Journal of Trauma and Acute Care Surgery, Publish Ahead of Print DOI: 10.1097/TA.0000000000000959
Early Propranolol after Traumatic Brain Injury is Associated with Lower
Mortality Ara Ko MD, Megan Y Harada BA, Galinos Barmparas MD, Gretchen M Thomsen PhD, Rodrigo F Alban MD, Matthew Bloom MD, Rex Chung MD, Nicolas Melo MD, Daniel R Margulies MD, Eric J Ley MD
Department of Surgery, Division of Trauma and Critical Care Cedars-Sinai Medical Center, Los Angeles, CA
TE
D
1
Submitted: July 31, 2015, Revised: October 21, 2015, Accepted: November 9, 2015.
EP
Author e-mail addresses: AK: [email protected]; MYH: [email protected]; GB: [email protected]; GMT: [email protected]; RFA: [email protected]; MB: [email protected]; RC: [email protected]; NM: [email protected]; DRM: [email protected]; EJL: [email protected]
The authors have no conflicts of interest to report and have received no financial support in relation to this manuscript.
C
This study was presented as a pôster at the 74th anual meeting of the American Association for the Surgery of Trauma, September 9-12, 2015, in Las Vegas, Nevada.
A
C
Correspondence to: Eric J. Ley, MD Cedars-Sinai Medical Center Department of Surgery 8700 Beverly Blvd, Suite 8215N Los Angeles, CA 90048 Tel: (310) 423-5874 Fax: (310) 423-0139 E-mail: [email protected]
Early Propranolol after Traumatic Brain Injury
1
Copyright © 2016 Wolters Kluwer Health, Inc. All rights reserved.
Abstract
Background: Beta adrenergic receptor blockers (BB) administered after trauma blunt the cascade of immune and inflammatory changes associated with injury. BBs are associated with improved outcomes after traumatic brain injury (TBI). Propranolol may be an ideal BB due to its nonselective inhibition and ability to cross the blood brain barrier. We determined if early
D
administration of propranolol after TBI is associated with lower mortality. Methods: All adults (age ≥ 18) with moderate to severe TBI (head AIS = 3-5) requiring
TE
intensive care unit (ICU) admission at a Level I trauma center from January 1, 2013 to May 31, 2015 were prospectively entered into a database. Administration of early propranolol was dosed within 24 hours of admission at 1 mg IV every 6 hours. Patients who received early propranolol after TBI (EPAT) were compared to those who did not (non-EPAT). Data including
EP
demographics, hospital length of stay (LOS), ICU LOS, and mortality were collected.
Results: Over 29 months, 440 patients with moderate to severe TBI met inclusion criteria. Early propranolol was administered to 25% (109/440) of patients. The EPAT cohort was younger (49.6 vs. 60.4 years, p < 0.001), had lower GCS (11.7 vs. 12.4, p = 0.003), lower head AIS (3.6 vs. 3.9,
C
p = 0.001), higher admission HR (95.8 vs. 88.4 bpm, p = 0.002), and required more days on the ventilator (5.9 vs. 2.6 days, p < 0.001). Similarities were noted in gender, ISS, admission SBP, hospital LOS, ICU LOS and mortality rate. Multivariate regression showed EPAT was
C
independently associated with lower mortality (AOR 0.25, p = 0.012).
A
Conclusions: After adjusting for predictors of mortality, early administration of propranolol after TBI was associated with improved survival. Future studies are needed to identify additional benefits as well as optimal dosing regimens.
Level of Evidence: Therapeutic Level IV
Keywords: Traumatic brain injury; Propranolol; Beta-blockers; Mortality
2
Copyright © 2016 Wolters Kluwer Health, Inc. All rights reserved.
