Author + information
- Received April 17, 1996
- Revision received August 14, 1996
- Accepted September 20, 1996
- Published online January 1, 1997.
- Barry L. Sharaf, MD, FACCB,*,
- David O. Williams, MD, FACCA,
- Nicholas J. MieleA,
- Robert P. McMahon, PhDB,
- Peter H. Stone, MD, FACCC,
- Preben Bjerregaard, MD, FACCD,
- Richard Davies, MD, FACCE,
- A.David Goldberg, MD, FACCF,
- Michael Parks, MDG,
- Carl J. Pepine, MD, FACCH,
- George Sopko, MDI,
- C.Richard Conti, MD, FACCH,
- for the ACIP Investigators1
- ↵*Dr. Barry L. Sharaf, Rhode Island Hospital, Division of Cardiology/APC 434A, 593 Eddy Street, Providence, Rhode Island 02903.
Objectives. The purpose of this Asymptomatic Cardiac Ischemia Pilot (ACIP) data bank study was to characterize angiographic features of coronary pathology of patients enrolled in the ACIP study.
Background. Ischemia during ambulatory electrocardiographic (AECG) monitoring is associated with increased morbidity and mortality. Reports relating AECG ischemia to severity or complexity of coronary artery disease are few in number and small in size and have produced conflicting results.
Methods. Coronary angiograms from patients with asymptomatic AECG ischemia enrolled in the ACIP study were reviewed at a central core laboratory. Quantitative measurement of percent stenosis and Thrombolysis in Myocardial Infarction flow grades were used to assess the severity of coronary artery disease. Lesions were also evaluated for the presence of intracoronary thrombus, ulceration and lumen contour as indicators of stenosis complexity. In addition, comparisons were made with 27 patients screened for the ACIP study, but who were found ineligible because they did not have AECG ischemia on 48-h Holter monitoring.
Results. A total of 329 (75%) of 439 patients with AECG ischemia had multivessel coronary artery disease. Proximal stenoses ≥50% diameter reduction were common in patients with AECG ischemia (62.2%), as were proximal stenoses ≥70% (38.7%). Features suggesting complex plaque were found in 50.1% of patients with AECG ischemia.
Conclusions. Multivessel coronary artery disease, severe proximal stenoses and features of complex plaque were observed frequently in patients who exhibited AECG ischemia. The presence of severe and complex coronary artery disease may explain, in part, the increased risk for adverse outcome associated with ischemia during activities of daily life.
(J Am Coll Cardiol 1997;29:78–84)>
Asymptomatic myocardial ischemic episodes detected by ambulatory ECG (AECG) monitoring are common events in patients with coronary artery disease (CAD). Recent reports have suggested a relation between asymptomatic ischemia recorded by AECG monitoring and adverse clinical outcomes ([1–14]).
Little is known about the details of the angiographic coronary pathology of patients with AECG ischemia. Reports relating AECG ischemia to the severity of CAD are few in number and small in size and have produced conflicting results ([14–19]). None of these studies has used central core laboratory quantitative angiographic review.
Recent pathologic data suggest that plaque rupture and thrombosis are common events leading not only to unstable clinical syndromes but also to progression of CAD in asymptomatic or stable patients ([20–26]). It is therefore important to investigate the relation between AECG ischemia and the presence of complex coronary artery lesions.
The Asymptomatic Cardiac Ischemia Pilot (ACIP) study investigated the efficacy and safety of therapy to alleviate ischemia, as well as the feasibility of a trial to compare the effect of therapies on clinical events ([27–30]). The trial provides a unique opportunity to examine the relation between coronary angiographic pathology and AECG ischemia in a well-characterized group of patients with stable CAD. The purpose of this study was to characterize, both quantitatively (lesion severity) and qualitatively (lesion complexity), the coronary angiographic pathology of patients enrolled in ACIP and to investigate the relation between coronary pathoanatomy and the presence of AECG ischemia.
