https://orcid.org/0000-0002-7918-058X
Figures
Abstract
Intradural spinal cord compression impairs perfusion pressure and is putatively rate-limiting for recovery after traumatic spinal cord injury (tSCI). After cervical tSCI, even minimally improved tissue preservation may help promote neurological recovery. To assess the nature and extent of spinal cord swelling and compression post-acute cervical tSCI, we evaluated several baseline MRI parameters including BASIC score, intramedullary lesion (IML) length, maximal canal compromise (MCC), maximal spinal cord compression (MSCC), extent of cord compression (ECC), maximal swollen anteroposterior diameter adjacent to injury site (Dmax), and maximal cord swelling (MCS) in 169 consecutive patients across 2 centers. In patients with either primarily intradural or combined (MSCC ≤5% or >5%, respectively) cord compression, we examined the predictive value of clinical and imaging admission parameters on American Spinal Injury Association Impairment Scale (AIS) severity and conversion up to 1-year follow-up. 37 (21.9%) patients presented with primarily intradural while 132 (78.1%) had combined cord compression. MSCC, MCS, and Dmax values differed significantly between the two groups (p < 0.0001, < 0.01 and < 0.001, respectively). MSCC was associated with age, MCC and MCS at baseline, while MCS was associated with age, MSCC and Dmax, on multivariable analysis. Logistic regression analysis of areas under receiver operating characteristic curve (AUROC) confirmed ECC (AUC 0.678) and MCS (AUC 0.922) as good and excellent predictors, respectively of AIS-conversion at 1-year for intradural compression participants. Additionally, MCS was significantly more accurate in predicting AIS-conversion in intradural group and the probability of AIS-conversion significantly decreased with each 1% increase in MCS (p = 0.003; OR 0.949), for both compression subtypes. In conclusion, baseline measures of cord swelling predict AIS-conversion likelihood up to 1-year. The deleterious effects of intradural cord compression, either isolated or presenting with extradural compression, may benefit from supplemental decompression strategies in addition to current standard-of-care.
Citation: Arora H, Darabi H, Toop N, Minnema A, Al-Sharshahi Z, Martin GM, et al. (2025) Spinal cord swelling and intradural compression predict neurological recovery after acute cervical traumatic spinal cord injury. PLoS One 20(8): e0325827. https://doi.org/10.1371/journal.pone.0325827
Editor: Alvan-Emeka K. Ukachukwu, Duke University Medical Center: Duke University Hospital, UNITED STATES OF AMERICA
Received: December 3, 2024; Accepted: May 20, 2025; Published: August 7, 2025
Copyright: © 2025 Arora et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: Data cannot be shared publicly because public availability would compromise privacy of the patients that sustained traumatic spinal cord injuries through various mechanisms. This restriction is in compliance with the IRB regulations set out at both participating institutions. Data will be made available by submitting reasonable requests to the OSU and/or UK Offices of Research Integrity for researchers who meet the criteria for access to confidential data. The contact information is as follows. OSU: IRBinfo@osu.edu UK: RS_ORI@uky.edu.
Funding: The author(s) received no specific funding for this work.
Competing interests: No authors have competing interests.
Introduction
The chronic functional impairment and disability resulting from tSCI place a considerable burden on individuals and their caregivers [1]. Along with varying degrees of motor, sensory, and autonomic dysfunction, multisystem early and delayed complications further complicate neurologic outcomes and increase mortality [2].
tSCI leads to varying degrees of local intraparenchymal spinal cord inflammation, progressive hemorrhagic necrosis, and swelling that potentially result in elevated intraspinal pressure (ISP). These processes in turn can compromise spinal cord perfusion pressure (SCPP), particularly when the cord swelling extends against the non-expansile dura mater, further reducing SCPP and potentially contributing to secondary injury [3]. This potential intradural compression is distinct from the extradural, or more precisely the combined spinal cord compression, that arises from fractures and/or dislocations of the spinal bony elements and the disco-ligamentous complex (see Fig 1).
The intradural subtype (shown in center) is characterized by primary spinal cord compression within the dura mater, underscoring that the contact on the spinal cord originates only from components contained in the intradural space. This internal pressure gives rise to a distinct localized egg-shaped appearance at the injury site, where the spinal cord diameter at the level of injury (di) is increased. With combined compression, there is both intradural compression and a discernible extrinsic pressure source (shown on right). This external compression suggests the involvement of factors lying outside the dura mater, such as bony fracture fragments, that significantly contribute to the spinal cord compression. The resultant combined compression typically results in an hourglass appearance where the diameter of the spinal cord at the injury level (di) is reduced as compared to the diameters above (da) and below (db) the injury. The extent of cord compression (ECC) measures the longitudinal span of spinal cord where both anterior and posterior cerebrospinal fluid signal are not evident. Dmax refers to the maximum anteroposterior spinal cord diameter adjacent to the site of spinal cord injury which, along with the da measurement, is used to calculate maximal cord swelling (MCS). Of note, the dashed lines refer to the levels at which the measurements are performed.
