Relationship between coronary flow reserve evaluated by phase-contrast cine cardiovascular magnetic resonance and serum eicosapentaenoic acid

This study included 237 patients with known or suspected CAD who were assessed by X-ray coronary angiogram and cardiovascular magnetic resonance (CMR) including cine CMR, PC cine CMR, late gadolinium enhancement (LGE) CMR. Figure 1 illustrates flow chart of patient enrollment in this study. We excluded the patients with dilated cardiomyopathy (n = 10), severe valvular disease (n = 8), hypertrophic cardiomyopathy (n = 7), sarcoidosis (n = 3) and amyloidosis (n = 1). We also excluded patients with history of coronary artery bypass graft surgery (CABG) (n = 10) and patients who demonstrated significant coronary arterial stenoses on X-ray CAG (n = 71). Finally, 127 patients (male, 116 (91%); mean age, 72.2 ± 7.4 years) were enrolled in the present study. Table 1 summarizes the characteristics of the included patients. The medical histories of 42 (33%), 64 (50%) and 72 (57%) of the patients included myocardial infarction, angina pectoris. We allocated the patients to groups with high (n = 64; EPA ≥ 75.8 μg/mL) and low (n = 63; EPA < 75.8 μg/mL) median serum EPA. Other characteristics including coronary risk factors, cardiovascular history and medication did not significantly differ between the groups (Table 1). None of them were taking purified EPA. All patients provided written, informed consent to participate in this study, which was approved by the local institutional review board.

CMR was performed on a 1.5-T MR system equipped with 32 channel cardiac coils (Achieva, Philips Healthcare, Best, The Netherlands). All patients were assessed by cine CMR, PC cine CMR and LGE CMR. Imaging was performed after overnight fasting. Serum fatty acid measurement was performed on the same day as CMR.

Vector-electrocardiographic (VCG) monitoring leads were positioned on supine patients and then imaging started. Scout images were acquired in three orthogonal planes for cardiac orientation. Vertical and horizontal long-axis cine CMR of the LV was acquired using a steady-state free precession (SSFP) sequence. The LV volume and mass were calculated from short-axis cine images of the LV acquired from the apex to the base (repetition time, 4.1 ms; echo time, 1.7 ms; flip angle, 55°; field of view, 350 × 350 mm; acquisition matrix, 128 × 128; slice thickness, 10 mm; and number of phases per cardiac cycle, 20).

Cine CMR in the axial plane was obtained through the atrioventricular groove to locate the coronary sinus (Figure 1). The imaging plane for blood flow measurement by PC cine CMR was positioned perpendicular to the coronary sinus 2 cm from the ostium of the coronary sinus. Phase-contrast cine CMR of the coronary sinus was acquired during suspended shallow breath-holding using a VCG triggered gradient echo sequence (repetition time, 7.3 ms; echo time, 4.4 ms; flip angle, 10°; field of view, 240 × 194 mm; acquisition matrix, 128 × 128; and number of phases per cardiac cycle, 20). Pharmacological stress was achieved by injecting ATP (160 μg · kg^–1 · min^–1) into the left antecubital vein for 4 min. PC cine CMR images of the coronary sinus were acquired during ATP stress and at rest. The duration between stress and resting image acquisition was at least 10 min. All patients were asked to refrain from caffeinated beverages for at least 24 hours prior to CMR.

After acquiring phase-contrast cine CMR, the patients were injected with 0.15 mmol/kg of gadopentetate dimeglumine (Magnevist, Bayer Healthcare, Leverkusen, Germany). Late gadolinium-enhanced CMR was obtained in the same planes as cine images 15 min later using an inversion recovery–prepared gradient-echo sequence. The imaging parameters for late gadolinium-enhanced imaging comprised: repetition time, 4.3 ms; echo time, 1.3 ms; flip angle, 15°; field of view, 380 × 380 mm; acquisition matrix, 256 × 180; and slice thickness, 10 mm. The null point of the normal myocardium was determined using a Look-Locker sequence.