The design, baseline characteristics and primary outcome data of patients in ACIP have been reported ([27–30]). Data are reported here for 439 ACIP patients. As previously reported (), because of unacceptable performance in the conduct of the trial at one clinical unit, it was determined that the data from that unit were not comparable to the data from the other clinical units. Therefore, all 60 patients from that clinical unit have been excluded from this report. Patients were eligible for inclusion if they met all the following criteria: one or more episodes of asymptomatic ischemia on 48-h ambulatory ECG; evidence of ischemia on stress testing; and revascularizable CAD. Coronary artery disease was defined as ≥50% diameter reduction of a major coronary artery or branch seen on a coronary angiogram. For the purposes of this data bank study, only 439 of 558 ACIP patients with angiograms performed within 90 days of the screening AECG (in the majority of cases within 4 weeks) were included. In almost all cases (≥95%), angiograms were performed electively in stable patients because of abnormal results of noninvasive testing. Reasons for exclusion from the trial were myocardial infarction within 4 weeks, coronary angioplasty within 6 months, coronary artery bypass surgery within 6 months, presence of unstable angina at study entry, heart failure with New York Heart Association functional class ≥III, significant noncardiac illness or inadequate control of angina on maximal medical therapy.
1.1 Coronary angiography
All coronary angiograms were reviewed in the central core laboratory. The cine film must have been of adequate quality to confirm eligibility and suitability for revascularization. Major coronary arteries were defined as the left anterior descending, left circumflex and right coronary arteries. Diagonal and obtuse marginal branches were considered major vessels if they supplied a large enough area of myocardium and were suitable for bypass surgery or percutaneous coronary angioplasty.
1.2 Severity of coronary artery disease
All coronary lesions in major arteries or branches were measured using an electronic cine projector–based “crosshair’s” technique (Vanguard Instrument Corp.). The technique is analogous to electronic calipers built into the hood of a cine projector, thus removing viewer parallax. This technique compares favorably with automated edge detection (). In our laboratory, the standard deviation of the mean difference in measured percent stenosis between the two methods was <8%. Percent stenosis was measured in a view demonstrating the most severe coronary artery narrowing. Flow was visually estimated by the method used in the Thrombolysis in Myocardial Infarction (TIMI) trial (). Stenoses with flow grade 0 or 1 were not measured but assigned a value of 100% and 99%, respectively. Interobserver variability in classifying flow grade in the core laboratory was low, with 85% of paired readings demonstrating complete agreement (r = 0.924). A random 5% of all films were interpreted twice for quality control. Collateral channels to the distal vessel beyond a stenosis were graded as none, partial or complete.
1.3 Stenosis complexity
Intracoronary thrombus was graded by a scoring system ranging from 0 to 4: grade 0 = no thrombus; grade 1 = possible thrombus where the lesion borders were hazy with a ground glass appearance; and grades 2 to 4 = definite thrombus defined as well-delineated discrete filling defects of increasing size. Grade 2 thrombus represents a filling defect with a greatest linear dimension ≤1/2 vessel diameter; grade 3 represents >1/2 but <2 vessel diameters in size; and grade 4, ≥2 vessel diameters in size. Interobserver variability with this scoring system is low, with 71% of paired readings demonstrating complete agreement and 23% differing by 1 score (r = 0.906). Lesion ulceration was graded from 0 to 2: grade 0 = no ulcer; grade 1 = possible ulcer with eccentric lesions with overhanging edges; and grade 2 = definite ulcer with a discrete outpocketing of contrast opacification. Lesion contour was graded from 0 to 2: grade 0 = smooth lumen borders; grade 1 = ill-defined borders; and grade 2 = irregular lumen borders. “Possible complex plaque” was considered present in lesions with any of the following criteria: a thrombus score ≥1 or an ulcer score ≥1 or a contour score ≥1. “Definite complex plaque” was considered present in lesions with any of the following criteria: a thrombus score ≥2 or an ulcer score ≥1 or a contour score of 2.
1.4 Statistical methods
Characteristics of patients with AECG ischemia were compared with a small sample of patients screened for ACIP, but without AECG ischemia using the chi-square test for categoric variables and the ttest or Wilcoxon rank-sum test for continuous variables. Odds ratios for the presence of AECG ischemia, given the presence of complex plaque, with and without adjustment for covariates, were calculated using logistic regression analysis. Percent diameter stenosis in lesions with and without complex plaque was compared using the generalized estimating equations method to account for correlation among lesions within the same patient (). To account for multiple comparisons in analyses, the ACIP protocol specified that p < 0.01 be considered as providing some evidence for the presence of differences, and p < 0.001 be considered as strong evidence.
