The Prevalence of Thyrotoxicosis-Related Seizures

Introduction

C ommon thyroid disorders may present with reversible neurologic signs and symptoms affecting the central and peripheral nervous systems (1). Seizures are among these signs and symptoms, and there are a number of case reports or small series of “idiopathic” seizures in patients with thyrotoxicosis (2 –6). The administration of thyroxine in several animal models of epilepsy reduces seizure threshold, whereas thyroidectomy protects against pharmacologically induced seizures (7 –9). Here we investigated the prevalence and clinical features of seizures that appeared to be related to the thyrotoxic state in a large series of patients with hyperthyroidism.

Materials and Methods

This study was conducted at the Severance Hospitals, which are located in Seoul, Korea. We retrospectively searched the diagnosis database for the period January 1990 to December 2008. We searched for all patients with thyrotoxicosis in all the discharge diagnoses and recorded how many patients had seizures. We reviewed the complete records of only those who had seizures as one of their final diagnoses. The epileptologist diagnosed patients with a seizure according to their history and the guardian's testimony in the medical records. We included patients who experienced seizure attacks with thyrotoxicosis and excluded patients with other seizure causes, such as an infection, metabolic disorders, neoplasm, and vascular disorders. We also excluded patients who had a history of epilepsy. The diagnosis of thyrotoxicosis was based on the laboratory tests for thyroid hormones in serum. If the patients had high triiodothyronine and free thyroxine with suppressed thyroid stimulating hormone, we diagnosed thyrotoxicosis. Radioactive iodine uptake, thyroid auto-antibodies, and a thyroid ultrasonography were performed as supportive tests in all patients. Brain imaging and an electroencephalography (EEG) were also performed in all patients with hyperthyroidism and acute seizures. After the diagnosis of the seizure and thyrotoxicosis, the follow-up studies included an EEG and measurements of serum thyroid hormone levels. When patients became euthyroid, we stopped the antiepileptic drug and re-examined the patient 6 to 12 months later.

Results

Records of 3382 admitted patients with hyperthyroidism were searched. Eleven patients (0.3%) were found to have hyperthyroidism and acute seizures. Among these, four patients were excluded, because two had a history of epilepsy, one had a history of sepsis, and one had a history of meningitis. This left seven patients (0.2%) with seizures and no obvious cause except for their thyrotoxicosis. Their average age was 39.6 years (range, 11–66 years old). Six patients had Graves' disease, and one was taking excess amounts of thyroid hormone. The semiology of seizures revealed primary generalized tonic-clonic seizures in four patients (57%), complex partial seizures with secondary generalized tonic-clonic seizures in two patients (29%), and a focal seizure in one patient (14%). Five patients underwent a computed tomography scan, and two patients underwent an magnetic resonance imaging (MRI). Based on the MRI results, one patient (Case 5) had diffuse brain atrophy, and the other patient (Case 4) had an old basal ganglia infarction. The electroencephalogram (EEG) was initially normal in two patients (29%), showed generalized slow activity in four patients (57%), and showed diffuse generalized beta activity in one patient (14%). The five patients who initially had abnormal EEG patterns were studied after they had been treated for their thyrotoxicosis and were euthyroid. After they were euthyroid, the EEG was normal in all of these five patients.

Four patients (58%) were treated with lorazepam for acute seizures. One patient (Case 2) was treated with valproate for 13 days before it was discontinued. Other patients (Case 1 and 6) were treated with phenytoin for 2–3 weeks. During the 18 to 24 month follow-up period in which the patients were euthyroid, none had recurrent seizures (Table 1).

Discussion

Jabbari and Huott reported that seizures occurred in 9% of all admissions for thyrotoxicosis and that thyrotoxicosis was the cause of the first seizure in 1.2% of all patients admitted for thyrotoxicosis (2). They believed that seizures are not rare in hyperthyroidism. In contrast, in our study only seven of the 3382 patients (0.2%) had seizures that appeared to be related to thyrotoxicosis. Another study noted a seizure rate of <1% in patients with thyrotoxicosis (10).

Little is known about the basic mechanisms underlying thyroid hormone action on excitability in the brain. Increased thyroid hormones influence the activity of sodium-potassium adenosine triphosphatase, leading to severely altered concentrations of sodium in neurons (11). Recently, the thyrotropin-releasing hormone (TRH) has been considered an important factor for anti-seizure treatment in thyrotoxicosis. The direct infusion of TRH to the hippocampus showed anticonvulsant effects in amygdale kindled rats by reducing the afterdischarge and seizure duration (12). These findings suggest that TRH in the hippocampus may be an endogenous anticonvulsant. Thyroid hormone alters tissue levels of TRH or its precursors in various limbic brain regions (13).

Table 1 presents clinical and laboratory characteristics of our patients and other patients in the literature who appeared to have thyrotoxicosis-related seizures (2,3,6,10,14 –20). None of these patients had concomitant abnormalities of glucose, serum electrolyte, or osmolarity, and there were no consistent changes on brain imaging studies. In most patients, the computed tomography was normal and MRI data were more limited. In one case, the MRI was normal and in two of our patients, one had mild cortical atrophy and the other had an old basal ganglia infarct.

The EEG patterns of those who had thyrotoxicosis-precipitated seizures varied. The frequency of an abnormal EEG was 78%, and the most common EEG abnormality was a generalized slowing (57%) in both the literature review cases and our cases (3,16,17). An excess of slow activity in the temporal regions in addition to the fast activity of the high amplitude in the frontal region is a characteristic of patients with thyrotoxicosis and seizures (17). There were no triphasic waves in patients with thyrotoxicosis, which is a representative feature of metabolic diseases. Moreover, the initial abnormal EEG mostly converted to normal after thyrotoxicosis treatment. In our study, the most common EEG abnormality was generalized slow activity in four patients (57%); one patient (Case 7) had beta activity but the other EEG findings were normal. These results are similar to those in a previous study (6).

As shown in Table 1, the type of seizure in thyrotoxicosis-associated seizures was variable, ranging from focal to generalized. Most patients showed generalized seizures, but five (22%) patients had focal seizure features. Two patients developed status epilepticus, which appeared to be precipitated by thyrotoxicosis.

Seizures associated only with thyrotoxicosis generally have a good prognosis. None of the patients were taking antiepileptic medication after they became euthyroid during their 18 to 24 month follow-up period. Therefore, we do not think it is necessary to maintain antiepileptic medication once these patients become euthyroid.

In conclusion, first-time seizures occur in a very small percentage of patients with thyrotoxicosis. Some such patients have brain abnormalities. These may have lowered their seizure threshold, but they are not enough to maintain seizures when the patients become euthyroid. Therefore, the prognosis of thyrotoxicosis-associated seizures is usually good if thyroid dysfunction is successfully treated.