41.2% ISR based on lesion length measurements (r2 0.82, linearfor types I to IV, respectively; 2 trend, 13.3; P .003). regression; Fig 2). Target lesion revascularization. Endovascular re- Multivariate analysis. We introduced diabetes, coro-treatment was required in three of 11 patients with type III nary artery disease, neurologic symptom status, recurrentISR, and in 10 of 17 patients with type IV ISR (Table III). ISR and primary atherosclerosis etiologies, stent type, stentThe mean interval between CAS and TLR was 18.2 number, and the pattern of ISR according to the intro-months. We observed a significant increase in TLR in duced B-mode ultrasound classification in a stepwise mul-association with increasing levels of ISR classification (0%, tiple logistic regression model to identify independent pre-0%, 27.3%, and 58.8% for types I to IV, respectively; 2 dictors of TLR. The only variables that independentlytrend, 29.4; P .001). Modalities used to treat ISR predicted TLR after CAS were a worsening pattern of ISRincluded balloon angioplasty, stenting, and cutting balloon according to our proposed ultrasound classification (oddsangioplasty alone or in conjunction with stenting. Proce- ratio, 5.1; P .003) and the presence of diabetes (oddsdural success was achieved in all these cases, without evi- ratio, 9.7; P .04). Coronary artery disease, neurologicdence of any abrupt arterial closure or neurologic events. symptoms status, recurrent ISR and primary atherosclerosisEndovascular treatment of ISR afforded similar percentage etiologies, stent type, and stent number were not signifi-diameter residual stenoses in all instances and was not cant predictors of TLR in this model.influenced by ISR class (Table III). Quantitative angiography vs B-mode imaging DISCUSSIONresults. Twenty-eight pairs of quantitative angiographic In this study we have described for the first time, to ourand B-mode ultrasound measurements were available for knowledge, the various anatomic patterns of ISR observedcomparative analysis of the geographic distribution of the after carotid stenting. We have proposed a classification ofISR lesions. The quantitative angiographic results verified these patterns that uses B-mode ultrasound imaging. The
  6. 6. JOURNAL OF VASCULAR SURGERY838 Lal et al November 2007scheme depends on the length and the geographic location ISR lesions (76.2%, type II, Table II) occurred in patientsof the intimal hyperplasia response in relation to the stent. undergoing CAS for primary atherosclerotic lesions (PThe B-mode definition of these patterns correlated with .02). Conversely, there was a trend towards fewer intrastentangiographic assessment, confirming that transcutaneous ISR lesions contributed by patients undergoing CAS forultrasound imaging allowed accurate recognition of the post-CEA restenosis. To our knowledge, this is the firstpatterns of ISR occurring after CAS. observation that ISR patterns may differ according to the The proposed classification (Fig 1) is noninvasive and etiology of the primary lesion being treated in the carotidcan be conveniently applied during each follow-up exami- artery. We have previously reported that stenting of thenation within a few minutes. It uses straight-forward B- carotid artery alters arterial wall biomechanical and hemo-mode imaging with power Doppler to outline the intimal dynamic properties.4 It is possible that stenting alters wallhyperplastic lesions, with length measurements performed properties differently in patients with calcified atheroscle-on-screen using standard software provided on most cur- rotic lesions compared with more compliant fibrotic post-rent ultrasound machines. Results can be printed and saved CEA restenotic lesions, thereby inducing differing patternsfor future comparisons during follow-up. The classification of ISR.incorporates prior observations that geographic location On univariate analysis, the type of stent used for CASand lesion length are important measures of the severity of was also observed to influence the pattern of ISR (P .03,the intimal hyperplastic response to coronary stenting.14 Table II). Intrastent patterns of ISR (both focal and diffuse,We observed that the higher level pattern (diffuse prolifer- types II and III) occurred more often in patients treatedative, type IV) of ISR after CAS predicted subsequent TLR. with the Acculink stent compared with the WallStent. WeTherefore, the intimal