2.1 Patient group
A total of 1,820 patients were screened for ACIP, 910 of whom had AECG ischemia, including 815 with at least one asymptomatic episode. From this group with asymptomatic AECG ischemia, 558 with a positive stress test and angiographically documented CAD amenable to revascularization were enrolled in ACIP ([27–29]). Thirty-three of the 558 patients had either prior coronary artery bypass graft surgery (n = 31) or an angiogram inadequate for analysis (n = 2) and were excluded from this study. An additional 86 patients underwent coronary angiography >90 days before their qualifying AECG visit and were also excluded. The remaining 439 ACIP patients comprise a group of patients with asymptomatic AECG ischemia included in this report. Among the 910 patients screened who did not have AECG ischemia, coronary angiograms were also reviewed in a second group of 40 patients with an abnormal exercise test, documented CAD and no prior coronary artery bypass graft surgery but no asymptomatic ischemia on 48-h AECG. Thirteen patients were excluded because more than 90 days elapsed from the coronary angiogram to the qualifying AECG. The remaining 27 ACIP-screened patients comprise a comparison group without symptomatic or asymptomatic AECG ischemia included in this report. This comparison group represents screened patients for whom angiograms could be obtained centrally and was not chosen by random sampling. Detailed clinical history data are not available for the 27 patients in this comparison group.
2.2 Clinical characteristics
Baseline characteristics for the group of patients with AECG ischemia are listed in Table 1. The mean age was 61.3 years, and the majority of patients were men. Over half the patients had angina in the 6 weeks before AECG screening. The patients had an average of 2.5 ± 2.8 ischemic episodes per 24 h and 21.4 ± 29.5 min of ischemia per 24 h.
2.3 Disease severity
Disease severity was characterized as the number of vessels with stenosis ≥50% diameter reduction, total number of stenoses ≥50% per patient, number of severe stenoses (≥95% diameter reduction) or occlusions and specifically by the presence or absence of proximal stenoses in a major artery (vs. branch). The number of vessels with stenosis ≥50% is depicted in Fig. 1. Approximately 75% of patients had multivessel CAD. Distributions of number of stenoses per patient are listed in Fig. 1. The majority of patients (81.5%) had two or more stenoses (≥50% diameter reduction), and 21.2% had two or more severe stenoses (≥95% diameter reduction).
Proximal (before major side branches) stenoses ≥50% in the left anterior descending, left circumflex or right coronary arteries were found frequently in patients with AECG ischemia (Fig. 2). Sixty-two percent of patients with AECG ischemia had at least one proximal stenosis ≥50%, and 38.7% of these patients had at least one proximal stenoses ≥70% diameter reduction. The most severe stenosis was 88.1 ± 14.2 (mean ± SD).
When comparing these patients with the group of ACIP-screened patients without AECG ischemia, the distribution of single-, double-, and triple-vessel disease was not different (p = 0.76). Proximal stenoses ≥50% were found more commonly in patients with AECG ischemia than in those without it (62.2% vs. 37.0%, respectively, p = 0.01), as were proximal stenoses ≥70% (38.7% vs. 18.5%, respectively, p = 0.04).
2.4 Features of complex plaque
Angiographic evidence for definite plaque ulceration was found in at least one lesion in 10.0% of patients with AECG ischemia. Definite thrombus was also infrequent; it was found in 3.0% of patients with AECG ischemia. At least one lesion meeting the angiographic criterion of “possible complex plaque” was found in 50.1% of patients with AECG ischemia, whereas 34.2% of patients with AECG ischemia had “definite complex plaque” (Fig. 3).
No patient without AECG ischemia had angiographic evidence of definite plaque ulceration or thrombus, and both possible (p = 0.02) and definite (p = 0.09) complex plaques were found less frequently than in patients with AECG ischemia (Fig. 3).
2.5 Angiographic findings in patients with and without angina and prior myocardial infarction
Compared with patients without angina, the number of coronary arteries with ≥50% stenosis was greater among ACIP patients with angina on the qualifying exercise treadmill test (p < 0.001) and tended to be more common among ACIP patients reporting angina within 6 weeks of the AECG (p = 0.04). Three-vessel disease was present in 102 (43%) of 238 patients with and in 60 (31%) of 193 patients without angina on the exercise test and among 110 (42%) of 262 patients with versus 53 (30%) of 177 patients without angina within 6 weeks of the AECG. The percentages of patients with proximal stenoses and complex plaque were similar among those with versus those without angina on the exercise test or reported within 6 weeks of the AECG (all p > 0.10). The distributions of number of arteries with stenosis ≥50%, presence of proximal stenosis, and presence of complex plaque were similar among patients with versus without a history of myocardial infarction (all p > 0.07).
