Evaluation of Levetiracetam for Early PostTraumatic Seizure Prophylaxis: A Level II Trauma Center Experience

Background: Traumatic brain injury (TBI) can induce early or late post-traumatic seizures (PTS). While PTS incidence is low, prophylaxis is used despite a lack of consensus on agent or duration. Levetiracetam (LEV) for early PTS prophylaxis is preferred due to its safety and ecacy. The purpose of this study was to evaluate LEV for early PTS prophylaxis. Methods: A single-center, retrospective chart review of TBI patients > 18 years who received LEV for early PTS prophylaxis between August 2018 - July 2019. The primary outcome was LEV duration. Secondary outcomes were incidence of seizure, intensive care unit (ICU) and hospital length of stay (LOS). Results: Of the 137 included, mean age was 59±20 years and 69.3% were male. The mean admission GCS was 13±4 and 77.4% had mild TBI. Median LEV duration was 7 (IQR 4-10) days and 13.9% met recommended 7-day duration. Those prescribed LEV > 7 days had more than twice the median LEV duration than those prescribed ≤ 7 days (10.25 (8.5-15.5) vs 4 (1.5-4.5) days, p < 0.0001). EEG-conrmed PTS occurred in 2.2%, with an early PTS incidence of 0.73%. Median ICU and hospital LOS were 2 (IQR 17) and 7 (IQR 3-16) days, respectively. Conclusions: The incidence of PTS was low as most patients in our study had mild or moderate TBI. Early PTS prophylaxis with LEV for 7 days is appropriate, although the majority of patients did not meet the recommended duration. Efforts to standardize and implement PTS prophylaxis protocols are needed. subclinical seizures cannot be ruled out. Our results indicate that for patients without a prior history of seizures, LEV monotherapy is sucient. The decision to change to alternative AED or initiate a combination of AED should be made on a patient specic basis.


Background
Traumatic brain injury (TBI) can induce early (≤ 7 days post-TBI) or late (> 7 days post-TBI) acute symptomatic seizures. 1 Severity of TBI is commonly determined using the Glasgow Coma Score (GCS): mild (GCS [13][14][15], moderate (GCS 9-12), and severe (GCS < 9). 1,2 Post-traumatic seizures (PTS) occur more often in severe TBI, meanwhile rates of subclinical seizures may be as high as 25%. 3 The rate of post-traumatic epilepsy and recurrent seizures occurring more than 7 days post-TBI are signi cantly higher when compared to epilepsy rates in the general population. 4,5 Risk factors for PTS include severe TBI, chronic alcohol use, prolonged post-traumatic amnesia, linear/depressed skull fracture, younger age (< 65 years), subdural hematoma, intracerebral hematoma, penetrating head injury, and immediate seizures. [6][7][8] Prophylaxis is routinely administered to prevent development of PTS, status epilepticus, chronic epilepsy, herniation, and death. 4,7 While timely initiation reduces complications of early PTS (i.e., pneumonia, acute respiratory distress syndrome, acute renal failure, pulmonary embolism, and elevated intracranial pressure), limited durations to avoid adverse drug events should be considered. 5 A randomized, doubleblind study of 404 TBI patients comparing phenytoin (PHT) to placebo for one year, found that PHY reduces seizures during the rst week following injury. 9 Similar results were seen in a pooled analysis of 6 clinical trials of antiepileptic drugs (AED) (carbamazepine, phenobarbital, and PHT) of 1405 patients showing that AEDs reduced early, but not late PTS. 10 Therefore, The Brain Trauma Foundation and American Association of Neurologic Surgeons recommend a 7-day duration for early PTS prophylaxis. 3 While the duration of PTS prophylaxis is well de ned, the choice of AED remains controversial. The Brain Trauma Foundation and American Association of Neurologic Surgeons recommend PHT for early PTS prophylaxis when "the overall bene t is felt to outweigh the complications associated with such treatment". 3 Despite the inability to recommend LEV for PTS prophylaxis due to lack of su cient evidence it is widely used. 3,5,6,11 A prospective, multicenter study of 813 consecutive blunt TBI patients, found LEV was not associated with differences in occurrence of early or late PTS, ADRs, or mortality when compared to PHT. 12 An equivalence study of LEV and PHT in 32 severe TBI patients found no differences in Glasgow Outcome Scale at 3 and 6 months or seizures, although LEV was associated with an increased seizure tendency on electroencephalography (EEG). 13 Similar results were seen in a prospective, randomized, single-blinded study nding no difference in seizure occurrence or death between LEV and PHT. Furthermore, LEV resulted in less undesirable side effects and better long-term outcomes for survivors. 14 Thus, LEV is preferred for its well-established e cacy and lack of required therapeutic drug monitoring, tolerability, and relatively minimal drug-drug/drug-food interactions.
When considering LEV for early PTS prophylaxis, dosing 500-1000 mg every 12 hours intravenous or oral is initiated immediately following injury. Low-dose LEV 500 mg every 12 hours has been shown to be effective in early PTS prevention and may be considered. 15 While a 7-day duration is appropriate for early PTS prophylaxis, this practice is not widely agreed upon nor implemented. Due to a lack of consensus regarding de nitive PTS prophylaxis agents, dosing, and duration, it is important to establish institutionspeci c protocols. In this study, we evaluated compliance with these recommendations pertaining to the use of LEV for early PTS prophylaxis at a tertiary care, academic, level II trauma center.

