Drug-Eluting Coronary Stents

TAXUS I 

The TAXUS I trial evaluated the safety and performance of an SR paclitaxel-eluting stent versus a noncoated stent in 61 patients with de novo native coronary lesions.153 The primary endpoint of 30-day major adverse cardiac events did not occur in either treatment arm (0%). By 6 months, the MACE rate was 0% in the coated stent arm versus 7% (n = 2) in the bare stent arm (P = NS) and at 12 months 10% versus 3% (P = NS). At 6-month angiography, the coated stent arm had larger MLDs (2.60 versus 2.19 mm, P = 0.007), smaller percentage stenosis (13.6% versus 27.2%, P < 0.001), and smaller late lumen loss (0.36 mm versus 0.71 mm, P = 0.008). On 6-month IVUS, mean minimal lumen area was larger in the coated stent arm (5.6 versus 4.8 mm2, P = 0.027), and neointimal hyperplasia was reduced (14.8 versus 21.6 mm3, P = 0.028).

TAXUS II 

The TAXUS II trial154 was designed as a prospective, randomized, triple-blind study evaluating the safety and efficacy of two consecutive but independent patient cohorts each randomized in a 1:1 ratio to a control uncoated stent versus a paclitaxel-eluting stent, in an SR or MR formulation. In Cohort 1, 131 patients were treated with the TAXUS™ (Boston Scientific, Natick, MA) SR stent and 136 patients were treated with a bare-metal stent. In Cohort 2, a total of 135 patients was treated with the TAXUS MR stent and 134 patients were treated with a bare-metal stent. The population included patients with stable angina undergoing a single large (reference diameter between 3.0 and 3.5 mm) native vessel intervention in a short de novo lesion <12 mm in length fully coverable with a single 15-mm stent.

The primary endpoint of in-stent volume obstruction as assessed by IVUS at 6 months was >60% lower in the paclitaxel-eluting stent arm in both the slow-release cohort (7.9% versus 23.2%, P < 0.0001) and the moderate release cohort (7.8% versus 20.5%, P < 0.0001) compared with a bare stent. At the 6-month follow-up, in-stent restenosis was lower in the paclitaxel-eluting stent arm compared with the bare stent arm, with both slow release (2.3% versus 17.9%, P = 0.0002) and moderate release (4.7% versus 20.2%, P < 0.0001). However, restenosis at the proximal and distal edges did not differ from the control in either cohort.

Minimal lumen diameters were larger in the slow-release paclitaxel-eluting stent arm in the stented segment (2.23 mm versus 1.79 mm, P < 0.0001), the proximal edge (2.40 mm versus 2.23 mm, P = 0.009), and the distal edge (2.19 mm versus 2.07 mm, P = 0.039). Similar results were seen in the moderate-release cohort: stented segment (2.24 mm versus 1.76 mm, P < 0.0001), the proximal edge (2.37 mm versus 2.19 mm, P = 0.005), and the distal edge (2.22 mm versus 2.06 mm, P = 0.013). Late lumen loss was lower in the paclitaxel-eluting stent arm in both the slow-release cohort (0.31 mm versus 0.79 mm, P < 0.001) and the moderate-release cohort (0.30 mm versus 0.77 mm, P < 0.001). MACE (cardiac death, MI, and TVR) at 6 months was lower in the paclitaxel-eluting stent arm in the slow-release cohort (8.5% versus 19.5%, P = 0.013) and the moderate-release cohort (7.8% versus 20.0%, P = 0.006), driven primarily by a TLR reduction (SR 4.6% versus 12.0%, P = 0.043; MR 3.1% versus 14.6%, P = 0.002).

Similar results were observed at 12-month follow-up: MACE (SR 10.9%, MR 9.9%, bare 21.7%, P = 0.002 for SR; P = 0.0048 for MR); TLR (SR 4.7%, MR 3.8%, bare 14.4%, P = 0.0035 for SR; P = 0.001 for MR). There were no deaths through 12 months in the slow-release or moderate-release groups.

A 24-month clinical (n = 524) and angiographic (n = 207) follow-up is now available.155 Between years 1 and 2, there was a single cardiac death in each of the control, slow-release, and moderate-release formulations. TLR occurred in eight control patients, one TAXUS slow-release patient, and no TAXUS moderate-release patients. TVR occurred in 19 patients in the control group, 2 patients in the slow-release, and 3 patients in the moderate-release formulations. The total 2-year MACE rates were 24.6% in the combined control group, 14.2% in the slow-release TAXUS group, and 14.2% in the moderate-release TAXUS group (P = 0.0178). No increased rate of TLR was noted over time. Survival free of TLR was 96.3% for the moderate-release formulation, 94.6 for the slow-release formulation, and 83.3% for the control arm (P = 0.0002). Six-month TLR rate for the slow-release formulation was 4.6% at 6 months, 4.7% at 1 year, and 5.5% at 2 years. With the moderate-release formulation, the TLR rate was 3.1% at 6 months, 3.8% at 1 year, and 3.9% at 2 years. For the control, bare-metal stent group, the TLR rate was 13.3% at 6 months, 14.4% at 1 year, and 15.5% at 2 years. No significant difference in late stent thrombosis was detected. Angiographic and IVUS subdata revealed that late loss and minimum lumen diameter remained stable over time. The primary endpoint of percentage in-stent volume obstruction was 17.07 ± 12.21% in the control group (n = 77), 10.59 ± 8.43% (n = 43) in the slow-release TAXUS stent, and 11.93 ± 9.67% (n = 41) for the moderate-release TAXUS stent (P = 0.0018 for control versus TAXUS slow release and P = 0.0139 for control versus TAXUS moderate release).

The results of this study showed that the beneficial effects of the TAXUS stent (both SR and MR) for the prevention of in-stent restenosis previously observed at 6 months and 1 year continued throughout 2 years of follow-up. There was no late catch-up effect and no increase in the percentage of late stent thrombosis or stent malopposition. Efficacy was noted by clinical, quantitative coronary angiography (QCA) and IVUS imaging analysis.

TAXUS IV 

The TAXUS IV trial is the largest paclitaxel-eluting stent trial to date156 using the slow formulation of the paclitaxel-eluting stent on the Express™ stent. In the study, there were 1326 patients with de novo lesions up to 32 mm in length, with 2.5- to 3.5-mm vessel diameter. The primary endpoint was Ischemic-driven TVR, adjudicated by the Clinical Event Committee. The secondary endpoints were MACE at 30 days and 1 year, defined as cardiac death, MI, or TVR; angiographic restenosis at 9 months; and IVUS at 9 months. Follow-up is available for 2 years157, 158 and is planned to continue further for a total of 5 years.

GP IIb/IIIa inhibitors were used in 56.7% of patients in the bare stent arm and 57.7% of patients in the paclitaxel arm (P = 0.74). Stent length did not differ between the treatment arms (21.7 mm in bare stent arm versus 21.9 mm in paclitaxel arm, P = 0.68), nor did the mean number of stents used (1.09 versus 1.08). There was no difference in 30-day death (0.5% versus 0.3%) or major adverse cardiac events (2.5% versus 2.9%, P = 0.73). The primary endpoint of TVR was lower in the paclitaxel arm (12.0% versus 4.7%, P < 0.0001), as was TLR (11.3% versus 3.0%, P < 0.0001).

The reduction in TVR was similar in many subgroups, including diabetic status; reference vessel diameter <3.0 mm and ≥3.0 mm; stent diameter 2.5, 3.0, and 3.5 mm; artery location; stent length; lesion length; and use of GP IIb/IIIa. There was no difference in 1-year cardiac death (1.3% versus 1.4%, P = 0.83) or MI (4.7% versus 3.5%, P = 0.31), but MACE was higher in the bare stent arm (20.0% versus 10.8%, P < 0.0001), driven by a reduction in TVR (17.1% versus 7.1%, P < 0.0001). Stent thrombosis occurred infrequently in both arms (0.8% with bare stent versus 0.6% with paclitaxel, P = 0.72).

Angiographic follow-up was prespecified in a subset of 732 patients and was complete in 559. Among these patients, percentage diameter stenosis in the analysis segment was lower in the paclitaxel arm compared with bare stent (39.8% versus 26.3%, P < 0.0001). Late lumen loss was smaller in the paclitaxel arm in both the in-segment (0.23 mm versus 0.61 mm, P < 0.0001) and the in-stent (0.39 mm versus 0.92 mm, P < 0.0001). Binary restenosis was also lower in the paclitaxel arm in both the in-segment (7.9% versus 26.6%, P < 0.0001) and the in-stent (5.5% versus 24.4%, P < 0.0001). Similar benefit in binary restenosis occurred in the subgroup analysis, including insulin-dependent diabetics (7.7% versus 42.9%, P = 0.007), small diameter vessels ≤2.5 mm (10.2% versus 38.5%, P < 0.0001), and long lesions >20 mm (14.9% versus 41.5%, P = 0.004). There was no difference in the risk of aneurysm formation or late total occlusion. A potential bias of routine angiographic follow-up is that patients may undergo TVR on the basis of angiographic findings alone, and not symptoms. This would be especially true in the plain stent group, as one would expect to find higher rates of binary restenosis in this group. This was confirmed in the TAXUS IV trial at 1 year. In the plain stent group, a significantly higher percentage of patients who underwent routine angiography had target lesion revascularization compared with those who did not have routine angiography (18.4% versus 12.8%, P = 0.04). This was not the case in the paclitaxel stent group (5.7% versus 3.3%, P = 0.18).

At 24 months, the freedom from TLR was 94.4% in the TAXUS group and 82.6% in the control group (P < 0.0001). Subgroup analysis including diabetes, infarct location, vessel diameter, and lesion length all favored the TAXUS stent. No significant difference in late stent thrombosis was noted. MACE rates between 1 and 2 years were not significantly different (7.3% control versus 5.1% TAXUS; P = 0.009) but the TLR rate was significantly reduced between 1 and 2 years in patients receiving the TAXUS stent (1.6% versus 3.6%; P = 0.03). Overall MACE rate at 2 years remained in favor of the TAXUS stent (24.9% control versus 14.7% TAXUS; P < 0.0001). The overall 2-year TLR rates were 17.4% in the control group and 5.6% in the TAXUS group (P < 0.0001).

Thus, the TAXUS IV trial demonstrated both angiographic and clinical benefits of the EXPRESS paclitaxel-eluting stent in a large number of patients. The angiographic restenosis rate in the bare-metal stent group was lower than that seen in the SIRIUS trial (∼25% versus 30%). Although this was further evaluated by head-to-head comparative clinical trials (details in the Studies Comparing the CYPHER and TAXUS Stents section), a restenosis rate of 25% is still higher than expected from previous stent trials. Similar to the sirolimus-eluting stent, the EXPRESS paclitaxel-eluting stent has acquired FDA approval for use in the United States.

TAXUS III 

The TAXUS III159 study was a small registry using the paclitaxel-eluting stent for 28 patients with in-stent restenosis, less than 30 mm in length, using the slow release formulation. A total of 28 patients with 28 target lesions were enrolled: Fourteen percent of patients had diabetes mellitus; the mean reference segment diameter was 2.75 mm; the mean lesion length was 13.6 mm; and 64% of lesions had a diffuse ISR pattern. Two stents were used in 46% of patients. The primary endpoint was MACE (death, MI, target-vessel repeat PCI, or CABG) at 6 and 12 months after the procedure.

Procedural success was achieved in 96% of patients. There were no cases of acute or subacute thrombosis. Six- and 12-month major adverse cardiac event rates were identical (29 and 32%), driven largely by TVR (six patients with PCI (21.4%) and one patient with CABG (3.6%)). Among the six patients who underwent repeat PCI, three patients met criteria for angiographic restenosis, while two patients had PCI due to incomplete stent apposition or underexpansion, and one underwent PCI due to symptoms with a small luminal diameter at follow-up.

Eighty-nine percent of patients underwent follow-up angiography at 6 months. The binary restenosis rate was 16% (four patients: two in a gap between two TAXUS stents, one in a bare-metal stent used to treat an edge dissection, and one within a TAXUS stent), with an average percentage diameter stenosis of 30.8%, and late loss of 0.54 mm. Further subgroup analyses demonstrated smaller minimal lumen diameter and larger diameter stenosis in patients receiving two stents compared to one stent.

Recently reported 3-year follow-up data160 show that two new MACE, a non-Q-wave MI in a nontarget vessel and death due to ventricular fibrillation, occurred in one patient at 688 days, resulting in a MACE rate of 39% at 2 years. A new repeat PTCA in the target vessel outside the target lesion at day 883 increased the 3-year MACE rate to 43%. There has been no reported stent thrombosis to date.

Although small, this study demonstrates the first experience with the TAXUS paclitaxel-eluting stent for the treatment of ISR. Deployment of the stent appeared safe, with reasonable late loss and restenosis rates. However, the MACE rate (driven primarily by repeat target revascularization) and restenosis rates were significantly greater in this trial than seen in other trials of drug-eluting stents to treat de novo lesions, despite the relatively short lesion length and low-risk profile of the patients enrolled, suggesting that there is a difference in restenotic response of ISR lesions compared to de novo lesions.

TAXUS VI 

The TAXUS VI trial161 evaluated paclitaxel-eluting stents compared with bare-metal stents for treatment of long (18 to 40 mm in length), de novo lesions. Four hundred forty-eight patients were randomized to bare-metal stent (n = 227) or paclitaxel-eluting moderate-release stent (n = 219). Stent lengths were 8, 16, 24, and 32 mm and diameters were 2.5, 3.0, and 3.5 mm.

Baseline characteristics were well matched between the treatment groups, with 58% hypertensives, 20% diabetics, and 72% with hyperlipidemia. Baseline minimum lumen diameter was 0.84 mm versus 0.87 mm (P = 0.39) and diameter stenosis was 70.2% versus 68.6% (P = 0.12) for the paclitaxel-eluting stent group versus the bare-metal stent group, respectively. Average lesion length was 20.9 mm versus 20.3 mm (P = 0.38). Procedural success was reported in 92.7% of the paclitaxel-eluting stent group and 95.2% of the bare-metal stent group (P = 0.32).

At 9-month angiographic follow-up, in-stent binary restenosis was lower in the paclitaxel-eluting stent group (9.1% versus 32.9%, P < 0.0001), as was late lumen loss (0.39 mm versus 0.99 mm, P < 0.0001). Percentage diameter stenosis at 9 months was smaller in the paclitaxel-eluting stent group both in-stent (22.2% versus 42.8%, P < 0.0001) and in the analysis segment (30.4% versus 45.4%, P < 0.0001).

The primary endpoint of target vessel revascularization at 9 months was lower in the paclitaxel-eluting stent group (9.1% versus 19.4%, P = 0.003). There was no significant difference in major cardiac adverse events at 9 months (16.4% versus 22.5%, P = 0.12). There was also no difference in cardiac death (0% versus 0.9%, P = 0.50), Q-wave MI (1.4% versus 1.3%, P = 1.0), or non-Q-wave MI (6.8% versus 4.8%, P = 0.42). Stent thrombosis occurred in 0.5% of the paclitaxel-eluting stent group and 1.3% of the bare-metal stent group (P = 0.62).

This study is the first large-scale trial of the paclitaxel-eluting stent to focus specifically on patients with long lesions. The earlier TAXUS I and II studies have included patients with shorter lesions of ≤12 mm. Despite the longer lesion length, the rate of safety events such as stent thrombosis was not increased with the paclitaxel-eluting stent compared with bare-metal stent and was comparable to other TAXUS trials of the same stent.

