Complete Versus Incomplete Revascularization
Definition of complete revascularization
There is no universally accepted definition for complete revascularization (CR). Ong et al characterized CR as anatomic/unconditional (all stenotic vessels are revascularized), anatomic/conditional (all stenotic vessels greater than a certain diameter are revascularized), functional (all stenotic vessels supplying viable myocardium are revascularized), and numerical (for CABG, the number of graft anastomoses equals the number of major diseased vessels) .
CABG studies tend to define CR as a graft to each major coronary artery with >50% stenosis. PCI studies have defined CR to include arteries with diameter greater than 1.5 , 2.0 , 2.25 , or 2.75 mm  with >50 or >70% stenoses. The Writing Group suggests the following definition of CR relevant to PCI: “revascularization of all significant arteries (as assessed by the angiographer) that threaten viable myocardium with stenoses either >70% diameter narrowing by angiography or of hemodynamic significance by stress testing or invasive assessment.” Artery size is not included in this definition, as there is no consensus on the size of an artery considered “significant,” and artery caliber is difficult to assess due to the diffuse nature of atherosclerosis.
Outcomes after CR (vs. incomplete revascularization)
CR is relevant to the extent that CR improves clinical outcomes. However, the clinical benefits of CR using angiographic criteria are uncertain, with some [65, 67–70] but not all [71–73] studies reporting superior outcomes with CR. CR does reduce the incidence of subsequent CABG [66, 74] or subsequent PCI  (Table II). CR using FFR testing to identify significant stenoses improved outcomes compared with CR based only on angiographic analysis . If an incomplete PCI revascularization strategy is used, stress testing may identify a high-risk subset of patients for whom additional revascularization procedures reduce subsequent ischemic events [86, 87].
Table II. Impact of Completeness of Revascularization by PCI on Subsequent Outcomes
|Bell ||1990||867||Balloon angioplasty||Stable angina||41||2.2||Similar incidence of death. Angina and CABG more frequent with IR|
|Bourassa ||1998||757||Balloon angioplasty||Stable angina and ACS||17||9||Similar incidence of death and MI. CABG more frequent with IR|
|Bourassa ||1999||896||Balloon angioplasty||Stable angina and ACS||65||5||Similar incidence of death and MI|
|Mariani ||2001||208||Balloon angioplasty and bare metal stents||Stable angina and ACS||24||1||Similar incidence of death, MI, and repeat revascularization|
|Kloeter ||2001||250||Balloon angioplasty and bare metal stents||Stable CAD||40||2.5||Similar incidence of cardiac events. Revascularization less frequent with IR|
|Van den Brand ||2002||576||Bare metal stents||Stable angina and ACS||70||1||Similar incidence of death or MI. CABG more frequent with IR|
|Brener ||2002||290||Bare metal stents||NSTE ACS||23||0.5||Similar incidence of death or MI|
|Nikolsky ||2004||352||Balloon angioplasty and bare metal stents||Stable angina and ACS||26.7||3.1||Higher incidence of death with IR|
|Palmer ||2004||151||Bare metal stents||NSTE ACS||47|| ||Similar incidence of death or MI. Angina more frequent with IR|
|Ijsselmuiden ||2004||219||Bare metal stents||Stable angina and ACS||51 (randomized trial)||4.6||Similar incidence of death, MI, CABG|
|Hannan ||2006||21,945||Bare metal stents||Stable CAD patients||31.1||3||Higher mortality with IR|
|Kong ||2006||1,982||Bare metal stents||Acute MI||32||In-hospital||Higher in-hospital mortality with IR|
|Kalarus ||2007||798||Bare metal and drug-eluting stents||Acute MI||24.2||1||Death and MACE more frequent with IR|
|Srinivas ||2007||1,781||Bare metal and drug-eluting stents||Both stable angina and ACS||17.7||1||Similar incidence of death and repeat revascularization|
|Shishehbor ||2007||1,240||Bare metal stents||NSTE ACS||39||2.3||Similar incidence of death and MI|
|Tamburino ||2008||508||Drug-eluting stents||Stable angina and ACS||41.7||2.25||Death, myocardial infarction, and target vessel revascularization more frequent with IR|
|Varani ||2008||399||Bare metal and drug-eluting stents||STEMI||46||30 days||Higher mortality with IR|
|Qarawani ||2008||120||Balloon angioplasty and bare metal stents||STEMI||79||1||Similar mortality. Reinfarction more frequent with IR|
|Hannan ||2009||11,294||Drug-eluting stents||Stable CAD patients||31||1.5||Higher mortality with IR|
|Yang ||2010||324||Bare metal and drug-eluting stents||ACS||31||1.5||Similar incidence of death, myocardial infarction, and repeat revascularization|
|Politi ||2010||214||Bare metal and drug-eluting stents||STEMI||66||2.5||More frequent MACE with IR|
One-Stage Versus Multistage PCI for MVCAD
Multivessel PCI can be performed in one or multiple stages. The second stage of a multistage PCI may be planned or unanticipated. This article focuses only on multivessel PCI that is planned to occur during a single stage or multiple stages.
