OUP user menu

Treatment of in-stent restenosis using a paclitaxel-eluting stent: acute results and long-term follow-up of a matched-pair comparison with intracoronary β-radiation therapy

Peter W Radke , Stefan Kobella , Axel Kaiser , Andreas Franke , Daniela Schubert , Eberhard Grube , Peter Hanrath , Rainer Hoffmann
DOI: http://dx.doi.org/10.1016/j.ehj.2004.02.014 920-925 First published online: 1 June 2004


Aims Intracoronary radiation therapy (ICR) has significantly improved the long-term outcome after treatment of diffuse in-stent restenosis (ISR). The efficacy of drug eluting stents in this setting remains less well defined. This matched-pair analysis compared the procedural and long-term clinical and angiographic outcome after treatment of diffuse ISR using a paclitaxel-eluting stent (PES) with intracoronary β-radiation therapy.

Methods and Results Twenty-two patients receiving 25 PES (ACHIEVETM, Cook, 3.1 μg paclitaxel per square millimeter, non-polymer based coating) for ISR underwent 6-month angiographic and 12-month clinical follow-up. From a database including 141 patients (174 lesions) undergoing intracoronary β-radiation for ISR, 25 lesions (25 patients) were pair-matched with the former group for lesion length and vessel size. PES implantation and ICR were successfull in all patients with a significantly lower postprocedural in-stent diameter stenosis in the PES group (8±12% vs. 18±8%, Math). Angiographic binary in-lesion restenosis at 6 month was 20% (5/25 lesions) in the PES group and 16% (4/25) in the ICR group (Math). PES implantation resulted in significantly higher in-stent MLD at FU (2.10±0.71 vs. 1.75±0.36, Math) and a higher in-stent net gain (PES: 1.19±0.69, ICR: 0.84±0.49, Math). Two patients in the PES group and 6 patients in the ICR group experienced a target lesion revascularisation at 12-month follow-up (Math).

Conclusion Implantation of a non-polymer based paclitaxel-elution stent and conventional ICR therapy for complex ISR lead to comparable acute and long-term clinical and angiographic follow-up results.

  • Stent
  • Restenosis
  • Brachytherapy
  • Drug-eluting stent
  • Paclitaxel

See Page 898 for the editorial comment on this article1


The prevention and treatment of in-stent restenosis (ISR) remain among the most challenging problems in interventional cardiology. The introduction of drug-eluting stents has significantly reduced the incidence of ISR as compared to bare metal stents in de novo lesions.1,2 Regarding the treatment of ISR, intracoronary radiation therapy (ICR) seems to be the only modality to effectively reduce the rate of recurrent restenosis.3 Radiation therapy, however, has potential side effects and requires a specific infrastructure.4,5 The efficacy of drug-eluting stents for the treatment of ISR lesions remains less well established. This study compared the acute procedural, 6-month angiographic, and 1-year clinical follow-up results after treatment of diffuse ISR using paclitaxel-eluting stents (PES) with intracoronary β-radiation therapy in a matched-pair analysis.


Patients and procedures

Forty-seven symptomatic patients (60±11 years, 79% male) presenting with 50 ISR lesions (Math50% diameter stenosis) in native coronary arteries treated at the Aachen University Hospital formed the study population.

Initially, 27 consecutive lesions in 24 patients were treated with a PES (AchieveTM, Cook, 3.1 μg paclitaxel per square millimeter, non-polymer-based coating). These patients were recruited from the DELIVER II trial, a prospective, non-randomised, multicentre, consecutive enrolment study evaluating non-polymer PES in the treatment of patients with lesions at high risk for revascularisation. The primary study endpoint was target-lesion revascularisation at 180 days. For the purpose of this comparative study, all patients underwent 6-month angiographic and 12-month clinical follow-up assessment. Inclusion criteria were lesion length requiring treatment with up to two PESs, and vessel diameter between 2.5 and 4.0 mm. Exclusion criteria were an ejection fraction Math30%, unprotected left main location, heavy calcifications, excessive tortuousity of a proximal vessel, life expectancy of less than 1 year, myocardial infarction within the previous 72 h, and previous intracoronary brachytherapy. Two patients withdrew consent for follow-up angiography. Neither of the 2 patients experienced a major adverse cardiac event (MACE) and both were symptom-free at 12 months of follow-up. Therefore, 22 patients with 25 lesions were available for angiographic and 12-month clinical follow-up analysis.

