Silent Brain Infarction and Risk of Future Stroke: A Systematic Review and Meta-Analysis

Abstract

Background and purpose: Silent brain infarction (SBI) on magnetic resonance imaging has been proposed as a subclinical risk marker for future symptomatic stroke. We performed a systematic review and meta-analysis to summarize the association between magnetic resonance imaging-defined SBI and future stroke risk.

Methods: We searched the medical literature to identify cohort studies involving adults with SBI detected by magnetic resonance imaging who were subsequently followed up for incident clinically defined stroke. Study data and quality assessment were recorded in duplicate with disagreements in data extraction resolved by a third reader. Strength association between magnetic resonance imaging-detected SBI and future symptomatic stroke was measured by an hazard ratio.

Results: The meta-analysis included 13 studies (14 764 subjects) with a mean follow-up ranging from 25.7 to 174 months. SBI predicted the occurrence of stroke with a random effects crude relative risk of 2.94 (95% confidence interval, 2.24-3.86, P<0.001; Q=39.65, P<0.001). In the 8 studies of 10 427 subjects providing hazard ratio adjusted for cardiovascular risk factors, SBI was an independent predictor of incident stroke (hazard ratio, 2.08 [95% confidence interval, 1.69-2.56; P<0.001]; Q=8.99; P=0.25). In a subgroup analysis pooling 9483 stroke-free individuals from large population-based studies, SBI was present in ≈18% of participants and remained a strong predictor of future stroke (hazard ratio, 2.06 [95% confidence interval, 1.64-2.59]; P<0.01).

Conclusions: SBI is present in ≈1 in 5 stroke-free older adults and is associated with a 2-fold increased risk of future stroke. Future studies of in-depth stroke risk evaluations and intensive prevention measures are warranted in patients with clinically unrecognized radiologically evident brain infarctions.

Keywords: brain infarction; infarction; magnetic resonance imaging; risk factors; stroke.

Figure 1

Forest plots of the association between MRI-determined silent brain infarction and future stroke in all included studies^–. Meta-analysis calculated using a random-effects mode, with crude relative risks pooled in panel A and adjusted-hazard ratios pooled in panel B. Squares represent point estimates for the effect sizes. The size of the squares is proportional to the inverse of the variance of the estimate. Diamond represents the pooled estimate and the horizontal lines represent the 95% confidence interval.