論文ID: CJ-20-1312
Background: This study investigated whether the percentage change (%Δ) in the eicosapentaenoic acid to arachidonic acid (EPA/AA) ratio is associated with cardiovascular event rates among acute coronary syndrome (ACS) patients receiving contemporary lipid-lowering therapy other than polyunsaturated fatty acids (PUFAs).
Methods and Results: This post hoc subanalysis of the HIJ-PROPER study included PUFA-naïve patients for whom EPA/AA ratio data were available at baseline and after 3 months. Patients were categorized into 2 groups based on the median %ΔEPA/AA ratio: Group 1, change less than the median; and Group 2, change greater than or equal to the median. The 3-year rates of the primary endpoint, a composite of all-cause death, non-fatal myocardial infarction, non-fatal stroke, and unstable angina pectoris, were compared between the 2 groups. The median %ΔEPA/AA ratio in Groups 1 and 2 was −26.2% (n=482 patients [49.9%]) and 42.2% (n=483 patients [50.1%]), respectively. At the 3-year follow-up, the occurrence of the primary endpoint was significantly lower in Group 2 than in Group 1 (29/483 [6.0%] vs. 53/482 [11.0%]; hazard ratio 0.53, 95% confidence interval 0.33–0.82; P=0.005). The same trend was observed after adjusting for patient factors (P=0.02).
Conclusions: Among ACS patients receiving contemporary lipid-lowering therapy other than PUFAs, a greater change in the EPA/AA ratio was associated with a lower incidence of cardiovascular events.
Polyunsaturated fatty acids (PUFAs) have been found to have pleiotropic protective benefits in humans.1,2 Eicosapentaenoic acid (EPA), known as omega-3 fatty acid, is an important PUFA that has been shown to have many beneficial effects, such as lowering blood triglyceride levels, reducing platelet aggregation, improving endothelial function, and promoting vasodilation and plaque stabilization, as well as anti-inflammatory and antiarrhythmic effects.1,2 Arachidonic acid (AA), known as omega-6 fatty acid, is a PUFA that is a precursor to several proinflammatory and proaggregatory mediators.3 Consequently, the EPA/AA ratio can be a predictor of risk for cardiovascular disease. Many studies have shown that a lower baseline EPA/AA ratio is associated with plaque vulnerability and a higher rate of clinical events.4–9 Furthermore, several clinical studies have investigated interventions for improving the EPA/AA ratio with an expectation of improving clinical endpoints, such as atherosclerotic lesions and mortality, in patients with cardiovascular disease.10–14 Humans are unable to synthesize PUFAs, and almost all PUFA requirements are fulfilled through dietary intake. It has been reported that Alaskan natives, Greenland Eskimos, and Japanese people who consume a large amount of fish have high EPA/AA ratios and a lower prevalence of cardiovascular disease.15,16 A change in the EPA/AA ratio may reflect a patient’s nutritional condition and the need for nutrition management. It remains unknown whether a change in the EPA/AA ratio in acute coronary syndrome (ACS) patients treated with contemporary lipid-lowering therapy other than PUFAs is associated with cardiovascular event rates.
Editorial p ????
In the Heart Institute of Japan-PRoper level of lipid lOwering with Pitavastatin and Ezetimibe in acute coRonary syndrome (HIJ-PROPER), we investigated the efficacy of intensive lipid-lowering therapy and compared it with that of conventional lipid-lowering therapy in ACS patients.17 The aim of the present study, a subanalysis of the HIJ-PROPER study, was to elucidate whether the percentage change in the EPA/AA ratio after receiving contemporary lipid-lowering therapy other than PUFAs is associated with cardiovascular event rates among ACS patients with dyslipidemia.
The present study was a post hoc subanalysis of the HIJ-PROPER study. Briefly, the HIJ-PROPER study was a multicenter prospective randomized open-label blinded endpoint trial with an active-control design that compared 2 lipid-lowering treatment strategies and was performed across 19 Japanese hospitals.17 Between January 2010 and April 2013, 1,734 patients were randomized to receive aggressive lipid-lowering therapy (pitavastatin+ezetimibe) or conventional lipid-lowering therapy (pitavastatin monotherapy).
