Feature article

Predicting Cognitive Outcomes After Heart Attack with MRI

Cardiac arrest, the sudden loss of heart function that often accompanies a heart attack, affects an estimated 559,000 people in the U.S. each year, according to the American Heart Association.5 A majority of patients who suffer heart attacks, stroke, cardiac arrest, and other head or medical traumas that cause anoxic brain damage resulting in widespread neuronal cell death and survive don’t always regain consciousness immediately afterwards.1-4 They may stay in a coma for hours, weeks, or longer making it difficult to determine their long-term recovery potential, which can include a range of outcomes, from full cognitive function to significant impairment requiring lifelong care in a nursing home, or even death.1-4

Connectome analysis of neurological activity using advanced multimodal MRI imaging techniques to map the brains of patients while in cardiac arrest-induced coma have recently been found to more accurately forecast favorable and unfavorable cognitive outcomes than currently used methods such as EEG or structural measures of tissue damage used in conventional MRI.1-4

Study researchers expect their findings to highlight a potential realm of precision medicine using network connectivity biomarkers that provide reliable and accurate predictions of long-term recovery, as well as help guide development of new therapies for targeted treatments that improve brain function.1-4

Clinicians have long sought improved assessment methods that allow for a better understanding of the magnitude of the neurological injuries and more accurate predictions of recovery.1-4 An inability to predict outcomes for patients with brain trauma makes treatment decisions difficult and impairs the ability to communicate with patient families about what to expect or how to prepare.1-4

Current methods for influencing and managing the magnitude of brain damage and prospects for recovery have been the timeliness and quality of resuscitation and bringing patients' body temperature down to about 33°C for a period of 24 hours to prevent fever, among others.1-4 Presently relied upon models for predicting the likelihood of cognitive recovery have limitations in efficacy and accuracy in determining which patients will emerge from a coma with or without impairment, and are inadequate even for patients with similarly manifesting characteristics who exhibit varying outcomes ranging from death to complete recovery.1-4

Early connectivity mapped with MRI

In a study aimed at improving methods used to predict brain healing and function in the months and years following cardiac arrest, researchers assessed functional activation using advanced imaging techniques to perform quantitative brain mapping with MRI data because of the ability of MRI to specifically and accurately identify changes in tissue structure, blood flow, and functional activation.1-4

Researchers scanned, mapped, compared and analyzed higher-order functional networks, the connectome comprised of different neuronal networks that work together to perform tasks, of 46 comatose patients within one month of having a heart attack with those of 48 healthy age-matched participants using advanced MRI technology and a multimodal MRI sequence including resting-state functional and structural MR imaging.1-4

Functional connectivity is determined by analyzing the correlation of activation in different parts of the connectome in order to establish the strength of connections between anatomically distinct regions.1-4 The degree of correlation can be measured within and between networks that work together to perform tasks.1-4 Resting-state functional MRI (fMRI) data analysis shows where brain network disruption occurs and how the disruptions relate to the likelihood of recovery from brain damage.1-4

Within- and between-network connectivity was measured using MRI advanced techniques including diffusion tensor imaging (DTI) and resting-state fMRI to focus on the brain’s large-scale functional integration.1-4

DTI is an emerging MR imaging technique for evaluating the microstructure of well-organized biologic tissues such as muscles and nerves by producing three-dimensional maps of water molecule movements.6 The two main parameters of measurement, fractional anisotropy and the diffusion coefficient, allow earlier detection of architectural changes occurring in injured tissue even when no abnormality may be seen with conventional MR imaging.6

Resting-state fMRI advantages over task-based fMRI include the ability to explore all resting-state networks and their inter-relationships in the connectome simultaneously compared to just one and its related network with task-based.7 Additionally, rest-state fMRI can be used with patients incapable of performing task-based fMRI such as children, dementia patients, or those who have been sedated or are in a coma.7

Large-scale neuronal integrations are organized into separate networks possessing features aligned with known sensorimotor and cognitive systems that change over time and in relation to processes such as sleep and learning.1-4 Using seed-based analysis of resting-state fMRI data, the analysis focused on four distinctive networks in the brain:1-4

  • Dorsal attention network (DAN), which is active when a person uses energy to focus attention)1-4
  • Default mode network (DMN), which is active when an individual is at rest and is the distinguishing network affected in Alzheimer’s disease, as well as other cognitive impairments1-4,8
  • Executive control network (ECN), which is active while initiating tasks and is associated with reward and inhibition1-4
  • Salience network (SN), a network that determines the importance of stimuli and may direct activation of other cognitive networks1-4

Structural changes identified with fluid-attenuated inversion recovery and diffusion-weighted imaging sequences were analyzed with validated morphologic scales.1-4 The association between connectivity measures, structural changes, and the principal outcome was analyzed using multiple variables to forecast possible outcomes.1-4

