Drug-eluting stents for the treatment of in-stent restenosis: A clinical review

1. Introduction 

In-stent restenosis (ISR) remains a major limitation of percutaneous coronary intervention (PCI) with bare-metal stent implantation. Rates of restenosis have been reported to occur in 10% to 50% in clinical practice [1], [2]. The treatment of ISR is associated with a high recurrence of restenosis, from 30% to 80% in complex lesions [3]. Intracoronary radiation therapy (IRT) has been the only proven therapy for ISR [4], [5]. However, drug-eluting stents (DES) hold promise as comparable therapy. Sirolimus-eluting stents (SES) and paclitaxel-eluting stents (PES) have been shown to reduce ISR and the need for repeat revascularisation to <10% in patients with de novo lesions [6], [7]. Promising results obtained in the primary prevention of ISR in de novo lesions has led to interest in extending DES application to more complex lesions, such as the treatment of ISR. The safety and efficacy of DES for the treatment of ISR are less defined. This paper reviews current evidence for DES implantation for the treatment of ISR.

2. In-stent restenosis 

The principal cause of ISR is neointimal hyperplasia resulting from the excessive proliferation of smooth muscle cells in response to injury during stent implantation [8]. Factors known to increase the risk of ISR include smaller vessel diameter, prior restenosis, length of stented vessel, and diabetes mellitus [3]. Effective treatment of ISR requires suppression of this intimal proliferation.

Sirolimus and paclitaxel suppress both smooth muscle proliferation and intimal hyperplasia [9], [10]. The efficacy of SES (Cypher, Cordis J&J) and PES (TAXUS, Boston Scientific) in reducing intimal hyperplasia in de novo lesions has been observed with intravascular ultrasound (IVUS) in several trials [6], [7]. The use of DES to treat ISR is an attractive option. Compared with IRT, stent implantation is simple and requires no additional personnel (e.g., radiation oncologist) or equipment. To date, a number of studies have shown the novel use of DES to treat ISR with SES and PES to be safe and effective (Table 1, Table 2; [11], [12], [13], [14], [15], [16], [17], [18], [19], [21], [22]).

Table 2. Summary of published studies on PES for the treatment of ISR

	Study	Iofina et al. [16]	Radke et al. [19]	Tanabe et al. [22]	Kastrati et al. [15]
	Drug	Nonpolymer (ACHIEVE, Cook)	Nonpolymer (ACHIEVE, Cook)	Polymer (TAXUS, Boston Scientific)	Polymer (TAXUS, Boston Scientific)
	Number of patients	24	22	28	100
	Diabetes mellitus (%)	21	23	14	31
	Diffuse ISR (%)	78	76	64	40
	Recurrent ISR (%)	N/A	N/A	N/A	N/A
	Mean lesion length (mm)	13.7	13.3	13.6	12.4
	Follow-up (months)	9	12	12	6
	Binary restenosis rate (%)	20	20	16	14.3
	Late loss (mm)	0.43±0.47	0.44±0.53	0.54±0.51	0.1
	TLR (%)	8	9	21.4	8
	MACE (%)	8	9	29	11

Table 1. Summary of published studies on SES for treatment of ISR

	Study	Degretekin et al. [11]	Sousa et al. [12]	Saia et al. [13], [18]	Werner et al. [14]	Kastrati et al. [15]	Iofina et al. [16]	Airoldi et al. [17]
	Number of patients	16	25	44	22	100	28	60
	Diabetes mellitus (%)	25	24	25	50	31	25	23
	Diffuse ISR (%)	81	68	58	100	76	71	86
	Recurrent ISR (%)	N/A	20	25	77	N/A	N/A	24
	Mean lesion length (mm)	18.4	13.6	17.5	23.4	12.4	10.6	15.5
	Follow-up (months)	9	12	9	12	6	9	6
	Binary restenosis rate (%)	20	4	14.6	14	14.3	13	13
	Late loss (mm)	0.21±0.62	0.36±0.46	0.17	0.39±0.54	0.1	0.29±0.52	0.35±0.73
	TLR (%)	8.3	0	16.3	14	8	11	10.9
	MACE (%)	18.7	0	18.5	14	8	14	10.9

3. Sirolimus-eluting stents 

Initial clinical data on SES for ISR arose from two small studies involving 16 patients from Rotterdam, the Netherlands, and 25 patients from São Paulo, Brazil, with ISR in native vessels [11], [12]. The lesion characteristics in the Rotterdam series were more complex than in the São Paulo series. Diffuse ISR was present in 68% of the São Paulo patients and 81% of Rotterdam patients. Mean lesion length was 18.4 mm in the former and 13.6 mm in the latter study. Four Rotterdam patients presented with ISR following failed IRT. These differences in risk profile were reflected in 1-year outcomes. One patient from São Paulo developed binary restenosis (≥50% diameter stenosis) on routine angiography, but there were no target lesion revascularisation (TLR) or major adverse cardiac events (MACE). By comparison, event rates were higher in the Rotterdam group, with three patients developing restenosis (TLR=8.3%, MACE=18.7%, with one myocardial infarction and two deaths).

