SHARP

In Review

Baigent C, et al. "The effects of lowering LDL cholesterol with simvastatin plus ezetimibe in patients with chronic kidney disease". Lancet. 2011. 377:2181-2192.
PubMed • Full text • PDF

Clinical Question

Among patients with moderate to severe chronic kidney disease, including patients on dialysis, does LDL-lowering therapy with simvastatin and ezetimibe reduce atherosclerotic events?

Bottom Line

LDL-lowering therapy reduces atherosclerotic events in patients with chronic kidney disease. This result was primarily driven by a reduction in ischemic strokes and arterial revascularization procedures, with no clear reduction in either myocardial infarction or coronary death. It is uncertain whether these results can be uniformly applied to all patients with CKD, especially those with Stage 5 disease and on dialysis.

Major Points

Multiple studies have demonstrated that statins reduce cardiovascular morbidity and mortality in the general population at high risk of cardiovascular disease. Previous trials and meta-analyses have shown similar results in patients with mild CKD (eGFR > 60 ml/min). However, these results were not felt to be generalizable to patients with more advanced CKD as the pathophysiology of disease changes with disease progression. As kidney function deteriorates, the importance of traditional cardiovascular risk factors (e.g. dyslipidemia, hypertension, smoking) becomes overshadowed by non-traditional risk factors including anemia, uremia, calcium disturbances, and sympathetic overactivity. This leads to alterations in vascular stiffness, inflammation, and structural heart disease. Trials conducted with patients on renal replacement therapy, including those on dialysis (4D, AURORA) and with a kidney transplant (ALERT), did not demonstrate any significant reduction in cardiovascular outcomes with the use of LDL-lowering therapy, supporting this theory. None of these trials, however, had included patients with more advanced CKD not requiring RRT.

Published in 2011, the SHARP (Study of Heart and Renal Protection) trial randomized over 9000 patients with moderate to severe CKD (including those on dialysis) to either a combination of ezetimibe and simvastatin or placebo. Participants were followed for an average of 5 years. Approximately one third of patients were on dialysis. Of those not on dialysis, about 40% of patients had an eGFR of 30-60, 40% between 15-30, and 20% less than 15. Treatment in the LDL-lowering arm was associated with a statistically significant reduction in the primary composite endpoint of atherosclerotic events including nonfatal MI, coronary death, non-hemorrhagic stroke, or any arterial revascularization procedure (11.3% vs. 13.4%, RR 0.83; ARR 2.1%; NNT 48). This was primarily driven by a reduction in ischemic stroke (NNT 111) and coronary revascularization procedures (NNT 83) with no significant difference observed in coronary deaths or nonfatal MIs. The SHARP study also demonstrated that LDL-lowering therapy does not slow the progression of renal disease.

Despite the overall positive findings of the SHARP trial, there is concern that benefit is not homogenous across eGFR subgroups and dialysis status. While the trial was not powered for such comparisons, there is a trend for heterogenous treatment effects by dialysis status and albuminuria level. In keeping with results from both 4D and AURORA, patients on dialysis did not appear to have as significant a risk reduction in the primary outcome compared to patients with CKD not on dialysis, although differences between groups were not statistically significant. A subsequent meta-analysis of hemodialysis patients receiving statin therapy including the 4D, AURORA, and SHARP trials did not come to a definitive conclusion on the utility of statin therapy, as results did not reach statistical significance even though CKD patients not on dialysis were found to have a statistically significant benefit.. It may be that patients with ESRD on dialysis are more at risk of causes of death that are not modified by statin therapy compared to patients with CKD not on dialysis.

Guidelines

KDIGO Clinical Practice Guidelines for Lipid Management in Chronic Kidney Disease (2013, adapted)

In adults aged ≥ 50 years, with eGFR < 60 mlmin/1.73m² not on chronic dialysis or with renal transplantation, we recommend treatment with a statin or statin/ezetimibe combination (Grade 1A).

In adults aged ≥ 50 years with CKD and eGFR ≥ 60ml/min/1.73m², we recommend treatment with a statin (Grade 1B).

In patients aged 18-49 years with CKD but not treated with chronic dialysis or kidney transplantation, statin treatment is recommended with one or more of the following (Grade 2A):

• known CAD
• Diabetes Mellitus
• Prior ischemic stroke
• CAD risk > 10% over 10 years

In adults with dialysis-dependent CKD, we suggest that statins or statin/ezetimibe combination not be initiated (Grade 2A).

In patients already receiving statins or statin/ezetimibe combination at the time of dialysis initiation, we suggest that these agents be continued (Grade 2C).

2019 ACC/AHA Guideline on the Primary Prevention of Cardiovascular Disease

Chronic Kidney disease (eGFR 15-59 ml/min/1.73m², not treated with dialysis or kidney transplant) is not an indication for lipid-lowering therapy/statin alone. It is considered a risk-enhancing factor beyond risk captured in pooled cohort equations, which can be considered for patients at borderline or intermediate risk by pooled cohort equations.

USPSTF

The USPSTF has not made specific recommendations regarding cardiovascular risk reduction with lipid-lowering therapy in patients with CKD.

