Clinical experience with intravenous valproate as first-line treatment of status epilepticus and seizure clusters in selected populations.

Objective: The aim of this study was to evaluate the efficacy and safety of intravenous valproate (i.v. VPA) as first-line treatment of status epilepticus (SE) and seizure clusters in selected patient populations. Methods: We enrolled 23 patients (11 females and 12 males; mean age: 61 years) with SE who received i.v. VPA as first-line therapy (25 mg/kg in 100 mL saline infused over 15 min). ECG tracing was monitored before, during, and after infusion. Liver function and serum ammonia tests were conducted after 24 and 72 h of treatment. We evaluate the response of SE to i.v. therapy and short-term outcome. Results: In 15 out of 23 patients (65%), i.v. VPA was effective. In our population, we retrospectively identified three different subgroups: patients with cardiorespiratory comorbidities discouraging the use of traditional SE first-line drugs, patients with specific epileptic subsyndromes (such as idiopathic generalized epilepsy), and patients affected by psycho-organic syndromes. No significant adverse effects were detected. Discussion: Our study shows the clinical relevance of i.v. VPA as first-line therapy of SE in patients with medical conditions contraindicating the use of traditional first-line antiepileptic drugs for SE, and in those presenting with specific forms of SE.

Keywords: status epilepticus; valproate; therapy; comorbidity

Status epilepticus (SE) is a medical emergency associated with significant mortality and morbidity, thus requiring prompt and effective treatment. Benzodiazepines (BDZ) and phenytoin (PHT) are traditionally used as first- and second-line therapy [[1]], but they may be inappropriate in patients with serious cardiovascular and respiratory diseases because of their side effects. Intravenous valproate (i.v. VPA) provides an interesting alternative for these patients due to its limited hemodynamic action and its modest effects on the central nervous system. We conducted an open-label study to evaluate the short-term efficacy and tolerability of i.v. VPA as first-line therapy of SE or seizure clusters (SC) without full recovery of neurologic functions in patients with severe clinical conditions or in those requiring a rapid titration of the drug in order to prevent the early recurrence of SE/SC and to start a chronic therapy.

We selected 23 patients (11 females and 12 males; age range: 14–88 years; mean age: 61 years) with convulsive or non-convulsive SE. All SE episodes were documented in the Epilepsy Unit of the "Sapienza" University of Rome within a period of 12 months; the patients came to our observation from both neurology and non-neurology wards, from our outpatients service, and from the emergency department of the Policlinico Umberto I. SE was defined as a prolonged seizure/cluster of recurrent seizures lasting ≥30 min or two or more sequential seizures without intervening recovery of consciousness or other impaired functions [[2]]. The inclusion criteria were: 1. video-EEG documented SE or SC; 2. i.v. VPA administration as first-line antiepileptic drug (AED); and 3. complete clinical data including general characteristics, medical comorbidities, current therapies, and also syndromic classification in epileptic patients. Patients with hepatic failure or any signs/symptoms of liver dysfunction were excluded. For each patient, SE etiology, electroclinical features, and neuroimaging findings were retrospectively collected and analyzed. SE duration was defined as the interval between the onset of symptoms (identified on the basis of the information reported by the patients' relatives or by medical staff) and the beginning of VPA administration. Assuming that SE duration is still an unsolved question in literature, we chose to take into consideration the interval between the onset of symptoms (identified on the basis of the information reported by the patients' relatives or by medical staff) and the beginning of VPA administration, in order to allow some statistical correlation between SE duration at first medical observation and response to treatment. Although the recommended protocol for SE treatment commonly used in our department indicates first-line therapy with BDZ with PHT as second-line AED, in the selected cases, specific clinical conditions guided us to select VPA therapy. According to the guidelines specified by the Italian League Against Epilepsy [[4]], i.v. VPA was administered at a loading dose of 25 mg/kg in 100 mL saline infused over 15 min, followed by a daily maintenance dose of 800–1600 mg (at an infusion rate of 0.8–1 mg/kg/h). In all patients, video-EEG monitoring (Micromed System Plus and Xltek devices), including ECG tracing to evaluate heart rate (HR) and RR interval, was conducted during i.v. VPA administration. Efficacy of i.v. VPA efficacy was considered as the cessation of seizures by either clinical or electrical evidence without the addition of another AED to control seizures. If such a clinical response was observed after the loading dose, the infusion of the maintenance dose was started, and video-EEG was repeated after 24 h as a follow-up. In case of a poor response to VPA, another AED was administered as add-on therapy. A blood sample (including blood count, hepatic and renal parameters, electrolyte balance, and VPA plasma level) was analyzed in all patients at the baseline and 24 and 72 h after administration of i.v. VPA.

