Coronary Flow Reserve During Dipyridamole Stress Echocardiography Predicts Mortality

Patients

The initial population was represented by 4,416 patients evaluated at 4 Italian Cardiology Institutions (Lucca, Mestre, Cesena, and Pisa) from August 2003 to June 2011 for enrollment in a study focused on assessing the prognostic value of CFR in the setting of known or suspected CAD. The study design was observational, not randomized, and prospective; patient data were entered into the databank at the time of initial assessment. All patients underwent stress echocardiography with wall motion and CFR assessment of LAD by using transthoracic Doppler ultrasound. Of the 4,416 patients initially selected, 103 (2%) were lost to follow-up. Thus, 4,313 (2,532 men; mean ± SD age 65 ± 11 years) formed the study group. Indications for stress echocardiography were suspected CAD in 2,766 (64%) subjects and risk stratification of known CAD (i.e., history of myocardial infarction, coronary revascularization, and/or angiographic evidence of >50% diameter coronary stenosis) in the remaining 1,547 (36%). According to individual needs and physician's choices, 2,515 (58%) patients were evaluated after anti-ischemic drugs had been discontinued, and 1,798 (42%) patients were evaluated during anti-ischemic treatment (Table 1). All patients gave their written informed consent before they underwent stress echocardiography. When patients signed the written informed consent form, they also authorized physicians to use their clinical data. Stress echocardiography data were collected and analyzed by stress echocardiographers not involved in patient care.

Stress echocardiography

Transthoracic stress echocardiographic studies were performed with commercially available ultrasound machines (Sonos 7500 or iE 33, Philips Ultrasound, Andover, Massachusetts; Vivid System 7, GE/Vingmed, Milwaukee, Wisconsin; Sequoia C256 Acuson Siemens Mountain View, California) equipped with multifrequency, phased-array sector scan probe (S3–S8 or V3–V7) and with second harmonic technology. Two-dimensional echocardiography and 12-lead electrocardiographic monitoring were performed in combination with high-dose dipyridamole administration (up to 0.84 mg over 6 min) (1). Echocardiographic images were semiquantitatively assessed using a 17-segment, 4-point scale model of the left ventricle (14). A wall motion score index was derived by dividing the sum of individual segment scores by the number of interpretable segments. Ischemia was defined as stress-induced new and/or worsening of pre-existing wall motion abnormality. Rest wall motion abnormality was akinetic or dyskinetic myocardium with no thickening during stress. CFR was assessed during the standard stress echocardiography examination by an intermittent imaging of both wall motion and LAD flow (1). Coronary flow in the mid-distal portion of LAD was searched in the low parasternal long-axis section under the guidance of color Doppler flow mapping (1). All studies were digitally stored to simplify off-line reviewing and measurements. Coronary flow parameters were analyzed off-line using the built-in calculation package of the ultrasound unit. Flow velocities were measured ≥2 times for each study; namely, at baseline and at peak stress (before aminophylline injection). At each time point, 3 optimal profiles of peak diastolic Doppler flow velocities were measured, and the results were averaged. CFR was defined as the ratio between hyperemic peak and basal peak diastolic coronary flow velocities. A CFR value ≤2.0 was considered abnormal (3). All investigators of contributing centers passed quality control criteria for regional wall motion and Doppler interpretation before entering the study as previously described (15). The previously assessed intraobserver and interobserver variability for measurements of Doppler recordings and regional wall motion analysis assessment were <10% (16).

Follow-up data

Follow-up data were obtained from review of the patient's hospital record, personal communication with the patient's physician and review of the patient's chart, a telephone interview with the patient or a patient's close relative conducted by trained personnel, or a staff physician visiting the patients at regular intervals in the outpatient clinic. Mortality was the only endpoint. To avoid misclassification of the cause of death (9), overall mortality was considered. Coronary revascularization (surgery or angioplasty) was also registered; however, it was not identified as a clinical event, and patients were censored.

