Computed tomography for patients with drug- or substance-induced seizures: a retrospective analysis

1. Introduction

Overdose and substance poisoning are major health issues globally. The World Health Organization estimated that 193,460 people died worldwide from unintentional poisoning in 2012 and nearly a million people die each year as a result of suicide [1]. Drug or substance poisoning can damage various organs. Seizure is a common poisoning-related neurological presentation in the emergency department (ED) [2]. An estimated 6% of new-onset seizures and up to 9% of status epilepticus cases are due to overdose or poisoning [3, 4].

Most drug/substance-induced seizures (DSSs) occur as a result of changes in the balance of excitatory and inhibitory neuronal activity, transmitters, and pathways, leading to involuntary discharge of the cerebral cortex [5]. The majority of DSSs manifest as generalized tonic-clonic seizures (GTCSs). Although most DSSs are not caused by actual brain damage, some studies have shown that poisoning might result in local brain injury and trigger seizures [6, 7].

Brain computed tomography (CT) is frequently used to evaluate intracranial lesions in patients with altered mental status or seizures in the ED. A high proportion of overdose and poisoned patients undergo brain CT [2, 8] and seizure is a common indication for CT examinations. CT is used widely because of its accessibility and reliability, but performing CT can be risky because it requires critically ill patients to leave closely monitored areas temporarily. Radiation exposure and the high cost are also of concern. Although brain CT in patients with a first GTCS is generally accepted [9, 10, 11, 12], it is not clear whether a patient with a DSS requires brain CT.

This study explored whether brain CT has a significant impact on the management and outcome in patients with DSS.

2. Materials & methods

3. Results

The study included 97 patients who met our criteria. The average age was 40.36 ± 12.95 years and 69 (71.13%) were males. Table 1 summarizes their demographic information and Table 2 lists the culprit drugs/substances. The most common cause of DSSs was alcohol withdrawal, accounting for more than half of the patients (57 patients, 58.76%), followed by carbon monoxide (12 patients, 12.37%), and stimulants (8 patients, 8.25%). While 69 patients (71.13%) had only a single GTCS, eight (8.25%) had status epilepticus. Fifteen (15.46%) patients required endotracheal intubation. Of the patients, 37 (37.76%) were admitted to an ordinary ward and 12 (12.26%) to the intensive care unit. Three patients ultimately died: two from the poisoning and the third from an underlying disease.

The 64 patients who underwent brain CT were assigned to Group A. All CT examinations were completed within 6 hours of hospital arrival. The 33 patients who did not undergo CT were assigned to group B. In Group A, four (6.25%) patients had an abnormal CT report (Table 3). CT images with the key findings of these four patients are shown in Fig. 1. The CT reports of patients one and two showed possible brain infarction and that of patient three was suspicious of both cerebral infarction and cerebral aneurysm. Patient one did not have any neurological symptoms and no further imaging study was arranged. Patients two and three underwent subsequent brain MRI or brain angiography and the results were negative for any acute abnormality. Patient four, who had a CT report of brain edema and increased intracranial pressure, was treated supportively with intravenous lorazepam, antiemetics, and hypertonic saline infusion (for hyponatremia). She did not receive any intracranial pressure-lowering therapies.

Fig. 1.CT images of four patients who had an abnormal CT report. (a) The brain CT report revealed a suspected cerebellar infarct. Refer to patient 1 in Table 3. (b) CT report revealed a suspected small infarct. Refer to patient 2 in Table 3. (c) CT report revealed a lacunar infarct and a possible left middle cerebral artery aneurysm at M1 segment. Refer to patient 3 in Table 3. (d) CT report revealed cerebral edema and increased intracranial pressure. Refer to patient 4 in Table 3.

Table 1 compares Groups A and B. Group A had worse GCS scores on arrival (p = 0.024) and higher rates of status epilepticus (p = 0.031). Groups A and B had similar admission rates, but once admitted, Group A had longer hospital stays (p = 0.032).

Approximately half of the patients (46) had a single GTCS either before ED arrival or in the ED and were not treated with any anticonvulsant; the other 51 cases were treated with anticonvulsants, and benzodiazepines were the most common first-line therapy (Table 4). Initially, 25 cases were treated with diazepam and 14 of them had repeat seizures that required re-dosing or other anticonvulsants. Lorazepam was given to 17 patients initially, and 6 had a repeat seizure. While patients initially treated with lorazepam had a lower rate of seizure recurrence, the treatment efficacies of the two drugs did not differ significantly (OR 2.33, 95% CI 0.65–8.30). The clinical parameters and disease severity of the patients initially treated with diazepam or with lorazepam did not differ, except the valium group was older (45.72 ± 15.63 vs. 35.94 ± 9.00 years, p = 0.014). Four patients were initially treated with midazolam; three of them had status epilepticus and received continuous midazolam infusions. Five patients were initially treated with a non-benzodiazepine antiepileptic; none had a witnessed GTCS on ED arrival. Because of the small numbers of cases and discrepant treatment indications, we did not include midazolam or other antiepileptic drugs in the efficacy comparison.