Design

• Multinational, double-blinded, double-dummy, randomized control trial
• N=9,270 (3,023 on dialysis)
  • Simvastatin 20mg daily plus ezetimibe 10mg daily (n=4,650)
  • Placebo (n=4,620)
• Setting: 200 centers in 10 countries
• Follow-up: Median 4.9 years
• Analysis: Intention-to-treat
• Primary outcome:
  • Annual incidence of major atherosclerotic events (non-fatal MI, coronary death, non-hemorrhagic stroke, arterial revascularization excluding dialysis access)
  • Original primary endpoint included all major vascular events (any stroke, in stead of non-hemorrhagic strokes only)

Population

Inclusion Criteria

• History of CKD
  • pre-dialysis
    • Men: Serum Cr ≥ 150umol/l (≥ 1.7mg/dL)
    • Women: Serum Cr ≥ 130 umol/l (≥ 1.5mg/dL)
  • on dialysis
• Age ≥ 40 years

Exclusion Criteria

• History of MI or coronary revascularization procedure
• Functioning renal transplant or planned living donor-related transplant
• less than 2 months since presentation of an acute uremic emergency
• History of chronic liver disease or abnormal liver function (ALT >1.5x ULN)
• Definite previous adverse reaction to a statin or ezetimibe
• Concurrent, peramenent treatment with a contraindicated drug: statin, fibrate, nicotinic acid, macrolide antibiotic, systemic imidazole or triazole antifungals, protease inhibitors, nefazodone, ciclosporin, ezetimibe
• Child-bearing potential (i.e. premenopausal without reliable contraception)
• Known poor compliance with medication or clinic visits
• Medical history that may prevent treatment success: cancer other than non-melanoma skin cancer, recent alcohol or substance misuse, other severe disease

Baseline Characteristics

Data taken from treatment group.

• Female: 37%
• Age: 62
• Smoker, current: 13%
• sBP: 139
• dBP: 79
• Total cholesterol: 4.88 mmol/L
• LDL cholesterol: 2.77 mmol/L
• HDL cholesterol: 1.12 mmol/L
• Triglycerides: 2.31 mmol/L
• BMI: 27.1 kg/m²
• Renal status:
  • Dialysis: 33%
    • Hemodialysis: 27%
    • Peritoneal dialysis: 6%
  • No dialysis: 67%
• MDRD eGFR (mL/min/1.73m²), excludes patients on dialysis:
  • ≥ 60: 1%
  • 30-59: 37%
  • 15-29: 41%
  • <15: 20%
• Urine ACR (mg/g), excludes patients on dialysis:
  • <30: 20%
  • 30-299: 37%
  • ≥ 300: 43%
• Comorbidities
  • Previous vascular disease: 15%
  • Diabetes: 23%

Interventions

• Run-in period prior to randomization with placebo tablets for six weeks, to ensure compliance
• Randomization to three groups, with double-dummy technique to preserve blinding:
  • Simvastatin 20mg daily + Ezetimibe 10mg daily (n=4,193)
  • Placebo (n=4,191)
  • Simvastatin 20mg daily: n = 1054
    • This arm was designed to monitor the safety profile of ezetimibe. After one year, the arm was re-randomized to simvastatin + ezetimibe (n=443) or placebo (n=443), leading to a total of n=4,650 in the treatment arm and n=4,620 in the placebo arm.
• Patients were followed for a median of 4.9 years
• Safety assessment of all patients between the three arms was determined at 1 year
• Biochemical efficacy (cholesterol profile, creatinine, proteinuria), was determined in a random 10% of patients at 1 and 4 years, and in all patients at 2.5 years.

Outcomes

Presented as simvastatin + ezetimibe vs. placebo. Patients re-randomized from the simvastatin only arm are included.

Primary Outcome

Major Atherosclerotic Events

Defined as nonfatal MI, coronary death, non-hemorrhagic stroke, any arterial revascularization procedure

11.3% vs. 13.4% (RR 0.83; 95% CI 0.74-0.94; p=0.0021; NNT = 48)

Secondary Outcomes

Any major coronary event

4.6% vs. 5.0% (RR 0.92; 95% CI 0.76-1.11; p=0.37)

Non-fatal MI

2.9% vs. 3.4% (RR 0.84; 95% CI 0.66-1.05; p=0.12)

CHD death

2.0% vs. 1.9% (RR 1.01; 95% CI 0.75-1.35; p=0.95)

Any non-hemorrhagic stroke

2.8% vs. 3.8% (RR 0.75; 95% CI 0.60-0.94; p=0.01; NNT = 100)

Ischemic strokes only

2.5% vs. 3.4% (RR 0.72; 95% CI 0.57-0.92; p=0.0073; NNT = 111)

Any Revascularization Procedures

6.1% vs. 7.6% (RR 0.79; 95% CI 0.68-0.93; p=0.0036; NNT = 67)

Coronary revascularization Procedures

3.2% vs. 4.4% (RR 0.73; 95% CI 0.59-0.90; p=0.0027; NNT = 83)

