HIP-HIP-HOPE-3! Expanded role for statin therapy for primary prevention


Written By

Reviewed By


Yusuf S, Bosch J, Dagenais G, et al. Cholesterol lowering in intermediate-risk persons without cardiovascular disease. N Engl J Med 2016; 374:2021-31.

Primary prevention is rooted at the foundation of public health by promoting interventions to decrease healthcare utilization and disease burden.  Cardiovascular disease (CVD) remains the leading cause of death not only in the United States but worldwide — causing 18 million global deaths annually and estimated to increase to 23 million by 2030.1,2 Stroke is the third leading cause of death in the United States but the second leading cause worldwide.1,2 Heart disease and stroke are costly and disabling. But there is HOPE!


The HOPE-3 trial sought to determine if blood pressure (BP) and cholesterol-lowering therapies are effective and safe as primary prevention strategies in intermediate risk patients. The results of the lipid-lowering arm of the HOPE-3 study were recently published.3  Preceding HOPE-3, the JUPITER and MEGA trials, as well as the CTT meta-analysis, demonstrated the effectiveness of statins to lower cardiovascular events rates; but the generalizability of the findings have been questioned due to the relatively homogenous populations enrolled.4,5,6 Moreover, the AFCAPS-TexCAPS, ASCOT-LLT, and CARDS trials proved statins to be beneficial in several high risk populations — but again, these studies did not enroll many women and few non-Caucasians.7,8,9  The HOPE-3 trial enrolled a much broader population, including women, Asian, and Hispanic patients, and considered non-traditional factors that may affect CV risk such as apolipoprotein B and high sensitivity-C-reactive protein (hs-CRP).3


The trial was multicenter, double-blind, randomized, and placebo-controlled using a 2-by-2 factorial design.  Participants were recruited and followed at 228 centers in 21 countries. Participants did not have CVD but were at intermediate risk with an annual risk of a major CV event of approximately 1%. The inclusion criteria:  men ≥55 years old and women ≥65 years old with at least 1 cardiovascular risk factor.  Risk factors included elevated waist-to-hip ratio; low high-density lipoprotein; tobacco use; dysglycemia; premature coronary disease in first degree relatives; mild renal dysfunction.   Women between 60 and 65 years of age could enroll if they had at least 2 risk factors. Patients were excluded if they had an indication for statin, angiotensin converting enzyme inhibitor, angiotensin receptor blocker, or thiazide diuretic therapy; symptomatic hypotension; chronic liver disease; abnormal liver function tests; inflammatory muscle disease; moderate renal dysfunction; or currently used cyclosporine or fibrate therapy.


A total of 12,705 participants enrolled in the study, consisting of 46% women, about 50% Asian, 20% Caucasian, 27% Hispanic, and 2% Black patients. The “average” participant was 66 years old, with a body mass index (BMI) of 27, systolic blood pressure of 138 mmHg, fasting plasma glucose of 95 mg/dL, total cholesterol of 200 mg/dL, LDL of 128 mg/dL, and hs-CRP of 2 mg/L. At the trial initiation, patients entered into a single-blind run-in phase, receiving active BP- and cholesterol- lowering therapy for 4 weeks. Patients who took at least 80% of the prescribed doses were then randomized to receive a fixed combination of candesartan 16 mg plus hydrochlorothiazide 12.5 mg daily or placebo (BP arm of study) and randomized to either rosuvastatin 10 mg daily or placebo (lipid-lower arm of study).  Patients were then followed for 5.6 years.


The first co-primary outcome was the composite of death from CV causes, nonfatal myocardial infarction (MI), or nonfatal stroke.  The second co-primary outcome included all these components plus resuscitated cardiac arrest, heart failure, and cardiac revascularization. The secondary outcome included the second co-primary outcome plus angina with evidence of ischemia. Additional outcomes included death from any cause, components of the co-primary and secondary outcomes, and new-onset diabetes. The expected event rate of 1% annually for the first co-primary outcome in the dual-placebo group with a mean follow-up of 5.5 years.  Other assumptions included a cumulative non-adherence rate of 23% in active treatment groups and a drop-in rate of 11% over 5 years.  Thus 12,700 participants were needed to achieve 80% power to detect a 22.5% reduction in event rate with rosuvastatin. P-values for significance varied: p<0.04 (first co-primary outcome); p<0.02 (second co-primary outcome); p<0.05 (all other analyses). The statistical analysis included an intention-to-treat population and a post-hoc analysis was performed to identify incidence of secondary events.