The control group was recruited from a prospective registry including all demographic, clinical, and angiographic data of 141 patients (174 lesions) undergoing ICR (90Sr Novoste system) for ISR between April 2000 and July 2002 at our institution. During enrolment of the DELIVER II study, patients only underwent ICR if they did not meet the DELIVER II inclusion criteria or if they refused to participate in the DELIVER II trial. Angiographic 6-month and clinical 12-month follow-up rates for the entire ICR patient population were 73% and 100%, respectively. Only patients fulfilling the DELIVER II inclusion and exclusion criteria qualified for the matching process. Of those, 25 lesions (25 patients) were matched with the PES group for lesion length and vessel size. For matching purposes, all other clinical and angiographic variables were blinded. From the 25 patients initially included in the matching process, 3 patients did not undergo follow-up angiography (all symptom-free, no major adverse cardiac event). In a second step, 3 further ICR patients were matched. All of these patients underwent follow-up angiography. The two patient groups are compared in Tables 1 and 2.

View this table:
Table 1

Baseline characteristics

View this table:
Table 2

Quantitative coronary angiography

In both groups, heparin was administered during the procedure according to standard practice. Aspirin (100 mg/day) and clopidogrel (300 mg loading dose) were started before the procedure. Postprocedure, in addition to aspirin, clopidogrel (75 mg/day) was administered for 3 months to the PES group and for 12 months to the ICR group.

All patients gave written informed consent. The protocols were approved by the ethics committee of the Medical School, University of Aachen.

Paclitaxel-eluting stent implantation

In the DELIVER II trial, direct stenting for the treatment of in-stent restenosis was encouraged. In case of predilatation before stent placement, a second wire was placed within the vessel to prevent balloon slippage in the restenotic lesion. Subsequently, the use of a stent longer than the initial balloon length was encouraged.

An ACHIEVETM stent (Cook) of 8, 13, 15, 18, 23, or 28 mm in length and 2.75, 3.0, 3.5, or 4.0 mm in diameter was used in this study. Using a proprietary process, paclitaxel (dose density 3.0 μg per square millimeter), was bound to the stents without using a polymer.

Intracoronary Math-radiation therapy

The Beta-Cath device (non-centred β-emitter, Novoste Europe SA/NV Brussels, Belgium) was used in all patients. This device consists of a hand-held, multiple-use transfer device, a 5F delivery catheter, and a 40- or 60-mm source train of strontium-90 seeds, depending on lesion length. The safety, feasibility and efficacy of this device has been previously evaluated for de novo and ISR lesions.6

Briefly, following conventional balloon angioplasty, the delivery catheter was advanced over a 0.014-inch guidewire. The delivery catheter was then positioned precisely across the dilated segment using the proximal and distal radiopaque markers. Confirmation of the delivery catheter position was provided by angiography. After the procedure, a final angiogram was performed. When new dissections or extensive elastic recoil (defined as ⩾30% acute recoil) could be documented, further balloon dilatation or stent placement was performed if deemed necessary by the investigator.

In-hospital and 12-month follow-up clinical outcomes

All clinical data were verified by independent review of the hospital charts and source documentation.

Procedural success was defined as final-diameter stenosis Math30% in the treated lesion and the absence of major clinical complications (in-hospital death, Q-wave myocardial infarction, or emergency coronary bypass surgery). In addition to the 6-month angiographic follow-up, all patients were contacted at 30 days and 12 months after the procedure for any major adverse cardiac event (MACE), defined as death, myocardial infarction, or the need for target-lesion revascularisation.

Quantitative coronary angiography

Quantitative angiographic analysis was performed at an independent core lab of the University of Aachen using the CAAS II System (PieMedical, Maastricht, The Netherlands) as described earlier for the treatment of in-stent restenosis.7 In brief, quantitative measurements included reference diameter, lesion length, and the minimal luminal diameter (MLD) before and after the procedure and at follow-up. A user-defined reference luminal diameter of a proximal, angiographically normal-appearing segment was chosen. After the procedure and at follow-up, the minimal luminal diameter was determined for the target lesion, which was defined as the in-stent segment plus the proximal and distal 5-mm edge segments, and for the stented segment without adjacent edge segments. Acute gain was defined as the change in MLD from pre- to postintervention, late lumen (LL) loss was defined as the change in MLD between postintervention and follow-up, and net lumen gain was defined as the change in MLD from preintervention to follow-up. In-stent restenosis was classified as previously described.8 Recurrent restenosis was defined as diameter stenosis ⩾50% within the target lesion at follow-up angiography.

