Correlation of ubiquitin C terminal hydrolase and S100β with cognitive deficits in young adults with mild traumatic brain injury
Correspondence Address: Source of Support: None, Conflict of Interest: None DOI: 10.4103/neuroindia.NI_884_15
Source of Support: None, Conflict of Interest: None
Keywords: Mild traumatic brain injury, post-concussion syndrome, S100B neuropsychological outcome, serum biomarkers, traumatic brain injury, ubiquitin C terminal hydrolase-L1
The number of people with a mild head injury and the subsequent post-concussion syndrome (PCS) is quite high. On an average, each year, 250–300 hospital admissions per 100,000 of population involve head injuries, of which at least 75% are mild.,,, Mild traumatic brain injury (mTBI) is the most common cause of cognitive impairment in the younger population. Many of these patients with a mild head injury have no symptoms whatsoever; however, at least half experience some PCS. These symptoms include headache, dizziness, memory impairment, and concentration problems. Most of them recover completely within 3 months of injury; however, approximately one-third have some persisting symptoms beyond this time. Approximately 8% have significant symptoms at one year, and in some cases, these symptoms are possibly permanent., Clearly, mTBI is not always a mild experience.,
Cognitive impairment is a major component of the PCS and can be measured objectively by neuropsychological assessment. Though it is rarely performed in the acute setting, it may have some value in predicting the development of symptoms., Cognitive impairments following mTBI are usually transient. Areas of impairment are generally similar to those involved in more severe injuries and include problems with attention, concentration, processing speed, executive function, and memory. Complaints are more common immediately following the injury; however, in a vast majority of cases, the impairments spontaneously resolve, allowing the patients to return to normal daily functioning., When these impairments are subtle and persistent, they become apparent only during increased cognitive demand.
The S100 calcium-binding (S100) proteins are a family of Ca 2+-binding proteins that help to regulate the intracellular levels of calcium. The S100B is a marker of astroglial injury and is the most widely used biomarker in TBI. The levels of S100B in serum are increased in patients with TBI and correlate with Glasgow Coma Scale (GCS) scores and neuroradiological findings on hospital admission., The levels of S100B can help to differentiate patients with mTBI from those with severe TBI and improve the predictions of their outcome., Ubiquitin C-terminal hydrolase (UCH-L1) is a marker of neuronal injury. The UCH-L1 levels increase in a manner similar to S100B in terms of correlation with other measures of the severity of TBI, as well as clinical outcome. Biomarkers have been evaluated for their ability to diagnose mTBI. Unfortunately, at present, there are no accurate prognostic biomarkers of mTBI outcome. These markers have been used to correlate outcome of mTBI with indicators of recovery such as return to work (RTW) and Glasgow Outcome Scale Extended (GOSE). These scales may not be appropriate for use in mTBI. Moreover, these scales, which are based on limited clinical symptom assessment, should be questioned because the potential for subjective interpretation is high. In the scenario of mTBI, neuropsychological assessment provides an objective measure of brain injury and recovery.
We hypothesized that the concentration of biomarkers during the acute phase after mTBI will have a correlation with the cognitive deficits at follow up.
Study design and patient population
A prior approval from the institute's ethical committee (NIMHANS/DO/SUB-COMMITTEE/2013/NIMHANS/IEC Sl. No. 1, Clinical Neurosciences) was obtained, and informed consent was taken from participants. This is a prospective observational study of patients with mTBI. The operational definition of mTBI used was “an acute alteration in brain function or loss of consciousness for 30 minutes or less caused by a blunt external force and a Glasgow Coma Scale (GCS) score of 14 to 15 at the time of presentation.” We did not include patients with a GCS of 13 because there is a controversy regarding the status of these patients. Although they are being designated as having a mild head injury, many believe that a GCS score of 13 should be clubbed with moderate TBI. The inclusion criterion was an age between 19–40 years, with the patients presenting to the casualty within 6 hours of trauma, and classified as mTBI. The following patients were excluded: Patients with multiple trauma, those with deterioration in GCS after inclusion, those with a pre-existing neurological disease, and those with a history of alcohol consumption. The controls comprised age, gender, and educational-status matched healthy volunteers without neurological or psychiatric disorders. The subjects, both cases and controls, were inhabitants of the city of Bangalore, southern India. The primary language of all of them was Kannada.