The present study enrolled participants from the HIJ-PROPER study for whom data regarding serum PUFA levels, including EPA and AA, at baseline and the 3-month follow-up were available. To investigate the relationship between changes in the EPA/AA ratio under lipid-lowering therapy other than PUFAs and cardiovascular events, patients prescribed PUFA supplements were excluded from the present study.
Laboratory tests at baseline were performed within 24 h of hospitalization. All laboratory analyses were performed exclusively at an external laboratory (SRL Inc., Tokyo, Japan).
Patient Grouping and ClassificationThe median (interquartile range [IQR]) percentage change in the EPA/AA ratio for the entire cohort from baseline to the 3-month follow-up was found to be 1.14% (−26.2%, 42.3%). Based on the median value, patients were divided into 2 groups: Group 1, patients with a change in the EPA/AA ratio <1.14%; and Group 2, patients with a change in the EPA/AA ratio ≥1.14%.
Study EndpointsThe primary endpoint was a composite endpoint of the first occurrence of one of the following: all-cause death, non-fatal myocardial infarction, non-fatal stroke, or unstable angina pectoris. Clinical outcomes were assessed over a 3-year follow-up period and were compared between the 2 groups. Furthermore, a comparison of the primary endpoint between the groups was performed based on whether patients had baseline serum triglyceride levels that were higher or lower than the median value (117 mg/dL; IQR 82–170 mg/dL).
Statistical AnalysisContinuous variables are reported as the mean±SD and non-normally distributed data are reported as the median with IQR. Categorical data are presented as absolute values and percentages. Comparisons between groups were made using Welch’s t-test, the Mann-Whitney U-test and Pearson’s Chi-squared tests for normally distributed continuous variables, non-normally distributed continuous variables, and categorical variables, respectively.
The significance of the difference in the time-to-first occurrence of events between groups was analyzed using the Kaplan-Meier method with a log-rank test and a conventional Cox proportional hazards model. Kaplan-Meier analysis between groups was performed according to patients’ baseline serum triglyceride levels (i.e., higher or lower than the median value). Cox regression analyses were performed after adjusting for randomized treatment assignment, age, sex, body mass index, clinical presentation of ACS, medication, history of diabetes, hypertension, myocardial infarction, revascularization, baseline glomerular filtration rate, and baseline levels of high-sensitivity C-reactive protein (hs-CRP), high-density lipoprotein cholesterol (HDL-C), low-density lipoprotein cholesterol (LDL-C), and triglycerides. Unless stated otherwise, 2-sided P<0.05 was considered to indicate statistical significance. All statistical analyses were performed using statistics JMP Pro 14 (SAS Institute, Cary, NC, USA).
Ethical ConsiderationsAll authors, external and internal, had full access to all data (including statistical reports and tables) in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.
The original study was conducted in accordance with the principles of the 1975 Declaration of Helsinki. The institutional review boards or the relevant ethics committees of each participating medical center approved the study protocol, and all patients provided written informed consent for trial enrollment. The original trial is registered with the UMIN Clinical Trials Registry (ID: UMIN000002742).
Of the 1,734 patients randomized in the HIJ-PROPER study, 13 patients were lost to follow-up. Thus, a total of 1,721 patients were analyzed in the original HIJ-PROPER study. In the present study, 43 patients who were treated with PUFA supplements during follow-up were excluded, as were 713 patients had no EPA/AA data either at baseline, the 3-month follow-up, or both. Thus, 965 patients were included in the present study (Supplementary Figure 1). Most baseline characteristics were balanced between patients excluded from the analysis and those included in the analysis (Supplementary Table), with the exception of the distribution of clinical classifications of ACS. In addition, patients excluded from the analysis used β-blockers more frequently.
Of the 965 patients included in the study, 482 (49.9%) were categorized into Group 1 and 483 (50.1%) were categorized into Group 2 based on the median percentage change in the EPA/AA ratio of 1.14% (IQR −26.2%, 42.3%; Supplementary Figure 2). The median percentage changes in the EPA/AA ratio in Groups 1 and 2 were −26.2% (IQR −41.6%, −12.2%) and 42.2% (IQR 19.1%, 77.3%), respectively.
