Cortical cerebral microinfarcts on 7T MRI: Risk factors, neuroimaging correlates and cognitive functioning – The Medea-7T study

Maarten H.t. Zwartbol, Ina Rissanen, Rashid Ghaznawi, Jeroen de Bresser, Hugo J. Kuijf, Kim Blom, Theo D. Witkamp, Huiberdina L. Koek, Geert Jan Biessels, Jeroen Hendrikse, Mirjam I. GeerlingsJournal of Cerebral Blood Flow & Metabolism 2021 · DOI

TL;DR

This study leverages high-resolution 7T MRI to investigate cortical cerebral microinfarcts (CMIs) in a diverse cohort of older adults. They found that CMIs, present in 10% of participants, are linked to vascular risk factors, other brain infarcts, and importantly, are associated with poorer global cognitive function, particularly in individuals with more than two CMIs.

What Problem Does This Paper Address?

Cortical cerebral microinfarcts (CMIs) are small, often "silent" lesions in the brain's cortex caused by localized ischemia. While their presence has been increasingly detected with advanced MRI techniques, their prevalence, associated risk factors, and impact on cognitive function remain areas of active research. Previous studies have yielded varying CMI frequencies, partly due to differences in MRI field strength and detection criteria. Understanding CMIs is crucial because they are thought to contribute to the cumulative burden of cerebrovascular disease, which is a major driver of cognitive decline and dementia in aging populations. This paper aims to provide a more precise understanding of CMI occurrence and their associations by using state-of-the-art 7T MRI in a well-characterized, diverse cohort.

Methods

The researchers utilized a 7T MRI scanner, which offers superior spatial resolution compared to lower field strength scanners (like 1.5T or 3T), allowing for more sensitive detection of small lesions like CMIs. CMIs were visually rated by a neuroradiologist blinded to patient information, following established criteria for strictly intracortical lesions visible on T1-weighted and T2-weighted images. They also assessed other neuroimaging markers of cerebrovascular disease (e.g., infarcts, lacunes, white matter hyperintensities, cerebral blood flow) using both 7T and 1.5T/3T MRI data. Vascular risk factors and cognitive function were assessed through questionnaires and standardized neuropsychological tests. Statistical analyses, including modified Poisson regression, were employed to determine the associations between CMIs and various factors, adjusting for age, sex, and other potential confounders.

Datasets

The study utilized data from the "Medea-7T study," a cohort study at the University Medical Center Utrecht, the Netherlands. This cohort comprised 386 participants recruited from four settings:

  1. General practice patients
  2. Participants of the PREDICT-MR study
  3. Participants of the SMART-MR study (patients with arterial disease)
  4. Memory clinic patients (with normal cognition, vascular disease, or mild cognitive impairment/early Alzheimer's disease).

This diverse recruitment strategy aimed to capture a broad spectrum of individuals, from healthy older adults to those with existing cerebrovascular or cognitive issues.

Key Insights and Ideas

  • 7T MRI is a Game Changer for CMI Detection: The use of 7T MRI significantly enhances the ability to detect CMIs, potentially explaining why their reported frequency in this study (10%) is lower than some previous 7T studies but still higher than many 3T studies. This highlights the importance of imaging technology in understanding the prevalence of these subtle lesions.
  • CMIs are a Marker of Broader Vascular Disease: The strong association of CMIs with other infarcts (cortical, lacunar, cerebellar) and decreased cerebral blood flow suggests they are not isolated events but rather part of a larger picture of cerebrovascular compromise, particularly related to large vessel disease.
  • Cognitive Impact is Dose-Dependent: The finding that having more than two CMIs, rather than just the presence of any CMIs, is significantly associated with lower global cognitive performance is a crucial insight. This suggests a threshold effect, where a cumulative burden of CMIs has a more pronounced impact on cognition.
  • Extracranial Atherosclerosis Links to CMIs: In patients with manifest arterial disease, markers of extracranial atherosclerosis (like intima-media thickness and carotid stenosis) were strongly associated with CMIs, reinforcing the idea that systemic vascular health is intrinsically linked to brain health.

