Benefits of Icosapent Ethyl Across the Range of Kidney Function in Patients With Established Cardiovascular Disease or Diabetes: REDUCE-IT RENAL

Supplemental Digital Content is available in the text.

D rug therapies that target low-density lipoprotein cholesterol in patients with established cardiovascular disease (CVD) or cardiovascular risk factors can improve survival, prevent first or subsequent cardiovascular events, and reduce the need for coronary revascularization. [1][2][3][4] Although hypertriglyceridemia is an independent predictor of cardiovascular events, randomized studies of medications that lower triglyceride levels, including niacin and fibrates, have had less consistent success in improving cardiovascular outcomes. [5][6][7][8][9] Contemporary studies of marine-derived long-chain polyunsaturated n-3 fatty acid mixtures, which can effectively lower triglyceride levels, have not demonstrated reductions in cardiovascular events among statin-treated patients. [10][11][12][13] However, clinical benefit may differ based on the particular lipid composition of the n-3 fatty acid formulation. Icosapent ethyl contains the ethyl ester of a single long-chain murine omega-3 fatty acid, eicosapentaenoic acid. REDUCE-IT (Reduction of Cardiovascular Events with Icosapent Ethyl-Intervention Trial) randomly assigned 8179 statin-treated patients with established CVD or diabetes and other cardiovascular risk factors to either 4 g daily of icosapent ethyl or matching placebo. [14][15][16][17][18][19][20] After a median follow-up period of 4.9 years, the study drug demonstrated a 25% relative risk reduction in the primary composite end point of cardiovascular death, myocardial infarction, stroke, coronary revascularization, or unstable angina. The study further demonstrated that patients treated with icosapent ethyl had a 26% relative risk reduction in the composite of cardiovascular death, myocardial infarction, or stroke.
Among patients with chronic kidney disease (CKD), CVD remains the leading cause of morbidity and mortality. [21][22][23] However, because of the gaps in data and negative study results, uncertainty exists over the benefits of applying proven CVD therapy in the general population to the CKD population, especially in patients with advanced kidney disease. [24][25][26][27] This analysis aimed to explore the effects of icosapent ethyl versus placebo across the range of kidney function among patients enrolled in the REDUCE-IT study.

Study Design and Patient Characteristics
The data that support the findings of this study may be made available from the corresponding author on reasonable request. The study design and main results of the REDUCE-IT trial have been published previously.
REDUCE-IT was an international, phase 3b, double-blind trial that randomly assigned patients to treatment with icosapent ethyl 4 g daily (2 g twice daily with food) or matching placebo. Verbal and written informed consent were obtained from all study participants, and all sites were approved by institutional review boards.
Patients met eligibility criteria for enrollment if they had established CVD (secondary prevention group) or if they had type 1 or type 2 diabetes on medical treatment, and if they were ≥50 years of age and had at least 1 other major cardiovascular risk factor (high-risk primary prevention group). All patients were required to be on statin therapy for at least 4 weeks and to have a low-density lipoprotein cholesterol level between 41 mg/dL and 100 mg/dL. In addition, study patients had baseline triglyceride levels between 135 mg/dL and 500 mg/dL. Key exclusion criteria included severe heart failure, planned coronary intervention or surgery, severe liver disease, hemoglobin A1c level >10%, a history of pancreatitis, or known hypersensitivity to fish, shellfish, or ingredients of icosapent ethyl or placebo.

Measurement of Kidney Function
Estimated glomerular filtration rate (eGFR) was calculated using the Chronic Kidney Disease Epidemiology Collaboration equation as follows: eGFR=141×min (Scr/κ, 1)α×max(Scr/κ, where Scr is serum creatinine in mg/dL, κ is 0.7 for women and 0.9 for men, α is -0.329 for women and -0.411 for men, min indicates the minimum of Scr/κ or 1, and max indicates the maximum of Scr/κ or 1.

