Effect of Completeness of Revascularization on Clinical Outcome in Patients with Multivessel Disease Presenting with Unstable An

ABSTRACT: We evaluated the current short- and medium-term outcomes of complete revascularization, compared to “culprit” lesion percutaneous coronary intervention (PCI), in patients with multivessel coronary disease presenting with unstable angina. One hundred fifty-one patients with multivessel coronary disease presented to a tertiary cardiothoracic center with unstable angina/non-ST elevation myocardial infarction (UA/NSTEMI) between January 2000 and September 2001. In group A (n = 71), the intended strategy was complete revascularization by multivessel PCI. In group B (n = 80), culprit lesion PCI was intended despite the presence of other lesions amenable to PCI (B1) or due to confounding anatomical factors (B2). Clinical variables and endpoints were collected from patient notes, a dedicated database and telephone follow-up, and included recurrent stable and unstable angina, need for repeat PCI or elective coronary artery bypass graft, incidence of non-fatal myocardial infarction (MI) and death. Baseline characteristics were similar in each group. Procedural success was achieved in over 95% of cases in both groups with high stent implantation rates (> 96%). There was no observed difference in mortality or incidence of MI between the groups. Compared to group A, more patients in group B1 had residual angina [22.8% (13/57) versus 9.9% (7/71); p = 0.041] and required further PCI [17.5% (10/57) versus 7.0% (5/71); p = 0.045]. There was a non-significant trend toward fewer readmissions for UA and less long-term antianginal medication in group A [38.0% (27/71) versus 52.6% (30/57); p = 0.043]. Complete and culprit lesion revascularization by PCI are safe methods of treating patients with multivessel coronary disease presenting with UA/NSTEMI. Reductions in residual angina, repeat PCI and need for antianginal therapies suggest that complete revascularization should be the strategy of choice when possible. J INVAS CARDIOL 2004;16:185–188 Key words: multivessel coronary disease, percutaneous coronary intervention, unstable angina Percutaneous coronary intervention (PCI) is an effective treatment for symptomatic, stable angina in patients with multivessel coronary artery disease and provides a less invasive alternative to coronary artery bypass grafting (CABG).^1–3 Although partial revascularization may provide good symptom relief, the goal of multivessel PCI is complete revascularization, unless there are anatomic limitations such as chronic total occlusions or an adverse lesion morphology.^4,5 In patients with multivessel disease presenting with unstable angina (UA), incomplete, “culprit” lesion, revascularization is often felt to be most appropriate given the relative hemodynamic instability and increased risks of acute vessel closure due to spasm, vessel dissection and thrombosis. In recent years, these complications have declined substantially as intracoronary stenting rates have increased and the routine use of intravenous glycoprotein (GP) IIb/IIIa inhibitors has become more widespread. This retrospective, single-center study evaluates the current short- and medium-term outcomes of the strategy of complete revascularization compared to culprit lesion PCI, in patients with multivessel coronary disease presenting with UA. Methods Study population. From January 2000 to September 2001, four hundred ninety-seven patients underwent PCI at our center following presentation with an acute coronary syndrome [UA or non-ST elevation myocardial infarction (NSTEMI)], of which 219 (44.1%) had multivessel disease, defined as coronary lesions of greater than 50% diameter stenosis in more than 1 of the major epicardial vessels or their major sidebranches. Of these 219 patients, fifteen were initially excluded due to the presence of lesions in right coronary arteries where a left dominant system existed, since these are not considered significant by the Coronary Artery Surgery Study algorithm.⁶ A further 40 were excluded due to a history of previous CABG or the presence of significant left main coronary artery disease, as both these features can confound the assessment of coronary artery disease and influence the PCI strategy.⁶ A further 13 patients were excluded due to lack of in-hospital and follow-up information. Therefore, one hundred fifty-one patients were included in this analysis. Acute coronary syndrome was typically defined as angina occurring at rest or early post-infarction angina, with raised cardiac markers (troponin) and/or electrocardiographic (ECG) changes, which recurred despite conventional antianginal therapy, including intravenous nitrates. Accepted ECG changes were persistent or labile ST-segment depression, transient ST elevation or minor T-wave changes. Baseline information, such as demographic and angiographic data, was obtained from patient records and from information entered prospectively onto a dedicated PCI database. The number of lesions and extent of stenosis were clearly described by the operator in all cases. The intended treatment strategy was noted and 2 groups were defined from this information. Group A included patients in whom the intended strategy was complete revascularization by multivessel PCI as either a single procedure (n = 68) or staged, where 1 or 2 lesions were treated at a later date (n = 3). Group B included those patients in whom single-vessel PCI to the culprit lesion only was intended (incomplete revascularization). This group was further subdivided into those patients who had lesions in other major vessels that were suitable for PCI (B1) and those in whom there were one or more lesions that were not attempted due to confounding anatomical factors, such as