Abstract
Objective:
Elevated intracranial pressure (ICP) is a complication of severe traumatic brain injury (TBI) that carries a risk of secondary brain injury. This study investigated the association between ICP burden and brain injury patterns on magnetic resonance imaging (MRI) in children with severe TBI.
Design, setting, patients:
Secondary analysis of the Approaches and Decisions in Acute Pediatric TBI (ADAPT) study, which included children with severe TBI (Glasgow Coma Scale score <9) who received a clinical MRI within 30 days of injury. We excluded patients who had ICP monitoring <24 hours, were missing ICP data for >40% of monitoring time, or who underwent craniectomy.
Interventions:
None
Measurements and Main Results:
ICP burden was defined as the trapezoidal area under the curve of hourly ICP >20 mmHg. ICP was standardized to total monitoring time, and patients were categorized to four levels of ICP burden. MRI was evaluated for number of diffuse axonal injury (DAI) microhemorrhages, intracerebral hemorrhage (ICH) volume, contusion volume, and number of regions with ischemia. Chi square or Fisher Exact Tests were used to test the independence between ICP burden and MRI injury amount. Of the 220 patients, 156 (71%) had DAI, 31 (14%) had ICH, 161 (73%) had contusions, and 70 (32%) had ischemia on MRI. Most patients (180, 82%) experienced episodes of ICP >20 mmHg. Contusion volume (p=0.02) and number of regions with ischemia (p=0.007) were associated with ICP burden, but we failed to identify such an association for DAI or ICH. Severe (but not mild or moderate) ICP burden was associated with presence of ischemia (odds ratio 4.64 [95% Confidence Interval 1.30, 19.5], p=0.02).
Conclusions:
Elevated ICP was prevalent in the ADAPT cohort. Ischemia and contusion were associated with the burden of ICP. Further research is needed to determine temporal relationships between elevated ICP and ischemia.
Keywords: traumatic brain injury, pediatrics, neuroimaging, intracranial hypertension, intracranial pressure, MRI
Internationally (1990–2021), traumatic brain injury (TBI) remains a leading cause of morbidity in children (1, 2). Historically, clinical cohorts covering the period 1976 through to 2011 support the idea that increased intracranial pressure (ICP) is one of the primary drivers of secondary brain injury after TBI and is associated with worse outcomes (3–8). The 2019 Brain Trauma Foundation (BTF) pediatric TBI guidelines suggest clinicians use ICP monitoring based on level 3 evidence (9). As such, the decision to place an ICP monitor is largely institution- and surgeon-dependent, with a general approach of placing an ICP monitor in children with severe TBI (i.e., Glasgow Coma Scale [GCS] less than 9) who do not have significant improvement in neurological examination after initial resuscitation. An area of ongoing investigation focuses on identification of early biomarkers that could improve a clinician’s ability to predict which patients are at highest risk of developing intracranial hypertension. Diagnostic computed tomography (CT) of the head is considered standard of care in severe TBI. However, CT findings have not been clearly demonstrated to predict which pediatric or adult patients go on to develop intracranial hypertension using datasets from 2002–2006 (10) and from 2000–2014 (11). In this context, brain magnetic resonance imaging (MRI) offers an opportunity to detect injury patterns (such as ischemia and diffuse axonal injury [DAI]), which may be missed on CT. Two publications in 2024 reported an association between the burden of ischemia and contusion identified on early clinical MRI and short-term and long-term functional outcomes in children with severe TBI. One report reviewed data from a single-center 2010–2021 (12), and the other report review data from the 2014–2017 ADAPT (Approaches and Decisions for Acute Pediatric TBI) multicenter, international trial dataset (13). Whether these patterns of injury are associated with elevated ICP remains unknown.
In this secondary analysis of the ADAPT trial MRI data (13), which recruited children with severe TBI undergoing ICP monitoring, we aimed to answer the following two questions: 1) what is the burden of elevated ICP in children with severe TBI?; and 2) is there an association between patterns of brain injury on MRI and ICP burden? In exploratory analysis we also aimed to determine if there is a temporal association between ICP burden and ischemia, which may be both a cause and a consequence of elevated ICP. We hypothesized that increased volume of contusion, increased volume of intracerebral hemorrhage (ICH), increased burden of DAI microhemorrhages, and increased burden of ischemia would be associated with increased burden of ICP.
