A. Balloon angioplasty

4. Balloon angioplasty with a balloon that is scored with blades (cutting balloon) or wires (Angiosculpt) allows the plaque to yield more easily, at lower inflation pressures. Theoretically, this increases the success of angioplasty and may be associated with less severe dissections. Also, a scored balloon is less likely to slip during balloon inflation. Standard, non-scored balloons tend to slip when inflated across a heavily fibrotic stenosis (“watermelon seeding”). Longer balloons, slower inflation, and scored balloons reduce this phenomenon.

B. Bare-metal stents (BMSs)

A stent addresses all three balloon angioplasty pitfalls. It covers and seals the dissection planes or, at least, prevents them from expanding and promoting abrupt vessel closure. It has enough radial strength to prevent early (elastic recoil) and late (restenosis) vessel constriction. Stenting improves the immediate and late PCI result, but is not without its own pitfalls:

1. Stent thrombosis: while preventing abrupt vessel closure, stenting is associated with a risk of stent thrombosis, mainly in the first month. This risk is 0.5–1% in patients receiving dual antiplatelet therapy (DAPT) with aspirin and clopidogrel, and 5–10% in patients receiving single antiplatelet therapy with aspirin.
   
2. In-stent restenosis (ISR): restenosis is reduced to 20–25% with stenting. Overall, stenting eliminates the late negative remodeling seen after plain angioplasty, but increases the degree of neointimal hyperplasia, which starts 1 month after stenting and progresses for up to 6–8 months. The rate of clinical restenosis after BMS, i.e., angina recurrence, is ~15%.

C. Drug-eluting stents (DESs)

(First-generation DESs: sirolimus [Cypher] and paclitaxel [Taxus, Ion]; second-generation DESs: everolimus [Xience V, Promus, Synergy], zotaralimus [Resolute]).

A DES consists of a stent platform similar to a BMS, with two additional components: a drug, and a polymer that controls the local drug delivery over 3–6 months. The drug inhibits cell proliferation and thus neointimal hyperplasia, with a restenosis rate of <5–10%. On the other hand, by delaying stent strut coverage, the risk of stent thrombosis used to persist beyond the first month. With the first-generation DESs, this risk of stent thrombosis remained significant for the first 12 months and mandated DAPT for 12 months. The second-generation DESs are associated with a dramatic reduction in the risk of stent thrombosis and target vessel MI (vs. first-generation DESs); the risk of stent thrombosis is very low even after clopidogrel interruption at 1–6 months.

While DES does not clearly reduce mortality or MI in comparison to BMS, it dramatically reduces restenosis and repeat revascularizations. DES is superior to BMS in all subsets of patients, but is particularly superior in the following cases:

1. One of five features indicative of a high risk of restenosis: diabetes, long lesions >20 mm, small vessel <3 mm, bifurcation stenting, chronic total occlusion (CTO).
   
2. In-stent restenosis.
   
3. Proximal LAD, left main, or multivessel disease, where restenosis is more consequential.

In complex lesions with several of those features, the restenosis rate remains high even with DES, ≥ 10%, with an even higher rate of target vessel revascularization related to disease progression outside the stented area.

BMS is very rarely used nowadays in any lesion subset as it has no advantage over DES, not even an earlier interruption of clopidogrel. In fact, 3 trials have shown that DES is superior to BMS even in high-bleeding-risk patients who stop clopidogrel at 1 month (ZEUS, LEADERS free, and SENIOR trials). Other studies have shown the safety of 1-to- 3-month DAPT vs. longer DAPT with DES (STOP-DAPT2, MASTER DAPT trials).

D. Rotational atherectomy (Rotablator) and orbital atherectomy

The atherectomy devices consist of a burr mounted over a catheter. As opposed to angioplasty, which expands the media without affecting plaque volume, atherectomy acts by removing plaque; it cuts the atheroma into small microparticles that are subsequently eliminated by the reticuloendothelial system. Those microparticles are typically very small, smaller than the red blood cells, and thus do not usually plug the microcirculation, as long as each atherectomy pass is limited to <30 seconds and vasodilators are used between passes. The Rotablator achieves better immediate results than angioplasty but is associated with a similar rate of late restenosis; the heat injury initiates both negative remodeling and neointimal hyperplasia. Atherectomy is useful in heavily calcified lesions that do not yield with high-pressure balloon angioplasty, meaning that the balloon never fully expands across the lesion. The lack of full balloon expansion is different from the more common elastic recoil, wherein the balloon expands but the vessel quickly collapses back. A stent should never be used for an unyielding lesion, i.e., a lesion where the balloon does not fully expand, as the stent will fail to appropriately expand and a catastrophic stent thrombosis may occur.

