Age and Sex Affect the Chronic Trajectory of Plasma Neurofilament Light and Glial Fibrillary Acidic Protein Levels in a Murine Thoracic Contusion Spinal Cord Injury Model

Abstract

Background:

Mouse models of traumatic spinal cord injury (SCI) are used to understand pathophysiology and test potential interventions. Experimental injury parameters, deficits on functional tasks, and histology are used to assess severity and recovery. Blood biomarkers may be a promising additional metric to assess severity and detect efficacy of interventions, but they have not been examined previously in mouse spinal cord injury (SCI).

Objectives:

To examine blood biomarkers in mouse SCI.

Methods:

We measured plasma levels of neurofilament light (NfL) and glial fibrillary acidic protein (GFAP) longitudinally following a thoracic contusion SCI in adolescent (3 month old) and aged (18 month old) male and female C57Bl6/J mice. Biomarkers were also assessed in comparably aged uninjured animals.

Results:

Three-month-old animals exhibited elevated plasma NfL and GFAP 1 month after injury. NfL levels decreased from 1 to 2 months post injury but remained elevated from baseline, while GFAP levels remained high. Adolescent males exhibited higher NfL levels than females post injury. In aged animals, NfL was comparably elevated at 1 and 2 months post injury. In aged females, GFAP was elevated at 1 and 2 months after injury, while levels in males did not increase from baseline until 2 months after injury. Values from uninjured animals show plasma NfL and GFAP increase with age in absence of injury.

Conclusion:

In a mouse SCI model, plasma NfL and GFAP are elevated chronically after injury. Sex and age at injury may affect biomarker trajectories, which may indicate underlying pathology relevant to treatment and recovery. Establishing the trajectory of NfL and GFAP after experimental injury may help to standardize injury paradigms, assess recovery, and detect efficacy of interventions.

Introduction

Rodent models of traumatic spinal cord injury (SCI) are a mainstay of fundamental and preclinical SCI research.¹ Experimental researchers endeavor to use metrics of severity and recovery that are comparable surrogates to human assessments.² Metrics of injury severity often include specifics of the injury model used (e.g., force of impact for contusion) and level of functional impairments. Recovery is frequently assessed by functional improvement or histological assays. Although these metrics have obvious parallels to the human injury course, they are not perfect corollaries, and there is heterogeneity in preclinical outcome assessments.³ As the clinical field develops blood-based biomarkers for SCI, there may be utility in exploring their use in rodents.

Identifying fluid biomarkers for assessing injury severity and prognosis in human SCI is an active area of research, with applications in clinical care and clinical trials (reviewed by Kwon et al.⁴). Biomarkers’ utility is reliant on their context of use, including established normative values and trajectories.⁵ Two blood-based biomarkers, neurofilament light (NfL) and glial fibrillary acidic protein (GFAP), are potential diagnostic and prognostic SCI biomarkers.^6–8 Elevated blood levels of GFAP indicate astrocyte structural damage, astrocyte reactivity, and inflammation; blood-brain barrier permeability may affect levels (reviewed by Abdelhak et al.⁹). Plasma NfL indicates axonal damage primarily in large myelinated neurons (reviewed by Gaetani et al.¹⁰).

Several studies have evaluated blood levels of neurofilaments and GFAP acutely following rat SCI (reviewed by Shool et al.¹¹). To our knowledge, no studies have examined biomarkers in mouse SCI, including at chronic time points. Establishing plasma biomarker trajectories in chronic mouse SCI could serve as a valuable metric for assessing outcomes and intervention efficacy in a common preclinical SCI model. Furthermore, as the incidence of SCI at advanced ages increases¹² and research reveals how sex modifies pathology (reviewed by Stewart et al.¹³), it is necessary to understand how age and sex affect injury course. This analysis reports trajectories for plasma NfL and GFAP in mice following thoracic contusion SCI. Age at time of injury and sex impact biomarker trajectories, and both biomarkers are elevated chronically. Plasma biomarkers may prove a relatively easy quantitative readout of recovery or intervention efficacy for preclinical scientists. Pending confirmation of their predictive utility and signal for efficacy in models and humans, they may provide a harmonized metric between experimental and clinical research. Preclinical research may benefit from adoption of blood biomarkers as a longitudinal assessment.

