Background
Seizure is one of the major complications of traumatic brain injury (TBI). Previous studies have indicated that post-traumatic seizures occur in approximately 16.7-28% of patients with blunt TBI [
1‐
5]. More importantly, nearly half of these were found to be non-convulsive that could only be detected by continuous electroencephalograph (EEG) monitoring [
2,
6]. Seizures after TBI may lead to elevated intracranial pressure, cerebral metabolic distress, additional brain damage, and eventually a worse outcome [
7,
8]. Compared to convulsive seizures, non-convulsive seizures (NCS) may have greater impact on patients’ outcome as they are not easy to detect.
The current guidelines for the management of severe TBI suggest prophylactic anti-seizure therapy to decrease the incidence of early posttraumatic seizures (within 7 days after injury) [
9]. Valproic acid (VPA) is a common anticonvulsant. It has been widely applied in neurointensive care units (NICU), although there is significant inter-individual variability in VPA plasma concentrations that requires monitoring for dose adjustment [
10]. Previous studies have indicated that the metabolism of VPA involves three major biotransformation pathways, which are conjugation with glucuronic acid, mitochrondrial β-oxidation, and CYP-catalyzed terminal desaturation and hydroxylation [
11]. Among these pathways, the most predominant one is glucuronidation conjugation, mainly mediated by UGT1A6 and UGT2B7 [
12‐
14]. CYP-catalyzed metabolism of VPA accounting for approximately 10-20% of the administrated dose [
15‐
17]. It has been demonstrated that the above genes encoding VPA metabolizing enzymes are highly polymorphic, with promoter single nucleotide polymorphisms (SNPs), which have significant impacts on drug metabolisms including VPA [
14,
18‐
21]. However, the reported contributions of genetic variations on VPA disposition were highly inconsistent, and their influences on early post-traumatic seizures and NCSs have not been explored. The aim of this study is to investigate the incidence of early post-traumatic seizures and the influence of genetic polymorphisms on VPA serum concentrations in a cohort of Chinese patients with severe TBI.
Discussion
Seizure is a common complication in all types of acute brain injuries. Previous studies have revealed that up to 28% of blunt TBI patients may have post-traumatic seizures [
1‐
5]. Continuous EEG monitoring reveals that approximately 10% of patients have nonconvulsive seizures, with 30% presenting in the first three days after injury [
2,
3,
35]. It has been recognized that seizure is associated with increased brain edema, excessive metabolic demand, elevated intracranial pressure, and worsened brain atrophy [
7,
8,
36]. Such pathologic process should be diagnosed and controlled as early as possible. In this study, even after excluding patients with history of seizures, still 83 (21.01%) of severe TBI patients had seizure during the first week after injury. More importantly, thirty out of 83 (36.14%) were identified as NCSs that could only be detected by continuous EEG monitoring. These findings were close to previous publications [
3‐
5]. In fact, we suspect that the actual frequency of NCSs could be higher than we identified. Some patients with subclinical seizures may not have been detected, as we only performed approximate 4 h continuous EEG monitoring for each patient due to limited equipment and conducting other examinations or therapies. The majority (67 of 83, 80.72%) cases had their seizure attacks within 48 h of monitoring. The peak incidence appeared in 12-24 h (25.30%) and 24-36 h (27.71%) after injury. Over 9% of patients did not have their first seizure until 48 h after injury, and 8.43% seizures were first detected at 72 h or later. This suggests that prolonged continues EEG monitoring for sever TBI patients is necessary.
Considering the high incidence, one may reasonably expect worse outcomes in patients with post-traumatic seizures. In this study, the impact of early post-traumatic seizures on clinical outcomes were reflected by prolonged ICULOS. This could be explained that patients with post-traumatic seizure requires longer ICU stay for EEG monitoring, extended periods for anti-seizure therapies, and other intensive care for complications. However, the HLOS showed no significant difference between groups. A possible explanation is that severe TBI patients were all transferred to rehabilitation centers for further therapies once their situations were stable. The mortalities and 3-month GOS also showed no significant difference between the two groups. This could due to the fact that the prognosis of sever TBI is influenced by many other factors, such as increased intracranial pressure, hospital acquired pneumonia, coagulation abnormalities, rehabilitation therapies, etc. To better determine the associations between seizures and patient outcomes, it would require further investigations with more sophisticated designs to screen each confounder.