METHODS
The ADAPT trial was a prospective multicenter observational study that enrolled 1000 children with severe TBI from February 2014 to September 2017. The study included sites in 8 countries (US, UK, Spain, the Netherlands, Australia, New Zealand, South Africa, India). Twenty-four out of the 51 ADAPT trial clinical sites participated in an MRI Sub-study (13). All sites obtained institutional human research review board approval (institutional review board [IRB] or equivalent), and the University of Pittsburgh received IRB approval to coordinate the study. All sites were permitted to perform data collection, including therapies that were administered as standard of care prior to informed consent. This secondary analysis was supported by the University of Wisconsin Health Sciences IRB, Protocol #2015–0185 titled “MRI Markers of Outcome after Severe Pediatric Traumatic Brain Injury (TBI)”, approved on 25th August, 2015. All research procedures followed the ethical standards of the University of Wisconsin Health Sciences IRB and the Helsinki Declaration of 1975. The trial followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline.
Briefly, the inclusion criteria for the ADAPT MRI sub-study were age <18 years, diagnosis of severe TBI with GCS <9, ICP monitor placement, and MRI performed within 30 days of injury (13). For this secondary analysis, we excluded patients who had ICP monitoring for <24 hours, those who had missing ICP values >40% of monitoring time, and those who underwent decompressive craniectomy.
ICP data analysis
ICP monitoring duration was defined as the time from ICP monitor placement to removal. ICP monitoring was performed according to standard clinical practice at participating sites using parenchymal fiberoptic ICP monitors and/or external ventricular drains. Hourly ICP readings were recorded for each patient. In subjects with multiple monitoring devices, the highest hourly ICP value was included in subsequent analyses. ICP trends were reviewed manually for all patients with ICP >50 mmHg to evaluate for potential data input errors and inconsistencies. To reduce the chance of inaccurate data input, ICP values >50 mmHg and more than 1.96 standard deviations (SD) above a sliding mean within a ± 5-hour window were considered outliers and excluded from the subject’s ICP trend. Outlier ICP values were found in 21 subjects.
Intracranial hypertension burden was defined as area under the curve (AUC) of all ICP measurements >20 mmHg using a trapezoidal approach as previously described (14, 15). Given the variable duration of monitoring for each patient, the ICP AUC was then standardized by dividing each AUC by the total number of hours of monitoring in each subject. AUC and standardized AUC (StdAUC) had high correlation (see Supplemental Digital Content [SDC] Figure S1) and StdAUC was used for all subsequent analyses. We also calculated mean ICP across all ICP measurements, excluding outliers, for each subject. Finally, ICP burden was stratified into 4 categories based on StdAUC cut-offs modified from prior work (15), and selected a priori for ease of clinical interpretation. That is: Category 1, no ICP burden, StdAUC = 0 mmHg; Category 2, mild ICP burden, StdAUC >0 to 1 mmHg (i.e., average hourly elevation of no more than 1 mmHg above 20 mmHg for the duration of monitoring); Category 3, moderate ICP burden, StdAUC 1–2 mmHg (i.e., average hourly elevation of between 1 to 2 mmHg above 20 mmHg for the duration of monitoring); and Category 4, severe ICP burden, StdAUC >2 mmHg (i.e., average hourly elevation of more than 2 mmHg above 20 mmHg for the duration of monitoring).
Brain MRI Analysis
As previously described, each participating site submitted the first brain MRI scan acquired within 30 days of injury in all ADAPT subjects enrolled at their site who had an MRI performed as part of standard clinical care (13). The MR scanning protocols were determined by the standard practice in use at each site. The details of the MRI analysis has been previously described, including intra- and interrater reliability (13, 16), and we followed the National Institute of Neurological Disorders and Stroke Common Data Elements for Neuroimaging (17).
Briefly, MRI scans were read by one of two board certified neuroradiologists, each with >20 years of clinical experience. MRI was assessed for the following: 1) ICH volume; 2) contusion volume; 3) number of DAI microhemorrhage lesions; and 4) number of regions affected by ischemia. Brain regions assessed included bilateral frontal, parietal, temporal, and occipital lobes, bilateral thalamus/basal ganglia and internal capsule, and the cerebellum and brainstem (total 14 regions). DAI was additionally assessed in the corpus callosum. All scans were included in the analysis, and the best evaluation was made based on the experience and expertise of the neuroradiologist, given the clinical image quality, imaging protocol, and magnetic field strength. For measurement of contusion and ICH volume, the greatest extent was measured in 3 orthogonal planes and an idealized ellipsoid volume was calculated (length*width*depth*pi/6). For quantification of DAI, microhemorrhages in white matter were counted in each brain region up to a total of 10 lesions, and regions with >10 lesions received a score of 11. Ischemia was defined as evidence of injury on diffusion-weighted imaging (DWI) consistent with a deficit between substrate demand and delivery, scored as present/absent in 14 brain regions and quantified as the number of brain regions affected by that lesion type.