According to the small size of the sample, the statistical analyses were conducted by means of Fisher's exact test correction when required. Values of p < 0.05 were considered statistically significant. All analyses were conducted with SPSS version 12.0 (SPSS Inc., Chicago, IL).

We retrospectively selected and analyzed 23 patients with SE. Eleven patients had a previous diagnosis of epilepsy: in six cases, SE was a symptom of brain tumors (three cases), ischemic lesions (two cases), and hippocampal sclerosis (subsequent amygdalohippocampectomy; one case), whereas three patients had a probability of symptomatic epilepsy. In two cases, SE represented the onset of an idiopathic generalized epilepsy (IGE). Among epileptic patients, SE was related to a poor compliance with therapy in two cases; in one, it might have been caused by the introduction of inadequate therapy (AED-induced worsening of seizures); in two cases, SE represented the evolution of the preexisting epileptic syndrome and, in another two, it was the first manifestation of epilepsy itself, as previously stated. In four epileptic patients, SE was considered "symptomatic" because it was caused by either an evolutive lesion (in three patients) or an intervening medical condition (pneumonia and fever associated with antibiotic treatment in one case). Among nonepileptic patients, SE was the result of: acute ischemic lesions (three patients), subacute ischemic lesion (one patient), cerebral hemorrhage (two patients), subdural hematoma (SDH; in one patient), cerebral hypoperfusion after cardiac surgery (one patient), sepsis (one patient), steroid-responsive encephalopathy associated with autoimmune thyroiditis (SREAT; one patient) and cerebral vasculitis (one patient; general characteristics are shown in Table 1).

Table 1.  General characteristics of the patients.

5 IGE, idiopathic generalized epilepsy; LEV, levetiracetam; OCBP, obstructive chronic bronchopneumopathy; OXC, oxcarbazepine; PB, phenobarbital; TPM, Topamax.

In our population, three different subgroups could be retrospectively indentified: 1. patients with comorbidities which discouraged the use of BDZ, PHT, or phenobarbital (PB), because of the higher risk of cardiorespiratory (i.e. cardiac arrhytmias, hypotension, and hypoventilation) and neurologic (central nervous system depression) complications; in addition, in one patient, VPA was preferred as BDZs were refused by relatives because of the occurrence of a severe intoxication during the treatment of a previous similar SE episode; 2. patients presenting with generalized SE and SC characterized by an electroclinical pattern highly suggestive of specific underlying epilepsy subsyndromes, such as IGE; and 3. patients affected by a psycho-organic syndrome developed as a consequence of different cerebral pathologies, such as SREAT, frontal neoplasm, brain surgery for the treatment of temporal lobe drug-resistant epilepsy (amygdalohippocampectomy), or cryptogenic epilepsy presenting with agitation and psychotic symptoms.

SE was convulsive in 15 patients (partial in 13 cases and generalized in two, the latter presenting as generalized myoclonic SE) and nonconvulsive in eight (complex partial in six cases, and simple partial in two). Classification scheme of SE subtypes adopted in this work has been reported in the Figure 1.

Graph: Figure 1. Classification scheme of SE population according to published proposals [[5]].

