Coagulopathy as a Surrogate of Severity of Injury in Penetrating Brain Injury

Abstract

Penetrating brain injury (PBI) is the most devastating type of traumatic brain injury. Development of coagulopathy in the acute setting of PBI, though common, remains of unclear significance as does its reversal. The aim of this study is to investigate the relationship between coagulopathy and clinical presentation, radiographical features, and outcome in civilian patients with PBI. Eighty-nine adult patients with PBI at a Level I trauma center in Chicago, Illinois who survived acute resuscitation and with available coagulation profile were analyzed. Coagulopathy was defined as international normalized ratio [INR] >1.3, platelet count <100,000 /μL, or partial thromboplastin time >37 sec. Median age (interquartile range; IQR) of our cohort was 27 (21–35) years, and 74 (83%) were male. The intent was assault in 74 cases (83%). The mechanism of PBI was gunshot wound in all patients. Forty patients (45%) were coagulopathic at presentation. In a multiple regression model, coagulopathy was associated with lower Glasgow Coma Scale (GCS)-Motor score (odds ratio [OR], 0.67; confidence interval [CI], 0.48–0.94; p = 0.02) and transfusion of blood products (OR, 3.91; CI, 1.2–12.5; p = 0.02). Effacement of basal cisterns was the only significant radiographical features associated with coagulopathy (OR, 3.34; CI, 1.08–10.37; p = 0.04). Mortality was found to be significantly more common in coagulopathic patients (73% vs. 25%; p < 0.001). However, in our limited sample, reversal of coagulopathy at 24 h was not associated with a statistically significant improvement in outcome. The triad of coagulopathy, low post-resuscitation GCS, and radiographical effacement of basal cisterns identify a particularly ominous phenotype of PBI. The role, and potential reversal of, coagulopathy in this group warrants further investigation.

Introduction

Penetrating brain injury (PBI) accounts for 1.5% of all traumatic brain injury (TBI) incidents in the United States and is by far the most lethal (42% of TBI deaths).¹ Most civilian cases of PBI are from gunshot wounds to the head (GSWH). Because of the highly destructive nature of ballistics to brain tissue, past studies on GSWH report a mortality rate of 71% on scene as well as an overall mortality rate of 88%.^1,3 Concurrently, civilian GSWH is poorly studied, and specific management remains elusive. No strict criteria exist for either surgical or medical treatment approaches,^4,5 which leads to treatment that is delayed, misapplied, or foregone. Identifying distinct clinical characteristics, phenotypes, and trajectories could improve the application of available medical and surgical treatments for PBI.

Coagulopathy, hypothermia, and acidosis are essential facets to the pathophysiology of traumatic injury.^6,7 However, the exact nature and course of coagulopathy, specifically in relation to TBI, remains only partially understood. Coagulopathy has been associated with progressive hematoma expansion.⁸ In comparison to blunt trauma, coagulopathy appears to be more common in PBI.² Consistent with other non-penetrating trauma mechanisms, the presence of coagulopathy in PBI correlates with higher mortality.^8–11 Nevertheless, a very limited number of studies have explicitly examined the role of coagulopathy in PBI. In this study, we sought to investigate the interaction between coagulopathy, as measured by conventional coagulation tests (CCTs), including serum platelet count, international normalized ratio (INR), and partial thromboplastin time (PTT), Glasgow Coma Scale (GCS) score on admission after initial resuscitation, radiographical variables of injury, and mortality in PBI patients. This investigation is unique to GSWH-related PBI.

Methods

Study cohort

This is a retrospective analysis of prospectively collected data from patients with PBI admitted to a single Illinois Department of Public Health Level I trauma center. Our institution maintained Level I trauma accreditation during the entire duration of the study. The institutional review board approved the study protocol under IRB-19-0220. Inclusion criteria for the study were: 1) presenting diagnosis of PBI and 2) age ≥16 years. Exclusion criteria were: 1) dead on arrival, 2) absence of dural penetration on imaging, and 3) absence of CCTs within 3 h of presentation.

