Acute

Balloon intervention suffers two major problems. Firstly, high pressure inflation induces “controlled” disruption of the atheromatous plaque, which sometimes becomes less controlled. At these times acute coronary closure may result. Published data suggest that outcome after surgery in such circumstances is worse than the outcome following stenting and worse than that after routine “cold” operations. Some authors3 have reported mortality rates of between 5–10% in such circumstances, depending on the patient's clinical condition at the time of urgent surgical referral. Kioka and colleagues4 reported an astonishing 15.8% mortality.

By the mid 1990s stent use was being advocated to improve outcome following acute coronary closure, the concept being that scaffolding of the loose and potentially obstructive intima would restore flow and reduce the need for emergency coronary surgery. While there are few large observation studies or any randomised trials, outcome following stent placement for bailout appeared good, with mortality rates of less than 5%.5 Roubin has published data on bailout stenting indicating an optimal angiographic result can be obtained in 93% of patients and a mortality rate of 1.7%.6 Clinical evidence suggests that the earlier the stent is deployed after a poor angiographic result, the better the outcome. Indeed, one reason for the increase in stent use was the desire not to wait until the artery closed, at which time stenting is more difficult and potentially unsuccessful. Two year survival following stenting for acute bailout has been reported to be 95%, but event-free survival was only 70.7%.7 Recent changes in stent design, deployment techniques, and treatments to prevent thrombotic occlusion and restenosis in higher risk patients will, it is hoped, improve this longer term outcome, but it highlights the issue of restenosis.

• Consensus—Acute closure after balloon angioplasty carries a high complication rate.

Chronic problems: restenosis following angioplasty

Initial concepts on recurrence or restenosis after balloon angioplasty centred on the neointimal (smooth muscle cell) response. However, clinical drug trials to limit the impact of smooth muscle cell hyperplasia were generally unsuccessful. Concepts evolved to include the importance of the acute luminal diameter gain, recoil, and the impact of negative remodelling8 (whereby the restenosing artery gets smaller rather than bigger to accommodate the intraluminal tissue) (fig 2). It was generally felt that such mechanical issues were more important than the impact of tissue in-growth. This meant that stent deployment could have an impact on recurrence since it would deal with three of the four factors thought to be important.

• Download figure
• Open in new tab
• Download powerpoint

Figure 2

Stents improve outcome by improving the acute gain and by reducing recoil and negative remodelling, although tissue growth is not affected and may even be greater with stents than balloon angioplasty. Even so, the eventual lumen diameter is still larger after stenting (A2 > A1).

The larger the lumen that can be achieved and maintained after PCI, the less impact any restenotic tissue might have.9Additionally 60% or so of the loss of lumen is caused by elastic recoil and negative remodelling. Quantitative angiographic data from a number of trials clearly showed that stents produced a significantly greater acute luminal gain,10 and prevented recoil. Although tissue response to stenting is exacerbated it has less impact since the arterial lumen is larger. Trial data10 indicate that the difference in final acute minimal luminal diameter can be increased from about 1.7 mm with angioplasty to about 2.7 mm (158%) with stents.

• Consensus—Evolving concepts on restenosis suggest a mechanical solution is likely to reduce recurrence rates.

Stenting versus angioplasty alone

Two major trials were designed to assess the medium term angiographic and clinical outcome following de novo (primary) stenting. Angioplasty in these trials was the preliminary procedure to stent delivery but could be used to optimise the post-stent lumen. The two trials (BENESTENT I and STRESS) have clearly shown that stenting in native vessels reduces the incidence of recurrence. In terms of reduction in restenosis rates, the results were remarkably similar in both trials.10 In the BENESTENT study, the primary clinical end points of myocardial infarction, need for CABG or re-PTCA, and stroke had a relative risk of 0.68 (95% confidence interval 0.5 to 0.92) in those patients randomised to stenting compared to those undergoing PTCA alone (p = 0.02). The angiographic restenosis rate, measured quantitatively on follow up angiogram, was 22% for stenting and 32% for PTCA. For the STRESS trial11 the restenosis rate was 29.1% for stenting versus 42% in the PTCA arm (p = 0.011).

A number of equivalence studies have now been published comparing newer stents with the stents used in these trials. Any stent which produces a good acute result leads to recurrence rates of between 15–20% compared to historical results of 35% for PTCA.12 The WEST,^w7 MUSIC,^w8 and FINESS^w9trials have confirmed even lower restenosis rates. Use of intravascular ultrasound to optimise the best possible result leads to restenosis rates of < 10%, but this cannot be justified in terms of time and costs.

Effect of stent use on clinical practice

While appropriately conducted clinical trials have shown a clear benefit of stenting compared to angioplasty, such trials are sometimes criticised for not being “real life” in that there is always some degree of patient selection. The case for stents in reducing the need for subsequent reintervention is supported further by indirect evidence from the British Cardiovascular Intervention Society (BCIS) audit data. From 1992 to 1998 there was an increase in angioplasties from 11 575 to 24 661 (113%) with an increase in case complexity. During the same period there was a reduction in reintervention rates for restenosis (from 11.6% to 5.2%) mirrored by an increase in stent use from 2.7% of all cases to 69%. This suggests that an increase in stenting reduces the clinical need for reintervention (table1).