Baseline CharacteristicsThe baseline clinical characteristics of the 2 groups are presented in Table 1. Compared with patients in Group 1, those in Group 2 were younger (66.1±11.7 vs. 64.6±11.4 years; P=0.04) and were more likely to be male (72.0% vs. 79.5%; P=0.007). A history of previous myocardial infarction and revascularization was less common in Group 2. Regarding clinical presentations of ACS, Group 2 had significantly more patients with ST-elevation myocardial infarction and significantly fewer patients with unstable angina pectoris than Group 1. The rates of β-blocker, aspirin, and ezetimibe use were significantly lower in Group 2. In terms of the lipid profile, patients in Group 2 had significantly higher LDL-C levels and lower EPA/AA ratios at baseline than those in Group 1.
| All patients (n=965) |
Group 1 (n=482) |
Group 2 (n=483) |
P valueA | |
|---|---|---|---|---|
| Age (years) | 65.3±11.5 | 66.1±11.7 | 64.6±11.4 | 0.047 |
| Male sex | 731 (76.8) | 347 (72.0) | 384 (79.5) | 0.007 |
| BMI (kg/m2) | 24.2±3.5 | 24.1±3.6 | 24.4±3.3 | 0.17 |
| eGFR (mL/min/1.73 m2) | 72.7±18.8 | 73.3±18.6 | 72.1±19.1 | 0.31 |
| Statin-naive | 801 (83.0) | 398 (82.6) | 403 (83.4) | 0.73 |
| Hypertension | 646 (66.9) | 318 (66.0) | 328 (67.9) | 0.54 |
| Diabetes | 279 (28.9) | 143 (29.7) | 136 (28.2) | 0.62 |
| Current smoker | 338 (35.0) | 153 (31.7) | 185 (38.3) | 0.04 |
| Previous MI | 63 (6.5) | 40 (8.3) | 23 (4.8) | 0.03 |
| Previous revascularization | 77 (8.0) | 49 (10.2) | 28 (5.8) | 0.01 |
| Type of index event | 0.004 | |||
| STEMI | 468 (48.5) | 211 (43.8) | 257 (53.2) | |
| NSTEMI | 101 (10.5) | 48 (10.0) | 53 (11.0) | |
| Unstable angina pectoris | 396 (41.0) | 223 (46.3) | 173 (35.8) | |
| Medication at enrollment | ||||
| β-blocker | 86 (8.9) | 53 (11.0) | 33 (6.8) | 0.02 |
| ACEIs/ARBs | 290 (30.1) | 148 (30.7) | 142 (29.4) | 0.67 |
| Calcium channel blocker | 276 (28.6) | 143 (29.7) | 133 (27.5) | 0.48 |
| Aspirin | 155 (16.1) | 89 (18.5) | 66 (13.7) | 0.04 |
| Ezetimibe as study medication | 478 (49.5) | 261 (54.2) | 217 (44.9) | 0.005 |
| Baseline cholesterol metabolism | ||||
| TC (mg/dL) | 206 [185–231] | 202 [184–228] | 210 [186–237] | 0.005 |
| HDL-C (mg/dL) | 47 [39–56] | 47 [40–57] | 46 [39–55] | 0.14 |
| LDL-C (mg/dL) | 130 [114–151] | 128 [113–146] | 132 [116–156] | 0.0005 |
| Triglyceride (mg/dL) | 117 [82–170] | 113 [79–163] | 121 [88–174] | 0.23 |
| PUFA | ||||
| EPA (μg/mL) | 52.6 [37.3–74.3] | 59.7 [43.2–84.4] | 47.2 [33–64.9] | <0.0001 |
| AA (μg/mL) | 162 [137–192] | 161 [136–190] | 162 [138–195] | 0.35 |
| EPA/AA ratio | 0.33 [0.22–0.47] | 0.38 [0.26–0.53] | 0.3 [0.19–0.41] | <0.0001 |
| hs-CRP (mg/L) | 8.12 [2.42–28.4] | 7.42 [2.15–23.0] | 8.93 [2.63–25.9] | 0.15 |
Unless indicated otherwise, data are given as the mean±SD, median [interquartile range], or n (%). Patients were divided into 2 groups based on the change in the eicosapentaenoic acid/arachidonic acid (EPA/AA) ratio from baseline to the 3-month follow-up: Group 1, patients with a change in the EPA/AA ratio <1.14%; Group 2, patients with a change in the EPA/AA ratio ≥1.14%. AP values refer to comparisons between Groups 1 and 2. ACEIs, angiotensin-converting enzyme inhibitors; ARBs, angiotensin II receptor blockers; BMI, body mass index; eGFR, estimated glomerular filtration rate; HDL-C, high-density lipoprotein cholesterol; hs-CRP, high-sensitivity C-reactive protein; LDL-C, low-density lipoprotein cholesterol; MI, myocardial infarction; NSTEMI, non-ST segment elevation MI; PUFA, polyunsaturated fatty acid; STEMI, ST-elevation MI; TC, total cholesterol.