An observer who was blinded to the results of CMR interpreted the conventional X-ray CAG using quantitative software (QangioXA, Medis, Inc., Raleigh, North Carolina). Intracoronary administration of isosorbide dinitrate (2 to 3 mg) was performed in all patients before contrast injection. We performed quantitative coronary angiography (QCA) analysis for evaluating the degree of coronary arterial stenosis on X-ray CAG. In our trial, significant coronary arterial stenosis was defined as reduction in luminal diameter ≥50% by QCA analysis.

Blood was sampled from all patients after an overnight fast for at least 12 h. Serum levels of EPA, AA, and docosahexaenoic acid (DHA) were measured using a gas chromatograph (GC-2010; Shimadzu, Kyoto, Japan) and a capillary column (BPX90; Wako, Osaka, Japan) at a central laboratory (BML, Tokyo, Japan).

Two observers used a workstation (Extend MR WorkSpace, Philips Healthcare) to analyze the cine MR, PC cine MR, and LGE images. All CMR images were rendered anonymous and reviewed in random order. To measure the LV cardiac mass, epi- and endocardial borders of the LV on the short axis cines were manually traced with exclusion of the papillary muscles at each anatomic level that encompassed the LV. The LV mass was calculated by the consensus of two observers as the sum of the myocardial volume areas multiplied by the specific gravity (1.05 g/mL) of the myocardial tissue [22].

The contours of the coronary sinus were manually traced on each frame of all PC cine images to quantify blood flow in the coronary sinus and velocity in the adjacent tissue was measured for phase-offset correction. Blood flow in the coronary sinus was calculated by integrating the product of the cross-sectional area and the mean velocity in the coronary sinus and corrected using mean velocity in the adjacent tissue for all cardiac phases in the cardiac cycle. In line with other studies, we also corrected coronary sinus blood flow using rate pressure products (RPP) [16, 19, 23‐26].

The ΔCS flow and CFR were calculated as:

The coronary vascular resistance (CVR) was calculated as follows.

Late gadolinium-enhancement was visually assessed and interpreted by the consensus of two observers who were blinded to the patients’ information. The mass with LGE was measured on short-axis slices using manual planimetry. The total mass of LGE was calculated by summing the late gadolinium-enhanced mass of all sections, and the ratio (%) of LGE is expressed as:

None of the patients had atypical enhancement defined as mid-myocardial or sub-epicardial enhancement with multifocal distribution.

Data were statistically analyzed using SPSS software, version 17.0 (SPSS, Inc., Chicago, IL, USA). Continuous values are presented as means ± standard deviation (SD). Skewed values are presented as medians with the interquartile range (IQR). Normality was determined by the Shapiro-Wilk test. Coronary flow reserve determined by CMR was compared between the high and low EPA groups. Differences between the groups were evaluated using an unpaired t-test for normally distributed variables, and the Mann–Whitney U test for skewed variables. The relationship between serum fatty acids and CFR determined by CMR was determined using Pearson’s correlation coefficient. Preserved CFR was defined as CFR > 2.5 by PC cine CMR, which is the lower limit of normal flow reserve without obstructive CAD [27]. Independent predictors of impaired CFR were evaluated using univariate analysis and multivariate stepwise analyses. Inter- and intra-observer reproducibility was evaluated using the intraclass correlation coefficient (ICC). P values < 0.05 were considered statistically significant. In addition, we allocated study population into 4 groups based on presence or absence of DM and serum EPA. We evaluated significance of difference between 4 groups by using one-way analysis of variance (ANOVA).

We acquired PC cine CMR from all patients in the resting state and during ATP stress (Figure 2). Figure 3 shows representative blood flow curves of the coronary sinus at rest and during infusion. Table 2 summarizes the CMR findings. Cine parameters such as LV end-diastolic volume index (LVEDVI), LV end-systolic volume index (LVESVI), LV ejection fraction (LVEF), LV mass index did not significantly differ between the low and high EPA groups. In addition, LGE parameters, such as presence of LGE and percentage of LGE (%) also did not significantly differ between the groups.