The incidence of the first co-primary outcome was significantly lower with 235 events occurring in the rosuvastatin group versus 304 in the placebo group (HR 0.76, p=0.002) and a number needed to treat (NNT) of 91. The incidence of the second co-primary outcome was also significantly lower with 277 events occurring in the rosuvastatin group versus 363 events in the placebo group (HR 0.75, p<0.001) and a NNT of 73. Other outcomes are listed in Table 1.


Table 1: Component Efficacy Outcomes3


Rosuvastatin Group (N=6361)

Placebo Group (N=6344)

Hazard Ratio   (95% CI)

P value


70 (1.1%)

99 (1.6%)



Ischemic stroke

Fewer ischemic strokes occurred in the rosuvastatin group (41 vs. 77)



Hemorrhagic stroke

Slightly more hemorrhagic strokes occurred in the rosuvastatin group (11 vs. 8)

Equal number of subarachnoid hemorrhages (4)




45 (0.7%)

69 (1.1%)



New-onset diabetes

232 (3.9%)

226 (3.8%)



Coronary heart disease

105 (1.7%)

140 (2.2%)




  CV deaths

154 (2.4%)

171 (2.7%)



  Non-CV deaths

180 (2.8%)

186 (2.9%)



  Total deaths

334 (5.3%)

357 (5.6%)




LDL (mg/dL)

Mean overall difference: 34.6 mg/dL (25.6%) in the rosuvastatin group


Triglycerides (mg/dL)

Mean overall difference: 21.1 g/L lower in the rosuvastatin group


Apolipoprotein B (g/L)

Mean overall difference: 0.23 mg/L (22%) lower in the rosuvastatin group


hs-CRP (mg/L)

Mean overall difference: 0.19 g/L lower in the rosuvastatin group



Safety data demonstrated that more patients experienced adverse effects, specifically myalgias and cataract surgeries, in the rosuvastatin group.  However, there was no significant difference in the rate of discontinuation due to muscle pain or rhabdomyolysis. The safety outcomes are listed in Table 2.


Table 2: Safety Outcomes3


Rosuvastatin Group (N=6361)

Placebo Group (N=6344)

Hazard Ratio   (95% CI)

P value

Hospitalizations for CV causes

281 (4.4%)

369 (5.8%)



Muscle pain

367 (5.8%)

296 (4.7%)



*No difference between groups for medication discontinuation rates due to muscle pain
[83 (1.3%) vs. 76 (1.2%), P=0.63]
or rhabdomyolysis (1 vs. 0)

Cataract surgery

241 (3.8%)

194 (3.1%)














Key strengths of HOPE-3 are the study design features (randomized, double blind, intention-to-treat analysis) as well as the strong representation of women and various ethic/racial groups.  Patients in this study were truly at intermediate risk for CVD.  However, a lack of information regarding medication adherence monitoring is a potential limitation.  Additionally, the study duration was relatively short in relation to the lifetime risk for developing CVD.  Thus the long-term risks and benefits aren’t known.


It should be noted that 50% of the participants in the HOPE-3 study were Asian.  While Asian patients are known to have a more robust response to statin therapy (in terms of LDL lowering), the LDL-C reduction seen in HOPE-3 was less than 30%.  Could poor adherence to therapy have been the culprit?


The trial demonstrates statin therapy is not associated with moderate to severe ADRs such as cancer, liver dysfunction, or muscle toxicity – adverse effects which commonly concern practitioners and patients. Unlike the JUPITER trial, new-onset diabetes did not occur more frequently in patients treated with rosuvastatin in the HOPE-3 study.3,5 However, the dose of rosuvastatin was lower (10mg vs. 20mg) and the patient population was at lower risk of developing type 2 diabetes (e.g. lower baseline BMI and inflammatory markers) in HOPE-3.


The American College of Cardiology and American Heart Association (ACC/AHA) guidelines support a risk-based approach to statin use but in intermediate risk patients the tradeoffs between benefit and risk were deemed “less clear.”11  The results of HOPE-3 trial provide strong evidence to support treating intermediate risk patients – moving the benefits from “less certain” to certain.


The HOPE-3 trial should expand the use of statins for primary prevention.  The data is particularly important for women and non-Caucasian patients. This intervention has the potential to significantly decrease global healthcare costs by minimizing cardiovascular disease burden, preventing hospitalizations, and minimizing polypharmacy.  Can these results be extrapolated to all moderate intensity statin therapy?  Or should we favor rosuvastatin use?  What do you think?


Evan Williams's picture

The non-significant occurrence of new onset diabetes is encouraging, but the author is correct to point out that the dose of rosuvastatin and risk factors associated with development of diabetes in the study population may have impacted the result. I also wonder what impact the high numbers of Asian and female patients has in the development of new onset diabetes. It will be interesting to see a longer term follow up of these patients.