Venous blood samples were collected at two time points (the first within 6 hours of injury, and the second 6–12 hours after the injury) for UCH-L1 and S100B measurement. Serum was separated and stored at −80°C until the time of analysis. Serum UCH-L1 and S-100B were measured by sandwich enzyme-linked immunosorbent assay (ELISA) using commercial kits (Sunred Biological Technology Co., Ltd, Shanghai) at our institute's neurochemistry laboratory. The manufacturer reported a range of between 1 ng/L and 300 ng/L and within-series coefficient of variance of <10% for S100B. For UCH-L1, the range was between 0.2 ng/mL and 30 ng/mL and within-series coefficient of variance of <10%. The sensitivity of the ELISA kits for S100B was 0.932 ng/L and for UCH-L1 was 0.125 ng/mL. The assays were carried out in duplicates. The mean values were taken in the study.
Neuropsychological assessment was done at 3 months after the injury using the National Institute of Mental Health and Neurosciences (NIMHANS) neuropsychological battery for head injury. These tests are standardized to the Indian population in Kannada language with normative data. The tests used are listed in [Table 1].
The data were analysed using the statistical software Statistical Package for the Social Sciences (SPSS) version 22 (IBM, USA). Descriptive statistics such as percentages were calculated for the categorical variables such as gender, whereas other descriptive indices, such as mean and standard deviation, were calculated for continuous data such as age, education, neuropsychological scores, and serum values of biomarkers. The groups were compared using Student's t-test (two tailed) because most of the data followed the normal distribution curve. Mann–Witney U test was used for the groups that did not follow normal distribution. The correlation coefficients were calculated using the Spearman's rho.
A total of 20 patients and 20 controls were enrolled in the study. The mean age of the patients with TBI was 30.5 years (range, 17–40 years) with 90% males (18 out of 20), and the mean duration of education was 8.7 years (range 2–12 years). The mean age of the controls was 30.5 years (range, 18–40 years) with 90% males and the mean duration of education was 9.1 years (range 4–12 years). The three most common injury mechanisms were road traffic accidents (45%), falls (15%), and motorcycle crashes (15%). The computed tomography scan revealed cerebral contusion in 8 (40%), extradural hematoma in 4 (20%), pneumocephalus in 3 (15%), subdural hematoma in 2 (10%), and subarachnoid hemorrhage in 2 patients. One (5%) patient had no brain parenchymal injury nor any skull fracture. However, none of the findings were significant enough to require intervention.
The average time to serum collection for patients was 3.7 hours for the first sample and 8.65 hours for the second sample. The second sample of 2 patients showed a technical error in the optical density, and hence were excluded from the study group for analysis. The difference in mean serum values were not statistically significant between patients and controls (P > 0.05) [Table 2].
Three patients could not complete the Stroop test and 1 patient could not complete the trail making test. One patient could neither complete the Stroop nor trail making test due to a low educational status. Three patients could not perform the complex figure test (CFT). The patients showed cognitive decline in focused attention and perpetual thinking (color trail test), working verbal memory (digit span test), verbal fluency (controlled oral word association tests), response inhibition (Stroop test), visual constructive ability and visual memory, and verbal learning and memory [Rey's auditory verbal learning test (AVLT)]. There were significant differences in the neuropsychological scores between the patients and the controls for all neuropsychological tests [Table 3].
Relationship between biomarkers and neuropsychological scores
S100 calcium-binding protein B
The patients' serum levels of S100B of the first sample (less than six hours of injury) correlated negatively with the scores of digit span test (rho = −0.481, P = 0.05) and AVLT (rho = −0.559, P = 0.01). The levels of the second sample (more than 6 hours of injury) correlated negatively with the digit span (rho = −0.457 P = 0.05) and controlled oral word association tests (COWA) [rho = −0.508, P = 0.05], and AVLT (rho = −0.602, P = 0.01) [Table 4].
Ubiquitin C terminal hydrolase
The patients' serum levels of UCH-L1 of the first sample (less than 6 hours of injury) showed a positive correlation with AVLT (rho = 0.529, P = 0.05). However, UCH-L1 levels of the second sample (more than 6 hours of injury) correlated negatively with CFT (rho = −0.478, P = 0.05), and positively with spatial span (rho = 0.650, P< 0.01) [Table 4].
Other neuropsychological scores did not correlate significantly with serum levels of biomarkers. These findings indicate that serum levels of S100B at an acute stage after mTBI correlate with deficits in working memory, verbal learning, and verbal fluency at 3 months after injury. The serum levels of UCH-L1 at the acute stage after mTBI correlate with deficits in verbal learning, visual memory, and working memory at 3 months after injury. This implies that more the concentration of serum levels of S100B and UCH-L1, more is the neuropsychological impairment.