Individuals with isolated intradural compression after tSCI offer a unique opportunity to analyze the evolution of spinal cord swelling, as the spinal cords in these circumstances are free from extradural constraints that separately limit cord expansion [4]. Careful distinction between these patterns of spinal cord compression is crucial, not only because it may allow for identification of baseline parameters with prognostic significance, but also because it may lead to the adoption of additional decompression strategies currently not included in the standard-of-care treatment regimen.
Despite decompression surgery being widely recognized as a mainstay intervention that can limit secondary injury, the exclusion of intradural or intraparenchymal decompression receives insufficient attention, in marked contrast to commonly-employed decompressive craniectomy protocols used in the management of traumatic brain injuries [5–7]. The use of measures aimed directly at the dura and spinal cord, such as expansion duroplasty and midline myelotomy have a longstanding history [3,8–12], and are currently being evaluated in a prospective trial [13]. Response to such additional measures may ultimately depend on the cord compression subtype, for instance with respect to the severity of intradural compression (either alone or when combined with extradural compression).
In this study, we combined independent tSCI cohorts from 2 academic institutions to identify the baseline demographic, clinical, and radiological characteristics of individuals who present with either primarily intradural versus combined (extradural and intradural) spinal cord compression, as evaluated on baseline standard-of-care MRI scans after tSCI. Using multivariable logistic regression analysis, we also (1) evaluate the association between these cord compression subtypes and other radiological measures of injury severity and (2) in a subset of subjects, we identify the combination of variables that are most predictive of neurological status and recovery.
Materials and methods
Study design
A retrospective cohort study was conducted across two academic level 1 trauma centers, the Ohio State University (OSU) and the University of Kentucky (UK). The study encompassed consecutive patients suspected of experiencing cervical spine trauma, spanning from January 2008 to May 2018 (OSU) and from May 2013 to May 2023 (UK). For OSU, as in a previous study [14], we used data from the institutional chapter of the Central Ohio Trauma System Registry (https://www.cotshealth.org/trauma) as well as from the Information Warehouse database while for UK we used the institutional trauma registry (https://ukhealthcare.uky.edu/doctors-providers/trauma). Long-term follow-up clinical evaluations were only available at OSU through participants’ electronic medical records and were primarily undertaken through scheduled visits with the Dodd SCI Rehabilitation Program, an awarded Center of the US-Spinal Cord Injury Model System. Ethical approval was obtained from the respective institutional review boards (protocol numbers 2015H0110 and 88464, respectively) and all guidelines were followed throughout the course of the study. Informed consent was waived given the retrospective nature of the study. Authors accessed electronic medical records for data collection and all individual participants’ data were de-identified prior to the start of data analysis. For the purposes of this research study, the data was accessed between August 15, 2020 and February 24, 2021 at OSU and between July 3, 2023 to March 11, 2024 at UK.
Study participants
The study adhered to predefined inclusion and exclusion criteria. Consecutive patients presenting with acute cervical tSCI within the study period were considered for inclusion. The inclusion criteria were: 1) ≥18 years of age; 2) sustained either American Spinal Injury Association impairment scale (AIS) complete A or incomplete (B, C, D) SCIs; 3) presented with various clinical patterns of tSCI including central cord syndrome (CCS); 4) were either admitted directly or initially stabilized at other facilities and subsequently transferred for further acute management; 5) underwent AIS assessments and imaging within 72 h after injury. Exclusion criteria included: 1) concomitant traumatic brain or brachial plexus injuries making AIS assessments not reliable; 2) lack of clear evidence of spinal cord trauma with lesions related to nontraumatic causes (infection, tumor, vascular and metabolic); 3) AIS E tSCI; 4) prior surgical intervention on the cervical spine at an outside facility; 5) penetrating SCI (e.g., gunshot injuries); 6) institutional post-injury MRI absent or not performed within 72 hours.
Data extraction
Various baseline parameters including the age on admission, sex, body mass index, and mechanism of injury were recorded. Injury mechanisms were classified into five categories, including motor vehicle collision (MVC), fall from height (high-energy), sports injury, mechanical fall (low-energy), and other high-energy trauma (OHT) that could not be classified otherwise into other groups. Injury and imaging times, injury severity score (ISS), Charlson Comorbidity Index (CCI), AIS on admission and on follow-up (when available), spinal fracture, presence of CCS, treatment details, length of stay (LOS), and in-hospital adverse events including death. Injuries were categorized according to the AIS scoring system with classifications ranging from AIS A – D indicating complete injury (A), incomplete injury with sensory sparing (B) and incomplete injury with motor sparing in less than and more than half the muscle groups below the level of injury (C and D, respectively) [15]. The neurological level of injury was defined as the most caudal spinal level at which all sensory and motor functions were normal.