Table 2 presents changes in the lipid profiles and hs-CRP levels from baseline to the 3-month follow-up in the 2 groups. There was no significant change in HDL-C levels from baseline to the 3-month follow-up in either group, whereas LDL-C levels decreased significantly in both groups. There was no significant difference in triglyceride levels between baseline and the 3-month follow-up in Group 1 (113 [IQR 79–163] and 114 [IQR 80–162] mg/dL, respectively; P=0.43), but there was a significant increase in triglyceride levels from baseline to the 3-month follow-up in Group 2 (121 [IQR 88–174] to 125 [IQR 89–184] mg/dL, respectively; P=0.001). hs-CRP levels decreased significantly from baseline to the 3-month follow-up in both groups.
| Baseline | 3 months | P value | % ChangeA | P valueB | |
|---|---|---|---|---|---|
| TC (mg/dL) | |||||
| Group 1 | 202 [184, 228] | 145 [128, 165] | <0.0001 | −29.4 [−36.9, −20.6] | |
| Group 2 | 210 [186, 237] | 152 [132, 172] | <0.0001 | −28.2 [−37.6, −18.2] | 0.09 |
| HDL-C (mg/dL) | |||||
| Group 1 | 47 [40, 57] | 47 [41, 56] | 0.96 | 0 [−10.9, 11.9] | |
| Group 2 | 46 [39, 55] | 47 [40, 56] | 0.10 | 1.5 [−8.7, 12.4] | 0.12 |
| LDL-C (mg/dL) | |||||
| Group 1 | 128 [113, 146] | 70 [56, 87] | <0.0001 | −45.9 [−56.4, −35.0] | |
| Group 2 | 132 [116, 156] | 74 [50, 90] | <0.0001 | −46.6 [−56.8, −33.6] | 0.37 |
| Triglyceride (mg/dL) | |||||
| Group 1 | 113 [79, 163] | 114 [80, 162] | 0.43 | 1.1 [−29, 44] | |
| Group 2 | 121 [88, 174] | 125 [89, 184] | 0.001 | 5.8 [−26.2, 48.8] | 0.37 |
| EPA/AA ratio | |||||
| Group 1 | 0.38 [0.26, 0.53] | 0.26 [0.18, 0.39] | <0.0001 | −26.2 [−41.6, −12.2] | |
| Group 2 | 0.30 [0.19, 0.41] | 0.43 [0.29, 0.60] | <0.0001 | 42.2 [19.1, 77.3] | <0.0001 |
| EPA (μg/mL) | |||||
| Group 1 | 59.7 [43.2, 84.4] | 43.5 [30.8, 63.4] | <0.0001 | −23.8 [−40.8, −9.6] | |
| Group 2 | 47.2 [33, 64.9] | 69.3 [48.9, 93.3] | <0.0001 | 41.3 [18.6, 83] | <0.0001 |
| AA (μg/mL) | |||||
| Group 1 | 161 [136, 190] | 167 [140, 198] | <0.0001 | 4.1 [−10.2, 17.8] | |
| Group 2 | 162 [138, 195] | 163 [135, 194] | 0.51 | −1.3 [−13.7, 13.5] | 0.001 |
| hs-CRP (mg/L) | |||||
| Group 1 | 7.4 [2.2, 23.0] | 0.52 [0.26, 1.31] | <0.0001 | −91.3 [−97.7, −64.0] | |
| Group 2 | 8.9 [2.6, 25.9] | 0.54 [0.27, 1.34] | <0.0001 | −92.5 [−97.8, −75.4] | 0.21 |
Unless indicated otherwise, data are expressed as the median [interquartile range]. Patients were divided into 2 groups based on the change in the EPA/AA ratio from baseline to the 3-month follow-up: Group 1, patients with a change in the EPA/AA ratio <1.14%; Group 2, patients with a change in the EPA/AA ratio ≥1.14%. APercentage change from baseline to the 3-month follow-up. BP values refer to the differences in the percentage changes between Groups 1 and 2. Abbreviations as in Table 1.