Data collection

For each patient, data regarding demographics, clinical presentation, GCS score after initial resuscitation, computed tomography scan of the head (HCT) on admission, Injury Severity Score (ISS), laboratory studies, hospital length of stay (LOS), Glasgow Outcome Score-Extended (GOSE) at the time of discharge, and discharge disposition were reviewed. The GCS recorded was based on the initial neurocritical care or neurosurgical evaluation in the trauma bay after initial hemodynamic stabilization; neurological evaluation was performed off sedation and after excluding residual neuromuscular blockade. In the absence of a universally accepted definition for coagulopathy after TBI, we defined coagulopathy according to previous studies' cutoffs, and based on local institutional laboratory ranges, as the presence of any of the following: INR >1.3, platelet count <100,000 /μL, or PTT >37 sec. The coagulation profile was documented on arrival to the emergency department (within 3 h) and at 24 h for all included patients. Of note, blood samples are sent on arrival and typically before transfusions. On the other hand, transfusions are given as guided by clinical condition at the trauma attending's discretion often before the coagulation labs result is received.

HCT images were first interpreted by a board-certified neuroradiologist and subsequently reviewed and classified by two neurocritical care fellows (A.L. and F.A.). Cisternal effacement was determined if all basal cisterns were completely effaced.

Statistical analysis

Descriptive statistics are presented as means with standard deviations (SDs) or medians with interquartile ranges (IQR = Q3–Q1) for continuous variables and as percentages for categorical variables. Comparisons between coagulopathic and non-coagulopathic groups were made using nonparametric Mann-Whitney U tests or two-sample t tests. Categorical variables were compared by chi-square (χ²) or Fisher's exact test. Multi-variable logistic regression was used to investigate interactions between GCS and coagulopathy. All analyses were performed with the use of R (version 3.6.1; R Foundation for Statistical Computing, Vienna, Austria) and Stata software (version 16.1; StataCorp LP, College Station, TX).

Results

Basic characteristics of the population

Within the duration of the study (May 2018 to May 2020), 130 patients with a PBI were treated at our institution. Thirty-eight patients who died during initial resuscitation, 2 pediatric and one stab-wound patient, were excluded from further analysis (Fig. 1). Eighty-nine patients were included in the main analysis. Median age (IQR) of our cohort was 27 (21–35) years of age, and 74 (83%) were male. Most patients were African American (n = 82; 92%), followed by 4 Caucasian (5%) and 3 Hispanic (3%). The intent was assault in 74 cases (83%), self-inflicted in 8 (9%), and accidental in 7 cases (8%). The mechanism of PBI was GSWH in all patients. Seventy-six patients had isolated PBI (85%) whereas 13 patients suffered polytrauma (15%). Median GCS score (IQR) at presentation was 6 (3–14).

FIG. 1.

Study design. ED, emergency department; GSWH, gunshot wounds to the head; PBIs, penetrating brain injuries.

Coagulopathy in penetrating brain injury cohort

At presentation, 40 patients were coagulopathic (45%), compared to 49 (55%) who were not. An INR of >1.3 was observed in 38 of 39 (97%) coagulopathic patients. PTT of >37 sec was observed in 32 of 39 (82%) coagulopathic patients, and a platelet count of <100,000 was observed in 7 of 40 (18%) coagulopathic patients. One patient did not have an INR value on presentation, and another patient did not have a PTT value on presentation.

Clinical, demographic, and outcomes in PBI cohorts with and without coagulopathy were compared (Table 1). For the coagulopathic cohort, median INR (IQR) was 1.6 (1.5–2.6), median PTT (IQR) was 45 (39–79), and median platelet count (IQR) was 156 (105–181). The non-coagulopathic cohort had a median INR (IQR) of 1.1 (1.0–1.2), median PTT (IQR) was 27 (24–29), and median platelet count (IQR) was 231 (208–272; see Table 1). All three variables were significantly different between the coagulopathic versus non-coagulopathic cohorts (Mann-Whitney U test, z = 7.8, 6.5, and 6.4, respectively; p < 0.001 for all three).