Background Traumatic brain injury (TBI) accounts for up to 30% of all injury-related deaths1 and significant morbidity, contributing to substantial long-term disability and economic burden on society.2 Guidelines for management of TBI emphasize prevention of secondary injury from hypoxemia and hypoperfusion3 as well as from damage due to excitatory neurotransmitters,
D
catecholamine surge, the inflammatory cascade, free radicals and calcium mediated injury.4 The post-head injury catecholamine surge and the associated intracranial and extracranial sequelae
TE
significantly impact mortality and morbidity in TBI patients.5,6
Beta-adrenergic receptor blockers may blunt the cascade of immune and inflammatory changes associated with injury. A number of retrospective database analyses demonstrate that
EP
beta-blocker exposure in the setting of TBI is associated with reduced mortality.7–11 Despite optimistic findings, early beta blockade use is not routine. Propranolol may be the ideal agent due to its nonselective inhibition and its lipophilic properties enabling it to penetrate the blood brain barrier.12 Small observational studies suggest that early low dose propranolol can be safely
C
administered and might improve outcomes after TBI13 or severe trauma.14 As most studies are
C
retrospective analyses, data evaluating standardized dosage and timing of beta-blocker administration is lacking.
A
The primary objective of this study was to prospectively investigate whether empiric
early administration of low dose propranolol in patients with moderate to severe TBI impacts outcomes. We hypothesize that administration of a standard low dose of propranolol within 24 hours after TBI is associated with lower mortality.
3
Copyright © 2016 Wolters Kluwer Health, Inc. All rights reserved.
Methods All TBI patients requiring ICU admission from January 1, 2013 to May 31, 2015 at Cedars-Sinai Medical Center, a Level I trauma center in Los Angeles, California were entered into an online database. The electronic medical record was reviewed to confirm TBI on imaging studies (CT head or brain MRI), extract patient demographics, and prospectively collect
D
outcomes data. Data included age, gender, admission Glasgow Coma Scale (GCS), admission systolic blood pressure (SBP), admission heart rate (HR), regional Abbreviated Injury Scale
TE
(AIS) score, Injury Severity Score (ISS), and type of brain injury as characterized by imaging studies. Outcomes data included types of neurosurgical interventions, hospital length of stay (LOS), intensive care unit (ICU) LOS, number of days requiring mechanical ventilation, and in-
EP
hospital mortality.
All adult trauma patients (age ≥ 18) admitted with TBI to the ICU were initially included in the study and followed prospectively. To exclude those with mild TBI (head AIS < 3) and non-survivable TBI (head AIS = 6) the trauma registry was utilized. Inclusion in the study
C
required ICU admission. Accordingly, those who expired in the ED were excluded. All patients
C
admitted to the ICU were eligible for early propranolol, even those who expired within 24 hours of ICU admission. Those whose data required additional privacy access authorization were
A
excluded as well. As administration of early propranolol is not routine, it was ordered at the discretion of the on-call trauma or neurosurgery attending. Propranolol may have been withheld if patients did not meet heart rate or blood pressure parameters, were away for tests or procedures, or if providers failed to remember to order early propranolol. Time from admission to dosing of propranolol was noted and only those who received the first dose of propranolol within 24 hours were included in the propranolol cohort. Those patients (5) who received
4
Copyright © 2016 Wolters Kluwer Health, Inc. All rights reserved.
propranolol after 24 hours were excluded from the study. Patients were administered low dose intravenous (IV) propranolol of 1mg every 6 hours, with hold parameters for HR < 60 beats per minute and SBP < 100 mmHg.13 No dose adjustments were made based upon tachycardia.14 When transferred to the floor, the dosage of propranolol was changed to the per os (PO) equivalent, at a dosage of 40mg twice daily. Patients who received early propranolol after TBI
D
(EPAT) were then compared to those who did not (non-EPAT). All patients were followed from ED admission until hospital discharge or death. Of note, all patients with TBI were cared for per
TE
standard of care according to the Brain Trauma Foundation guidelines. The same group of
trauma surgeons and neurosurgeons cared for all the patients in the study cohort. Most patients were admitted under the Trauma team, but there were instances in which admission was to the
EP
Neurosurgery team. Regardless of who was the primary care provider, patient care was a collaborative effort between the teams.
