Electrographic status epilepticus or encephalopathy in baclofen intoxication?

Abstract

Baclofen is a synthetic derivative of the inhibitory neurotransmitter gamma-aminobutyric acid (GABA). It is considered first-line treatment for spasticity, particularly in spinal cord-related diseases such as multiple sclerosis, spinal cord injuries, and cerebral palsy. Doses between 80 and 2500 mg are associated with intoxication and more than 200 mg is associated with intensive care unit admission.¹ Commonly reported effects of baclofen overdose include muscle weakness, hemodynamic instability, cardiac arrhythmias, respiratory depression, CNS depression, and epileptic seizures.² The exact mechanism responsible for the proconvulsant effect of high baclofen doses is unknown. Indeed, low-dose baclofen rather exhibits antiepileptic effects. Thus, the clinical proconvulsant effects properties seem paradoxical. They may be explained by GABA receptors located presynaptically on GABAergic axonal terminals modulating neural activity.¹ Activation of presynaptic GABA receptors can lead to the inhibition of inhibition providing a potential mechanism for neuronal excitation.

Some early reports described the electroencephalographic features of baclofen toxicity as nonconvulsive status epilepticus (NCSE),³ while others described them as consistent with baclofen-induced encephalopathy.⁴ Since then, detailed Salzburg consensus criteria⁵ and American Clinical Neurophysiology Society's (ACNS) Standardized Critical Care EEG Terminology criteria⁶ for the diagnosis of NCSE have been defined. We describe the electroencephalographic findings in a 49-year-old woman with a reduced level of consciousness caused by intoxication with 300 mg of baclofen.

The patient had chronic pain, asthma, ischemic heart disease, hypertension, and cervical disc herniation. She was found unconscious with tongue bite, soaked with urine and with blood on her face. Paramedics administered 2.5 mg of midazolam. Neurological evaluation revealed small twitches in the left arm and shoulder and no other focal signs. Consciousness was reduced (GCS 11 [E3, V3, M5]). Due to clinical suspicion of NCSE, 5.5 g of levetiracetam was administered i.v., followed by 3.5 g of valproate and 200 mg of lacosamide. There was no clinical improvement. Drug screening was negative and there was no ethanol in the blood. Lactate, standard biochemistry, magnetic resonance imaging of the brain, and spinal fluid analysis including examination for autoimmune encephalitis antibodies was unremarkable.

EEG recorded the day after admission showed sharp generalized periodic discharges at a frequency of 26/10 seconds (Figure 1A) fulfilling Salzburg consensus and ACNS criteria for NCSE.^{5, 6} However, this activity was terminated by auditory stimulation (Figure 1B) and followed by diffuse theta activity. The activity shown in Figure 1A was present during the entire recording except for stimulations, when it was interrupted for approximately 10 seconds. Repeat EEG at Day 4 of admission (Figure 1C) showed diffuse slowing and generalized sharp waves occurring at irregular intervals.

Consciousness gradually improved, and the patient was fully awake 9 days after admission. The patient reported having autonomously increased the baclofen dose and ingested approximately 30 baclofen tablets (10 mg each) daily prior to admission. There were no suicidal thoughts. The patient was in her premorbid functional state at clinical follow-up two months later. Informed consent was obtained from the patient.

This is the first report of patient intoxicated with baclofen with electrographic features that fulfill Salzburg and ACNS criteria for NCSE. However, we noted that this activity was terminated by stimulation and replaced by diffuse theta activity. In a recent review of generalized periodic discharges in patients with encephalopathy, it was suggested that stimulus-induced wakefulness with transient improvement of the EEG is a significant feature that should be considered as a major criterion to determine that the EEG is not ictal in addition to established criteria for NCSE.⁷ To further complicate the distinction between metabolic encephalopathy and NCSE, the former can be associated with epileptic seizures.⁸

Previous case reports have described generalized sharp waves in baclofen intoxication.^{9, 10} The sharp waves were described as pseudo-periodic or periodic, and the frequency ranged from <1⁹ to 2 Hz.¹⁰ In addition, there was slowing of background activity. Epileptiform activity disappeared following baclofen discontinuation.

In 14 patients with baclofen intoxication identified through a hospital EEG database, 10 had generalized triphasic waves occurring at a frequency of 1–2 Hz.¹¹ None of the recordings fulfilled criteria for NCSE. Further, there was generalized slowing or suppression of background activity in all patients. Experiments in rats showed that electroencephalographic changes are dose dependent ranging from generalized periodic discharges or bi- or triphasic complexes with a 1- to 2-s interval to burst suppression and eventually suppression.¹²

In conclusion, baclofen intoxication may induce electroencephalographic features which fulfill Salzburg and ACNS criteria for NCSE. However, considering that the epileptiform activity was terminated by auditory stimulation and the good clinical outcome, we interpret the electrographic findings as a pattern of toxic encephalopathy. The strict application of Salzburg criteria may lead to overestimation of NCSE in patients intoxicated with baclofen.

The authors declare that they have no conflicts of interest.