Radiological variables
Sagittal and axial T1 and T2-weighted cervical MRI views were used to measure and derive the various baseline parameters. These parameters included intramedullary lesion (IML) length, spinal canal diameters (Di, Da and Db), maximal canal compromise (MCC), maximum spinal cord compression (MSCC), spinal cord anteroposterior diameters (di, da, db, and Dmax), extent of cord compression (ECC), and maximal cord swelling (MCS) (see Fig 1 and Table 2) [3,16–19]. For each institutional series, two independent assessors that were blinded to the neurological outcomes performed the various measurements and the average values were used for the final analysis. Using axial T2-weighted views at the site of injury, the Brain and Spine Injury Centre (BASIC) [20] scores were also determined by two independent observers with any disagreements resolved by the senior author (who was also blinded to the neurological outcomes).
Outcome variables
AIS grades of A and B (compared with AIS C and D) were defined as “severe neurological deficit.” Individuals with positive AIS conversion of at least one grade were grouped separately compared with those with no change or worsening in their AIS. Severe neurological deficit on admission as well as severe neurological deficit and AIS conversion on follow-up were analyzed as primary outcome variables.
Statistical analysis
Continuous variables were reported as median with interquartile range (IQR) while frequency distributions were employed for categorical variables. Interclass correlation coefficients (ICC) were calculated for the radiological parameters for 66/169 patients as measures of interrater reliability [21]. Categorical variables were reported as the number of individuals within each cohort. The Shapiro-Wilk test was performed for all continuous variables to assess for normality of distribution. Comparisons between cohorts were conducted using the Mann-Whitney U test for continuous variables and Fisher’s Exact test for categorical variables, as the data did not follow a normal distribution. Spearman correlation was performed to assess the relationship between MSC, MSCC, ECC, Dmax and MCS. Univariate regression analysis was used to assess for the association between MSCC or MCS and the various baseline demographic and radiological measurements. A stepwise multivariable regression analysis was performed (including variables with p < 0.15 on univariate) to identify the relationship between the various baseline parameters and either MSCC or MCS. p < 0.05 was considered statistically significant. Further, to assess for multicollinearity, variance inflation factors (VIFs) were calculated for all included variables in the multivariable regression. Finally, Area under the Receiver Operating Characteristic (ROC) curve (AUC) analysis was employed to assess the association between the various radiological parameters and neurological severity rates, as well as AIS conversion rates, at 6-month and 1-year follow-up. All data were analyzed using R Studio Version 4.2.0 [22].
Results
Population characteristics
A total of 262 participants were identified across the centers. Following screening and application of eligibility criteria, 169 individuals (103 from OSU, 66 from UK) and were included in the final analysis (see Fig 2). Overall, 37 (21.9%) were classified on admission MRI as exhibiting primarily intradural spinal cord compression and 132 (78.1%) with combined compression (Table 1). The Shapiro-Wilk test confirmed a non-normal distribution for the demographic variables and radiological parameters of the study participants. Baseline demographic factors including female/male sex ratio (7/30 vs 33/99, p = 0.517) and BMI (26.0 [IQR = 23.6–31.6] vs 28.0 [IQR = 24.1–31.9], p = 0.357) were not significantly different between groups. Injury mechanism was significantly different between groups (p = 0.013) with MVCs representing the most common form (n = 59), followed by mechanical falls (n = 50) and falls from height (n = 31). Admission injury severity scores (p = 0.173), AIS grades (p = 0.665), neurological level of injury (p = 0.444) and, the mean time-of-injury to MRI interval (9.1 [IQR = 6.1–19.1] vs 11.0 [IQR = 7.1–20.9] hrs, p = 0.230) were not significantly different between groups. Surgical decompression was performed in 147/169 (87.0%) patients and was achieved within 24 hours in 51/169 (30.2%) of cases, no significant differences were noted in surgery rates between groups.
The flow diagram outlines the enrollment of study participants based on the screening process and the defined eligibility criteria.
MRI parameters dichotomized by cord compression subtype
Table 2 displays statistical analyses comparing MCC, MSCC, IML length, BASIC score, ECC, and MCS between intradural and combined cord compression subtypes. The ICC revealed highest interrater reliability of measurements for ECC followed by average canal-based measurements (Di, Da, Db), average cord-based measurements (di, da, db), Dmax, and IML at 0.651, 0.645, 0.590, 0.569, and 0.565, respectively. These values are representative of moderate range interrater reliability [19]. Individuals with intradural cord compression had similar MCC as compared to those with combined cord compression injuries (32.3 [IQR = 17.4–42.0] vs 33.9 [IQR = 17.7–53.3], respectively; p = 0.083). However, lower Di, Da and Db variables were observed within the combined cord compression group (p = < 0.0001, 0.002 and < 0.0001, respectively). Likewise, the combined group had a significantly higher extent of MSCC (−6.1 [IQR = −19.1- −1.6] vs 22.9 [IQR = 14.0–31.4], p < 0.0001). The negative value indicates a spinal cord diameter that is increased from baseline at the level of injury. Rostrocaudal IML length was similar between cord compression groups (26.8 [IQR = 16.9–49.7] mm vs 23.7 [IQR = 12.7–35.9] mm, p = 0.161). BASIC score distributions, as assessed on axial T2-weighted views, were also not different between groups (p = 0.197). The most prevalent BASIC score was 2 in both groups, observed in 13/37 (35.1%) and 47/132 (35.6%) patients, respectively.