TAXUS V 

The TAXUS V trial162 is the North American counterpart of the TAXUS VI trial evaluating the efficacy of the slow-rate release polymer-based paclitaxel-eluting stent PES compared with bare-metal stent in patients with single complex coronary lesions. Patients undergoing nonemergent stent implantation of a single lesion, visually estimated to be ≥10 and ≤46 mm in length, with reference vessel diameter ≥2.25 and ≤4.0 mm, were randomized in a double-blind manner to either paclitaxel-eluting stent (n = 577) or bare-metal stent (n = 579). Stents were available in diameter of 2.25 to 4.0 mm and in lengths of 8 to 32 mm. Patients were followed for 1 year. Angiographic follow-up was performed at 9 months. Baseline characteristics were well matched between the treatment groups, with 31% diabetics and 31% female. Multiple stents were used in 33% of patients, and the mean stent length was 28 mm. The primary endpoint of target vessel revascularization at 9 months was lower in the paclitaxel-eluting stent group compared with the bare-metal stent group (12.1% versus 17.3%, P = 0.018), driven by a reduction in target lesion revascularization (8.6% versus 15.7%, P = 0.0003) with no difference in non-TLR (4.8% versus 4.2%, P = 0.67). As a result of the reduction in target vessel revascularization, the overall MACE rate was also lower in the paclitaxel-eluting stent group (15.0% versus 21.2%, P = 0.008). There was no difference in the other components of the MACE endpoint, including cardiac death (0.5% versus 0.9%), Q-wave MI (0.5% versus 0.2%), or non-Q-wave MI (4.8% versus 4.4%). Stent thrombosis occurred in 0.7% of patients in each group.

At angiographic follow-up, in-stent late lumen loss was lower in the paclitaxel-eluting stent group compared with bare-metal stent (0.49 mm versus 0.90 mm, P < 0.0001) as was in-stent binary restenosis (13.3% versus 31.9%, P < 0.0001). In the subgroup of patients who underwent IVUS, in-stent net volume obstruction was lower in the paclitaxel-eluting stent group (13.1% versus 31.8%, P < 0.001).

In a subgroup analysis of patients who were treated with multiple stents (n = 379; n = 326 with stent overlap), the rate of 30-day MI was significantly higher in the paclitaxel-eluting stent group (8.3% versus 3.3%, P = 0.047). It was suggested this was due to an increase in side branch TIMI flow reduction, which was more frequent in the paclitaxel-eluting stent group (41.9% versus 28.6%, P = 0.016). Nonetheless, the increase in 30-day MI in the multiple stent group warrants further monitoring. Patients in the TAXUS V trial will be followed clinically for 5 years.

TAXUS Stent in the “REAL WORLD” 

The paclitaxel-eluting (TAXUS) stent global registries include the Milestone II Registry, an observational registry of drug-eluting stent selection and use with data pending, and the WISDOM Registry (Web-Based TAXUS Inter-Continental Observational Data Transitional Registry Program). The WISDOM Registry is an internet-based data collection of patient outcomes following paclitaxel-eluting stent implantation in nine countries, with 26 sites and 35 physicians worldwide. Enrollment was complete in July 2003, with 778 patients and 992 stents deployed. Type B2 and C lesions occurred in 71%, CTO in 7%; average lesion length was 15.6 mm and reference vessel size 2.9 mm. Multiple lesions were treated in 15% of patients, with SVG in 2% and lesions greater than 20 mm in 14%. The overall MACE rate at 6 months was 4.4%, with TVR in 2.1%. In the United States, the ARRIVE Registry, a peri-approval registry of the TAXUS stent examining long-term clinical outcomes, is ongoing.

ASPECT and ELUTES Trials 

The ASPECT163 (ASian Paclitaxel-Eluting stent Clinical Trial) and ELUTES164 (European EvaLUation of pacliTaxel-Eluting Stent) studies were designed as dose-finding studies to evaluate a nonpolymeric paclitaxel-eluting stent bound directly to the outer surface of the metal struts.

In the ASPECT163 study, 176 patients were randomized to the Supra G-stent (Cook, Inc., Bloomington, IN) coated with high-dose paclitaxel (3.1 μg/mm2), low-dose (1.3 μg/mm2) paclitaxel, or to an uncoated stent. The primary endpoint of percentage diameter stenosis decreased in both the low- and the high-dose paclitaxel-eluting stent arms (39% in the control group versus 23% in the low-dose group and 14% in the high-dose group; P < 0.001). Late lumen loss was also reduced in a dose-dependent manner in the eluting-stent arm (1.04 mm in controls, 0.57 mm in the low-dose group, and 0.29 mm in the high-dose group; P < 0.001).

Binary restenosis occurred more frequently in the bare stent arm compared with the eluting stent arms (27% in the control arm, 12% in the low-dose arm, and 4% in the high-dose arm; P < 0.001). Neointimal hyperplasia at follow-up was higher in the bare stent arm compared with the eluting stent arms (31 mm3 in the controls, 18 mm3 in the low-dose group, and 13 mm3 in the high-dose group; P < 0.001) in the 81 patients in the intravascular ultrasound substudy.

Three patients in the high-dose arm had a subacute thrombosis compared with only one in the low-dose arm and none in the bare stent arm. Event-free survival at 6 months was 90% in the high-dose arm, 93% in the low-dose arm, and 95% in the bare stent arm.

The ASPECT study continues to add to the growing body of evidence along with the TAXUS trials that paclitaxel-eluting stents reduce restenosis and late lumen loss compared with bare stents. The eluting stents were nonpolymeric, which did show a reduction in late lumen loss, but did not show a significant reduction in the primary endpoint of target vessel failure defined as death, MI, or target lesion revascularization. Nonetheless, despite the positive angiographic findings in ASPECT, more adverse clinical events occurred than had been expected. This might have been due to the antiplatelet therapy used after stent placement, rather than the stent itself. The lack of a polymer for paclitaxel release may also have influenced outcomes. There are no plans to bring this stent platform to the United States.

The ELUTES164 dose-finding study involved 190 patients randomized to bare stent (n = 39) or V-Flex Plus (Cook, Inc., Bloomington, IN) coronary stents (Cook Inc.) coated with escalating doses of paclitaxel (0.2 μg/mm2, n = 37; 0.7 μg/mm2, n = 40; 1.4 μg/mm2, n = 39; and 2.7 μg/mm2, n = 37) applied directly to the abluminal surface of the stent. The paclitaxel was located directly on the stent (ie, no polymer used to modulate the release of the drug).

The primary endpoint of percentage diameter stenosis at 6-month angiographic follow-up was greater in the bare stent arm (33.9%) compared with the 2.7 μg/mm2 group (14.2%, P = 0.006), but there was no difference between the bare stent arm and the lower dose coated stent groups (32.8% for 0.2 group, 27.5% for 0.7 group, and 23.3% for 1.4 group). Late loss was lower in the bare stent arm versus the 2.7 group (0.73 mm versus 0.11 mm, P = 0.002), but there was no difference between the bare stent arm and the lower dose coated stent groups (0.71 mm for 0.2 group; 0.47 mm for 0.7 and 1.4 groups). Binary restenosis trended lower in the 2.7 group versus the bare stent group (3.2% versus 20.6%, P = 0.056), but again, there were no significant differences in the lower paclitaxel doses (20.6% with 0.2 group, 14.3% with 0.7 group, and 13.5% with 1.4 group).

There was no difference in freedom from major adverse cardiac events at 1 year for any dose group versus bare stent (82% bare stent versus 86% for 2.4 group, 90% for 1.4 group, 92% for 0.7 group, and 95% for 0.2 group; P = NS for each versus bare stent). There was one subacute thrombosis in the highest dose group, and one in the bare stent group.

The paclitaxel V Flex Plus PTX stent (Cook Inc.) has been approved for use in Europe. Due to patent issues, there are no plans to market the V Flex Plus PTX stent in the United States.

The DELIVER Trial 

The DELIVER trial165 evaluated the safety and efficacy of the paclitaxel-coated RX ACHIEVE™ Drug-Eluting Coronary Stent System (CSS) compared with the metallic MULTI-LINK RX PENTA™ CSS in 1043 patients with de novo coronary lesions <25 mm in length, in 2.5- to 4.0-mm vessels. Patients were randomized to paclitaxel-coated (3.0 μg/mm2) RX ACHIEVE™ Drug-Eluting CSS (coated stent, n = 522) or the metallic MULTI-LINK RX PENTA™ CSS (bare stent, n = 519).

The primary endpoint was target vessel failure (TVF) at 270 days, defined as death, MI, or target lesion revascularization. Angiographic follow-up was performed in a subset of 442 patients (ACHIEVE n = 228; ML PENTA n = 214). Prespecified endpoints were a 40% reduction in target-vessel failure at 9 months (primary clinical endpoint) and a 50% reduction in binary restenosis at 8 months (major secondary endpoint). Baseline clinical characteristics were comparable between the groups. Patients in ACHIEVE had more type C lesions and a larger reference diameter.

At 270-days follow-up, TVF did not differ significantly between the coated and bare stent arms (11.9% versus 14.5%, P = 0.12), nor did 240-day in-stent angiographic binary restenosis (14.9% versus 20.6%, P = 0.076). There was no difference in mortality (1.0% versus 1.0%, P = NS) or MI (1.0% versus 1.2%, P = NS). In-stent late loss at 240 days was smaller in the coated stent arm compared with the bare stent arm (0.81 mm versus 0.98 mm, P = 0.003). There was no difference in subacute (0.2% for each arm, P = 1.0) or late thrombosis (0.2% for each arm, P = 1.0).

Thus the ACHIEVE paclitaxel-coated stent decreased neointimal proliferation compared with the bare-metal PENTA stent; however, this reduction was insufficient to meet the prespecified primary endpoint of target-vessel failure and the secondary endpoint of binary restenosis.

DELIVER II registry 

The Prospective, Non-randomized, Multicenter Evaluation of the Achieve™ Paclitaxel Eluting Coronary Stent System in the Treatment of Lesions with a High Risk of Revascularization (DELIVER II) registry166 enrolled 1500 patients to evaluate the paclitaxel nonpolymeric-coated ACHIEVE™ Drug-Eluting Coronary Stent System (Guidant) in the treatment of lesions with a high risk of revascularization (including multivessel disease, chronic total occlusions, bifurcation, long lesions, small vessels, and restenosis, including ISR) from 86 sites in Europe, the Middle East, and Africa. All patients were followed for clinical events for 6 months, and a subset of 500 patients will be followed for 1 year.

Mean number of lesions treated per patient was 1.3, and 71.6% were de novo lesions. Complex lesion characteristics included restenotic in 28.5%, chronic or subtotal occlusion in 16.5%, small lesions (<2.75 mm) in 55.2%, long lesions in 16.7%, and bifurcation lesions in 29.0%. Target vessel was the left anterior descending (LAD) in 48% of lesions.

In a per-lesion analysis, the primary endpoint of TVR occurred in 10.5% of lesions by 6 months (8.8% TLR with PCI and 2.5% TLR with coronary artery bypass graft). In a per patient analysis, the 6-month major adverse cardiac event rate was 15.7%; TVF, 16.7%; mortality, 2.3%; Q-wave MI, 2.3%; and non-Q-wave MI, 2.6%.

Risk factors for TLR at 6 months on multivariate analysis included postprocedure minimum lumen diameter (MLD), LAD lesions, number of diseased vessels, restenotic lesions, and total stent length. Diabetes was not identified as a risk factor for TVR at 6 months.

Thus, among patients with complex coronary lesions implanted with the paclitaxel nonpolymeric ACHIEVE stent, postprocedure MLD, LAD lesions, number of diseased vessels, restenotic lesions, and total stent length were identified as risk factors for TLR at 6 months. Because the trial was a registry, it was not designed to assess the efficacy of the paclitaxel nonpolymeric ACHIEVE stent since there was no control group. This trial did evaluate efficacy and demonstrated that the ACHIEVE stent was not associated with a significant reduction in TVF or angiographic binary restenosis compared with control. Given the lack of efficacy in the DELIVER trial, it is unclear if the risk factors would be applicable to the more widely used polymeric coated stents, which other trials have shown to be efficacious in preventing TVR.

Chronic Total Occlusions 

It is well known that success rates for treatment of chronic total coronary artery occlusions with BMSs are lower than for nonoccluded arteries. Even when an occlusion is crossed and dilated successfully, restenosis and reocclusion are more frequent than in nonoccluded arteries.167, 168, 169 In addition, several randomized trials have demonstrated that stent implantation decreases restenosis and reocclusion rates170, 171, 172, 173, 174, 175 and confers a long-term survival advantage.176 However, restenosis remains the major complication limiting late outcome after PCI. For this reason there has been an increasing interest in the use of DES implantation in patients with chronic total occlusions (CTO) and several reports and small studies have recently became available.

Nakamura and coworkers177 investigated procedural and 6- and 12-month angiographic outcomes (analyzed by QCA) and left ventricular function in 60 patients who received SESs and 120 patients who received BMSs. Minimum luminal diameter did not differ immediately after recanalization (SES group, 3.04 ± 0.50 mm versus BMS group, 3.12 ± 0.48 mm). After 6 months, the SES group still had significantly larger luminal diameters (3.04 ± 0.44 mm versus 1.94 ± 0.98 mm) and significantly lower restenosis and reocclusion rates (2 and 0%, respectively) than did the BMS group (32 and 6%, respectively). Late loss was significantly smaller in the SES group than in the BMS group. At follow-up, the SES group had fewer cardiac events, including target lesion revascularization (P < 0.001) than did the BMS group.

Furthermore, Hoye and coworkers178 reported the use of SES for the treatment of a de novo CTO in 56 patients. This CTO cohort was compared with a similar group of patients (n = 28) treated in the preceding 6-month period with bare-metal stent. At 1 year, the cumulative survival-free rate for major adverse cardiac events was 96.4% in the SES group versus 82.8% in the BMS group (P < 0.05). At 6-month follow-up, 33 patients (59%) in the SES group underwent angiography with a binary restenosis rate (>50% diameter stenosis) of 9.1% and in-stent late loss of 0.13 ± 0.46 mm. One patient (3.0%) at follow-up was found to have reoccluded the target vessel.

Another small trial, the Sirolimus-eluting Stent In Chronic Total Occlusion (SICTO),179 was designed to assess the feasibility and restenosis/reocclusion rates of coronary stenting with the Cypher SES in patients with chronic total occlusion. A total of 25 patients being treated with the Cypher stent were recruited in this nonrandomized, prospective, multicenter study. Clinical follow-up was planned at 30 days and 6, 12, 18, and 24 months. Patients also underwent repeat angiography and IVUS at 6-month follow-up. The baseline angiographic characteristics showed a mean lesion length of 30.2 mm and reference vessel diameter of 2.6 mm. Vessels treated included the left anterior descending artery in 40%, right coronary artery in 32%, and circumflex in 28%. The immediate postprocedure in-stent percentage stenosis was 15.7%, compared with a 6-month follow-up stenosis of 19.3%. The in-stent late loss was −0.1 ± 0.3 mm and percentage stent plaque volume was 13.1 ± 18.4%. There were no MACE, and TVR was observed in only two patients (8%) at 6 months. One patient had proximal and distal stenosis outside the stent and one patient had a distal dissection at the index procedure, which was treated at follow-up. The SICTO study shows that in this feasibility phase the CYPHER stent was very effective in the treatment of CTO, with very low rates of TVR and MACE compared with historical data from bare-metal stents. In addition, the IVUS data showed that the CYPHER stent significantly reduced luminal hyperplasia within the stents at long-term follow-up, with minimum luminal area only 0.3 mm2 less at follow-up compared with immediately after the procedure.

More recently, Ge and coworkers180 compared 158 consecutive patients who underwent PCI in CTO lesion with SES (SES group) with 290 consecutive patients with CTO lesion treated with BMS in the same length period immediately before the introduction of SES (BMS group). The patients of the SES group more frequently had established risk factors of coronary heart disease, more multivessel diseases, and longer lesions and received more stents. Clinical follow-up at 6 months was complete for all patients and angiographic follow-up was complete for 85 patients in the SES group (71%) and 204 patients in the BMS group (70%). At 6-months follow-up, the cumulative rate of MACE was 17.5% in the SES group and 31.4% in the BMS group (P = 0.005). The incidence of in-segment restenosis was 11% in the SES group and 33.9% in the BMS group (P < 0.001). TLR at 6 months was 7.5% in the SES group and 23.4% in the BMS group (P < 0.001), respectively. By multivariate logistic regression analysis, usage of BMS (odds ratio (OR): 5.46; 95% confidence interval (CI): 2.33 to 12.8; P < 0.001) and multivessel coronary disease (OR: 2.17; 95% CI: 1.17 to 3.99; P = 0.01) were identified as predictors of MACE during 6-month follow-up.