hyperplastic response was most se- believe this is the first observation that ISR patterns mayvere in patients who presented with the higher level of the vary according to the type of stent deployed in the carotidclassification (type IV). We infer that the classification artery. The results will require substantiation with a largeradequately captures the magnitude of the intimal hyper- number of patients. Stent biomechanical properties varyplastic response to CAS. according to their material, geometric design, and dimen- In a previous study, we reported that ISR occurs in a sions before and after deployment. These factors may in-significant proportion of patients after CAS.5 Although duce variable changes in carotid arterial wall biomechan-most lesions were moderate in severity (40% to 79% steno- ics,22 thereby influencing ISR patterns. Findings from thesis), 6.4% were hemodynamically significant ( 80% steno- current investigation may therefore impact future stentsis) and required reintervention. There is currently no way designs and material predict which of the low-grade or moderate-grade le- Despite optimal endovascular treatment with similarsions will progress to need reintervention. The present final diameter stenosis, some individuals had an enhancedinvestigation demonstrates that a classification of ISR pat- intimal hyperplastic response to carotid stenting. As stated,terns can independently predict the development of high- univariate analysis demonstrated that several factors couldgrade ISR necessitating reintervention. influence the magnitude of the intimal hyperplastic re- The only other independent predictor was diabetes, sponse and subsequent need for reintervention. These in-which is a well-known predictor of early and aggressive cluded the pattern of ISR after CAS, diabetes, prior ISR,intimal hyperplasia and ISR after coronary stenting.14,18 the etiology of the original lesion, and the type of stent usedOne report has observed an increased incidence of ISR in (Tables I and II). On multivariate analysis, however, thediabetic patients undergoing carotid stenting.19 However, pattern of ISR, reflecting the severity of the intimal hyper-the relationship between diabetes and the need for repeat plastic response to injury, independently predicted futurecarotid revascularization was not evaluated. Our study con- target vessel failure (odds ratio, 5.1; P .003). Prior ISRfirms that diabetes is associated with severe ISR (type IV, did not independently predict subsequent TLR. This ob-Table I) and that it is an independent predictor of target servation is different from findings in the coronary artery.18vessel failure and subsequent reintervention. Because prior ISR does result in more severe ISR after CAS, Prior ISR has been hypothesized to predict future we infer that the pattern of ISR, influenced by severalsevere recurrent ISR,5 but this has been refuted by oth- mechanisms, determines the risk for target vessel failure anders.20 Our study provides evidence that prior ISR predis- TLR in the carotid artery.poses to severe recurrent ISR (type IV, Table II). The The proposed classification will facilitate the identifica-reason for this exaggerated intimal hyperplastic response to tion of patients with severe patterns of ISR (type IV)a reintervention is unknown. In the laboratory, a severe noninvasively by ultrasound imaging. This will enable earlyintimal hyperplastic response can, however, be reproduc- selection of these patients for aggressive monitoring oribly stimulated in animal models by inducing vascular additional therapy in conjunction with stenting. Endovas-injury.21 The continuing presence of an intravascular stent cular treatment was successfully achieved without compli-may afford a similar injury stimulus for intimal hyperplasia cations in all instances of reintervention for high-gradein patients. ( 80%) ISR (Table III). We have previously reported that The most frequently occurring pattern of ISR in our post-CAS ISR is a neointimal hyperplastic lesion that is notcohort was the focal end-stent type (type I, Table I). On associated with the same risk of atheroembolic complica-univariate analysis, we observed that most of the intrastent tions as in primary atherosclerotic lesions.5 The multiple
