In 2002–2003, drug-eluting stents (DES) were approved by regulatory bodies in Europe and the USA after initial studies showed a dramatic reduction in rates of restenosis compared with bare metal stents (BMS). Subsequent randomized trials in more challenging lesions have confirmed this benefit with a reduction in restenosis of 60–80%, irrespective of the type of lesion or clinical syndrome. Accordingly, the use of DES has been swiftly embraced with market penetration of up to 90% in certain countries. Nevertheless, clinical event rates after implantation of DES are certainly not zero.
So why do DES 'fail'? Not all of these clinical failures are due to 'failure' of the DES, since many patients have progression of atherosclerosis elsewhere in the coronary tree. However, restenosis still occurs even with DES and finally there are questions regarding the long-term clinical outcome with particular concerns about stent thrombosis.
Shortly after the introduction of DES into clinical practice, the authors reported on a series of 19 patients with angiographic restenosis in 20 lesions after implantation of sirolimus-eluting stents (Cypher; Cordis Europa NV).Thirty per cent of the cases demonstrated edge restenosis, the majority (83%) of which were caused by local injury outside the stent. Edge restenosis occurred more frequently in the proximal than in the distal stent border, suggesting that higher downstream drug concentrations may be protective.
Restenosis within the stent also occurred, accounting for 70% of cases. Most of these (86%) were focal (restenosis length <10mm) and presented a peculiar angiographic pattern manifested by a very localized stenotic site bordered by segments without evidence of lumen compromise. In half of these cases, a gap between stents or a stent fracture could be identified, implying a reduced local drug concentration. Incomplete lesion coverage in ostial lesions or bifurcations was also found.
Subsequently, the authors' clinical practice has changed to accommodate DES, and it is now routine to avoid balloon dilation in a non-stented area and to strive for complete lesion coverage, including the treatment of any residual dissections. Although many of these clinical failures are potentially modifiable, in-stent restenosis nevertheless may occur due to unequal stent strut distribution or disruption of the polymer coating, resulting reductions in local drug availability. Restenosis is clinically silent in approximately half the patients and usually presents as a recurrence of angina.
Also of concern is the issue of stent thrombosis (ST), which although less common than restenosis, causes far more serious clinical events including myocardial infarction and death. Although effective in reducing neointimal hyperplasia, the drugs used are potent cytostatic or cytotoxic agents with detrimental effects on endothelialisation. The incidence of late stent thrombosis was evaluated by two large institutions (Bern and Rotterdam) which implemented the unrestricted use of DES shortly after the products received their CE Mark, thus providing an opportunity to evaluate the technology in all-comers, with the longest possible follow-up available. At the European Society of Cardiology annual conference in September 2006, Dr Wenaweser from Bern presented the findings that identified 152 patients with angiographically documented ST amongst a cohort of 8,146 patients. The cumulative incidence of stent thrombosis was 2.9% after three years. Early ST (within 30 days of the procedure) was observed in 91 (60%) patients, and late ST in 61 (40%) patients. At the time of ST, dual antiplatelet therapy was taken by 87% of early and 23% of late ST patients. Late stent thrombosis was encountered steadily at a constant rate, with an incidence of 0.6% per year between 30 days and three years with no evidence of diminution up to three years of follow-up. This phenomenon was observed with both sirolimus- and paclitaxel-eluting stents.
Some of the most useful mechanistic data regarding stent thrombosis come from pathology. A recent histopathological report described 23 patients who died more than one month (mean 223 ± 253 days) after DES implantation. Fourteen of these patients suffered from stent thrombosis, of the group, seven were receiving dual antiplatelet therapy. Amongst the DES patients, those with and without ST had significant differences in the degree of endothelialization (27.1% vs 66.1%), mean stent length, (38.8mm vs. 23.6mm) and mean number of stents implanted (1.74 vs 1.25).The DES ST patients all had evidence of delayed healing. Additionally, three demonstrated chronic inflammation or hypersensitivity, in three patients the procedure involved an ostium of bifurcation, malapposition was found in two cases, and in-stent restenosis in two. Compared with matched BMS patients, DES showed persistent fibrin deposition, poorer endothelialization, (55.8 ± 26.5% struts endothelialized vs 89.8 ± 20.9, p=0.0001) and more eosinophils (5.6 ± 1.1 vs 0.6 ± 2.3, p=0.01).
For the moment we can minimize DES failures by utilizing good procedural practice (complete lesion coverage, adequate stent deployment etc.) and maintaining long-term clopidogrel therapy. Apart from possible reactions to the polymer, the first generation of DES have done exactly what they were designed to do. However, the goals have shifted and future technology clearly needs to address the failings of the first generation of DES.
The primary goal of the next generation of DES is no longer the complete cytotoxic inhibition of neointimal hyperplasia but the restitution of a healthy, functionally active endothelial lining capable of modulating the healing process without a permanent metallic implant; biodegradable DES are currently undergoing clinical trials (see Figures 1 and 2).
Additionally, the era of multiple drug elution has started with the dual release of heparin benzylconiumsirolimus, palcitaxel-pimecrolimus, and zotarolimusdexamethasone. These anti-inflammatory drugs (dexamethasone, pimecrolimus) mitigate the cytotoxic effect of the anti-proliferative drugs while having a synergistic effect on smooth muscle cell proliferation.
Possible future options include stents which capture endothelial progenitor cells (EPCs) together with the use of drugs which promote re-endothelialization, such as nitric oxide donors, estrogens, or vascular endothelium growth factor. Using the same principle, another alternative would be to create a combination of EPC capture and release of cytotoxic drugs, but at low concentrations. The ultimate dream may be a biodegradable platform attracting endothelial cells and eluting drugs.