The technical goal of coronary stenting should be to eliminate or lessen ischemic symptoms in the short/intermediate term and possibly reduce the risk of death or myocardial infarction, while not exposing the patient to excess short- or long-term risk. Yet major adverse cardiac events (MACE) occur in 4% to 20% of stented sites within a year,^1,2 and stent underexpansion^3,4 and axial misalignment (“geographic miss”)⁵ account for nearly half of these. Although a discussion of choosing percutaneous intervention instead of bypass surgery or optimal medical therapy alone is beyond the scope of this chapter, we will address various issues to be considered to provide the patient with an optimal technical result.

Improvements in guide catheter and wire support, as well as balloon and stent deliverability, have facilitated small-caliber device intervention and, in particular, radial access–based intervention. The result has been a dramatic reduction in bleeding complications and red cell transfusion–induced inflammation and perhaps even a reduction in patient mortality under some circumstances (especially primary percutaneous coronary intervention [PCI] for ST-segment elevation myocardial infarction [STEMI]).⁶ Except when anatomy precludes it, radial access should be the preferred route of vascular access. Femoral, brachial, and axillary access can at times be reasonable alternatives.

Both antiplatelet and antithrombin therapies are required for safe stent implantation and healing. Aspirin and clopidogrel are usually sufficient antiplatelet therapy for patients with stable ischemic symptoms. Second-generation P2Y₁₂ inhibitors have been shown to be superior to clopidogrel in unstable angina patients but not, in general, in stable patients.^7,8

Although direct thrombin inhibitors have several potential advantages over unfractionated heparin as the requisite antithrombin agent to accompany coronary stenting, with good dual antiplatelet therapy, heparin alone is almost always adequate to prevent complications⁹ and is cost effective.¹⁰ The situation with STEMI patients is perhaps less clear, with a recent post HEAT-PPCI (Unfractionated Heparin Versus Bivalirudin in Primary Percutaneous Coronary Intervention) meta-analysis showing no difference between bivalirudin and heparin for the end points of death or myocardial infarction, but more stent thrombosis and less bleeding with bivalirudin.¹¹ Although disputed by some,¹² most operators use activated clotting time monitoring to assure satisfactory antithrombin effect (200-250 seconds for the commonly used HemoTec) (Medtronic, Dublin, Ireland system) (the Hemochron [Accriva, Piscataway, NJ] system measures 50 seconds longer),¹³ with higher levels for complex interventions (eg, acute coronary syndrome)¹⁴ or those requiring multiple guide wires in the same coronary artery.

Stent Length

In general, stents should be chosen to cover the lesion, from (angiographic) “normal to normal” vessel. Assessing where “normal” begins and ends is sometimes challenging, and intravascular imaging can often be helpful. Stent ends should not be deployed in areas of obvious plaque buildup, in vessel bends, or ideally in sites compressing thin-capped plaque or areas of necrotic debris (“lipid-rich plaque”).^15,16 Stents should also be deployed to cover the length of artery traumatized by predilatation ballooning or atherectomy, so as to avoid the “geographic miss” found in the STLLR (Stent Deployment Techniques on Clinical Outcomes of Patients Treated With the Cypher Stent) study that doubled the risk of MACE (Fig. 29-1).⁵ For longer lesions, it is often difficult to anticipate the stent length needed. One can use the length of the radiopaque part of the coronary guide wire (30 mm for most) or the known length of the predilation balloon for a “ruler.”

Diameter

To a large extent, the degree of luminal late loss after stenting is related to biologic processes such as the presence of diabetes and not to the extent of stent expansion (within limits)¹⁷; hence, the general adage has become “bigger is better.” But how big is big enough? Studies from a decade or more ago reliably found that what appeared angiographically to be a very good stent result (<10% stenosis) not infrequently left the stent undersized, underexpanded, or both when viewed by intravascular ultrasound (IVUS).¹⁸ In addition, meta-analysis of the 7 moderate-sized randomized controlled trials of IVUS versus angiographic guidance for bare metal stent implantation found that IVUS modestly reduced restenosis¹⁹; hence, we rely a lot on IVUS findings to answer this question. Importantly, IVUS has shown that balloon postdilation at 14 to 16 atm for a nominally sized stent delivery balloon typically results in the stent achieving only about 75% of predicted (out of the body) diameter (Fig. 29-2).²⁰ In the drug-eluting stent era, Hong et al²¹ found that the best IVUS cross-sectional area (CSA) cutoff value discriminating risk of restenosis or no restenosis was 5.5 mm² (length also was a risk factor). However, the implications of achieving that value are not the same for a patient with a 2.25- or 3.5-mm diameter vessel and further leave the patient with an average rather than ideal risk of restenosis. Other criteria have been proposed, ranging from the classic but difficult to achieve MUSIC (Multicenter Ultrasound Stenting in Coronaries) criteria (complete stent apposition and CSA ≥80% of the average of proximal and distal “normal” reference areas²²), achieved in only about 60% of cases,²³ and the easier to achieve ≥55% CSA of the reference segment proposed by Moussa et al.²⁴

IVUS studies have also linked poor stent expansion to risk of early stent thrombosis, finding, for instance, increased risk with failure to achieve 80% reference CSA or CSA <5 mm.^25-27 Dimensions obtained via optical coherence tomography (OCT) are likely helpful in this regard, because they more closely reflect actual dimension compared with IVUS, which overestimates vessel diameter by 10% to 15%.