Methods

Methods are reported in accordance with ARRIVE guidelines.¹⁴

Animals

All experiments and protocols were approved by institutional authorities in compliance with Canadian Council on Animal Care and institutional guidelines. Data reported here are a part of a larger study using purchased C57Bl6/J animals (for adolescent experiments, Jackson Laboratories, strain 000664, RRID:IMSRJAX:000664) or in-house bred C57Bl6/J mice with a Cre-recombinase reporter (for aged experiments, Jackson Laboratories, strain 021162, RRID:IMSR_JAX:02116). Injured animals underwent injuries and plasma collection (outlined below). Age-matched uninjured animals underwent plasma collection (outlined below). Comparisons are between animals of the same genotype. Up to 5 animals were housed per open-top wire lid cage on a reverse light-dark cycle with ad libitum access to food and water. Domes, crinkle paper, suspended rings, and tubes provided enrichment. Animals were housed in the same room to limit environmental confounds. Approximately equal numbers of males and females were used; sexes are noted on the figures.

Injuries

Animals were injured at 3 months of age or 18 months of age. Mice were initially anesthetized with 3% isoflurane in oxygen. After reaching surgical plane, animals were maintained on ~1.2% isoflurane on heat support for the remainder of the procedure. Animals received subcutaneous injections of lactated Ringer's (20 mL/kg), buprenorphine (0.05 mg/kg), and bupivacaine (locally at the incision site, 8 mg/kg). The back was shaved and disinfected with successive chlorhexidine and 70% ethanol washes. An incision revealed the lower thoracic spinal column to permit a thoracic level 9 full laminectomy. The Infinite Horizon Impactor (Precision Systems) delivered a 70 kDyne force impact with no dwell, commensurate with previous moderate severity injuries in our laboratory.¹⁵ Incisions were closed with resorbable suture (6-0 DemeCAPRONE, DemeTECH, cat. no. MO176011F13M). Animals awoke in a 32 °C humidified recovery cage. Animals were given moistened food and hydration support on the floor of their home cage. Supportive fluids and buprenorphine were administered every 8 hours for 3 days. Bladders were expressed manually 2 to 3 times per day until voluntary micturition returned.

Plasma collection

Blood was collected from the saphenous vein using EDTA-coated tubes (Sarstedt, cat. no. NC9976871) the week prior to injury for baseline (BL) and at regular intervals post injury. Blood from uninjured animals was collected at comparable timepoints and ages to injured counterparts. Researchers collecting blood could not be blinded to animal age or injury condition. A lack of blinding was not anticipated to affect sample collection.

Biomarker detection

Plasma levels of murine GFAP were quantified by a homebrew immunoassay using MesoScale Discovery (MSD) platform as previously published¹⁶ using rabbit anti-GFAP antibodies (Abcam, cat. no. AB242692 RRID:AB_3099647). Plasma was diluted 4 times with phosphate buffered saline containing 1% bovine serum albumin (Sigma Aldrich, cat. no. A7906). The assay's lower limit of detection (LLOD) threshold of 3.05 pg/mL yields a sample LLOD of 12.21 pg/mL after correcting for dilution factor; samples below the LLOD are reported as 12.21 pg/mL.¹⁷ One 3-month-injured animal's 1 month post injury (mpi) sample and a different 3-month-injured animal's 2 mpi sample had insufficient volume for the GFAP analysis; the reduced n is noted in the figure legends. Plasma NfL levels were measured by the MSD Neurofilament-light assay kit (Mesoscale, cat. no. K1517XR-2) following manufacturer's instruction. Plasma was diluted 6 times with MSD Diluent 12. Researchers conducting biomarker analysis were blinded to age and injury conditions. All reported comparisons result from samples analyzed within the same run.