The current study was designed to evaluate the role of biomarkers in predicting the cognitive deficits in mTBI after 3 months of injury. We chose 3 months after injury for assessment because many patients improve within 3 months, and we wanted to determine whether or not the patients with persistent symptoms had elevated biomarkers in the acute stage after injury., We used the standard biomarker S100B and the relatively new biomarker UCH-L1, as being representative of glial and neuronal injury, respectively, after mTBI. The serum levels of both the markers was not significanly different compared to their levels in controls. This indicates that the biomarker measurement is not a sensitive tool to differentiate persons with mTBI from the normal population. In literature, there are conflicting reports regarding the usefulness of biomarkers as a predictor for PCS.
S100 calcium-binding protein B
The concentration level of S100B was found to be significantly higher in patients with mTBI than in controls by Iverson et al. The S100B levels were associated with brain abnormalities detected by computed tomography scans, and the prediction of a poor outcome as measured by the Glasgow Outcome Scale (GOS) and return to work (RTW). There is little literature that supports the role of biomarkers in consistently predicting the extent or persistence of cognitive impairment after mTBI. The first large study suggesting a relationship between serum S100B concentration and neuropsychological outcome after mTBI was published by Ingebrigtsen, et al. A total of 50 patients with mTBI underwent S100B level estimation within 12 hours of injury. Patients with a detectable S100B concentration showed a trend toward impaired neuropsychological functioning on measures of information-processing speed, memory, and attention. In another study, serial S100B concentrations were analysed during the first three days after TBI of all severity. Most of these patients had mTBI. Twenty nine out of 69 patients with a median GCS score of 13 (3–15) were followed up at six months after injury. Patients with neuropsychological disorders, defined as a performance of less than one standard deviation below (age adjusted) normal data in at least three cognitive domains, had significantly higher S100B serum concentrations during the first three days after TBI. However, other studies failed to show a relationship between serum S100B concentration after mTBI and outcome. In a study by Stapert et al., comprising 50 patients with mTBI, serum S100B concentration were measured within 6 hours of injury, and the outcome was assessed at a median of 13 days (range 7–21 days) after injury by a validated neuropsychological test battery. There was lack of a difference in outcome between patients with low and high S100B concentrations. In another study by de Boussard, et al., comprising 97 patients with mTBI, S100B concentration were measured in the emergency department. S100B concentrations was measured on the next day, and subsequently at 14 days, and 3 months after injury. The outcome measures included computerized neuropsychological testing as well as cognitive testing and assessment of post-concussive symptoms. No significant association between S100B concentration and symptoms or signs of cognitive impairment were found. Instead of cut-off values for S100B concentrations, we correlated the serum levels with neuropsychological scores as this is a continuous variable. Levels of S100B at an early acute stage (less than six hours) after mild TBI correlated with deficits in working memory, as tested by digit span test. Levels of S100B in late acute stage (more than six hours) after mTBI correlated with deficits in working memory as tested by digit span test, in verbal learning as tested by AVLT, and in verbal fluency as tested by COWA [Table 4]. This suggests the usefulness of early as well as late S100B levels in predicting working memory, though it did not correlate with neuropsychological scores of all cognitive domains.