At the 3-year follow-up, the primary endpoint had occurred significantly less often in Group 2 (29/483; 6.0%) than in Group 1 (53/482; 11.0%; hazard ratio [HR] 0.53; 95% confidence interval [CI] 0.33–0.82; P=0.005; Figure 1). The difference between groups remained significant even after adjusting for patients’ characteristics and laboratory tests (adjusted [a] HR 0.57; 95% CI 0.35–0.91; P=0.02; Table 3). In addition, when analyzing patients who were ≥65 years old, the incidence of the primary endpoint was lower in Group 2 than in Group 1 (HR 0.54; 95% CI 0.32–0.89; P=0.017; Supplementary Figure 3). Regarding individual components of the primary endpoint, all-cause mortality was significantly lower in Group 2 than in Group 1 (16/483 [3.3%] vs. 28/482 [5.8%], respectively; aHR 0.50; 95% CI 0.26–0.94; P=0.03) and non-fatal myocardial infarction occurred significantly less often in Group 2 than in Group 1 (1/483 [0.2%] vs. 7/482 [1.5%], respectively; aHR, 0.09; 95% CI 0.01–0.58; P=0.01).
Incidence of the primary endpoint in the 2 groups. Group 1, patients with a change in the eicosapentaenoic acid/arachidonic acid (EPA/AA) ratio <1.14%; Group 2, patients with a change in the EPA/AA ratio ≥1.14%.
| Group 1 (n=482) |
Group 2 (n=483) |
Crude model | Adjusted model | |||||
|---|---|---|---|---|---|---|---|---|
| HR | 95% CI | P value | HR | 95% CI | P value | |||
| Primary endpoint | 53 (11.0) | 29 (6.0) | 0.53 | 0.33–0.82 | 0.005 | 0.57 | 0.35–0.91 | 0.02 |
| Individual component of events | ||||||||
| All-cause death | 28 (5.8) | 16 (3.3) | 0.56 | 0.29–1.02 | 0.06 | 0.50 | 0.26–0.94 | 0.03 |
| Non-fatal MI | 7 (1.5) | 1 (0.2) | 0.14 | 0.01–0.79 | 0.02 | 0.09 | 0.01–0.58 | 0.01 |
| Non-fatal stroke | 7 (1.5) | 6 (1.2) | 0.84 | 0.27–2.53 | 0.76 | 1.01 | 0.30–3.43 | 0.99 |
| Unstable angina pectoris | 18 (3.7) | 12 (2.5) | 0.65 | 0.30–1.33 | 0.24 | 0.77 | 0.35–1.65 | 0.51 |
| All-cause death/non-fatal MI/non-fatal stroke | 39 (8.1) | 22 (4.6) | 0.55 | 0.32–0.91 | 0.02 | 0.52 | 0.30–0.90 | 0.02 |
Unless indicated otherwise, data are given as n (%). The primary endpoint was defined as a composite endpoint of all-cause death, non-fatal MI, non-fatal stroke, or unstable angina pectoris. The adjusted model was adjusted for age, sex, BMI, history of diabetes, hypertension, MI, and revascularization, the clinical presentation of acute coronary syndrome, medication, glomerular filtration rate, baseline hs-CRP, HDL-C, LDL-C, and triglyceride levels, and randomized treatment arm. CI, confidence interval; Group 1, patients with a change in the EPA/AA ratio <1.14%; Group 2, patients with a change in the EPA/AA ratio ≥1.14%; HR, hazard ratio. Other abbreviations as in Table 1.