chronic total occlusions or extensively calcified lesions (B2). Procedural details. All procedures were performed, or closely supervised, by experienced Consultant Interventional Cardiologists. Percutaneous transluminal coronary angioplasty was performed via the femoral artery route in all cases. Guide catheters ranged from 6–8 French depending on operator choice. When incomplete revascularization was intended, the culprit lesion was clearly identified in 50% of cases by the ECG changes and/or the appearance of the lesion at angiography. The use of stents and/or infusions of GP IIb/IIIa inhibitors was left to the discretion of the cardiologist in charge of the patient. High-pressure (> 12 atmospheres) stent implantation was performed in all suitable lesions. Balloon angioplasty alone was performed on lesions in relatively small caliber vessels or in ostial or bifurcation lesions where stent implantation was not felt to be possible. Successful revascularization of a lesion by stent implantation or balloon angioplasty alone was defined as a residual stenosis less than 30% with no ischemic event during the hospitalization period. All patients received oral antiplatelet therapy (aspirin/clopidogrel) and heparin (intravenous/subcutaneous) as part of their initial management and during the procedure. PCI was performed immediately following diagnostic angiography (ad hoc), due to operator preference or clinical need, or within a few days (elective). Clinical endpoints measured. Details of procedural success and in-hospital complications were obtained from patient records. In all patients, follow-up information regarding angina recurrence, need for repeat PCI or elective CABG, incidence of non-fatal MI and death was available through follow-up clinic letters and hospitalization details (63%) or by patient self-report via telephone follow-up (37%). Statistical analysis. Fisher’s Exact test was used to assess differences in proportion of baseline characteristics and the proportion of patients experiencing an event during follow-up. For continuous variables, means and standard deviations were calculated followed by Students t-test to compare the groups. A p-value of Baseline characteristics. Table 1 displays the baseline characteristics for patients in each group. Compared to patients in whom the final goal was complete revascularization, patients with an intended strategy of incomplete revascularization had a slightly higher proportion of triple-vessel disease [21.1% (12/57) versus 11.3% (8/71); p = 0.053]. Patients in whom incomplete revascularization was performed because of anatomical limitations (group B2) had greater proportions of triple-vessel disease, lesion complexity, previous percutaneous revascularization, history of MI and congestive heart failure. In addition, this group tended to contain more patients who had severe, inoperable coronary disease, significant concomitant disease and lesions that were high risk for revascularization due to extensive disease in other major vessels. Procedural results. Table 2 demonstrates procedural details, such as stent implantation rates and GP IIb/IIIa receptor inhibitor use, and in-hospital and follow-up events of the two groups. The revascularization strategy selected was not influenced by the timing of the procedure in relation to the initial presentation (Figure 1). Procedural success was achieved in over 95% of cases in both groups. In group A, one patient suffered an acute MI following abrupt closure of a thrombotic proximal left circumflex coronary artery lesion that could not be reopened. In addition, one patient required repeat PCI two hours after the initial procedure because of abrupt vessel closure caused by a distal stent edge tear. In group B1, one patient who had undergone PCI suffered an MI in a different epicardial territory. One patient required urgent CABG surgery due to failure to cross a subtotal occlusion. High stent implantation rates (> 96%) occurred in both groups. One patient in group A and two patients in group B1 had successful balloon angioplasty alone to distal lesions in small caliber vessels. In-hospital events were slightly, although not significantly, higher when incomplete revascularization was performed for anatomical reasons (group B2). Clinical outcomes. During follow-up, there was no observed difference in mortality. The single death that did occur was due to chronic congestive heart failure. There were no significant differences in the incidence of MI between the groups, although a trend toward more surgical revascularization in group B2 existed. Compared to group A, significantly more patients in group B1 had residual angina [22.8% (13/57) versus 9.9% (7/71); p = 0.041] and required further PCI [17.5% (10/57) versus 7.0% (5/71); p = 0.045] to treat de novo lesions (60%) or in-stent restenosis (40%). All repeat PCIs performed in group A were for in-stent restenosis. There was a non-significant trend (p = 0.065) toward fewer readmissions for UA in group A. This group also required less long-term antianginal medication [38.0% (27/71) versus 52.6% (30/57); p = 0.043]. Overall, 85.9% (61/71) of patients in group A were free of major adverse coronary events at follow-up compared to 71.9% (41/57) in group B1. Discussion Approximately 30–40% of patients presenting with UA or NSTEMI have multivessel coronary artery disease.⁷ In some patients, anatomic limitations, primarily total occlusions, restrict the ability to carry out complete revascularization by PCI. However, even when the anatomy of multiple lesions is favorable, incomplete culprit lesion revascularization is frequently the strategy of choice for operators who decide to treat the lesion that is felt to be responsible for the clinical status of the patient while minimizing the procedural risks that result from the prothrombotic state, the presence of unstable, friable, atherosclerotic plaque, the relative hemodynamic instability and the increased arrhythmic tendency.