Post-traumatic epilepsy in Cameroon: a retrospective study in a referral hospital

Post-traumatic epilepsy is defined as the onset of at least one seizure beyond the first week following a traumatic brain injury (TBI). High prevalence of TBI in our setting may contribute to the burden of epilepsy in adult population. This is a retrospective review of medical records of patients admitted from January 1st, 2010 to December 31st, 2019) at Douala General Hospital. We included patients aged ≥ 18 years with seizure onset at least one week after TBI. Incomplete files and previously known epilepsy were excluded. Data on sociodemography, clinical and para-clinical features, treatment and outcome were analysed using R software version 36.2. We finally included 65 patients with post-traumatic epilepsy among 993 medical records of epilepsy. The mean age was 35.1 ± 12.6 years, with 64.6% of male. Road traffic accident was the main aetiology of brain trauma (78.5%), resulting in haemorrhagic contusions (21.5%), sub-dural haematoma (15.4%), and diffuse axonal lesions (15.4%) mainly. Seizure onset was within 2 years post-trauma in 73.8% of cases. Generalized tonic–clonic seizures were the commonest seizure’s type. Electroencephalogram was abnormal in 81%, including 47% of focal discharges. Antiepileptic drugs were mainly sodium valproate, carbamazepine, and phenobarbital. Seizure freedom was obtained in 67.7% of cases. Post-traumatic epilepsy is a heterogeneous, frequent and often disabling complication of traumatic brain injury. Road traffic accident is the main cause of brain trauma. It affects a young and active population. About half of cases presented GTCS. With antiepileptic drugs, more than two-thirds of patients become seizure-free.


Background
The International League Against Epilepsy (ILAE) defines epilepsy as a chronic neurological disease characterized by the occurrence of at least two unprovoked or reflex epileptic seizures (ES) spaced more than 24 h apart, or one unprovoked seizure with a high probability (≥ 60%) of occurrence of another ES in the next ten years or the existence of an epileptic syndrome [1]. Traumatic brain injury (TBI) is a physical shock on the skull, which may lead to lesions of the brain [2]. Post-traumatic epilepsy (PTE) is defined as the occurrence of at least one ES beyond the first week following a TBI. Early ES (immediate < 24 h or delayed < 1 week) are acute symptomatic events corresponding to the brain's response to the physical effects of TBI [3].
In Cameroon, TBI mainly affects young and active people [11]. Intracranial lesions such as sub-dural haematoma (SDH), epidural haematoma (EDH), diffuse axonal injury (DAI), contusions, craniocerebral wounds, increase the burden of PTE in adult population. Patients with PTE suffer double the consequence of physical trauma and psychological trauma associated with significant stigma associated with epilepsy in Cameroon. In this study, we aimed to determine characteristics of post-traumatic epilepsy (PTE) in a referral hospital of Cameroon.

Study design
We retrospectively reviewed medical records of patients admitted or consulted from January 1st, 2010 to December 31st, 2019 for epilepsy in the neurology and neurosurgery units of the Douala General hospital (DGH). We included all adult cases (≥ 18 years old) of post-traumatic epilepsy defined as onset of epileptic seizure at least one week after a traumatic brain injury (TBI). Incomplete file and previously known patients with epilepsy were excluded from this study. A file was considered incomplete when missing at least two of the following data (TBI mechanism, time to onset of the first seizure, neuroimaging of the trauma, type of ES).

Data collection
We carefully reviewed patients' data in archives, outpatient registers, and through call. Data collected included: socio-demographic, comorbidities and risk factors of seizures, clinical description of seizures, neuroimaging and EEG, treatment and outcome.

Statistical analysis
We carried out a descriptive analysis of the variables using R software version 36.2. Categorical variables were expressed as frequencies with percentage. Continuous variables were presented as mean with standard deviation (SD). The Chi-square test was used to look for associations between our variables. The results were considered significant for a p-value < 0.05.

Results
Among the 993 medical records of patients with epilepsy, we found 81 cases with post-traumatic epilepsy (PTE). We finally included 65 patients (we excluded 5 incomplete files and 11 patients aged less than 18 years old). Male represented 64.6% of this population, the mean age (SD) was 35.1 ± 12.6 years old, and 18 to 28 years was the main age group (Table 1). The history of traumatic brain injury (TBI) was relevant for initial loss of consciousness (70%), early seizures (27.7%). In this study, road traffic accident (RTA) was the TBI aetiology in 78.5% of cases. TBI was classified as severe (GCS < 8) in 61.5% of patients ( Table 2). The frequency of seizures onset decreased over time: 43% during the first year to 7.7% between 6 and 10 years after the trauma. Generalized tonic-clonic seizures (GTCS) were the commonest seizure's type. 50.8% of patients had a biannual seizure frequency ( Table 3). The brain imaging was performed either at the time of TBI (n = 51) or after the diagnosis of PTE (n = 14). The commonest lesions at the time of TBI were: haemorrhagic contusion (21.5%), diffuse axonal injury (15.4%), and sub-dural haematoma (15.4%); see Table 4. EEG was done in all cases resulting in abnormal findings in 81.5%. Anti-seizure medications (ASMs) were prescribed in all patients. Sodium valproate and carbamazepine were the frequent ASMs used. 13.8% of cases received a combination of two ASMs. No patient underwent surgical procedure for epilepsy management. Seizure frequencies drop after initiation of ASMs, with 67.7% of seizure freedom ( Table 5). The mean duration time for seizure freedom was 2.5 ± 1.2 years. No patient died in this population.

