http://informahealthcare.com/bij ISSN: 0269-9052 (print), 1362-301X (electronic) Brain Inj, 2014; 28(3): 341–346 ! 2014 Informa UK Ltd. DOI: 10.3109/02699052.2013.865270
ORIGINAL ARTICLE
Clinical correlations of proton magnetic resonance spectroscopy findings in acute phase after mild traumatic brain injury Sˇ. Siva´k1, M. Bittsˇansky´2, J. Grossmann1, V. Nosa´l’1, E. Kantorova´1, J. Siva´kova´3, A. Demkova´1, P. Hnilicova´2, D. Dobrota2, & E. Kurcˇa1 Clinic of Neurology, 2Department of Medical Biochemistry, and 3Clinic of Gynecology and Obstetrics, Jessenius Faculty of Medicine, Comenius University, Martin, Slovak Republic Abstaract
Keywords
Introduction: Standard brain magnetic resonance imaging (MRI) is typically normal in most patients after mild traumatic brain injury (MTBI). Proton magnetic resonance spectroscopy (1H-MRS) is more sensitive to detect subtle post-traumatic changes. The aim of the study was to evaluate the clinical correlations of these changes in the acute phase (within 3 days) after MTBI. Methods: Twenty-one patients with MTBI and 22 controls were studied. Both groups underwent neuropsychological testing and single-voxel 1H-MRS examination of both frontal lobes and upper brainstem. Results: Significant decrease in NAA was found in both frontal lobes and in NAA/Cre ratio in the right frontal lobe (p50.05). Correlation analysis showed a correlation of NAA in the left frontal lobe with Backward Digit Span (p ¼ 0.022) and Stroop test A (p ¼ 0.0034) and a weak correlation with TMT B time (p ¼ 0.046). The NAA/Cre in the right frontal lobe correlated with Stroop test A (p ¼ 0.007) and with the total score of Digit Span (p ¼ 0.016). Lower NAA was found in the upper brainstem (p ¼ 0.0157) in the sub-group of patients with post-traumatic unconsciousness. Conclusions: This study found a correlation of 1H-MRS metabolite changes with cognitive decline and presence or absence of loss of consciousness in the acute phase after MTBI.
Magnetic resonance spectroscopy, mild traumatic brain injury, neuropsychological testing
Introduction Mild traumatic brain injury (MTBI) is a very common neurotraumatological diagnosis. It is known that, in addition to different subjective symptoms, many patients develop neuropsychological dysfunction with objective impairment of attention, memory and selected executive functions [1]. Computer Tomography (CT) is still considered to be the gold standard of imaging in acute management of patients after MTBI. Its main purpose is to exclude life-threatening intracranial complications and to allow early discharge to outpatient care [2]. Magnetic resonance imaging (MRI) is not routinely used in patients after MTBI, despite its proven greater sensitivity and specificity in detecting intra-parenchymal lesions compared with CT [3]. It has been shown that patients after MTBI with obvious traumatic brain MRI lesions have more severe neuropsychological deficits than patients without traumatic lesions [4, 5]. Proton magnetic resonance spectroscopy (1H-MRS) is a non-invasive technique that allows detection and
Correspondence: Sˇtefan Siva´k, MUDr., PhD, Clinic of Neurology, Jessenius Faculty of Medicine, Comenius University, Kolla´rova 2, Martin 03659, Slovak Republic. Tel: +421 43 4203 209.Fax: +421 43 4131 005. E-mail: [email protected]
History Received 8 July 2013 Revised 8 November 2013 Accepted 9 November 2013 Published online 20 December 2013
quantification of several brain metabolites including N-acetylaspartate (NAA)—a marker of neuronal functional integrity, choline-containing compounds (Cho)—markers of cellular membrane integrity, and total creatine (Cre)—a marker of cellular energy status [6]. 1H-MRS has shown greater sensitivity to injury and can detect even that missed by convectional structural MRI of the brain [6]. 1H-MRS has been used several times to evaluate MTBI [7–18], but only a limited number of studies have been dealing with patients in the acute phase within the first days after mild brain trauma [12, 13, 17, 18]. These studies have demonstrated significant metabolite changes in patients after MTBI when compared to control groups. However, only one study evaluated associations between these metabolite changes and clinical variables in a group of university athletes within 6 days after MTBI [18]. In this study, 1H-MRS changes significantly correlated with symptom severity, but did not correlate with neuropsychological deficit. It is a common observation that most athletes with post-traumatic neuropsychological impairment recover spontaneously over 7 days after concussion [19]. This study evaluated only patients with no traumatic changes on standard brain MRI within just 3 days after MTBI. The authors were expecting to show differences not only in 1 H-MRS metabolites and subjective symptoms but also in neuropsychological parameters. The aim of this study was to
20 14
Brain Inj Downloaded from informahealthcare.com by Karolinska Institutet University Library on 01/27/15 For personal use only.
1
342
Sˇ. Siva´k et al.
determine the association of clinical or neuropsychological performance with 1H-MRS metabolite changes in this very acute phase after MTBI.
Materials and methods
Brain Inj Downloaded from informahealthcare.com by Karolinska Institutet University Library on 01/27/15 For personal use only.
Participants MTBI is defined as any blow to the head resulting in short-term impairment of neurological functions (unconsciousness lasting 30 minutes or less, confusion, disorientation and posttraumatic amnesia lasting 24 hours or less), which can be accompanied by other symptoms and signs (headache, nausea, vomiting, drowsiness, dizziness, emotional changes and cognitive deficits) [20]. The exclusion criteria for subjects were: aged under 18, previous traumatic brain injury, traumatic lesions detected by standard MRI of the brain, history of chronic alcohol or drug abuse, pre-existing neurological disorder (e.g. cerebrovascular disease, epilepsia, multiple sclerosis), any psychiatric disorder, arterial hypertension and diabetes mellitus. The exclusion criteria were assessed in the interview with the subjects and from their medical records. All patients after MTBI were routinely sent from the Emergency department to the Traumatology clinic as inpatients for observation. All the subjects who agreed to participate and met the criteria then underwent standard neurological examination, neuropsychological testing and brain MRI with 1 H-MRS one after another. All the tests were performed within 2 hours. The patients were examined 24–72 hours after the injury. The control group consisted of volunteers who responded to an appeal of the Neurology clinic to participate in the study and met the criteria. The study was approved by the local ethics committee and all participants provided informed consent before participation. Post-concussion symptom scale (PCSS) This questionnaire assesses severity of 19 common symptoms rated on a scale from 0 (none) to 6 (severe), with a maximum score of 114 [21]. The subjects filled PCSS on the test day as part of neuropsychological testing. Sub-scores for somatic (som), cognitive (cog), emotional (emot) and sleep (sleep) symptoms were counted. Neuropsychological testing To avoid the ceiling effect, this study only focused on cognitive functions which are typically more affected in patients after MTBI [22]. A battery of standard neuropsychological tests for selected cognitive functions was used. Trial Making Test (TMT) provides information about visual search speed, speed of processing, mental flexibility, as well as executive functioning [23]. Part A (TMT-A) requires the individual to draw lines to connect 25 encircled numbers distributed on a page. Part B (TMTB) is similar, except the person must alternate between numbers and letters and it is believed to be more difficult and takes longer to complete. Both sections are timed and the number of errors is counted. Concentration and Attention Test (CAT) measures complex reaction time when selective attention to letters is
Brain Inj, 2014; 28(3): 341–346
required [24]. The tested person compares two columns of letters. The task is to identify as many differences as possible during 270 seconds. The quantity score (QS) is defined as the number of identified differences. The score of errors (SEr) is defined as the number of wrong responses. The quality score (QualS) is the number of correct responses divided by the total number of responses. Wechsler Memory Scale–third edition (WMS-III) sub-tests assess different functions of memory [25]. Forward (FDS) and backward (BDS) digit-span scores and Letter–number sequencing score (LNS) measure functions of working memory. 12-Word List Test assesses verbal learning using score of the first recall (S1R), total recall score (TR), learning slope score (LSS), score of interference (SI), short and long delay recall (SDR, LDR), recognition score (RS) and percentage retention (PR). Stroop Test (ST) is a tool for evaluation of executive functions measuring selective attention, interference effect, cognitive flexibility and processing speed [26]. Time (T) and number of errors (Err) were studied in all three paradigms—A, neutral; B, congruent; and C, incongruent. MRI and 1H-MRS Patients and volunteers underwent a magnetic resonance imaging and a spectroscopy session in the hospital MRI unit using a 1.5 T Siemens Symphony scanner. The imaging part comprised of routine MRI protocols (transversal T1-w, T2-w, FLAIR, T2*-w and diffusion-weighted sequences) in order to exclude patients with visible traumatic lesions of the brain. A lesion was defined as traumatic in the presence of the following focal changes: haemorrhages (T2*-GRE) or changes in water diffusion due to cytotoxic and/or vasogenic oedema (DWI/ADC, FLAIR, which reflect cortical contusions and DAI). Patients with traumatic lesions were excluded from the study. Patients and controls with small, non-specific, hyperintense, punctate lesions recognizable only on T2/ FLAIR images were not excluded from the study. All 1HMRS voxels were placed in brain tissue that appeared normal. Three specific spectroscopic voxels were measured using single-voxel 1H-MRS: one in the upper brainstem and two symmetrically in both sides of the frontal lobes in the dorsolateral pre-frontal area (dlPFA) (see Figure 1). Voxels were chosen on the assumption that working memory, attention, executive functions (dorsolateral prefrontal cortex and its connections) and loss of consciousness (ascending reticular activating system in the brainstem where originate arousal pathways to cortex) are typically affected after concussion. Spectra were shimmed individually using a combination of field mapping and manual shim adjustments, and they were acquired using a PRESS sequence with these parameters: TE/TR ¼ 135/4000 ms, 100 averages, CHESS water suppression, voxel volume of 12 ml (cortex) and 9 ml (upper brainstem). For each voxel, a separate water-unsuppressed MR spectrum was measured: TE/TR ¼ 30/10 000 ms, two averages. The magnetic resonance session took 50 minutes. Data were processed automatically using LCModel software [27]. Metabolite concentrations were calculated in arbitrary units (a.u.) and scaled using water signal for variable coil sensitivity. This enabled direct comparison of metabolite
DOI: 10.3109/02699052.2013.865270
Correlations of 1H-MRS in acute phase after MTBI
343
Aspin-Welch test for unequal variance and non-parametric Mann-Whitney U-test were used to evaluate significant differences between groups. Correlations between 1H-MRS and neuropsychological variables were tested using Pearson or Spearman rank correlation analysis. Correlation analyses were performed only with variables sensitive enough to distinguish between patient and control groups. The significance level was set to ¼ 0.05. Statistical software package NCSS 2004 (Number Cruncher Statistical System, Kaysville, UT) was used for the calculations.