In contrast, significant differences were noted between groups in MCS (13.3 [IQR = 4.8–23.0] vs 6.9 [IQR = −1.4–17.0], p = 0.008) and associated Dmax values (8.1 [IQR = 7.3–9.4] vs 7.2 [IQR = 6.5–8.1], p = 0.0008), even though the ECC was not significantly different (11.5 [IQR = 7.7–21.0] mm vs 11.9 [[IQR = 5.2–23.3] mm, p = 0.805).
Inter-relationships between baseline variables
Spearman correlations were performed to explore potential relationships between the various baseline radiological parameters (Fig 3). The analysis shows a strong correlation between MCS and Dmax (r = 0.56, p ≤ 0.0001). A weak positive correlation is noted between MSCC and ECC (r = 0.17, p = 0.032). Negative correlations are shown between MSCC and MCS (r = −0.23, p = 0.003) and MSCC and Dmax (r = −0.25, p = 0.001).
Correlation heat-plot matrix depicting the Spearman correlation coefficients (r) for the various radiological variables assessed in the study. Blue intensity corresponds to the degree of positive correlation (r > 0) while red intensity corresponds to the degree of negative correlation (r < 0). No correlation is represented by white (r = 0). Significant associations between two variables (p < 0.05) are represented with asterisks (*) in the respective cells.
Regression analyses were conducted to analyze the association between various baseline variables and either MSCC or MCS (Table 3). Significant associations are shown on multivariable analysis between MSCC and age (p < 0.0001), MCC (p = 0.005), MCS (p = 0.0002), and ECC (p = 0.032). With respect to MCS, significant associations on multivariable analysis are shown for MSCC (p = 0.0001) and Dmax (p = 0.048). Further, the VIFs for all variables across both regression models were all < 2, indicating no evidence of multicollinearity.
To determine whether there is a relationship between cord swelling and time of injury to MRI, individuals were stratified into subgroups based on whether the MRI study was performed ≤ 12 hrs, between 12–24 hrs, or between 24–72 hrs from the time of injury. While no significant differences on ANOVA were observed in either Dmax or MCS, ECC was significantly different across the three defined time intervals (p = 0.025).
Baseline MRI variables and outcome analysis
Follow-up AIS evaluations were available at 6 months and at 1-year post-injury intervals for a total of 103 individuals (see Table 4 and Supplementary Fig 1). No differences were noted between tSCI compression subtypes in mortality or follow-up rates (p = 0.678). All baseline MRI variables were analyzed for their ability to predict neurological severity (Table 5) and AIS conversion (Table 6) within both the intradural and combined compression tSCI groups using AUC analysis. Each variable’s predictive strength was analyzed and compared for both the intradural and combined compression tSCI groups at both follow-up time points. MCC and MSCC are poor predictors (AUC range 0.294 to 0.653) while ECC is a fair predictor (AUC range 0.659 to 0.824) of neurological severity for both compression groups. In contrast, MCS is a good predictor (AUC range 0.700 to 0.850) while IML length and BASIC are excellent predictors (AUC ≥ 0.889) at both 6 months and 1-year. There were no significant differences with respect to the neurological severity predictive abilities of these various measures at the two follow-up intervals.
Similar findings are noted for these MRI variables as predictors of AIS conversion with the intrinsic cord-signal and swelling-based values again performing more robustly as compared to MCC and MSCC. While MCC and MSCC again performed poorly overall (AUC range 0.233 to 0.641), the predictive ability of MSCC is significantly better for individuals with combined compression at both the 6-month (p = 0.036) and 1-year (p = 0.011) follow-up time points. Once again, while ECC is a fair predictor (AUC range 0.641 to 0.713), IML length and BASIC are excellent predictors particularly at 1-year follow-up (AUC range 0.840 to 0.956). In contrast to all the other robust predictors of AIS conversion, MCS performed significantly better at predicting AIS conversion for individuals with primarily intradural compression at both the 6-month (0.889 vs 0.692, p = 0.048) and 1-year follow-up time points (0.922 vs 0.700, p = 0.018).
Finally, we performed binary ordinal logistic regression analysis to evaluate the probability of AIS conversion based on ECC and MCS. For both compression subtypes, we observe that the chance of AIS conversion significantly decreases with every 1 mm increase in ECC (p = 0.004; OR 0.938) and with each 1% increase in MCS (p = 0.003; OR 0.949).
Discussion
Secondary biochemical injury processes after tSCI begin immediately and progress over hours and days [23]. This phase is characterized by a wide range of pathophysiological mechanisms including progressive hemorrhagic necrosis, inflammation, reduced blood flow, toxic effects from excessive glutamate release, ionic imbalance, and oligodendrocyte death [24–26]. Spinal cord swelling, an integral component of secondary injury [27], leads to elevated ISP and reduced SCPP, potentially further compounding injury severity [28–30].