Figure 2 shows Kaplan-Meier analysis of the primary endpoint in patients stratified according to baseline serum triglyceride levels (i.e., higher or lower than the median of 117 mg/dL [IQR 82–170 mg/dL]). There were no significant differences in the primary endpoint between the 2 groups stratified according to baseline serum triglyceride levels lower than the median (aHR 0.71; 95% CI 0.36–1.38; P=0.32; Figure 2A; Table 4). Among patients whose baseline serum triglyceride levels were higher than the median, the primary endpoint had occurred significantly less often in Group 2 (13/254; 5.1%) than in Group 1 (27/223; 12.1%; aHR 0.41; 95% CI 0.19–0.84; P=0.01; Figure 2B; Table 4). In addition, when patients were stratified according to a median EPA/AA ratio of 0.33 at baseline, there was no significant difference in the primary endpoint between patients with baseline EPA/AA ratio ≤0.33 and >0.33 (Figure 3).
Incidence of the primary endpoint in the 2 groups (Group 1, patients with a change in the eicosapentaenoic acid/arachidonic acid [EPA/AA] ratio <1.14%; Group 2, patients with a change in the EPA/AA ratio ≥1.14%), stratified according to baseline serum triglyceride levels that were (A) higher and (B) lower than the median (i.e., 117 mg/dL).
| Group 1 | Group 2 | HR | 95% CI | P value | |
|---|---|---|---|---|---|
| Patients with baseline serum triglyceride <117 mg/dL | |||||
| No. of patients | 259 | 229 | |||
| Primary endpoint | 26 (10.0) | 16 (7.0) | 0.71 | 0.36–1.38 | 0.32 |
| All-cause death | 16 (6.2) | 11 (4.8) | 0.68 | 0.29–1.53 | 0.35 |
| Non-fatal MI | 2 (0.8) | 0 | – | – | – |
| Non-fatal stroke | 5 (1.9) | 3 (1.3) | 0.50 | 0.10–2.51 | 0.40 |
| Unstable angina pectoris | 8 (3.1) | 5 (2.2) | 1.19 | 0.34–3.99 | 0.77 |
| All-cause death/non-fatal MI/non-fatal stroke | 20 (7.7) | 13 (5.7) | 0.65 | 0.30–1.37 | 0.26 |
| Patients with baseline serum triglyceride >117 mg/dL | |||||
| No. of patients | 223 | 254 | |||
| Primary endpoint | 27 (12.1) | 13 (5.1) | 0.41 | 0.19–0.84 | 0.01 |
| All-cause death | 12 (5.4) | 5 (2.0) | 0.27 | 0.08–0.81 | 0.02 |
| Non-fatal MI | 5 (2.2) | 1 (0.4) | 0.02 | 0.01–0.73 | 0.03 |
| Non-fatal stroke | 2 (0.9) | 3 (1.2) | 2.83 | 0.29–43.5 | 0.38 |
| Unstable angina pectoris | 10 (4.5) | 7 (2.8) | 0.57 | 0.19–1.62 | 0.29 |
| All-cause death/non-fatal MI/non-fatal stroke | 19 (8.5) | 9 (3.5) | 0.37 | 0.15–0.85 | 0.02 |
Unless indicated otherwise, data are given as n (%). All hazard ratios (HRs) were adjusted for age, sex, BMI, history of diabetes, hypertension, MI, and revascularization, clinical presentation of acute coronary syndrome, medication, glomerular filtration rate, baseline hs-CRP, HDL-C, LDL-C, and triglyceride levels, and randomized treatment arm. CI, confidence interval; Group 1, patients with a change in the EPA/AA ratio <1.14%; Group 2, patients with a change in the EPA/AA ratio ≥1.14%. Other abbreviations as in Table 1.
Incidence of the primary endpoint in patients with an eicosapentaenoic acid/arachidonic acid (EPA/AA) ratio at baseline <0.33 or ≥0.33.