VPA, administered i.v., was effective in 15 out of 23 patients, determined by the disappearance of EEG abnormalities and ictal signs/symptoms without recurrence within 24 h; in most responders, VPA proved to be effective within 1 h from the start of i.v. treatment. All nonresponsive patients had a symptomatic SE, associated with tumors (two patients), intracranial hemorrhage (one patient), subdural hematoma (one patient), subacute ischemic lesion (one patient), vasculitic lesions (one patient), sepsis (one patient), and pneumonia plus antibiotic therapy (one patient). A relapse of SE within 24 h was detected in three early-responsive patients, in whom SE developed as a consequence of sepsis, cerebral vasculitis and SDH. Three non responder patients presenting with acute symptomatic SE died because of the evolution of the underlying pathology (sepsis, hemorrhage, SDH, etc.) that had caused SE in the first place. Two other patients died due to intervening medical conditions although SE had been controlled (for SE findings, response to VPA, and outcome, see Table 2). Although strongly influenced by the small size of the sample, statistical analysis showed a relationship within the limits of statistical significance among SE duration ( p = 0.06), SE etiology (p = 0.05), cardiorespiratory comorbidities (p = 0.05), and response to therapy (Table 3). Response to treatment was not influenced by SE semeiologic subtypes and age at onset ( p = 0.09; Table 3).

Table 2.  SE findings, response to VPA, and outcome.

6 AED, antiepileptic drug; BDZ, benzodiazepine; CBZ, carbamazepine; CLB, clobazam; CLZ, clonazepam; ECC, extracorporeal circulation; GCSE, generalized convulsive status epilepticus; HE, Hashimoto encephalopathy; IGE, idiopathic generalized epilepsy; maint, maintenance; PCSE, partial convulsive status epilepticus; PCNCSE, partial complex non-convulsive status epilepticus; PHT, phenytoin; PSNCSE, partial simple non-convulsive status epilepticus; th, therapy; VPA, valproate acid.

No significant adverse effects were detected among the survivors: no changes in HR or RR were observed during i.v. VPA administration; no renal nor liver alterations emerged from the blood samples, except for a case of transient hyperammonemia (160 mg/dL).

The results of our study are surely affected by some limitations: (1) the small number of patients and (2) the uncontrolled, retrospective design of the study. Moreover, patients from the intensive care unit (ICU) were not included in our sample; therefore, VPA was never used in more severe types of SE, such as generalized tonic–clonic SE (GTCSE). Thus, we cannot rule out a bias due to patient selection.

Since the approval of i.v. VPA in 1980s, an increasing number of reports has supported its use as a therapeutic option for SE. PB, BDZ, and PHT are traditionally used in early and established SE, but serious side effects might occur; especially in patients with cardiorespiratory comorbidities, PHT can provoke hypotension and cardiac arrhythmias while PB and BDZ may cause hypoventilation and vigilance reduction [[1]],[7]]. Conversely, VPA has no central depressant effect and has a mild hemodynamic action: the most frequent side effects, which occurred in <10% of the patients treated, are hypotension, dizziness, and thrombocytopenia. Several studies confirm the safety profile of the VPA i.v. formulation, although cases of severe adverse effects, such as acute encephalopathy or liver failure in patients with mitochondrial disease, have been rarely reported [[8]]. VPA has proven to be effective in several types of SE [[10]], and is now widely recommended for SE treatment after failure of first-line therapy; recent studies have shown no significant differences among VPA, PHT, or BDZ [[10]]. On the contrary, reliable evidence on VPA as first-line AED is still lacking; a few

Table 3.  Response to i.v. VPA according to age, SE duration, SE etiology, SE subtypes, and comorbidities.