Given shared pathologic CNS injury mechanisms involving edema and elevated intercompartment pressure [12,31], standard-of-care management for both traumatic brain injury and tSCI include decompressive surgical techniques including decompressive hemicraniectomy, discectomy, corpectomy, and laminectomy. However, while cranial procedures typically include opening and enlargement of the dural compartment (expansive duroplasty), this component has not historically been included in spinal procedures. Incorporation of more aggressive options like myelotomy and expansive duroplasty into the management stream remains a matter of debate and requires further evaluation [10,32]. This concept is clearly worth exploring, given our decades-long use and proven benefit of decompressive hemicraniectomies in TBI care, which always involve dural opening (Fig 1) [12,13,33,34].
In this paper, we delineate two distinct compression patterns of tSCI to explore the related differences in canal and spinal cord-based measurements. In line with previous work [9], we distinguished tSCIs based on their corresponding MSCC ratios; a ≤ 5% cutoff was used to define intradural compression while all other injuries were assigned to the combined group (see Fig 1). To our knowledge, our study presents the first analysis of the relationship between these compression subtypes (dural-only versus combined dural and extrinsic/ discoligamentous) and neurological outcomes. This dichotomization also allows for a novel evaluation of distinct cord-based measures of swelling applying standard MRI techniques. Specifically, Dmax and MCS have been recognized as relevant measures of the degree of anteroposterior cord swelling [18]. Both measures are significantly greater and are more robustly predictive of neurologic recovery specifically in the intradural group. Of note, these two measures showed a significant inverse correlation with MSCC (Fig 3). This is in line with the observation that the intradural group has higher Dmax values than the combined group given the absence of extrinsic compression that allows the spinal cord to swell further in the anteroposterior dimension. As such, we conclude that the absence of extrinsic compression is relatively more permissive of the injured spinal cord’s tendency to swell post-tSCI and thus allows for detection of the relevance of the associated secondary injury processes in predicting neurologic recovery. In our cohort, combined compression apparently did not allow for a similar degree of cord swelling despite a similar profile of underlying tSCI mechanisms and severity on admission.
ECC is a measure of cord swelling in the longitudinal axis [9], and may therefore similarly track post-SCI intraparenchymal changes as compared to Dmax and MCS. However, while we failed to identify a significant correlation between ECC and MCS (p = 0.065), a significant negative correlation was identified between ECC and MSCC (p = 0.032) in our cohort. Further, ECC is not significantly different between the compression subtypes in our study. This observation actually aligns with prior studies that demonstrated that the extent of cord swelling, for which ECC serves as a surrogate, may correlate with injury severity following tSCI [35,36]. Since injury severity on admission (e.g., baseline ISS and AIS) is similar between our two study groups, it appears that this longitudinal-based measure does not as faithfully track cord swelling as compared to the injury site axial-based Dmax and MCS measures.
To our knowledge, our study also suggests for the first time that younger patients are more prone to develop primarily intradural compressive cervical tSCI (see Table 1). This may in part be attributable to the tendency of older patients to suffer from fractures and/or dislocations of the spine [37,38], potentially manifesting as central cord syndrome in cases with pre-existing cervical spondylosis with spinal canal narrowing and spinal cord compression (see Table 3). As the age demographic of tSCI is changing over time [39,40], we believe that this important compression subtype may remain underrecognized.
IML length and BASIC scoring have emerged as effective surrogate markers for assessing the severity of tSCI, based on sagittal and axial imaging characteristics, respectively [14,16,20]. In our cohort, these measures demonstrated significant utility for predicting neurologic severity and AIS conversion for both compression subtypes up to one-year follow-up. The one-year mark is pivotal for individuals with tSCI, as it generally marks the transition to a chronic recovery state following which significant recovery is less likely [41]. In line with the critical importance of mid-sagittal “tissue bridges” to enable neurological recovery, these findings underscore the relevance of early neuroprotection strategies though sufficient spinal cord decompression [14,42]. ‘Time (to decompression) is spine’ relates to an overarching paradigm mirroring the span of time over which the swollen spinal cord is strangulated by its local environment. Here, we propose and validate pragmatic, real-world MRI measures to assess different phenotypes of spinal cord swelling being predictive of neurologic recovery after tSCI. The represented MRI-markers characterizing post traumatic edema are also candidates to further improve stratification for interventional trials and prognostication models.
Consistent with the most recent guidelines [43,44], many trauma centers have adopted an early decompression surgical strategy in their management of patients with acute tSCI. While the benefits of this approach are increasingly recognized, it may lead to underappreciation of the extent of spinal cord swelling that is expected to reach a peak stage at 3–5 days after injury [9,45–47]. In our study, a significantly greater proportion of subjects with combined compression underwent surgery ≥24 hours post-injury, which may be attributed to both extrinsic (e.g., logistical or transportation-related limitations) and intrinsic factors (e.g., need for stabilization). To assess for a potential relationship between the progression of spinal cord swelling and time of injury, we categorized the study participants into 3 groups based on their injury-MRI time intervals and identified significant time-dependent differences in ECC measurements. Further controlled studies will be needed to confirm this finding and more accurately define the timescale of cord swelling following acute tSCI.