Is there too much stenting?

It has been suggested that stenting is used uncritically and unnecessarily.13 A study designed to assess the potential added value of stenting once a good angioplasty result had been obtained has been published.^w10 In this trial (involving only 116 patients) if residual stenosis was less than 0.3 mm, patients were randomised to either stent or no stent; 13.5% of patients developed recoil at 30 minutes and crossed over to the stent group. The results showed that there was no difference in restenosis rate, event-free survival, and target lesion revascularisation between the two groups at follow up. There are a number of important issues related to this study. Firstly, interventionists do not generally stent when they achieve a stent-like result. There are certain other messages, however. Of 953 angioplasties considered for this trial only 116 (12%) met the criteria for the study, suggesting that achieving a “stent-like” result in everyday practice is difficult. It is unclear from the cost benefit analysis whether the excess number of balloons that are frequently required to achieve a stent-like result have been included. Finally, one of the stents that was used predominantly was the GR II stent, which is a coil stent and as such has been associated with a higher restenosis rates. Certainly for the lesions included, stent restenosis rates would be expected to be lower than that cited in this study.

It is clear that if a “stent-like” result can be achieved with one or two balloons then stenting (which with contemporary units requires only one predilatation balloon) need not be necessary. Such immediate success with angioplasty alone is, however, uncommon and attempts to obtain the + 5% to −5% residual stenosis routinely seen after stenting is not without attendant risks of dissection and acute vessel closure, since larger balloons and higher pressure may be needed.

• Consensus—Stenting improves outcome compared to balloon angioplasty alone. Achieving a stent-like result with angioplasty alone is difficult.

Stenting following previous CABG

Long waiting lists for first time surgery, the success of the procedure, and the lower mortality and morbidity associated with angioplasty compared to re-do operation makes non-surgical intervention an attractive method of treating post-CABG patients (about 15% of all PCI). Restenosis rates after angioplasty to vein grafts are known to be higher than in native vessels, however, although the reasons are unclear. Rates of up to 60% have been reported.^w17 In some small observational studies the rate of restenosis following stenting has been shown to be less (range 14–34%).^w18 A further observational study by the Palmaz-Schatz study group^w19suggests that the clinical outcome and restenosis rates for stented grafts are comparable to those for angioplasty undertaken on native vessels. In other words, the excess rates associated with angioplasty to vein grafts may be “normalised” by stenting. A recently published study (saphenous vein de novo trial) confirmed a better clinical outcome in that the end points of freedom from death, myocardial infarction, repeated bypass surgery or revascularisation were reached less frequently in stented patients (73%v 58%, p = 0.03).^w20 One further challenge for the treatment of old grafts with stents is the incidence of no reflow, where atherosclerotic debris occludes the distal vessel. A recent trial (SAFER^w21) suggests that the outcome (in terms of enzyme defined myocardial infarction) is better if a distal protection device is used.

Stents for treating occluded vessels

There is good evidence that using stents in the setting of balloon angioplasty for chronic total occlusions improves the outcome. This was shown in the SICCO trial20 which reported that 57% of 119 stented patients were angina-free compared to 24% treated with angioplasty alone (p < 0.001). Angiographic restenosis occurred in 32% of stented patients compared to 74% of the PTCA group (p < 0.001) and reocclusion occurred in 12% and 26%, respectively (p = 0.058). Target lesion revascularisation within 300 days was also less frequent in the stented patients (22%v 42%, p = 0.025). The TOSCA study^w22 confirmed the value of stenting in chronic total occlusions. Stenting produced a 45% reduction in clinically driven target lesion revascularisation at six months (15.4%v 8.4%, p = 0.03).

Stents in unstable angina.

Despite some evidence that lesions in patients with unstable angina are more thrombogenic and are therefore potentially at higher risk, intervention for acute coronary syndromes accounts for 40% of the angioplasty workload. Support for the use of stents in the setting of unstable angina is available in the literature. Marzocchi and colleagues have recently published their experience of stenting in the setting of unstable angina.^w23 Of 266 consecutive patients stented either electively (24%), for bailout (11%) or for suboptimal angioplasty result (65%), the overall 30 day mortality was 0.3% (one patient with cardiogenic shock), and the incidence of non-fatal myocardial infarction was 2.2% (six patients). At longer term follow up (mean (SD) 17.7 (9.4) months) the cardiac mortality was 0.4%, the myocardial infarction rate was 1.5%, and the target lesion revascularisation rate was 9.3%. Again certain subgroups, such as patients with diabetes or those who had longer stents implanted, adversely influenced the outcome.