Brain Inj Downloaded from informahealthcare.com by Karolinska Institutet University Library on 01/27/15 For personal use only.
Results
Figure 1. Voxel localizations in the right (a) and left (b) frontal lobes and in the brainstem (c).
signals between different subjects and voxels. Apart from the metabolic concentrations calculated in the LCModel, following metabolite ratios were used for statistical evaluation: NAA/Cre, NAA/Cho and Cho/Cre. Statistics All values are expressed as mean standard deviation (SD). Assumption of normality was assessed using Shapiro-Wilk test and D’Agostino normality tests. Homogeneity of the variance was determined with modified Levene’s test. Parametric Student two-tailed t test for equal variance,
Twenty-two patients after MTBI and 23 healthy volunteers were screened for the study. One patient (56-year-old man) and one control (53-year-old man) were excluded from the analysis due to information about a previous concussion, long-lasting arterial hypertension and diabetes mellitus. The final patient group included 21 persons with MTBI. Patients were examined 48.9 16.0 hours (median 47.5 hours) after being injured. The final control group included 22 healthy volunteers. The demographic data of the patient and control groups are summarized in Table I. The two groups did not differ in age (p ¼ 0.65) and level of education (p ¼ 0.20). Education level was scored from 1 (primary) to 6 (postgraduate tertiary). Glasgow coma scale (GCS) score at admission was 15 in 11, 14 in five and 13 in five patients. Post-traumatic amnesia (PTA) was reported in all 21 patients with mean duration 212 326 minutes (median ¼ 60 minutes, range ¼ 5–1200 minutes). Loss of consciousness (LOC) was reported by witnesses as unresponsive state following injury. LOC was present in seven patients, not present in seven patients and there was no reliable information about LOC in seven patients. Mean duration of LOC in seven patients was 5.9 8.6 minutes (median ¼ 2 minutes, range ¼ 1–25 minutes). MTBI had different causes: falls (n ¼ 6), fighting (n ¼ 5), vehicle accidents (car (n ¼ 3); motorbike (n ¼ 2); pedestrian hit by car (n ¼ 1)), sporting accidents (ice hockey (n ¼ 1); snowboard (n ¼ 2)), and being struck by falling object (n ¼ 1). The neuropsychological findings are summarized in Table II. Higher number and intensity of subjective symptoms (PCSS total, PCSS som) were reported in the patient group (p ¼ 0.033, p ¼ 0.015). Patients after MTBI performed significantly worse in terms of immediate recall (S1R, p ¼ 0.003; TR, p ¼ 0.002), short delayed recall (SDR, p ¼ 0.037), longdelayed recall (LDR, p ¼ 0.003), memory retention (PR, p ¼ 0.014), concentration and working memory scores (BDS, p ¼ 0.033; total score of digit span, p ¼ 0.032; LNS, p ¼ 0.013) and executive functions (TMT B, p ¼ 0.049; Stroop test A, p ¼ 0.023; Stroop test C, p ¼ 0.025). Standard brain MRI ruled out traumatic lesions in all patients and controls. Non-specific, punctate T2/FLAIRhyperintense lesions were found—nine in three patients (aged 26, 52 and 54), 10 in four controls (aged 29, 51, 52 and 55). Evaluated with LCModel, the typical line width of the spectra was 0.04 and 0.06 ppm and typical SNR was 24 and 8 for the dorsolateral pre-frontal area and brainstem voxels, respectively. In comparison with the control group, the values
344
Sˇ. Siva´k et al.
Brain Inj, 2014; 28(3): 341–346
Table I. Demographic and clinical parameters.
Brain Inj Downloaded from informahealthcare.com by Karolinska Institutet University Library on 01/27/15 For personal use only.
Patients n Gender (M/F) Age Mean SD Range (years) Education (level) GCS at admission Score 15/14/13 Amnesia Yes/No Duration, mean SD (minutes) Range (minutes) LOC Yes/No/Unknown Duration, mean SD (minutes) Range (minutes)
Table II. Neuropsychological findings. Controls
p
21 19/2
22 20/2
33.2 12.4 20–58 3 1.18
31.5 11.7 20–55 3.5 1.1
0.65
11/5/5
–
–
21/0 212 326 min.
–
–
–
–
–
0.2
5–1200 7/7/7 5.9 8.6 1–25
GCS, Glasgow Coma Scale score; LOC, loss of consciousness; Education level ranges from 1 ¼ primary to 6 ¼ postgraduate tertiary.
of the patients were significantly lower in the NAA/Cr ratio in the right frontal lobe (p ¼ 0.012), NAA in the right and left frontal lobes (p ¼ 0.029, p ¼ 0.002), and Cho/Cre in the right frontal lobe (p ¼ 0.047). When the Bonferroni adjustment for number of comparisons was applied (n ¼ 3, p50.016), this significance was found only in NAA/Cre ratio in the right and in NAA in the left frontal lobe. Other metabolites and their ratios in both frontal lobes and upper brainstem did not show any significant differences (Table III). Correlation analysis showed significant correlations of NAA in the left frontal lobe with Backward Digit Span score (p ¼ 0.022, Ro ¼ 0.55) and Stroop test A time (p ¼ 0.0034, Ro ¼ 0.67) and its weak inverse correlation with TMT B time (p ¼ 0.046, Ro ¼ 0.48). The NAA/Cre in the right frontal lobe correlated significantly with Stroop test A time (p ¼ 0.007, Ro ¼ 0.63) and with the total score of Digit Span (p ¼ 0.016, r ¼ 0.58). Significant correlations were not observed between NAA in the right frontal lobe and the affected cognitive functions. No correlations of metabolite levels with the interval between the injury and MRS were found (p40.05). Patients were divided into two groups based on the presence or absence of LOC. Both groups consisted of seven patients and were similar in age (p ¼ 0.56). Significantly lower NAA was found in the upper brainstem (p ¼ 0.0157) in patients with LOC (5.07 0.38 a.u.) compared to patients without LOC (5.94 0.73 a.u.). When compared to the whole control group, the value of NAA in patients with LOC was lower (p ¼ 0.009), but it showed no significant difference in patients without LOC (p ¼ 0.32). After Bonferroni adjustment the differences still remain significant (n ¼ 3, p50.016). Other metabolites and their ratios in the brainstem and the frontal lobes did not show significant differences (p40.05).