However, in the current era of cost containment, IVUS or OCT cannot be justifiably used in all cases, so many advocate their use principally to clarify angiographic ambiguities, and evaluate pre- or poststent anatomy, and treated lesions that are at high risk of restenosis (diabetes, long lesions) or are anatomically high risk (left main stenosis).

Although perhaps not as well validated as one might like, we favor using angiographic step-up and step-down as a surrogate for adequate stent expansion, as initially put forward by Antonio Colombo’s group¹⁸ and, more recently, supported by work of Haldis et al.²⁸ This approach has led to a 90% likelihood of successful implant by MUSIC criteria, albeit with an increased risk of OCT-detected small edge dissections.²⁹

The role of predilatation/pretreatment before stenting should be to both allow stent delivery and also full expansion, without extending the treatment area via dissection or slippage. Balloon inflation, sized at 0.25 to 0.50 mm less than vessel diameter, usually suffices. Several devices other than plain old balloon angioplasty (POBA) have been proposed to improve the likelihood of achieving this goal, but all come with increased cost and hence should be used judiciously. Of these, perhaps the easiest to use is the AngioSculpt (Angioscore, Fremont, CA). Studies have shown greater stent expansion with pretreatment using this device compared with POBA (eg, 88% vs 77% of expected CSA in moderately complex lesions, irrespective of lesion morphology²⁰; Fig. 29-3). It should be noted that there were few heavily calcified lesions in this study, and we and others have encountered difficulties with the device in heavily calcified lesions, including device entrapment.³⁰ The Cutting Balloon (Boston Scientific, Marlborough, MA) cuts more deeply and is less deliverable than the AngioSculpt, but in general, it has the same benefits and limitations. Both devices are particularly useful to prevent device slippage in the setting of in-stent restenosis (ISR) and for side branch treatment in conjunction with provisional stenting of bifurcation lesions.

More heavily calcified or difficult to cross lesions often require rotational atherectomy with the Rotablator (Boston Scientific) or Orbital Atherectomy Device (Cardiovascular Systems Inc [CSI], Saint Paul, MN) before stenting. An arc of calcium >180° by IVUS has been associated with poor stent expansion,³¹ but this may be hard to predict with angiography alone (suggested with somewhat limited sensitivity and specificity by angiographic calcium visible across the diameter using still frame imaging)³² and is more common in patients with renal insufficiency and in the elderly. A randomized trial comparing Rotablator (mean burr size, 1.5 mm) versus POBA in calcified lesions before drug-eluting stent implantation showed an increase in acute gain that was largely offset by more late loss and no difference in restenosis or stent thrombosis.³³ However, because poor stent expansion is a risk factor for stent thrombosis and restenosis, atherectomy for lesions at risk for poor stent expansion may be beneficial.

The CSI speed-expansive rotational atherectomy device competes with the older device, being simpler to set up and use and providing for generally more ablation, but with greater risk of vessel perforation (1.8% in the ORBIT-2 [Evaluate the Safety and Efficacy of OAS in Treating Severely Calcified Coronary Lesions] study).³⁴

We will often resort to a trial of balloon inflation to 6 to 10 atm to see if the vessel “yields” if we are uncertain about the need for atherectomy. Although there is a small risk of creating a dissection without adequate vessel enlargement with this maneuver, atherectomy with the Rotablator seems to be safe in this setting.

A more complete discussion of the role of rotational atherectomy is provided in Chapter 34.

For moderately stenotic noncalcified lesions, direct stenting without predilatation or atherectomy has been advocated, largely as a time- and cost-saving approach (generally about 20% for both).³⁵ A recent meta-analysis of randomized controlled trials evaluating this approach found a significant 20% to 25% reduction in periprocedural myocardial infarction or death, but no difference in late target vessel revascularization.³⁶ The downsides to this approach are occasionally being unable to deliver the stent, potentially misjudging the length of the stent due to lack of the balloon marker “ruler,” and rarely, discovering a difficult-to-dilate lesion after the stent has been placed. Used judiciously, this approach appears to have merit.

Both failing to fully dilate the edges of a stent and dilating beyond the edges (“geographic miss”) are fairly common.⁵ To minimize risk of edge trauma, the delivery balloon should generally be inflated to only 10 to 14 atm (the low end for somewhat oversized stents). Higher pressure inflation risks “dog-boning” and edge dissection, because the balloons on which contemporary stents are mounted are semi-compliant. Postdilatation is generally required unless angiographic step-up/step-down has already been achieved. Noncompliant balloons with a diameter approximately 0.25 mm larger than apparent vessel diameter should be used, inflated to nearly maximal pressure, for postdilatation. High-quality angiographic imaging (Stent Boost [Philips, Amsterdam, the Netherlands], or equivalent) should be used to assure that dilatation (outside edge of balloon marker) has extended to, but not beyond, the stent edge.

The role of adjunctive imaging techniques is addressed in Chapter 26. Fractional flow reserve (FFR) has been shown in several setting to be highly useful in discerning which lesions require revascularization,³⁷ but the optimal role of the 3 imaging modalities remains uncertain for routine and uncomplicated situations. We favor the use of IVUS for high-risk anatomic situations (eg, left main stenoses), for use in settings at high risk for restenosis (eg, long lesions, small but important vessels, diabetics), evaluation of the etiology of ISR, and in situations of ambiguous anatomy before or after PCI by angiography alone. OCT yields much better near-field resolution, often finds poststent luminal disruptions, and may be helpful in assessing which peristent dissections need further stents (Fig. 29-4).³⁹ Infrared spectroscopy allows for identification of lipid-rich plaque, the placement of a stent into which leads to an increased risk of debris embolization.