The new generation of anti-epileptic medications such as levetiracetam, topiramate, lamotrigine appears to be increasing in use for seizure prophylaxis. However, to date there have been insufficient data to support their superiority or inferiority compared to other anticonvulsants [
9]. For patients with severe TBI and in coma, VPA for injection has the advantage of being easy to administrate and affordable. For example, VPA in East China is approximately $0.02 per 0.2g, $0.2 per 0.5g of sustained release tablets, and $12 per 0.4g of injection. Levetiracetam pills, in the contrast, are approximately $1.4 per 0.5g. Levetiracetam injections are not available in China. Topiramate and Lamotrigine pills are both around $0.7 per 50mg. Therefore, in many countries, particularly for patients in less developed areas, VPA it is still the first line antiepileptic drug. As mentioned earlier, the antiepileptic effects of VPA requires stable plasma concentrations, which are associated with genetic polymorphisms affecting VPA pharmacokinetics and pharmacodynamics. Compared to previous studies, there were no significant differences in the frequencies of the UGT1A6, UGT2B7, CYP2C9, CYP2C19 genotypes between our observations and other reported Chinese populations [
29,
37‐
39]. During the first two days after injury, when plasma concentrations were initially not available, all patients received similar dosages of VPA injection in our study. Under this circumstance, patients with UGT1A6 double heterozygosities at nucleotide positions 19T > G, 541A > G and 552A > C had lower adjusted plasma VPA concentrations compared to those with wild type or single heterozygosity. Accordingly, these patients received higher dose of VPA injection to achieve targeting plasma concentrations during later days of monitoring. Similarly, patients with wild type CYP2C9 (EMs) were associated with lower adjusted VPA plasma concentrations during the first 2 days after injury and higher VPA dosages at later time. Among the other studied SNPs, the adjusted plasma VPA concentrations and VPA dosages showed no significant difference. These results suggested that UGT1A6 double heterozygosities had higher enzyme activity and intrinsic clearance. The robust VPA terminal desaturation and hydroxylation in patients with CYP2C9 EMs may also be responsible for decreased adjusted plasma VPA concentrations.
Previously, there have been studies investigating SNPs in genes relating to VPA metabolism and disposition either alone or in combination. However, the results were not entirely consistent. For example, our findings on UGT1A6 SNPs were supported by two recent publications showing that homozygous carriers of the variant UGT1A6 19T > G, 541A > G and 552A > C allele have relatively low VPA concentrations [
39,
40]. However, similar results were not achieved in other studies [
41]. In addition, while there have been studies demonstrating the impact of CYP2C9 and/or CYP2C19 polymorphisms on VPA pharmacokinetics [
28,
42], other studies claimed no association between the CYP2C9 genotypes and VPA plasma concentrations or incidence of seizure [
29,
43,
44]. We suppose such discrepancy lies in the fact that VPA is metabolized by multiple UGTs and CYPs, and the disposition could be related to other factors such as sample size, race, and patient age. In our studies we found patients over 65 years old had higher adjusted VPA plasma concentrations and required less adjusted VPA dosages.
Our study has several limitations. Although we have recruited potential patients for a period of four years, the number of cases with certain genotypes, such as CYP2C9 PMs, were still limited, which led to statistically significant difference in neither VPA concentrations nor VPA dosages. Similarly, although it is identified that male, age over 65, and UGT1A6 double heterozygosities were associate with higher incidence of early post-traumatic seizures, the powers of the difference were less than 0.80. Therefore, further single- or multi-center investigations with larger sample size are required to increase the power and to confirm the difference. Another undeniable fact is that the pathogenesis of seizure and its influence factors are not fully understood. Although VPA is a broad-spectrum antiepileptic drug, reaching the effective blood concentration does not equal to avoidance of seizure attack. New antiepileptic medications have been more popular, yet their pharmacokinetics and relationship to SNPs are mostly complicated and remain unclear. In our study, early post-traumatic seizures were treated using first and second line anti-seizure drugs [
32]. We did not change the therapeutic regime after genetic polymorphism analyzing. This is partly due to genetic polymorphism analyzing requires several days in our institution, too long for a window of 7 days of early post-traumatic seizures. However, this would be an attractive issue to address in the future study. In sum, the above questions reinforce the need for further investigations and to promote precision medicine in the area of anti-seizure therapy.
Acknowledgements
We thank Dr. Zhuoying Du and Dr. Zhiqi Li for supporting this study.