Discussion In this study of patients in the acute phase of MTBI, a decreased NAA and NAA/Cre ratio was found in the frontal lobes in patients compared to healthy volunteers.
Test PCSS total som emot cog sleep Word list S1R TR SDR LSS SI LDR PR CAT QS SEr QualS TMT A_time (seconds) A_err B_time (seconds) B_err Digit span FDS BDS Total score LNS Score Stroop test A_time (seconds) A_err B_time (seconds) B_err C_time (seconds) C_err
Patients
Controls
p
33.4 28.2 9.5 8.7 5.2 6.1 12.8 12.5 5.9 5.6
10.9 11.5 3.1 4.6 1.7 1.6 3.8 3.8 2.3 2.2
0.033 0.015 0.074 0.11 0.075
5.1 1.9 31.3 5.7 6.3 1.9 4.4 1.5 3.2 1.7 5.3 2.5 54.4 23.0
7.1 1.3 37.8 4.7 8 2.4 3.8 1.3 2.8 1.8 8.3 2.5 75.6 19.7
0.003 0.002 0.037 0.53 0.61 0.003 0.014
64.3 20.6 4.8 4.4 0.94 0.06
69.1 19.5 2.7 1.4 0.97 0.02
0.52 0.13 0.22
40.6 16.8 0.06 0.24 103.1 41.8 1.1 1.5
39.3 18.8 0.25 0.45 83.6 45.2 0.9 1.2
0.68 0.13 0.049 0.91
8.6 1.6 6.3 2.1 14.9 2.6
9.5 1.7 7.7 1.6 17.2 2.7
0.17 0.033 0.032
9.4 2.7
11.8 1.7
0.013
20.5 4.3 0.2 0.5 25 7.9 0.22 0.5 47 15.7 1.8 3.0
17.3 2.6 00 22.4 3.0 0.08 0.3 37.2 5.0 0.3 0.7
0.023 0.24 0.23 0.5 0.025 0.063
PCSS, Post-concussion Symptom Scale with its somatic (som), cognitive (cog), emotional (emot) and sleep (sleep) symptom sub-scores; 12Word List Test sub-scores: score of the first recall (S1R), total recall score (TR), learning slope score (LSS), score of interference (SI), short and long delay recall (SDR, LDR), percentage retention (PR); Concentration and Attention Test (CAT) with quantity score (QS), score of errors (SEr) and quality score (QualS); Trial Making Test (TMT A and B) with time for completion (time) and number of errors (err); forward (FDS) and backward (BDS) Digit-Span Scores; Letter– Number Sequencing Score (LNS); Stroop Test (A, B and C part) with time for completion (time) and number of errors (err).
These 1H-MRS results correlated with selected neuropsychological tests. The small sub-group of patients with posttraumatic unconsciousness had lower levels of NAA in the upper brainstem than the sub-group without unconsciousness. So far, 1H-MRS has been used in several studies to evaluate patients with mild traumatic brain injuries at different times after head trauma. In most of the studies, patients were examined 1 month [7–13] or longer [13–17] after the injury. This study examined patients between 24–72 hours (median ¼ 47.5 hours) after the brain injury. Apart from this study, there is only a limited number of studies dealing with patients in the acute period within the first days or the first week after MTBI [12, 13, 17, 18]. Vagnozzi et al. [12], in their multi-centre study, found lower NAA/Cre and NAA/Cho ratios in frontal lobes in 40 patients on the 3rd day after the injury. In the following study of 11 patients after MTBI, Vagnozzi et al. [13] found higher NAA/Cre and Cho/Cre
Correlations of 1H-MRS in acute phase after MTBI
DOI: 10.3109/02699052.2013.865270
345
Table III. Single-voxel 1H-MRS findings.
Right frontal lobe Patients Controls p Left frontal lobe Patients Controls p Pons Patients Controls p
Cho (a.u.)
Cho/Cre
Cre (a.u.)
NAA (a.u.)
NAA/Cre
NAA/Cho
0.99 0.16 1.04 0.12 0.29
0.35 0.04 0.38 0.04 0.047
2.8 0.2 2.7 0.2 0.39
5.5 0.45 5.8 0.44 0.029
1.98 0.19 2.14 0.21 0.012*
5.7 1.0 5.7 0.7 0.93
0.94 0.2 1.0 0.1 0.08
0.36 0.06 0.37 0.04 0.4
2.6 0.3 2.8 0.2 0.077
5.3 0.5 5.8 0.3 0.002*
2.03 0.17 2.1 0.19 0.21
5.9 1.3 5.7 0.5 0.92
1.3 0.2 1.3 0.1 0.67
0.5 0.08 0.5 0.07 0.61
2.6 0.3 2.6 0.3 0.92
5.8 0.8 5.7 0.5 0.7
2.2 0.4 2.2 0.3 0.55
4.6 0.57 4.44 0.48 0.26
Brain Inj Downloaded from informahealthcare.com by Karolinska Institutet University Library on 01/27/15 For personal use only.
a.u., arbitrary units; *significant after Bonferroni adjustment.
ratios and a lower NAA/Cho ratio. The authors concluded that there is a concomitant decrease in NAA and a more pronounced decrease of Cre levels on the 3rd day [13]. The advantage of this study was inclusion of the unsuppressed water measurement of each voxel. While Vagnozzi et al. [13] had to hypothesize which of the two metabolites in a ratio was actually responsible for the change, the correction here for water signal enabled this study to directly identify each single metabolite. The results support their findings only partially. Similarly, lower NAA was found in both frontal lobes, but this study cannot support their findings of the pronounced decrease in Cre levels and the increase of Cho levels. It is supposed that the most important cause of the differences between the study results is the use of different 1H-MRS techniques and post-processing methods. Data of Vagnozzi et al. [12, 13] were evaluated by simple integration of the metabolite peaks, which is sufficient for longer echo proton spectra. However, the approach of the LCModel, accounting for the complexity of the signals and using regularized line widths, is considered to be more accurate, especially in cases when two peaks are not completely differentiated, such as Cre and Cho. Henry et al. [17, 18], in their two studies, found a lower average NAA/Cre ratio in the dorsolateral prefrontal cortex and M1 area. In contrast to Henry et al., this study did not average metabolite values for both hemispheres together, but it evaluated each frontal lobe voxel separately. Their findings of bilateral affection of frontal lobes is, however, supported, as this study found differences between the hemispheres in subtle metabolite levels that were also supported by cognitive testing. A major strength of this study is that it studied the cohort at a very acute stage after MTBI. The patients were examined on the 2nd or 3rd day, which is earlier than other studies (Vagnozzi et al. [12, 13] on the 3rd day, Henry et al. [17, 18] within 6 days). This could be another possible explanation of the differences in results caused by development of neurometabolic changes several days after concussion [28]. There are several studies which evaluated and found clinical correlations of 1H-MRS metabolite changes in patients with moderate or severe traumatic brain injury [29–32]. This study focused only on patients with mild traumatic brain injury without notable traumatic changes on standard brain MRI within 3 days after trauma. As had been expected, it was found that 1H-MRS changes in the
dorsolateral pre-frontal area significantly correlate with post-traumatic cognitive dysfunction in terms of executive functions, speed of information processing (TMT-B, Stroop test A) and working memory (Digit Span). Several studies in patients after MTBI attempted to assess clinical and neuropsychological value of metabolite changes detected by single voxel 1H-MRS or multi-voxel MR spectroscopic imaging (MRSI) [9–11, 18]. Apart from this study, only Henry at al.’s [18] study focused on patients in the acute period within the first week after MTBI. Gasparovic et al. [9] showed a correlation of Cre in the splenium and white matter with executive function and emotional distress scores. Govindaraju et al. [11] found a weak correlation between global NAA/Cho ratio and Glasgow Outcome Score on discharge. Other two studies found a weak correlation of self-report symptoms with white matter glutamate-glutamine (Glx) [10] and a significant correlation of self-report symptoms with a NAA/Cre ratio in M1 area [18]. Yeo et al. [10] and Henry et al. [18] did not find any correlations between selected neuropsychological tests and the metabolite changes that they detected. However, it is important to notice that, according to their results, both groups of authors used neuropsychological tests, which were not sensitive enough to detect differences between the control and patient group at the time of examination. Several experimental studies in animal models of concussion with LOC demonstrated structural alterations in the brainstem [33, 34]. This study can support these findings with the in-vivo human study in patients after MTBI with LOC. These patients had lower NAA in the upper brainstem when compared to patients without LOC and controls. In conclusion, proton magnetic resonance spectroscopy in patients after MTBI is sensitive enough to detect posttraumatic metabolic changes in brain tissue that standard MRI detects as normal. This study has found a correlation of 1 H-MRS metabolite changes with cognitive decline and presence or absence of loss of consciousness in the acute phase after MTBI.