Taken together, our study supports the notion that intradural compression remains an underappreciated source of injury to the spinal cord that may independently affect outcomes. The reliability of our study findings is supported by the relatively large number of patients included over two independent trauma centers. Potential selection bias was minimized through the establishment of uniform criteria and standardized data collection procedures. Another potential limitation is the moderate inter-rater reliability observed for spinal cord radiological measurements. While these measurements support the overall trends and findings, the moderate agreement may introduce some degree of variability. This may in turn impact the precision or reliability of the results, underscoring the need for future studies with larger sample sizes to assess these radiological variables in acute cervical tSCI and their clinical associations. Among the 169 patients, 37 (21.9%) had primarily intradural compression at baseline, which is consistent with the work of Saadoun and colleagues (25.8%) [9] but higher than the estimate provided by Aarabi and colleagues [10] when evaluating the proportion of patients that underwent “inadequate decompression” with noted absence of cerebrospinal fluid on post-operative imaging (9.6%). An ongoing clinical trial evaluating the potential benefits of expansive duroplasty may shed further insight into the clinical relevance of intradural spinal cord compression [13].
Conclusion
This study reveals the presence of a distinct group of individuals with tSCI who experience spinal cord compression primarily related to swelling within the intradural compartment without associated significant compromise of the spinal canal. This distinct cohort highlights the clinical relevance of baseline Dmax and MCS measures as additional predictors of the likelihood of neurological recovery after tSCI. Since spinal cord swelling may be variably noted in both tSCI cord compression modes, our findings provide a rationale for careful evaluation of these measures during the first few days after tSCI. More importantly, our findings suggest that individuals with tSCI may benefit from additional decompression of the intradural compartment by mitigating the secondary injury associated with intradural cord compression and ischemia. Further confirmation of these findings over time will provide a rationale for a tailored surgical strategy as defined by the patient-specific pathophysiological tSCI features.
Supporting information
Supplementary Fig 1. Relationship of cord compression subtype with neurologic recovery over the first year after tSCI.
Neurological recovery profiles after tSCI of individual subjects (represented as blue lines) are displayed at 6 and 12 months after injury in comparison to baseline AIS scores for both intradural (left) and combined (right) compression subtypes.
https://doi.org/10.1371/journal.pone.0325827.s001
(TIF)
References
- 1. Keihanian F, Kouchakinejad-Eramsadati L, Yousefzadeh-Chabok S, Homaie Rad E. Burden in caregivers of spinal cord injury patients: a systematic review and meta-analysis. Acta Neurol Belg. 2022;122(3):587–96. pmid:35157242
- 2. Krassioukov A. Autonomic function following cervical spinal cord injury. Respir Physiol Neurobiol. 2009;169(2):157–64. pmid:19682607
- 3. Saadoun S, Chen S, Papadopoulos MC. Intraspinal pressure and spinal cord perfusion pressure predict neurological outcome after traumatic spinal cord injury. J Neurol Neurosurg Psychiatry. 2017;88(5):452–3. pmid:27864426
- 4. Phang I, Zoumprouli A, Papadopoulos MC, Saadoun S. Microdialysis to Optimize Cord Perfusion and Drug Delivery in Spinal Cord Injury. Ann Neurol. 2016;80(4):522–31. pmid:27463064
- 5. Cooper DJ, Rosenfeld JV, Murray L, Arabi YM, Davies AR, D’Urso P, et al. Decompressive craniectomy in diffuse traumatic brain injury. N Engl J Med. 2011;364(16):1493–502. pmid:21434843
- 6. Cooper DJ, Rosenfeld JV, Murray L, Arabi YM, Davies AR, Ponsford J, et al. Patient Outcomes at Twelve Months after Early Decompressive Craniectomy for Diffuse Traumatic Brain Injury in the Randomized DECRA Clinical Trial. J Neurotrauma. 2020;37(5):810–6. pmid:32027212
- 7. Vieira E, Guimarães TC, Faquini IV, Silva JL, Saboia T, Andrade RVCL, et al. Randomized controlled study comparing 2 surgical techniques for decompressive craniectomy: with watertight duraplasty and without watertight duraplasty. J Neurosurg. 2018;129(4):1017–23. pmid:29148904
- 8. Allen A. Surgery of experimental lesion of spinal cord equivalent to crush injury of fracture dislocation of spinal column a preliminary report. JAMA. 1911;57(11):878–80.