Ubiquitin C terminal hydrolase
The UCH-L1 is highly abundant in the brain, present exclusively in neurons, making it attractive as a brain biomarker. The concentration of this protein is at least 50 times higher in the brain than any other tissue. Given the massive preponderance of UCH-L1 in the brain, it seems reasonable to argue for its improved specificity over S100B. It is a novel biomarker of TBI. The cerebrospinal fluid (CSF) and serum concentrations of UCH-L1 are elevated after TBI. There are a few reports on the utility of UCH-L1 in severe TBI.,,, However, the utility of UCH-L1 as a biomarker in mTBI is limited. In a study by Papa et al., the authors compared early serum levels of UCH-L1 from patients with mild and moderate TBI with uninjured and injured controls and examined their association with traumatic intracranial lesions seen on CT scan and the need for neurosurgical intervention. Their patient population included 86 patients with mTBI. The average time to serum collection for TBI patients was 2.7 hours. The lower limit of detection (LOD) was determined to be 0.03ng/mL. UCH-L1 was detectable in serum within an hour of injury. Early UCH-L1 levels demonstrated significant differences between patients with a GCS score of 15 versus the uninjured control patients (P < 0.001), and between patients with a GCS score of 15 versus the trauma control patients (P < 0.022). Area under the curve (AUC) was calculated from the receiver operating characteristics (ROC) curves constructed to assess the performance of early UCH-L1 levels in distinguishing TBI from control patients. The AUC for UCH-L1 to differentiate TBI patients with a GCS score 15 from uninjured controls was 0.87 (excellent). The authors concluded that UCH-L1 is associated with measures that indicate the severity of injury, including the GCS score, CT lesions, and neurosurgical intervention in patients with mild and moderate TBI. Contrary to this study Puvenna et al., did not find the utility of UCH-L1 in mTBI. In their study, they recruited football players with sub-concussive head hits, patients with mTBI, and healthy volunteers. Serum levels of S100B, UCH-L1, and beta-2 transferrin were measured within 6 hours of injury. They found that low levels of S100B were able to rule out mTBI and high S100B levels correlated with the TBI severity. However, UCH-L1 did not display any interpretable change in football players or in individuals with mTBI. There was no correlation between UCH-L1 levels and mTBI patients. They concluded that the significance of UCH-L1 changes in mTBI needs to be further elucidated. In the present study, the control subjects had UCH-L1 levels comparable to patients with mTBI. We did not find any significant difference of UCH-L1 levels between the controls and patients with mTBI. None of the previous studies addressed the utility of UCH-L1 in determining outcome of mTBI. In the present study, levels of UCH-L1 at the early acute stage (less than six hours) after mTBI correlated with deficits in verbal learning as tested by AVLT, and in working memory as tested by the digit span test, when tested three months after injury. Levels of UCH-L1 in the late acute stage (more than six hours) after mTBI correlated with deficits in working memory as tested by spatial span test, and in visual memory as tested by CFT [Table 4]. However, serum levels of UCH-L1 did not correlate with neuropsychological scores of all cognitive domains.
The inconsistency in the results of studies of biomarkers in mTBI is due to many factors. The first of these factors is the lack of consensus regarding the definition of mTBI. Mild TBI is defined very differently by different investigators, making comparison of studies difficult. The second factor is the heterogeneity of the inclusion and exclusion criteria., The third factor is the heterogeneity in the outcomes being evaluated., While there are multiple validated ways to measure the neuropsychological outcome, it is difficult to directly compare the tests to one another. The final factor is the lack of adequate control subjects for comparison of outcome. The outcome after mTBI is affected by several pre-injury, injury, and post injury factors; and the controls should be matched for all of them. A group of control patients with non-cranial injuries besides healthy adults should also be included as the possible neuropsychological outcomes can be affected due to injury and hospitalization rather than TBI per se.
The major limitation of this study is the smaller size of the convenient sample rather than the inclusion of consecutive patients. There was difficulty in getting patients motivated to undergo the neuropsychological assessment because they apparently felt well. Second, we did not find significant differences in the serum levels of biomarkers between patients with mTBI and controls. As mentioned earlier, we did not use controls with similar experiences of trauma, but without TBI. The elevated serum biomarkers as well as neuropsychological impairment and PCS can be seen in patients with extracranial injuries as well. As the hospital in which this study was conducted is a primary neurological hospital, we do not get cases with isolated extracranial injuries. Although the serum markers were not significantly elevated, they could still be correlated with neuropsychological impairment. Prospective studies with larger cohorts are required to determine the accuracy of the predictability of biomarkers in determining neuropsychological impairment after mTBI.
There was insignificant change in the serum S100B and UCH-L1 levels in patients with mTBI. Patients with mTBI had significant cognitive deficits at 3 months after injury, which is suggestive of involvement of diffuse areas of the brain, particularly the premotor, prefrontal, and medial inferior frontal lobes and the basitemporal region. The correlation of biomarkers with cognitive deficits in mild head injury was found in the following domains: working memory, verbal learning, verbal fluency, and visual memory in the short term. The serum biomarkers may differentiate patients with mTBI from the normal subjects and have correlation with the neuropsychological outcome in the selected domains.
Declaration of interest
Subir Dey received a grant for this study from “NEUROCON2011” fund. The other authors report no declarations of interest.
Financial support and sponsorship
Conflicts of interest
There are no conflicts of interest.
[Table 1], [Table 3], [Table 2], [Table 4]