Statistical Analyses
Demographics and injury characteristics were summarized using descriptive statistics. Chi square or Fisher Exact Tests were used to test the independence between ICP burden and MRI injury amount. The association between ICP burden severity (with reference as StdAUC of 0 mmHg, which is no ICP burden) and presence of ischemia was then analyzed in univariate regression. P-values were considered significant if p <0.05. These data are summarized as the odds ratio with 95% confidence interval (OR [95%CI]).
RESULTS
At the ADAPT participating sites between 2014–2017, 355 patients with severe TBI underwent brain MRI within 30 days of admission. Of those, 338 had ICP monitoring for at least 24 hours with ICP values available for 60% of monitoring time. After excluding 118 patients who underwent craniectomy, the final cohort for this secondary analysis included 220 patients (see SDC Figure S2). The median (interquartile range, IQR) age was 6.7 (IQR 2.6, 13.4) years, 129/220 (59%) were male, median motor GCS was 3 (IQR 1, 4), and 179/220 (81%) had bilaterally reactive pupils (see SDC Table S1).
ICP Results
Out of 220 patients, 40 (18.2%) did not have any ICP values above 20 mmHg, with a corresponding StdAUC of 0 mmHg (Category 1) and a median ICP of 10 (IQR 8, 12) mmHg (Table 1). Overall, StdAUC had more dynamic range and increased variance compared to mean ICP for patients with any ICP value above 20 mmHg (see SDC Figure S3). Category 2 (StdAUC >0–1 mmHg) included 146 (66.4%) patients with median StdAUC of 0.13 (IQR 0.03, 0.34) mmHg, which corresponded to a median ICP of 13 (IQR 11, 15) mmHg and median 6 (IQR 2, 14) hours with elevated ICP. Category 3 (StdAUC >1–2 mmHg) included 20 (9.1%) patients with median StdAUC of 1.14 (IQR 1.09, 1.59) mmHg corresponding to median ICP 18 (IQR 16, 19) mmHg and median 34 (IQR 31, 49) hours above 20 mmHg. Finally, category 4 (StdAUC >2 mmHg) included 14 (6.4%) patients with median StdAUC 3.67 (IQR 2.7, 12.54) mmHg, which corresponded to median ICP 21 (IQR 20, 32) mmHg and median 56 (IQR 36, 78) hours above 20 (Table 1). Representative tracings of each ICP category are shown in Figure 1.
Table 1.
Description of Standardized AUC (StdAUC) ICP Bins.
| StdAUC Bin | |||||
|---|---|---|---|---|---|
| Characteristic | Entire cohort | 0 mmHg | >0–1 mmHg | >1–2 mmHg | >2 mmHg |
| Number of patients (number, %) | 220 (100) | 40 (18.2) | 146 (66.4) | 20 (9.1) | 14 (6.4) |
| Duration of monitoring, hours | 130 (81, 162) | 87 (56, 150) | 131 (85, 161) | 162 (150, 164) | 121 (69, 157) |
| Number of hours >20 mmHg | 5 (1, 17) | 0 | 6 (2, 14) | 34 (31, 49) | 56 (36, 78) |
| ICP, mmHg | 13 (11, 15) | 10 (8, 12) | 13 (11, 15) | 18 (16, 19) | 21 (20, 32) |
| StdAUC, mmHg | 0.10 (0.01, 0.51) | 0 | 0.13 (0.03, 0.34) | 1.14 (1.09, 1.59) | 3.67 (2.70, 12.54) |
Data represented as median (IQR). AUC area under the curve, ICP intracranial pressure
Figure 1.
Representative tracings from individual patients from the four ICP categories.