Buyers Guide_950x250.jpg

Functional connectivity abnormalities and anti-correlation

One year later, researchers assessed the cardiac arrest-induced coma patients using the Cerebral Performance Category (CPC) scale, a commonly used measure of neurological function.1-4 The test uses a scale from one to five, with one indicating minimal to no disability and five indicating brain death.1-4 Eleven of the initial 46 patients were assessed as favorable outcomes or a one or two on the CPC scale.1-4

Researchers saw a distinct contrast when they compared the brain imaging results of patients who had favorable outcomes with those who did not.1-4 While data showed abnormalities in the resting networks of the 46 comatose patients compared to the 48 healthy participants in the control group, there were also abnormalities found between the 11 patients who recovered with full or almost full brain function and those that did not.1-4 Patients assessed as having experienced a favorable outcome had greater/stronger preservation of functional connectivity within the resting networks, particularly within the resting DMN and SN.1-4 Subjects who experienced the most favorable results had higher within-DMN connectivity, suggesting that the higher connectivity was linked to a higher likelihood of cognitive recovery.1-4

Greater anti-correlation (when one is active the other is not) between the SN and ECN and the SN and DMN, was observed when comparing the 11 favorable outcome patients with patients who had an unfavorable outcome (CPC greater than two) suggesting anti-correlation was preserved in patients who recovered and eliminated in those who didn’t.1-4 This effect remained even after adjusting for multiple variables.1-4

A key predictor with the highest accuracy was found to be the anti-correlation of SN-DMN, suggesting relative preservation of this anti-correlation was the most robust signal of a favorable outcome.1-4 These two networks are normally anti-correlated, meaning that as the DMN becomes more active, activity is reduced in the SN, and vice versa.1-4 Anti-correlation of SN-DMN predicted outcomes with higher accuracy than fluid-attenuated inversion recovery or diffusion-weighted imaging scores.1-4

Results of the study demonstrated that anatomical and functional abnormalities in long-range connectivity occur early after cardiac arrest within and between brain networks in the acute phase of anoxic brain injury and are independently associated with long-term functional outcome.1-4 Researchers suggest the early changes in network connectivity could represent a biomarker of recovery potential in patients with cardiac arrest-related brain damage.1-4

Future connectome outcomes

Though the sample size was small, the implications of could soon be huge.1-4 Clinicians may not only depend on advanced MRI techniques to improve treatment selection and outcomes but also as an accurate resource to provide patients and their families with much-needed information about their neurological health.1,2,4 One study left the researchers with the confidence that clinicians can in fact depend on these advanced MRI techniques.3

Researchers don’t expect connectome analysis to be the cure-all solution for predicting cognitive outcomes, but it could increase physicians’ abilities to provide a much better likelihood of what to expect.1-4

Additional implications of fMRI and connectome analysis include guiding treatment and aiding in the development of therapeutic interventions for neurologically disabled patients.1-4


References

  1. Early Functional Connectome Integrity and 1-Year Recovery in Comatose Survivors of Cardiac Arrest. Radiology Journal. https://pubs.rsna.org/doi/abs/10.1148/radiol.2017162161?journalCode=radiology Accessed 5/29/2018
  2. Mapping Brain Connectivity with MRI May Predict Outcomes for Cardiac Arrest Survivors, Study Finds. Johns Hopkins Medicine. https://www.hopkinsmedicine.org/news/media/releases/mapping_brain_connectivity_with_mri_may_predict_outcomes_for_cardiac_arrest_survivors_study_finds. Accessed 5/29/2018
  3. Brain MRI images can predict cognitive function after heart attack. Health Data Management. https://www.healthdatamanagement.com/news/brain-mri-images-can-predict-cognitive-function-after-heart-attack Accessed 5/29/2018
  4. MRI may predict neurological outcomes for cardiac arrest survivors. RSNA on Science Daily. https://www.sciencedaily.com/releases/2017/10/171018090224.htm. Accessed 5/29/2018
  5. Cardiac Arrest Statistics. American Heart Association. http://cpr.heart.org/AHAECC/CPRAndECC/General/UCM_477263_Cardiac-Arrest-Statistics.jsp. 5/29/2018
  6. Diffusion Tensor Imaging in Musculoskeletal Disorders. RSNA RadioGraphics. https://pubs.rsna.org/doi/full/10.1148/rg.343125062 Accessed 5/31/2018
  7. Resting-State Functional MR Imaging: A New Window to the Brain. RSNA Radiology. https://pubs.rsna.org/doi/10.1148/radiol.14132388. Accessed 5/31/2018
  8. Functional Disintegration of the Default Mode Network in Prodromal Alzheimer’s Disease. Journal of Alzheimer’s Disease. https://content.iospress.com/articles/journal-of-alzheimers-disease/jad161120. Accessed 5/31/2018