A recent report181 of single center experience in Italy described 75 consecutive patients with documented myocardial ischemia and/or viability in occluded artery-related myocardial segments who underwent SES implantation after successful CTO (>30 days) recanalization. Vessel diameter was in the 2.5- to 30-mm range and lesion length was in the 8- to 33-mm range. A Cipher stent was successfully implanted in all lesions. No complications occurred during the procedure. In five patients (6.8%) an additional SES has been implanted due to a flow-limiting dissection at the distal edge. No MACEs occurred during the hospital stay. Six-month clinical follow-up obtained in all patients showed one noncardiac death, no MI, and no need for repeat TLR. Six patients had residual angina in absence of demonstrated target vessel-related myocardial ischemia. Twenty-three patients (31%) had PCI for other vessel, without target vessel restenosis. Eight-month angiographic follow-up, obtained in 67 patients (89%), showed four (5.9%) focal ISR. No diffuse or proliferative ISR pattern or reocclusion was seen.

The STENT Registry included a total of 158 CTO procedures (3.3% of all procedures, with a range across centers of 1.4 to 5.9%). Of these, 61 patients received DES alone for a CTO. Acute, 3-month, and 9-month MACE and TVR rates are being examined.182 These are the first emerging, multicenter, long-term data concerning DES for CTO in the United States.

In addition, analysis from the e-CYPHER registry183 assessed the outcomes of 415 patients with a chronic total occlusion (defined as an occlusion >3 months old) treated with Cypher SES. The in-hospital outcomes throughout 30-day and 180-day follow-up compared with those seen for the rest of the patients enrolled in the registry (n = 10,962); there was no difference regarding death (0.6%CTO versus 1.4% registry), myocardial infarction (1.1% versus 0.9%), target lesion revascularization (1.4% versus 1.1%), or major adverse cardiac events (3.1% versus 3%). These encouraging preliminary results suggest that the use of the CYPHER™ stent is safe and appears effective in the treatment of CTO lesions.

The use of PES (TAXUS) in CTO was evaluated in the PACTO Study (Paclitaxel in Chronic Total Occlusion Study).184 In 48 consecutive patients a CTO (duration >2 weeks) was successfully recanalized with implantation of TAXUS stent. Each of these patients was matched to a patient treated with BMS in the preceding time period based on lesion characteristics and clinical parameters. Clinical and demographic data were similar in both groups with a high rate of diabetes in 32 and 34%. Duration of CTOs was >3 months in 50% of patients. The number of stents was 1.7 ± 0.9 in both groups, with a mean diameter of the implanted stent of 3.0 ± 0.3 mm. The mean stented segment length was similar in the BMS group with 36 ± 17 mm and the TAXUS group with 40 ± 20 mm. Periprocedural MACE consisted only of an asymptomatic increase of CKMB in 4.5% with BMS and 5% with TAXUS. At 6-months repeat angiography the TVF was significantly lower with PES than with BMS (8% versus 51%; P < 0.001). Notably, the rate of late reocclusion was lower with TAXUS in 2% versus 23% (P < 0.001). The late loss was significantly reduced by TAXUS with 0.19 ± 0.62 mm versus 1.21 ± 0.70 mm (P < 0.001). All nonocclusive restenotic lesions in the TAXUS group were focal and underwent repeat PCI. No patient in the TAXUS group, but 13% in the BMS group with lesion recurrence, required bypass surgery. The overall MACE rate at 12 months was 13% with Taxus and 48% with BMS (P < 0.001). The clinical success persisted in all 28 patients who completed 18-months follow-up.

Interestingly, a comparison between SES (Cypher™) and PES (TAXUS™) on the outcome of patients with CTO has recently been reported from a Multicenter Registry in Asia.185 A prospective analysis of 308 patients was performed with 358 CTOs (128 Cypher™ and 180 TAXUS™) after successful recanalization (defined as TIMI flow trade 0 and the age of occlusion more than 3 months; Cypher™: LAD 53.3%, LCX 28.0%, RCA 18.7%; TAXUS™: LAD 52.4%, LCX 21.6%, RCA 26.0%). The baseline clinical characteristics between the two groups were similar. The procedural success was 100% in both groups. There was no MACE at 30 days. At 12 months the restenosis rate was 1.5% versus 1.7%, TVR = 2.3% versus 2.2%, MACE at 12 months = 2.3% versus 2.2% for Cypher and TAXUS, respectively (all P’s nonsignificant). These results support the use of both stents in patients with CTO as equally safe with low acute complication and low incidence of restenosis.

These small studies and observations showed that implantation of DES was feasible, safe, and promising for the management of CTO. Randomized comparative studies comparing DES with BMS, and SES with PES, in this high-risk patient subset are, however, lacking.

Sidney C. Smith, Jr.: Randomized comparative studies comparing PES with BMS, and SES with PES could provide valuable information to assist with PCI strategies for CTO.

Bifurcation Lesions 

Treatment of bifurcation lesions is associated with both an increase in early complications (particularly compromise of the main or branch vessel) and an increase in restenosis, regardless of which device approach is used. A number of stent techniques have been used to treat bifurcation lesions including “T-stenting,” “reverse y-stenting,” “trouser-leg stenting,” and stent implantation of the major branch with angioplasty or atherectomy (or both) of the side branch.186, 187 Observational studies suggest no advantage in stenting both branches of the bifurcation lesion compared with stenting one branch and performing angioplasty of the other branch188, 189, 190; in fact, the outcome appears to be worse when stents are placed in both branches. Recommendations and lessons learned from the bare stent era may become obsolete in today’s practice. In the current era of drug-eluting stents, the question is raised as to whether systematic stenting of the side branch could improve outcome and whether strategies recommended to date should be changed.

The SIRIUS Bifurcation trial191 was a multicenter randomized controlled study of SES in true bifurcation lesions. Eighty-six patients were randomly assigned to either stenting both branches or stenting the main branch (MB) with provisional stenting of the side branch (SB). Data were analyzed by actual treatment received, not by intention to treat. Crossover was very high: 22 patients crossed over from stent/PTCA to stent/stent (51.2%), and two patients crossed over from stent/stent to stent/PTCA (4.7%) groups. Lesion angiographic success was ascertained in 59 cases in the stent/stent group (93.6%) and 17 cases in the stent/PTCA group (77.3%). Restenosis at 6-month angiographic follow-up did not differ significantly between the stent/stent (28.0%) and the stent/PTCA (18.7%) groups (P = NS). In-stent late luminal loss trended larger in the stent/stent group versus the stent/PTCA group for the MB (0.28 mm versus 0.14 mm, P = 0.19), but did not differ for the SB (0.50 mm versus 0.37 mm, P = 0.41). During the 6-month follow-up, there was one death in the stent/stent group and none in the stent/PTCA group. There was no significant difference between groups in Q-wave MI (1.6% versus 4.5%), non-Q-wave MI (9.5% versus 4.5%), target vessel revascularization with PTCA (11.1% versus 9.0%), or target vessel failure (19.0% versus 13.6%). There were three cases of stent thrombosis, all of which were in the stent/stent group (4.8%, 3/63). Although these results show a mild reduction in restenosis compared to historical results observed with bare-metal stents, subacute stent thrombosis (4.8%) was higher than that observed in other lesion settings, where the safety profile of SES was much better. In addition, the study had several limitations: the crossover rate was high (>50%); there was not a bare stent comparison arm; the study was not blinded; and the data were analyzed by treatment received rather than intent-to-treat.

Another randomized study, Pan and coworkers192 compared the two same strategies for the SES treatment of bifurcation lesions in 91 patients with true coronary bifurcation lesions. All patients received an SES at the main vessel, covering the SB. Patients from group A (n = 47) were assigned to balloon dilation of the involved SB; patients in group B (n = 44) were randomized to receive a second stent at the SB origin. There were no differences between groups regarding baseline clinical and angiographic data. MACE occurred in three patients from group A (two non-Q-wave myocardial infarctions and one target lesion revascularization). Six-month angiographic reevaluation was obtained in 80 patients (88%). Restenosis of the main vessel was observed in one (2%) patient from group A and in four (10%) from group B. Restenosis of the SB appeared in two (5%) patients from group A and in six (15%) patients from group B. Thus, both strategies are effective in decreasing restenosis rates, with no differences in terms of clinical outcome.

Another observational study193 confirmed these findings in 174 consecutive patients who underwent PCI of bifurcational lesions with SES (one-branch stenting group; n = 57 and two-branch stenting group; n = 117). The incidence of MACE was evaluated in the hospital and at 9-month follow-up. There were no statistically significant differences between the two groups with regard to the incidence of TLR (5.4% versus 8.9%, P = 0.76), TVR (5.4% versus 11.1%, P = 0.51), and cumulative MACE (18.9% versus 23.3%, P = 0.76) at 9 months.

From these studies there is no evidence that stenting both branches with SES is better than provisional stenting. However, in clinical practice, a number of bifurcational lesions will ultimately require stents in both branches. Additionally, in the first two randomized studies, restenosis at the SB origin was two to three times higher in the SB stenting group than in the balloon dilation group. This might reflect incomplete coverage of the ostium, thereby reducing the efficacy of the drug elution. Therefore a technique capable of ensuring full coverage of the ostium is needed whenever the two-stent approach is used.

In an observational study of the SES in a consecutive group of patients (58 patients with 65 de novo bifurcation lesions), restenosis occurred particularly at the ostium of the side branch following the use of T-stenting.194

In another real-world study from France195 82 patients were treated with a Cypher stent, the treatment of bifurcation lesions. Patients with distal left main disease or acute MI with cardiogenic shock were excluded. Patients had multivessel disease in 71% of cases and 0.69 ± 0.71 lesions other than the bifurcation were treated during the same procedure. Bifurcations involved mainly LAD/Diagonal (71%). The proximal reference diameter of the main branch was 2.92 ± 0.46 mm, interpolated reference 2.33 ± 0.50, side branch 1.96 ± 0.34 mm. Stenting of the main branch with provisional T-stenting of the side branch was the strategy in 97% of cases (side branch was stented in 22.5% of cases). A final kissing balloon inflation was performed in 99% of cases. Angiographic success was obtained in 100% of cases for the main branch and 98% for main and side branch. In-hospital outcome was uneventful except for asymptomatic CPK elevation in two patients (2.5%). Seven-month clinical follow-up was obtained in all cases. There were no cases of stent thrombosis, MI, or death. Two patients (2.5%) had TVR (focal restenosis of both branches in one case and focal main branch restenosis in one case). Thus Implantation of the SES at coronary bifurcations using a strategy of main branch stenting with provisional T-stenting of the side branch followed by kissing balloon inflation is associated with a high rate of success and looks promising.

Nonetheless, the T-stenting and modified T-stenting techniques cannot guarantee full ostium coverage, especially in cases with a narrow bifurcation angle. To avoid incomplete SB stenting, Colombo et al196 proposed and reported in 2003 a new technique consisting of crushing the SB stent inside the parent vessel stent (the Crush technique) as a technically straightforward method that ensures complete coverage of the side-branch ostium. In a feasibility and safety study of 20 patients, an in-hospital adverse event occurred in three (two MI, one re-PTCA related to dissection of the main vessel distal to the bifurcation), with no further events at 1 month. In particular, there were no episodes of stent thrombosis.

Airoldi and coworkers recently reported the clinical and angiographic results of 73 consecutive bifurcations from 70 patients treated using the “crush” technique with SES.197 Angiographic success was reached in all lesions. Final kissing inflation was performed in 28 lesions (38%). During the hospital stay, no patient died; four (5.7%) patients had MI. At 6-months follow-up, no patients died and one patient who stopped double antiplatelets after 1 month had a ST-elevation MI. TLR was performed in 16/70 (23%) patients. The 6-month angiographic follow-up rate was available in 61/73 (83%) lesions. Restenosis rate was 33% (20/61: 7% main and side branch and 26% only side branch). No difference was observed in restenosis rate on the main branch between lesions treated with final kissing balloon inflation (KY) and lesions without final kissing inflation (KN) (4% in the KY group versus 8% in the KN group, P = 1.00). The restenosis rate on the side branch was lower in the KY group (17%) in comparison to the KN group (42%) (P = 0.046). This mid-term clinical follow-up indicates a good safety profile of the procedure. The angiographic outcomes show a low rate of restenosis on the main branch and a relatively high restenosis rate on the side branch that could be lowered, accomplishing the procedure with a final kissing inflation.

Another report evaluated the treatment of bifurcation lesions with SES using the “Crush” and “V” techniques.198 This included a total of 169 patients bifurcation SES implantation using either the Crush (n = 111) or “V” (n = 58) technique. A final kissing inflation was performed in 45% of Crush patients. The mean patient age was 65 years and 31% have diabetes. Lesions involved the LAD 68%, LCx 17%, a protected LM 7%, and the RCA 8% of the time. Procedural success was 100% for both branches. Non-Q MI occurred in 12% patients postprocedure. There was no in-hospital revascularization (TLR), thrombosis, or deaths. There were three thromboses—one resulting in cardiac death. Follow-up was obtained in 97 eligible patients. TLR occurred in 14.4%. MACE (defined as death, MI, TVR) occurred in 18.6%. TLR occurred in main branch alone in 0%, side branch alone in 79%, and both branches in 21%. With follow-up still accruing, Crush stenting and V-stenting had similar TLR (15.8% versus 11.8%). Final kissing dilatation in this report in Crush did not decrease TLR rates (14.8% no kiss versus 22% with kiss).

Although treatment of bifurcational lesions using crush stent technique is feasible, there is rising a concern about risk of thrombosis due to three-layer struts at the site of carina. A recent report199 states 178 consecutive patients who underwent PCI in bifurcational lesions by crush stent technique with either SES (n = 106) or PES (n = 72). There were no significant differences between the two groups in procedural complications and in-hospital outcome. Final kissing balloon was performed in 60% patients of the SES group and 78% patients of the PES group (P = 0.01). During 6 months of follow-up, the rate of TLR on either branch was 6.6% in the SES group and 5.6% in the PES group (P = 1.0). The cumulative MACE rate at 6 months was 18% (18.9% of the SES group and 16.7% of the PES group, P = 0.8). Five (2.8%) patients sustained stent thrombosis after the index procedure (two (1.9%) patients of the SES group and three (4.2%) of the PES group, P = 0.7). The two patients of the SES group had thrombosis after premature discontinuation of dual antiplatelet therapy (day 28 and 55, respectively). No significant differences were observed in the incidence of subacute thrombosis (0.9% of SES group versus 0% of PES group, P = 1.0) and late thrombosis (0.9% of SES group versus 4.2% of PES group, P = 0.3) between the two groups. Independent predictors of thrombosis were calcified lesions (OR = 4.71; 95%CI, 1.07 to 20.7, P = 0.04), diabetes (OR = 2.20; 95%CI, 1.15 to 4.23, P = 0.02), lesion length (OR: 1.09; 95% CI: 1.01 to 1.19, P = 0.03), and left main bifurcation (OR: 0.94; 95% CI: 0.90 to 0.94, P = 0.004). Particular attention needs to be directed to this complication when multiple predisposing factors are present.