MRI Results
Of the 220 patients, 19 (8.6%) had MRI obtained prior to ICP monitor placement. Of the remaining 201 patients, median time to MRI after ICP monitor placement was 4.8 (IQR 1.4, 8.6) days (see SDC Table S1). Of the 220 patients, 156 (70.9%) had evidence of DAI with median 42 (IQR 18, 65) lesions. ICH was present in 31 (14.1%) patients who had median ICH volume of 4.55 (IQR 1.9, 23.5) cm3. Ischemia was present on the MRI of 70 (31.8%) patients with median of 6 (IQR 2, 10) regions involved. Finally, contusion was found in 161 (73.2%) patients with median volume of 15.1 (IQR 3.6, 34.1) cm3 (see SDC Table S2). Distribution of imaging findings is found in SDC Figure S4.
Association between MRI features and ICP burden
We next tested for association between MRI findings and ICP burden. Of the imaging features, contusion volume and number of regions with ischemia were associated with ICP burden. However, we failed to identify an association between the number of DAI lesions or volume of ICH, and ICP burden (Figure 2, Table 2). Because ischemia can precede or be a result of intracranial hypertension, we further tested for possible temporal association between ischemia and ICP burden based on the timing of MRI. Early MRI (before or <72 hours post-ICP monitor placement) was obtained in a total of 81 patients. Late MRI (>7 to 14 days after ICP monitor placement) was obtained in 51 patients. Ischemia detected on early MRI was associated with ICP burden (p=0.03) (see SDC Table S3). We failed to identify such an association in the late MRI group. Finally, we tested if increased burden of ICP was associated with presence of ischemia on MRI. Severe ICP burden, in comparison with no ICP burden, was associated with greater odds of presence of ischemia on MRI (OR 4.64 [95%CI 95% 1.30, 19.5], p=0.02) (Table 3). Regarding, moderate ICP burden, in comparison with no ICP burden, we failed to identify such an association. However, closer inspection of the point estimate (OR 1.24 [95%CI 0.40, 3.74]) shows that we are unable to exclude the possibility that moderate ICP burden was associated with up to a 3.74-fold greater odds of ischemia presence.
Figure 2.
Relative frequency of MRI injury burden across four ICP categories (None = StdAUC 0 mmHg; Mild = StdAUC >0–1 mmHg; Moderate = StdAUC >1–2 mmHg; Severe = StdAUC >2 mmHg). A) Contusion volume; B) Number of regions with ischemia; C) DAI lesion number; and D) ICH volume. Circles represent relative fraction of patients in each of the four ICP categories, such that a bigger circle reflects a higher relative fraction. Non-adjusted p-values are included in the Right upper corner.
Table 2.
Association between MRI features with ICP burden.
| StdAUC Bins | |||||||
|---|---|---|---|---|---|---|---|
| MRI Feature | 0 mmHg | >0–1 mmHg | >1–2 mmHg | >2 mmHg | p-value | adjusted p-value* | |
| Number of DAI lesions | None | 13 (32.5) | 39 (26.7) | 6 (30) | 6 (42.9) | 0.30 | 0.30 |
| 1–41 | 15 (37.5) | 56 (38.4) | 6 (30) | 1 (7.1) | |||
| >41 | 12 (30) | 51 (34.9) | 8 (40) | 7 (50) | |||
| ICH volume, cm3 | None | 37 (92.5) | 122 (83.6) | 20 (100) | 10 (71.4) | 0.23 | 0.30 |
| >0–4.55 | 2 (5) | 12 (8.2) | 0 (0) | 2 (14.3) | |||
| >4.55 | 1 (2.5) | 12 (8.2) | 0 (0) | 2 (14.3) | |||
| Contusion volume, cm3 | None | 15 (37.5) | 35 (24) | 5 (25) | 4 (28.6) | 0.02 | 0.06 |
| >0–15.1 | 19 (47.5) | 55 (37.7) | 4 (20) | 3 (21.4) | |||
| >15.1 | 6 (15) | 56 (38.4) | 11 (55) | 7 (50) | |||
| Number of regions with ischemia | None | 26 (65) | 108 (74) | 12 (60) | 4 (28.6) | 0.007 | 0.03 |
| 1–6 | 8 (20) | 22 (15.1) | 3 (15) | 3 (21.4) | |||
| 7–14 | 6 (15) | 16 (11) | 5 (25) | 7 (50) | |||
Data presented as number of patients in each category and % of patients in each StdAUC bin.
MRI magnetic resonance imaging, ICP intracranial pressure, StdAUC standardized area under the curve, DAI diffuse axonal injury, ICH intracranial hemorrhage
p-values adjusted for multiple comparisons using the Hochberg method
Table 3.