Since April 2002, the Rotterdam University Hospital Thoraxcenter had adopted SES implantation as the default strategy for all coronary interventions as part of the Rapamycin-Eluting Stent Evaluation at Rotterdam Cardiology Hospital (RESEARCH) Registry. From this registry, 44 consecutive patients without previous IRT were treated with SES for ISR [13]. At baseline, 42% of the lesions were focal, 21% diffuse, 26% proliferative, and 11% occluded, with small vessel size (reference diameter ≤2.5 mm) in 49% and a mean lesion length of 17.5 mm. Binary restenosis occurred in 14.6% of the lesion at 6-month angiographic follow-up. At 9-month clinical follow-up, there were no deaths, 4.7% myocardial infarction, and 16.3% TLR. Of note, no restenosis was observed at follow-up in focal lesions. There were no differences in the rates of repeat restenosis between the diffuse (22.2%), proliferative (25%), or chronic total occlusions (20%), in which rates of repeat restenosis have been reported to be 35%, 50%, and 85%, respectively, with conventional treatment [3]. SES may be less effective in nonfocal lesions but appeared to be equally effective in diffuse, proliferative, and occlusive ISR.

The e-CYPHER Registry: Real World use of Sirolimus-Eluting Stents for the Treatment of In-Stent Restenosis has been presented (AHA 2004, New Orleans). This international, Internet-based e-CYPHER Registry was established in April 2002 to assess the performance of the SES in a real-world setting. At the time of presentation, there were 1827 patients (12% of registry) with ISR-treated with SES. Patient characteristics were comparable with the registries described above, with 30% diabetes mellitus, 77% AHA/ACC Type B2 or C lesions, and mean lesion length of 18.2 mm. Six-month outcomes were encouraging, with 2.1% TLR and 3.8% MACE.

4. SES in complex ISR 

Patients with diabetes mellitus and diffuse and recurrent ISR are associated with the highest risk of recurrent stenosis [3]. A study of 22 patients with a high prevalence of these risk factors was performed by Werner et al. [14]. In this cohort of patients, 50% had diabetes mellitus, 77% had more than one previous ISR, and all had diffuse or occlusive ISR. The mean lesion length of 23 mm was longer than the lesions studied in the Rotterdam (18.4 mm), São Paulo (13.6 mm), and RESEARCH (17.5 mm) registries [11], [12], [13]. Despite these high-risk characteristics, angiographic restenosis, TLR, and MACE (all 14%) were comparable to the registry results shown above.

The SECURE (Compassionate use of Sirolimus-Eluting Stents) trial enrolled 252 patients who had complex lesions with no acceptable alternative treatment, including IRT and coronary arterial bypass grafting (CABG). In this study, the majority of patients had ISR (87% had >1 episode of restenosis, 72% had previous IRT, 39% had diabetes mellitus, and 21% had bypass grafts). At 6 months, a higher incidence of MACE was observed in patients who had a previous IRT failure (23.5%) than in patients without previous IRT (8.5%). MACE and TLR in the group treated for bypass grafts (20% and 16.7%) were identical to outcomes in patients who received SES (21.5% and 20%) for native vessel ISR. These favourable results were the first experience of DES in bypass graft ISR.

5. DES versus other treatments of ISR 

A number of studies compared the efficacy of DES in the treatment of ISR with other percutaneous treatment modalities, such as balloon angioplasty (PTCA), cutting balloon (CB) angioplasty, and IRT.

6. Paclitaxel-eluting stents 

The first experience with PES in ISR came from a negative study, which used a paclitaxel derivative eluting stent (QuaDS-QP2, Quanum Medical) on 15 patients with ISR from two centers [21]. Although 6-month angiographic follow-up demonstrated a restenosis rate of 13.3%, by 12 months, this had deteriorated to 61.5%. TLR at 12 months was an excessive 60%.

Two studies evaluated the use of nonpolymer paclitaxel stents (ACHIEVE, Cook) in the treatment of ISR. The study by Iofina et al. [16] described above showed that angiographic restenosis at 9 months in patients treated with PTCA, PES, and SES were 61%, 20%, and 13%, respectively (P=.042). A second study had 22 patients treated with PES and 141 patients treated with IRT pair matched in lesion length and vessel diameter [19]. Angiographic binary restenosis at 6 months was 20% in the paclitaxel group and 14% in the IRT group (P=NS). At 12 months, 9% of patients in the PES group and 24% of patients in the IRT group underwent TLR (P=NS).

The TAXUS III feasibility trial examined PES (TAXUS, Boston Scientific) for ISR in 28 patients [22]. The vessels treated were between 3 and 3.5 mm, with a mean lesion length of 13.6 mm (64% diffuse ISR). At 1 year, MACE occurred in 29% and TLR in 21.4% of patients. This study demonstrated that TAXUS stent was safe and efficacious in the treatment of ISR.

