It has been reported that more than 50 million people suffer from traumatic brain injury (TBI) each year, of which severe TBI (sTBI) is a significant contributor to high mortality and morbidity worldwide (1). To the best of our knowledge, patients with sTBI are heterogeneous and encompass a cascade of brain disorders and secondary injuries (2, 3). With the goal of continuously optimizing the conditions for brain recovery, active brain function monitoring can provide a reliable basis for mechanism-guided management of sTBI (4).

With the advances in intensive care, especially neurocritical management, there is great potential for objectively evaluating the pathophysiology of patients with sTBI (5). However, in most developing countries, people have always been concerned that the treatment and cure come at the cost of an increase in the economic burden and the survivors with severe disabilities. Confounded by uncertainty in sTBI prognoses, many helpless trauma centers and family members have made decisions to withdraw from life-sustaining therapies. It has been reported that early withdrawal of life-sustaining therapies does not have worse outcomes than those making later decisions (6). In a recent study that included 41 randomized trials, the total mortality rate of sTBI was 23%, of which 63% was related to the withdrawal of life-sustaining therapies (7). Therefore, there is an unmet clinical need to identify and establish accurate sTBI predictive biomarkers that correlate with the sTBI pathologies at the acute phase (8).

Clinicians gradually realize that spontaneous brain activity in a resting state is distributed spatially and temporally in multiple brain regions of patients with sTBI (9, 10). With the help of functional magnetic resonance imaging (fMRI), people can identify and describe the widespread disruption of brain network mechanisms, such as the changes in functional connectivity and abnormal brain activation in patients with sTBI (11–13). However, most of the assessments were carried out in the rehabilitation stages of sTBI, and the timeliness of assessment is lost in the acute phase of sTBI. In recent days, studies have reported that there was a compensatory mechanism that several brain regions can converge into abnormal functional hubs in the acute phase of patients with sTBI, which provides a potential approach to predict the long-term outcome of patients (14). However, it is still not clear about the abnormal brain function patterns in which local brain regions play an important role in the formation of this compensation mechanism.

In the study of fMRI, a fractional amplitude of low-frequency fluctuation (fALFF) approach was proposed to reveal the spontaneous activity of the brain regionally (15). In addition, regions of interest (ROI)-based functional connectivity (FC) can be calculated to discrete the various brain functional states of the whole brain spatially (16). It has recently been found that sTBI patients who present a better prognosis showed a significant increase in regional and spatial values of the brain cortical network (11). Based on this further, we hypothesized that (1) changes in residual brain spontaneous activity after sTBI will disrupt the functional pattern of the whole brain, and (2) this early compensatory mechanism can be used for the long-term prognostic prediction of sTBI.

The most important innovation of this study is that we present a compensation mechanism at the acute phase of sTBI; that is, several parts of the brain regions can spontaneously activate and further lead to a decrease in functional connectivity (FC) to the whole brain, which may help to provide information for early diagnosis and prognostic prediction. Another innovation of this study is that we have carried out a gradual functional analysis framework of the sTBI patients’ residual brain activity; we use fALFF to identify the residual brain “activation” and use FC to describe its large-scale “connectome.” Significantly, several activations of brain areas reported in this study provide the early intervention targets for sTBI treatment.

To the best of our knowledge, both local activation and spatial connectivity form the prototype of various brain diseases (16, 18). There are two main methods for investigating the brain functional status in sTBI patients: one is to calculate the functional activity changes in the local ROI—within regions, and the other is to compare the functional interactions between regions, which can calculate the ROI-based FC to the whole brain over time—between regions (19). In this study, we found that fALFF values were mainly increased in the posterior cortex (e.g., occipital lobe, posterior cingulate gyrus, and parietal gyrus) of sTBI patients who present with good outcomes, indicating that these regions with high spontaneous brain functional activity are crucial for sTBI neurological function rehabilitation. The important clinical significance is that these results provide beneficial intervention targets for the early treatment of sTBI. It also reveals that these are essential compensatory brain regions that enable integration and coupling across the whole brain at the acute phase of sTBI (20). These findings consistently revealed that increased brain functional activity is associated with better cognitive performance in patients with chronic TBI (21). In addition, it has been reported that sTBI patients have lower values of local efficiency of their brain than healthy controls, implying altered brain functional organization (22). Our findings extend these findings by revealing an increased fALFF in the posterior cortex of sTBI patients who present with good outcomes, indicating that “from backward to forward” functional reorganization was involved in the good recovery of sTBI.