Association between ICP burden severity and presence of ischemia on MRI.
| ICP Burden Severity (StdAUC) | OR (95% CI) | p-value |
|---|---|---|
| No ICP (0, ref) | - | - |
| Mild ICP (>0–1 mmHg) | 0.65 (0.31, 1.41) | 0.26 |
| Moderate ICP (>1–2 mmHg) | 1.24 (0.40, 3.74) | 0.70 |
| Severe ICP (>2 mmHg) | 4.64 (1.30, 19.5) | 0.02 |
StdAUC standardized area under the curve, OR odds ratio, CI confidence interval, MRI magnetic resonance imaging, ICP intracranial pressure
DISCUSSION
The 2014–2017 ADAPT cohort dataset provides unique insight into the significance of ICP burden and patterns of brain injury on MRI in children with severe TBI. We found that over 80% of the ADAPT children who underwent ICP monitoring for >24-hours had elevated ICP at least once during monitoring time. On MRI, DAI and contusion were detected in over 70% of the patients, while ischemia and ICH were less frequent, affecting 32% and 14%, respectively. We found that contusion volume and number of regions with ischemia were associated with ICP burden, but we failed to identify such an association in those with DAI lesions and ICH.
Burden of intracranial hypertension in children with severe TBI has previously been described in two small, now historical cohorts from the 1970s to 1980s: one with a prevalence of elevated ICP of 20/40 patients (18), and the other 18/56 (19). More recent data about ICP-thresholds (and therefore prevalence of raised ICP), trajectory of exposure, and associated mortality is given by the results of ICP monitoring in a single center (2012–2016) in 262 pediatric patients, 87 of whom has suffered TBI (20). Our observations in the 2014–2017 ADAPT study international cohort confirm the risk of elevated ICP in children with severe TBI. In order to better characterize the burden of elevated ICP in our cohort, we employed a previously described method of trapezoidal interpolation of AUC exposure, which has better agreement with higher temporal resolution of data compared to staircase and spline interpolation (14). By employing this method, we aimed to capture ICP burden in two dimensions – time and magnitude – that may more effectively capture secondary brain insult when compared to methods that use maximal ICP value or time spent above certain threshold. It is important to note that intermittent, hourly resolution of ICP data may still underestimate the true burden of secondary brain insult. With ongoing advancement in big data collection and analysis, future research should aim to incorporate higher temporal resolution of data. Our threshold of ICP 20 mmHg is the suggested treatment threshold in the 2019 BTF pediatric TBI guidelines (9). Even so, this threshold may underestimate the degree of secondary injury that may occur in infants and younger children, whose optimal ICP thresholds may be lower than in older children and adults (20, 21). We also recognize that high ICP values may only partially explain secondary brain insults. The variability in ICP was studied in a cohort of pediatric TBI cases seen in two centers in the United Kingdom (up to 2003) and compared with a European dataset from four centers seen 2009–2013 (21). Differences may have reflected the adaptive capacity of the injured brain, which has improved correlation with survival and functional outcome compared to mean ICP in adult patients with TBI (11). Future studies should incorporate measures of ICP variability and their association with severity of brain injury and outcomes in children with TBI.
The focus of our study was to examine the association between ICP burden and MRI patterns of injury. We analyzed four main lesion types, and it is important to recognize the temporal evolution of these injuries in relation to intracranial hypertension. Hemorrhagic DAI, which represents injury from shearing forces occurring at the time of initial impact, can be visualized on MRI early after the injury, and is stable over weeks to months post-injury. ICH and contusion occur during the initial injury, evolve over the initial days post injury, and the associated edema and blood products represent space occupying lesions which may directly affect ICP. In contrast, ischemia may result from the initial injury due to systemic hypoperfusion or vascular injury leading to inadequate cerebral perfusion, and the associated cerebral edema may result in increased ICP. Alternatively, ischemia may be a manifestation of secondary brain injury resulting from inadequate cerebral perfusion due to increased ICP developing over the hours to days following the initial insult. We excluded patients who underwent neurosurgical intervention and therefore the ICH and contusions found in our cohort were likely deemed inoperable or too small to intervene upon with craniotomy. While we failed to identify an association between ICH and ICP burden, contusion volume was associated with ICP burden. This discrepant observation may be related to sample size, or other confounders, or the known risk of progressive edema with traumatic cerebral contusions, which can result in increased ICP – as described in studies of adult patients (22–24).