Left Main Coronary Artery Disease 

Patients with disease of the left main coronary artery (LMCA) can be divided into two groups: those with a “protected” LMCA (patients with a patent graft to the left circumflex or left anterior descending coronary artery), and those with an “unprotected” LMCA (patients with no patent grafts to the left coronary system). Studies of angioplasty of patients with LMCA stenoses reported varying degrees of procedural success but poor long-term results.200, 201, 202 Stenting of the LMCA is primarily performed in patients with a protected LMCA.203 However, stenting is also performed with increasing frequency in patients with an unprotected LMCA, and the results are improved compared with previous studies of balloon angioplasty alone.204, 205, 206 The outcome in these patients also depends on left ventricular function. In patients who undergo stent placement in the LMCA and have normal left ventricular function (ie, those who would have been excellent surgical candidates), the results of stenting are superior to results in those who undergo stenting but are poor surgical candidates or have inoperable lesions.204 However, the potentially fatal complication of LMCA restenosis requires that patients who undergo LMCA stenting require careful follow-up, usually with repeat angiography and serial noninvasive testing.

Sidney C. Smith, Jr.: This is a very important point. Several reports with results of unprotected LMCA stenting have called attention to the higher short-term risk for serious adverse events. Most have recommended early assessment of results within 6 months after tenting using coronary arteriography or stress imaging. In this group at higher risk for post-stent adverse events, patients should be strongly cautioned to adhere to recommended antiplatelet therapy.

There are few reports regarding implantation of drug-eluting stents for LMCA stenosis. Two articles from the RESEARCH registry reported favorable results of SES implantation in LMCA stenoses.207, 208 However, those studies involved small populations, included emergent interventions or protected LMCA stenoses, and had limited angiographic follow-up. Recently, however, more data of DES use in the treatment of unprotected LMCA became available.

Preliminary data on the initial US experience on treatment of LMCA disease with SES from the e-Cypher Post-Marketing Surveillance Registry209 involved 46 patients (93 lesions) with LM disease treated with SES. The mean age was 65.9 years, 65.2% were male, and 45.5% were diabetic. Multivessel disease was present in 87% of patients, and an additional 47 lesions (other than the LM) were treated. Clopidogrel was given to 93.3% and IIb/IIIa inhibitors were used in 41.3% of patients. There were no in-hospital death, repeat revascularization, or Q-wave MI. One patient had a periprocedural non-Q-wave MI. There was one death within 30 days postprocedure. There were no acute or subacute stent thromboses. Another patient died between 30-days and 6-months follow-up. The overall event-free survival rate was 93.5% at 6-months follow-up. There were one TVR and no late stent thrombosis.

A recent report from Italy by Di Salvo and coworkers210 included a series of 80 patients with ULM stenosis who underwent PCI with Sirolimus- (31.3%) or Paclitaxel- (68.7%) eluting stents. Seven patients (8.7%) were affected by in-stent restenosis after a previous BMS. Mean patient age was 65 ± 9 years; the ejection fraction was 51 ± 12; 25 patients (31%) were diabetic; and 40 (50%) had unstable angina. Bifurcation was treated in 55% of patients. Stenosis in other vessels was in 89% of patients. Procedural success was obtained in 98.7%. In-hospital MACE were three (3.7%) deaths (one acute thrombosis, one subacute thrombosis, one cardiac failure not related to stenting) and one CABG for ostial recoil. Non-Q-wave acute myocardial infarction (AMI) occurred in 1 patient (1.2%). At 1-month clinical follow-up (76 patients) there was no recurrence of symptoms, and one patient died of cardiac failure. At 6-month clinical follow-up one patient died of noncardiac disease, and two patients (4.2%) had TLR (one re-PTCA and one CABG). Six-month angiography achieved in 58 patients showed restenosis in 4 patients (6.9%).

Another report from Italy by Sheiban and coworkers211 evaluated 148 consecutive patients with unprotected LMCA stenosis: 86 patients (Group I) LMCA stenosis were treated with BMS and 62 patients (Group II) were treated with DES. Demographic, clinical, and angiographic characteristics were similar in both groups. In group I (86 patients), procedural success was 97.6%. There were two in-hospital deaths (procedural and in-hospital mortality was 2.2%). There were four (4.4%) periprocedural non-Q AMI all in multivessel PCI patients, and no stroke or urgent CABG. Clinical follow-up varied from 7 to 36 months. During the follow-up period, there were three cardiac deaths (3.5%) due to heart failure in two patients with EF <30% and one sudden death. A noncardiac death occurred in 10 patients (13.2%), all at least after 7 months following the procedure. Non-Q AMI occurred in 3 patients (3.5%), and 15 patients (17.4%) had recurrence of angina. Repeat revascularization was required in 14 (Re-PCI in all 14). At the end of the follow-up period, 76% of contacted patients were free of events (65% of them with completed 6-month angiography and no restenosis). Group II (62 patients) procedural success was 100%. One in-hospital death occurred (1.6%) for renal failure. At least 6 months of clinical follow-up was completed in all. Recurrence of angina occurred in three patients (4.8%) and TLR/TVR in one (1.6%). Total MCE was 5.7%. Clinical follow-up ranged from 6 to 22 months: No death, stroke, or CABG. During this period, recurrence of angina occurred in five patients (8.1%); TLR/TVR occurred in three patients (4.8%), and total MACE was 12.9%. Nine-month angiographic follow-up was completed in 31 patients (50%). In-segment restenosis was 4.4%.

In addition, 2-year results from a multicenter registry in Japan212 for 42 patients who received Cypher™ stent in patients with LMCA (male 71.4%, mean age 68.8 years, average ejection fraction 50.5%) suggest that the clinical benefit provided by the Cypher™ stent is durable and late benefit appears to be incremental. In this group the lesion location of LM was ostial in 10 cases (23.8%), mid shaft in 12 cases (28.6%), and distal in 22 cases (47.6%). There was no MACE at 30 days. On follow-up angiography at 6, 12, and 24 months the MLD was 3.40 ± 0.77, 3.38 ± 0.82, and 3.42 ± 0.88 (mm), respectively. The percentage restenosis rate was 4.8% at 6 months and did not change at 12 or 24 months.

In comparison, clinical benefit provided by the TAXUS stent was demonstrated by the Left Main TAXUS pilot study,213 which was conducted in France. This study involved 130 patients with unprotected LM stenosis. Preliminary results were recently presented. The patients were 68 ± 11 years old, 75% male, with unstable angina in 27% and diabetes in 30%. The ejection fraction was 62 ± 13% and Euroscore was 4.6 ± 3.1 (estimated CABG mortality rate of 7.6 ± 17.9%). The number of other treated lesions was 1.1 ± 0.9. Reference diameter of the LM was 3.76 ± 0.49 mm. The lesion was distal in 76% of cases. All patients were successfully stented on the LM with 1.1 ± 0.3 stents (stent length 16.8 ± 6.0 mm). In patients with distal LM, a second stent was used in 31% of cases (length 11 ± 3 mm) and a final kissing balloon inflation was performed in 98% of cases. The total MACE rate at hospital discharge (4.5 ± 3.1 days) was 3.1%. One patient had acute stent thrombosis and died (0.8%) and three patients had a non-Q-wave MI (2.3%). There were no events between hospital discharge and 1-month follow-up (n = 112). A total of 46 patients have reached the 6-month follow-up and 32 had a coronary angiogram. Two patients died from noncardiac causes (total death rate 6.7%, total cardiac death 2.2%) and one patient had restenosis of the circumflex artery ostium (3.1%).

Moreover, two nonrandomized small studies214, 215 have recently been published. In the first study214 Park and coworkers from Korea examined outcomes in 102 patients who received a SES for unprotected LMCA disease, comparing these with outcomes in a historical control group that had received a bare-metal stent. Compared to the BMS group, the SES group received more direct stenting, had fewer debulking atherectomies, had a greater number of stents, had more segments stented, and underwent more bifurcation stenting. The procedural success rate was 100% for both groups. There were no incidents of death, stent thrombosis, Q-wave MI, or emergent bypass surgery during hospitalization in either group. Despite less acute gain (2.06 ± 0.56 mm versus 2.73 ± 0.73 mm, P < 0.001) in the SES group, SES patients showed a lower late lumen loss (0.05 ± 0.57 mm versus 1.27 ± 0.90 mm, P < 0.001) and a lower 6-month angiographic restenosis rate (7.0% versus 30.3%, P < 0.001) versus the BMS group. At 12 months, the rate of freedom from death, MI, and target lesion revascularization was 98.0 ± 1.4% in the SES group and 81.4 ± 3.7% in the BMS group (P = 0.0003).

The second study215 by Chieffo and coworkers included 85 patients who underwent elective DES implantation and 64 who received bare-metal stents for LMCA disease. Both sirolimus-eluting stents (n = 41) and paclitaxel-eluting stents (n = 44) were used. The decision to perform PCI instead of surgery was made on the basis of suitable anatomy and patient/physician preference or contraindications for surgery. Patients treated with DES had lower ejection fractions (51.1 ± 11% versus 57.4 ± 13%, P = 0.002) and were more often diabetics (21.2% versus 10.9%, P = 0.12) with more frequent distal left main involvement (81.2% versus 57.8%, P = 0.003). Furthermore, in the DES group, smaller vessels (3.33 ± 0.6 versus 3.7 ± 0.7 mm, P = 0.0001) with more lesions (2.94 ± 1.6 versus 2.25 ± 1.3, P = 0.004) and vessels (2.03 ± 0.69 versus 1.8 ± 0.72, P = 0.05) were treated with longer stents (24.3 ± 12 versus 15.8 ± 8.6 mm, P = 0.0001). Despite the higher risk patients and lesion profiles in the DES group, the incidence of major cardiac events at a 6-month clinical follow-up was lower in the DES than in the BMS group (20.0% versus 35.9%, respectively; P = 0.039). Moreover, cardiac deaths occurred in three DES patients (3.5%) as compared with six (9.3%) in the BMS group (P = 0.17).

More recently, a recent report from the RESEARCH and T-SEARCH registries216 described 181 patients that underwent PCI for LM stenosis. The first cohort consisted of 86 patients (19 protected LM) treated with bare-metal stents (pre-DES group); the second cohort comprised 95 patients (15 protected LM) treated exclusively with DES. The two cohorts were well balanced for all baseline characteristics. At a median follow-up of 503 days (range, 331 to 873 days), the cumulative incidence of MACE was lower in the DES cohort than in patients in the pre-DES group (24% versus 45%, P = 0.01). Total mortality did not differ between cohorts; however, there were significantly lower rates of both MI (4% versus 12%, P = 0.006) and TVR (6% versus 23%, P = 0.004) in the DES group. On multivariate analysis, use of DES, Parsonnet classification, troponin elevation at entry, distal LM location, and reference vessel diameter were independent predictors of MACE events.

Data from these studies and reports encourage the undertaking of a large, long-term, multicenter randomized study to compare SES implantation and bypass surgery for unprotected LMCA stenosis.

Small Vessels 

Atherosclerotic lesions of small coronary arteries are frequently found in patients undergoing revascularization. However, the revascularization of small coronary arteries is a problem for bypass surgery because it is technically difficult and associated with a high failure and mortality rate217, 218 and for PCI because it is associated with high rates of acute complications and restenosis after standard balloon angioplasty219, 220, 221 and stent implantation.222, 223 Restenosis in small coronary arteries may be as high as 50%, as an inverse relationship between vessel size and angiographic restenosis has reported.222, 223 Potential explanations for the lack of efficacy of coronary angioplasty with or without stent implantation in preventing restenosis in small vessels may be related to characteristics of patients harboring atherosclerotic small-vessel lesions, ie, women, diabetic patients, the elderly, and patients with peripheral vascular disease, all of whom are associated with a higher risk of restenosis.224 Another possible explanation may be related to the narrow diameter of the vessels, which cannot accommodate even minimal neointimal hyperplasia after angioplasty or stent deployment without becoming restenotic. Given their ability to deliver prolonged and sufficient intramural drug concentrations to target coronary segments, drug-eluting stents are able to dramatically reduce neointimal hyperplasia, and this specific mechanism may be particularly useful in reducing restenosis in small coronary arteries.

For this reason there has been great interest in the use of drug-eluting stents for small vessels.225 There is an expanding body of knowledge about this group, from subset analyses of larger multicenter randomized clinical trials and registries and from randomized clinical trials specifically aimed at evaluating small vessels. In the multicenter randomized clinical trial SIRIUS, vessel size was broken down into terciles: small (approximately 2.3 mm), medium (approximately 2.8 mm), and large (approximately 3.3 mm).138 In the bare-metal stent group (control), there was a dramatic relationship between vessel size in terciles and restenosis, ranging from 42.9% in the smallest vessels to 30.2% in the largest vessels. There was also a relationship in the SES-treated patients: the larger vessels had an in-segment restenosis rate of 1.9%, whereas in small vessels it was 18.6%. Although the latter rate remains elevated, it is still significantly improved when compared with that seen with bare-metal stents.

As mentioned earlier there have been two specific randomized clinical trials of SES in the treatment of small-vessel disease (the E-SIRIUS and C-SIRIUS studies). In the E-SIRIUS Schofer and coworkers143 randomly assigned 352 patients undergoing treatment of de novo lesions 15 to 32 mm in length and 2.5 to 3.0 mm in diameter to either an SES or a bare-metal stent (mean vessel diameter was 2.55 mm). There was no difference in initial procedural success rates, which were 100 and 99.4%, respectively, for SES and bare-metal stents. There was also no difference in follow-up events of death or myocardial infarction. There was, however, a dramatic reduction in target lesion revascularization (4.0% versus 20.9%; 95% CI −23.6, −10.2; P < 0.0001), a dramatic reduction in binary restenosis (5.9% versus 42.3%; P = 0.0001), and at 8 months, the minimal lumen diameter was significantly larger with the SES compared with the bare-metal stent (2.22 mm versus 1.33 mm; P < 0.0001). This larger follow-up MLD was accompanied by a marked improvement in the major adverse cardiac event rate.

In the smaller C-SIRIUS study, Schampaert and coworkers145 randomized 100 patients with a vessel size of 2.5 to 3.5 mm to either an SES or a bare-metal stent. As was true with the experience of Schofer and coworkers, at 270 days there was no difference in the hard endpoints of death or myocardial infarction, but a dramatic reduction in clinically driven target lesion revascularization (4.0% versus 18.0%; 95% CI −26.0, −2.0) and angiographic restenosis (2.3% versus 52.3%; P < 0.0001) was observed. Late lumen loss both within the stent and within the treated segment was also dramatically improved with SES compared with bare-metal stents.

The SVELTE226 (Study in Patients with De Novo Coronary Artery Lesions in Small Vessels Treated with the Cypher Stent) Trial was a multicenter, nonrandomized, intravascular ultrasound controlled study conducted in 101 patients with de novo coronary artery lesions in small vessels (RVD 2.25 to 2.75 mm visually estimated) treated with the Cypher SES compared to BMS and SES arms of similar patients in the SIRIUS trial. Patient matching was based on RVD and lesion length. In-lesion late lumen loss and binary restenosis determined by quantitative coronary angiography, as well as in-stent neointimal hyperplasia and MACE rate, were evaluated at 8 months. Outcomes of these patients were compared with outcomes of patients who had small vessel disease treated in the SIRIUS study with either the uncoated Bx Velocity stent (BMS Control, n = 323) or the Cypher SES stent (n = 350). Clinical characteristics were similar between SIRIUS and SVELTE patients. Baseline QCA analysis showed a comparable mean lesion length (13.6 ± 4.3 mm BMS, 13.6 ± 4.6 mm SES, 14.8 ± 6.6 mm SVELTE, P = 0.8) and stent length (22.2 ± 8 mm versus 22.7 ± 9.5 mm versus 24.5 ± 8 mm, P = 0.3) but smaller RVD in SVELTE (2.54 ± 0.3 mm, 2.56 ± 0.3 mm versus 2.37 ± 0.3 mm, P = 0.0003). The LL was significantly reduced with the use of sirolimus stent from 0.81 ± 0.64 to 0.26 ± 0.5 to 0.20 ± 0.38 mm, P < 0.0001, resulting in rates of in-lesion restenosis of 39, 11.6, and 6.3%, respectively, P < 0.0001. The IVUS analysis confirmed a significant reduction in the mean neointimal area with the use of sirolimus stent (2.73 ± 1.6 mm2 BMS, 0.58 ± 0.8 mm2 SES, 0.08 ± 0.1 SVELTE mm2, P < 0.0001). MACE-free survival at 8 months was 85% in BMS, 92% in SES, and 95% in SVELTE (P = 0.001), and TLR was 12.7% versus 4.6% versus 0%, P < 0.0001, respectively.