Sample size, inclusion, and exclusion

Inclusion criteria included no neurological or physiological abnormalities and consistent injury force and displacement values (eFigure 1). One outlier was removed for a high force.

As we could find no published values for plasma biomarkers in murine SCI, previous values obtained in our laboratory for mice with and without traumatic brain injury were used as expected means. Previous studies found 7 C57Bl6 mice had a mean NfL of 120.7 pg/mL ± 92.1 versus a mean of 741.3 pg/mL ± 707.3 post injury (Cohen's d = 1.23), and a mean of 31.5 pg/mL ± 15.561 GFAP in uninjured animals versus a mean of 3013.9 pg/mL ± 1897 post injury (Cohen's d = 2.22). With alpha = 0.05 and beta = 0.8, sample sizes of 4 per group would permit us to detect meaningful effect sizes.

Fifteen 3-month-old animals were injured, and one animal was euthanized for high injury force; subsequently this analysis represents data from 14 animals. For the 18-month SCI group, 14 of 25 animals did not reach experimental endpoint (n = 9 humanely euthanized for sequelae of injury and/or age reaching humane endpoints, n = 3 died during experimental procedures, n = 2 unknown). The force value for one aged animal's impact was an outlier (ROUT analysis with Q=1%) and was removed from analysis. Subsequently, this analysis includes 10 aged animals. For the uninjured groups, six 3-month-old animals and six 18-month-old animals were used; one animal was euthanized for loss of condition at 19.5 months and is omitted from the 20-month analysis. For practical reasons, all experiments were not conducted in parallel. Statistical comparisons are only between groups that occurred concurrently. Subsequently, no randomization was used to determine which animals were grouped into injured versus uninjured. As this analysis examines repeated measures to establish trajectories after injury, there are no control groups.

Statistical analysis

Statistics were conducted using Prism 10 (GraphPad, RRID:SCR_002798) and R packages afex¹⁸ and emmeans.¹⁹ Data are represented as individual points representing individual animals, with summary means where appropriate. Simple linear regressions were used to assess the relationship between injury parameters and biomarker levels. Data were transformed by y=log(y) due to magnitude differences breaking normality, permitting means to be used for summary statistics. Repeated measures (RM) 1-way or 2-way analysis of variance (ANOVA) tests with Tukey's corrections were used to compare means over time and between sexes. For instances of missing data (see above), values were omitted for ANOVA. Mixed effects models were used to predict means as a confirmatory analysis (eFigure 2). Pearson's correlations were used to identify correlations between NfL and GFAP levels. Significance is designated as P < .05.

Results

To establish the trajectory of plasma biomarker levels following a moderate thoracic contusion SCI in adolescent (3 month old) and aged (18 month old) mice, blood was collected at 1 and 2 mpi (Figure 1). Injuries were consistent, with acceptable variation within each group based on force and displacement (eFigure 1). Biomarker levels underwent log transformation to compare means (Figure 1). Estimated means were generated using mixed effects models to confirm findings (eFigure 2).

Figure 1.