Declaration of interest The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper. This work was supported by the grants co-financed from EU
346
Sˇ. Siva´k et al.
resources and European social fund—ITMS 26110230067 and 2012/31-UKMA-8.
Brain Inj Downloaded from informahealthcare.com by Karolinska Institutet University Library on 01/27/15 For personal use only.
References 1. Grindel SH, Lovell MR, Collins MW. The assessment of sportrelated concussion: The evidence behind neuropsychological testing and management. Clinical Journal of Sport Medicine 2001;11:134–143. 2. Vos PE, Alekseenko Y, Battistin L, Ehler E, Gerstenbrand F, Muresanu DF, Potapov A, Stepan CA, Traubner P, Vecsei L, et al. Mild traumatic brain injury. European Journal of Neurology 2012; 19:191–198. 3. McCullough BJ, Jarvik JG. Diagnosis of concussion: The role of imaging now and in the future. Physical Medicine Rehabilitation Clinics of North America 2011;22:635–652. 4. Hughes DG, Jackson A, Mason DL, Berry E, Hollis S, Yates DW. Abnormalities on magnetic resonance imaging seen acutely following mild traumatic brain injury: Correlation with neuropsychological tests and delayed recovery. Neuroradiology 2004;46: 550–558. 5. Kurcˇa E, Siva´k S, Kucˇera P. Impaired cognitive functions in mild traumatic brain injury patients with normal and pathologic magnetic resonance imaging. Neuroradiology 2006;48:661–669. 6. Gonzalez PG, Walker MT. Imaging modalities in mild traumatic brain injury and sports concussion. PM&R 2011 Oct;3(10 Suppl 2): S413–424. 7. Johnson B, Zhang K, Gay M, Neuberger T, Horovitz S, Hallett M, Sebastianelli W, Slobounov S. Metabolic alterations in corpus callosum may compromise brain functional connectivity in MTBI patients: An 1H-MRS study. Neuroscience Letters 2012;509:5–8. 8. Johnson B, Gay M, Zhang K, Neuberger T, Horovitz SG, Hallett M, Sebastianelli W, Slobounov S. The use of magnetic resonance spectroscopy in the subacute evaluation of athletes recovering from single and multiple mild traumatic brain injury. Journal of Neurotrauma 2012;29:2297–2304. 9. Gasparovic C, Yeo R, Mannell M, Ling J, Elgie R, Phillips J, Doezema D, Mayer AR. Neurometabolite concentrations in gray and white matter in mild traumatic brain injury: An 1H-magnetic resonance spectroscopy study. Journal of Neurotrauma 2009;26: 1635–1643. 10. Yeo RA, Gasparovic C, Merideth F, Ruhl D, Doezema D, Mayer AR. A longitudinal proton magnetic resonance spectroscopy study of mild traumatic brain injury. Journal of Neurotrauma 2011;28: 1–11. 11. Govindaraju V, Gauger GE, Manley GT, Ebel A, Meeker M, Maudsley AA. Volumetric proton spectroscopic imaging of mild traumatic brain injury. American Journal of Neuroradiology 2004; 25:730–737. 12. Vagnozzi R, Signoretti S, Cristofori L, Alessandrini F, Floris R, Isgro` E, Ria A, Marziale S, Zoccatelli G, Tavazzi B, et al. Assessment of metabolic brain damage and recovery following mild traumatic brain injury: A multicentre, proton magnetic resonance spectroscopic study in concussed patients. Brain 2010;133: 3232–3242. 13. Vagnozzi R, Signoretti S, Floris R, Marziali S, Manara M, Amorini AM, Belli A, Di Pietro V, D’urso S, Pastore FS, et al. Decrease in N-Acetylaspartate following concussion may be coupled to decrease in Creatine. Journal of Head Trauma Rehabilitation 2013 Jul–Aug;28(4):284–92. 14. Kirov II, Tal A, Babb JS, Reaume J, Bushnik T, Ashman T, Flanagan SR, Grossman RI, Gonen O. Proton MR spectroscopy correlates diffuse axonal abnormalities with post-concussive symptoms in mild traumatic brain injury. Journal of Neurotrauma 2013; 30:1200–1204. 15. Kirov II, Tal A, Babb JS, Lui YW, Grossman RI, Gonen O. Diffuse axonal injury in mild traumatic brain injury: A 3D multivoxel proton MR spectroscopy study. Journal of Neurology 2013;260: 242–252.
Brain Inj, 2014; 28(3): 341–346
16. Cohen BA, Inglese M, Rusinek H, Babb JS, Grossman RI, Gonen O. Proton MR spectroscopy and MRI-volumetry in mild traumatic brain injury. American Journal of Neuroradiology 2007;28: 907–913. 17. Henry LC, Tremblay S, Leclerc S, Khiat A, Boulanger Y, Ellemberg D, Lassonde M. Metabolic changes in concussed American football players during the acute and chronic postinjury phases. BMC Neurology 2011;11:105. 18. Henry LC, Tremblay S, Boulanger Y, Ellemberg D, Lassonde M. Neurometabolic changes in the acute phase after sports concussions correlate with symptom severity. Journal of Neurotrauma 2010;27: 65–76. 19. Belanger HG, Vanderploeg RD. The neuropsychological impact of sports-related concussion: A meta-analysis. Journal of the International Neuropsychological Society 2005;11:345–357. 20. Carroll LJ, Cassidy JD, Holm L, Kraus J, Coronado VG; WHO Collaborating Centre Task Force on Mild Traumatic Brain Injury. Methodological issues and research recommendations for mild traumatic brain injury: The WHO Collaborating Centre Task Force on Mild Traumatic Brain Injury. Journal of Rehabilitation Medicine 2004 Feb;(43 Suppl):113–125. 21. Aubry M, Cantu R, Dvorak J, Graf-Baumann T, Johnston K, Kelly J, Lovell M, McCrory P, Meeuwisse W, Schamasch P. Concussion in Sport Group. Summary and agreement statement of the First International Conference on Concussion in Sport, Vienna 2001. Recommendations for the improvement of safety and health of athletes who may suffer concussive injuries. British Journal of Sports Medicine 2002;36:6–10. 22. Johnston KM, McCrory P, Mohtadi NG, Meeuwisse W. Evidencebased review of sport-related concussion: Clinical science. Clinical Journal of Sport Medicine 2001;11:150–159. 23. Reitan RM. Validity of the Trail Making test as an indicator of organic brain damage. Perceptual & Motor Skills 1958;8: 271–276. 24. Kucˇera M. Test koncentrace pozornosti. Bratislava: Psychodiagnosticke´ a didakticke´ testy; 1980. 25. Wechsler D. Wechsler Memory Scale. 3rd ed (WMS-III, Slovak version). Bratislava: Psychodiagnostika; 1999. 26. Stroop JR. Studies of interference in serial verbal reactions. Journal of Experimental Psychology 1935;18:643–662. 27. Provencher SW. Estimation of metabolite concentrations from localized in vivo proton NMR spectra. Magnetic Resonance Medicine 1993;30:672–679. 28. Giza CC, Hovda DA. The neurometabolic cascade of concussion. Journal of Athletic Training 2001;36:228–235. 29. Marino S, Ciurleo R, Bramanti P, Federico A, De Stefano N. 1 H-MR spectroscopy in traumatic brain injury. Neurocritical Care 2011;14:127–133. 30. Govind V, Gold S, Kaliannan K, Saigal G, Falcone S, Arheart KL, Harris L, Jagid J, Maudsley AA. Whole-brain proton MR spectroscopic imaging of mild-to-moderate traumatic brain injury and correlation with neuropsychological deficits. Journal of Neurotrauma 2010;27:483–496. 31. Garnett MR, Blamire AM, Corkill RG, Cadoux-Hudson TA, Rajagopalan B, Styles P. Early proton magnetic resonance spectroscopy in normal-appearing brain correlates with outcome in patients following traumatic brain injury. Brain 2000;123: 2046–2054. 32. Ariza M, Junque´ C, Mataro´ M, Poca MA, Bargallo´ N, Olondo M, Sahuquillo J. Neuropsychological correlates of basal ganglia and medial temporal lobe NAA/Cho reductions in traumatic brain injury. Archives of Neurology 2004;61:541–544. 33. Jane JA, Steward O, Gennarelli T. Axonal degeneration induced by experimental noninvasive minor head injury. Journal of Neurosurgery 1985;62:96–100. 34. Brown WJ, Yoshida N, Canty T, Verity MA. Experimental concussion. Ultrastructural and biochemical correlates. American Journal of Pathology 1972;67:41–68.