Another trial, the Sirolimus-Eluting Stent and a Standard Stent in the Prevention of Restenosis in Small Coronary Arteries(SES-SMART) Trial,227 has also been reported with even smaller vessels (mean reference vessel diameter 2.22 mm), which were randomly assigned to either SES (n = 129) or bare-metal stent (n = 128). There was no difference in percentage stenosis at baseline (66.9% for SES versus 66.8% for bare stent, P = NS) or postprocedure (22.4% versus 22.9%, P = NS). At 8-month angiographic follow-up, the primary endpoint of binary in-lesion stenosis was dramatically lower in the SES group compared with the bare stent arm (9.8% versus 53.1%, P < 0.001). Percentage diameter stenosis was also smaller in the SES arm (29.7% versus 50.8%, P < 0.001), while minimum lumen diameter was larger in the SES arm (1.7 mm versus 1.1 mm, P < 0.001). Major adverse cardiac events were lower in the SES arm (9.3% versus 31.3%, P < 0.001), driven primarily by an impressive reduction in target vessel revascularization (7% versus 21.1%, P = 0.002) and myocardial infarction (1.6% versus 7.8%, P = 0.04). By 8 months, four subacute stent thromboses occurred in the bare stent group, and one in the SES group. These data extend the findings of the SIRIUS, C-SIRIUS, and E-SIRIUS trials, which evaluated use of SESs in larger vessels (2.5 to 3.5 mm diameter). The postprocedure percentage stenosis was relatively high in both arms (∼22%), which is likely due to the difficulty of dilatation and stenting in these patients with small arteries.

The postmarketing SES performance in “a real-world” SES implantation in SV PCI comes from e-CYPHER International Registry.228 Among 14,316 patients, 9% have received a SES in at least one lesion with a reference diameter ≤2.5 mm. The incidence of females, diabetics, and multivessel disease was higher than for patients treated in larger vessels. Patients were older, with a higher proportion of previous cardiovascular history. The lesions were shorter with a lower number of restenotic lesions. In 50% of cases, the SV PCI was associated with another treated site, resulting in a larger number of stents used. Antiplatelet preregimen (included GIIb/IIIa inhibitors) was similar in the two cohort groups. Clinical follow-up was obtained in 84% of this population. Comparing the SV group (n = 1333) with other patients (n = 12,983), the results at 180 days were as follows: MACE 4% versus 3.1%, P = NS; death 1.45% versus 1.49%, P = NS; MI 1.83% versus 0.89%, P = 0.0032; TLR 1.54% versus 1.22%, P = NS and stent thrombosis 1.54% versus 0.85%, P = 0.02, respectively. From these results reference diameter is still a strong predictor factor of stent thrombosis and MI. In view of these conclusions, the duration of antiplatelet regimen may have an important impact on this type of PCI.

Additionally, in the RESEARCH Registry experience of 91 patients with 112 lesions with a reference diameter of 1.88 ± 0.34 mm treated with a 2.25-mm SES, the outcome was also excellent.229 The binary restenosis rate was only 10.7%; target lesion revascularization at 12 months was 5.5%, and late lumen loss was only 0.07 ± 0.48 mm.

Long Lesions 

Long coronary lesions represent a difficult lesion subset from both a technical standpoint as well as a clinical outcome.230, 231, 232, 233, 234, 235, 236, 237 Treatment of long lesions is associated with an increase of restenosis after both angioplasty and stent placement. In addition, treatment of complex coronary artery stenosis with a long segment of bare-metal stent is associated with high restenosis rates and poorer clinical outcome. Lesion length, stent length, and placement of multiple stents are all independent predictors of restenosis. Therefore, in contrast to shorter lesions, stent placement for long diffusely diseased coronary segments is frequently avoided.

Degertekin and coworkers238 evaluated the outcomes of a predefined study group composed of patients receiving overlapping sirolimus-eluting stents implanted over a length >36 mm to treat native de novo coronary lesions from the RESEARCH registry. During the RESEARCH registry enrollment, SESs were available at lengths of 8, 18, and 33 mm. Therefore, because of the availability of stent lengths, all included patients had a combination of at least two overlapping stents at a minimum length of 41 mm (ie, one 33-mm SES overlapping an 8-mm SES). The study group comprised 96 consecutive patients (102 lesions). Clinical follow-up was available for all patients at a mean of 320 days (range 265 to 442). In all, 20% of long-stented lesions were chronic total occlusions, and mean stented length per lesion was 61.2 ± 21.4 mm (range 41 to 134). Angiographic follow-up at 6 months was obtained in 67 patients (71%). Binary restenosis rate was 11.9% and in-stent late loss was 0.13 ± 0.47 mm. At long-term follow-up (mean 320 days), there were two deaths (2.1%), and the overall incidence of major cardiac events was 8.3%.

Another study, the Multicenter Prospective Nonrandomized Registry Study for Drug-Eluting Stents in Very Long Coronary Lesions (Cypher versus Taxus) (LONG-DES) trial, was recently presented.239 This study was a nonrandomized registry comparing angiographic and clinical outcomes in 637 patients undergoing stent placement with either BMS (n = 177), the Cypher rapamycin drug-eluting stent (C-DES) (n = 294), or the Taxus paclitaxel drug-eluting stent (P-DES) (n = 166), in long coronary lesions. Baseline characteristics were similar in the three arms including the number of high-risk and diabetic patients. Stent length was similar (42.8 mm C-DES versus 43.1 mm -DES; P = NS). DES patients received more stents with longer lengths when compared to BMS. Patients in the Cypher arm had smaller reference vessel diameter (2.80 mm C-DES versus 2.90 mm P-DES). At 6-month angiographic follow-up, completed in approximately 80% of all patients, there was 65% less in-stent late loss in the C-DES patients compared to P-DES (0.27 mm versus 0.78 mm, P < 0.0001), and 65% less in-segment restenosis in C-DES patients compared to P-DES (7.4% versus 21.3%, P < 0.001). Diameter stenosis was less in the C-DES group compared to P-DES (10.4% versus 29.3%, P < 0.001). The rates of TLR were significantly lower with DES compared to BMS (2.7% C-DES, 5.4% P-DES, and 18.6% BMS; P < 0.001) but the difference between C-DES and P-DES was not statistically significant (P = 0.14).

Stent length as a predictor of restenosis after long SES implantation was evaluated by a substudy240 of the LONG-DES trial. This substudy included 336 de novo long coronary lesions that were treated with long SESs (total stent length = 28 mm). Of 233 lesions with 6-month angiography, in-segment restenosis occurred in 17 lesions (7%). Compared to the group without restenosis, the group with restenosis had small reference diameter (2.6 ± 0.4 mm versus 2.8 ± 0.4 mm, P = 0.032), small postprocedural minimal lumen diameter (2.5 ± 0.3 mm versus 2.7 ± 0.4 mm, P = 0.021), long lesion length (43.1 ± 13.1 mm versus 33.0 ± 13.3 mm, P = 0.006), and long total stent length (49.4 ± 15.3 mm versus 41.3 ± 16.0 mm, P = 0.047) in univariate analysis. Clinical and procedural characteristics were similar in the two groups with or without restenosis. Total stent length (OR 1.04, 95% CI 1.01 to 1.07, P = 0.018) and reference diameter (OR 0.18, 95% CI 0.03 to 0.92, P = 0.039) were independent determinants of restenosis in multivariate analysis.

Moreover recent data from a multicenter registry in Asia241 of elective PCI patients who had long (>25 mm) lesions treated with a single or multiple overlapped Cypher SES included a total of 168 lesions that were treated with 268 Cypher SES. The primary endpoint was the occurrence of MACE; the secondary endpoint included restenosis rate and TLR. The number of stents used was 1.6 ± 0.6 per lesions. The maximum pressure used was 18.8 ± 4.8 atm and balloon artery ratio was 1.1 ± 0.1. The in-hospital procedural success was 100%. No MACE at 30 days. Reference diameter was 2.90 ± 0.38 mm, post-MLD 2.78 ± 0.44 mm, and lesion length 35.8 ± 10 mm. Angiographic follow-up at 12 months showed MLD 2.50 ± 0.50 mm, restenosis rate 3.0%, and TVR of 3.0%.

Stenotic Vein Grafts 

Treatment of stenotic vein grafts accounts for up to 10% of percutaneous interventional procedures.242 Major problems associated with balloon dilatation in vein grafts include increased acute complications such as distal embolization and “no reflow” (which can cause MI) and also high restenosis rates.243, 244, 245 Bare-metal stents have shown superiority over balloon angioplasty for SVG lesions,245, 246, 247, 248, 249, 250 with improvement in the initial success rates and reduced need for target vessel revascularization and have become the standard of care for the treatment of SVG disease in conjunction with distal protection devices. The application of DES to SVG lesions has not been systematically studied in randomized trials to date, and there are theoretic concerns about the efficacy of DES in these lesions,251 because of the expected delay in endothelial healing after DES placement, perhaps increasing the risk of thrombosis. However, these patients experience higher clinical restenosis rates than patients with de novo lesions and are an important group in whom DES might have an impact on long-term outcomes. It will be essential, however, to document that DES for SVG disease is not associated with a higher subacute stent thrombosis rate or with thromboembolic complications and to document the impact on late restenosis.

Nonetheless there are an increasing number of reports suggesting favorable outcomes with DES compared to conventional BMS in the treatment of SVG. Chu and coworkers252 recently reported 56 patients with 70 SVG lesions who underwent standard PCI with SES compared with 511 patients with 721 SVG lesions with BMS. All patients received distal protection devices (DPD) during SVG intervention. Baseline clinical and procedural characteristics were balanced between both groups; however, the SES group had more ostial lesions (22.7% versus 12.7%, P = 0.024) and restenotic lesions (18.6% versus 10.3%, P = 0.035). The patients in the SES group had lower in-hospital major CK-MB elevations. At 30 days, non-Q-wave MIs were significantly lower in the SES group. This benefit continued to be lower up to 6 months, although death, Q-wave MI, TLR, TVR, and MACE were similar between both groups.

Another report by Sharma and coworkers253 compared 47 patients who underwent SES implantation in SVGs with 52 consecutive patients who received BMS in the 9 months prior to SES availability. Baseline variables between both groups were similar. Patients undergoing SES implantation more frequently received embolic distal protection (69.4% SES versus 46.2% BMS). Mean stent diameter was smaller in the SES treatment group (3.15 ± 0.4 versus 3.43 ± 0.5mm), with a trend toward increased total stent length in the SES cohort (27.2 ± 18.1 versus 22.0 ± 10.4 mm). At a mean follow-up of 12.3 months, the SES group had a 56% reduction in MACE (15.0% versus 34.2%, P = 0.03), primarily driven by an 84% reduction in clinical restenosis (4.3% versus 26.9%, P = 0.002).

In addition, Ge and coworkers254 reported 76 consecutive patients who underwent PCI in SVG lesions treated with DES over a 2-year period. A control group was composed of 89 consecutive patients with SVG lesions treated with BMS in the same period immediately before the introduction of DES. Among patients treated with DES, 44 patients received sirolimus-eluting stents and 32 patients received paclitaxel-eluting stents. There were no significant differences between the two groups in angiographic success rate and in-hospital outcome. Clinical follow-up at 6 months was complete for all patients and angiographic follow-up was complete for 63.6% in the DES group and 65.2% in the BMS group. Cumulative MACE during 6 months was 12.7% in the DES group and 28.1% in the BMS group (P = 0.04). Patients treated with DES had a significantly lower incidence of in-segment restenosis (10.9% versus 27.7%, P = 0.03), target lesions revascularization (3.2% versus 17.5%, P =0.005), and TVR (6.5% versus 20.8%, P = 0.01). By multivariate logistic regression analysis, diabetes, usage of BMS, age of SVG, maximal inflation pressure, and baseline reference vessel diameter were identified as predictors of MACE during 6-month follow-up.

In addition to these small reports, data analyses from larger registries describing “real-world” use of DES in SVG lesions have recently become available. In a preliminary report from the American College of Cardiology–National Cardiovascular Data Registry (ACC-NCDR),255 17,287 PCI procedures involving bypass grafts were submitted from 313 institutions from January 2003 to April 2004. In 14,010 (81.0%) of these procedures a stent was used (BMS, DES, or both). Although DES patients had a lower adjusted mortality (0.6% versus 1.0%; P < 0.0001), BMS were still predominantly used in treatment of patients presenting with STEMI or high-risk lesions in that period.

Another report256 evaluated the efficacy of SES in 126 patients with arterial and venous bypass grafts based on data extracted from e-CYPHER registry. The cumulative 6-month MACE (Death, Q or Non-Q-Wave MI, Emergency CABG, TLR) was 6.3% (8/126) with TLR rate of 4.0% (5/126).

Additionally, a recent report257 from the e-CYPHER registry compared the outcomes of SES treatment in 14,068 patients with native coronary lesions (Native) to 248 with SVG lesions. SVG patients presented with worst risk profile: older patients, more male patients, more diabetics, more hypertension, more hyperlipidemia, and more restenotic lesions. Regarding procedural characteristics, direct stenting was more often applied in the SVG group (43.9% versus 33.2%; P = 0.0001), and the number of stents per lesion was similar in both groups (1.39 ± 0.7 versus 1.34 ± 0.6; P = 0.273). Reference vessel diameter was larger for the SVG group (3.0 ± 0.37 versus 2.86 ± 0.35; P = 0.0001), and lesion length was similar (17.1 ± 10 versus 17.2 ± 8.8; P = 0.782). At 6-month clinical follow-up TVR was 2.51% for Native group versus 0.81% in the SVG group, TLR 5 (2.51%) versus 132 (1.23%), Subacute/late thrombosis 1 (0.5%)/1 (0.5%) versus 67 (0.6%)/16 (0.15%), Total MACE 13 (6.53%) versus 361 (3.35%); P = 0.014. In this “real-world” experience, SES treatment of SVG lesions was associated with a low rate of 6-month TLR (2.5%). The higher MACE rates compared to the native group was explained by the higher TVR (non-TLR) rates. This still represents a striking improvement when compared with historical BMS data.

On the other hand, a report on data extracted from the American College of Cardiology National Cardiovascular Data Registry database and adjunct DES data registry on utilization, angiographic, and outcome data from 203 patients (85 BMS, 118 DES), 217 procedures (88 BMS, 129 DES), and 346 lesions (118 BMS, 228 DES) involving implantation of a stent in a SVG suggested that there is no indication that there is a lower restenosis rate in the long term with DES.258 In this report patients with both BMS and DES, or with an additional vessel dilated, were excluded. The distribution of SVG lesion locations was 7% in the aortic anastomosis, 74% in the body of the graft, and 19% in the distal anastomosis. There was no difference in patient demographics, graft age, use of embolic protection, glycoprotein IIb/IIIa inhibitor usage, or incidence of acute MI. Post-TIMI III flow was similar between the two groups (99.8% BMS, 99.6% DES). There was no difference in in-hospital coronary artery bypass (0% BMS, 0.8% DES) or death (1.1% BMS, 0% DES). TLR at 6 months was 6% BMS versus 8% DES (P = NS).

