The results of this study demonstrate that LA influences clinical outcome after ischemic stroke and that this effect appears to be independent of the volume of brain tissue that is irreversibly injured by the acute insult. Our findings also suggest that LA volume may mediate at least a portion of the influence of age on stroke outcome as the effect of age increases when LA is removed as a predictor of clinical outcome.
Despite the fact that LA has consistently been shown to predict development of poststroke dementia,9
prior studies of LA and stroke outcome have been inconclusive; a direct effect of LA on stroke morbidity and mortality has been demonstrated in some20,21
but not other7,22
studies. These discrepancies, in part, resulted from the use of visual scales to quantify LA burden instead of precise volumetric assessment. Prior studies also varied in the scales used to measure stroke outcome. Furthermore, LA burden was often measured using CT rather than MRI, which is far less sensitive to its detection,23
and poses particular challenges in assessing its severity. The present study quantitatively measured LA volume on MRI as opposed to employing a visual rating scale. Our findings demonstrate a small but significant shift in mRS toward less favorable outcomes with increasing LA volume. The relationship between LA volume and outcome persisted after adjustment for a comprehensive list of previously validated predictors of poststroke clinical outcome including age, initial stroke severity, and infarct volume. The small correlation between LA volume and clinical outcome (r
= 0.19) probably reflects the fact that individual risk factors in multifactorial conditions such as stroke outcome can only account for a small proportion of the overall variance in the risk. Nevertheless, the highest quartile of LA volume was associated with a 1.5-fold increase in mRS as compared with the lowest quartile. The absolute mean difference in mRS between the lowest and highest LA quartiles was 0.8. This is a robust difference and compares well with the data from several recent therapeutic stroke trials.24,25
For instance, in the National Institute of Neurological Disorders and Stroke tissue plasminogen activator trial, the mean mRS was 2.66 in the treatment and 3.19 in the placebo groups, revealing an absolute difference of only 0.53.26
The relationship between LA and clinical outcome could involve multiple mechanisms. It is well known that LA regions have reduced vascular density27
and cerebral blood flow,28
which may lead to infarct growth in the acute setting by preventing peripheral compensation during ischemic stress.29
Alternately, the presence of dysfunctional neuronal networks in patients with LA may be partially responsible for the association between high LA burden and unfavorable outcome. An intact system of intrahemispheric and interhemispheric connectivity is essential for favorable functional recovery following stroke.30
Neuropathologic and functional neuroimaging studies have shown the presence of decreased neuronal connectivity in patients with LA,31,32
which is most likely due to the demyelination, loss of axons and oligodendrocytes, and astrocytic gliosis.15
This may result in decreased functional connectivity of distant cortical regions,33
which could impair plasticity and inhibit recovery. This mechanism is supported by the observations that LA is associated with poststroke dementia9
and that subcortical lesions can reduce effective interhemispheric interactions in a manner that correlates with a patient’s level of motor impairment.30
Finally, LA is a well-known risk factor for developing poststroke cognitive impairment and depression, which may ultimately adversely affect patients’ compliance with treatment and recovery programs.9,34
Further research using objective measures of connectivity, infarct evolution, and baseline cognitive functioning is needed to elucidate the exact mechanism of the relationship between LA and clinical outcome.
Age is one of the most consistent clinical variables associated with clinical outcome in the literature; however, most published studies supporting the role of age on stroke outcome included mortality as part of the outcome measure.2,35
Our findings demonstrate that the effect of age on outcome is of varying degrees and is dependent on the inclusion of mortality as an outcome measure. The deleterious effect of age on clinical outcome appears to be mediated in part by LA volume; age is not an independent predictor of clinical outcome when LA volume is accounted for. In contrast, age is a strong predictor of outcome regardless of LA volume if the outcome measure includes mortality. There may be multiple reasons for this finding. First, the strong correlation between LA volume and age might have decreased the statistical power of the regression models. An alternative explanation could be that age has a differential influence on recovery vs mortality following stroke, especially given the fact that a significant portion of deaths in patients with stroke are not attributed to the stroke but to age-related comorbid conditions.19
The present study has a number of strengths and limitations. The strengths include a relatively large sample size, blind assessment of DWI and LA volumes with respect to clinical outcome, volumetric assessment of LA burden, rigorous adjustment for potential confounders, and the use of mRS as an ordinal variable to prevent the loss of information that occurs during dichotomization.36
An important limitation of our study is its observational prospective nature that resulted in several inevitable dropouts; patients did not consent or could not be scanned because of MRI contraindications or were not available for follow-up assessment. The consent and follow-up bias might have resulted in a systematic error toward selection of a more favorable outcome population37,38
; the present cohort was considerably skewed toward milder strokes at baseline (median admission NIHSS score: 6) and more favorable clinical outcome at 6 months (median mRS: 1) as compared with other published consecutive stroke cohorts.1
Nevertheless, because LA volume correlates with stroke severity, our inability to include severe strokes likely resulted in an underestimation of the observed magnitude of relationship between LA volume and clinical outcome. The current study was restricted to patients who had DWI and FLAIR obtained within 24 hours of symptom onset. As a result, patients with severe strokes who were not clinically stable enough for an early MRI, patients with mild stroke with delayed hospital admission, patients with grave short-term prognosis, and patients with contraindications for MRI were missed. However, neither baseline stroke severity nor age, historically the two strongest predictors of stroke outcome, differed among patients with or without admission MRI, arguing against a significant bias in patient selection.
It has long been known that advanced age hinders recovery from acute ischemic stroke. While it is generally assumed that the mechanisms underlying the effect of advanced age on stroke recovery are associated with age-related comorbidities such as osteoporosis, osteoarthritis, and coronary artery disease, it is clear that changes within the brain itself contribute substantially to the brain’s response to injury. LA is emerging as a potent manifestation of the cerebrovascular aging process and a marker of risk for stroke as well as age-related cognitive and gait deterioration.9,39
Our data suggest that the conditions that manifest as LA also contribute to the brain’s recovery from stroke and, because of their relationship to the aging process, underlie a portion of the effect of age on stroke outcome. These observations suggest that therapies that slow the progression of LA hold promise not only for reducing the incidence of stroke, but also reducing stroke severity and improving clinical outcome when it occurs. Indeed, as LA volume appears to be determined only partially by the same vascular risk factors that predispose to symptomatic stroke, such therapies could well involve novel mechanisms currently not exploited in stroke prevention. Given the prevalence of LA among our aging population, such therapies may transform the aging process for future generations.