We failed to identify an association between DAI microhemorrhage lesion number and ICP burden. A small number of previous studies in adult patients with TBI has revealed mixed results when examining the association between ICP and DAI lesions (25–27). In one study of 19 patients, the number of hemorrhagic DAI lesions on MRI was positively correlated with maximum ICP (25). This investigation was limited by small sample size and lack of robust analysis of ICP burden. A larger study of 52 pediatric and adult patients (age range 9–61 years) showed that increased number of non-hemorrhagic DAI lesions in the central brainstem were associated with increased proportion of time with ICP >20 mmHg (27). Overall, DAI was frequently detected in our cohort, affecting 71% of patients, but we failed to identify an association between number of DAI lesions and burden of ICP.
We found that cerebral ischemia was associated with ICP burden. Ischemia is an important imaging finding that is associated with worse short-term and long-term outcomes in children with severe TBI (12, 13). When taking our entire cohort of 220 patients, there was a clear signal between presence of ischemia on MRI and greater odds of severe ICP burden (OR 4.64). While there was no clear signal between the presence of ischemia on MRI and moderate ICP, we could not exclude the possibility of there being up to 3.74-fold greater odds of moderate ICP. This difference may represent a dose-effect but, importantly, it raises the issue that both moderate and severe ICP burden may ensue in relation to MRI-identified ischemia. To better understand whether ischemia was a result or a consequence of intracranial hypertension, we performed a subgroup analysis on patients who had early MRI (obtained before ICP monitor or within the first 72 hours of monitoring) versus late MRI (obtained at 7 to 14 days after initiation of ICP monitoring). A surprisingly high number of patients (81 out of 220) had early MRI, while MRI in the 7 to 14-day window was obtained in 51 out of 220 patients. We found that ischemia on early (but not late) MRI was associated with ICP burden. It is still unclear, however, if ischemia detected on MRI in the first 72 hours was a result of initial impact versus secondary brain insult from high ICP, or other clinically important factors, such as hypoxia or hypotension, in these critically ill patients. We failed to identify an association between mild or moderate elevations in ICP above 20 mmHg and presence of ischemia, and not every patient with severe intracranial hypertension had evidence of ischemia. Overall, we cannot draw clear conclusions because the analysis was underpowered with less than 10 patients in the groups with severe ICP burden. Further studies are needed to understand etiologies and potentially modifiable risk factors for early ischemia in severe TBI. Nonetheless, the association between ischemia and ICP that we observed highlights the need for vigilant ICP monitoring and treatment of intracranial hypertension in the hope of preventing secondary brain injury in all children with severe TBI.
There are some limitations to our study. We had access to only the first clinical MRI and as such the temporal evolution of the lesions was unknown. MRI was obtained at the attendant clinician’s discretion which may introduce selection bias and may reflect the clinical team’s suspicion of brain injury. For MRI analysis, ICH volume was not normalized to total brain volume, which may be a limitation given the range of ages in the ADAPT trial cohort. Furthermore, we acknowledge that late MRIs may miss ischemic lesions, which could explain why we did not identify an association between ICP burden and ischemia on late MRIs. Next, we excluded patients a priori who had <24 hours of monitoring data (n=11) and those who had more than 40% of monitoring data missing (n=6), which may have introduced another bias. Given that our analysis of the ADAPT dataset was exploratory, evaluating an association between MRI findings and ICP burden (without clear temporal relationship), we did not test for clinical confounders. This approach will be an important next step in a future prospective study. Given the exploratory nature of our study, we did not account for hospital-level differences. Finally, severe intracranial hypertension was seen in a small number of patients and as such our results need to be replicated and externally validated in a larger cohort.
In conclusion, in this secondary analysis of the 2014–2017 ADAPT trial dataset we have found that elevated ICP was prevalent in these children with severe TBI. Furthermore, on detailed examination of the MRIs we found that increased number of brain regions showing ischemia and brain contusion volume were associated with increased burden of ICP. We failed to identify such an association between number of DAI lesions or ICH volume and increased burden of ICP. Further research is needed to determine the temporal association between elevated ICP and ischemia.
Supplementary Material
INFORMATION BOXES.
Research in context
Elevated intracranial pressure (ICP) can lead to secondary brain insult after traumatic brain injury (TBI).
Clinical and injury characteristics on admission do not adequately predict which patients develop increased ICP.