Data were collected onto an electronic spreadsheet and Statistical Package for the Social Sciences (Version 22, SPSS Inc., Chicago, IL) was used for statistical analyses. Descriptive
C
statistics were reported as raw percentages or means and standard deviations. A Student’s t-test or Mann-Whitney test was used when appropriate to compare means for parametric or non-
C
parametric data respectively. A chi-square test or Fisher’s exact test was performed for
A
comparison of categorical variables. P < 0.05 was considered statistically significant. Variables were dichotomized for regression analysis including age ≥ 65, GCS ≤ 8, SBP < 90mmHg, and ISS ≥ 25, consistent with methods of existing literature. After identifying possible confounding variables, such as those with p-values less than 0.2 on bivariate analysis or risk factors that were considered theoretical confounders, multiple regression analysis was performed to determine
5
Copyright © 2016 Wolters Kluwer Health, Inc. All rights reserved.
independent risk factors associated with in-hospital mortality. This study was approved by the Cedars Sinai Medical Center Institutional Review Board.
Results Over the 29-month study period, 535 patients presented to Cedars Sinai Medical Center
D
and required ICU admission for TBI. A total of 481 were confirmed to have TBI on imaging. Twenty seven patients were excluded (5 patients had age < 18 years, 5 patients had propranolol
TE
administered >24 hours after admission, 17 patients had charts that were locked due to privacy reasons requiring an extra layer of security and precluding prospective data collection, and 14 patients had a head AIS < 3 or = 6). After exclusions, 440 patients remained in the study
EP
population (Figure 1). Early propranolol (EPAT) was administered to 25% (109/440) of patients. The average time from admission to administration of propranolol was 5.5 ± 3.8 hours. On average, patients received 15 ± 20.0 doses during the hospital stay. In terms of patient characteristics, the EPAT group was younger compared to the non-
C
EPAT group (49.6 vs. 60.4 years, p < 0.001 respectively). Those who received propranolol had a
C
lower GCS (11.7 vs. 12.4, p < 0.001), lower head AIS (3.6 vs. 3.9, p = 0.001), and higher admission HR (95.8 vs. 88.4, p = 0.002). Similarities between the groups were noted in gender,
A
ISS and admission SBP (Table 1). Compared to the non-EPAT cohort, the EPAT group had significantly more subarachnoid (60.6% vs. 43.8%, p = 0.002) and intraparenchymal hemorrhage (40.4% vs. 21.8%, p < 0.001), but less subdural hemorrhage (52.3% vs. 64.4%, p = 0.025) (Table 2). Those in the EPAT cohort were more likely than the non-EPAT group to have an intracranial pressure (ICP) monitor or bolt placed (14.7% vs. 6.9%, p = 0.014). The use of external ventricular drains (EVD) were similar in both groups (7.3% EPAT vs. 6.9% non-EPAT, p =
6
Copyright © 2016 Wolters Kluwer Health, Inc. All rights reserved.
0.890). There was a trend toward requiring less craniotomies or craniectomies in the EPAT group (11% vs. 15.4%, p = 0.255) (Table 2). Analysis of outcomes demonstrated that EPAT patients tended to have longer hospital LOS (12.5 vs. 9.2 days, p = 0.165) and ICU LOS (6.6 vs. 4.6 days, p = 0.056), but these differences were not statistically significant (Table 3). EPAT patients did, however, have
D
significantly more days on the ventilator (5.9 vs. 2.6 days, p < 0.001). The lower mortality rate observed in EPAT did not reach statistical significance (6.4% vs. 13.0%, p = 0.061).