Multivessel Disease 

Insight on the impact of DES in multivessel disease patients comes from the Second Arterial Revascularization Therapy Study (ARTS II) trial,259 which is a registry involving a nonrandomized comparison of consecutive multivessel patients treated with CYPHER stents (n = 607), compared to the original surgical coronary artery bypass arm of ARTS I trial260 (a randomized trial comparing surgical coronary artery bypass (CABG) (n = 605) versus multivessel bare-metal stent implantation (n = 600)) and using similar inclusion and exclusion criteria to ARTS I by the same clinical sites involved in ARTS I separated in time and thus will allow a comparison of a more contemporary DES cohort to the bypass group from the initial trial. Similar to ARTS I, ARTS II is evaluating patients who had not undergone bypass surgery or angioplasty with stable angina, unstable angina or silent ischemia, AND de novo lesions in different vessels and territories that were amenable to stent implantation. The primary endpoint is MACE-free survival at 1 year: death, cerebrovascular event, nonfatal myocardial infarction, repeat revascularization (either percutaneous or surgical).

Baseline differences were found when comparing ARTS II to ARTS I. In ARTS II, patients were more frequently diabetic (26.2% versus 18.2%) and they more often had three-vessel disease present (54% versus 28%). Lesions were more frequently complex (13.9% versus 7.5% for Type C lesions). More stents were implanted per patient than in ARTS I (3.7 versus 2.8) and longer lengths of stent were implanted (73 mm versus 48 mm). Data at 6 months261 showed that the incidence of major adverse cardiac complications was lower in ARTS II registry patients when compared to the patients enrolled in the randomized portion of the ARTS I trial (6.4, 9.0, and 20.0% for ARTS II, ARTS I CABG, and ARTS I PCI groups). Freedom from revascularization was more common in the ARTS II registry (94.5%) compared to patients undergoing PCI in ARTS I (84.7%)(P < 0.0001) and similar to CABG patients in ARTS I (97.3%). The incidence of stent thrombosis was low in these patients (0.8% in ARTS II versus 2.8% in ARTS I).

More recently data at 1-year follow-up became available.262 At 1 year, there was no difference in the incidence of MACE comparing the ARTS II SES registry patients with the CABG randomized patients in the ARTS I trial (10.4% versus 11.6%, P = 0.46) but MACE was lower in ARTS II SES compared with BMS patients in ARTS I (26.5%). There was also no difference in 1-year death (1.0% for ARTS II registry SES patients versus 2.7% for ARTS I CABG patients), cerebrovascular events (0.8% versus 1.8%), MI (1.2% versus 3.5%), or revascularization with CABG (2.0% versus 0.7%) or PCI (5.4% versus 3.0%).

Thus, among patients suitable for either CABG or PCI, this registry experience demonstrates that Cypher Sirolimus eluting stent placement substantially reduces the need for repeat revascularization. In patients undergoing multivessel intervention, the rates of repeat revascularization are substantially reduced when compared to patients undergoing bare-metal stent implantation and approach those of patients undergoing CABG in the ARTS I trial. These improved outcomes are striking despite the increased length and complexity of lesions treated and stents implanted in this group. Clinical adverse events were rare in this group and, despite a larger average number of stents implanted (3.7) and longer average stent length implanted (73 mm), rates of stent thrombosis were rare. This registry further confirms the impressive reductions in repeat revascularization rates associated with implantation of DES in multivessel disease. Despite attempts to control for confounding variables by using similar inclusion and exclusion criteria between the registry and the much earlier trial, unidentified modifiers may affect these results. Conclusions regarding superiority of a given therapy would need to be addressed in randomized trials.

In addition, there are observational series data from single sites available. A report of 155 patients undergoing multivessel real-world CYPHER stenting observed very low rates of MACE to 30 days, and low rates of death or myocardial infarction at 6 months.263 Interestingly, the rate of TVR was 18% in this small, complex cohort of patients. Obviously, the most reliable comparison of multivessel DES versus bypass surgery will require a randomized trial. The Future Revascularization Evaluation in Patients with Diabetes Mellitus: Optimal Management of Multivessel Disease (FREEDOM) trial will be performed in the United States and will enroll patients with diabetes and multivessel disease in a head-to-head comparison of PCI with DES versus bypass surgery. FREEDOM will be an incredibly important study in a complex group of patients, which will provide data to answer the question of DES efficacy in multivessel disease.

In-Stent Restenosis 

ISR is a novel pathobiologic process, histologically distinct from restenosis after balloon angioplasty and comprised largely of neointima formation. As percutaneous coronary intervention increasingly involves the use of stents, ISR is also becoming correspondingly more frequent. In-stent restenosis has been classified on the basis of length of restenosis in relation to stented length.264 The following four types of ISR have been defined: (1) focal (≤10 mm length); (2) diffuse (ISR >10 mm within the stent); (3) proliferative (ISR >10 mm extending outside the stent); and (4) occlusive ISR. Type I has been further subdivided into types IA to ID based on the site of focal ISR in relation to the stent. An additional type of ISR has been proposed, that of “aggressive ISR,” defined as ISR that is longer and/or more severe than the original lesion.265 This type is noteworthy in that the clinical course is not benign, with patients more likely to have more severe symptoms and higher rates of myocardial infarction.

The main problem associated with ISR is the difficulty of finding appropriate treatment modalities to reduce the excessive risk of recurrence. Plain balloon angioplasty has been the first-line treatment option for in-stent restenosis, yet its results have often been disappointing with a recurrence rate above 40%.266 Repeated use of bare stents appears to further exacerbate the risk of recurrence. Alternative interventional options, including rotational atherectomy, excimer laser angioplasty, directional coronary atherectomy, and cutting balloon, have not provided additional benefits.267, 268 Although brachytherapy is currently the treatment approach most supported by evidence for in-stent restenosis,269 the complexity of its application, concerns about the prolonged risk of vessel occlusion,267 and the decrease in benefit over time270 have limited use of this strategy.

The frequent failure of local approaches in the treatment of in-stent restenosis has stimulated major interest in the use of systemic therapies. In the Oral Sirolimus to Inhibit Recurrent In-Stent Stenosis (OSIRIS) trial271 a strategy based on a 2-day pretreatment with a high dose of sirolimus followed by a usual maintenance dose for a week after balloon angioplasty recently resulted in a significant reduction of recurrences in patients with in-stent restenosis (22.1% versus 42.2%). However, the need for a 2-day pretreatment and, consequently, the impossibility to perform “ad-hoc” interventions will most probably prevent the widespread use of this approach.

There has been an increasing interest in applying DES to ISR. Initial clinical data on DES for ISR has come from small feasibility studies. The first reported experience includes a combined series of sirolimus for ISR from Rotterdam, The Netherlands and Sao Paulo, Brazil.251 In this feasibility study, 41 patients (25 from Sao Paulo, 16 from Rotterdam) with ISR were treated with the sirolimus stent. Vessel sizes ranged from 2.5 to 3.5 mm and included one or two 18-mm CYPHER stents. Aspirin and clopidogrel were used for 60 days. ISR lesion patterns included focal in 40%, diffuse in 32%, and proliferative in 28%. The angiographic late loss was shown to be 0.17 mm in the Sao Paulo patients and 0.51 mm (owing to two total occlusions) in the Rotterdam patients. At 1 year, restenosis occurred in three patients (7.3%), with one MI and two deaths. These early results suggested that acute complications might arise in very complex diffuse ISR lesions (such as those with previous brachytherapy) but that very low long-term restenosis is possible in selected patients. As mentioned earlier, the TAXUS III feasibility trial examined the TAXUS stent for ISR in a total of 30 patients. MACE occurred at 1 year in 28.6%, with TLR in 21.4%. MI occurred in 3.6%. There was no stent thrombosis or mortality in this series.

Additional data on ISR comes from the Treatment of Patients With an In-stent Restenotic Native Coronary Artery Lesion (TROPICAL) Study.272 TROPICAL is a multicenter, nonrandomized study of the CYPHER stent in the treatment of patients with ISR. The objective of the study was to compare the CYPHER stent in ISR of native coronary lesions with the historical group of the combined GAMMA I and II Trials. A total of 162 patients with ISR were consecutively treated with the CYPHER stent. Angiographic follow-up was planned in all patients. The primary endpoint was in-lesion late loss and was shown to be 0.08 mm for the CYPHER-treated patients, compared with 0.68 mm in the historical control group from the GAMMA I/II Trials. The clinical outcomes at 180 days for the TROPICAL study showed 4.9% MACE, 0.6% deaths, 1.8% MI, 2.5% TLR, and 0.6% late stent thrombosis. This compared favorably with the historical control group of the GAMMA I/II Trials, particularly in relation to the rate of TLR 2.5% versus 14%, P < 0.001). This study demonstrated that sirolimus-eluting stents were highly effective in ISR, with an average in-lesion late loss of less than 0.1 mm, which translated into a very low restenosis rate when compared with historical results of brachytherapy.

Further data on the safety and efficacy of SES in the treatment ISR comes from the US internet e-CYPHER Registry.273 Of 2067 patients treated with SES, 142 had placement of SES for treatment of ISR. Demographic, procedural, and follow-up data were collected: mean patient age was 64 years; 47% of the patients had prior myocardial infarction; 35% were diabetic; 37% had triple vessel disease; 23% had ostial disease; and 28% of the lesions were class C. In- and out-of-hospital to 30 and 180 days events were as follows: MACE (death, MI, emergent CABG, TLR) 4.9% versus 16.9%; death 0% versus 2.1%; target vessel failure 4.9% versus 18.3%; TLR 4.9% versus 15.1%; stent thrombosis 3.5% versus 3.5%, respectively.

Another trial, the ISAR-DESIRE trial (Intracoronary Stenting and Angiographic Results: Drug-Eluting Stents for In-Stent Restenosis)274 evaluated the use of DES compared with balloon angioplasty for treatment of ISR and compared the sirolimus-eluting stent with the paclitaxel-eluting stent in that setting. This was a randomized, open-label, active-controlled trial conducted among 300 patients with angiographically significant ISR. After pretreatment with 600 mg of clopidogrel for at least 2 hours before intervention, all patients were randomly assigned to one of three treatment groups: sirolimus stent, paclitaxel stent, or PTCA (100 patients in each group). The primary endpoint was restenosis in-segment at 6-month angiography, defined as stenosis ≥50%, while the secondary endpoints were net lumen gain at 6-month angiography and clinical event at 9 months, including death, MI, and TVR. Postprocedure minimum lumen diameter was smaller in the PTCA group (2.06 mm) than the sirolimus DES group (2.5 mm) or paclitaxel DES group (2.6 mm; P < 0.001). Almost half of the interventions were performed on the left anterior descending artery (45%). Repeat angiography at 6 months was performed in 92% of patients. The primary endpoint of in-segment restenosis on 6-month angiography was smaller in the DES groups than the PTCA groups (14% sirolimus DES, 22% paclitaxel DES, 45% PTCA; P < 0.001 for sirolimus DES versus angioplasty, P = 0.001 for paclitaxel DES versus PTCA). Likewise, net lumen gain was larger in the DES groups than the PTCA groups (1.11 mm sirolimus DES, 0.93 mm paclitaxel DES, 0.46 mm PTCA; P < 0.001 for sirolimus DES versus PTCA, P < 0.001 for paclitaxel DES versus PTCA). Clinical target vessel revascularization occurred less often in the DES groups than the PTCA groups (8% sirolimus DES, 19% paclitaxel DES, 33% PTCA; P < 0.001 for sirolimus DES versus PTCA, P = 0.024 for paclitaxel DES versus PTCA). There was no difference in mortality (2% sirolimus DES, 1% paclitaxel DES, 2% PTCA) or death or MI by 9 months (3% sirolimus DES, 3% paclitaxel DES, 2% PTCA). For the analysis comparing the two DES, in-segment late lumen loss was lower in the sirolimus DES group compared with the paclitaxel DES group (0.45 mm versus 0.66 mm, P = 0.02), as was in-stent late lumen loss (0.21 mm versus 0.48 mm, P = 0.006). Angiographic restenosis was directionally but nonsignificantly lower in the sirolimus DES group compared with the paclitaxel DES group (14% versus 22%, P = 0.19). Target vessel revascularization occurred less frequently in the sirolimus DES group compared with the paclitaxel DES group (8% versus 19%, P = 0.02). The ISAR-DESIRE trial demonstrated that a strategy based on sirolimus- or paclitaxel-eluting stents is superior to conventional balloon angioplasty for the prevention of recurrent restenosis in patients with in-stent restenosis. The lack of a brachytherapy group as a control in this study is a major limitation. The study also suggested that in this high-risk subset of patients SESs might be superior to PESs.

Moreover, a recent study275 examined the use of DES for failed Intracoronary Radiation Therapy (IRT) patients and compared their outcome to repeat IRT. A cohort of 65 patients who failed IRT for the treatment of ISR was studied. Of these, 14 patients were treated with DES with either Taxus or Cypher and 51 with repeat radiation using either beta or gamma sources. All patients prescribed at least 12 months of clopidogrel postprocedure. The baseline characteristics of the patients were similar among the treated groups. At present mean follow-up available on the DES group is 8.9 ± 3.2, while the repeat radiation completed 9-months follow-up. Overall DES was resulted with higher events rate when compared to VBT at 9-months follow-up (MACE 50.0% for DES group versus 21.6% for VBT group, P = 0.009).

Presently, there is an ongoing randomized trial of sirolimus-eluting stents versus brachytherapy for ISR, entitled the SISR Trial (Sirolimus-Eluting BX Velocity Balloon Expandable Stent versus Intracoronary Brachytherapy in the Treatment of Patients with In-Stent Restenotic Coronary Artery Lesions). This is a multicenter (26 sites, 400 patients), randomized study (2:1) of CYPHER stent versus brachytherapy. The primary endpoint is target vessel failure at 9 months postprocedure, and efficacy results are expected by mid-2005. Despite the lack of randomized trial data at this time, the initial results of the TROPICAL and the ISAR-DESIRE studies as well as emerging data from other registries suggest very favorable results of DES for the treatment of ISR. Clearly, however, additional data are needed from larger registries and prospective clinical series in the future.

Sidney C. Smith, Jr.: This is an excellent review of the evolving evidence for use of DES for ISR. Increasingly, DES is becoming a front-line therapy for ISR.

Calcified Lesions 

Coronary stent implantation in severely calcified lesions remains a challenge in interventional practice owing to difficulties in stent delivery and expansion. In such patients, lesion preparation with high-pressure balloon inflation might occasionally succeed but is often insufficient to overcome vessel-wall resistance. Rotational atherectomy (RA) has proven to be the preferred strategy to ablate calcified plaque, but despite the high procedural success rate, late stenosis recurrence remains high when it is used as a stand-alone treatment.276 The use of DES in these patients has therefore an intriguing potential for prevention of restenosis. Nonetheless, data on DES use in calcified lesions are limited.

In a substudy of TAXUS IV, the impact of culprit lesion calcification was evaluated.277 By core lab analysis, 247 lesions (19%) were moderately or severely calcified. TLR at 1 year was reduced by 56% in TAXUS-randomized patients with calcified lesions (11.9% versus 5.1%, P = 0.09), and by 75% in TAXUS-randomized patients with noncalcified lesions (15.7% versus 4.3%, P < 0.0001). By interaction testing, the efficacy of the paclitaxel-eluting stent in reducing 1-year TLR was similar in calcified and noncalcified lesions (P = 0.30). The 9-month angiographic follow-up data showed a nonsignificant trend was present (P = 0.10) for interaction between the presence of calcium and treatment assignment on the occurrence of analysis segment restenosis.