The efficacy of the TAXUS stent in ISR was confirmed in the randomised controlled ISAR-DESIRE trial [22] described above. Compared with PTCA, receiving a TAXUS stent was associated with a relative risk of target vessel revascularisation of 0.58 (95% CI=0.35–0.94). Although ISAR-DESIRE was not designed to compare the efficacy between SES and PES, the secondary analysis of the study was the first to show that SES had a significantly lower TLR (8% vs. 19%, P=.02) and late lumen loss in-stent (2.45 vs. 2.21, P<.05) than did PES. Compared with PES, receiving an SES was associated with a reduced risk of target vessel revascularisation of 0.42 (95% CI=0.19–0.92). This suggested that SES may be superior to PES for the treatment of ISR.

7. Other DES 

Other DES, such as everolimus-eluting stent (Guidant), ABT-578-eluting stents (Abbott Vascular and Medtronic), A9 Biolimus-eluting stents (Terumo, Biosensors), and Tacrolimus-eluting stents, are currently in various phases of development. Their efficacy in de novo lesions will need to be proven first before attempts to extend their applications to more complex lesions, such as ISR.

8. DES for failed brachytherapy 

Recurrent ISR after IRT occurs in 17–32% of patients at 1 year [23]. The option of a second IRT procedure to treat ISR after IRT failure has been evaluated in 51 patients [24]. A TLR rate of 23.5% and a MACE rate of 29.4% were reported at 9-month follow-up. The most common pattern of restenosis after IRT failure was focal (<10 mm; [23]).

Three small studies evaluated the use of SES for ISR after IRT failure ([25], [26], [27]; Table 3). In a study of 12 patients from the RESEARCH registry, at 9-month follow-up, TLR rate was 25% and MACE was 42% [25]. A second study of 15 patients showed 13% TLR and 20% MACE at 12-month follow-up [26]. A third study of 25 patients used IVUS analysis to assess SES implantation in patients with either first episode of ISR or recurrent ISR after failed IRT [27]. The results showed that SES was equally effective in both groups. Late lumen loss and intimal hyperplasia cross-sectional areas in the group with first episode of ISR were 0.2±0.7 mm and 0.2 mm2 and in the recurrent ISR after IRT group were 0.5±0.5 mm and 0.3 mm2, respectively (P=NS). These results were comparable with the results observed in studies on SES in de novo ISR [11], [12], [13]. Results from these three studies suggest SES implantation to be effective in the treatment of ISR after IRT failure.

Table 3. Summary of published studies on DES for failed intracoronary radiation

	Study	Saia et al. [25]	Iakovou et al. [26]	Shiele et al. [27]
	Drug	Sirolimus (CYPHER, Cordis)	Sirolimus (CYPHER, Cordis)	Sirolimus (CYPHER, Cordis)
	Number of patients	12	15	25
	Diabetes mellitus (%)	25	27	20
	Diffuse ISR (%)	75	N/A	N/A
	Recurrent ISR (%)	100	93	100
	Previous IBT (%)	100	100	100
	Mean lesion length (mm)	N/A	25.8	15.4
	Follow-up (months)	9	12	6.9
	Binary restenosis rate (%)	40	17	8
	Late loss (mm)	0.68±1.2	N/A	0.5±0.5
	TLR (%)	25	13	4
	MACE (%)	41.6	20	N/A

9. In-stent restenosis in drug-eluting stents 

DES has significantly reduced the incidence of ISR; however, it has not been eliminated. As the indications for DES continue to expand and more DES are implanted, the prevalence of ISR may, in fact, increase. Studies have shown that the pattern of restenosis associated with DES implantation has changed to a predominantly (84% to 100%) focal pattern and the lesions are mostly in-stent rather than edge restenosis [28], [29]. Areas of stent discontinuity are particularly at risk; thus, overlapping DES is recommended.

Currently, there are no data on optimal treatment of ISR in DES. Therapeutic options include repeat DES implantation with the same or a different DES, IRT, or CABG. A study of 24 patients with post-SES restenosis undergoing repeat PTCA (11%), repeat bare-metal stent (4%), or repeat DES (85%) showed high rates of recurrent restenosis of 42.9% overall [30]. Debulking techniques such as rotablation, DCA, or CB angioplasty have no proven additional advantage.

10. Conclusion 

SES and PES show promise for the treatment of ISR. SES may be superior to PES for the treatment of ISR. There is more evidence for the use of SES in high-risk patients with diffuse and recurrent ISR, with or without prior IRT. However, the efficacy of DES in complex ISR is, as expected, less effective than in focal ISR and in de novo lesions. The durability of DES treatment benefit beyond 12 months remains uncertain. There may be late “catch-up” (delayed restenosis), mandating the need for long-term follow-up. Evidence of the preserved efficacy of DES in de novo lesions at 3 years is encouraging [31]. Despite limitations of IRT, there is currently no head-to-head comparison between DES and IRT to support the superiority of either treatment strategy. Ongoing randomised controlled trials will help decide the role of DES and IRT in the treatment of ISR. However, given the ease of use and availability of DES, they are likely to become the predominant strategy for ISR. As DES implantation becomes more widespread, ISR in DES will continue to challenge interventional cardiologists.