• 7 *p = 0.09; ** p = 0.06; ***p = 0.05 (Fisher's exact test correction)
• 8 CSE, convulsive status epilepticus; NCSE, non-convulsive status epilepticus.

open-label controlled studies evaluating VPA in early SE showed good success rates, but some of these suffered from selection bias or low statistical power [[14]]. Our study highlights the clinical relevance of i.v. VPA as the first-line choice in treating SE or SC in patients from specific populations. Firstly, it proved to be a therapeutic alternative in patients unfit to receive traditional first-line therapy because of the high risk of side effects due to concomitant medical problems (for details, see Table 1). Moreover, the lack of sedative effects allowed correct evaluation of patient's state of consciousness, which might be extremely useful in both comatose and awake patients when it is recommendable to follow their mental status over time and –if needed –to differentiate whether the impairment of consciousness is related to either SE or the underlying clinical condition. This attribute appears to be crucial, especially in old patients affected by stroke or other cerebrovascular diseases (that account for approximately one third of our population), in whom the continuous monitoring of consciousness levels is usually recommended. In addition, i.v. VPA was successfully used in two patients whose prolonged cluster seizures represented the onset of an IGE, which is known to be highly sensitive to VPA; in such cases, on one hand, the i.v. formulation enabled quick drug titration, which contributed to prevention of early recurrences of seizures and, on the other, the change to oral VPA allowed us to easily start a chronic therapy regimen. Finally, in SE patients presenting with psychiatric disturbances or symptoms due to a psycho-organic syndrome developed as a consequence of different cerebral pathologies, VPA was administered not only to treat SE, but also because of its good psychotropic effect. Its chronic administration also allowed to achieve a better long-term behavioral control. In our study, i.v. VPA was effective in controlling SE in 15 out of 23 patients (65%). Seizure control was obtained within 3 h after drug infusion, which is in disagreement with some previous studies reporting a response within 15–20 min or even during drug administration [[15], [17], [19]]; however, as mentioned earlier, in most patients, VPA proved to be effective within 1 h from the start of i.v. treatment. The delayed response might be explained by both the slightly lower loading dose and infusion rate (25 mg/kg in 100 mL saline in 15 min at a dose of 1.32–1.65 mg/kg/min), although plasma levels remained within range in all patients. In contrast with other reports [[20]], we adopted submaximal doses not only in accordance with the official recommendations of the Italian League Against Epilepsy [[4]] and the indications of the manufacturer, but also considering the clinical condition of our population, characterized by less severe forms of SE, as stated earlier (i.e. no ICU patients with GTCSE). As expected, epileptic patients in whom SE was related to the evolution of epileptic syndromes (e.g. IGE) or to inadequate antiepileptic therapy (worsening seizures/AED withdrawal) presented a good response to treatment (see Table 2). As regards patients with CSE, according to published data [[23], [25]], the aetiology likely influenced their worse response; indeed, in most cases (10 out of 15), CSE was acutely symptomatic of cerebral/systemic conditions which led to death in three patients. Nonetheless, we cannot rule out the role of the submaximal dose of VPA used in our study for determining the poor response of CSE patients. Patients with comorbidities showed a worse response to therapy and a poor outcome, both presumably related, once again, to patients' age and SE etiology. On the whole, i.v. VPA has proven to be well tolerated and safe, even in patients with concomitant medical conditions; no serious adverse effects or alterations in blood parameters were detected, except for a slight and transient hyperammonemia in one patient. Our work presents some limitations, mainly related to its retrospective, uncontrolled design and the small number of enrolled subjects. These limitations, also including the heterogeneity of the population and the unavoidable selection bias, strongly depend on the nature of the study itself, based on a real-life clinical approach. In conclusion, our study shows the clinical relevance of i.v. VPA as first-line therapy of SE in patients with medical conditions contraindicating the use of traditional SE first-line AEDs, and in those presenting with specific forms of SE; however, considering its duration for action, the use of i.v. VPA as first-line therapy should be prudently evaluated in more severe forms of SE that require fast-acting AEDs.

The authors declare that they have no conflict of interests. The authors alone are responsible for the content and writing of this article.