⁴ Data from the NHLBI registry in the era of conventional balloon angioplasty demonstrated that patients presenting with UA or NSTEMI were much less likely to undergo complete revascularization as the intended strategy.⁸ Non-randomized comparisons of the degree of revascularization have demonstrated no difference in mortality and MI rates, but similar² or higher rates^9–12 of recurrent angina, subsequent CABG and exercise-induced ischemia when incomplete revascularization is undertaken. It is therefore probable that the relatively high restenosis rate observed following multivessel PTCA had less impact on recurrence of symptoms than progression of residual, untreated lesions. Stent implantation and the use of effective oral and intravenous antiplatelet therapies has improved the safety of PCI.^13,14 Multivessel disease can more readily be treated as the problems of abrupt vessel closure, thrombotic occlusion and dissection can be almost completely overcome, while more anatomically complex lesions can be approached. High procedural success rates, fewer in-hospital events and good long-term outcomes have been consistently reported for these patients.^15,16 Greater confidence in acute and long-term vessel patency enables all suitable lesions to be effectively treated during the initial procedure. This observational, non-randomized study compared the strategies of incomplete and complete revascularization in high-risk patients with UA and multivessel disease in the current setting of high stent implantation rates and relatively high use of intravenous antiplatelet agents. Unlike other studies, patients with confounding anatomical factors for complete revascularization were analyzed as a separate group, since in this group the operator has little option but to treat the culprit lesion with the knowledge that residual symptoms are likely. However, it was evident that 38% of patients underwent culprit lesion PCI alone despite significant lesions in other major vessels that were amenable to PCI. Complete revascularization was the initial strategy in only 47% of patients. Operator preference appeared to be the main reason for each approach, since patient co-morbidity, pattern of presentation and lesion complexity appeared similar between the groups. Although there was a trend toward more triple-vessel disease in the incomplete revascularization group, this was not predictive of a culprit lesion only approach. The timing of the procedure in relation to onset of symptoms, ongoing symptoms that were refractory to medical treatment, or hemodynamic instability may have influenced the strategy employed. However, it was evident that complete revascularization was performed to a similar degree in circumstances where some operators elected to stabilize the situation by treating the culprit lesion alone. As with previous studies, no differences in in-hospital or follow-up mortality or non-fatal MI rates were noted between complete and incomplete revascularization groups. However, significant benefit with respect to residual angina and the need for further revascularization was noted when complete revascularization was undertaken. In addition, less long-term anti-ischemic medication was required in this group. The widespread use of stenting in UA has enabled better acute and long-term results to be obtained compared to the early trials of multivessel PCI.^17,18 A recent study of 208 patients with multivessel disease who underwent either single or multiple lesion PCI demonstrated that similar risks of death or other major ischemic events occurred at 1-year follow-up, suggesting that incomplete revascularization is an acceptable strategy for these patients.¹⁹ Clearly, this is reassuring for those patients with confounding anatomical factors preventing complete revascularization. As demonstrated in our study, this group tends to include higher risk patients with more extensive coronary artery disease, greater left ventricular impairment, previous MI and concomitant disease. However, the relatively high incidence of recurrent angina and need for further revascularization procedures seen in both incomplete revascularization groups demonstrates that the degree of revascularization in patients with multivessel disease has an important influence on long-term outcome. These results resemble those of comparisons of PTCA and CABG where significantly less angina and repeat revascularization were noted when surgery was undertaken compared to PTCA, where incomplete revascularization occurred more frequently.²⁰ Study limitations. As this was an observational, non-randomized study performed in a single tertiary center, confirmation of our findings in a randomized, controlled trial setting would be desirable. However, it is our opinion that limited enthusiasm for such a trial would exist as interventionalists opt for complete revascularization, particularly when drug-eluting stents can be implanted. As this was a retrospective study, the intended strategy for revascularization was based upon data entered following the procedure. Ideally, prospective recording of the treatment strategy should have been performed. Conclusion. These results demonstrate that both complete and incomplete culprit lesion revascularization by PCI appear to be safe methods for treating patients with multivessel coronary disease presenting with UA or NSTEMI. However, in the current clinical setting, the reductions in residual angina, repeat PCI and need for multiple antianginal therapies observed when complete revascularization is undertaken suggest that this should be the strategy of choice where possible.