Trajectories of plasma neurofilament light (NfL) and glial fibrillary acidic protein (GFAP) with age and after thoracic contusion injury at 3 or 18 months of age. Three-month-old (a, b) or 18-month-old animals (c, d) had baseline (BL) blood sampling prior to a thoracic level 9 (T9) spinal cord injury (SCI), with subsequent sampling at 1 and 2 months post injury (mpi). In animals injured at 3 months, plasma levels of NfL (a) and GFAP (b) are both elevated at 1 month post injury as compared to BL. In animals injured at 18 months of age, NfL (c) and GFAP (d) are elevated at 1 mpi and remain elevated at 2 mpi. Plasma levels of NfL were not statistically significantly different in uninjured animals from 3 to 5 months of age (e), though there was high variability. GFAP levels (f) were undetectable in most animals. Uninjured animals sampled from 18 to 20 months of age showed an increase over that time in plasma NfL (g) and a trend of an increase in GFAP (h). Data are presented as a log transformation to accommodate changes in magnitude. Dashed lines denote lower limit of detection (LLOD) for GFAP measurements. Significance values represent comparisons between means. *P < .05. **P < .001. ***P < .0005. ****P < .0001. 3-month injured: n = 14 animals for NfL measurements; n = 12 animals for GFAP measurements due to low sample volumes. 18-month injured: n = 11 animals. 3- to 5-month uninjured: n = 6. 18- to 20-month uninjured: n = 5-6.

Chronically elevated biomarkers after SCI at 3 months of age

One month after a thoracic contusion SCI at 3 months of age, plasma levels of NfL were significantly elevated from baseline (BL) in both males and females (geometric mean [GM] for males at BL = 314.897 pg/mL vs. 5529.016 pg/mL at 1 mpi.) See Figure 1a for 2-way RM ANOVA table (df = 7, P < .0001, with Tukey's corrections; for females GM at BL = 246.090 pg/mL vs. 4250.011 pg/mL at 1 mpi, df = 5, Tukey's P = .0001). Levels decreased from 1 mpi to 2 mpi in both males and females but were still elevated from BL (male 2 mpi GM = 1753.794 pg/mL vs. 1 mpi, df = 7, P < .0001; female 2 mpi GM = 1215.788 pg/mL, df = 5, P < .0001; BL vs. 2 mpi df = 7, Tukey's P = .0003 for males, df = 5, P = .0003 for females). Mean levels of NfL were higher in males than females at 1 mpi (df = 9.782, P = .0107) and 2 mpi (df = 5.722, P = .0057).

Plasma GFAP was below detection threshold at baseline (Figure 1b). There were no differences between males and females in GFAP levels, but overall GFAP levels increased post injury (2-way RM ANOVA with values omitted to accommodate 2 samples with insufficient volume; Figure 1b). In both males and females, mean GFAP values increased from BL to 1 mpi (BL below detection threshold; GM = 12.21 pg/mL for males and females; 1 mpi males GM = 411.756 pg/mL, df = 10, P < .0001; females GM = 380.977 pg/mL, df = 10, P < .0001) and remained elevated at 2 mpi as compared to BL (males 2 mpi GM = 559.658 pg/mL, P < .0001; females 2 mpi GM = 979.798 pg/mL, P < .0001). There was no difference between 1 mpi and 2 mpi values (male 1 mpi vs. 2 mpi, P = .7851; female 1 mpi vs. 2 mpi, P = .05741). GFAP and NfL levels are not correlated at 1 or 2 months post injury (eFigure 3A, B; nonsignificant by Pearson correlation). These data show that in animals injured at 3 months, plasma GFAP and NfL are elevated at 1 mpi. NfL begins to decrease at 2 mpi, whereas GFAP remains elevated. Females exhibit lower NfL levels than males post injury.