One-stage multivessel PCI is reasonable when the following conditions have been met: (a) multiple vessels have hemodynamically significant lesions (either angiographically severe or, if intermediate, proven by stress testing or invasive testing to be significant), (b) indications for PCI are present (e.g., to relieve symptoms in stable angina or to prevent death or recurrent ischemic events for acute coronary syndrome patients), (c) adequate informed consent was obtained and consideration of alternatives has occurred, (d) the first stage of PCI is uncomplicated and without excessive radiation or contrast doses, (e) the patient and the operator are willing to proceed with multivessel PCI, and (f) the impact of resulting delays for other patients and operators has been considered.
Strategies of one-stage versus multistage PCI are difficult to compare by retrospective analysis of large databases. Databases do not easily distinguish a planned staged PCI from an unplanned multisession PCI, where the second session is due to acute closure, early restenosis, or an additional significant stenosis detected by further invasive testing at the time of the initial procedure. Also, databases cannot identify a planned multistage PCI for which the second session is aborted due to complications from the first stage, complete relief of symptoms, or patient preference.
Multivessel PCI in the same session as diagnostic catheterization
PCI is frequently undertaken at the time of diagnostic catheterization, termed “ad hoc” PCI. This is the most convenient for patients, and compared with a staged strategy, it is much preferred by patients. However, when ad hoc PCI is multivessel, there are special risks to consider [88, 89]. First, contrast and radiation used during the diagnostic catheterization may limit the additional doses that can be used during PCI. Second, the complexity of decision making increases with the number of lesions and vessels considered for PCI. Ad hoc multivessel PCI thus requires complex decision making that may not be optimal in the ad hoc scenario. Third, the informed consent before diagnostic catheterization may be inadequate for multivessel PCI. Patients scheduled for “cath possible PCI” should be informed about the risks of average single-vessel PCI, but it is unlikely that in-depth discussions of the pros, cons, and risks of multivessel PCI and the alternatives of medical therapy and bypass surgery have occurred. Fourth, logistical concerns may pose a special problem. If the patient is scheduled for a diagnostic catheterization time slot, the laboratory must be able to adjust to the addition of a multivessel PCI without undo delay to other patients, physicians, and catheterization laboratory staff. Finally, multivessel PCI done ad hoc does not allow input from cardiac surgeons, other specialists, or family members that might contribute to optimal decision making by the patient and interventionist. Recent guidelines have strongly encouraged a “heart team” approach for patients with unprotected left main or complex MVCAD [17, 20]. The “heart team” approach includes giving the patient, interventionist, and cardiac surgeon an opportunity to talk about various revascularization strategies. This cannot occur when PCI is performed ad hoc at the time of diagnostic catheterization.
Necessity of the second stage of a multistage PCI
When PCI of the primary stenosis is completed, the interventionist must consider whether to proceed to a second stenosis. Even if a second stenosis is proven to be hemodynamically significant by FFR testing, the clinical need for additional vessel PCI may remain questionable and a rational strategy may be to defer the second PCI. This allows additional time to reassess the patient's symptoms and their improvement after the first PCI, as well as to provide a trial of medical therapy or additional testing to confirm the need for the second procedure. A second informed consent process is necessary before proceeding with the second stage.
Unfavorable risk-to-benefit ratio for the second stage of a multistage PCI
Occasionally, the risk-to-benefit ratio for PCI is favorable for one lesion but unfavorable for additional lesions. For example, a patient with a 95% type A proximal right coronary lesion and a 70% type C distal circumflex calcified bifurcation lesion may gain complete relief of symptoms with right coronary artery PCI. The safest approach may be the single-vessel PCI, with a second stage only if needed for refractory angina.