4. Discussion

In this study, approximately two-thirds of adult patients with a suspected DSS underwent brain CT despite the absence of head trauma, focal neurological symptoms, or focal seizures. Few of these brain CT scans (6.25%) had abnormal findings. Some of these abnormalities led to further imaging studies but ultimately had no major impact on clinical management.

The literature reports abnormal emergency CT findings in 34–56% of patients with a first seizure [14, 15, 16, 17, 18, 19]. It is generally accepted that brain CT should be performed in patients with a first GTCS. However, few articles have specifically discussed the role of brain CT in DSSs. In a prospective study, 43 of 152 poisoned patients with altered consciousness underwent brain CT, including eight scans done in 13 patients with seizures, and none of these examinations revealed acute abnormalities [2]. Although poisoning may cause intracranial lesions, such as intracranial hemorrhage or cerebral infarction [6], most of these diseases usually have clinical findings, such as focal neurological deficits. In our series, only 1 of the 40 poisoned patients (patient four in Table 3) had abnormal CT findings, which did not result in a significant change in management.

The usefulness of brain CT in alcohol-withdrawal seizures is also debatable. Earnest et al. [20] reported intracranial lesions in 16 of 259 (6%) patients with a first alcohol-related seizure and found no correlation with focal neurological symptoms. They regarded CT as an important test in these patients. However, 9 of these 16 patients had a minor head injury or minor focal neurological deficits that met our exclusion criteria. By contrast, Feussner et al. [21] found 1% reversible cerebral lesions in patients with suspected alcohol-related seizures who had no neurological symptoms, and concluded that CT does not improve the evaluation of these patients. Schoenenberger et al. [15] reported that, in patients who probably had alcohol-related seizures without focal neurological deficit, emergency brain CT may not be absolutely necessary. None of 35 patients with these characteristics in their series had a focal lesion on brain CT. In our study, after carefully excluding patients with a focal neurological deficit or with head trauma, only 3 of 57 patients suspected of having alcohol-withdrawal seizures had abnormal CT findings, and none of these abnormalities had a significant clinical impact.

Benzodiazepines are a commonly used first-line pharmaceutical therapy for DSSs [5, 22]. No randomized controlled study has compared the efficacies of individual benzodiazepines. A Cochrane review of status epilepticus of any cause found that intravenous lorazepam was better than intravenous diazepam for stopping status epilepticus [23]. For alcohol-withdrawal seizures, diazepam seems to have the most favorable pharmacokinetic profile (faster onset and longer half-life) [24], but studies failed to show its superiority over other benzodiazepines [25]. In our study, although patients initially treated with lorazepam had a lower rate of recurrent seizure, the result did not reach statistical significance.

DSS patients who underwent brain CT had a lower initial level of consciousness, higher rate of status epilepticus, and longer hospital stay once admitted. This implies that DSS patients with greater disease severity are more likely to undergo brain CT. This is consistent with several guidelines for new-onset seizures [9, 10, 11, 12], and may be strongly related to the clinical risk assessment of emergency physicians.

This study has several limitations. First, it was a retrospective chart review and defined DSS according to the judgement of the emergency physician in charge. DSS cases might have been mistakenly included or excluded. The drug or substance exposure relied highly on the exposure history, which might have been incorrect or misunderstood. Besides, the extent of the physician’s suspicion for DSS at the time of deciding to obtain a brain CT could not be well evaluated. Second, cases of DSS are relatively rare and an accurate diagnosis can be difficult. Although this is the largest reported case series, the sample size is still too small to provide strong evidence. Third, the CT abnormalities were based on the formal radiology report, not a dedicated team, which might have led to bias in interpretation.

5. Conclusions

Based on the results of our case series, brain CT examination did not provide additional information in the management of patients with suspected DSSs and did not have concomitant head trauma or significant neurological deficits. Multi-center prospective studies are needed to minimize the bias and obtain stronger evidence.

Author contributions

SFL and MNH designed the study, collected the data and composed the original draft. CHH performed the statistical analysis, composed the original draft. CKC reviewed the chart and collected the data. PCC collected the data and did the data validation. CJS performed the statistical analysis and edited the original draft. HYC designed and supervised the study, reviewed the chart, and edited the manuscript.

Ethics approval and consent to participate

This study was approved by Chang Gung Institutional Research Board (IRB: 20160883B0).

Acknowledgment

Thank numerous individuals participated in this study, and thanks to all the peer reviewers for their opinions and suggestions.

Funding

This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors. At the time of data acquisition, the authors were employed at Keelung branch of Chang Gung memorial hospital, Linkou branch of Chang Gung memorial hospital, and En Chu Kong hospital, Taiwan.

Conflict of interest

The authors declare no conflict of interest.

Data availability

The data used to support the findings of this study are available from the corresponding author upon request.