Chronically elevated biomarkers after SCI at 18 months of age

In aged animals, NfL values were statistically significantly elevated from BL levels in males and females at 1 mpi (Figure 1c; 2-way RM ANOVA, male BL GM = 612.729 pg/mL vs.1 mpi GM = 6180.145 pg/mL, Tukey's df = 3, P = .0005; female BL GM = 802.305 pg/mL vs. 1 mpi 4814.536 pg/mL, df = 5, P = .0008). There were no differences between males and females (minimum P = .3392). Unlike in the adolescent animals, NfL values at 2 mpi in aged animals did not decrease from 1 mpi (male 2 mpi GM = 4112.952 pg/mL: 2 mpi vs. BL df = 3, P = .0061, 2 mpi vs. 1 mpi df = 3, P = .1807; female 2 mpi GM = 4886.326 pg/mL: 2 mpi vs. BL df = 5, P = .0010, 2 mpi vs. 1 mpi df = 5, P = .9852). In aged animals, plasma GFAP was elevated post injury, with no differences between males and females (Figure 1d; 2-way RM ANOVA table). Plasma GFAP values did not statistically significantly increase in males from BL to 1 mpi (BL GM = 418.034 pg/mL; 1 mpi GM = 2842.742 pg/mL, Tukey's df = 3, P = .0908) but did increase from BL to 2 mpi (GM = 3144.247 pg/mL, df = 3, P = .0435). Values did not decrease from 1 mpi to 2 mpi in males (df = 3, P = .7807). In females, GFAP levels were elevated from BL levels at 1 mpi and 2 mpi (BL GM = 136.673 pg/mL, 1 mpi GM = 3289.359 pg/mL, df = 5, P = .0001; 2 mpi GM = 3825.928 pg/mL vs. BL, df = 5, P = .0001) but did not statistically significantly change from 1 to 2 mpi (df = 5, P = .6067). Levels of GFAP and NfL do not correlate at 1 mpi in the aged injured animals (eFigure 3c), but there is a positive correlation between the 2 biomarkers at 2 mpi [eFigure 3d; Pearson's correlation r(8) = 0.6697, P = .0341]. Overall, in aged animals post injury, GFAP and NfL are both chronically elevated at 2 months post injury, and there are fewer sex effects than in adolescent injured animals.

As experiments were conducted asynchronously for the 2 age groups, comparisons between them should be interpreted carefully. However, as their plasma samples were analyzed within the same run, it can be noted that the aged animals have higher baseline levels of NfL (GM = 720.296 pg/mL) than 3-month-old animals (GM = 283.321 pg/mL). Comparably, in the adolescent group, BL GFAP was below the LLOD in all animals (Figure 1b) but was detected in all 18-month-old animals (GM of 213.744 pg/mL). Interestingly, aged male BL GFAP was not statistically different than GFAP at 1 mpi (Figure 1d).

Biomarkers in uninjured animals

To establish the trajectories of biomarkers in absence of injury, naïve uninjured animals were sampled at 1-month intervals commensurate to the injured experiment timeline. Even though there was no statistically significant difference at any time point in 3-month-old uninjured animals, baseline NfL values were highly variable, driven by the males; their values decreased at subsequent timepoints (Figure 1e; nonsignificant by 1-way RM ANOVA). Sample size precludes testing for sex differences. Plasma GFAP was below the level of detection for uninjured animals at 3 and 4 months of age (Figure 1f). Three of the 6 animals had detectable GFAP at 5 months of age (GM = 24.15 pg/mL), and there were no effects of time on GFAP levels (Figure 1f; nonsignificant by RM ANOVA).

In the aged uninjured animals, one animal failed to reach experimental endpoint; removal of this datapoint and analysis by RM ANOVA found a significant effect of time overall on NfL levels from 18 to 20 months of age (Figure 1g). Mean NfL values increased from 18 months (GM = 648.2 pg/mL) to 20 months (GM = 1682 pg/mL, df = 4, P = .0047). Plasma GFAP could be detected in five of the six 18-month-old uninjured animals (GM = 109.59 pg/mL; Figure 1h). There was no effect of time on GFAP levels in the aged animals (RM ANOVA with omitted values [Figure 1h]), 19-month GM = 297.79 pg/mL, 20-month GM = 284.8 pg/mL). There was no correlation between plasma GFAP levels and NfL in the uninjured aged animals at any time point (eFigure 2e-h).

Plasma GFAP decreases chronically after SCI at 3 months of age

To establish the trajectories of biomarkers in an extended chronic mouse SCI model, 4 animals injured at 3 months of age were sampled for an additional 7 months post injury (Figure 2). The NfL data show considerable variability (Figure 2a) with no statistically significant effect over time (linear regression Y = 226.6*X - 746.8, F = 2.962, P = .1837). GFAP levels decrease from 3 to 9 mpi (Figure 2b; linear regression Y = -33.24*X + 292.3, F = 28.98, P = .0126), and several animals had values below detection threshold.