2.6 Correlation of features of complex plaque with disease severity
When combining the two groups of patients with and without AECG ischemia (n = 466), associations were noted between the presence of “definite complex plaque” and several measures of disease severity (Table 2). Of patients with “definite complex plaque,” 45.2% had three-vessel disease compared with 32.8% of patients without “definite complex plaque” (p = 0.009). In addition, 70.3% of patients with complex plaque had three or more stenoses as opposed to only 51.8% of those without complex plaque (p < 0.001). There was a trend for patients with complex plaque to have more proximal stenoses (≥50% and ≥70% diameter reductions) than those without complex plaque (p = 0.02 and 0.03, respectively). The site of complex plaque, however, was not necessarily the site of the most severe stenosis. Percent diameter stenosis in lesions with and without complex plaque was compared using the generalized estimating equations method to account for correlation among lesions within the same patient. The mean (±SE) difference in percent stenosis between lesions with and without “definite complex plaque” was −0.5 ± 1.1 (p = 0.67); the mean (±SE) difference between lesions with and without “possible complex plaque” was −1.8 ± 1.0 (p = 0.07).
Logistic regression analysis was performed to examine the relation of angiographic variables with the presence or absence of AECG ischemia. After adjusting for the number of vessels with ≥50% stenosis and the presence of proximal lesions ≥50%, a trend toward an association remained between the presence of AECG ischemia and the presence of “possible complex plaque” (odds ratio [OR] 2.91, 99% confidence interval [CI] 0.88 to 9.54, p = 0.02). However, there was less evidence for an association between AECG ischemia and the presence of “definite complex plaque” after adjusting for severity of disease (OR 2.26, 99% CI 0.60 to 8.56, p = 0.11).
Among the 439 patients with AECG ischemia, the Wilcoxon rank-sum test was performed to examine the association between the number of ischemic episodes per 24 h and angiographic findings. No significant association was found between the number of ischemic episodes and the presence of possible complex plaque (p = 0.61), definite complex plaque (p = 0.86), presence of proximal stenosis ≥50% (p = 0.81) or number of vessels with stenosis ≥50% (p = 0.11). We noted a trend toward more ischemic episodes per 24-h with the presence of more stenoses ≥50% (p = 0.04) (Fig. 4).
There was no correlation between the severity of CAD, as measured by the number of vessels with ≥50% stenosis and 12-week, 6-month or 12-month adverse outcome (ischemic episode by AECG, hospital admission for angina, myocardial infarction or death). In addition, the presence of possible or definite complex plaque did not correlate with adverse outcome. Among patients with an adverse outcome, 179 (43.3%) of 415 had a proximal stenosis ≥70% versus 41 (33.6%) of 123 patients without an adverse outcome (p = 0.06).
This ACIP data bank study characterized angiographic severity and complexity of CAD in patients with asymptomatic ambulatory cardiac ischemia. The ACIP enrollment criteria, requiring an abnormal exercise test and AECG ischemia, identified a group of patients with critical and complex CAD.
3.1 Relation between angiographic disease severity and asymptomatic cardiac ischemia
Other investigators have attempted to measure disease severity and to correlate this with the presence of AECG ischemia. Deedwania and Carbajal () evaluated 107 patients with stable angina with AECG monitoring. No significant difference in severity of CAD was found. Fifty-three percent of patients with AECG ischemia had three-vessel CAD compared with 43% of those patients without AECG ischemia. Yeung et al. () studied 138 patients with stable angina. Although there was a trend toward more patients with AECG ischemia having three-vessel CAD than those without AECG ischemia (42% vs. 26%, respectively), this difference was not significant. Kunkes et al. () studied a heterogeneous group of 50 patients with stable angina, chest pain after myocardial infarction and chest pain of uncertain etiology. Those with no CAD or single-vessel CAD less frequently had AECG ischemia than those with two- and three-vessel disease (44% vs. 100%, respectively). Of note, over one-third of the patients in their study had no significant CAD (defined as ≥50% stenosis), yet 38% of these patients had AECG ischemia. Common to these prior reports was a relatively simple definition of CAD severity. When ACIP-enrolled patients were analyzed in a central angiographic core laboratory, severe disease was noted with 75% having multivessel and 58% with at least one severe stenosis (≥95%) giving rise to abnormal flow (TIMI flow grade ≤2). Like these other reports, we found that when measuring severity of disease either by total number of vessels with stenosis ≥50%, total number of lesions ≥50% or number of stenoses ≥95% (including occlusions) patients with or without AECG ischemia were indistinguishable. The simple grading system of no, single-, double- or triple-vessel disease is a weak predictor of the presence of AECG ischemia.