Statistical analysis

Continuous variables are expressed as mean ± SD. Differences between groups were compared using the Student t test and the chi-square test, as appropriate. Mortality rates were estimated with Kaplan-Meier curves and compared using the log-rank test. Patients undergoing coronary revascularization were censored at the time of the procedure. Annual mortality was obtained from Kaplan-Meier estimates to take censoring of the data into account. The association of selected variables with outcome were assessed with the Cox proportional hazards model using univariate and stepwise multivariate procedures. In the Cox model, the time-dependent revascularization effect was estimated. A significance of 0.05 was required for a variable to be included in the multivariate model, and 0.1 was the cutoff value for exclusion. Hazard ratios (HRs) with the corresponding 95% confidence intervals (CIs) were estimated. The following covariates were analyzed: age, sex, left bundle branch block, family history of CAD, diabetes mellitus, arterial hypertension, hypercholesterolemia, smoking habit, previous myocardial infarction, previous coronary artery bypass graft, previous percutaneous revascularization, ongoing anti-ischemic therapy, resting wall motion abnormality, resting wall motion score index, ischemia at stress echocardiography, wall motion score index at peak stress, test positivity, and CFR on LAD.

Statistical significance was set at p < 0.05. SPSS version 20 (SPSS Inc., Chicago, Illinois) was used for analysis.

Stress echocardiographic findings

No complication occurred during the test. Mean CFR on LAD was 2.35 ± 0.68. At individual patient analysis, 1,419 (33%) subjects had CFR ≤2 and 2,894 (67%) had CFR >2. Stress echocardiography was positive for ischemia in 765 (18%) individuals. Ischemia was assessed in 516 patients with CFR ≤2 and in 249 patients with CFR >2 (36% vs. 9%; p < 0.0001) (Table 1).

Patients with CFR ≤2 were older, more frequently men, and had a higher incidence of diabetes mellitus, arterial hypertension, hypercholesterolemia, left bundle branch block, previous myocardial infarction, and coronary revascularization (Table 1); in addition, they were more frequently tested under anti-ischemic medication than patients with CFR >2.

Survival analysis

During a median follow-up of 19 months (interquartile range: 8 to 36 months), 146 (3.4%) patients died.

According to the physician's judgment, 907 (21%) underwent myocardial revascularization (205 surgery and 702 percutaneous intervention) after a median of 111 days (1st quartile 30 days; 3rd quartile 250 days) from the index stress echocardiography. There were 470 revascularizations in patients with ischemia and 437 in those without ischemia at stress echocardiography (61% vs. 12%; p < 0.0001). In addition, there were 635 revascularizations in patients with CFR ≤2 and 272 revascularizations in those with CFR >2 (45% vs. 9%; p < 0.0001).

The 4-year mortality was significantly higher in patients with CFR ≤2 than in patients with CFR >2, both considering the group with ischemia (39% vs. 7%; p < 0.0001) and the group without ischemia at stress echocardiography (12% vs. 3%; p < 0.0001) (Fig. 1). Interestingly, in both patient groups, CFR ≤2 was associated with markedly greater annual mortality in subjects with known CAD as well as in subjects with suspected CAD (Fig. 2). Considering patients without ischemia and CFR >2, annual mortality was only 0.8% in those with known CAD and 0.6% in those with suspected CAD. In the subset of patients with suspected CAD and not undergoing medical therapy at the time of testing, the annual mortality was 0.6% in those with CFR >2 and 7.2% in those with CFR ≤2 (p < 0.001).

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

Wall Motion, CFR, and Risk

The study population separated on the basis of presence (+) or absence (−) of ischemia at stress echocardiography (SE) and coronary flow reserve (CFR) on left anterior descending artery ≤2 or >2. Assuming a hazard ratio of 1 for the lowest mortality rate subset (SE−/CFR), the hazard ratio with relative 95% confidence intervals are reported in the graphic. The number of patients per year is shown.

• Download figure
• Open in new tab
• Download powerpoint

Figure 2

Annual Mortality Rate

Annual mortality in the group of patients with known coronary artery disease (CAD) and suspected CAD separated on the basis of presence (+) or absence (−) of ischemia at SE and CFR on left anterior descending artery ≤2 or >2. Abbreviations as in Figure 1.

Univariate indicators of mortality are listed in Table 2. At multivariable analysis, CFR on LAD ≤2 (HR: 3.31 [95% CI: 2.29 to 4.78]; p < 0.0001), ischemia at stress echocardiography (HR: 2.40 [95% CI: 1.65 to 3.48]; p < 0.0001), left bundle branch block (HR: 2.26 [95% CI: 1.50 to 3.41]; p < 0.0001), age (HR: 1.08 [95% CI: 1.06 to 1.10]; p < 0.0001), resting wall motion score index (HR: 3.52 [95% CI: 2.38 to 5.21]; p < 0.0001), male sex (HR: 1.74 [95% CI: 1.12 to 2.52]; p = 0.003), and diabetes mellitus (HR: 1.47 [95% CI: 1.03 to 2.08]; p = 0.03) were independent predictors of mortality. Revascularization procedure did not exert any significant effect (HR: 0.586 [95% CI: 0.80 to 4,284]) when considered as a time-dependent covariate in the Cox model.

