Reduction in stroke risk: no role of pleiotropic effects of cholesterol-lowering therapies

Takeaway

• A meta-regression of cholesterol-lowering trials (including PCSK9 inhibitors) showed that a decrease in total cholesterol corresponds to a log-linear relationship with stroke reduction.
• Cholesterol lowering is associated with a proportional decrease in the risk for stroke, without the need of invoking 'pleiotropic' effects of the therapy.

Why this matters

• The causal role of cholesterol in coronary heart disease has been well established; however, the relationship between cholesterol levels and stroke is not clearly understood.
• Although trials with statin have demonstrated a reduced incidence of stroke in treated individuals, trials with non-statin drugs or therapies have not done so, thereby supporting the theory that such effects are attributable to the 'pleiotropic' properties of statins.
• The ability of the PCSK9 inhibitors evolocumab, bococizumab and alirocumab to reduce total and low-density lipoprotein (LDL) cholesterol by a mechanism completely independent of the mevalonate pathway was assessed by theFOURIER,SPIRE-1/SPIRE-2 andODYSSEY OUTCOMEStrials, respectively.

Study design

• A comprehensive meta-regression of all reported trials on cholesterol lowering (FOURIER, SPIRE-1/2 and ODYSSEY OUTCOMES).
• Key outcomes included:
  • Establishing the relationship between cholesterol lowering and the risk for total stroke and
  • Presenting a revised equation to predict the extent of total stroke reduction per each definite change in total cholesterol.
• Funding: None.

Key results

• FOURIER
  • Observed relative risk [RR] for total stroke in the original study: 0.79.
  • Predicted RR for total stroke RR using the new meta-regression: 0.79.
• SPIRE-1 and -2
  • Observed RR for total stroke in the original study: 0.60.
  • Predicted RR for total stroke RR using the new meta-regression: 0.79.
• ODYSSEY OUTCOMES
  • Observed RR for total stroke in the original study: 0.79.
  • Predicted RR for total stroke RR using the new meta-regression: 0.84.
• The revised equation for predicting the saving of strokes as a function of total cholesterol lowering is LnRR = -0.061–0.005*(% total cholesterol reduction).
• Based on the equation, a RR of 0.851, 0.810 and 0.770 can be expected for a 20%, 30% and 40% reduction in total cholesterol, respectively.

Limitations

• The analysis was based on total cholesterol and not LDL cholesterol; however, LDL cholesterol has a better correlation with stroke.
• Total stroke was used as an endpoint without the distinction of ischaemic and haemorrhagic strokes; however, only ischaemic stroke has a positive relation with total or LDL cholesterol.