TE
Multivariate regression was performed for further analysis after identifying potential
confounders. As GCS ≤ 8 correlates with severe brain injury, this value was chosen to control for such head injury that could independently predict mortality. A Receiver Operating Characteristic
EP
(ROC) curve for ISS ≥ 16 and ISS ≥ 25 was created, which indicated that a regression adjusting for an ISS ≥ 25 yielded a higher area under the curve (AUC). The AUC was 0.91 (CI 0.86 to 0.97, p
Early Propranolol after Traumatic Brain Injury is Associated with Lower
Mortality Ara Ko MD, Megan Y Harada BA, Galinos Barmparas MD, Gretchen M Thomsen PhD, Rodrigo F Alban MD, Matthew Bloom MD, Rex Chung MD, Nicolas Melo MD, Daniel R Margulies MD, Eric J Ley MD
Department of Surgery, Division of Trauma and Critical Care Cedars-Sinai Medical Center, Los Angeles, CA
TE
D
1
Submitted: July 31, 2015, Revised: October 21, 2015, Accepted: November 9, 2015.
EP
Author e-mail addresses: AK: [email protected]; MYH: [email protected]; GB: [email protected]; GMT: [email protected]; RFA: [email protected]; MB: [email protected]; RC: [email protected]; NM: [email protected]; DRM: [email protected]; EJL: [email protected]
The authors have no conflicts of interest to report and have received no financial support in relation to this manuscript.
C
This study was presented as a pôster at the 74th anual meeting of the American Association for the Surgery of Trauma, September 9-12, 2015, in Las Vegas, Nevada.
A
C
Correspondence to: Eric J. Ley, MD Cedars-Sinai Medical Center Department of Surgery 8700 Beverly Blvd, Suite 8215N Los Angeles, CA 90048 Tel: (310) 423-5874 Fax: (310) 423-0139 E-mail: [email protected]
Early Propranolol after Traumatic Brain Injury
1
Copyright © 2016 Wolters Kluwer Health, Inc. All rights reserved.
Abstract
Background: Beta adrenergic receptor blockers (BB) administered after trauma blunt the cascade of immune and inflammatory changes associated with injury. BBs are associated with improved outcomes after traumatic brain injury (TBI). Propranolol may be an ideal BB due to its nonselective inhibition and ability to cross the blood brain barrier. We determined if early
D
administration of propranolol after TBI is associated with lower mortality. Methods: All adults (age ≥ 18) with moderate to severe TBI (head AIS = 3-5) requiring
TE
intensive care unit (ICU) admission at a Level I trauma center from January 1, 2013 to May 31, 2015 were prospectively entered into a database. Administration of early propranolol was dosed within 24 hours of admission at 1 mg IV every 6 hours. Patients who received early propranolol after TBI (EPAT) were compared to those who did not (non-EPAT). Data including
EP
demographics, hospital length of stay (LOS), ICU LOS, and mortality were collected.
Results: Over 29 months, 440 patients with moderate to severe TBI met inclusion criteria. Early propranolol was administered to 25% (109/440) of patients. The EPAT cohort was younger (49.6 vs. 60.4 years, p < 0.001), had lower GCS (11.7 vs. 12.4, p = 0.003), lower head AIS (3.6 vs. 3.9,
C
p = 0.001), higher admission HR (95.8 vs. 88.4 bpm, p = 0.002), and required more days on the ventilator (5.9 vs. 2.6 days, p < 0.001). Similarities were noted in gender, ISS, admission SBP, hospital LOS, ICU LOS and mortality rate. Multivariate regression showed EPAT was
C
independently associated with lower mortality (AOR 0.25, p = 0.012).
A
Conclusions: After adjusting for predictors of mortality, early administration of propranolol after TBI was associated with improved survival. Future studies are needed to identify additional benefits as well as optimal dosing regimens.
Level of Evidence: Therapeutic Level IV
Keywords: Traumatic brain injury; Propranolol; Beta-blockers; Mortality
2
Copyright © 2016 Wolters Kluwer Health, Inc. All rights reserved.