ORIGINAL ARTICLE
Clinical correlations of proton magnetic resonance spectroscopy findings in acute phase after mild traumatic brain injury Sˇ. Siva´k1, M. Bittsˇansky´2, J. Grossmann1, V. Nosa´l’1, E. Kantorova´1, J. Siva´kova´3, A. Demkova´1, P. Hnilicova´2, D. Dobrota2, & E. Kurcˇa1 Clinic of Neurology, 2Department of Medical Biochemistry, and 3Clinic of Gynecology and Obstetrics, Jessenius Faculty of Medicine, Comenius University, Martin, Slovak Republic Abstaract
Keywords
Introduction: Standard brain magnetic resonance imaging (MRI) is typically normal in most patients after mild traumatic brain injury (MTBI). Proton magnetic resonance spectroscopy (1H-MRS) is more sensitive to detect subtle post-traumatic changes. The aim of the study was to evaluate the clinical correlations of these changes in the acute phase (within 3 days) after MTBI. Methods: Twenty-one patients with MTBI and 22 controls were studied. Both groups underwent neuropsychological testing and single-voxel 1H-MRS examination of both frontal lobes and upper brainstem. Results: Significant decrease in NAA was found in both frontal lobes and in NAA/Cre ratio in the right frontal lobe (p50.05). Correlation analysis showed a correlation of NAA in the left frontal lobe with Backward Digit Span (p ¼ 0.022) and Stroop test A (p ¼ 0.0034) and a weak correlation with TMT B time (p ¼ 0.046). The NAA/Cre in the right frontal lobe correlated with Stroop test A (p ¼ 0.007) and with the total score of Digit Span (p ¼ 0.016). Lower NAA was found in the upper brainstem (p ¼ 0.0157) in the sub-group of patients with post-traumatic unconsciousness. Conclusions: This study found a correlation of 1H-MRS metabolite changes with cognitive decline and presence or absence of loss of consciousness in the acute phase after MTBI.
Magnetic resonance spectroscopy, mild traumatic brain injury, neuropsychological testing
Introduction Mild traumatic brain injury (MTBI) is a very common neurotraumatological diagnosis. It is known that, in addition to different subjective symptoms, many patients develop neuropsychological dysfunction with objective impairment of attention, memory and selected executive functions [1]. Computer Tomography (CT) is still considered to be the gold standard of imaging in acute management of patients after MTBI. Its main purpose is to exclude life-threatening intracranial complications and to allow early discharge to outpatient care [2]. Magnetic resonance imaging (MRI) is not routinely used in patients after MTBI, despite its proven greater sensitivity and specificity in detecting intra-parenchymal lesions compared with CT [3]. It has been shown that patients after MTBI with obvious traumatic brain MRI lesions have more severe neuropsychological deficits than patients without traumatic lesions [4, 5]. Proton magnetic resonance spectroscopy (1H-MRS) is a non-invasive technique that allows detection and
Correspondence: Sˇtefan Siva´k, MUDr., PhD, Clinic of Neurology, Jessenius Faculty of Medicine, Comenius University, Kolla´rova 2, Martin 03659, Slovak Republic. Tel: +421 43 4203 209.Fax: +421 43 4131 005. E-mail: [email protected]
History Received 8 July 2013 Revised 8 November 2013 Accepted 9 November 2013 Published online 20 December 2013
quantification of several brain metabolites including N-acetylaspartate (NAA)—a marker of neuronal functional integrity, choline-containing compounds (Cho)—markers of cellular membrane integrity, and total creatine (Cre)—a marker of cellular energy status [6]. 1H-MRS has shown greater sensitivity to injury and can detect even that missed by convectional structural MRI of the brain [6]. 1H-MRS has been used several times to evaluate MTBI [7–18], but only a limited number of studies have been dealing with patients in the acute phase within the first days after mild brain trauma [12, 13, 17, 18]. These studies have demonstrated significant metabolite changes in patients after MTBI when compared to control groups. However, only one study evaluated associations between these metabolite changes and clinical variables in a group of university athletes within 6 days after MTBI [18]. In this study, 1H-MRS changes significantly correlated with symptom severity, but did not correlate with neuropsychological deficit. It is a common observation that most athletes with post-traumatic neuropsychological impairment recover spontaneously over 7 days after concussion [19]. This study evaluated only patients with no traumatic changes on standard brain MRI within just 3 days after MTBI. The authors were expecting to show differences not only in 1 H-MRS metabolites and subjective symptoms but also in neuropsychological parameters. The aim of this study was to
20 14
Brain Inj Downloaded from informahealthcare.com by Karolinska Institutet University Library on 01/27/15 For personal use only.
1
342
Sˇ. Siva´k et al.
determine the association of clinical or neuropsychological performance with 1H-MRS metabolite changes in this very acute phase after MTBI.
Materials and methods
Brain Inj Downloaded from informahealthcare.com by Karolinska Institutet University Library on 01/27/15 For personal use only.
Participants MTBI is defined as any blow to the head resulting in short-term impairment of neurological functions (unconsciousness lasting 30 minutes or less, confusion, disorientation and posttraumatic amnesia lasting 24 hours or less), which can be accompanied by other symptoms and signs (headache, nausea, vomiting, drowsiness, dizziness, emotional changes and cognitive deficits) [20]. The exclusion criteria for subjects were: aged under 18, previous traumatic brain injury, traumatic lesions detected by standard MRI of the brain, history of chronic alcohol or drug abuse, pre-existing neurological disorder (e.g. cerebrovascular disease, epilepsia, multiple sclerosis), any psychiatric disorder, arterial hypertension and diabetes mellitus. The exclusion criteria were assessed in the interview with the subjects and from their medical records. All patients after MTBI were routinely sent from the Emergency department to the Traumatology clinic as inpatients for observation. All the subjects who agreed to participate and met the criteria then underwent standard neurological examination, neuropsychological testing and brain MRI with 1 H-MRS one after another. All the tests were performed within 2 hours. The patients were examined 24–72 hours after the injury. The control group consisted of volunteers who responded to an appeal of the Neurology clinic to participate in the study and met the criteria. The study was approved by the local ethics committee and all participants provided informed consent before participation. Post-concussion symptom scale (PCSS) This questionnaire assesses severity of 19 common symptoms rated on a scale from 0 (none) to 6 (severe), with a maximum score of 114 [21]. The subjects filled PCSS on the test day as part of neuropsychological testing. Sub-scores for somatic (som), cognitive (cog), emotional (emot) and sleep (sleep) symptoms were counted. Neuropsychological testing To avoid the ceiling effect, this study only focused on cognitive functions which are typically more affected in patients after MTBI [22]. A battery of standard neuropsychological tests for selected cognitive functions was used. Trial Making Test (TMT) provides information about visual search speed, speed of processing, mental flexibility, as well as executive functioning [23]. Part A (TMT-A) requires the individual to draw lines to connect 25 encircled numbers distributed on a page. Part B (TMTB) is similar, except the person must alternate between numbers and letters and it is believed to be more difficult and takes longer to complete. Both sections are timed and the number of errors is counted. Concentration and Attention Test (CAT) measures complex reaction time when selective attention to letters is
Brain Inj, 2014; 28(3): 341–346
required [24]. The tested person compares two columns of letters. The task is to identify as many differences as possible during 270 seconds. The quantity score (QS) is defined as the number of identified differences. The score of errors (SEr) is defined as the number of wrong responses. The quality score (QualS) is the number of correct responses divided by the total number of responses. Wechsler Memory Scale–third edition (WMS-III) sub-tests assess different functions of memory [25]. Forward (FDS) and backward (BDS) digit-span scores and Letter–number sequencing score (LNS) measure functions of working memory. 12-Word List Test assesses verbal learning using score of the first recall (S1R), total recall score (TR), learning slope score (LSS), score of interference (SI), short and long delay recall (SDR, LDR), recognition score (RS) and percentage retention (PR). Stroop Test (ST) is a tool for evaluation of executive functions measuring selective attention, interference effect, cognitive flexibility and processing speed [26]. Time (T) and number of errors (Err) were studied in all three paradigms—A, neutral; B, congruent; and C, incongruent. MRI and 1H-MRS Patients and volunteers underwent a magnetic resonance imaging and a spectroscopy session in the hospital MRI unit using a 1.5 T Siemens Symphony scanner. The imaging part comprised of routine MRI protocols (transversal T1-w, T2-w, FLAIR, T2*-w and diffusion-weighted sequences) in order to exclude patients with visible traumatic lesions of the brain. A lesion was defined as traumatic in the presence of the following focal changes: haemorrhages (T2*-GRE) or changes in water diffusion due to cytotoxic and/or vasogenic oedema (DWI/ADC, FLAIR, which reflect cortical contusions and DAI). Patients with traumatic lesions were excluded from the study. Patients and controls with small, non-specific, hyperintense, punctate lesions recognizable only on T2/ FLAIR images were not excluded from the study. All 1HMRS voxels were placed in brain tissue that appeared normal. Three specific spectroscopic voxels were measured using single-voxel 1H-MRS: one in the upper brainstem and two symmetrically in both sides of the frontal lobes in the dorsolateral pre-frontal area (dlPFA) (see Figure 1). Voxels were chosen on the assumption that working memory, attention, executive functions (dorsolateral prefrontal cortex and its connections) and loss of consciousness (ascending reticular activating system in the brainstem where originate arousal pathways to cortex) are typically affected after concussion. Spectra were shimmed individually using a combination of field mapping and manual shim adjustments, and they were acquired using a PRESS sequence with these parameters: TE/TR ¼ 135/4000 ms, 100 averages, CHESS water suppression, voxel volume of 12 ml (cortex) and 9 ml (upper brainstem). For each voxel, a separate water-unsuppressed MR spectrum was measured: TE/TR ¼ 30/10 000 ms, two averages. The magnetic resonance session took 50 minutes. Data were processed automatically using LCModel software [27]. Metabolite concentrations were calculated in arbitrary units (a.u.) and scaled using water signal for variable coil sensitivity. This enabled direct comparison of metabolite
DOI: 10.3109/02699052.2013.865270
Correlations of 1H-MRS in acute phase after MTBI
343
Aspin-Welch test for unequal variance and non-parametric Mann-Whitney U-test were used to evaluate significant differences between groups. Correlations between 1H-MRS and neuropsychological variables were tested using Pearson or Spearman rank correlation analysis. Correlation analyses were performed only with variables sensitive enough to distinguish between patient and control groups. The significance level was set to ¼ 0.05. Statistical software package NCSS 2004 (Number Cruncher Statistical System, Kaysville, UT) was used for the calculations.