Decline in Renal Function

p=NS for all endpoints (Initiation of maintenance dialysis or transplantation, ESRD or death, ESRD or doubling SCr)

Subgroup Analysis

By eGFR

≥ 60: 6.8% vs. 6.8% (HR 0.84; 95% CI 0.17-4.18)

≥ 30-60: 7.9% vs. 10.4% (HR 0.75; 95% CI 0.57-1.00)

≥ 15-30: 10.2% vs. 12.7% (HR 0.78; 95% CI 0.62-0.98)

<15: 10.9% vs. 13.3% (HR 0.82; 95% CI 0.59-1.13)

By Urinary ACR (mg/g)

30: 8.1% vs 9.4% (HR 0.85; 95% CI 0.57-1.26)

≥ 30 <=300: 9.0% vs. 11.9% (HR 0.75; 95% CI 0.57-0.97)

>300: 9.5% vs. 11.9% (HR 0.72; 95% CI 0.57-0.91)

By Dialysis Status

No Dialysis: 9.5% vs. 11.9% (HR 0.78; 95% CI 0.67-0.91)

All dialysis: 15.0% vs. 16.5% (HR 0.90; 95% CI 0.75-1.08)

Hemodialysis: 15.2% vs. 15.9% (HR 0.95; 95% CI 0.78-1.15)

Peritoneal dialysis: 14.0% vs. 19.7% (HR 0.70; 95% CI 0.46-1.08)

Age at randomization

40-49y: 5.8% vs. 5.5% (HR 1.08; 95% CI 0.74-1.58)

50-59y: 7.3% vs. 10.4% (HR 0.70; 95% CI 0.53-0.91)

60-69y: 13.3% vs. 13.7% (HR 0.98; 95% CI 0.79-1.21)

>70y: 17.1% vs. 21.2% (HR 0.78; 95% CI 0.65-0.93)

LDL cholesterol (mmol/L)

<2.5: 11.4% vs. 12.1% (HR 0.94; 95% CI 0.78-1.15)

2.5 -<3.0: 10.9% vs. 13.0% (HR 0.82; 95% CI 0.64-1.05)

≥ 3: 11.4% vs. 15.4% (HR 0.73; 95% CI 0.60-0.88)

Subgroup analyses showed treatment tended to favor those without a history of smoking, high levels of total cholesterol and LDL cholesterol, and higher BMI.

Please see full text of article to see subgroup analyses on other groups, including gender, diabetes status, vascular disease status, BP and full cholesterol profile.

Adverse Events

Muscle pain, any

21.3% vs. 20.8% (p=NS)

Muscle pain, treatment stopped

1.1% vs. 0.06% (p=0.02)

Increased CK

5-10x ULN: 1.1% vs. 1.0% (p=NS)

10-40x ULN: 0.4% vs. 0.3% (p=NS)

>40x ULN: 0.1% vs. 0.1% (p=NS)

Persistently increased transaminases

0.6% vs. 0.6% (p=NS)

Hepatitis

0.5% vs. 0.4% (p=NS for total, infective, non-infective)

Gallstones

Complicated: 1.8% vs. 1.6% (p=0.55)

Uncomplicated: 0.5% vs. 0.6% (p=0.25)

Pancreatitis without gallstones

0.3% vs. 0.6% (p=0.02)

Cancer incidence

9.4% vs. 9.5% (p=NS)

Cancer mortality

2.8% vs. 2.5% (p=NS)

Other

Medication Compliance (at 4 years)

68% in treatment arm

14% in placebo arm taking statins

LDL cholesterol difference (mmol/L) (at 4 years)

-0.84 vs. 0.08

Criticisms

• Lack of statin-only arm makes it unclear if there is a benefit of adding ezetimibe independent of LDL-lowering effects.
• The primary outcome was primarily driven by patients with stage 3 and 4 CKD.
  • Patients with stage 5 CKD (without hemodialysis) were both fewer in number and did not have a statistically significant outcome on subgroup analysis.
  • Subgroup analysis does not show statistically significance of the primary outcome in patients on hemodialysis. The non-statistical relative reduction (approximately 10%) in this group is similar to a previous study, AURORA, which only included patients on hemodialysis. Medication compliance may have been lower in this group.
• Authors state this study shows an RRR of 19%/1 mmol/L LDL, without accounting for the difference in point estimates in different subgroups of CKD (dialysis and stage 5 CKD with non-significant effects)
• SHARP investigators changed the primary endpoint after a statistical check on the trial revealed higher rates of events not known to be affected by LDL cholesterol reductions (33% - noncoronary cardiac deaths, hemorrhagic strokes) than expected, and a lower reduction in LDL cholesterol with treatment than expected (0.85 mmol/L vs. 1.01 mmol/L). This was done to avoid a false negative outcome, although analyses at the end of the study showed both old and new primary endpoints showed similar results at the end of the study.

Funding

• Merck/Schering-Plough (Vytorin, brand name of ezetimibe/simvastatin)
• British Heart Foundation
• National Health and Medical Research Council, Australia
• Medical Research Council