Statistical analysis

Due to the exploratory nature of this study, no adjustments to the significance level were made. Data were not normally distributed, but departures from the normality assumption were modest. All continuous data are presented as mean±SD for the convenience of the reader. All tests were two-sided. Comparisons between groups were performed with the paired Student t test for continuous variables. Qualitative data are presented as frequencies and quantitative data were compared using Fisher's exact test. A Math-value of Math0.05 was considered statistically significant.


Patient characteristics

A comparison of the two patient populations is provided in Tables 1 and 2. Both groups were well balanced for lesion length and vessel size as a result of the matching process. In addition, further preprocedural angiographic variables, demographic data, type of ISR, and length of the primarily implanted stent were also comparable. Seventy-six percent lesions in either group were classified as “diffuse.”

In the PES group, predilatation was performed in 20 lesions and direct stenting in 5 lesions. Four lesions (16%) were each treated with 2 paclitaxel-eluting stents.

In the ICR group, the mean prescribed dose was 21.1±3.1 Gy at a distance of 2 mm from the radiation source (range 16.0–25.3 Gy); the dwell time was 228±45 s. Atheroablative techniques and the cutting balloon were not used. Three patients (12%) received an additional bare-metal stent at the stent margins due to dissection.

Clinical outcome

Clinical 30-day and 12-month follow-up data were available for all patients. The procedural success was 100% and there was no MACE at 30 days in both groups. No subacute stent thrombosis occurred in either group. During the 12-month follow-up, 2/22 (9%) patients in the PES group and 6/25 (24%) patients in the ICR group experienced a MACE due to recurrent restenosis (all target-lesion revascularisations, Math). One of the patients undergoing ICR and receiving additional stents at the margins experienced a MACE due to recurrent restenosis (85% diameter stenosis at angiographic follow-up) and subsequent target-lesion revascularisation. There were no deaths and no myocardial infarctions.

Angiographic outcome

All 47 patients underwent 6-month follow-up angiography (100%, Table 2). Binary restenosis was documented in 5/25 lesions (20%) treated with a PES and in 4/25 lesions (16%) receiving ICR (Math). Paclitaxel-eluting stent implantation lead to a higher postprocedural in-stent lumen diameter (2.54±0.42 vs. 2.29±0.35, Math) and less in-stent diameter stenosis (8±12 vs. 18±8, Math). With a comparable late lumen loss in both treatment groups, this resulted in a significantly higher in-stent minimal lumen diameter (2.10±0.71 vs. 1.75±0.36, Math) and a higher net lumen gain (1.19±0.69 vs. 0.84±0.49, Math) in the PES group.


This study compared a non-polymer-based paclitaxel-eluting stent (PES) with intracoronary β-radiation for the treatment of complex ISR.

Recent reports have suggested beneficial effects of sirolimus9,10 and paclitaxel-eluting stents11 in the treatment of ISR. At present, however, only intracoronary radiation therapy has been shown to effectively reduce the incidence of recurrent ISR.3,12 Intracoronary radiation therapy, however, requires a specific infrastructure and carries the risk of potentially fatal side effects (i.e., late thrombosis5) Given the ease of use and availability of drug-eluting stents, they are likely to become the predominant treatment for diffuse in-stent restenosis, even when angiographic and clinical results obtained are comparable to ICR. Despite a lack of randomised trials thus far, this matched-pair analysis provides further support for the potential benefit of drug-eluting stents in the treatment of ISR.

This study was neither randomised nor powered to detect clinical treatment differences. In the PES group, however, the MACE rate was very low (9%). It is interesting that the clinical event rate in the TAXUS III study using a polymer-bound, paclitaxel-eluting stent for the treatment of diffuse in-stent restenosis was higher than in this series (29%). Differences in study populations (i.e., 18% bypass interventions in the TAXUS III study) and a higher proportion of patient in TAXUS III receiving 2 or more stents (46% vs. 16%) probably account for this. Notably, the size of the study population in both trials was relatively small.