PUFAs are essential nutrients that cannot be synthesized in the human body, and almost all PUFA requirements are filled through dietary intake. Usually, the EPA/AA ratio reflects a patient’s nutritional status. Several previous clinical studies have reported that adding PUFAs to statin therapy improves the prognosis in patients with cardiovascular disease.10,11,14 The results of the present study add new insights to the previous data. Even among patients who were treated with current lipid-lowering therapy other than PUFAs, the incidence of cardiovascular events was significantly lower in those with a percentage change in the EPA/AA ratio that was greater than the median value. Although some fundamental studies have reported an association between an increase in the EPA/AA ratio and parameters of atherosclerosis and arterial stiffness,18 few studies have investigated the association between changes in the EPA/AA ratio and clinical outcomes in patients with cardiovascular disease.
In the present subanalysis, there were several differences in baseline characteristics between the groups: patients in Group 2 were younger and more likely to be male. It is well known in epidemiological studies that a younger age and male sex are associated with favorable cardiovascular outcomes among patients with ACS.19,20 In the present study, after adjusting for patients’ characteristics, including age and sex, patients in Group 2 had a significantly lower cardiovascular event rate than patients in Group 1. However, the difference in age was decisive and difficult to adjust for; thus, we performed multivariate Cox analysis after adjustment in patients who were aged ≥65 years. In this analysis, the same trend was observed, indicating the reliability of the present results. With regard to baseline laboratory findings, mean LDL-C levels were higher and the EPA/AA ratio was lower in Group 2. Moreover, triglyceride levels increased significantly from baseline to the 3-month follow-up in Group 2 but not in Group 1. Generally, higher LDL-C levels are associated with a poor clinical outcome in patients with cardiovascular disease.21,22 Furthermore, elevated triglyceride levels are reportedly associated with increased mortality.23 The results of the present study suggest that the improved clinical outcome in Group 2 was independent of baseline LDL-C levels and the change in triglyceride levels.
One of the surprising results of the present study was that the incidence of cardiovascular events decreased in Group 2, even though the baseline EPA/AA ratio was low in this group. No significant difference was observed in the incidence of cardiovascular events between patients with high and low EPA/AA ratios at baseline. Comparable results were obtained in the previous subanalysis of the HIJ-PROPER study, which was performed on a larger number of patients than in the present study.24 This suggests that a change in the EPA/AA ratio may be a significant predictor for the incidence of cardiovascular events in ACS patients.
Considering the individual components of the primary endpoint in the present study, the lower mortality and lower incidence of myocardial infarction in Group 2 were the main factors contributing to the difference in the rate of the primary endpoint between the 2 groups. In contrast, no difference was observed in the incidence of stroke between the 2 groups. Several studies have reported an association between the EPA/AA ratio and mortality or the incidence of myocardial infarction in patients with cardiovascular disease or high-risk patients. However, a few studies have reported an association between the EPA/AA ratio and the incidence of stroke. In agreement with previous reports, our results suggest that the EPA/AA ratio may be a prognostic factor for death and myocardial infarction, but it is less relevant for stroke events.
It was of note that serum triglyceride levels at baseline affected the predictive value of the percentage change in the EPA/AA ratio in the present study. High triglyceride levels are related to comorbidities such as obesity and diabetes. Even after adjusting for these comorbidities, the significant difference between Group 1 and Group 2 was retained in patients with high triglyceride levels. A recent randomized controlled trial that focused on patients who had elevated triglyceride levels under statin treatment found that the addition of EPA to statin therapy reduced cardiovascular events among patients with cardiovascular disease or diabetes.14 Although it is difficult to analyze the detailed mechanism in the present study, it is hypothesized that there may be an interaction between the increase in the EPA/AA ratio and high triglyceride levels at baseline, and that this is related to a lower incidence of cardiovascular events. Patients treated with PUFA supplements were excluded from the present study; thus, the difference in the change in the EPA/AA ratio between Groups 1 and 2 is likely due to changes in the patients’ diet. Patient awareness of dietary effects may have increased the EPA/AA ratio, resulting in a lower incidence of cardiovascular events. The residual risks that may be addressed to improve the long-term cardiovascular outcome in patients with ACS remain unclear. Although few reports have examined whether PUFAs can improve the prognosis of cardiovascular events in ACS patients, the present study suggests that the administration of PUFAs improves the prognosis of ACS patients, especially those with high triglyceride levels. Further prospective studies are needed to validate and understand the present findings.