Background Traumatic brain injury (TBI) accounts for up to 30% of all injury-related deaths1 and significant morbidity, contributing to substantial long-term disability and economic burden on society.2 Guidelines for management of TBI emphasize prevention of secondary injury from hypoxemia and hypoperfusion3 as well as from damage due to excitatory neurotransmitters,
D
catecholamine surge, the inflammatory cascade, free radicals and calcium mediated injury.4 The post-head injury catecholamine surge and the associated intracranial and extracranial sequelae
TE
significantly impact mortality and morbidity in TBI patients.5,6
Beta-adrenergic receptor blockers may blunt the cascade of immune and inflammatory changes associated with injury. A number of retrospective database analyses demonstrate that
EP
beta-blocker exposure in the setting of TBI is associated with reduced mortality.7–11 Despite optimistic findings, early beta blockade use is not routine. Propranolol may be the ideal agent due to its nonselective inhibition and its lipophilic properties enabling it to penetrate the blood brain barrier.12 Small observational studies suggest that early low dose propranolol can be safely
C
administered and might improve outcomes after TBI13 or severe trauma.14 As most studies are
C
retrospective analyses, data evaluating standardized dosage and timing of beta-blocker administration is lacking.
A
The primary objective of this study was to prospectively investigate whether empiric
early administration of low dose propranolol in patients with moderate to severe TBI impacts outcomes. We hypothesize that administration of a standard low dose of propranolol within 24 hours after TBI is associated with lower mortality.
3
Copyright © 2016 Wolters Kluwer Health, Inc. All rights reserved.
Methods All TBI patients requiring ICU admission from January 1, 2013 to May 31, 2015 at Cedars-Sinai Medical Center, a Level I trauma center in Los Angeles, California were entered into an online database. The electronic medical record was reviewed to confirm TBI on imaging studies (CT head or brain MRI), extract patient demographics, and prospectively collect
D
outcomes data. Data included age, gender, admission Glasgow Coma Scale (GCS), admission systolic blood pressure (SBP), admission heart rate (HR), regional Abbreviated Injury Scale
TE
(AIS) score, Injury Severity Score (ISS), and type of brain injury as characterized by imaging studies. Outcomes data included types of neurosurgical interventions, hospital length of stay (LOS), intensive care unit (ICU) LOS, number of days requiring mechanical ventilation, and in-
EP
hospital mortality.
All adult trauma patients (age ≥ 18) admitted with TBI to the ICU were initially included in the study and followed prospectively. To exclude those with mild TBI (head AIS < 3) and non-survivable TBI (head AIS = 6) the trauma registry was utilized. Inclusion in the study
C
required ICU admission. Accordingly, those who expired in the ED were excluded. All patients
C
admitted to the ICU were eligible for early propranolol, even those who expired within 24 hours of ICU admission. Those whose data required additional privacy access authorization were
A
excluded as well. As administration of early propranolol is not routine, it was ordered at the discretion of the on-call trauma or neurosurgery attending. Propranolol may have been withheld if patients did not meet heart rate or blood pressure parameters, were away for tests or procedures, or if providers failed to remember to order early propranolol. Time from admission to dosing of propranolol was noted and only those who received the first dose of propranolol within 24 hours were included in the propranolol cohort. Those patients (5) who received
4
Copyright © 2016 Wolters Kluwer Health, Inc. All rights reserved.
propranolol after 24 hours were excluded from the study. Patients were administered low dose intravenous (IV) propranolol of 1mg every 6 hours, with hold parameters for HR < 60 beats per minute and SBP < 100 mmHg.13 No dose adjustments were made based upon tachycardia.14 When transferred to the floor, the dosage of propranolol was changed to the per os (PO) equivalent, at a dosage of 40mg twice daily. Patients who received early propranolol after TBI
D
(EPAT) were then compared to those who did not (non-EPAT). All patients were followed from ED admission until hospital discharge or death. Of note, all patients with TBI were cared for per
TE
standard of care according to the Brain Trauma Foundation guidelines. The same group of
trauma surgeons and neurosurgeons cared for all the patients in the study cohort. Most patients were admitted under the Trauma team, but there were instances in which admission was to the
EP
Neurosurgery team. Regardless of who was the primary care provider, patient care was a collaborative effort between the teams.