Comparison with previous studies

The value of stress-induced wall motion abnormalities to predict mortality has been extensively documented in patients with known or suspected CAD (11–13). Our findings confirm and expand those previous experiences, demonstrating an additive prognostic value of CFR assessment on LAD over wall motion analysis. In fact, a CFR ≤2 identified a subset of patients at increased risk of mortality both in the presence and in the absence of ischemia at stress echocardiography.

The current study has major relevant differences when compared with prior studies addressing the issue of prediction of future events with stress echocardiography. First, the study design is large-scale and multicenter, which provides a realistic picture of stress echocardiographic results obtained from heterogeneous nonacademic cardiologic institutions with on-site stress echocardiography interpretation; this setting is able to evaluate the effectiveness of pharmacological stress echocardiography in prognosis. Second, the sample size consisted of 4,313 patients, which is the largest population studied to date with combined CFR assessment, large enough to provide prognostic information only analyzing hard endpoints such as total mortality with no need to include surrogate or composite endpoints (2–7) to increase the power of prognostication. Third, the type of population analyzed (i.e., low risk) was ideal to test the performance of stress tests in identifying only those patients at higher risk for mortality. Still, the high rate of revascularizations, especially in the ischemic result subset, may have lowered the power of stratification of stress testing.

Clinical implications

There are several tests and strategies for the evaluation of patients with known or suspected CAD, but no single strategy has been demonstrated to be superior overall. Nonetheless, the approach to these patients with physiological testing is sound, rational, and effective. Nowadays, the availability of high-tech imaging and its increasing and unrestricted use (17) is shadowing simpler and less expensive technologies such as ultrasounds. In most cases, the new technologies are deployed into clinical practice before their incremental and additive value has been demonstrated through the conventional process of health technology assessment (18,19). However, in a new environment of healthcare cost-containment, minimization of risks related to imaging such as ionizing radiations (20), stress echocardiography with dual imaging seems to be the appropriate tool to be implemented in the everyday clinical practice due to its high negative predictive value, low cost, lack of ionizing radiations, and wide availability. In fact, in large series of unselected patients referred to high-volume stress echocardiography laboratories, stress test results should influence clinical management. The benign prognostic implications of a nonischemic stress echocardiography with normal CFR on LAD should reduce the need for coronary angiography, representing the essential step for coronary revascularization only in case of extensive test positivity or markedly reduced CFR. The presence of a severely reduced CFR with no inducible ischemia by wall motion criteria may identify a group with relevant microvascular impairment that is not prognostically neutral as also shown in patients with hypertension (5), diabetes (3), and normal or near-normal coronary arteries (4). Dual-imaging stress echocardiography with dipyridamole has reached the status of an established technique that needs to be implemented in clinical echocardiography laboratories with appropriate training and technology upgrading. All these echocardiography parameters are entwined to provide an accurate and powerful risk stratification with the perfect tailoring of the test to the individual patient's needs.

Study limitations

In this study, there was no central reading. Stress echocardiography and CFR measurement were interpreted in the peripheral centers, and data were entered directly into the databank. This system allowed substantial sparing of human and technological resources, but it was also the logical prerequisite for a large-scale study, designed to represent the realistic performance of the test rather than the results of a single laboratory or even a single person working in a highly dedicated echocardiography laboratory. Because the assessment of the echocardiograms was qualitative and subjective, variability in reading the echocardiograms might have modulated the results of individual centers (21). However, all our readers in individual centers had a long-lasting experience in echocardiography, passed the quality control in stress echocardiography reading as previously described (15). The CFR was sampled only on LAD. There is no doubt that the 3 coronaries approach would be more fruitful, but at present it remains too technically challenging for a large-scale assessment. Forty-two percent of the patient population underwent dipyridamole stress echocardiography while receiving anti-ischemic medical therapy, which may have offset myocardial ischemia. Nevertheless, therapy exerts a negligible effect on the diagnostic (5) and prognostic (7) value of CFR. In the current study, follow-up was censored at the time of revascularization, and no data are available on the effect of treatments (surgery or angioplasty).