Discordance between LDL-C and apolipoprotein B is associated with large-artery-atherosclerosis ischemic stroke in patients ≤70 years of age

Background and aims: Low density lipoprotein (LDL-C) and other atherogenic lipoproteins are coated by apolipoprotein B100 (apoB). The correlation between LDL-C and apoB is usually thight, but in some cases LDL-C underestimates apoB levels and residual cardiovascular risk. We aimed to assess if a discordance of LDL-C-levels with apoB levels is associated with LAA stroke. Methods: We included patients with an acute ischemic stroke from two prospective studies enrolled at the University Hospital Bern, Basel and Zurich, Switzerland. LDL-C and apoB were measured within 24 h of symptom onset. By linear regression, for each LDL-C, we computed the expected apoB level assuming a perfect correlation. Higher-than-expected apoB was defined as apoB level being in the upper residual tertile. Results: Overall, we included 1783 patients, of which 260 had a LAA stroke (15%). In the overall cohort, higher-than-expected apoB values were not associated with LAA. However, a significant interaction with age was present. Among the 738 patients <= 70 years of age, a higher-than-expected apoB was more frequent in patients with LAA- versus non LAA-stroke (48% vs 36%, p = 0.02). In multivariate analysis, a higher-than-expected apoB was associated with LAA stroke (aOR = aOR 2.48, 95%CI 1.14-5.38). Among those aged <= 70 years and with LAA, 11.7% had higher than guideline-recommended apoB despite LDL-C <= 1.8 mmol/L (<70 mg/dl), compared to 5.9% among patients with other stroke etiologies (p = 0.04). A triglyceride cut-off of >= 0.95 mmol/L had, in external validation, a sensitivity of 71% and specificity of 52% for apoB >= 0.65 g/L among patients with LDL-C <1.8 mmol/L. Conclusions: Among patients aged <= 70 years, a higher-than-expected apoB was independently associated with LAA stroke. Measuring apoB may help identify younger stroke patients potentially benefiting from intensified lipid-lowering therapy.