Additionally, randomized trials of drug-eluting stents have excluded patients with calcified lesions requiring RA prior to stent implantation. A recent report278 however analyzed all PCI patients undergoing RA for de novo calcified lesions in native vessels, followed by BMS from May 2002 to April 2003 (n = 284) and compared to patients undergoing RA followed by DES from May 2003 to April 2004 (n = 246) at one institution. Baseline clinical and angiographic characteristics (especially prior MI, diabetes mellitus, LVEF, multivessel disease, LAD lesion) were similar in two groups; GP IIb/IIIa inhibitors were used in >80% of cases. In the DES group, stenting the area that underwent balloon dilatation post-RA was routinely done. Procedural success (<30% diameter obstruction post) was 98.2% versus 99.2% for RA ; BMS versus RA ; DES, respectively. Clinical success (procedural success in the absence of Q-wave MI, urgent revascularization, or death) was 96.8% versus 97.7%, 30-day MACE (MI with CK-MB >3×, urgent revascularization or death) was 6.7% versus 5.3% (P = 0.62) and target TLR was 15.5% versus 5.3% (P ≤ 0.01). Multivariate predictors of TLR were diabetes (OR 3.2; 95% CI 2.8 to 3.8), lesion length (OR 2.4; 95% CI 1.2 to 3.7), and use of DES (OR 0.5; 95% CI 0.2 to 0.9). These data demonstrate that DES is superior to BMS in the calcified coronary lesions even following RA.

More Complex Lesions 

Further data on DES in complex lesions comes from the recently presented SCANDSTENT trial (Stenting of Coronary Arteries in Non-Stress/Benestent Disease).279 In this trial 322 patients with angina pectoris, non-ST-segment-elevation MI, and complex lesions were randomized to PCI with the SES or a BMS. Complex lesions were defined as those with occlusions longer than 15 mm, lesions with significant side branches, ostial lesions, and angulated lesions. Patients with an MI within 3 days of PCI were excluded from randomization. Six-month angiographic outcomes demonstrated improvements in minimal lumen diameter in patients treated with the SES (2.48 mm for SES versus 1.63 mm for BMS, P < 0.0001) and significant reductions in diameter stenosis (19.3 mm versus 43.8 mm, P < 0.0001). There was also a dramatic reduction in binary restenosis (2.0% versus 31.9%, P < 0.0001). At 12-months clinical follow-up there were no differences in the rates of death, MI, or stent thrombosis, but patients treated with the Cypher stent had significantly lower rates of TLR compared with the bare-metal-stent group (2.4% versus 29.8%, P < 0.0001).

Eric R. Bates: This encyclopedic review of DES in complex lesions dramatically demonstrates the consistent benefit of DES over BMS.

Diabetes Mellitus 

Several clinical, angiographic, and biological features are associated with CAD in diabetic patients that constitute potential risk factors and confer a poor prognosis.280 Endothelial dysfunction,281, 282, 283, 284, 285, 286, 287 platelet and coagulation abnormalities,288, 289, 290, 291, 292, 293, 294, 295 and metabolic disorders295, 296, 297, 298, 299, 300, 301, 302, 303, 304, 305, 306, 307 associated with diabetes mellitus (DM) play a major role in the accelerated atherosclerotic process and in the formation of coronary thrombosis, and they contribute substantially to the complex healing process after arterial wall injury. Angiographic features related particularly to diffuse and distal coronary disease may lead to incomplete revascularization or increase the risk of surgical or percutaneous intervention in these patients.308, 309, 310, 311, 312, 313, 314, 315, 316, 317, 318, 319, 320, 321 The risk of morbidity and mortality is also increased by several unfavorable clinical characteristics and comorbidities that are more common in diabetic patients.280 These comorbidities include hypertension, dyslipidemia, systolic and diastolic heart failure, autonomic dysfunction, peripheral vascular disease, cerebrovascular disease, microvascular disease, a prothrombotic state, and nephropathy.

Although coronary stent deployment was observed to improve both angiographic and clinical late outcomes in diabetic cohorts when compared with standard balloon angioplasty, late restenosis and the requirement for revascularization following coronary stent deployment remains significantly more common in diabetics versus nondiabetics.322, 323, 324, 325, 326, 327, 328, 329

DES offer a potential solution to the higher rates of repeat revascularization and restenosis observed following conventional stent deployment in diabetics. In the initial experience from the RAVEL trial with the SES stent, preferential benefit for reduction in binary (>50%) angiographic restenosis was observed in the diabetic cohort (0% versus 41.7%, respectively; P = 0.002) for the CYPHER SES versus the Bx Velocity BMS.135 These salutary results were confirmed in the diabetic cohort (n = 279) of the SIRIUS trial, which demonstrated a reduction in binary angiographic restenosis (in-lesion) for the SES versus the BMS (17.6% versus 50.5%, respectively; P < .001). Similarly, TLR at 1-year follow-up was reduced by the SES versus the BMS (8.4% versus 26.5%, respectively; P = 0.0002).139 Indeed, SES stent deployment (versus BMS) was associated with a similar magnitude of reduction in 12-month TLR (70 to 80%) irrespective of lesion length, reference vessel size, or diabetic status. Presently, a wealth of clinical and angiographic data from randomized clinical trials as well as postmarket surveillance registries offer support for these initial observations of durable benefit following SES deployment in diabetic patients.

Diabetic patients enrolled in the SIRIUS trial demonstrated significant reductions in both angiographic (binary restenosis, late coronary lumen loss) and clinical (TLR) restenosis at late follow-up in favor of the SES versus BMS.139 In the SIRIUS trial, although diabetics had a higher incidence of adverse outcomes following coronary stent deployment versus nondiabetics, substantial benefit was observed in diabetic patients treated with the SES compared with BMS. Indeed, no difference in MACE-free survival to 9 months was observed by diabetic status (P = 0.30) in patients treated with the SES. Conversely, in patients who were treated with BMS, the presence of diabetes was associated with a significant (P = 0.018) reduction in MACE-free survival. As noted previously, the magnitude of reduction in clinical coronary restenosis (TLR) by the SES was not influenced by lesion length, reference vessel diameter, or diabetic status. This salutary effect of the SES on clinical restenosis in diabetic patients has been consistent across clinical trials as well as postmarket surveillance registries. In the cumulative clinical experience involving diabetic patients treated with SES, TLR was observed in only 0 to 7.0% and the reduction in TLR conferred by SES (versus BMS) ranged from 70 to 100%.

These observations are supported by from the Direct Stenting Using the Sirolimus-Eluting Stent (DIRECT) trial, in which the strategy of direct stenting (versus predilatation) with the SES stent was employed.330 Patients enrolled into the DIRECT trial were compared with the historical cohort of patients treated with the SES from the SIRIUS trial. Quantitative coronary angiography was performed at 8 months following stent deployment in both the DIRECT and the SIRIUS trials. The strategy of direct stenting versus predilatation was associated with lower rates of binary (in-lesion) restenosis in nondiabetics (5.3% versus 6.0%; P = 1.00), non-insulin-dependent diabetics (10.3% versus 13.8%; P = 0.76), and, particularly, insulin-dependent diabetics (0% versus 35.0%, P = 0.03). The DIRECT trial suggests the importance of operator technique for deploying the CYPHER stent, especially in diabetic patients. These data support the concept that minimizing the extent of endoluminal injury beyond the confines of the drug-delivery platform by eliminating the predilatation process (when possible) has clinical/angiographic benefit.

The DIABETES study331 was the first randomized study of drug-eluting stent implantation that including large numbers of diabetic patients who underwent angiographic follow-up. In this multicenter trial, the investigators randomized 160 patients (221 lesions) with diabetes mellitus (oral hypoglycemic or insulin-treated) and de novo native coronary artery lesions to the SES versus its BMS counterpart. The two groups were matched for baseline clinical and angiographic characteristics, and one-third of the patients were insulin-treated in both groups. The primary endpoint was in segment late lumen loss as assessed by QCA at 9-month angiographic follow-up. Clinical follow-up was scheduled at 1, 9, and 12 months. An average of 1.4 lesions was treated per patient and 1.7 stents were implanted per patient. At 9-month angiographic follow-up, the angiographic restenosis rate was reduced by over 70% in the sirolimus stent arm (7.7% versus 33.0%; P < 0.0001). Similarly, late lumen loss was reduced by more than 80% in the sirolimus stent group (0.08 ± 0.4 mm versus 0.44 ± 0.5 mm; P < 0.0001). At 9-month follow-up there were no differences in the clinical endpoints of death or MI. There was no late stent thrombosis. Target lesion revascularization was more frequent in diabetic patients who received BMS compared to those who received the Cypher sirolimus eluting stent (31.3% (n = 25) versus 7.5% (n = 6); P < 0.0001). Angiographic analyses showed most lesions to be located in proximal or mid vessel segments and the reference vessel diameter to be 2.34 ± 0.6 mm. Lesion length was 15.0 ± 8.1 mm and an average stent length was 22 ± 10 mm. This beneficial effect was extended to 1-year follow-up332 (TLR: 11.1% in SES group versus 41.3% in BMS group, P < 0.0001) (data for 117 patients (73%) who have completed 1-year follow-up). There were no differences between groups (SES versus BMS) in death (1.9% versus 3.2%; P = NS) or myocardial infarction rates (3.8% versus 6.5%; P = NS) at 1 year. The need for new revascularization due to progression of atherosclerosis in nontarget segments was higher, but not significant, in the SES group (11.1% in SES group versus 3.2% in BMS group; P = 0.14).

In a subset of 165 lesions (SES = 85, BMS = 80) serial IVUS analysis (automated motorized pullback at a speed 0.5 mm/sec) was performed after stent implantation and at 9-month follow-up.333 Vessel, luminal, and stent mean areas and volumes were evaluated at proximal and distal edges and within the stent. To date, 69 lesions have been analyzed by an independent core lab. At 9-months follow-up, a significant reduction in mean in-stent luminal area was evidenced in the BMS group as compared with SES group (8.2 ± 2.8 mm2 versus 6.3 ± 2.3 mm2; P = 0.003) due to a significant increase in mean neointimal hyperplasia in the BMS group (2.3 ± 1.6 mm2 versus 0.3 ± 0.8 mm2; P < 0.001). At distal edges, mean lumen area was significantly reduced in the BMS group (delta lumen: −1.3 ± 2.1 mm2, P = 0.004), whereas it showed a trend to increase in the SES group (delta lumen: 0.9 ± 2.3 mm2, P = 0.07). At proximal edges no significant changes were observed in either group. Acquired stent malapposition was observed more often in the SES group (P = 0.03). Thus, this study demonstrated a striking reduction in angiographic parameters of restenosis and the need for repeat revascularization. In addition, there were no differences based on type of diabetes (non-insulin-requiring versus insulin-requiring). Clinical consequences of acquired malapposition of those stents remain to be elucidated. In addition, larger patient trials will be needed to formally conclude the exact role of SES for the management of insulin-dependent diabetic patients.

Additionally, preliminary results are available from the RECORD334 trial (Rapamycin and Paclitaxel Drug-Eluting Stents in Diabetic Patients), another randomized multicenter trial in which 60 patients with diabetes (mean age 67 ± 10 year old, 55% male) with coronary artery disease were randomly assigned to PCI with RES or PES. Forty-five percent of patients were women, and all were high-risk patients (70% acute coronary syndrome, 64% HTN, 62% dyslipidemia, 22% previous MI); 1.4 DES were used per patient with a 98% success rate by intention to treat. There were no early episodes of acute or subacute thrombosis of any stented artery. There were no MACE or major in-hospital complications at a mean follow-up of 6.1 ± 2.7 months. One patient was readmitted for angina. Stent delivery to target lesion was significantly better for PES than for SES (100% versus 87%; P < 0.05).

Moreover, in the Strategic Transcatheter Evaluation of New Therapies (STENT) group registry,335 all diabetic patients were included from May 2003 to June 2004 and stratified according to insulin-dependence. Primary endpoints of analysis included MACE (death, MI, and TVR) and subacute thrombosis at 3- and 9-month follow-up. Three-month clinical follow-up has been completed for a total of 1339 DES procedures, of which 399 (30%) were DM patients (69% NIDDM, 31% IDDM). For DM patients with DES procedures, the average age was 63, with coronary artery bypass surgery in 15%, acute coronary syndrome in 71%, ST-elevation myocardial infarction in 8%, urgent or emergent PCI in 75%, and mean ejection fraction in 51%. Patients received 1.4 DES per procedure with an average total stent length of 27.4 mm. In-hospital procedural success was 98%. Cumulative MACE at 90 days was 3.8% for all DM, including 4.9% for IDDM and 3.3% for NIDDM. TVR at 90 days occurred in 1.0% for all DM, with 0.8% for IDDM and 1.1% for NIDDM. In-stent subacute thrombosis was 1.3% (n = 5) for DM. Thus diabetic patients receiving DES appear to have favorable early clinical outcomes in this diverse “real-world” population.

Further “real-world” clinical practice experience with SES comes from the e-CYPHER registry, which has been updated in 3438 patients.336 Despite the fact that diabetics were older and had more associated risk factors, including multivessel disease and number of stents deployed per patient as well as smaller reference vessel diameters and longer lesion lengths versus nondiabetic patients enrolled in this registry, clinical TLR to 6-months follow-up was observed in only 1.4% of diabetics and 0.9% of nondiabetics (P = not significant). Total MACE to 6 months were observed in only 4.2% of 2716 diabetic patients followed up for 6 months post-SES deployment. These data are similar to the preliminary real-world experience from the BRIDGE registry of SES stenting in diabetic patients. In BRIDGE, 6-month clinical follow-up in 547 diabetic patients with average reference vessel diameter of 2.80 mm and lesion length of 18.4 mm demonstrated TLR in 3.5% and MACE in 5.9%.337 The expanding database for SES use in diabetics suggests safe and durable clinical benefit in the form of low MACE and TLR rates as well as minimal late coronary lumen loss (0.20 to 0.30 mm) by quantitative coronary angiography.

Furthermore a single center experience338 of 1407 patients (1618 lesions) underwent PCI with SES for treatment of various lesions, including complex lesions and vein grafts, included 160 patients (11.4%) with IDDM, 332 (23.6%) with NIDDM, and the remaining 915 (65.1%) nondiabetics (Non-DM). Baseline characteristics were similar except that a higher number of patients in the IDM group were women and diabetic patients had higher body mass indexes and concomitant traditional coronary risk factors. IDDM patients had higher history of renal insufficiency and in-hospital complications including death and CABG. Six-month follow-up showed higher incidence of TLR-MACE and combined rate of death and Q-wave MIs (P = 0.03) in diabetic patients (IDDM 5.4%, NIDDM 6.3%, Non-DM 2.7%).

In addition, the PES has also demonstrated efficacy for the treatment of diabetic patients. Late (9-month) quantitative angiographic follow-up in the TAXUS IV trial, in which patients were treated with either the PES or the BMS, demonstrated a significant reduction in binary restenosis in the analysis segment from 34.5 to 6.4% (P = 0.0001) in all medically treated diabetic patients and from 42.9 to 7.7% (P = 0.0065) in insulin-requiring diabetics.157 Furthermore, clinical restenosis as reflected by the requirement for TVR to 1-year follow-up was reduced in both insulin-requiring (from 19.4 to 6.2%; P = 0.07) and oral medication-treated (21.6 to 7.9%; P = 0.005) diabetic patients following treatment with the express versus the TAXUS stent, respectively. The presence of diabetes mellitus was not an independent predictor for late TLR by multivariable analysis following PES deployment.339 The reduction in angiographic and clinical restenosis in PES-treated patients was associated with a lower incidence of late (not periprocedural) myocardial infarction as well.