Stents and acute myocardial infarction

Thrombolysis remains the mainstay of treatment for acute infarction. Intervention in the form of primary angioplasty has been shown to be better than angioplasty in terms of patency and longer term follow up.^{w24 w25} However, primary angioplasty has a major drawback. Most patients with acute infarcts present to district general hospitals with no interventional facilities. However, in those that do undergo PCI a number of studies have shown that the use of stents appears to confer significant advantage despite the potentially thrombogenic combination of the stent and disrupted thrombogenic atheromatous plaque. The GRAMI trial^w26 randomised 104 patients to either balloon angioplasty alone or to balloon angioplasty followed by stent deployment. While procedural success was equally high in both groups (94.2% v 98%, respectively), in-hospital adverse events were significantly less in the stented group (3.8% v 19.2%, p < 0.03), and at late follow up event-free survival was 83% for the stented group versus 65% for the angioplasty group (p < 0.002). Other small observational studies appear to show that restenosis rates in patients stented during acute infarction are no higher than for routine stented patients.^{w27 w28}

The use of angioplasty with stenting, if considered necessary, as compared to repeat thrombolytic or conservative treatment in patients with failed thrombolysis is being tested in the UK based REACT trial being funded by the British Heart Foundation. A similar trial (MERLIN) is comparing repeat thrombolysis and intervention.

Small vessel disease

Previous trials have suggested high restenosis rates (> 45%) in those patients with small vessel disease. Four recent studies have produced conflicting results. A number of studies have recently been published. The SISA trial^w39 (n = 325) showed a trend only towards less adverse outcome with stent (Bestent) versus balloon in vessels < 2.9 mm. This was because the acute gain with stenting (1.37 mm v 0.91 mm, p = 0.0001) was offset by the greater late loss in the stent group (28.5%v 18.4%, p = 0.0002). In the setting of small vessels such loss cannot be accommodated so easily. The BESMART study^w40 (n = 381) on the other hand (also Bestent) showed a restenosis rate of 22.7% in the stent group versus 48.8% in balloon group (p < 0.0001) and a target lesion revascularisation rate of 13% versus 25%, respectively (p = 0.016). ISAR-SMART showed no benefit from stenting (35.7% v 37.4%, vessel size 2.0–2.8 mm) whereas RAP (Garcia) (vessel size 2.2–2.7 mm) demonstrated a restenosis rate 27% in the stent group versus 37% in the balloon group. The average restenosis rates for these studies in stents deployed in vessels of between 2.2–2.9 mm is 27.5%, double that of BENESTENT lesions. It is clear that physicians will wish not to exclude patients from the potential benefit of stenting based on vessel size. It is likely therefore that it is in this group that drug eluting stents may have most impact since intimal hyperplasia will always have greater impact on the smaller lumen.

Left main stem stenting

It has always been regarded as taboo to undertake PCI on unprotected left main stem disease. However, a number of groups world wide are, through the use of registries, identifying the risk and the patients in whom such intervention could be deemed acceptable. In general early studies such as that by Park suggested excellent results in those at low risk (100% success rate, 17% clinical recurrence at six months and only one death),^w41 although Barragan reported three deaths out of 15 high risk patients.^w42Ellis has reported on the ULTIMA registry and the outcome appeared dependent on patient characteristics, ranging from those with stable angina to those deemed inoperable, as well as left ventricular function status. Two hundred and seventy nine consecutive patients who had left main stem PCI at one of 25 sites between 1993 and 1998 were studied. Forty six per cent of these patients were deemed inoperable or at high surgical risk. Thirty eight patients (13.7%) died in hospital, and the rest were followed for a mean of 19 months. The one year incidence was 24.2% for all cause mortality, 20.2% for cardiac mortality, 9.8% for myocardial infarction, and 9.4% for CABG. Independent correlates of all cause mortality were: left ventricular ejection fraction (LVEF) ⩽ 30%, mitral regurgitation grade 3 or 4, presentation with myocardial infarction and shock, creatinine ⩾ 2.0 mg/dl, and severe lesion calcification. For the 32% of patients aged < 65 years, with LVEF > 30% and without shock, the prevalence of these adverse risk factors was low. No periprocedural deaths were observed in this low risk subset, and the one year mortality was only 3.4 %. It is likely that PCI for unprotected left main stem disease will be kept for those at high risk from surgery.

Direct stenting

Direct stenting may have significant advantages over routine procedure where there is predilatation with a balloon first. There appears little doubt that tissue response to the vessel wall is the result of the damage induced by the balloon inflation (and stent deployment). If this damage could be reduced, less tissue response should occur. Additionally the procedure would be cheaper requiring a balloon-stent unit only. There have been a number of reports outlining the feasibility of such a strategy. Hamon recently published the outcome of 122 “carefully selected” patients. Factors such as calcification and tortuosity need to be taken into account when direct stenting is considered. Procedural success was 96%.^w44 In five cases it was not possible to deliver the stent through the undilated lesion and the stent was retrieved, but in two cases the stent was lost in the peripheral circulation. The authors rightly reported the need for a controlled trial and one UK study (the SLIDE trial) is underway.