Severe proximal stenoses were noted frequently in the ACIP patient group with AECG ischemia. This association has been reported previously. Mody et al. () found that patients with AECG ischemia for longer than 60 min were more likely to have proximal stenoses than patients with no AECG ischemia (OR 6.75, p < 0.002). This finding is in accordance with the concept that patients with larger amounts of myocardium jeopardized by severe proximal stenoses are expected to demonstrate more readily detectable AECG ischemia.
Complex plaque was found frequently in this group with AECG ischemia, with definite complex features noted in 34% and possible complex features in 50% of patients. A relation between AECG ischemia and complex coronary morphology has been suggested by other investigators, but only in the context of acute or unstable clinical syndromes. Bugiardini et al. () studied 88 patients with unstable angina. Complex coronary morphology was defined visually by the presence or absence of lesion ulceration or thrombus. Those with complex plaque had a longer duration of AECG ischemia and a higher likelihood of demonstrating residual AECG ischemia after aggressive medical therapy. Pozzati et al. () followed 86 patients with unstable angina (some of whom may have been included in the preceding report). Both at admission and after 3 days of medical treatment, AECG ischemia was more frequent and more severe in patients with complex stenosis morphology.
The significance of an association between complex plaque and AECG ischemia found in this and other smaller studies remains unclear. It is possible that in the ACIP study of clinically stable patients, the presence of complex stenosis morphology merely identified a group of patients more likely to have severe CAD and thus more likely to have AECG ischemia. This concept is supported by the fact that the association between the presence of complex plaque and AECG ischemia was dependent in part on severity of disease.
Alternatively, specific pathophysiologic events involved with plaque rupture may give rise to AECG ischemia. Vasospasm, partial or transient thrombosis or other dynamic mechanisms may be involved. Support for this concept is found in recent pathologic studies, which suggest that plaque rupture and thrombosis are the common events leading not only to acute coronary syndromes but also to progression of CAD in asymptomatic or stable patients ([20–26]). Autopsy of suicide and accident victims reveals plaque disruption unassociated with symptoms (). Asymptomatic plaque disruption has been found in ∼9% of noncardiac deaths ().
Ambulatory ECG ischemia may be a sensitive clinical marker for atherosclerotic disease activity or progression, which is manifested angiographically as complex plaque. As such, AECG ischemia may reflect total ischemic burden more accurately than angina, indicative of the presence of severe (proximal) or progressive (complex) CAD. This would explain the increased risk for adverse outcomes often observed in these patients with AECG ischemia ([1–14, 25]). Our findings of severe and complex CAD in the majority of ACIP patients gives some insight into the preliminary finding that 12-week and 1-year outcomes were improved with revascularization as compared with the two medical treatment arms; however, whether routine coronary angiography and revascularization, risk factor modification including lipid lowering therapy or interventions directed at plaque stabilization such as antiplatelet or antioxidant therapy will have an impact on clinical outcome in patients with AECG ischemia remains unknown and needs to be addressed in a larger clinical trial involving symptom-versus AECG ischemia–guided therapy ([29, 30]).