Table 1.

Univariate Analysis Comparing Patients with and without Coagulopathy

PBI, penetrating brain injury; OSH, Outside Hospital; GCS, Glasgow Coma Scale; SBP, systolic blood pressure; HR, heart rate; ISS, injury severity scale; Plt, platelets; INR, International Normalized Ratio; PTT, Partial thromboplastin time; IVH, intraventricular hemorrhage; LOS, length of stay; GOSE, Glasgow outcome scale extended; IQR, interquartile range; SD, standard deviation.

Median age (IQR) in the coagulopathic cohort was 28 (21–32) years versus 27 (21–40) in the non-coagulopathic cohort. Thirty-four of 40 coagulopathic patients (85%) were male compared to 40 males (82%) in the non-coagulopathic cohort. Injury was isolated to the head in 33 (83%) coagulopathic patients versus 43 (88%) in the non-coagulopathic cohort. None of those variables was significantly different across the two groups.

Association between GCS-M and coagulopathy

Median GCS (IQR) in the coagulopathic cohort was 3 (3–5) versus 11 (5–15) in the non-coagulopathic cohort (p < 0.001). Median GCS-Motor (GCS-M; IQR) in the coagulopathic cohort was 1 (1–2) versus 6 (3–6) in the non-coagulopathic cohort (p < 0.001). Conversely, patients with GCS-M of 1 had higher rates of coagulopathy (28 of 34; 82%) compared to patients with GCS-M of 6 (4 of 29 [14%]; χ² = 37.2; p < 0.001).

Clinical and radiographical features associated with coagulopathy

Mean systolic blood pressure (SD) of the coagulopathic cohort was 119 (35) versus 130 (27) in the non-coagulopathic cohort (p = 0.11). Mean heart rate (HR; SD) in coagulopathic patients was 111 (31) and 88 (21) in the non-coagulopathic group (t_{(87 df)} = 4.2; p < 0.001). Median (IQR) Injury Severity Scale (ISS) was 70 (25–75) in the coagulopathic versus 25 (16–25) in the non-coagulopathic patients (Mann-Whitney U test, z-score = 4.3; p < 0.001). Thirty-one coagulopathic patients (80%) received at least one blood-product transfusion versus 17 (35%) non-coagulopathic patients (χ² = 18.8; p < 0.001).

Basal cisterns were effaced in 29 (73%) coagulopathic patients and in 16 (33%) non-coagulopathic patients (χ² = 12.4; p < 0.001). Bihemispheric involvement was observed in 23 (58%) coagulopathic versus 17 (35%) non-coagulopathic patients (χ² = 3.8; p = 0.05). Thalamic and brainstem lesions were observed respectively in 17 (43%) and 9 (23%) coagulopathic patients versus 8 (16%) and 2 (4%) non-coagulopathic patients (χ² = 6.2; p = 0.01 and χ² = 5.3, p = 0.02, respectively). Posterior fossa was injured in 11 (28%) coagulopathic patients versus 3 (6%) non-coagulopathic patients (χ² = 6.1; p = 0.01). Frontal, parietal, temporal, and occipital lobe were involved in 28 (70%), 23 (58%), 18 (45%), and 11 (28%) coagulopathic versus 29 (59%), 24 (49%), 14 (29%), and 8 (16%) non-coagulopathic patients, respectively (p > 0.1 for all). Intraventricular hemorrhage was present in 19 (48%) coagulopathic patients versus 11 (22%) non-coagulopathic patients (χ² = 5.1; p = 0.02).