  7. 7. JOURNAL OF VASCULAR SURGERYVolume 46, Number 5 Lal et al 839modalities used (plain/cutting balloon angioplasty, pri- for intensive monitoring and treatment. By providing amary stenting or a combination of these) indicate an ab- standardized method of describing restenotic lesions, it willsence of consensus on the optimal method of treatment. also facilitate further investigations into adjunctive treat-Intravascular radiation23 and local24 or systemic25 pharma- ments for ISR and improved stent design.cologic agents are under active investigation in high-riskcoronary ISR patients and may prove beneficial in their AUTHOR CONTRIBUTIONStreatment. Early and appropriate identification and treat- Conception and design: BLment of these high-risk ISR patients using our ultrasound Analysis and interpretation: BL, EK, RHclassification may improve cost-effective patient manage- Data collection: BL, EK, SC, IK, RHment and help develop novel revascularization strategies. Writing the article: BL, RH Limitations. This retrospective analysis is subject to Critical revision of the article: BL, EK, SC, IK, RHthe limitations related to such an investigation. Data were Final approval of the article: BL, EK, SC, IK, RHretrieved from our CAS Quality Assurance Database. The Statistical analysis: BLdatabase is updated prospectively and incorporates the Obtained funding: BL, RHsame information for all carotid stenosis patients undergo- Overall responsibility: BLing endovascular interventions, regardless of clinical indi-cations and outcomes. Because selective carotid angiogra- REFERENCESphy is associated with a risk of atheroembolic stroke,26 thismodality was not used for routine follow-up. Duplex ultra- 1. Hobson RW 2nd, Goldstein JE, Jamil Z, Lee BC, Padberg FT Jr, Hanna AK, et al. Carotid restenosis: operative and endovascular man-sonography is readily available at our certified vascular agement. J Vasc Surg 1999;29:228-35; discussion 235-8.laboratory. It produces high-resolution images of this rel- 2. Hobson RW 2nd, Lal BK, Chakhtoura E, Goldstein J, Haser PB,atively superficial artery, is inexpensive,12,27 and is therefore Kubicka R, et al. Carotid artery stenting: analysis of data for 105 patientsideally suited for the follow-up of patients undergoing at high risk. J Vasc Surg 2003;37:1234-9. 3. Hobson RW 2nd, Lal BK, Chakhtoura EY, Goldstein J, Kubicka R,CAS.5 Only persistently high velocities indicative of ISR Haser PB, et al. Carotid artery closure for endarterectomy does not 80%, or a recent rapid increase in velocities, were indica- influence results of angioplasty-stenting for restenosis. J Vasc Surgtions for diagnostic angiography. Although intravascular 2002;35:435-8.ultrasonography may provide accurate information on inti- 4. Lal BK, Hobson RW 2nd, Goldstein J, Chakhtoura EY, Duran WN.mal hyperplastic lesions, the risk of atheroembolic stroke Carotid artery stenting: is there a need to revise ultrasound velocity criteria? J Vasc Surg 2004;39:58-66.from additional instrumentation across the treated arterial 5. Lal BK, Hobson RW 2nd, Goldstein J, Geohagan M, Chakhtoura E,segment has resulted in limited use of this modality for Pappas PJ, et al. In-stent recurrent stenosis after carotid artery stenting:follow-up by most CAS operators. life table analysis and clinical relevance. J Vasc Surg 2003;38:1162-8; This analysis includes 14 patients that underwent CAS discussion 1169. 6. Ohki T, Veith FJ. Carotid artery stenting: utility of cerebral protectionwithout antiembolic protection early in our experience. To devices. J Invasive Cardiol 2001;13:47-55.the best of our knowledge, however, use of an antiembolic 7. Roubin GS, New G, Iyer SS, Vitek JJ, Al-Mubarak N, Liu MW, et al.device does not alter ISR and should therefore not influ- Immediate and late clinical outcomes of carotid artery stenting inence results of this study. patients with symptomatic and asymptomatic carotid artery stenosis: a Carotid occlusion is a rare outcome after CAS5 and 5-year prospective analysis. Circulation 2001;103:532-7. 8. Vitek JJ, Roubin GS, New G, Al-Mubarek N, Iyer SS. Carotid angio-occurred in only one patient from this series of 255 CAS plasty with stenting in post-carotid endarterectomy restenosis. J Inva-procedures; therefore, it is not possible to determine pre- sive Cardiol 2001;13:123-5; discussion 158-70.dictors for this outcome. Of note, this patient remained 9. Yadav JS, Wholey MH, Kuntz RE, Fayad P, Katzen BT, Mishkel GJ, etasymptomatic despite the occlusion. al. Protected carotid-artery stenting versus endarterectomy in high-risk patients. N Engl J Med 2004;351:1493-501.CONCLUSION 10. NACPTAR-Investigators. Update of the immediate angiographic re- sults and in-hospitalcentral nervous system complications of cerebral ISR after CAS presents with various ultrasound pat- percutaneous transluminal angioplasty. Circulation 1995;92:383.terns that provide important prognostic information. As 11. FDA. Circulatory system devices panel. Panel transcript. http://www-determined by the classification introduced in this study, 2004. 