Key Results

  • Prevalence: CMIs were found in 10% (n=35) of the 386 participants. Frequencies varied by cohort, ranging from 0.6% in general practice to 13.2% in the SMART-MR cohort.
  • Risk Factors: Increased CMI risk was associated with older age (RR=1.79 per 10 years), male sex (RR=3.08), and a history of stroke or TIA (RR=4.03).
  • Neuroimaging Correlates: CMIs were strongly associated with the presence of other infarcts, particularly lacunes (RR=5.66) and cortical infarcts (RR=5.28). Decreased cerebral blood flow was also linked to increased CMI risk (RR=1.35 per -100 ml/min).
  • Cognitive Functioning: Participants with more than 2 CMIs had significantly lower global cognitive functioning (mean difference = -0.48 SD) compared to those without CMIs. This effect was most pronounced for executive functioning and working memory.
  • Extracranial Atherosclerosis (SMART-MR cohort): Higher intima-media thickness, ≥50% carotid stenosis, and ankle-brachial index <0.8 were all associated with increased CMI risk.

Limitations and Future Work

The authors acknowledge several limitations:

  • Cross-sectional Design: The study cannot establish causality; it's unclear if CMIs cause cognitive decline or if shared risk factors lead to both.
  • Treatment Effects: Prior treatment of vascular risk factors might have influenced associations.
  • Cohort Pooling: While increasing generalizability, pooling diverse cohorts with different inclusion/exclusion criteria could lead to under- or overestimation of associations.
  • MRI Protocol Differences: While efforts were made to standardize, variations in MRI quality across cohorts might have affected lesion detection. Ideally, all markers would be assessed with 7T MRI.
  • Statistical Power: The relatively small number of participants with 1-2 CMIs or >2 CMIs might limit the power to detect associations in specific cognitive domains.

Future work could involve longitudinal studies to establish causality, investigate the role of cerebral metabolism, and explore the impact of CMIs on specific cognitive trajectories in more detail. The study also highlights the need for standardized CMI detection criteria across different MRI field strengths.

Summary updated

Abstract

We determined the occurrence and association of cortical cerebral microinfarcts (CMIs) at 7 T MRI with risk factors, neuroimaging markers of small and large vessel disease, and cognitive functioning. Within the Medea-7T study, a diverse cohort of older persons with normal cognition, patients with vascular disease, and memory clinic patients, we included 386 participants (68 ± 9 years) with available 7 T and 1.5 T/3T brain MRI, and risk factor and neuropsychological data. CMIs were found in 10% of participants and were associated with older age (RR = 1.79 per +10 years, 95%CI 1.28-2.50), history of stroke or TIA (RR = 4.03, 95%CI 2.18-7.43), cortical infarcts (RR = 5.28, 95%CI 2.91-9.55), lacunes (RR = 5.66, 95%CI 2.85-11.27), cerebellar infarcts (RR = 2.73, 95%CI 1.27-5.84) and decreased cerebral blood flow (RR = 1.35 per -100 ml/min, 95%CI 1.00-1.83), after adjustment for age and sex. Furthermore, participants with >2 CMIs had 0.5 SD (95%CI 0.05-0.91) lower global cognitive performance, compared to participants without CMIs. Our results indicate that CMIs on 7 T MRI are observed in vascular and memory clinic patients with similar frequency, and are associated with older age, history of stroke or TIA, other brain infarcts, and poorer global cognitive functioning.
BibTeX
@article{zwartbol2021cortical,
  author = {Maarten H.T. Zwartbol and Ina Rissanen and Rashid Ghaznawi and Jeroen de Bresser and Hugo J. Kuijf and Kim Blom and Theo D. Witkamp and Huiberdina L. Koek and Geert Jan Biessels and Jeroen Hendrikse and Mirjam I. Geerlings},
  title = {{Cortical cerebral microinfarcts on 7T MRI: Risk factors, neuroimaging correlates and cognitive functioning – The Medea-7T study}},
  journal = {Journal of Cerebral Blood Flow \& Metabolism},
  year = {2021},
  doi = {10.1177/0271678x211025447},
}
Figure