3.2 Study limitations
The study patients enrolled in ACIP had revascularizable CAD with at least one 50% stenosis. The comparison group of ACIP-screened patients without AECG ischemia was small, which reduced the power to detect differences. For example, it was estimated that with alpha = 0.01, the power would be just under 80% to detect a 60% reduction in prevalence among comparison group patients for a characteristic that was present in half of the patients with AECG ischemia. The comparison group was not random but was limited to patients with a positive exercise test and an available angiogram, which in most cases had been recently performed for clinical purposes. The requirement for angiography may have led the comparison group to include a higher proportion of patients with severe disease than would be seen in a representative sample of patients with CAD with a positive exercise test but no ischemic episodes during AECG monitoring. This bias in sampling would result in an underestimation of average coronary morphologic differences between patients with and without AECG ischemia. Thus, the important finding that proximal stenoses and complex plaque were seen more frequently in patients with AECG ischemia than in the comparison group will need to be confirmed through a detailed angiographic analysis from a larger study comparing patients with and without AECG ischemia.
The ACIP study is the largest and only multicenter study characterizing coronary angiographic findings in patients with AECG ischemia. The majority of patients enrolled had severe and complex CAD. Proximal stenoses and complex plaque were seen more frequently in patients with AECG ischemia than in those without it. The presence of severe and complex CAD may explain, in part, the increased risk for an adverse outcome known to be associated with the presence of AECG ischemia during activities of daily living.
We acknowledge Craig Pratt, MD, and Nancy L. Geller, PhD, for comments and suggestions, and Reneé Brodeur for assistance in the preparation of the manuscript.
↵1 Address for reprints: ACIP Clinical Coordinating Center, Maryland Medical Research Institute, 600 Wyndhurst Avenue, Baltimore, Maryland 21210.
↵fn1 This study was funded by the National Heart, Lung, and Blood Institute, Cardiac Diseases Branch, Division of Heart and Vascular Diseases, National Institutes of Health, Bethesda, Maryland, by Research Contracts HV-90-07, HV-90-08, HV-91-05 to HV-91-14. Study medications and placebo were donated by Zeneca Pharmaceuticals Group, Wilmington, Delaware; Marion-Merrell Dow, Kansas City, Missouri; and Pfizer, New York, New York. Support for electrocardiographic data collection was provided in part by Applied Cardiac Systems, Laguna Hills, California; Marquette Electronics and Mortara Instrument, Milwaukee, Wisconsin; and Quinton Instruments, Seattle, Washington. Some centers had partial support from General Clinical Research Center grants.
- Asymptomatic Cardiac Ischemia Pilot study
- ambulatory electrocardiographic
- coronary artery disease
- Thrombolysis in Myocardial Infarction
- Received April 17, 1996.
- Revision received August 14, 1996.
- Accepted September 20, 1996.
- The American College of Cardiology
- Gottlieb SO,
- Weisfield ML,
- Ouyang P,
- Mellits ED,
- Gerstenblith G
- Nademanee K,
- Intarachot V,
- Josephson MA,
- Rieders D,
- Mody FV,
- Singh BN
- Von Arnim T,
- Gerbig HW,
- Krawietz W,
- Hofling B
- Langer A,
- Freeman MR,
- Armstrong PW
- Deedwania PC,
- Carbajal EV
- Rocco MB,
- Nabel EG,
- Campbell S,
- et al.
- Yeung AC,
- Barry J,
- Orav J,
- Bonassin E,
- Raby KE,
- Selwyn AP
- Quyyumi AA,
- Mockus L,
- Wright C,
- Fox KM
- Mody FV,
- Nademanee K,
- Intarachot V,
- Josephson MA,
- Robertson HA,
- Singh BN
- Pozzati A,
- Bugiardini R,
- Borghi A,
- et al.
- Ambrose JA,
- Winters SL,
- Arora RR,
- et al.
- Badimon J,
- Fuster V,
- Chesebro J,
- Badimon L
- Ambrose JA,
- Tannenbaum MA,
- Alexopoulos D,
- et al.
- Davies MJ,
- Thomas AC
- Davies MJ,
- Bland J,
- Hangartuer J,
- Angelini A,
- Thomas AC
- Davies MJ
- for the ACIP Investigators,
- Pepine CJ,
- Geller NL,
- Knatterud GL,
- et al.
- for the ACIP Investigators,
- Knatterud GL,
- Bourassa MG,
- Pepine CJ,
- et al.
- for the ACIP Investigators,
- Rogers WJ,
- Bourassa MG,
- Andrews TC,
- et al.
- Conti CR,
- Knatterud GL,
- Sopko G
- Liang KY,
- Zeger SL