Brain Inj Downloaded from informahealthcare.com by Karolinska Institutet University Library on 01/27/15 For personal use only.
Results
Figure 1. Voxel localizations in the right (a) and left (b) frontal lobes and in the brainstem (c).
signals between different subjects and voxels. Apart from the metabolic concentrations calculated in the LCModel, following metabolite ratios were used for statistical evaluation: NAA/Cre, NAA/Cho and Cho/Cre. Statistics All values are expressed as mean standard deviation (SD). Assumption of normality was assessed using Shapiro-Wilk test and D’Agostino normality tests. Homogeneity of the variance was determined with modified Levene’s test. Parametric Student two-tailed t test for equal variance,
Twenty-two patients after MTBI and 23 healthy volunteers were screened for the study. One patient (56-year-old man) and one control (53-year-old man) were excluded from the analysis due to information about a previous concussion, long-lasting arterial hypertension and diabetes mellitus. The final patient group included 21 persons with MTBI. Patients were examined 48.9 16.0 hours (median 47.5 hours) after being injured. The final control group included 22 healthy volunteers. The demographic data of the patient and control groups are summarized in Table I. The two groups did not differ in age (p ¼ 0.65) and level of education (p ¼ 0.20). Education level was scored from 1 (primary) to 6 (postgraduate tertiary). Glasgow coma scale (GCS) score at admission was 15 in 11, 14 in five and 13 in five patients. Post-traumatic amnesia (PTA) was reported in all 21 patients with mean duration 212 326 minutes (median ¼ 60 minutes, range ¼ 5–1200 minutes). Loss of consciousness (LOC) was reported by witnesses as unresponsive state following injury. LOC was present in seven patients, not present in seven patients and there was no reliable information about LOC in seven patients. Mean duration of LOC in seven patients was 5.9 8.6 minutes (median ¼ 2 minutes, range ¼ 1–25 minutes). MTBI had different causes: falls (n ¼ 6), fighting (n ¼ 5), vehicle accidents (car (n ¼ 3); motorbike (n ¼ 2); pedestrian hit by car (n ¼ 1)), sporting accidents (ice hockey (n ¼ 1); snowboard (n ¼ 2)), and being struck by falling object (n ¼ 1). The neuropsychological findings are summarized in Table II. Higher number and intensity of subjective symptoms (PCSS total, PCSS som) were reported in the patient group (p ¼ 0.033, p ¼ 0.015). Patients after MTBI performed significantly worse in terms of immediate recall (S1R, p ¼ 0.003; TR, p ¼ 0.002), short delayed recall (SDR, p ¼ 0.037), longdelayed recall (LDR, p ¼ 0.003), memory retention (PR, p ¼ 0.014), concentration and working memory scores (BDS, p ¼ 0.033; total score of digit span, p ¼ 0.032; LNS, p ¼ 0.013) and executive functions (TMT B, p ¼ 0.049; Stroop test A, p ¼ 0.023; Stroop test C, p ¼ 0.025). Standard brain MRI ruled out traumatic lesions in all patients and controls. Non-specific, punctate T2/FLAIRhyperintense lesions were found—nine in three patients (aged 26, 52 and 54), 10 in four controls (aged 29, 51, 52 and 55). Evaluated with LCModel, the typical line width of the spectra was 0.04 and 0.06 ppm and typical SNR was 24 and 8 for the dorsolateral pre-frontal area and brainstem voxels, respectively. In comparison with the control group, the values
344
Sˇ. Siva´k et al.
Brain Inj, 2014; 28(3): 341–346
Table I. Demographic and clinical parameters.
Brain Inj Downloaded from informahealthcare.com by Karolinska Institutet University Library on 01/27/15 For personal use only.
Patients n Gender (M/F) Age Mean SD Range (years) Education (level) GCS at admission Score 15/14/13 Amnesia Yes/No Duration, mean SD (minutes) Range (minutes) LOC Yes/No/Unknown Duration, mean SD (minutes) Range (minutes)
Table II. Neuropsychological findings. Controls
p
21 19/2
22 20/2
33.2 12.4 20–58 3 1.18
31.5 11.7 20–55 3.5 1.1
0.65
11/5/5
–
–
21/0 212 326 min.
–
–
–
–
–
0.2
5–1200 7/7/7 5.9 8.6 1–25
GCS, Glasgow Coma Scale score; LOC, loss of consciousness; Education level ranges from 1 ¼ primary to 6 ¼ postgraduate tertiary.
of the patients were significantly lower in the NAA/Cr ratio in the right frontal lobe (p ¼ 0.012), NAA in the right and left frontal lobes (p ¼ 0.029, p ¼ 0.002), and Cho/Cre in the right frontal lobe (p ¼ 0.047). When the Bonferroni adjustment for number of comparisons was applied (n ¼ 3, p50.016), this significance was found only in NAA/Cre ratio in the right and in NAA in the left frontal lobe. Other metabolites and their ratios in both frontal lobes and upper brainstem did not show any significant differences (Table III). Correlation analysis showed significant correlations of NAA in the left frontal lobe with Backward Digit Span score (p ¼ 0.022, Ro ¼ 0.55) and Stroop test A time (p ¼ 0.0034, Ro ¼ 0.67) and its weak inverse correlation with TMT B time (p ¼ 0.046, Ro ¼ 0.48). The NAA/Cre in the right frontal lobe correlated significantly with Stroop test A time (p ¼ 0.007, Ro ¼ 0.63) and with the total score of Digit Span (p ¼ 0.016, r ¼ 0.58). Significant correlations were not observed between NAA in the right frontal lobe and the affected cognitive functions. No correlations of metabolite levels with the interval between the injury and MRS were found (p40.05). Patients were divided into two groups based on the presence or absence of LOC. Both groups consisted of seven patients and were similar in age (p ¼ 0.56). Significantly lower NAA was found in the upper brainstem (p ¼ 0.0157) in patients with LOC (5.07 0.38 a.u.) compared to patients without LOC (5.94 0.73 a.u.). When compared to the whole control group, the value of NAA in patients with LOC was lower (p ¼ 0.009), but it showed no significant difference in patients without LOC (p ¼ 0.32). After Bonferroni adjustment the differences still remain significant (n ¼ 3, p50.016). Other metabolites and their ratios in the brainstem and the frontal lobes did not show significant differences (p40.05).
Discussion In this study of patients in the acute phase of MTBI, a decreased NAA and NAA/Cre ratio was found in the frontal lobes in patients compared to healthy volunteers.