After balloon angioplasty fails to expand a lesion, the patient is brought back 4–6 weeks later for rotational atherectomy. It is not performed in a patient with dissection planes soon after angioplasty, as a major vessel perforation may be created by atherectomy.

E. Aspiration thrombectomy and distal embolic protection

As an adjunct to balloon and stenting, aspiration thrombectomy catheters (suction catheters: Penumbra, Pronto catheters) can be used to aspirate thrombi.

Large STEMI trials (TASTE and TOTAL) did not show any clinical benefit from the systematic use of thrombectomy and there was a small stroke hazard. The selective use of thrombectomy in STEMI patients with a large thrombus burden may still be justified.

The Angiojet rheolytic thrombectomy catheter may be used for a heavy thrombus burden, and is mostly valuable in lower extremity acute limb ischemia, often as an adjunct to local thrombolysis. It is associated with a high risk of embolization.

Wires with a distal basket, called distal embolic protection devices, are mainly useful during PCI of venous grafts where they reduce no-reflow and peri-PCI MI, not mortality (SAFER trial), and during femoro-popliteal atherectomy. They have not shown a benefit in acute MI: native coronary arteries have numerous side branches, such that a distal basket will prevent embolization into the distal vessel but will favor embolization into side branches, some of which are large. Also, placement of the distal protection may create arterial injury per se (especially if the basket moves during device manipulations).

A. Thrombosis

Stent thrombosis usually occurs within 1 month of stent placement but can occur later with DES or BMS. It is prevented with dual antiplatelet therapy and adequate stent expansion. Stent thrombosis typically leads to STEMI, the prognosis of which is much worse than that of spontaneously occurring STEMI (high clot burden), with a 15–45% risk of mortality.

Timing of stent thrombosis: acute (within 24 hours), subacute (between 1 and 30 days), late (between 1 and 12 months), and very late (>12 months). Late and very late stent thromboses are rare, but may be seen with DES or BMS, especially if both antiplatelet agents are discontinued. The risk of late stent thrombosis is probably similar with second-generation DES and BMS(0.2–0.4% per year) (DAPT trial). After the first month, stent thrombosis usually occurs ≥ 7 days after clopidogrel discontinuation, implying that brief clopidogrel interruption for non-cardiac surgery may be safe.

There are three categories of risk factors for stent thrombosis (Table 38.1).

B. Restenosis

C. Neoatherosclerosis

Beyond 1 year, atherosclerosis starts to grow over the stent neointima; this is referred to as neoatherosclerosis. As opposed to neointimal hyperplasia, neoatherosclerosis is characterized by a lipid-laden plaque with foamy macrophages and a fibrous cap. It tends to appear earlier and more frequently with DES (30–50% at 1–2 years) than BMS (16% at 5 years). It explains late restenosis and partly explains the very late stent thrombosis seen with both types of stents.

D. Mechanisms of recurrent target vessel disease after stenting (Table 38.2)

A. Multivessel PCI

According to clinical trials, multivessel PCI compares favorably with CABG if the stenoses are technically amenable to PCI. The presence of a chronic total occlusion or one or more technically difficult lesions or long lesions (angiographic SYNTAX score ≥ 23) should favor CABG, especially because CABG provides a more complete revascularization. Diabetes favors CABG as well, as CABG provides lower rate of repeat revascularization and better survival compared with PCI in diabetes.

B. Long lesions and diffuse disease (multiple lesions in series)

Normally, the stent should be implanted from normal reference to normal reference if possible (starting 2 mm before and 2 mm after the lesion); this will avoid edge dissections and edge plaque shift. The reference should have <40% plaque burden on IVUS. In case of diffuse disease with tandem lesions in series, the most severe lesions are stented (spot stenting), while the intermediate ones may be left untreated. FFR may be used to guide PCI in diffuse disease: stent the areas where the pressure drop is largest, as assessed by FFR pullback tracing. FFR may be reassessed after stenting each lesion. If possible, the distal lesion should be stented first, followed by the proximal lesion; this obviates the need to cross the proximal stent with the distal stent.

While DES length does not increase restenosis rate as strongly as BMS length, very long stenting (“full metal jacket”), even with DES, is associated with a high risk of restenosis and is preferably avoided. Furthermore, long stenting may compromise future placement of bypass grafts and is preferably avoided, especially in the LAD.

C. Bifurcation lesions

Side branch (SB) narrowing may occur after main branch (MB) PCI. This is related to plaque shift (snowplow phenomenon). The likelihood of this depends on the degree of ostial SB narrowing. When the ostium of the SB has >50% stenosis (= true bifurcation lesion), there is a risk of SB occlusion with MB stenting (14–35%). Bifurcation stenoses are approached as follows:

• Double wiring, i.e., wiring of both SB and MB, is indicated when the SB is large (>2–2.5 mm) and has significant disease. Occasionally, if the supplied territory is very large, SB is wired even in the absence of significant disease.
  
