Left Main Coronary Artery Angioplasty 

Abstract & Commentary

As techniques, equipment, and adjuvant therapy have evolved in the 25 years since the first balloon angioplasty was performed, the application of percutaneous coronary intervention (PCI) has expanded widely. While once reserved for uncomplicated lesions in patients with single vessel disease and limited comorbidity, PCI is now performed in a wide variety of clinical and pathoanatomic scenarios. For example, the current generation of coronary stents, with the addition of atheroablative therapy, intravascular ultrasound imaging catheters, and glycoprotein IIb/IIIa inhibitors, in particular, have contributed to make previously high-risk, low success rate lesions, such as those with heavy calcium, large thrombus burden, eccentricity or tortuosity, safely approachable by the experienced interventionalist, even in the setting of acute myocardial infarction, multivessel disease, or severely reduced left ventricular systolic function. In contrast, the presence of left main coronary artery (LMCA) stenosis remains a relative contraindication to PCI in most clinical settings. In fact, the most recent ACC/AHA Guidelines for PCI,1 list left main disease (particularly in the patient who is a candidate for coronary artery bypass graft surgery) as a class 3 condition for PCI (condition for which there is evidence and/or general agreement that the procedure is not useful/effective, and in some cases may be harmful) with level of evidence B (data derived from a single randomized trial or nonrandomized studies). However, despite these recommendations, there is some evidence that PCI of the unprotected LMCA might be effective and appropriate therapy in certain subsets of patients in need of coronary revascularization.

Tan and colleagues present the long-term results of the ULTIMA multicenter registry of patients undergoing unprotected LMCA PCI. This paper describes the outcomes of 279 consecutive patients between July 1993 and July 1998 at 25 high-volume clinical centers. Forty-six percent of these patients were believed to be inoperable or at high risk for poor surgical outcomes, and the majority of the remaining patients received PCI due to patient preference. Eighty nine (32%) patients were characterized as low risk, defined as age < 65 years, left ventricular ejection fraction (LVEF) > 30%, and absence of acute MI or cardiogenic shock. Of these, 65.2% of patients received stenting as primary therapy, and an additional 3.6% received stenting for a bailout indication. Almost 30% received atheroablative therapy (rotational atherectomy or directional atherectomy) and 46% of procedures were performed with intra-aortic balloon counterpulsation support. Of note, only 4.3% of patients received treatment with a glycoprotein IIb-IIIa inhibitor (abciximab).

Thirty-eight patients died in-hospital. At 1 year of follow-up, the rate of total mortality was 24%, cardiac death was 20%, MI was 9.8%, CABG was 9.4%, and repeat PCI was 24%. After the index hospitalization, the majority of adverse events occurred within the first 3 months of follow-up. By multivariate analysis, significant predictors of mortality were LVEF < 30% (hazard ratio [HR] = 4.21), severe mitral regurgitation (HR = 3.66), cardiogenic shock (HR = 3.56), creatinine > 2 mg/dL (HR = 3.1), and severe lesion calcification (HR = 2.32). In contrast, there were no in-hospital deaths among patients in the low-risk subgroup, the 1-year actuarial incidence of death was 3.4% and of MI was 2.3%, with no deaths or MI occurring beyond 4 months after hospital discharge. Among low-risk patients, repeat revascularization rates were comparable with 11.4% undergoing CABG and 20.4% undergoing repeat PCI within 1 year.

Tan et al comment that patient selection remains of utmost importance for both the surgeon and the interventional operator, but that the low-risk patients are likely to have good outcomes with either LMCA PCI or CABG. They conclude that, despite the limitations of a study based on a registry population, LMCA PCI might be considered in low-risk patients as already outlined, in addition to those patients deemed inoperable. Due to the clustering of adverse events early in the follow-up period, they also recommend routine surveillance angiography be performed in all patients at 2 and 4 months after LMCA PCI to evaluate for restenosis (Tan WA, et al. Circulation. 2001;104:1609-1614).

Comment by Sarah M. Vernon, MD

PCI for LMCA disease remains controversial in part due to well-established data demonstrating survival benefit for patients with LMCA disease undergoing CABG, with overall low rates of in-hospital and 1 year mortality. In addition, earlier reports from the ULTIMA registry, raised concern because of high rates of procedural complications and early mortality.2 However, as in the present report, these outcomes are at least in part attributable to extremely high-risk patient subsets including advanced age, acute MI, cardiogenic shock, and patients deemed to be inoperable, making comparison between the 2 methods of revascularization difficult, particularly for high-risk patients. In these patients, particularly those felt not to be surgical candidates, LMCA PCI, while high risk, remains by definition "the lesser of 2 evils." The present report suggests that, in lower-risk patients, PCI using more contemporary therapy (albeit infrequent use of platelet inhibitors) may represent a truly viable alternative to bypass surgery. Further evidence that PCI might be an acceptable approach in low-risk patients also comes from a recent publication by Park and associates, who report a 2-year survival rate of 97% and angiographic restenosis in 19% in 127 patients undergoing LMCA stenting.3 They also suggest that angiographic outcomes can be optimized using IVUS guidance and atheroablative debulking prior to stent implantation. While it is unlikely that the use of LMCA PCI will ever be clearly established in a randomized clinical trial, accumulating evidence suggests that in time PCI may ultimately prove to be the procedure of choice in some subsets of patients with LMCA stenosis.


1 Smith SC, et al. J Am Coll Cardiol. 2001;37: 2215-2238.

2 Ellis SG, et al. Circulation. 1997;96:3867-3872.

3. Park S, et al. J Am Coll Cardiol. 2001;38:1054-1060.