Test PCSS total som emot cog sleep Word list S1R TR SDR LSS SI LDR PR CAT QS SEr QualS TMT A_time (seconds) A_err B_time (seconds) B_err Digit span FDS BDS Total score LNS Score Stroop test A_time (seconds) A_err B_time (seconds) B_err C_time (seconds) C_err
Patients
Controls
p
33.4 28.2 9.5 8.7 5.2 6.1 12.8 12.5 5.9 5.6
10.9 11.5 3.1 4.6 1.7 1.6 3.8 3.8 2.3 2.2
0.033 0.015 0.074 0.11 0.075
5.1 1.9 31.3 5.7 6.3 1.9 4.4 1.5 3.2 1.7 5.3 2.5 54.4 23.0
7.1 1.3 37.8 4.7 8 2.4 3.8 1.3 2.8 1.8 8.3 2.5 75.6 19.7
0.003 0.002 0.037 0.53 0.61 0.003 0.014
64.3 20.6 4.8 4.4 0.94 0.06
69.1 19.5 2.7 1.4 0.97 0.02
0.52 0.13 0.22
40.6 16.8 0.06 0.24 103.1 41.8 1.1 1.5
39.3 18.8 0.25 0.45 83.6 45.2 0.9 1.2
0.68 0.13 0.049 0.91
8.6 1.6 6.3 2.1 14.9 2.6
9.5 1.7 7.7 1.6 17.2 2.7
0.17 0.033 0.032
9.4 2.7
11.8 1.7
0.013
20.5 4.3 0.2 0.5 25 7.9 0.22 0.5 47 15.7 1.8 3.0
17.3 2.6 00 22.4 3.0 0.08 0.3 37.2 5.0 0.3 0.7
0.023 0.24 0.23 0.5 0.025 0.063
PCSS, Post-concussion Symptom Scale with its somatic (som), cognitive (cog), emotional (emot) and sleep (sleep) symptom sub-scores; 12Word List Test sub-scores: score of the first recall (S1R), total recall score (TR), learning slope score (LSS), score of interference (SI), short and long delay recall (SDR, LDR), percentage retention (PR); Concentration and Attention Test (CAT) with quantity score (QS), score of errors (SEr) and quality score (QualS); Trial Making Test (TMT A and B) with time for completion (time) and number of errors (err); forward (FDS) and backward (BDS) Digit-Span Scores; Letter– Number Sequencing Score (LNS); Stroop Test (A, B and C part) with time for completion (time) and number of errors (err).
These 1H-MRS results correlated with selected neuropsychological tests. The small sub-group of patients with posttraumatic unconsciousness had lower levels of NAA in the upper brainstem than the sub-group without unconsciousness. So far, 1H-MRS has been used in several studies to evaluate patients with mild traumatic brain injuries at different times after head trauma. In most of the studies, patients were examined 1 month [7–13] or longer [13–17] after the injury. This study examined patients between 24–72 hours (median ¼ 47.5 hours) after the brain injury. Apart from this study, there is only a limited number of studies dealing with patients in the acute period within the first days or the first week after MTBI [12, 13, 17, 18]. Vagnozzi et al. [12], in their multi-centre study, found lower NAA/Cre and NAA/Cho ratios in frontal lobes in 40 patients on the 3rd day after the injury. In the following study of 11 patients after MTBI, Vagnozzi et al. [13] found higher NAA/Cre and Cho/Cre
Correlations of 1H-MRS in acute phase after MTBI
DOI: 10.3109/02699052.2013.865270
345
Table III. Single-voxel 1H-MRS findings.
Right frontal lobe Patients Controls p Left frontal lobe Patients Controls p Pons Patients Controls p
Cho (a.u.)
Cho/Cre
Cre (a.u.)
NAA (a.u.)
NAA/Cre
NAA/Cho
0.99 0.16 1.04 0.12 0.29
0.35 0.04 0.38 0.04 0.047
2.8 0.2 2.7 0.2 0.39
5.5 0.45 5.8 0.44 0.029
1.98 0.19 2.14 0.21 0.012*
5.7 1.0 5.7 0.7 0.93
0.94 0.2 1.0 0.1 0.08
0.36 0.06 0.37 0.04 0.4
2.6 0.3 2.8 0.2 0.077
5.3 0.5 5.8 0.3 0.002*
2.03 0.17 2.1 0.19 0.21
5.9 1.3 5.7 0.5 0.92
1.3 0.2 1.3 0.1 0.67
0.5 0.08 0.5 0.07 0.61
2.6 0.3 2.6 0.3 0.92
5.8 0.8 5.7 0.5 0.7
2.2 0.4 2.2 0.3 0.55
4.6 0.57 4.44 0.48 0.26
Brain Inj Downloaded from informahealthcare.com by Karolinska Institutet University Library on 01/27/15 For personal use only.
a.u., arbitrary units; *significant after Bonferroni adjustment.
ratios and a lower NAA/Cho ratio. The authors concluded that there is a concomitant decrease in NAA and a more pronounced decrease of Cre levels on the 3rd day [13]. The advantage of this study was inclusion of the unsuppressed water measurement of each voxel. While Vagnozzi et al. [13] had to hypothesize which of the two metabolites in a ratio was actually responsible for the change, the correction here for water signal enabled this study to directly identify each single metabolite. The results support their findings only partially. Similarly, lower NAA was found in both frontal lobes, but this study cannot support their findings of the pronounced decrease in Cre levels and the increase of Cho levels. It is supposed that the most important cause of the differences between the study results is the use of different 1H-MRS techniques and post-processing methods. Data of Vagnozzi et al. [12, 13] were evaluated by simple integration of the metabolite peaks, which is sufficient for longer echo proton spectra. However, the approach of the LCModel, accounting for the complexity of the signals and using regularized line widths, is considered to be more accurate, especially in cases when two peaks are not completely differentiated, such as Cre and Cho. Henry et al. [17, 18], in their two studies, found a lower average NAA/Cre ratio in the dorsolateral prefrontal cortex and M1 area. In contrast to Henry et al., this study did not average metabolite values for both hemispheres together, but it evaluated each frontal lobe voxel separately. Their findings of bilateral affection of frontal lobes is, however, supported, as this study found differences between the hemispheres in subtle metabolite levels that were also supported by cognitive testing. A major strength of this study is that it studied the cohort at a very acute stage after MTBI. The patients were examined on the 2nd or 3rd day, which is earlier than other studies (Vagnozzi et al. [12, 13] on the 3rd day, Henry et al. [17, 18] within 6 days). This could be another possible explanation of the differences in results caused by development of neurometabolic changes several days after concussion [28]. There are several studies which evaluated and found clinical correlations of 1H-MRS metabolite changes in patients with moderate or severe traumatic brain injury [29–32]. This study focused only on patients with mild traumatic brain injury without notable traumatic changes on standard brain MRI within 3 days after trauma. As had been expected, it was found that 1H-MRS changes in the
dorsolateral pre-frontal area significantly correlate with post-traumatic cognitive dysfunction in terms of executive functions, speed of information processing (TMT-B, Stroop test A) and working memory (Digit Span). Several studies in patients after MTBI attempted to assess clinical and neuropsychological value of metabolite changes detected by single voxel 1H-MRS or multi-voxel MR spectroscopic imaging (MRSI) [9–11, 18]. Apart from this study, only Henry at al.’s [18] study focused on patients in the acute period within the first week after MTBI. Gasparovic et al. [9] showed a correlation of Cre in the splenium and white matter with executive function and emotional distress scores. Govindaraju et al. [11] found a weak correlation between global NAA/Cho ratio and Glasgow Outcome Score on discharge. Other two studies found a weak correlation of self-report symptoms with white matter glutamate-glutamine (Glx) [10] and a significant correlation of self-report symptoms with a NAA/Cre ratio in M1 area [18]. Yeo et al. [10] and Henry et al. [18] did not find any correlations between selected neuropsychological tests and the metabolite changes that they detected. However, it is important to notice that, according to their results, both groups of authors used neuropsychological tests, which were not sensitive enough to detect differences between the control and patient group at the time of examination. Several experimental studies in animal models of concussion with LOC demonstrated structural alterations in the brainstem [33, 34]. This study can support these findings with the in-vivo human study in patients after MTBI with LOC. These patients had lower NAA in the upper brainstem when compared to patients without LOC and controls. In conclusion, proton magnetic resonance spectroscopy in patients after MTBI is sensitive enough to detect posttraumatic metabolic changes in brain tissue that standard MRI detects as normal. This study has found a correlation of 1 H-MRS metabolite changes with cognitive decline and presence or absence of loss of consciousness in the acute phase after MTBI.