• Pre-dilate SB if it is significantly diseased at baseline.
  
• Avoid planned dual stenting (NORDIC, BBC-1, NORDIC 4, and PERFECT), except when SB is very large with diffuse (not just ostial) disease.
  
• Post-dilate SB only if occlusion or impaired flow occurs (NORDIC 1 and 3), or ± severe narrowing (>90%) develops (BBC-1, PERFECT). Avoid post-dilating a side branch that was not severely diseased at baseline, even if it looks severely narrow after stenting. CROSS study shows that post-dilatation of SB that only narrows after MB stenting is harmful, with more MB restenosis and no SB benefit over the long-term.
  
• Stent SB only if occlusion, poor flow, or +/- severe narrowing (>90%) persists despite post-dilatation. SB is more readily post-dilated or stented when it is large (e.g., LCx in distal left main stenting). Indeed, a planned 2-stent strategy has only improved outcomes in 1 trial, DK-CRUSH VI, which recruited true LM bifurcations. After dual stenting, final kissing balloon inflation is necessary to resolve MB stent distortion.

FFR studies have shown that SB narrowing that occurs after stenting is often functionally non-significant (~75% of the time). The narrowing is partly due to a benign geometric kinking of SB when MB is straightened by the stent. In addition, provisional rather than planned stenting is associated with a lower periprocedural MI and similar long-term outcomes. Stent-induced occlusion of a large branch may result in significant myocardial ischemia/necrosis, though in most patients the long-term prognosis is good and many initially occluded SBs are patent on follow-up.

D. Chronic total occlusion (CTO)

CTO implies a total coronary occlusion that is >3 months old. An occlusion may develop progressively over time and lead to an insidious, chronic angina presentation. It may also develop or progress acutely/subacutely leading to STEMI or NSTEMI; when this acute occlusion is not treated early on, it enters the category of CTO 3 months later. Well-developed collaterals may provide flow equivalent to a 50–90% stenosis, which helps maintain myocardial viability and prevents resting myocardial ischemia. When the coronary occlusion develops progressively rather than abruptly, the myocardial function may be normal at rest or depressed, but not infarcted (hibernating rather than necrotic myocardium). Hibernation is distinguished from infarction by the persistence of angina, lack of Q waves, lack of echocardiographic thinning, and ischemia or >50% uptake on nuclear imaging. MRI may be used when viability is questionable.

CTO PCI is indicated in symptomatic patients whose angina is refractory to medical therapy. Four randomized trials have shown that CTO recanalization does not reduce the long-term risk of death or MI. It does not improve overall wall motion or EF (Explore trial), and does not clearly improve the segmental motion of dysfunctional segments (REVASC trial). It only improves angina and quality of life, as per Euro-CTO trial. In addition, opening an occlusion that was associated with a large transmural infarct (STEMI or Q-wave MI), and a myocardium that is now akinetic and minimally ischemic, is not beneficial and is not indicated (OAT trial). The success rate of CTO PCI can exceed 80% when using a hybrid algorithm, with mortality rates of 0–2% (see Question 7). Restenosis rate is markedly reduced with DES (~10%).

Six main features predict failure of the antegrade approach to CTO PCI (the first 5 are J-CTO score):

• Any visible calcification across the CTO.
  
• Sharp angle/tortuosity >45° within the CTO site.
  
• Presence of a side branch at the occlusion point without any stump (the wire will preferentially go into the side branch). This, or the presence of thin bridging collaterals, defines an “ambiguous proximal cap”.
  
• Long CTO >2 cm.
  
• Failed prior PCI attempt
  
• Presence of a network of thin bridging collaterals around the occlusion (caput medusa). These are sometimes hard to distinguish from intralesional microchannels, which, to the contrary, predict a high success rate. CTO with antegrade, microchannel flow is called functional CTO; the vessel continues to opacify past the CTO in an antegrade fashion.

Polymer-coated wires may be used in a functional CTO to slide through the microchannels. Otherwise, drilling wires with progressively heavier tips are used, with the support of a catheter advanced to the occlusion site. Subintimal passage and distal reentry into the true lumen is the preferred technique when multiple adverse features are present, especially length >2 cm. Double arterial access with engagement of the contralateral coronary is usually necessary. Contralateral injections allow collateral retrograde filling of the index vessel and thus proper visualization of the wire progression. Antegrade vessel opacification usually ceases once the wire and catheter are advanced to the CTO site, making contralateral guidance critical in most cases.