Discussion
This retrospective study conducted in a referral hospital in Cameroon, aimed to describe different features of post-traumatic epilepsy (PTE) for the first time. In about 6% of patients with epilepsy, traumatic brain injury (TBI) is thought to be the cause [12]. In this study, patients with PTE represented 8.1% of the patients followed up for epilepsy in our institution. The male predominance found in this study is consistent with the findings of Ogunrin [13,14]. This concordance of results could be explained by the fact that men are more active and carry out more risky activities exposing them to TBI. Two patients out of five were within the 18 to 28 age group. These results are similar to those found in China by Zhao et al. with a mean age of 40.07 years and the most represented age group varying between 20 and 29 years [14]. Similar findings were reported by Ogunrin et al. found also a mean age of 38.6 years this agreement of the results is explained by the fact that the young population is more active and consequently more at risk of undergoing TBI [13]. In Cameroon, young people are mainly involved in motorcycle taxi with very few regulations. This contributes to the burden of trauma in general, and TBI in particular.
In this study, road traffic accidents (RTA) were the TBI mechanism in more than three-quarters of cases. This is consistent with other studies which reported RTA as the main mechanism of TBI in different countries [15][16][17]. TBI was classified as severe in close to two-thirds of cases. Severe TBI has been found to be associated with PTE [18]. Similar findings were also reported by other studies [15,19]. The time to the first seizure, which corresponds to the duration of epileptogenesis, varies depending on the individual. In this study, more than two patients out of five underwent their first seizures within one year following the TBI, and close to three-quarters of cases within the first 2 years. These results are similar to other studies who reported 80% of patients with first seizure during the first 2 years following the TBI [13,15,20,21].
In this study, close to half of cases presented a generalized tonic-clonic seizures (GTCS). Other seizures types included focal seizures, and focal seizures with secondary generalization. Similar results were reported by Haltiner et al. with 51% of GCTS, 33% of focal seizures, and 17% focal seizures with secondary generalization [22]. Englander et al. found a higher frequency of GTCS [23]. However, our findings were different from the results of da Silva et al. who described 54.6% of focal seizure, and 37.6% of GTCS in a military population [24]. The focal of seizure could be overshadowed by the spectacular presentation of GTCS. Families of patients tend to report secondary generalization more frequently. In addition, the mechanisms of TBI in the military population are different from RTA.
Neuroimaging (brain CT scan or MRI) allows after a TBI to visualize the type and extent of brain damage. After epileptogenesis, neuroimaging contributes to rule out other causes, and to demonstrate TBI sequelae. In this study, the commonest traumatic lesions were brain contusions, diffuse axonal injuries (DAI), and epidural haematoma (EDH). Studies conducted in Africa reported mainly depressed skull fracture, and EDH [13,15]. In these latest studies depressed skull fracture were associated with contusions. In any type of epilepsy, EEG is very useful not only for diagnosis, but also the monitoring of patients. This was demonstrated in this study where we found abnormal EEG findings in more than four patients out of five (81.5% of cases). These EEG abnormalities were: generalized spikes-and-waves discharges (41.6%), frontal spikes-and-waves discharges (30.8%), temporal spikes-and-waves discharges (6.2%), and focal slow wave activity (3%).
The choice of ASMs in PTE depends on the availability, seizure type, and tolerance and drugs interactions [25,26]. In a resource limited setting like Cameroon, the panel of ASMs available is narrow. Based on their availability, commonest ASMs are sodium valproate, carbamazepine, phenobarbital, and diazepam. More than half of patients were on monotherapy with either sodium valproate, carbamazepine, or phenobarbital. For Ogunrin et al. patients received mainly phenytoin, carbamazepine, and sodium valproate as molecules [13]. Right after a moderate to severe TBI, patients can be placed on ASMs in order to prevent the progression of epileptogenesis to late seizures and epilepsy [27]. Phenytoin is the ASMs with most evidence in seizure prevention [28]. In this study, two patients received seizure prevention medication with phenobarbital. phenytoin is almost not available in our setting. The use of ASMs in prophylaxis is still controversial. According to Pingue et al., the use of ASMs was associated with a poor rehabilitation outcome, independently of the onset of epilepsy during treatment [29].
The use of ASMs in this study showed some benefits with more than two-thirds of cases with seizures freedom after a mean treatment duration of 2.5 years. For other patients, there was a reduction in seizure frequency. In Nigeria, Ogunrin et al. found 87% of patient's seizures free after 3 months of ASMs and 100% after 6 months [13]. In this latest study, phenytoin was frequently prescribed. This demonstrates the effectiveness of phenytoin in seizure control in patients with PTE. When ASMs do not significantly reduce seizures, epilepsy surgery and vagus nerve stimulation may improve seizures control [16,30]. In this study, no patients underwent surgery, which is not available in Cameroon.

Conclusion
In Cameroon, post-traumatic epilepsy affects mainly young adults of the male gender. Road traffic accident are by the far, the leading mechanism of injury, leading to severe traumatic brain injury in about two-thirds of cases. Contusions and epidural haematoma were the commonest brain lesions on imaging. Generalized tonic-clonic seizures are the most frequent clinical seizure's type. Despite the limited choice in antiseizure medications, we achieve a seizure