Association between acute patterns of brain injury on magnetic resonance imaging (MRI) and elevated ICP in children with severe TBI has not been previously investigated.
What this study means
In the 2014–2017 ADAPT study cohort, children with severe TBI were at a significant risk of elevated ICP.
Increased burden of ischemia and contusion volume were associated with increased burden of elevated ICP.
The temporal association between ischemia and elevated ICP requires further investigation.
Acknowledgements
We acknowledge all the children and families who generously participated in the ADAPT trial. We also acknowledge all ADAPT MRI Investigators (listed in SDC Table S4).
Financial support for the study:
Research reported in this publication was supported by the National Institute of Neurological Disorders and Stroke of the National Institutes of Health under Award Number U01 NS081041 (M.J.B.), K08NS078113 (P.A.F.), and R01NS092870 (P.A.F.), and the National Heart, Lung, Blood Institute under Award Number K24 HL168225 (T.D.B.), and was also supported in part by a core grant to the University of Wisconsin Waisman Center from the National Institute of Child Health and Human Development (U54HD090256). Additional support was provided by a University of Colorado Pediatric Critical Care Research Award (A.M.J.).
Copyright Form Disclosure:
Drs. Janas, Broman, Bennett, Rebsamen, Rosario, Bell, Alexander, and Ferrazzano received support for article research from the National Institutes of Health (NIH). Dr. Bennett’s institution received funding from National Heart, Lung, and Blood Institute, the Centers for Disease Control and Prevention, and the National Institute of General Medical Sciences. Drs. Bell’s and Alexander’s institutions received funding from the NIH. Dr. Ferrazzano’s institution received funding from the National Institute of Neurological Disorders and Stroke. Dr. Field has disclosed that he does not have any potential conflicts of interest.
Footnotes
Conflicts of interest: None
*Group Information: The ADAPT MRI Investigators appear in Supplemental Digital Content.
Financial disclosures: None
REFERENCES
- 1.Yan J, Wang C, Sun B: Global, regional, and national burdens of traumatic brain injury from 1990 to 2021. Front Public Health 2025; 13:1556147. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Zhong H, Feng Y, Shen J, et al. : Global Burden of Traumatic Brain Injury in 204 Countries and Territories From 1990 to 2021. Am. J. Prev. Med. 2025; 68:754–763 [DOI] [PubMed] [Google Scholar]
- 3.Alberico AM, Ward JD, Choi SC, et al. : Outcome after severe head injury. Relationship to mass lesions, diffuse injury, and ICP course in pediatric and adult patients. J. Neurosurg. 1987; 67:648–656 [DOI] [PubMed] [Google Scholar]
- 4.Sharples PM, Stuart AG, Matthews DS, et al. : Cerebral blood flow and metabolism in children with severe head injury. Part 1: Relation to age, Glasgow coma score, outcome, intracranial pressure, and time after injury. J Neurol Neurosurg Psychiatry 1995; 58:145–152 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.White JR, Farukhi Z, Bull C, et al. : Predictors of outcome in severely head-injured children. Crit. Care Med. 2001; 29:534–540 [DOI] [PubMed] [Google Scholar]
- 6.Jagannathan J, Okonkwo DO, Yeoh HK, et al. : Long-term outcomes and prognostic factors in pediatric patients with severe traumatic brain injury and elevated intracranial pressure. J Neurosurg Pediatr 2008; 2:240–249 [DOI] [PubMed] [Google Scholar]
- 7.Alkhoury F, Kyriakides TC: Intracranial Pressure Monitoring in Children With Severe Traumatic Brain Injury: National Trauma Data Bank-Based Review of Outcomes. JAMA Surg 2014; 149:544–548 [DOI] [PubMed] [Google Scholar]
- 8.Bennett TD, Riva-Cambrin J, Keenan HT, et al. : Variation in intracranial pressure monitoring and outcomes in pediatric traumatic brain injury. Arch. Pediatr. Adolesc. Med. 2012; 166:641–647 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Kochanek PM, Tasker RC, Carney N, et al. : Guidelines for the Management of Pediatric Severe Traumatic Brain Injury, Third Edition: Update of the Brain Trauma Foundation Guidelines. Pediatr Crit Care Med 2019; 20:S1–S82 [DOI] [PubMed] [Google Scholar]
