Temporal Profile of Microtubule-Associated Protein 2: A Novel Indicator of Diffuse Brain Injury Severity and Early Mortality after Brain Trauma

Abstract

This study compared cerebrospinal fluid (CSF) levels of microtubule-associated protein 2 (MAP-2) from adult patients with severe traumatic brain injury (TBI) with uninjured controls over 10 days, and examined the relationship between MAP-2 concentrations and acute clinical and radiologic measures of injury severity along with mortality at 2 weeks and over 6 months. This prospective study, conducted at two Level 1 trauma centers, enrolled adults with severe TBI (Glasgow Coma Scale [GCS] score ≤8) requiring a ventriculostomy, as well as controls. Ventricular CSF was sampled from each patient at 6, 12, 24, 48, 72, 96, 120, 144, 168, 192, 216, and 240 h following TBI and analyzed via enzyme-linked immunosorbent assay for MAP-2 (ng/mL). Injury severity was assessed by the GCS score, Marshall Classification on computed tomography (CT), Rotterdam CT score, and mortality. There were 151 patients enrolled-130 TBI and 21 control patients. MAP-2 was detectable within 6 h of injury and was significantly elevated compared with controls (p < 0.001) at each time-point. MAP-2 was highest within 72 h of injury and decreased gradually over 10 days. The area under the receiver operating characteristic curve for deciphering TBI versus controls at the earliest time-point CSF was obtained was 0.96 (95% CI 0.93-0.99) and for the maximal 24-h level was 0.98 (95% CI 0.97-1.00). The area under the curve for initial MAP-2 levels predicting 2-week mortality was 0.80 at 6 h, 0.81 at 12 h, 0.75 at 18 h, 0.75 at 24 h, and 0.80 at 48 h. Those with Diffuse Injury III-IV had much higher initial (p = 0.033) and maximal (p = 0.003) MAP-2 levels than those with Diffuse Injury I-II. There was a graded increase in the overall levels and peaks of MAP-2 as the degree of diffuse injury increased within the first 120 h post-injury. These data suggest that early levels of MAP-2 reflect severity of diffuse brain injury and predict 2-week mortality in TBI patients. These findings have implications for counseling families and improving clinical decision making early after injury and guiding multidisciplinary care. Further studies are needed to validate these findings in a larger sample.

Keywords: biomarkers; diffuse axonal injury; microtubule-associated protein; mortality; neuronal injury; outcome; severe traumatic brain injury; traumatic brain injury.

FIG. 1.

(A) Temporal profile of microtubule-associated protein 2 (MAP-2) in severe traumatic brain injury patients measured over 10 days versus uninjured control patients. MAP-2 was detectable at the earliest time-point measured in the cerebrospinal fluid (CSF; i.e., within 6 h of injury) and was significantly elevated compared to controls (p < 0.001) at each time-point. Lines represent means ± standard error. The number of available samples at each time-point was 21 controls, 69 at enrollment, 23 at 6 h, 57 at 12 h, 82 at 18 h, 89 at 24 h, 87 at 48 h, 74 at 72 h, 66 at 96 h, 58 at 120 h, 53 at 144 h, 50 at 168 h, 48 at 192 h, 49 at 216 h, and 41 at 240 h. (B) Temporal profile of MAP-2 is compared in survivors versus non-survivors at 2 weeks. CSF levels are significantly higher in non-survivors compared to survivors for the first 48 h after injury (p < 0.001). Listed below the figure are the areas under the curve for predicting mortality at 2 weeks with MAP-2 at each time-point collected post-injury.

FIG. 2.

Boxplot of initial microtubule-associated protein 2 (MAP-2) levels within 24 h versus dichotomized post-resuscitation Glasgow Coma Scale (GCS) score. Patients with a post-resuscitation GCS score of 3–5 had significantly higher initial cerebrospinal fluid (obtained after ventriculostomy; p = 0.040) and maximal 24-h levels (p = 0.024) of MAP-2, compared with patients with a GCS 6–8. Boxplots represent medians with interquartile ranges.

FIG. 3.

(A) Boxplot of levels microtubule-associated protein 2 (MAP-2) levels versus Rotterdam CT (computed tomography) score. Concentrations of MAP-2 in CSF increased with risk of mortality but did not reach statistical significance. Boxplots represent medians with interquartile ranges. (B) Pattern of biomarker release over 10 days versus Rotterdam CT score categories. The pattern of biomarker release in the different risk categories of the Rotterdam CT score is displayed. The levels of MAP-2 increase dramatically once the risk of death is over 63%. Lines represent mean values of MAP-2 over time.

FIG. 4.

(A) Boxplot of levels microtubule-associated protein 2 (MAP-2) levels versus Marshall Classification. Overall, there were significant differences between the groups for both initial MAP-2 levels (p = 0.05) and maximal 24-h levels in cerebrospinal fluid (p = 0.008). Levels were highest among patients with diffuse injury, compared with those with mass lesions. Among the diffuse injury group, those with Diffuse Injury III-IV had higher initial (p = 0.115) and maximal (p = 0.007) MAP-2 levels than those with Diffuse Injury I-II. Boxplots represent medians with interquartile ranges. (B) Pattern of biomarker release over 10 days versus Marshall Classification. There is a graded increase in the overall levels and the peaks of MAP-2 as the degree of diffuse injury increases. This is mostly seen within the first 120 h post-injury. Also, MAP-2 levels were slightly higher in non-evacuated mass lesions than evacuated mass lesions. Lines represent mean values of MAP-2 over time.

FIG. 5.

(A) Boxplot of levels microtubule-associated protein 2 (MAP-2) levels versus survival at 6 months. Cerebrospinal fluid MAP-2 levels are compared in those who did and did not survive to 6 months post-injury. Initial (p = 0.015) and maximal 24-h (p = 0.006) levels were significantly higher in those in those who did not survive. All MAP-2 levels drawn within the first 24 h were significantly higher in non-survivors at 6 months (p ≤ 0.05). (B) Boxplot of levels MAP-2 levels versus time of death. There were 79 patients who survived to 6 months, five who died within 48 h, 15 who died between 48 h to 1 week, seven who died 1 week to 1 month, and four who died between 1 to 3 months. Levels of MAP-2 relative to time of death showed insignificant differences in initial (p = 0.073) but significant differences in maximal MAP-2 levels (p = 0.011) between groups. The highest concentrations of MAP-2 were in those who died within 48 h. Boxplots represent medians with interquartile ranges.