Data were collected onto an electronic spreadsheet and Statistical Package for the Social Sciences (Version 22, SPSS Inc., Chicago, IL) was used for statistical analyses. Descriptive
C
statistics were reported as raw percentages or means and standard deviations. A Student’s t-test or Mann-Whitney test was used when appropriate to compare means for parametric or non-
C
parametric data respectively. A chi-square test or Fisher’s exact test was performed for
A
comparison of categorical variables. P < 0.05 was considered statistically significant. Variables were dichotomized for regression analysis including age ≥ 65, GCS ≤ 8, SBP < 90mmHg, and ISS ≥ 25, consistent with methods of existing literature. After identifying possible confounding variables, such as those with p-values less than 0.2 on bivariate analysis or risk factors that were considered theoretical confounders, multiple regression analysis was performed to determine
5
Copyright © 2016 Wolters Kluwer Health, Inc. All rights reserved.
independent risk factors associated with in-hospital mortality. This study was approved by the Cedars Sinai Medical Center Institutional Review Board.
Results Over the 29-month study period, 535 patients presented to Cedars Sinai Medical Center
D
and required ICU admission for TBI. A total of 481 were confirmed to have TBI on imaging. Twenty seven patients were excluded (5 patients had age < 18 years, 5 patients had propranolol
TE
administered >24 hours after admission, 17 patients had charts that were locked due to privacy reasons requiring an extra layer of security and precluding prospective data collection, and 14 patients had a head AIS < 3 or = 6). After exclusions, 440 patients remained in the study
EP
population (Figure 1). Early propranolol (EPAT) was administered to 25% (109/440) of patients. The average time from admission to administration of propranolol was 5.5 ± 3.8 hours. On average, patients received 15 ± 20.0 doses during the hospital stay. In terms of patient characteristics, the EPAT group was younger compared to the non-
C
EPAT group (49.6 vs. 60.4 years, p < 0.001 respectively). Those who received propranolol had a
C
lower GCS (11.7 vs. 12.4, p < 0.001), lower head AIS (3.6 vs. 3.9, p = 0.001), and higher admission HR (95.8 vs. 88.4, p = 0.002). Similarities between the groups were noted in gender,
A
ISS and admission SBP (Table 1). Compared to the non-EPAT cohort, the EPAT group had significantly more subarachnoid (60.6% vs. 43.8%, p = 0.002) and intraparenchymal hemorrhage (40.4% vs. 21.8%, p < 0.001), but less subdural hemorrhage (52.3% vs. 64.4%, p = 0.025) (Table 2). Those in the EPAT cohort were more likely than the non-EPAT group to have an intracranial pressure (ICP) monitor or bolt placed (14.7% vs. 6.9%, p = 0.014). The use of external ventricular drains (EVD) were similar in both groups (7.3% EPAT vs. 6.9% non-EPAT, p =
6
Copyright © 2016 Wolters Kluwer Health, Inc. All rights reserved.
0.890). There was a trend toward requiring less craniotomies or craniectomies in the EPAT group (11% vs. 15.4%, p = 0.255) (Table 2). Analysis of outcomes demonstrated that EPAT patients tended to have longer hospital LOS (12.5 vs. 9.2 days, p = 0.165) and ICU LOS (6.6 vs. 4.6 days, p = 0.056), but these differences were not statistically significant (Table 3). EPAT patients did, however, have
D
significantly more days on the ventilator (5.9 vs. 2.6 days, p < 0.001). The lower mortality rate observed in EPAT did not reach statistical significance (6.4% vs. 13.0%, p = 0.061).
TE
Multivariate regression was performed for further analysis after identifying potential
confounders. As GCS ≤ 8 correlates with severe brain injury, this value was chosen to control for such head injury that could independently predict mortality. A Receiver Operating Characteristic
EP
(ROC) curve for ISS ≥ 16 and ISS ≥ 25 was created, which indicated that a regression adjusting for an ISS ≥ 25 yielded a higher area under the curve (AUC). The AUC was 0.91 (CI 0.86 to 0.97, p