Figure 2.

Plasma levels of neurofilament light (NfL) and glial fibrillary acidic protein (GFAP) 3 to 9 months post injury at 3 months of age (n = 4). Animals injured at 3 months of age had regular blood draws until 9 months post injury (mpi). Plasma levels of NfL (a) did not statistically change in this time, though there is a trend of an increase. Plasma levels of GFAP (b) decreased over 3 to 9 mpi. Dashed lines denote lower limit of detection (LLOD) for GFAP measurements. Points denote values for individual animals. Males are represented as hexagons and females as squares.

Discussion

As blood biomarkers are developed for the diagnosis and prognosis of SCI, parallel efforts in rodent SCI models may provide a useful translational tool. Efforts to harmonize protocols and use metrics that best recapitulate clinical metrics are hypothesized to improve efficacy of preclinical research (e.g., in diabetes research²⁰) and can be catalyzed through open data sharing initiatives.²¹ A standardized quantitative measurement of injury severity could provide an additional metric to ensure consistent and repeatable experimental injuries. As markers of recovery and outcome, biomarkers could provide efficacy signal for interventions, and they may help address outcome heterogeneity found in experimental SCI.³

This study describes significantly elevated plasma NfL and GFAP levels months after a thoracic contusion in both adolescent and aged mice as compared to preinjury levels. Chronic plasma biomarkers have not been studied previously in rodents. Acute studies in rats describe a peak and resolution of GFAP and NfL post injury. Peak GFAP values occur within days post injury and return to baseline by 1 week post injury.²² Phosphorylated neurofilament heavy values peaked 3 days post injury and decreased at 7 days post injury, though they are still elevated from baseline levels.²³ An acute peak and resolution of biomarker levels does not preclude elevated biomarkers chronically. It may be that an initial peak and recovery of biomarker levels after SCI is followed by chronic release of structural or inflammatory markers. This interpretation is supported in part by findings from a recent study tracking NfL and GFAP in the blood of people with SCI for several months after injury.⁸ Though the mechanisms of biomarker release into the blood and the underlying pathology remain unknown, NfL peaked ~1 week post injury and was still elevated at ~5 mpi.⁸ This trajectory is similar to the data described here in mice. In contrast to the GFAP results described here, GFAP peaks within hours of injury, and levels resembled healthy controls at subsequent follow-ups.⁸ Importantly, injury course may vary between species, including the timing of inflammatory responses and glial reactivity (differences between rats and mice reviewed by Perez et al.²⁴). Further studies are needed to ascertain differences and similarities in trajectories of plasma biomarkers in models and humans to improve our understanding of the generalizability between species.

Elevated GFAP persisted for months, though the mean decreased from 3 to 9 mpi, suggesting peak GFAP levels may occur at 2 to 3 mpi in mice. Chronically elevated levels may be indicative of ongoing degeneration or inflammation. Further studies need to confirm these findings, ascertain the underlying pathology, and correlate those findings with functional outcomes. With validation, regular assessment of blood biomarkers may be a reliable and easy way to assess efficacy of interventions in chronic studies.

As more injuries are sustained at advanced age,¹² it is necessary to understand age's effect on injury course. In aged mice, NfL and GFAP levels remained elevated at 2 mpi. Although GFAP was not detectable in the blood of adolescent mice prior to injury, there was detectable GFAP at baseline in aged mice. The increase in GFAP from BL to 1 mpi in aged male mice failed to reach significance, potentially due to higher baseline levels or variation. Studies in normative human populations describe increasing levels of biomarkers as adults age.^25–27 Older individuals have higher levels of NfL in absence of injury and also display more elevated levels after brain injury²⁸; similar findings were observed here in uninjured and injured aged mice. The findings presented here may indicate age-related increases in biomarker levels in combination with injury. Despite the age-related increase in biomarkers described here, age did not modify acute biomarker levels in people with SCI.⁷ Age-related increases in biomarker levels may be too low to be detected amidst injury-induced increases, or the relatively young population in that study (mean 42.9 years) may have prevented the detection of age effects. As age was not included as a confound for chronic biomarker levels in humans,⁸ it remains to be seen if the age-related effect on trajectories described here is also observed in humans.