Outcomes associated with coagulopathy

In terms of outcomes, median (IQR) hospital length of stay (LOS) for coagulopathic patients was 2 (1–13) days versus 7 (3–16) in the non-coagulopathic cohort (Mann-Whitney, z-score = 2.9; p = 0.004). Median (IQR) GOSE was 1 (1–2) in the coagulopathic cohort versus 4 (3–6) in the non-coagulopathic cohort (Mann-Whitney U test, z-score = 4.69; p < 0.001). Mortality rates were 73% versus 25% (χ² = 18.5; p < 0.001). GOSE was calculated on discharge or at time of death. Median number of days to discharge was 3.5 (1–13) and was 13 (7–21) days in survivors and 1 (1–3) in patients who expired.

In summary, GCS score on presentation, HR on presentation, transfusion of at least one blood product, and ISS were significantly different between the two cohorts (p < 0.01). Radiographical variables different across the two groups included basal cistern effacement, thalamic, brainstem, posterior fossa involvement, and intraventricular hemorrhage (p < 0.01). All three outcome surrogates: GOSE at discharge, LOS, and mortality were significantly different between the two cohorts.

Clinical predictors of coagulopathy

Multiple variable logistic regression models were fit to further assess variables associated with coagulopathy. The variables included were: GCS-Motor (GCS-M), receipt of a transfusion, and ISS. Higher GCS-M was associated with a 33% reduction in likelihood of coagulopathy (odds ratio [OR], 0.67; 95% confidence interval [CI; 0.48, 0.94] per unit increase in GCS-M score; p = 0.02), whereas receipt of a transfusion was associated with a ∼4 times increase in odds of coagulopathy (OR, 3.91; 95% CI [1.2, 12.5]; p = 0.02). The ISS was not independently associated with coagulopathy after adjusting for these two variables.

Radiographical predictors of coagulopathy

Of the imaging features associated with coagulopathy on single variable comparison, only basal cistern effacement was significantly associated in the multiple variable analysis, with a ∼3.5 times more likelihood of coagulopathy with effacement (OR, 3.34; 95% CI [1.08, 10.37]; p = 0.04; Table 2).

Table 2.

Multiple Variable Logistic Regression for Imaging Variables in Relation to Presence of Coagulopathy on Presentation

IVH, intraventricular hemorrhage.

Coagulopathy and mortality

Mortality was found to be significantly more common in coagulopathic patients (73% vs. 25%; p < 0.001). GCS-M is the single most established predictor of mortality after PBI. Therefore, and given the association observed between coagulopathy and GCS, we sought to investigate the contribution of coagulopathy when controlling for GCS-M and isolated brain injury. We performed a multiple variable logistic regression with GCS-M-isolated brain injury, and coagulopathy as covariates, finding the OR of GCS-M (OR = 0.27; 95% CI [0.15, 0.46]; p < 0.001) and isolated brain injury (OR = 11.03; 95% CI [1.41, 86.58]; p = 0.02) significant, but not coagulopathy (OR = 1.34; 95% CI [0.22, 8.10]; p = 0.75).

Coagulopathy at 24 h

Of the 40 patients who were coagulopathic on presentation, 25 survived to complete at least 24 hours in the intensive care unit. Coagulopathy was reassessed at 24 h. Of the 4 patients who were no longer coagulopathic at hospital day 1, 3 survived (75%) compared to 8 of the 21 patients whose coagulopathy remained (38%). This difference was not statistically significant (p = 0.29); however, the power for detecting an association is limited.

Discussion

In this article, we examined characteristics and outcomes of coagulopathy in civilian GSWH-PBI and report on a number of clinically relevant observations.. According to the definition used and after excluding patients dead on arrival, half of the patients were classified as coagulopathic. Coagulopathic patients had significantly lower total and motor GCS, and higher ISS scores, suggesting a potential coupling between the clinical severity of injury as reflected by the clinical assessment (GCS) and a serum marker (coagulopathy). Basal cistern effacement, bihemispheric, thalamic, brainstem, and posterior fossa involvement were all more frequent in the coagulopathic cohort. Mortality in the coagulopathic cohort was exceedingly high (73%).