Moreover, a meta-analysis340 evaluated the use of PES compared with BMS in diabetic patients who were enrolled in TAXUS II, IV, or VI trials. Patients were categorized as nondiabetic (n = 906 for bare stent group and n = 925 for paclitaxel-eluting stent group), diabetic patients (n = 242 for bare stent group and n = 216 for paclitaxel-eluting stent group), and insulin-dependent (n = 83 for bare stent group and n = 71 for paclitaxel-eluting stent group). Diabetes was defined as requiring medical therapy. Reference vessel diameter was smaller in diabetic patients compared with nondiabetic patients (2.69 mm versus 2.76 mm, P = 0.0002), while lesion length was longer (14.8 mm versus 13.9 mm, P = 0.02). Consequently, stent length was longer in diabetic patients (23.5 mm versus 22.1 mm, P = 0.008). Target lesion revascularization was lower in the paclitaxel-eluting stent group compared with bare-metal stent in nondiabetics (4.6% versus 14.2%, P < 0.0001), diabetics, and insulin-dependent patients. In-stent binary restenosis was also lower in the paclitaxel-eluting stent group compared with bare-metal stent in nondiabetics (4.9% versus 22.9%, P < 0.0001), diabetics (4.4% versus 32.9%), and insulin-dependent patients, as was in-segment binary restenosis. In-stent lesion length was shorter in the paclitaxel-eluting stent group compared with bare-metal stent in nondiabetics (0.36 mm versus 0.86 mm), diabetics (0.36 mm versus 1.03 mm), and insulin-dependent patients (0.33 mm versus 1.01 mm), as was percentage diameter stenosis (19.0% versus 36.2% for nondiabetics; 19.4% versus 41.8% for nondiabetics). This meta-analysis further supports the finding that paclitaxel-eluting stents were associated with improved angiographic outcomes compared with BMS in the cohort of diabetic patients.

Despite these findings, recent concerns have been raised whether DES are really superior to thin-strut BMS for preventing repeat revascularization in patients with diabetes.341 First, these studies and registry observations are inadequately powered to make this determination. Second, studies have used thick-strut stents, which are known to have high restenosis rates as controls. Third, low angiographic follow-up in some studies underestimates the true TVR rate because of the high incidence of silent ischemia in patients with diabetes. Moreover, the importance of medical therapy to achieve recommended glycemic control targets and management of usual risk factors in these patients cannot be overstated and represent important potential confounding factors in interpreting any data from these trials. No information regarding the level of glycemic control of these patients as well as their medical regimen (eg, number taking a statin) was given, and medical regimens were simplified into insulin-requiring versus non-insulin-requiring. The validity of subset analysis is uncertain given these uncontrolled-for variables.

ST Elevation Acute Myocardial Infarction 

Given the suspected delayed endothelial healing of DES compared with bare-metal stents, concerns have arisen regarding the risk of acute or subacute thrombosis of DES when implanted in the actively thrombotic environment of the STEMI lesion. In addition, STEMI patients have been categorically excluded from any of the randomized DES trials completed to date. Registries and clinical trials with STEMI data include the RESEARCH Registry, the e-CYPHER Registry, the STENT Group Registry, the TYPHOON Trial, and the DESCOVER Registry.251

The first of these to examine STEMI patients was the RESEARCH Registry from Rotterdam, The Netherlands. Within this patient population there were 120 patients in the pre-SES phase (October 2001 to April 2002) and 94 patients in the SES phase (April 2002 to October 2002) with STEMI.146 These patient populations were compared with respect to baseline clinical characteristics and long-term outcome. The cumulative incidence of death, MI, or TVR at 9 months was 15.8% in the pre-SES phase, compared with 8.5% in the SES phase (95% CI 0.2 to 1.0, P = 0.07). There was no increase in the incidence of subacute stent thrombosis during the SES compared with the pre-SES phase. Angiographic follow-up was performed in 60% of the SES patients and showed that for these STEMI lesions, averaging 16.9 mm in length and with a mean vessel diameter of 2.73 mm, the long-term late loss was minimal, with 0% binary restenosis. Although the sample size of 94 STEMI patients is rather small, these data from the RESEARCH Registry are highly favorable and indicate that routine SES implantation seemed to be safe for unselected patients with STEMI as compared with patients from the preceding 6 months with bare-metal stents.

Other emerging clinical data from Europe include results from the e-CYPHER Registry.342 Of 12,108 patients enrolled in this registry, 803 were treated in the setting of AMI (<72 hours). The infarct-related arteries were left anterior descending (52.6%), left circumflex (19.8%), left main (0.9%), and right (26.8%). Thrombus was visible in 36.9%. A mean of 1.34 SES was implanted per patient, with direct stenting in 27.6%. Prior to SES implantation, 413 patients (47%) were given thienopyridine. Preprocedure, TIMI flow-0/1 was present in 38.7% of patients and TIMI-2 in 14.3%. Postprocedure TIMI flow-3 was achieved in 97.2%. Follow-up is available for 665 patients at 6 months (83% of those eligible). In-hospital MACE rate was 1.0% (death 0.2%, recurrent MI 0.7%, TLR 0.3%). At 180 days the MACE rate was 5.3% (death, 3.5%, cardiac death 2.4%, MI 1.7%, TLR 1.7%). Stent thrombosis occurred in 10 patients (1.5%): 2 (0.3%) acute, 7 (1.1%) subacute, and 1 (0.2%) late.

The first US multicenter data on DES and STEMI come from the STENT Registry.343 STEMI as the primary indication for PCI accounted for 12% of all procedures. Clinical outcomes of STEMI patients from the STENT Registry for 412 patients with follow-up through March 2004 show excellent acute and 3-month outcomes, with no significant difference between DES and bare-metal stents for MACE or subacute stent thrombosis. The data also show that bare-metal stents are being used in a higher percentage of STEMI patients than DES during this period, and it is too early to determine any difference in rates of TVR.

Another “Real World” experience comes from a recent report344 that compared 312 consecutive STEMI who received either an SES159 or a BMS153 with a total of 223 (BMS) and 194 (SES) stents successfully deployed. Patients with SES had a 0.6% in-hospital and 30-day mortality, compared to a 5.2% (in-hospital and 30-day, P = 0.016) mortality rate in the BMS patients. The 6-month mortality rates were 2.1% (SES) and 10.2% (BMS) (P = 0.002). The causes of death were as follows: SES, n = 3; one in-hospital (anoxic brain injury (ABI) after out of hospital cardiac arrest), two heart failure deaths at 6 months; BMS, n = 15; eight in-hospital (four ABI, three cardiogenic shock, one noncardiac), seven after discharge (two noncardiac, five sudden death). At 6 months, patients receiving BMS were more likely to have repeat, unplanned revascularization (1.4% versus 7.8%, P = 0.005) and reinfarction (0.6% versus 4.1%, P = 0.05) than patients receiving SES. At 6 months, the overall cardiac endpoint of death, reinfarction, and revascularization was 3.4% SES versus 16.3% BMS (P < 0.0001). No subacute thrombosis occurred in the SES group at 6 months.

Two important randomized clinical trials are underway for DES in STEMI. The TYPHOON trial is a multicenter (44 sites), prospective, single-blind, randomized study of the CYPHER stent versus bare-metal stent in STEMI. The plan is to enroll 700 STEMI patients in Europe with the primary endpoint of TVF at 1-year postprocedure. As of May 2004, 556 of the 700 patients were enrolled; 6-month data will be presented at the 2005 meeting of the American College of Cardiology, and 12-month data will be available at the American Heart Association 2005 meeting. The HORIZONS Trial is a 3400 STEMI patient trial of the TAXUS stent versus bare-metal stent. This trial will also examine bivalirudin versus unfractionated heparin plus a glycoprotein IIb/IIIa inhibitor, as well as the impact of complete revascularization (immediate stenting of all diseased vessels) versus staged stenting.

Eric R. Bates: Other high-risk subsets requiring study because of disappointing results with BMS include renal dialysis patients and cardiac transplant patients with transplant vasculopathy.

FUTURE I Feasibility Study 

The FUTURE I trial351 was the first-in-man clinical trial with the Biosensor everolimus-eluting stent (EES), which evaluated both safety and feasibility of this stent design in treatment of de novo coronary lesions. FUTURE I was a prospective, single-center, single-blinded, randomized trial including 27 and 15 patients with native de novo coronary lesions allocated for EES, coated with a bioabsorbable polymer, and BMS (control) groups, respectively. Patients were included if angiography showed lesion length less than 18 mm, diameter stenosis between 50 and 99%, and vessel diameter between 2.75 and 4.0 mm. Patients were excluded if they had diabetes mellitus, an acute myocardial infarction within the 4 weeks prior to intervention, in-stent restenosis, or a left ventricular ejection fraction less than 30%. The primary endpoint was 30-day survival, free from MACE. MACE was defined as death from any cause, Q-wave MI, target vessel revascularization, or stent thrombosis. Clinical evaluation was conducted at 1, 6, and 12 months after stent implantation. Coronary angiographic and intravascular ultrasound imaging was performed before and after stent implantation and at 6-month follow-up visit. Baseline demographics and lesion characteristics were similar between both study groups. At 30 days after stent implantation, there was no incidence of MACE or stent thrombosis in either cohort, indicating a comparable safety profile for the EES relative to the uncoated control stent. At 6 months, the MACE rate was 7.7% (two events in 26 patients) for the everolimus group compared with 7.7% (one event/13 patients) in the control group. The two MACE in the everolimus group were a pulmonary disease-related noncardiac death and a target lesion revascularization due to in-segment restenosis at the distal margin of the study stent. There were no additional MACE events between 6- and 12-month follow-up in either group. Follow-up angiography was performed in 25 everolimus patients (93%) and 11 control patients (73%) at 6 months. All quantitative angiographic indices (minimum lumen diameter, % diameter stenosis, lumen loss) were significantly (P < 0.001) improved in the everolimus stent group, compared with the control. The binary in-stent restenosis rate was 0.0% in the EES group versus 9.1% in the control. In-segment restenosis rates were 4.0% (1/25) versus 18.2% (2/11), respectively. In-stent late loss decreased from 0.85 mm (BMS) to 0.11 mm (EES). Intravascular ultrasound examination at 6 months revealed a significant reduction of percentage neointimal volume in the everolimus group (2.9 ± 1.9) compared with the control group (22.4 ± 9.4). In addition, there were no late acquired incomplete stent appositions in either group. With these results, an acceptable safety profile was established for the EES with a biodegradable coating system, with a very low MACE rate up to 6 months postimplantation. With all patients treated for 3 months with clopidogrel, no stent thromboses were reported in either group. The everolimus stent group demonstrated significant and concordant improvements in the sensitive IVUS and QCA parameters, with an 88% reduction of in-stent late loss in the EES group and an 87% reduction of percentage neointimal volume. Moreover, the in-stent late loss of 0.11 mm suggests that the dosage as well as the release pattern of everolimus disrupts the restenotic cascade, while allowing sufficient neointimal growth to promote healing and avoid late stent thrombosis.

FUTURE II Feasibility Study 

To extend these results, the multicenter FUTURE II352 study was conducted, which included patients with diabetes. FUTURE II was a three-center study with a total of 64 patients randomized in a single-blind 1:2 fashion (21 patients in the everolimus group; 43 patients in the control group). The purpose of the “reverse” randomization was to equalize the total numbers of active and control arms included in both the FUTURE I and the FUTURE II studies. Baseline characteristics of both groups in FUTURE II were similar, with diabetics representing 23.8% (everolimus group) and 27.9% (control) of patients. At 30-day follow-up, there were no adverse cardiac events in the everolimus group versus one event in the control (2.3%; non-Q-wave MI). At 6 months post-stent implantation, the MACE rate was 4.8% versus 17.5% in the everolimus and control groups, respectively. Only one target lesion revascularization was observed in the everolimus stent group, due to a proximal edge (in-segment) restenosis. The 6-month binary restenosis rate was 0% (in-stent) and 4.8% (in-segment) for EES compared to 19.4 and 30.4% in the control. IVUS analysis revealed a significant reduction of the percentage neointimal volume and no evidence of late stent malapposition in either treatment group.

Given these data, FUTURE II supported and extended the safety and feasibility results observed in FUTURE I by expanding treatment to a higher risk patient population, which included diabetics and patients with longer lesions. Finally, the pooled data analysis revealed an in-stent late lumen loss of 0.12 mm versus 0.85 mm as well as a binary restenosis rate of 0.0% versus 17.0% (in-stent) and 4.3% versus 27.7% (in-segment) in the combined everolimus and control groups, respectively. Based on these findings, Guidant Corporation entered into an exclusive agreement with BioSensors to enhance the everolimus-eluting S-Stent. By placing the S-Stent on the MULTI-LINK VISION® everolimus delivery system to improve performance, this device was renamed the “CHAMPION™” stent. In this stent design, the drug is loaded on the abluminal stent side and demonstrates an elution profile of 70% in approximately 30 days and 85% in 90 days. Porcine over-stretch models recently performed and reported have consistently shown safety and efficacy of this stent design with completeness of re-endothelialization of the CHAMPION™ stent similar to the control bare stent. Anticipating favorable results of these tests, two clinical studies have been planned, FUTURE III and FUTURE IV.

FUTURE III—The Superiority Study 

FUTURE III353 is planned as a multicenter randomized study for evaluation of the efficacy of the everolimus-eluting CHAMPION™ Everolimus-Eluting Coronary Stent System compared to the bare-metal MULTI-LINK ZETA® Coronary Stent System in a total of 840 patients. Up to 80 sites will be included in this international superiority study. Primary endpoint is defined as in-stent late loss at angiographic follow-up. FUTURE III will be conducted as a series of three sequential randomized trials with different intervals of follow-up for angiography and IVUS imaging. In the first series, 120 patients will be randomized (CHAMPION™ versus ZETA®) and will undergo follow-up at 4 months. The next 360 randomized patients will undergo angiography at 6-month follow-up and the final 360 patients will have angiography at 12 months postimplantation. Substudies will examine the impact of diabetes as well as inflammatory markers (hs-CRP). Long-term follow-up will be performed in all patients and the results pooled for analysis. FUTURE III enrollment started in April 2004.

FUTURE IV—The US Pivotal Noninferiority Study 

To compare the EES system with already approved drug-eluting stents, FUTURE IV353 will be conducted as a randomized noninferiority US pivotal clinical trial. FUTURE IV is designed to secure US FDA approval and will compare the CHAMPION™ EES System to either the paclitaxel-eluting TAXUS™ or the sirolimus-eluting CYPHER™ stent in a total of 975 patients at up to 80 study sites. Primary endpoint is an angiographic in-segment late loss at 8 months. Secondary endpoint is clinically driven target vessel failure at 9 months. An angiographic and IVUS follow-up is scheduled at 8 months poststent implantation. Clinical follow-up will be performed at 1-, 6-, 8-, 9-, and 12-months postimplantation and then yearly for up to 5 years.

Everolimus was also investigated in the SPIRIT FIRST trial,354, 355 a multicenter randomized controlled trial assessing the feasibility and performance of the Guidant MULTI-LINK VISION-E® RX Drug Eluting Stent System in the treatment of patients with de novo native coronary artery lesions. Sixty eligible patients were randomized in the proportion of 1:1 to stent implantation with a Guidant MULTI-LINK VISION-E RX Drug Eluting Stent System or with an uncoated Guidant MULTI-LINK VISION RX Coronary Stent System.

At 1-month follow-up, two MACE events had occurred in the EES group (one Q-wave MI and one TLR by PCI), whereas none were reported in the control group. There were no additional MACE events between 1- and 6-month follow-up in the EES group. At 6-month follow-up, the rate of total MACE in the control group was 21.4%, which was considerably higher than the total MACE rate of 7.7% in the EES group. There were no reported incidences of acute, subacute, or late stent thrombosis, and no cardiac deaths occurred during the course of follow-up in either arm of the study.

Angiographic in-stent late loss at 6 months (primary endpoint) was significantly lower in the EES group than in the control group (0.10 versus 0.84, respectively; P < 0.0001), accounting for an 88% reduction versus control. The reduction remained significant when measuring late loss at the proximal and distal edges. The rates of both in-stent and in-segment percentage diameter stenosis and binary restenosis were also significantly lower in the EES group than in the control group. By IVUS analysis, the use of the EES reduced neointimal volume by 73% and reduced in-stent volume obstruction by 70% compared with control (P < 0.001 for both comparisons). The rate of in-stent late loss reported in SPIRIT FIRST compared well with late loss rates reported in other drug-eluting stent trials.