Increased NfL levels in adolescent males compared to females post injury may reflect underlying sex differences in murine injury course (reviewed by Stewart et al.¹³), such as increased neuroprotection in females. In humans with SCI, biological sex did not affect acute biomarker levels.⁷ Interestingly, though aged male mice trended towards higher GFAP levels than females at baseline, there were no sex difference in aged animals. These findings may represent an age-dependent loss of the purported neuroprotection or worse injury response overall with age.

Results from uninjured adolescent and aged mice support the assertion that mice exhibit similar age-related increases to those observed in humans. NfL levels increased from 18 to 20 months of age. There was no statistically significant difference in NfL levels from 3 to 5 months of age. Of note, high variability and relatively high values were observed at 3 months of age. The male 3-month-old uninjured animals were observed repeatedly fighting in the days prior to baseline sampling, necessitating separation; fighting and social stress are implicated in serum levels of inflammatory responses.^29,30 These data may imply that as in humans,²⁵ other factors affect NfL levels. These data also illustrate a limitation of the current study's design: Without a sham injury condition, we cannot ascertain the contribution of surgical procedures to the increase in plasma biomarkers. Understanding which confounding factors impact biomarker levels (such as additional injuries³¹) will be necessary for interpretation of results.

The high mortality in the aged group mimics the high mortality rates observed in aged individuals with SCI.³² Only data from animals that reached the experimental endpoint were included. Analyzing and including data from all animals may have affected the reported means, as values from excluded animals may have been altered by injury sequelae or comorbidities.

Practical and ethical considerations prevented us from acquiring blood samples acutely post injury. Acute studies are needed to harmonize with human pathology and provide insight into the diagnostic and prognostic value of biomarkers in mice. This analysis is also limited to a single injury paradigm and severity. These data show that with a consistent mouse thoracic contusion paradigm, fairly consistent plasma levels of GFAP and NfL are achieved. In older animals, the NfL levels post injury may be sensitive to small expected random variation of force (eFigure 1g). The overall lack of correlation between injury parameters and biomarker levels in animals with comparable injuries suggests that biomarker levels are comparable as well. The lack of correlation between NfL and GFAP (except for 2 mpi 18-month-injured animals; eFigure 3) may indicate different underlying pathologies. Additional future studies should assess acute biomarker levels across a range of severities and determine their predictive value for later biomarker levels, function, and histological findings.

This relatively time- and cost-effective quantitative metric could easily be incorporated into ongoing fundamental and preclinical SCI experiments. Blood can be acquired repeatedly from the same animal in sufficient volumes to assess biomarkers longitudinally. Future studies are needed to replicate the findings described here and establish values across a range of severities and timepoints post injury. Acquisition of blood samples in the hours post injury could provide metrics on injury severity and consistency and is relevant to harmonize with ongoing clinical research. Preclinical researchers can incorporate chronic longitudinal blood sampling as a metric of recovery and intervention efficacy. The present findings describe trajectories of plasma biomarker levels over time relative to standard values unique to each analysis and thus are not normative. Establishing normative values and trajectories will be aided by deposition of preclinical data in open-access repositories.²¹ Developing the use of blood biomarkers for mouse SCI research, in harmony with their development in human trials, may provide a useful tool for understanding the pathophysiology of SCI and testing new interventions.

Funding Statement

Financial Support B.R.K. is supported by grant 3195 from Paralyzed Veterans of America Research Foundation.