Hypocoagulable hemostatic disturbance presents early after severe TBI; it has been reported in up to two thirds of patients and is associated with increased mortality.⁽¹²⁾ The primary drivers appear to be platelet dysfunction, endothelial activation, disturbed fibrinolysis, endogenous anticoagulation, and inflammation. The first few hours after TBI are characterized by coagulation cascade dysfunction that is likely secondary to unregulated thrombin production.¹³ Coagulation cascade dysfunction often occurs at a time when the risk of intracranial lesion progression is high, suggesting possible contribution to hemorrhagic expansion. Talving and colleagues were some of the first to separate penetrating from blunt TBI by coagulopathy, finding that patients with isolated PBI were significantly more coagulopathic by conventional coagulopathy tests.¹⁴ They also found that patients with isolated severe TBI and coagulopathy had a significantly higher mortality rate compared to those without coagulopathy (50% vs. 7%). Folkerson and colleagues had similar findings, with 60% of PBI patients experiencing a coagulopathy, and coagulopathy being an independent predictor of mortality.¹⁵

The high mortality in our coagulopathic cohort is closely related to the low GCS scores. This aligns with the covariates used in the Survival After Acute Civilian Penetrating Brain Injuries (SPIN) score, a logistic, regression-based, risk-stratification scale for estimating in-hospital and 6-month survival after civilian PBI.⁹ Specifically, in the more recent SPIN validation study, the three strongest predictors of survival were pupillary reactivity, GCS-M, and INR.¹⁶ The SPIN model did not consider radiographical variables, which is an important limitation given that early triage decisions and plans of care are heavily reliant on extent and location of injuries as identified on non-contrast HCT, often before the coagulation studies have resulted. In penetrating injuries, ballistic trajectory has been shown to play a significant role in the morbidity and mortality of these patients, particularly those sustaining bihemispheric, transventricular, or posterior fossa injuries.^17–20

In our cohort of patients, coagulopathy, low GCS, and ominous radiographical findings were strongly correlated, suggesting a particular phenotype of severe injury with very high mortality. Early recognition of a severely affected clinical trajectory facilitates triage decisions, offers an opportunity for immediate and aggressive resuscitation in order to preserve organ function and possibly alter outcomes, and informs the approach to family members regarding future care for the patient. The identification of a high-acuity, high-mortality group also allows for future targeted use of selected therapies.

Limitations

We report our experience from a significant number of civilian PBI patients; however, this is a single-center study where triage and treatment decisions were made case by case by a multi-disciplinary neurotrauma team. In the absence of a universally accepted definition of what constitutes coagulopathy after TBI, we used a definition according to our own institutional laboratory limits and in accordance with previous relevant literature. Viscoelastic assays, such as thromboelastography (TEG) and rotational thrombelastometry, are not specifically studied in TBI let alone PBI. TEG data could further elucidate the nature and prevalence of coagulopathy, impact stratification of patients, and guide resuscitation to minimize coagulopathy.²¹ Unfortunately, although a trend was observed, our current sample size was not sufficient to examine the effect of coagulopathy reversal on clinical outcomes, and this remains an open question. The higher rate of survival in patients who ceased to be coagulopathic merits further study to better elucidate the possible causal relationship or lack thereof, as well as the role of active reversal of coagulopathy in PBI specifically. Finally, time between the onset of injury and arrival to the emergency department was not available in our data set.

Conclusions

Similar to previous studies, we found high prevalence of coagulopathy in a large, single-center cohort of civilian PBI patients. Coagulopathy significantly correlated with a low GCS, radiographical indicators of extensive, severe brain injury, and exceedingly high mortality. The combination of these features combined can be used early to identify a severely affected phenotypic variant of penetrating brain injury with ominous prognosis. Future work should address whether specific interventions, such as coagulopathy reversal or early surgical intervention, could alter patient outcomes potentially impacting mortality, morbidity, or hospital LOS.

Funding Information

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Author Disclosure Statement

No competing financial interests exist.