Methods
This is a single-center, retrospective chart review of patients ≥ 18 years of age presenting with TBI who received LEV for early PTS prophylaxis between August 1, 2018, and July 31, 2019, at St. Joseph's University Medical Center in Paterson, NJ. Those with a known hypersensitivity to LEV, pre-admission use of chronic AED, history of seizure disorder, or lack of documented LEV duration were excluded. Patients who left against medical advice or died within 7 days of admission were also excluded. Patient demographics, home medications, past medical history, social history, reason for admission, mechanism of traumatic brain injury, admission unit and in-hospital mortality were evaluated. Medical records were also reviewed for AED (drug, dose, frequency, and duration) ordered during hospitalization and EEG. This study was reviewed and approved by the St. Joseph's Health institutional review board (EX2019-22). All methods were carried out in accordance with institutional guidelines and regulations.
The primary outcome was duration of LEV for PTS prophylaxis. Duration was de ned as the total number of days LEV was prescribed (loading and maintenance doses) and included in-hospital days plus the number of days prescribed at discharge. Duration prescribed at discharge was obtained from discharge medication reconciliation, physician discharge notes, and prescription ll histories. Secondary outcomes included incidence of EEG-con rmed seizure, hospital and intensive care unit (ICU) length of stay (LOS). Additional secondary outcomes evaluated were time to administration of rst LEV dose, administration of LEV loading dose, LEV maintenance dose (milligrams; mg) and frequency, and co-administration or transition to an alternative AED. All durations, time to initiation, and LOS were measured in calendar days. Subgroup analyses were performed to identify differences in LEV duration for TBI severity, ICU admission, ICU LOS (< 7 vs ≥ 7 days), and hospital LOS.
Demographic factors and clinical characteristics were summarized with counts (percentages) for categorical variables, means and standard deviations for normally distributed continuous variables, and medians and interquartile ranges for non-normally distributed variables. Median data was compared using the Mann-Whitney U test and nominal data was compared using the Chi-Square test. All analyses were performed with Microsoft Excel Tool Pak 2015. Signi cance was de ned as p-value ≤ 0.05.

Discussion
Incidence of early PTS varies with TBI severity, though prophylaxis in the rst 7 days may reduce the risk of post-TBI seizure disorders. In this study, most participants received LEV 500 mg twice daily for at least 7 days, starting on the day of admission. While this dosing scheme and time to initiation coincide with current recommendations, the total duration in our study did not. Factors contributing to longer duration of LEV therapy included ICU admission, longer hospital LOS, and higher TBI severity. While it is unclear why, in-hospital and total LEV duration were greater with longer hospital LOS. This may be due to disease severity or concern for seizures after day 7, although only 13.9% of the study population presented with severe TBI. While not directly assessed, other potential reasons for LEV duration > 7 days include prescriber specialty (critical care, neurosurgery, neurology), standardized AED duration of 30 days in computerized physician order entry system, and accidental renewals by ordering providers. Of note, twofths of the study population were discharged on LEV with a median total duration of 7.5 days.
The incidence of PTS in our study is signi cantly lower than prior reports. This may be due to a smaller percent of patients presenting with severe TBI or lack of detection. Only one-quarter of patients had EEG, thus the potential for subclinical seizures cannot be ruled out. Our results indicate that for patients without a prior history of seizures, LEV monotherapy is su cient. The decision to change to alternative AED or initiate a combination of AED should be made on a patient speci c basis.
There are several limitations to our study. First, this was a retrospective, single-center cohort study therefore all data collected was obtained through electronic medical records. As a result, differences in prescribing patterns among providers were not assessed. Second, despite documentation of LEV at discharge, we cannot exclude the possibility that patients may not have been continued on or may not have been adherent to therapy. Third, most patients included had mild TBI therefore our results may not represent PTS incidence for all TBI severities. Forth, in-hospital subclinical seizures were not evaluated in most patients. Use of EEG should be considered in patients with more severe TBI and risk factors for PTS. Lastly, our study did not evaluate for late PTS therefore the incidence may be under-reported.

Conclusions
Most patients in our study had mild or moderate TBI, therefore the incidence of PTS was low. Early PTS prophylaxis with LEV for 7 days is appropriate, although the majority of patients did not meet the recommended duration. Efforts to standardize and implement PTS prophylaxis protocols are needed. Furthermore, prospective, multi-center, randomized studies are needed to further evaluate LEV for PTS prophylaxis and the relation to TBI severity. Consent for publication: This study was granted exemption for informed consent. Need for informed consent was waived by the St. Joseph's Health ethics committee due to the retrospective nature of the study and less than minimal risk to study participants.

Abbreviations
Availability of data and materials: The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.
Competing interests: The authors declare that they have no competing interests.
Funding: The authors received no nancial support for the research, authorship, or publication of the study.
Authors' contributions: TA, SF, OA and JJ contributed to the study design. TA obtained all data necessary for the study. TA, SF, AW, JJ analyzed and interpreted the data. TA, SF, OA and JJ contributed to the writing of the manuscript. All authors read and approved the nal manuscript Median LEV duration measured in calendar days were compared among all patients, those with durations < 7, = 7 days or > 7 days. All patients had a median LEV duration of 7 days. Of those with LEV durations > 7 days, total median LEV duration was 10.3 days.

Figure 2
Median LEV duration measured in calendar days were compared among all patients, those with hospital LOS < 7 days, = 7 day and > 7 days. Of those who had hospital LOS > 7 days, LEV duration was approximately twice as long compared to those with hospital LOS < or = 7 days.

Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download.