- 10.Bailey BM, Liesemer K, Statler KD, et al. : Monitoring and prediction of intracranial hypertension in pediatric traumatic brain injury: clinical factors and initial head computed tomography. J Trauma Acute Care Surg 2012; 72:263–270 [DOI] [PubMed] [Google Scholar]
- 11.Jha RM, Elmer J, Zusman BE, et al. : Intracranial Pressure Trajectories: A Novel Approach to Informing Severe Traumatic Brain Injury Phenotypes. Crit. Care Med. 2018; 46:1792–1802 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Janas AM, Miller KR, Stence NV, et al. : Utility of Early Magnetic Resonance Imaging to Enhance Outcome Prediction in Critically Ill Children with Severe Traumatic Brain Injury. Neurocrit Care 2024; 41:80–90 [DOI] [PubMed] [Google Scholar]
- 13.Ferrazzano PA, Rebsamen S, Field AS, et al. : MRI and Clinical Variables for Prediction of Outcomes After Pediatric Severe Traumatic Brain Injury. JAMA Netw Open 2024; 7:e2425765. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Hemphill JC 3rd, Barton CW, Morabito D, et al. : Influence of data resolution and interpolation method on assessment of secondary brain insults in neurocritical care. Physiol. Meas. 2005; 26:373–386 [DOI] [PubMed] [Google Scholar]
- 15.Vik A, Nag T, Fredriksli OA, et al. : Relationship of “dose” of intracranial hypertension to outcome in severe traumatic brain injury. J. Neurosurg. 2008; 109:678–684 [DOI] [PubMed] [Google Scholar]
- 16.Ferguson NM, Rebsamen S, Field AS, et al. : Magnetic Resonance Imaging Findings in Infants with Severe Traumatic Brain Injury and Associations with Abusive Head Trauma. Children (Basel) 2022; 9 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Haacke EM, Duhaime AC, Gean AD, et al. : Common data elements in radiologic imaging of traumatic brain injury. J. Magn. Reson. Imaging 2010; 32:516–543 [DOI] [PubMed] [Google Scholar]
- 18.Bruce DA, Raphaely RC, Goldberg AI, et al. : Pathophysiology, treatment and outcome following severe head injury in children. Childs Brain 1979; 5:174–191 [DOI] [PubMed] [Google Scholar]
- 19.Barzilay Z, Augarten A, Sagy M, et al. : Variables affecting outcome from severe brain injury in children. Intensive Care Med. 1988; 14:417–421 [DOI] [PubMed] [Google Scholar]
- 20.Woods KS, Horvat CM, Kantawala S, et al. : Intracranial and Cerebral Perfusion Pressure Thresholds Associated With Inhospital Mortality Across Pediatric Neurocritical Care. Pediatr Crit Care Med 2021; 22:135–146 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Güiza F, Depreitere B, Piper I, et al. : Visualizing the pressure and time burden of intracranial hypertension in adult and paediatric traumatic brain injury. Intensive Care Med. 2015; 41:1067–1076 [DOI] [PubMed] [Google Scholar]
- 22.Chen DY, Wu PF, Zhu XY, et al. : Risk factors and predictive model of cerebral edema after road traffic accidents-related traumatic brain injury. Chin. J. Traumatol. 2024; 27:153–162 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Kawamata T, Katayama Y: Cerebral contusion: a role model for lesion progression. Prog Brain Res 2007; 161:235–241 [DOI] [PubMed] [Google Scholar]
- 24.Kawamata T, Katayama Y, Mori T, et al. : Mechanisms of the mass effect of cerebral contusion: ICP monitoring and diffusion MRI study. Acta Neurochir. Suppl. 2002; 81:281–283 [DOI] [PubMed] [Google Scholar]
- 25.Yanagawa Y, Sakamoto T, Takasu A, et al. : Relationship between maximum intracranial pressure and traumatic lesions detected by T2*-weighted imaging in diffuse axonal injury. J Trauma 2009; 66:162–165 [DOI] [PubMed] [Google Scholar]
- 26.Tomei G, Sganzerla E, Spagnoli D, et al. : Posttraumatic diffuse cerebral lesions. Relationship between clinical course, CT findings and ICP. J. Neurosurg. Sci. 1991; 35:61–75 [PubMed] [Google Scholar]
- 27.Abu Hamdeh S, Marklund N, Lewén A, et al. : Intracranial pressure elevations in diffuse axonal injury: association with nonhemorrhagic MR lesions in central mesencephalic structures. J. Neurosurg. 2018; 131:604–611 [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.