This study has some limitations. First, it was a retrospective study and based on a subgroup analysis of a previous prospective study. Second, the PUFA measurements were recorded only at baseline and 3 months; thus, there was uncertainty regarding whether the change in the EPA/AA ratio was stable during the follow-up period. Third, the primary endpoints of the original HIJ-PROPER study were all-cause death, non-fatal myocardial infarction, non-fatal stroke, unstable angina pectoris, and revascularization. Revascularization events were removed in the present analysis to make the primary endpoint strict. Although, the endpoint of the present study was reviewed by an independent expert endpoint committee, the results of this study should be interpreted with caution. Fourth, we did not examine the patients’ dietary intake during the follow-up period.
In conclusion, among ACS patients receiving contemporary lipid-lowering therapy other than PUFAs, a greater change in the EPA/AA ratio is associated with a decreased incidence of cardiovascular events, including all-cause death and non-fatal MI, especially in patients who have high baseline serum triglyceride levels.
The authors thank the HIJ-PROPER participants and the investigators and administrative staff of the HIJ-PROPER study for their contributions. The HIJ-PROPER Steering Committee had full access to all data in the study and had final responsibility for the decision to submit the article for publication. The clinical centers that participated in this study were: Tokyo Women’s Medical University, Sakakibara Heart Institute, Saisei-Kai Kumamoto Hospital, Cardiovascular Center of Sendai, Seirei Hamamatsu General Hospital, Saisei-Kai Kurihashi Hospital, National Yokohama Medical Center, Tokyo Metropolitan Tama Medical Center, Kosei General Hospital, NTT-East Kanto Medical Hospital, Tokyo Metropolitan Tama-Hokubu Medical Center, Shin-Matsudo Central General Hospital, JCHO Sagamino Hospital, Nishiarai Heart Center, Ogikubo Hospital, Shiseikai-Daini Hospital, Tokyo Metropolitan Ebara Hospital, Tokyo Women’s Medical University Medical Center East, and Tokyo Women’s Medical University Yachiyo Medical Center. Finally, the authors thank Editage (www.editage.com) for English language editing of this article and publication support.
The HIJ-PROPER trial was funded by the Japan Research Promotion Society for Cardiovascular Diseases. No additional extramural funding was used to support this work.
All members of the HIJ-PROPER study group report having received research support to perform clinical trials from the Japan Research Promotion Society for Cardiovascular Diseases, which is sponsored by Abbott Vascular Japan Co., Ltd., AstraZeneca K.K., Bayer Yakuhin Ltd., Boston Scientific Corporation, Bristol-Myers K.K., Daiichi Sankyo Company, Limited, Kowa Pharmaceutical Co., Ltd., Mochida Pharmaceutical Co., Ltd., MSD K.K., Nippon Boehringer Ingelheim Co., Ltd., Novartis Pharma K.K., Pfizer Japan Inc., Sanofi K.K., and Takeda Pharmaceutical Company Limited.
N. Hagiwara reports having received honoraria from Bristol-Myers K.K. and Nippon Boehringer Ingelheim Co., Ltd., and grants from Astellas Pharma Inc., Daiichi Sankyo Company, Limited, Eisai Co., Ltd., Mitsubishi Tanabe Pharma Corporation, Otsuka Pharmaceutical Co., Ltd., Shionogi & Co., Ltd., and Takeda Pharmaceutical Company Limited. N. Hagiwara is also a member of Circulation Journal’s Editorial Team.
J. Yamaguchi belongs to a division (Clinical Research Division for Cardiovascular Catheter Intervention) that is financially supported by donations from Abbott, Boston Scientific, Medtronic, and Terumo.
The remaining authors have no conflicts of interest to declare.
This study was approved by the Institutional Review Board of the Tokyo Women’s Medical University (Reference no. 1741). The original trial is registered with the UMIN Clinical Trials Registry (ID: UMIN000002742).
Data, including patients’ characteristics and the results of this study, will be shared for 2 years after publication. Data supporting the findings of this study are available from the corresponding author according to the rules of the Ethics Committee upon reasonable request. For any purpose, data will be shared as Excel file via E-mail.
Please find supplementary file(s);
http://dx.doi.org/10.1253/circj.CJ-20-1312
