Lessons learned in the past five years in relation to drug-eluting stents (DES) include their efficacy and safety, particularly sirolimus-eluting stents (SES), compared to bare metal stents (BMS) as supported by the best available data. Other lessons are:

• Stent thrombosis is not a new problem related to DES.
• Mechanisms of stent thrombosis is not fully understood.
• Very low incidence of stent thrombosis, yet very significant consequences.
• Stent thrombosis occurs in BMS, although much more rare versus DES.
• Improvements needed in the outcomes of modern percutaneous coronary interventions (PCI), but BMS are not the answer.

The goal is no risk of stent thrombosis, which requires continued basic and clinical research.

2006: A pivotal year

Prior to summer 2006, it was clear that DES were an effective technology without emerging problems, stated Costa. In 2003 the major known problem was in-stent thrombosis, which was addressed successfully by DES.  First-in-man 5-year clinical data with SES from Costa and colleagues showed no concern regarding restenosis or stent thrombosis (ST). Although this was a single center study with only 100 patients, Costa stated it reflects the power of the technology.

Large clinical trials comparing BMS and DES in pivotal randomized clinical trials showed significant benefit for freedom from major adverse clinical events (MACE) through 4 years.

Then, data presented at the ESC in 2006 called into question the safety of DES along with stating they may cause more harm than benefit. This began the “dark year for DES” stated Costa. Data from the SCAAR Registry was presented in 2006 and published in 2007 showing a higher mortality with DES than BMS after 6 months, and pointed to late and very late stent thrombosis. CYPHER data from Europe showed that late and very late ST were associated with a very high risk of complications, although the incidence was unclear.

However, these data led to the a valid, agreed definition of stent thrombosis led by the Academic Research Consortium (ARC).

Mechanism of stent thrombosis

Delayed endothelialization is proposed as one mechanism for stent thrombosis. One hypothesis is that endothelial function returns to normal after BMS placement (Kawaguchi, Costa et al 2007). In relation to DES, there is an hypothesis that endothelial function never returns to normal, while yet another proposes that endothelial function is lulled after DES placement but will normalize over time and in about 5 years anti-platelet therapy can be stopped. Costa stated that the latter hypothesis was not biologically possible. A third hypothesis in relation to DES is that there is a “calm period” from 9 months to about 18-24 months during which the ST rate is none or low and then ST occurs; endothelial function improves but then deteriorates over time. Costa noted this makes some biological sense because late inflammation can deteriorate endothelial function that was initially recovered and thereby cause ST independent of the initial process.

Deployment causes ST

Improper deployment was determined to be the cause of the very high rate of ST (15-20%) seen with BMS in the early 1990s, when multiple anti-coagulation medications were required. Costa believes deployment technique with DES may be contributing to the ST seen with DES, and this is suggested by the 1-year safety data from STLLR.

Turning tide

In 2007, emerging data presented a different picture of DES compared to the data at the ESC in 2006.  The SCAAR data reported at the ESC in August 2007 showed that DES reduced the risk of myocardial infarction (MI) and death. This was followed by two reports from Europe showing DES improved death and myocardial infarction, and that SES was better than paclitaxel-eluting stents (PES).

The Ontario Registry 3-year follow-up data showed DES reduced death and MI compared to BMS, and there was no incremental risk of mortality with DES (Tu, NEJM 2007).

Mauri and colleagues showed in 2-year government performance data (10,882 patients, matched risk differences) that mortality decreased to 8% with DES compared to 11% with BMS and revascularization decreased to 22% with DES from 24% with BMS.

National Medicare data in the United States compiled by Cohen showed in the 2004 data (with the 75% having DES) that adjusted and unadjusted survival, MI, and repeat revascularization were improved, with a direct impact on cost efficacy.

The Collaborative Network Meta-Analysis with some 60,000 patients from major randomized controlled trials of BMS versus DES and SES versus PES with 4-year follow-up showed that the incidence of ARC-defined ST was similar for BMS and SES. ST was increased with PES, particularly versus BMS. MI was reduced with SES, while the rates were similar for PES and BMS. All-cause mortality was not different between BMS, SES, and PES.

A summary of DES and BMS studies and registries with 115,399 patients followed for 2 years or more showed that no study showed significant harm with DES versus BMS. Nine of the 11 studies showed at least a trend for a safety advantage with DES and 6 of the 11 registries showed a significant safety advantage with DES.

Data from the prospective Asian-Multicenter DES-LMT Registry in 448 patients with unprotected left main trunk disease treated with either a sirolimus eluting stent (SES; n=248) or a paclitaxel eluting stent (PES; n=200), presented by Dr. Sunao Nakamura, New Tokyo Hospital, revealed at 3 years:

Safe and feasible treatment of unprotected left main coronary artery disease with the SES and PES.

• Significantly greater minimal lumen diameter (MLD) on QCA at 12 months in the SES group at 3.4% compared to 3.1% in the PES group (p=0.06).
• Lower rate of restenosis in the SES group (9.3% vs 14% with PES; p<0.05).
• Lower rate of target lesion revascularization with SES (8% vs 12.5% PES;p<0.05).
• Durable results to at least 3 years with the SES and PES.

The registry included consecutive patients symptomatic for chest pain or demonstrating inducible ischemia with angiographic evidence of ≥50% diameter stenosis of the left main coronary artery, who underwent successful PCI with an SES (from March 2000) or a PES (from March 2003). Follow-up was  36 months. The primary endpoint was major adverse cardiac events (death, MI and CABG) and 36 months and the secondary endpoint was the 12-months restenosis rate.

Antiplatelet therapy was clopidogrel 75 mg daily plus aspirin 100 mg daily for 12 months and then aspirin 100 mg daily thereafter. Angiographic and statistical analysis was performed at New Tokyo Hospital. IVUS was used in all cases in Japan.

The patients (average age 70 years old, 32%  women) were from Japan, Malaysia, Korea, Indonesia, and Thailand. At baseline the SES and PES groups were well matched: hypertension in 55%, hyperlipidemia in 29%, diabetes in 32-37%, 50% current smokers, prior MI in 8%, and the mean left ventricular ejection fraction (LVEF) was 49%.  

No significant differences in lesion characteristics were seen between groups:

• 2-vessel disease in 65% of SES and 58% of PES patients.
• 3-vessel disease in 35% of SES and 42 of PES patients.
• Stable angina on presentation in 89% of both groups.
• Mean proximal reference diameter 3.8 mm and 3.7 mm in the SES and PES groups.
• Bifurcated lesions in 74% of both groups.

Significant differences between groups were the mean lesion length (20.4 mm SES vs 21.7 mm PES; p=0.07) and the use of IVUS during the procedure (36% SES, 31% PES; p=0.07). Lesion location was ostial in 8.5%, midshaft in 12,5%, and distal in 79%.

At 30-days, the success rate was 100%, with no major adverse cardiovascular events (MACE), myocardial infarction (MI) or subacute thrombosis was seen in either group, or any significant differences between the SES and PES groups.

Bifurcated lesions were the primary site of restenosis, at 9% of SES and 13% of PES groups. In the SES group, no in-stent thrombosis was observed, compared to 2% of the PES group. No late stent thrombosis was seen at 3 years.

On multivariate analysis, ejection fraction was a strong predictor of death, and significant predictors of restenosis were diabetes, ejection fraction, and bifurcated lesion.