End Points and Follow-Up
The primary efficacy end point was a composite of cardiovascular death, nonfatal myocardial infarction, nonfatal stroke,

Clinical Perspective
What Is New?
• Icosapent ethyl reduced cardiovascular events among patients with elevated triglycerides and well-controlled low-density lipoprotein cholesterol on statin therapy across a wide range of baseline kidney function.
What Are the Clinical Implications?
• Despite having a well-controlled low-density lipoprotein cholesterol on statin therapy, patients with elevated triglycerides have significant residual risk for coronary events. • Treatment with icosapent ethyl has been shown to significantly reduce cardiovascular events and mortality in this patient population. • These findings are applicable to patients with chronic kidney disease across the spectrum of baseline kidney function. At a median follow-up of 4.9 years, both the primary and secondary composite end points were significantly reduced among patients treated with icosapent ethyl. This therapy led to consistent reduction in both the pri-mary and key secondary composite end points across baseline eGFR categories ( Figure 1, Figure S1). Patients with eGFR <60 mL·min -1 ·1.73 m -2 treated with icosapent ethyl had the largest absolute and similar relative risk reduction for the primary composite end point (icosapent ethyl versus placebo, 21 Figure 2, Figure S3). A post hoc analysis categorizing patients by commonly used eGFR categories corresponding to CKD stage also revealed a consistent reduction in primary and key secondary outcome event rates across eGFR categories of patients treated with icosapent ethyl ( Figure S2).

Nonstandard Abbreviations and Acronyms
Application of eGFR categories to prespecified hierarchical testing demonstrated consistent reductions, in general, in event rates across eGFR categories with icosapent ethyl as supported by nonsignificant interaction P values ( Figure S3). Icosapent ethyl was associated with significant reductions in cardiovascular death or nonfatal myocardial infarction in eGFR <60 mL·min - P=0.004) categories, with numeric reduction in the ≥90 mL·min -1 ·1.73 m -2 category. Similar differences were observed for fatal or nonfatal myocardial infarction and need for urgent or emergent revascularization.
The relative risk reductions in the primary composite and key secondary composite end points, in general, were consistent among icosapent ethyl-treated patients with either diabetes and risk factors for CVD (high-risk primary prevention cohort) or with established CVD (secondary prevention cohort) across eGFR subgroups ( Figures S4 and S5). Event rates and absolute risk reductions were numerically highest among patients with eGFR <60 mL·min -1 ·1.73 m -2 in the established CVD cohort (23.6% versus 33.2%; P<0.0001). Among patients with diabetes and risk factors, higher event rates were observed in the lowest eGFR group, but a higher absolute risk reduction was not consistently observed.
In each of the eGFR categories, we observed consistent risk reductions in the primary and key secondary end points among icosapent ethyl-treated patients with triglyceride levels ≥200 mg/dL or <200 mg/dL ( Figure S6).
A safety profile similar to the full cohort was observed for icosapent ethyl compared with placebo across eGFR subgroups. Adverse event rates rose with decreasing eGFR, but total adverse events occurred at similar rates with icosapent ethyl versus placebo.
Among icosapent ethyl-treated patients with eGFR <60 mL·min -1 ·1.73 m -2 , there was a higher rate of bleeding-related disorders (18.0% versus 13.3%; P=0.007). The highest rate of serious bleeding events was observed in the eGFR <60 mL·min -1 ·1.73 m -2 cat-egory (5.4% versus 3.6%; P=0.07; Table 2). However, HRs for all bleeding and serious bleeding events were similar regardless of eGFR cutoff, with no significant interaction observed (P interactions for all bleed-ing=0.68 and serious bleeding=0.76; Figure S7). No significant differences in gastrointestinal or cen-   Percentages are based on the number of patients randomly assigned to each treatment group. In general, the baseline value is defined as the last nonmissing measurement obtained before randomization. The baseline LDL-C value obtained through preparative ultracentrifugation was used unless it was missing. If the LDL-C preparative ultracentrifugation value was missing, then another LDL-C value was used, with prioritization of values obtained from LDL-C direct measurements followed by LDL-C derived by the Friedewald calculation method (only for patients with triglycerides <400 mg/dL) and LDL-C derived using the calculation published by investigators at The Johns Hopkins University. For all other lipid and lipoprotein marker parameters, wherever possible, baseline was derived as the arithmetic mean of the randomization visit 2 (day 0) value and the preceding visit 1 (or visit 1.1) value. If only one of these values was available, the single available value was used as baseline. Tertiles for LDL-C and triglycerides are based on the overall intention-to-treat population. eGFR indicates estimated glomerular filtration rate; HDL-C, high-density lipoprotein cholesterol; and LDL-C, low-density lipoprotein cholesterol. *The P value for age between treatment groups for the eGFR <60 mL·min -1 ·1.73 m -2 group was significant at P=0.0272. For all other baseline characteristics across groups, the P values were nonsignificant. To assess balance between treatment groups, P values were from a χ 2 test for categorical variables and Wilcoxon test for continuous variables.
tral nervous system bleeding events were observed between icosapent ethyl and placebo across eGFR categories. In addition, no significant differences in serious bleeding-related adverse events between icosapent ethyl and placebo across eGFR categories were observed (Tables 2-4).
In each eGFR category, no significant difference was observed in rates of treatment emergent adverse events of atrial fibrillation or atrial flutter (Tables 5-7, Figure S7), although a significant difference had been noted in the trial overall. Atrial fibrillation/flutter requiring hospitalization was an adjudicated end point, and rates of positively  Figure S7). The relative risk for atrial fibrillation among icosapent ethyl-treated patients was similar among eGFR categories, with no significant interaction observed (P-interaction=0.92; Figure S7). There were no significant differences between icosapent ethyl and placebo in overall or serious treatment emergent adverse events (Table S1). Severe treatment emergent adverse events occurred more commonly among patients with eGFR <60 mL·min -1 ·1.73 m -2 but did not differ between icosapent ethyl and placebo. Microalbuminuria was reported at low rates, more commonly with placebo (14 versus 3; P=0.01), and expressed no clear trend across eGFR subgroup.