Table 3 compares coronary sinus blood flow at rest and during ATP infusion, and CFR in the low and high EPA/AA groups. The coronary sinus blood flow measurements were reproducible, with ICCs of 0.93 and 0.95 for inter- and intra-observer reproducibility, respectively. No significant difference between low and high EPA groups was found in HR, SBP, DBP and RPP at rest and during ATP infusion. SBP and DBP were significantly decreased by ATP infusion both in low and high EPA groups. Corrected coronary sinus blood flow was significantly augmented by ATP infusion in the low and high EPA groups (from 79.1 ± 32.4 to 191.5 ± 87.2 mL/min, p < 0.001 and from 82.3 ± 35.0 to 229.5 ± 95.0 mL/min, p < 0.001, respectively), whereas ΔCS flow was significantly lower in the group with low EPA (112.4 ± 73.7 vs. 147.2 ± 77.7 mL/min, p = 0.036). Furthermore, CFR was significantly reduced in the low, compared with the high EPA group (2.54 ± 1.00 vs. 2.91 ± 0.98, p = 0.038). We found significant difference in CVR during ATP infusion between high and low EPA groups (0.53 ± 0.25 vs. 0.44 ± 0.26, p = 0.049). In addition, we divided study population into 4 groups based on presence or absence of DM and serum EPA (group 1, n = 44: DM(-) & EPA ≥75.8 μg/mL, group 2, n = 44: DM(-) & EPA <75.8 μg/mL, group 3, n = 20: DM (+) & EPA ≥75.8 μg/mL, group 4, n = 19: DM (+) & EPA <75.8 μg/mL. CFR was highest in group 1 and lowest in group 4 (mean CFR: 3.04 ± 1.00 in group1, 2.69 ± 1.11 in group2, 2.67 ± 1.04 in group 3, 2.39 ± 0.83). However, difference between 4 group did not show significant difference (p = 0.16 by one-way ANOVA).

Figure 4 shows relationships between serum fatty acids (EPA, DHA, AA) and CFR evaluated by CMR. Serum levels of EPA and DHA correlated significantly and positively (R = 0.35 and p < 0.001, R = 0.29 and p = 0.001, and R = 0.35 and p < 0.001, respectively), whereas those of AA correlated significantly and negatively (R = -0.18, p = 0.046) with CFR.

Table 4 shows the results of the univariate and multivariate analyses of predictors of CFR > 2.5 determined by PC cine CMR. Univariate analysis selected history of diabetes mellitus, serum EPA, LVMI on cine CMR as significant predictors of CFR > 2.5. Variables with p values < 0.05 in the univariate analysis were included in the multiple stepwise regression models. Table 4 shows that the EPA was a significant determinant CFR > 2.5 (odds ratio (OR), 1.01; 95% confidence interval (CI), 1.00 – 1.02; p = 0.008). Other independent variables comprised history of diabetes mellitus (OR, 0.39; 95%CI, 0.17 – 0.89; p = 0.026), LVMI (OR, 0.97; 95%CI, 0.95 – 0.99; p = 0.024).

The results of multivariate regression analysis in this study revealed that presence of DM, LVMI and serum EPA were the independent predictors of preserved CFR (CFR > 2.5). The CFR is an index of coronary microvascular function in patients without epicardial coronary arterial stenosis, therefore, we assessed the relationship between microvascular function and serum EPA concentration. It’s well known that the women more frequently develop microvascular disease (MVD) in comparison to men, however, men and women who have coronary MVD often have conventional risk factors, such as hypertension and DM [28]. The presence of DM was an independent predictor of CFR in this study, and the result is in line with previous study. However, presence of hypertension and presence of dyslipidemia were not independent predictors of CFR. The main reason is that the ACEI/ARB treatment for hypertension or statin treatment for dyslipidemia was already started. These drugs potentially have benefirial effects to MVD [28]. The serum EPA might play an important role for the regulation of CFR, and EPA drug treatment potentially improve the microvascular function in patients with MVD.