Clinical long-term follow-up in this study identified only 2 PES patients who experienced a MACE (both target-lesion revascularisation). Both patients showed angiographic restenosis at 6 months of follow-up and events occurred in the first 7 months after the index procedure (5 and 7 months, respectively). In the ICR group, however, 4 patients with angiographic restenosis at 6 months were identified. Of those, 3 patients had to undergo target-lesion revascularisation and 1 patient showed a non-significant 60% stenosis. In addition, 3 further patients without angiographic restenosis at 6 months of follow-up required target-lesion revascularisation due to delayed restenosis in the first year after the index procedure (10–12 months). Indeed, recent preclinical data suggest that clonogenic inactivation is the mechanism by which radiation therapy significantly delays the recurrence of ISR rather than completely preventing this process.13 These findings are further supported by clinical observations showing a late catch-up phenomenon in patients undergoing ICR for ISR.14 Of note, this phenomenon has also been described in a paclitaxel-derivate-eluting polymer stent system (QuaDS),15 but was not observed in this study using a paclitaxel-eluting (non-polymer-based coating) stent. Clearly, this observation underscores the complexity of drug-eluting stent systems. Both the drug carrier systems (i.e., polymers) and the antiproliferative substances used (i.e., paclitaxel, paclitaxel-derivatives, sirolimus) determine the therapeutic potential of the stent system to reduce/suppress neointimal hyperplasia.

Regarding the clinical long-term outcome, PES implantation not only compares favourably to the matched ICR group (24% MACE), but also to (1) previously reported data from randomised trials using β- and γ-radiation (15–28% MACE3,12) and to (2) stent-in-stent therapy using uncoated stents (22–44% MACE16).

The angiographic analysis showed a significantly lower postprocedural in-stent diameter stenosis in the PES group compared to the ICR group (8±12% vs. 18±8%, Math). This observation deserves special attention as a recent meta- regression analysis has revealed that postprocedural diameter stenosis is the strongest predictor of the frequency of MACE in patients undergoing treatment for in-stent restenosis.16 This may not apply for ICR as a recent report has suggested that the residual stenosis after angioplasty preceding ICR does not affect clinical outcome, at least until 6 months after the procedure.17

In-lesion angiographic analyses did not show differences between treatment groups. A significant edge-stenosis phenomenon, however, was not observed in the PES group. As a result of small patient numbers, the advantageous procedural result for the PES group probably did not translate into a superior long-term outcome in this study. Of note, previous studies have also shown better postprocedural lumen dimensions using the stent-in-stent technique as compared to balloon angioplasty alone.18

Lumen loss, as a surrogate for the amount of recurrent neointimal hyperplasia, was 0.44±0.53 mm in the PES group and, thus, comparable to the ICR group (0.55±0.60 mm). In addition, these data are also well in line with the findings of large ICR trials (i.e., 0.44±0.69 mm in the INHIBIT trial3) or recently published data using a slow-release, polymer-bound, paclitaxel-coated stent (0.54±0.51 mm11).

As the implantation of additional stents should be minimised when using ICR due to an increased rate of adverse events,12,19 PES implantation for ISR seems to particularly benefit from improved procedural results in combination with antiproliferative effects, leading to a significantly higher in-stent net lumen gain in this study.

Finally, this study has intrinsic limitations. First, the study was not randomised, but both study groups were recruited from prospective registries with prespecified endpoints. Furthermore, the group sizes are relatively small. The next level of evidence supporting the use of drug-eluting stents for the treatment of in-stent restenosis, however, will likely be obtained by randomised studies rather than an increase in patient populations of case-control studies. Intravascular ultrasound was not routinely performed during the procedures. Therefore, further information regarding the mechanism of acute lumen gain and recurrent restenosis in the PES group is not available.

In summary, PES implantation combines advantageous procedural results with a low late lumen loss. Even though total patient numbers treated with drug-eluting stents for diffuse ISR are small at the moment, these favourable mechanistic results clearly warrant reasonably powered randomised trials with clinically relevant endpoints. If comparable results are confirmed, drug-eluting stents may become the preferred treatment modality for patients with diffuse in-stent restenosis.


The expert statistical assistance of Robert Kwiecien, PhD, Institute for Medical Statistics, University Hospital of Aachen, is gratefully acknowledged.


  1. [1]
  2. [2]
  3. [3]
  4. [4]
  5. [5]
  6. [6]
  7. [7]
  8. [8]
  9. [9]
  10. [10]
  11. [11]
  12. [12]
  13. [13]
  14. [14]
  15. [15]
  16. [16]
  17. [17]
  18. [18]
  19. [19]
View Abstract