DISCUSSION
The REDUCE-IT study demonstrated a 25% reduction in the risk of the primary composite end point of cardiovascular death, nonfatal myocardial infarction, nonfatal stroke, coronary revascularization, or unstable angina    among statin-treated patients randomly assigned to 4 g of icosapent ethyl (2 g twice daily) compared with those who were randomly assigned to placebo. This corresponds to a number needed to treat of 21 to prevent a cardiovascular event. In the current prespecified and post hoc analyses, we found that icosapent ethyl consistently reduced cardiovascular events across the full spectrum of baseline renal function categories in the REDUCE-IT study. Subgroup analyses demonstrated consistent reductions in the composite primary and key secondary end points for patients with eGFR <60 mL·min -1 ·1.73 m -2 , 60 to <90 mL·min -1 ·1.73 m -2 , and ≥90 mL·min -1 ·1.73 m -2 . The greatest absolute reduction in composite pri-mary and key secondary event rates was seen among patients with eGFR <60 mL·min -1 ·1.73 m -2 . Among patients treated with icosapent ethyl with eGFR <60 mL·min -1 ·1.73 m -2 , there was a 29% relative and 7.1% absolute reduction in the primary composite end point, corresponding to a number needed to treat of 14. Significant reductions in cardiovascular mortality or nonfatal myocardial infarction occurred among patients in the icosapent ethyl group with eGFR <60 mL·min -1 ·1.73 m -2 and 60 to <90 mL·min -1 ·1.73 m -2 . We additionally observed numeric reductions in patients on this treatment with eGFR ≥90 mL·min -1 ·1.73 m -2 . Tolerability and safety remained consistent within the entire study cohort. Adverse events occurred at similar rates with icosapent ethyl and placebo but more fre-  Percentages are based on the number of subjects in the intent-to-treat population within each treatment group (N). All adverse events are coded using the Medical Dictionary for Regulatory Activities (MedDRA Version 20.1).
*Includes atrial fibrillation/flutter treatment emergent adverse events and excludes positively adjudicated events. P value was based on Fisher's exact test.
†Includes atrial fibrillation/flutter treatment emergent adverse events meeting seriousness criteria. P value is based on Fisher's exact test.
‡Includes positively adjudicated atrial fibrillation/flutter requiring ≥24 hours of hospitalization clinical events by the clinical end point committee. P value is based on stratified log-rank test. Percentages are based on the number of subjects in the intent-to-treat population within each treatment group (N). All adverse events are coded using the Medical Dictionary for Regulatory Activities (MedDRA Version 20.1).
*Includes atrial fibrillation/flutter treatment emergent adverse events and excludes positively adjudicated events. P value was based on Fisher's exact test.
†Includes atrial fibrillation/flutter treatment emergent adverse events meeting seriousness criteria. P value is based on Fisher's exact test.
‡Includes positively adjudicated atrial fibrillation/flutter requiring ≥24 hours of hospitalization clinical events by the clinical end point committee. P value is based on stratified log-rank test. Percentages are based on the number of subjects in the intent-to-treat population within each treatment group (N). All adverse events are coded using the Medical Dictionary for Regulatory Activities (MedDRA Version 20.1).
*Includes atrial fibrillation/flutter treatment emergent adverse events and excludes positively adjudicated events. P value was based on Fisher's exact test.
†Includes atrial fibrillation/flutter treatment emergent adverse events meeting seriousness criteria. P value is based on Fisher's exact test.
‡Includes positively adjudicated atrial fibrillation/flutter requiring ≥24 hours of hospitalization clinical events by the clinical end point committee. P value is based on stratified log-rank test.
