Baseline predictors and clinical outcomes of incident infarcts in the year after a mild stroke

Abstract

Introduction

Factors associated with incident infarcts after stroke are unclear. We aimed to determine whether subsequent incident infarcts continue to develop one year post-stroke and how incident infarcts relate to baseline imaging features, vascular risks, and cognitive outcomes.

Methods

We recruited patients with non-disabling stroke. After diagnostic MRI, we repeated MRI at 3-6 monthly intervals for 12 months, visually assessing incident infarcts on DWI or FLAIR. We used logistic regression to determine associations with incident infarcts, including baseline vascular risks, SVD score, and index stroke subtype. We quantified 7-level ordinal cognitive outcome status at one year, using MoCA/telephone MoCA and modified Rankin Scale.[1,2] We used ordinal regression to determine whether cognitive outcomes associated with incident infarcts and baseline age, mRS, MoCA, and WMH volume.

Results

We recruited 229 participants, mean age 65.9 (SD 11.1) years; 77/229 (33.6%) female; 130/229 (56.8%) index lacunar stroke. From baseline to one-year MRI, we detected 117 incident infarcts in n=57/229 participants at 80 visits. Most were small subcortical infarcts: 86/117 (73.5%) infarcts in n=38/57 (66%). N=39 participants had incident infarcts at one visit; n=14 at two visits; n=3 at three visits, and n=1 at four visits. Nineteen participants had multiple incident infarcts at a single visit. Baseline summary SVD score was the strongest predictor of incident infarcts (aOR 1.74, 95%CI 1.29-2.41, Figure 1). At one year, 10/218 (4.6%) participants met criteria for single-domain and 62/218 (28.4%) for multi-domain neurocognitive disorder; 15/218 (6.8%) for mild dementia; 2/218 (0.9%) for moderate dementia; none severe; 3/229 (1.3%) had died. Participants’ odds of impaired one-year cognition increased for every one-unit increment in baseline mRS (aOR=2.03 [1.30-3.10]) and decreased for every one-unit increment in baseline MoCA (aOR=0.78 [0.71-0.85]). For participants with incident small subcortical infarcts, the odds of impaired cognition were 23% higher than for incident cortical infarcts, though not statistically significant (aOR 1.23 [0.55-2.1])(Figure 2).

Discussion

In a mild stroke population, incident infarcts, mostly small subcortical, occur in one quarter and associate with worse baseline SVD. Minor neurocognitive disorder occurs in one third and associates with baseline mRS, MoCA, and trends towards incident small subcortical rather than cortical infarcts.