ABSTRACT: Coronary artery aneurysm (CAA) often occurs after percutaneous coronary intervention, and it could be recognized more often in coronary intervention with directional coronary atherectomy (DCA). However, it has been uncertain and the natural history of CAA after DCA remains obscure. Thus, we examined the clinical course after DCA. This study included 792 lesions in which a follow-up angiogram was completed at mid- or long-term (3 months [mos.] or more than 1 year after DCA). The mean average of the angiographic follow-up period was 24.8 mos. (range 3–128 mo.), and clinical follow-up period was 45.6 mos. (range 3 to 144 mos.). CAA was defined as 1.5 > DCA site diameter / reference diameter by quantitative coronary angiography (QCA). CAAs were detected in 21 lesions (2.7%). There was no significant difference in the target lesion revascularization rate between CAA and non-CAA lesion (19.0% vs. 24.6%). More than twice as many follow-up coronary angiograms were performed in 15 lesions among 21 CAA lesions. The mean duration of the follow-up angiograms was 19.6 mos. There was no significant difference in lumen diameter between the value of QCA at first and final follow-up coronary angiography (3.9 ± 0.9 vs. 4.0 ± 0.8 mm). Acute coronary events or coronary perforations did not occur at all. This study showed the frequency of DCArelated CAA and revealed acceptable short- or long-term prognosis. DCArelated CAA had not been progressing during the follow-up period. We concluded that almost all CAAs should be managed conservatively. J INVASIVE CARDIOL 2008;20:159–160

       Coronary artery aneurysms (CAAs) represent a relatively uncommon finding in patients who undergo percutaneous coronary intervention (PCI). However, the prevalence of CAAs has been shown to be higher in patients undergoing PCI utilizing directional coronary atherectomy (DCA) in whom CAAs have been observed in up to 10% of patients.^1–5 The higher prevalence of CAAs in patients undergoing DCA has led practitioners to question the prognostic implications of the occurrence of this epiphenomenon, which to date still remain uncertain. The aim of the present study was to investigate the short- and long-term impact of CAAs occurring in patients undergoing DCA.

Methods
       From February 1992 to 2004, standalone DCA was performed in 792 lesions (590 patients). All patients were on aspirin (81–100 mg per day) for at least 1 week before the procedure. During the procedure, heparin was given as a bolus 120 U/kg with additional boluses to 1000 U/hour. DCA was performed using a 6 Fr or 7 Fr graft Simpson AtheroCath, or Flexi-Cut (Abbott Vascular, Santa Clara, California). Intravascular ultrasound was performed in 712 lesions (89.9%). All patients included in this study underwent angiographic follow up, which was performed at 24.8 ± 22.3 months. Clinical follow up was extended to 45.6 ± 32.8 months.
       Serial quantitative coronary angiography (QCA) analysis was performed using the Cardiovascular Measurement System (CMS-MEDIS Medical Imaging System, Leiden, The Netherlands) after the procedure and at follow up after intracoronary administration of 2.5 mg of isosorbide dinitrate. Offline QCA was conducted using the same postprocedure and follow-up view that best revealed the CAA site. CAA was defined as coronary dilatation > 1.5 times larger compared with an adjacent healthy reference segment by QCA.^6,7
       Statistical analysis. Statistical analysis was performed using StatView 5.0 (SAS Institute). The chi-square test was used to evaluate categorical variables (i.e., restenosis rate). Continuous variables were expressed ± standard deviation for each measurement. Differences in continuous variables were assessed using the paired Student’s t-test (i.e., CAA size). A p-value < 0.05 was considered statistically significant.

Results

Table 2. Procedural results of percutaneous coronary intervention using DCA.

Table 1. Lesion location of coronary artery aneurysm after DCA.

       CAAs were detected in 21 of the 792 lesions treated (2.7%). Standalone DCA on the left anterior descending artery (LAD) was the most common intervention (Table 1). Procedural characteristics of patients with and without CAA are summarized in Table 2.
       During clinical follow up, a total of 3 deaths (1 cardiac and 2 noncardiac) occurred. The cardiac death occurred following an acute ST-elevation myocardial infarction, the culprit vessel of which was different from that of the location of the CAA. No other acute coronary events occurred in the CAA population. The 2 noncardiac deaths were due to malignancy. The target lesion revascularization (TLR) rate was similar in the CAA and non-CAA groups (19.0% vs. 24.6%; nonsignificant). More than twice as many follow-up coronary angiograms were performed in 15 of the 21 CAA lesions. The mean duration between the first follow up and the final coronary angiography of the 15 lesions was 19.6 ± 16.6 months. There was no difference in the size of the CAAs between the two (3.9 ± 0.9 mm vs. 4.0 ± 0.8; nonsignificant).

Discussion
       The present study reports on the largest series of patients who underwent DCA and serial angiographic follow up. The results of this report demonstrate an overall low prevalence of CAA formation following DCA compared with other studies, and importantly, show an overall benign outcome of this finding. In fact, CAA formation was not associated with the occurrence of acute ischemic complications and presented a rate of TLR comparable to lesions also treated with DCA, but that did not develop CAA. In addition, individuals in whom CAAs were identified showed a lack of progression of ectasia over time, as assessed by serial follow-up angiography.
       Prior reports have suggested thrombotic development at the site of CAA formation,^8–10 which has been attributed to hypoperfusion at the site of ectasia.¹¹ However, in our series of patients with CAA, there was no incidence of acute ischemic complications due to thrombotic occlusions. Of note, our patients were treated only with aspirin for secondary prevention of atherothrombotic complications. Such antithrombotic treatment was proven to provide protection against thrombotic complications despite the presence of a “nitch” potentially vulnerable to thrombosis.
       Previous series have noted a higher risk of event rates in patients undergoing DCA with stent implantation.¹² Animal studies have demonstrated arterial medial thinning at the sites of the stent struts. Therefore, it has been suggested that bare-metal stents may increase the risk of aneurysm formation due to radial forces at the site of the stent struts.¹³ However, in our series of patients treated with DCA and stent implantation (n = 99), there was no incidence of CAA. The favorable findings of our results may be attributed to our routine use of intravascular ultrasound (IVUS) during DCA, which precluded stent implantation if deep cuts were observed in order to avoid future development of CAAs. Because it has been shown that standalone DCA with IVUS guidance can safely achieve optimal angiographic results with low residual plaque mass, and this was associated with a low restenosis rate,¹⁴ IVUS should be used in all DCA cases with or without bare-metal stents in order to retain the feasibility and safety of DCA. In addition, although prior reports showed cases of CAA rupture,¹⁵ this was not observed in our series, corroborating the results of other groups.^6,16
       In our study, the TLR rate was low and the location of restenosis was proximal or distal to the CAA. This may be attributed to incomplete dilatation due to fear of CAA perforation.
       Clinical implications. We demonstrated the benefits of DCA in bifurcation lesions with large sidebranches, with acceptable longterm results of DCA for LAD proximal lesions (TLR 9.5 %) and no left circumflex artery ostium stenosis.¹⁷ Indeed, the patients who underwent standalone DCA did not need to take anti-platelet drugs (except for aspirin) and had no late thrombosis. We had already demonstrated that DCA prior to DES makes it possible to avoid complex stenting in bifurcated lesions and that there was no restenosis in left main trunk bifurcations.18 Thus, we believe that, even in the DES era, DCA is a feasible strategy in some lesions.

Conclusions
       The present study shows an overall low prevalence of CAA formation following DCA procedures that are judiciously performed, which implies the routine use of IVUS. CAA formation is benign and associated with favorable long-term clinical outcomes. These findings suggest the need for conservative management of CAA.