quently among patients with lower eGFR. A safety profile similar to the full cohort was observed for icosapent ethyl compared with placebo across eGFR subgroups, including an increase in total bleeding events without an increase in serious central nervous system or gastrointestinal bleeding, and an increase in the rates of atrial fibrillation or atrial flutter. Overall, bleeding rates were higher with decreasing eGFR, but the relative risks for all and serious bleeding were similar across eGFR categories, with no significant interaction observed. Atrial fibrillation/flutter event rates were higher in the lowest eGFR group, although absolute and relative risk differences were similar to those observed in other eGFR subgroups with no significant interaction observed.
CKD is strongly associated with dyslipidemia, and CVD remains the leading cause of mortality among patients with CKD. 25,26,28 Commonly used cardiovascular medications that treat dyslipidemia may be ineffective among patients with severe CKD. The benefit of statin-based therapy decreases as eGFR declines. Limited data exist on the clinical benefit of nonstatin medications, such as niacin, among patients with CKD. Both gemfibrozil and fenofibrate are renally cleared so they either require dose reduction or should be avoided depending on the severity of CKD. The dyslipidemia of CKD is characterized primarily by hypertriglyceridemia and reduced levels of high-density lipoprotein cholesterol. Therefore, therapies targeting triglyceride reduction may modify cardiovascular risk. Three studies have examined the relationship between marine derived n-3 fatty acids and cardiovascular outcomes; however, all were performed in patients on hemodialysis and used a less tailored n-3 fatty acid formulation than icosapent ethyl. [29][30][31] Thus, there remains a critical need to test therapies for hypertriglyceridemia and related cardioprotection and ascertain their efficacy and safety among patients across the broad range of kidney disease.
Formal statistical testing did not demonstrate heterogeneity for the primary and key secondary composite end points with respect to baseline renal function, and therefore, the overall results of the REDUCE-IT study apply to the entire study population. It is reassuring that the benefits of icosapent ethyl seen in the initial study manifest across eGFR categories, given that other commonly used cardiovascular medications may have less efficacy and greater adverse events among patients with CKD.
There are limitations to the present analysis. The REDUCE-IT study was not powered specifically for subgroup analyses. A creatinine clearance <30 mL·min -1 or the need for renal replacement therapy excluded patients from the REDUCE-IT study. Therefore, a small number of patients with severe CKD were enrolled in the study. Based on these enrollment numbers, the power to detect the potential benefits, safety, or risk of icosapent ethyl among this cohort was more limited. We performed analyses on the basis of prespecified subgroups and post hoc analyses, as well, that used eGFR staging categories. Last, urine samples were not collected routinely in REDUCE-IT. Therefore, microalbuminuria and other adverse events relying on specimen analysis may be underreported in both treatment arms.
Overall, the results of the REDUCE-IT RENAL analyses are consistent with the overall study results. Among statin-treated patients randomly assigned to icosapent ethyl 4 g daily (2 g twice daily), there were significant reductions in the primary and key secondary composite end points regardless of baseline eGFR. These benefits extend to significant reductions in myocardial infarction and cardiovascular death.