- 9. Saadoun S, Werndle MC, Lopez de Heredia L, Papadopoulos MC. The dura causes spinal cord compression after spinal cord injury. Br J Neurosurg. 2016;30(5):582–4. pmid:27080553
- 10. Aarabi B, Chixiang C, Simard JM, Chryssikos T, Stokum JA, Sansur CA, et al. Proposal of a Management Algorithm to Predict the Need for Expansion Duraplasty in American Spinal Injury Association Impairment Scale Grades A-C Traumatic Cervical Spinal Cord Injury Patients. J Neurotrauma. 2022;39(23–24):1716–26. pmid:35876459
- 11. Kim KD, Lee KS, Coric D, Chang JJ, Harrop JS, Theodore N, et al. A study of probable benefit of a bioresorbable polymer scaffold for safety and neurological recovery in patients with complete thoracic spinal cord injury: 6-month results from the INSPIRE study. J Neurosurg Spine. 2021;34(5):808–17. pmid:33545674
- 12. Papadopoulos MC, Saadoun S. Letter to the Editor. The INSPIRE studies for spinal cord injury. J Neurosurg Spine. 2021;35(5):684–5. pmid:34388712
- 13. Saadoun S, Grassner L, Belci M, Cook J, Knight R, Davies L. Duroplasty for injured cervical spinal cord with uncontrolled swelling: protocol of the DISCUS randomized controlled trial. Trials. 2023;24(1):497.
- 14. Farhadi HF, Kukreja S, Minnema A, Vatti L, Gopinath M, Prevedello L, et al. Impact of Admission Imaging Findings on Neurological Outcomes in Acute Cervical Traumatic Spinal Cord Injury. J Neurotrauma. 2018;35(12):1398–406. pmid:29361876
- 15. Roberts TT, Leonard GR, Cepela DJ. Classifications in brief: American spinal injury association (ASIA) impairment scale. LWW. 2017.
- 16. Aarabi B, Sansur CA, Ibrahimi DM, Simard JM, Hersh DS, Le E, et al. Intramedullary Lesion Length on Postoperative Magnetic Resonance Imaging is a Strong Predictor of ASIA Impairment Scale Grade Conversion Following Decompressive Surgery in Cervical Spinal Cord Injury. Neurosurgery. 2017;80(4):610–20. pmid:28362913
- 17. Fehlings MG, Furlan JC, Massicotte EM, Arnold P, Aarabi B, Harrop J, et al. Interobserver and intraobserver reliability of maximum canal compromise and spinal cord compression for evaluation of acute traumatic cervical spinal cord injury. Spine (Phila Pa 1976). 2006;31(15):1719–25. pmid:16816769
- 18. Skeers P, Battistuzzo CR, Clark JM, Bernard S, Freeman BJC, Batchelor PE. Acute Thoracolumbar Spinal Cord Injury: Relationship of Cord Compression to Neurological Outcome. J Bone Joint Surg Am. 2018;100(4):305–15. pmid:29462034
- 19. Miyanji F, Furlan JC, Aarabi B, Arnold PM, Fehlings MG. Acute cervical traumatic spinal cord injury: MR imaging findings correlated with neurologic outcome--prospective study with 100 consecutive patients. Radiology. 2007;243(3):820–7. pmid:17431129
- 20. Talbott JF, Whetstone WD, Readdy WJ, Ferguson AR, Bresnahan JC, Saigal R, et al. The Brain and Spinal Injury Center score: a novel, simple, and reproducible method for assessing the severity of acute cervical spinal cord injury with axial T2-weighted MRI findings. J Neurosurg Spine. 2015;23(4):495–504. pmid:26161519
- 21. Bobak CA, Barr PJ, O’Malley AJ. Estimation of an inter-rater intra-class correlation coefficient that overcomes common assumption violations in the assessment of health measurement scales. BMC Med Res Methodol. 2018;18(1):93. pmid:30208858
- 22. Ihaka R, Gentleman R. R: A Language for Data Analysis and Graphics. Journal of Computational and Graphical Statistics. 1996;5(3):299–314.
- 23. Ahuja CS, Nori S, Tetreault L, Wilson J, Kwon B, Harrop J, et al. Traumatic Spinal Cord Injury-Repair and Regeneration. Neurosurgery. 2017;80(3S):S9–22. pmid:28350947
- 24. Garcia VB, Abbinanti MD, Harris-Warrick RM, Schulz DJ. Effects of Chronic Spinal Cord Injury on Relationships among Ion Channel and Receptor mRNAs in Mouse Lumbar Spinal Cord. Neuroscience. 2018;393:42–60. pmid:30282002
- 25. Hachem LD, Fehlings MG. Pathophysiology of Spinal Cord Injury. Neurosurg Clin N Am. 2021;32(3):305–13. pmid:34053718
- 26. Xu G-Y, Hughes MG, Ye Z, Hulsebosch CE, McAdoo DJ. Concentrations of glutamate released following spinal cord injury kill oligodendrocytes in the spinal cord. Exp Neurol. 2004;187(2):329–36. pmid:15144859
- 27. Werndle MC, Saadoun S, Phang I, Czosnyka M, Varsos GV, Czosnyka ZH, et al. Monitoring of spinal cord perfusion pressure in acute spinal cord injury: initial findings of the injured spinal cord pressure evaluation study*. Crit Care Med. 2014;42(3):646–55. pmid:24231762
- 28. Cao X j, Feng S q, Fu C f, Gao K, Guo JS, Guo XD. Repair, protection and regeneration of spinal cord injury. Neural Regen Res. 2015;10(12):1953–75.