LV hypertrophy is another important mechanism of CFR impairment. presence of significant coronary arterial stenosis that results in dilatation already at rest. The patients with LV hypertrophy have a higher level of auto-regulated blood flow at rest to meet the demand of an increased myocardial mass [29]. Therefore, CFR is reduced because resting flow is already increased closer to the maximal level. The LVMI was an independent predictor of CFR in the current study, which is in line with previous study.

Epidemiological studies have shown that the incidence of acute myocardial infarction is substantially lower among non-emigrated inhabitants of Greenland (5.3%), than in individuals from the United States (40.4%) and Denmark (34.7%), with age- and sex adjusted death rates for ischemic heart disease [30]. This is mainly due to the composition of Eskimo food in north western Greenland, which is dominated by n-3 PUFA, such as EPA and DHA (ratios of n-3 PUFA in total Eskimo and Danish dietary intake of fatty acids: 13.7% and 2.8%, respectively) [30]. The randomized controlled GISSI-Prevenzione trial showed that dietary n-3 PUFA intake can reduce the rate of cardiovascular death among patients after a recent (≤ 3 months) myocardial infarction [13]. The large, randomized JELIS control trial found that that administering pure EPA together with statin therapy reduced the risk of developing coronary events to 19% [14]. However, precisely how n-3 PUFA prevents CAD remains unknown. Some reports have shown favorable effects of n-3 PUFA on the cardiovascular system. Leaf et al. showed that n-3 PUFA helped to prevent arrhythmic deaths, including sudden cardiac death [31]. Ghio et al. showed that n-3 PUFA significantly increases LV systolic function in patients with symptomatic heart failure of any etiology [31]. In addition, Moertl et al. found that n-3 PUFA administered for 3 months exerts a dose-dependent increase in the LVEF of patients with chronic heart failure. Moreover, high-dose n-3 PUFA intervention significantly improves endothelial function and decreases interleukin 6 levels [32]. We found that the serum EPA is an independent predictor of impaired CFR evaluated by PC cine CMR. This finding indicates that improved CFR might be an important mechanism through which n-3 PUFA prevents cardiovascular events.

In this study, we enrolled the CAD patients without ≥50% diameter stenosis on X-ray CAG, and assessed the microvascular function by calculating the CFR by PC-cine CMR. It is important to evaluate the disease severity in patients with non-obstructive CAD, because the prognosis of MVD is not as benign as commonly thought [28]. Randomized placebo-controlled studies have demonstrated that angiotensin-converting enzyme inhibitors, statins etc. relieve symptoms, vascular dysfunction, or both [28]. In the current study, most of our patients were administered with aspirin and statins, and had relatively low LDL cholesterol levels. Regardless, the EPA significantly correlated with CFR, and it was an independent predictor of reduced CFR. These results suggest that EPA therapy will help to improve CFR and prevent cardiovascular events through different mechanisms from statins or antiplatelet drugs. A further prospective interventional study is required to clarify this point.

First, this was a single center study with a relatively small patient cohort. Although CFR assessment by PC cine CMR is non-invasive and useful, CMR is contraindicated for patients implanted with pacemakers or cardiovascular defibrillators and those with claustrophobia. Such patients were excluded from the present study. Second, we excluded the CAD patients with significant coronary arterial stenosis (diameter reduction ≥50% by QCA analysis). Therefore, the findings of this study can only be applied to the CAD patients without ≥50% diameter stenosis. The reason to exclude the patients with significant CAD was that the fall of blood pressure induced by ATP should have too much compromised coronary flow with the risk to produce an unwanted myocardial ischemia. Third, CVR assessed in the current study mean the resistance throughout the entire cardiac cycle, included the systolic period during which a compressive resistance is added to the resistance due to the coronary vasomotor tone, usually indicated as coronary “vascular” resistance (auto-regulative resistance).