A. Mortality

Mortality after diagnostic angiography is ~0.1%, but this increases significantly in severe AS or left main disease (up to 0.5%), where any transient arrhythmia or hypotension, sometimes induced by sedation, may initiate an irreversible, vicious circle of myocardial ischemia and further hypotension. Also, in left main disease, the catheter abutting a left main plaque may create a left main dissection.

Mortality after PCI is <1% (~0.5%). Complications and mortality after PCI are related to both the patient and the lesion characteristics:

1. Patient characteristics: ACS presentation, unstable hemodynamics/emergent case, severe HF or severe LV dysfunction, older age, CKD
   
2. Lesion characteristics: severe calcifications, tortuosity, multivessel CAD, CTO, degenerated venous graft (especially left venous graft, where guide support is difficult)

Patient characteristics are a stronger determinant of complications than lesion characteristics.

B. Coronary dissection

A coronary dissection is seen after 30% of balloon angioplasties. It may also occur at the edges of a stent, when overinflation of the stent’s balloon damages the edges. The frequency of stent edge dissection is 5% angiographically, 10% by IVUS, and 30% by OCT (the latter frequently detects some small and clinically insignificant dissections). Coronary dissection is more common in diffuse or calcified disease. Coronary dissection may also result from wire manipulations or from a deep-seated guiding catheter. Dissections typically propagate distally (antegradely).

C. Periprocedural MI

Two types of periprocedural MI may be seen:

1. Q-wave transmural MI occurs in ~1% of PCIs. It is often the result of abrupt vessel closure, thrombus formation, or occlusion of a major side branch.
   
2. Non-Q-wave MI is usually due to either:
   
   
   
   1. Distal microembolization, microvascular spasm, and microcellular reperfusion injury in a successfully opened vessel. A sluggish flow in spite of a patent epicardial artery (TIMI 2 flow or less) is called no reflow.
      
   2. Loss of small side branches.
      
   3. Transient procedural complications, such as thrombus or flow-limiting dissection.

No reflow is reduced with the use of IIb–IIIa inhibitors, and with the use of filter devices during venous graft interventions. Once it occurs, it is treated with intracoronary adenosine, verapamil, or nitroprusside.

According to the 2018 ACC universal MI definition, periprocedural non-Q-wave MI is defined by an increase in troponin I to >5× the upper limit of normal, along with: (1) prolonged chest pain, or (2) ischemic ST changes or new Q waves, or (3) new wall motion abnormality, or (4) angiographic evidence of procedural, flow-limiting complications. However, as opposed to the more recognized prognostic value of post-PCI CK-MB elevation (>3–10× normal), the degree of troponin rise used in the definition does not clearly carry an independent negative prognostic value. Therefore, an expert consensus document has suggested the use of CK-MB elevation ≥ 10× normal or troponin I elevation ≥ 70× normal to define periprocedural MI. Lesser degrees of CK-MB elevations, as well as isolated troponin elevation post-PCI, are of questionable prognostic value.

The above definition applies for patients with normal baseline biomarkers. If the biomarker values are elevated but stable or falling, periprocedural MI is diagnosed by ≥ 20% reincrease of a previously downward trending troponin or CK-MB value. In patients who present with elevated biomarkers that are on the rise and who undergo PCI within 24 hours of presentation, periprocedural MI is defined by the occurrence of procedural thrombotic/occlusive complications, prolonged ischemic symptoms ≥ 30 minutes, new ischemic ST abnormalities, or Q waves after PCI, along with an increase of the next troponin I or CK-MB by at least 50% (this is the definition used in trials, e.g., ACUITY, EARLY ACS, CHAMPION Phoenix trials).

The increase in biomarkers reflects a more extensive and unstable atherosclerotic burden that predisposes to future ischemic events. The latter is likely the main reason for survival impairment, rather than the minor injury induced by PCI. This explains why the unadjusted increase in mortality with periprocedural MI is more striking than the adjusted increase in mortality, and that only marked biomarker rise post-PCI carries an independent prognostic value.

Note that spontaneous NSTEMI carries a much stronger independent prognostic value than periprocedural MI, despite less biomarker elevation (3× higher mortality, according to ACUITY trial data). In fact, in NSTEMI, the adverse outcome is related not only to the minor myocardial injury but to the ruptured, active plaques that carry a high risk of a large infarction, which is not the case in controlled periprocedural MI.

Large periprocedural MI (CK-MB>5-10x normal) currently occurs in ~2.5% of PCI cases and the incidence is reduced with the current antiplatelet regimens. Most periprocedural MIs are clinically silent. Cardiac markers should be checked 12–24 hours after all complicated PCIs and in patients with ischemic symptoms during or after PCI. Cardiac markers may also be checked routinely, 12–24 hours after any PCI (class IIa recommendation).

D. Coronary perforation