12. Willfort-Ehringer A, Ahmadi R, Gruber D, Gschwandtner ME,focal end-stent, focal intrastent, diffuse intrastent, diffuse Haumer A, Haumer M, et al. Arterial remodeling and hemodynamics inproliferative, and occluded ISR represent a spectrum of carotid stents: a prospective duplex ultrasound study over 2 years. J Vascincreasing intimal hyperplastic response to CAS. The pat- Surg 2004;39:728-34.tern of ISR and a history of diabetes are important deter- 13. Piamsomboon C, Roubin GS, Liu MW, Iyer SS, Mathur A, Dean LS, et al. Relationship between oversizing of self-expanding stents and late lossminants of long-term outcome after CAS. In contrast, index in carotid stenting. Cathet Cardiovasc Diagn 1998;45:139-43.other comorbidities, prior neurologic symptom status, and 14. Mehran R, Dangas G, Abizaid AS, Mintz GS, Lansky AJ, Satler LF, et al.immediate technical results appear to be less important. Angiographic patterns of in-stent restenosis: classification and implica-These observations imply that intrinsic biologic character- tions for long-term outcome. Circulation 1999;100:1872-8.istics (ISR pattern and diabetes) have more influence over 15. ICAVL. The intersocietal commission for the accreditation of vascular laboratories: standards for accreditation in noninvasive vascular testing.the long-term results of CAS than the immediate technical Columbia, MD: ICAVL; 2005.results of the treatment. The classification also offers an 16. Hobson RW 2nd, Strandness DE Jr Carotid artery stenosis: what’s inopportunity for the early identification of high-risk patients the measurement? J Vasc Surg 1993;18:1069-70.
  8. 8. JOURNAL OF VASCULAR SURGERY840 Lal et al November 200717. Beneficial effect of carotid endarterectomy in symptomatic patients with 23. Bhargava B, Karthikeyan G, Tripuraneni P. Intravascular brachyther- high-grade carotid stenosis. North american symptomatic carotid end- apy: indications and management of adverse events. Am J Cardiovasc arterectomy trial collaborators. N Engl J Med 1991;325:445-53. Drugs 2004;4:385-94.18. Abizaid A, Kornowski R, Mintz GS, Hong MK, Abizaid AS, Mehran R, 24. Kastrati A, Mehilli J, von Beckerath N, Dibra A, Hausleiter J, Pache J, et et al.The influence of diabetes mellitus on acute and late clinical al. Sirolimus-eluting stent or paclitaxel-eluting stent vs balloon angio- outcomes following coronary stent implantation. J Am Coll Cardiol plasty for prevention of recurrences in patients with coronary in-stent 1998;32:584-9. restenosis: a randomized controlled trial. Jama 2005;293:165-71.19. Willfort-Ehringer A, Ahmadi R, Gessl A, Gschwandtner ME, Haumer 25. Brito FS Jr, Rosa WC, Arruda JA, Tedesco H, Pestana JO, Lima VC. A, Lang W, et al. Neointimal proliferation within carotid stents is more Efficacy and safety of oral sirolimus to inhibit in-stent intimal hyperpla- pronounced in diabetic patients with initial poor glycaemic state. Dia- sia. Catheter Cardiovasc Interv 2005;64:413-8. betologia 2004;47:400-6. 26. Endarterectomy for asymptomatic carotid artery stenosis. Executive20. Setacci C, de Donato G, Setacci F, Pieraccini M, Cappelli A, Trovato RA, committee for the asymptomatic carotid atherosclerosis study. JAMA et al. In-stent restenosis after carotid angioplasty and stenting: a challenge 1995;273:1421-8. for the vascular surgeon. Eur J Vasc Endovasc Surg 2005;29:601-7. 27. Lal BK, Hobson RW 2nd, Pappas PJ, Kubicka R, Hameed M, Chakh-21. Clowes AW, Schwartz SM. Significance of quiescent smooth muscle toura EY, et al. Pixel distribution analysis of b-mode ultrasound scan migration in the injured rat carotid artery. Circ Res. 1985;56:139-45. images predicts histologic features of atherosclerotic carotid plaques. J22. Garcia LA, Hosley SE, Baim DS, Carrozza JP Jr. In vivo assessment of Vasc Surg 2002;35:1210-7. stent recoil in normal porcine arteries: evaluation of contemporary stent designs. Catheter Cardiovasc Interv 2001;53:277-80. Submitted Jun 21, 2007; accepted Jul 17, 2007.