- 29. Hogg FRA, Gallagher MJ, Kearney S, Zoumprouli A, Papadopoulos MC, Saadoun S. Acute Spinal Cord Injury: Monitoring Lumbar Cerebrospinal Fluid Provides Limited Information about the Injury Site. J Neurotrauma. 2020;37(9):1156–64. pmid:32024422
- 30. Zhu YK, Lu FT, Zhang GD, Liu ZP. A review of strategies associated with surgical decompression in traumatic spinal cord injury. Journal of Neurological Surgery Part A: Central European Neurosurgery. 2023.
- 31. Seblani M, Decherchi P, Brezun J-M. Edema after CNS Trauma: A Focus on Spinal Cord Injury. Int J Mol Sci. 2023;24(8):7159. pmid:37108324
- 32. Phang I, Werndle MC, Saadoun S, Varsos G, Czosnyka M, Zoumprouli A, et al. Expansion duroplasty improves intraspinal pressure, spinal cord perfusion pressure, and vascular pressure reactivity index in patients with traumatic spinal cord injury: injured spinal cord pressure evaluation study. J Neurotrauma. 2015;32(12):865–74. pmid:25705999
- 33. Hutchinson PJ, Kolias AG, Timofeev IS, Corteen EA, Czosnyka M, Timothy J, et al. Trial of Decompressive Craniectomy for Traumatic Intracranial Hypertension. N Engl J Med. 2016;375(12):1119–30. pmid:27602507
- 34. Rossini Z, Nicolosi F, Kolias AG, Hutchinson PJ, De Sanctis P, Servadei F. The History of Decompressive Craniectomy in Traumatic Brain Injury. Front Neurol. 2019;10:458. pmid:31133965
- 35. Hayashi K, Yone K, Ito H, Yanase M, Sakou T. MRI findings in patients with a cervical spinal cord injury who do not show radiographic evidence of a fracture or dislocation. Paraplegia. 1995;33(4):212–5. pmid:7609978
- 36. Meves R, Avanzi O. Correlation among canal compromise, neurologic deficit, and injury severity in thoracolumbar burst fractures. Spine (Phila Pa 1976). 2006;31(18):2137–41. pmid:16915102
- 37. Dong Y, Peng R, Kang H, Song K, Guo Q, Zhao H, et al. Global incidence, prevalence, and disability of vertebral fractures: a systematic analysis of the global burden of disease study 2019. Spine J. 2022;22(5):857–68. pmid:34906740
- 38. Freitas SS, Barrett-Connor E, Ensrud KE, Fink HA, Bauer DC, Cawthon PM, et al. Rate and circumstances of clinical vertebral fractures in older men. Osteoporos Int. 2008;19(5):615–23. pmid:18038109
- 39. Aarabi B, Albrecht JS, Simard JM, Chryssikos T, Schwartzbauer G, Sansur CA, et al. Trends in Demographics and Markers of Injury Severity in Traumatic Cervical Spinal Cord Injury. J Neurotrauma. 2021;38(6):756–64. pmid:33353454
- 40. Jain NB, Ayers GD, Peterson EN, Harris MB, Morse L, O’Connor KC. Traumatic spinal cord injury in the United States, 1993-2012. JAMA. 2015;313(22):2236–43.
- 41. Kirshblum S, Snider B, Eren F, Guest J. Characterizing Natural Recovery after Traumatic Spinal Cord Injury. J Neurotrauma. 2021;38(9):1267–84. pmid:33339474
- 42. Emmenegger TM, Pfyffer D, Curt A, Schading-Sassenhausen S, Hupp M, Ashburner J, et al. Longitudinal motor system changes from acute to chronic spinal cord injury. Eur J Neurol. 2024;31(4):e16196. pmid:38258488
- 43. Fehlings MG, Tetreault LA, Hachem L, Evaniew N, Ganau M, McKenna SL, et al. An Update of a Clinical Practice Guideline for the Management of Patients With Acute Spinal Cord Injury: Recommendations on the Role and Timing of Decompressive Surgery. Global Spine J. 2024;14(3_suppl):174S-186S. pmid:38526922
- 44. Sattari SA, Antar A, Theodore JN, Hersh AM, Al-Mistarehi A-H, Davidar AD, et al. Early versus late surgical decompression for patients with acute traumatic central cord syndrome: a systematic review and meta-analysis. Spine J. 2024;24(3):435–45. pmid:37890727
- 45. Hale C, Yonan J, Batarseh R, Chaar R, Jonak CR, Ge S, et al. Implantable Osmotic Transport Device Can Reduce Edema After Severe Contusion Spinal Cord Injury. Front Bioeng Biotechnol. 2020;8:806. pmid:32754586
- 46. Weirich SD, Cotler HB, Narayana PA, Hazle JD, Jackson EF, Coupe KJ, et al. Histopathologic correlation of magnetic resonance imaging signal patterns in a spinal cord injury model. Spine (Phila Pa 1976). 1990;15(7):630–8. pmid:2218708
- 47. Zusman BE, Kochanek PM, Jha RM. Cerebral Edema in Traumatic Brain Injury: a Historical Framework for Current Therapy. Curr Treat Options Neurol. 2020;22(3):9. pmid:34177248