We further found that the injured brain can compensate itself by enhancing the fALFF signal of the local region, thereby reducing communication with the whole brain. Our ROI-based FC analysis results mainly focus on the brainstem region, indicating the contribution of subcortical regions to large-scale cortical region reorganization and the importance of FC between the cortex and subcortical regions in maintaining physiological function (23). For example, the functional pathway between the brainstem reticular ascending activation system and the cerebral cortex plays a critical role in wake-promoting at the acute phase of TBI (24). Accordingly, our results were not merely a consequence of the decrease in a few FC values, which was significantly associated with poorer balance of brain functional performance. Particularly, sTBI patients with poor outcomes displayed abnormally increased FC to massive brain regions; this reorganization phenomenon caused by injury reveals the importance of brain region coordination and cooperation in maintaining normal brain function (25).

It has been recently reported that the formation of brain functional architecture is an optimal trade-off between promoting communication efficiency and reducing wiring costs (26). At this point, these findings indicated that the synchronization of the whole brain is compromised in the acute phase of sTBI (27). Most interestingly, the dispersed functional hubs (increased in various FCs) could form a less efficient cortical network in the acute phase of sTBI. To the best of our knowledge, previous sTBI studies mainly emphasized a decrease in connectivity between the two ICNs (DMN and SN) (11, 28), and the residual FC could reappear in the chronic phase of sTBI (3 months after injury)(29). However, little is known about the functional compensation mechanisms at the acute phase of sTBI. Accordingly, our findings revealed that massive increased FC in the brain was associated with the severity of brain dysfunction at the acute phase of sTBI, which plays an important role in the poor prognosis in the long-term follow-up. To further investigate the clinical value of such abnormal brain functional patterns, we performed the predictive analysis of altered fALFF and FC on the long-term outcome of sTBI patients. Significantly, our findings suggested that sTBI patients with poor outcomes tend to have increased FC than those with good outcomes; it reveals that the less the compensatory effect of FC, the better the residual brain function of sTBI patients at the acute phase.

In the past, a lot of prognostic models have been developed for patients with sTBI, and the AUC of a model that separates the different outcomes in TBI is approximately 0.8. Although several complex models showed slightly better discrimination ability, most of them are broadly accepted in sTBI research rather than clinical practice (30). In this study, we developed severe high-quality indicators through fMRI scanning in the acute phase of sTBI. This objective evaluation method from the real clinical setting is comparable to complex prediction models of sTBI. Of which fALFF increase in the temporal lobe, occipital lobe, precuneus, and middle temporal gyrus shows the highest predictive ability (AUC = 0.883). These findings indicated that prognostic assessment based on brain functional data has great potential as it could reflect patient characteristics from multiple dimensions. As expected, we will try more schemes (functional and structural) in future research and compare the performance of different combinations.

This study has several limitations that should be highlighted. First, the sample size of this study was relatively small because of the complexity and difficulty of data collection during the acute phase of sTBI. Second, we excluded subjects with motion > 2 mm or 2 degrees of rotation in this study and its potential effect on signals analysis needs further clarification, and the macro-scale fMRI BOLD signals analysis cannot capture the compensation mechanism of sTBI at the neural circuits and molecular level. Third, although many potential confounding factors (cerebrospinal fluid, white matter fibers) have been adjusted, the data bias and the different injury forms or sites may affect the findings. In addition, whether these findings could be extended to other types of brain injury (e.g., cerebral hemorrhage or infarction) is not yet known.

In summary, we present a compensatory mechanism that several regions of the brain can be spontaneously activated at the acute phase of sTBI in those who present with a good prognosis in the 6-month follow-up. The residual brain activities of sTBI could be dominated by the increased fALFF in the local region and decreased FC to the whole brain. Interestingly, such abnormal brain function pattern has a potential capability for predicting the 6-month outcome of patients with sTBI, of which fALFF increased in the temporal lobe, occipital lobe, precuneus, and middle temporal gyrus has the highest distinguishability.

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

The studies involving humans were approved by the Department of Emergency, the Second Affiliated Hospital of Nanchang University, Nanchang, Jiangxi, China. The studies were conducted in accordance with the local legislation and institutional requirements. Written informed consent for participation in this study was provided by the participants' legal guardians/next of kin.

BL: Data curation, Formal analysis, Writing – original draft. WL: Data curation, Formal analysis, Writing – original draft. YG: Conceptualization, Writing – original draft, Methodology, Resources. YW: Writing – original draft, Formal analysis, Funding acquisition. LX: Methodology, Data curation, Formal analysis, Writing – review & editing. YY: Writing – original draft. SX: Writing – review & editing. ZT: Conceptualization, Investigation, Supervision, Writing – original draft. YM: Writing – review & editing. KW: Conceptualization, Data curation, Investigation, Methodology, Project administration, Writing – original draft, Writing – review & editing.