Declaration of interest The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper. This work was supported by the grants co-financed from EU
346
Sˇ. Siva´k et al.
resources and European social fund—ITMS 26110230067 and 2012/31-UKMA-8.
Brain Inj Downloaded from informahealthcare.com by Karolinska Institutet University Library on 01/27/15 For personal use only.
References 1. Grindel SH, Lovell MR, Collins MW. The assessment of sportrelated concussion: The evidence behind neuropsychological testing and management. Clinical Journal of Sport Medicine 2001;11:134–143. 2. Vos PE, Alekseenko Y, Battistin L, Ehler E, Gerstenbrand F, Muresanu DF, Potapov A, Stepan CA, Traubner P, Vecsei L, et al. Mild traumatic brain injury. European Journal of Neurology 2012; 19:191–198. 3. McCullough BJ, Jarvik JG. Diagnosis of concussion: The role of imaging now and in the future. Physical Medicine Rehabilitation Clinics of North America 2011;22:635–652. 4. Hughes DG, Jackson A, Mason DL, Berry E, Hollis S, Yates DW. Abnormalities on magnetic resonance imaging seen acutely following mild traumatic brain injury: Correlation with neuropsychological tests and delayed recovery. Neuroradiology 2004;46: 550–558. 5. Kurcˇa E, Siva´k S, Kucˇera P. Impaired cognitive functions in mild traumatic brain injury patients with normal and pathologic magnetic resonance imaging. Neuroradiology 2006;48:661–669. 6. Gonzalez PG, Walker MT. Imaging modalities in mild traumatic brain injury and sports concussion. PM&R 2011 Oct;3(10 Suppl 2): S413–424. 7. Johnson B, Zhang K, Gay M, Neuberger T, Horovitz S, Hallett M, Sebastianelli W, Slobounov S. Metabolic alterations in corpus callosum may compromise brain functional connectivity in MTBI patients: An 1H-MRS study. Neuroscience Letters 2012;509:5–8. 8. Johnson B, Gay M, Zhang K, Neuberger T, Horovitz SG, Hallett M, Sebastianelli W, Slobounov S. The use of magnetic resonance spectroscopy in the subacute evaluation of athletes recovering from single and multiple mild traumatic brain injury. Journal of Neurotrauma 2012;29:2297–2304. 9. Gasparovic C, Yeo R, Mannell M, Ling J, Elgie R, Phillips J, Doezema D, Mayer AR. Neurometabolite concentrations in gray and white matter in mild traumatic brain injury: An 1H-magnetic resonance spectroscopy study. Journal of Neurotrauma 2009;26: 1635–1643. 10. Yeo RA, Gasparovic C, Merideth F, Ruhl D, Doezema D, Mayer AR. A longitudinal proton magnetic resonance spectroscopy study of mild traumatic brain injury. Journal of Neurotrauma 2011;28: 1–11. 11. Govindaraju V, Gauger GE, Manley GT, Ebel A, Meeker M, Maudsley AA. Volumetric proton spectroscopic imaging of mild traumatic brain injury. American Journal of Neuroradiology 2004; 25:730–737. 12. Vagnozzi R, Signoretti S, Cristofori L, Alessandrini F, Floris R, Isgro` E, Ria A, Marziale S, Zoccatelli G, Tavazzi B, et al. Assessment of metabolic brain damage and recovery following mild traumatic brain injury: A multicentre, proton magnetic resonance spectroscopic study in concussed patients. Brain 2010;133: 3232–3242. 13. Vagnozzi R, Signoretti S, Floris R, Marziali S, Manara M, Amorini AM, Belli A, Di Pietro V, D’urso S, Pastore FS, et al. Decrease in N-Acetylaspartate following concussion may be coupled to decrease in Creatine. Journal of Head Trauma Rehabilitation 2013 Jul–Aug;28(4):284–92. 14. Kirov II, Tal A, Babb JS, Reaume J, Bushnik T, Ashman T, Flanagan SR, Grossman RI, Gonen O. Proton MR spectroscopy correlates diffuse axonal abnormalities with post-concussive symptoms in mild traumatic brain injury. Journal of Neurotrauma 2013; 30:1200–1204. 15. Kirov II, Tal A, Babb JS, Lui YW, Grossman RI, Gonen O. Diffuse axonal injury in mild traumatic brain injury: A 3D multivoxel proton MR spectroscopy study. Journal of Neurology 2013;260: 242–252.
Brain Inj, 2014; 28(3): 341–346
16. Cohen BA, Inglese M, Rusinek H, Babb JS, Grossman RI, Gonen O. Proton MR spectroscopy and MRI-volumetry in mild traumatic brain injury. American Journal of Neuroradiology 2007;28: 907–913. 17. Henry LC, Tremblay S, Leclerc S, Khiat A, Boulanger Y, Ellemberg D, Lassonde M. Metabolic changes in concussed American football players during the acute and chronic postinjury phases. BMC Neurology 2011;11:105. 18. Henry LC, Tremblay S, Boulanger Y, Ellemberg D, Lassonde M. Neurometabolic changes in the acute phase after sports concussions correlate with symptom severity. Journal of Neurotrauma 2010;27: 65–76. 19. Belanger HG, Vanderploeg RD. The neuropsychological impact of sports-related concussion: A meta-analysis. Journal of the International Neuropsychological Society 2005;11:345–357. 20. Carroll LJ, Cassidy JD, Holm L, Kraus J, Coronado VG; WHO Collaborating Centre Task Force on Mild Traumatic Brain Injury. Methodological issues and research recommendations for mild traumatic brain injury: The WHO Collaborating Centre Task Force on Mild Traumatic Brain Injury. Journal of Rehabilitation Medicine 2004 Feb;(43 Suppl):113–125. 21. Aubry M, Cantu R, Dvorak J, Graf-Baumann T, Johnston K, Kelly J, Lovell M, McCrory P, Meeuwisse W, Schamasch P. Concussion in Sport Group. Summary and agreement statement of the First International Conference on Concussion in Sport, Vienna 2001. Recommendations for the improvement of safety and health of athletes who may suffer concussive injuries. British Journal of Sports Medicine 2002;36:6–10. 22. Johnston KM, McCrory P, Mohtadi NG, Meeuwisse W. Evidencebased review of sport-related concussion: Clinical science. Clinical Journal of Sport Medicine 2001;11:150–159. 23. Reitan RM. Validity of the Trail Making test as an indicator of organic brain damage. Perceptual & Motor Skills 1958;8: 271–276. 24. Kucˇera M. Test koncentrace pozornosti. Bratislava: Psychodiagnosticke´ a didakticke´ testy; 1980. 25. Wechsler D. Wechsler Memory Scale. 3rd ed (WMS-III, Slovak version). Bratislava: Psychodiagnostika; 1999. 26. Stroop JR. Studies of interference in serial verbal reactions. Journal of Experimental Psychology 1935;18:643–662. 27. Provencher SW. Estimation of metabolite concentrations from localized in vivo proton NMR spectra. Magnetic Resonance Medicine 1993;30:672–679. 28. Giza CC, Hovda DA. The neurometabolic cascade of concussion. Journal of Athletic Training 2001;36:228–235. 29. Marino S, Ciurleo R, Bramanti P, Federico A, De Stefano N. 1 H-MR spectroscopy in traumatic brain injury. Neurocritical Care 2011;14:127–133. 30. Govind V, Gold S, Kaliannan K, Saigal G, Falcone S, Arheart KL, Harris L, Jagid J, Maudsley AA. Whole-brain proton MR spectroscopic imaging of mild-to-moderate traumatic brain injury and correlation with neuropsychological deficits. Journal of Neurotrauma 2010;27:483–496. 31. Garnett MR, Blamire AM, Corkill RG, Cadoux-Hudson TA, Rajagopalan B, Styles P. Early proton magnetic resonance spectroscopy in normal-appearing brain correlates with outcome in patients following traumatic brain injury. Brain 2000;123: 2046–2054. 32. Ariza M, Junque´ C, Mataro´ M, Poca MA, Bargallo´ N, Olondo M, Sahuquillo J. Neuropsychological correlates of basal ganglia and medial temporal lobe NAA/Cho reductions in traumatic brain injury. Archives of Neurology 2004;61:541–544. 33. Jane JA, Steward O, Gennarelli T. Axonal degeneration induced by experimental noninvasive minor head injury. Journal of Neurosurgery 1985;62:96–100. 34. Brown WJ, Yoshida N, Canty T, Verity MA. Experimental concussion. Ultrastructural and biochemical correlates. American Journal of Pathology 1972;67:41–68.
