Aberrant brain-heart coupling is associated with the severity of post cardiac arrest brain injury

Abstract

Objective: To investigate autonomic nervous system activity measured by brain-heart interactions in comatose patients after cardiac arrest in relation to the severity and prognosis of hypoxic-ischemic brain injury.

Methods: Strength and complexity of bidirectional interactions between EEG frequency bands (delta, theta, and alpha) and ECG heart rate variability frequency bands (low frequency, LF and high frequency, HF) were computed using a synthetic data generation model. Primary outcome was the severity of brain injury, assessed by (i) standardized qualitative EEG classification, (ii) somatosensory evoked potentials (N20), and (iii) neuron-specific enolase levels. Secondary outcome was the 3-month neurological status, assessed by the Cerebral Performance Category score [good (1-2) vs. poor outcome (3-4-5)].

Results: Between January 2007 and July 2021, 181 patients were admitted to ICU for a resuscitated cardiac arrest. Poor neurological outcome was observed in 134 patients (74%). Qualitative EEG patterns suggesting high severity were associated with decreased LF/HF. Severity of EEG changes were proportional to higher absolute values of brain-to-heart coupling strength (p < 0.02 for all brain-to-heart frequencies) and lower values of alpha-to-HF complexity (p = 0.049). Brain-to-heart coupling strength was significantly higher in patients with bilateral absent N20 and correlated with neuron-specific enolase levels at Day 3. This aberrant brain-to-heart coupling (increased strength and decreased complexity) was also associated with 3-month poor neurological outcome.

Interpretation: Our results suggest that autonomic dysfunctions may well represent hypoxic-ischemic brain injury post cardiac arrest pathophysiology. These results open avenues for integrative monitoring of autonomic functioning in critical care patients.

Figure 1

Computation of brain–heart interactions through a synthetic data generation model. Bidirectional brain–heart interactions between the brain within the EEG frequencies of interest (delta, theta, and alpha) and the heart within low‐frequency (LF) and high‐frequency (HF) bands were computed using a synthetic data generation model. The model estimates the functional interplay between the brain and the heart by assuming a communication loop in which ongoing EEG activity (EEG_t) modulates autonomic activity (LF_t+1 and HF_t+1), while in turn, ongoing autonomic activity (LF_t and HF_t) modulates EEG activity (EEG_t+1). For each EEG‐ECG frequency band pair, brain–heart coupling strength and complexity are expressed by the median and refined composite multiscale entropy (RCMSE) of the coupling coefficients over the 5 min recording.

Figure 2

Flowchart.

Figure 3

Brain‐to‐heart coupling strength and complexity according to EEG background ACNS classification. Brain‐to‐heart coupling strength (median) in (A) and complexity (RCMSE) in (B) according to EEG background activity following Westhall classification: highly malignant (HM), malignant (M), and benign (B). Overall comparisons were performed using permutation‐based Kruskal–Wallis tests with 10,000 random Monte Carlo permutations and post hoc two‐by‐two group comparisons using Wilcoxon tests. Y‐axes for the brain‐to‐heart coupling strength (median) are in logarithmic scales for ease of visualization, but statistics were performed on raw data. *p ≤ 0.05, **p ≤ 0.01.

Figure 4

Brain‐to‐heart coupling strength according to SSEP and Day 3 NSE levels. (A) Brain‐to‐heart coupling strength according to the results of the somatosensory evoked potentials (SSEP): bilaterally absent N20 (N20‐) versus uni‐ or bilateral N20 presence (N20+). Groups were compared using permutation‐based Wilcoxon tests with 10,000 random Monte Carlo permutations. (B) Permutation‐based Spearman rho (ρ) correlations with 10,000 random Monte Carlo permutations between median brain‐to‐heart coupling indices and NSE levels at Day 3 after CA. All Y‐axes are in logarithmic scales for ease of visualization, but statistics were performed on raw data.

Figure 5

Heart‐to‐brain coupling strength according to EEG patterns, SSEP, and NSE levels. (A) Heart‐to‐brain coupling strength (median) according to EEG background activity (following Westhall classification: highly malignant (HM), malignant (M) and benign (B)) and, (B) according to the results of the somatosensory evoked potentials (SSEP): uni‐ or bilateral N20 presence (N20+) or bilaterally absent N20 (N20‐). (C) Spearman rho (ρ) correlations between median heart‐to‐brain coupling and NSE levels at Day 3 after CA. All Y‐axes are in logarithmic scales for ease of visualization, but statistics were performed on raw data with permutation tests with 10,000 random Monte Carlo permutations. *p ≤ 0.05, **p ≤ 0.01.

Figure 6

Poor outcome predictive performances of NSE, LF/HF, EEG alpha power and alpha‐to‐heart markers. Receiver Operating Channels curves and corresponding AUC (with 2000 bootstrap replicates confidence interval) for poor outcome prediction at 3 month (CPC 3 to 5) of the best clinical marker (NSE at day 3 > 60 μg/L) and its combination with the significant HRV marker (LF/HF ratio), the significant EEG power marker (alpha power), or the significant brain–heart metrics (alpha‐to‐LF and alpha‐to‐HF strength and complexity). The association of brain–heart metrics to NSE was significantly more performant that NSE alone (p = 0.0075) or than NSE combined with EEG alpha power (p = 0.0171), but not from NSE combined with LF/HF (p = 0.0789).

Figure 7

Heartbeat‐evoked potentials according to patients' neurological outcome. (A) Group median (and median absolute deviation) heartbeat‐evoked potential (HEP) time courses with respect to the R‐peak (0 sec) over the left and right central and prefrontal channels in the good and bad outcome groups and the difference between the two (good minus bad). (B) HEP scalp topography distribution in the interval 200–400 ms with respect to the R‐peak of the group median difference (left) and the corresponding statistics based on cluster permutation analyses of Wilcoxon rank sum test (right). No significant differences were found between the patients with good and bad neurological outcome (all z‐values within [−1.96; 1.96]).