Isolated Seizures

Peter Wolf

Introduction

All epilepsies begin with a first seizure. However, a first seizure does not necessarily mean that epilepsy will develop (Chapters 7 and 122). It may turn out to be an isolated event, symptomatic of some other pathologic condition (Chapter 8), or indicative of an inherent increased risk for development of epilepsy and the need to prevent such development.

This chapter does not deal with the first of these possibilities—the epileptic seizure as an acute symptomatic event—but rather with the second, which is often discussed under the heading of “first unprovoked seizures.” The discussion of these has mostly addressed two questions: How probable is it that a first “unprovoked” seizure signifies epilepsy, and should treatment begin after a first seizure? In this literature, treatment is always conceived of as pharmacotherapy, and a thorough definition of seizure provocation is rarely given. Important as these aspects may be, the theoretically more interesting question of nosology—the relation of isolated seizures to the established epileptic syndromes—is asked rarely, if at all, although apart from dealing with the occasional patient in such a situation, physicians hear about isolated seizures perhaps most frequently when taking the family histories of patients with epilepsy.

Historical Perspectives

The statement of Gowers¹⁰ that “we have no means of ascertaining, on any considerable scale, the frequency with which a single epileptic fit occurs without successors” clearly shows that he and his contemporaries were quite aware of the possibility of isolated seizures. However, this was not a major concern with the classic epileptologic writers. Apart from the specific subgroup of febrile seizures in infants, isolated seizures mostly were mentioned in epidemiologic and genetic investigations. These were briefly reviewed by Janz¹⁶ in a section on Gelegenheits-Epilepsien. His remarkable conclusion was that this is “at the same time the most frequent and mildest form of epilepsy, and the one with the highest rate of heredity.”

It was mainly for epidemiologic, prognostic, and similar statistical purposes and quantitative studies of therapy that epilepsy was defined as a minimum of two unprovoked seizures. This definition left isolated seizure events as a separate group that logically—and provided that the diagnosis of an epileptic seizure was reliable—had to be included in the classification of epilepsies and epileptic syndromes.⁴ The new International League Against Epilepsy (ILAE) and International Bureau for Epilepsy (IBE) definition of epilepsy as a disorder “characterized by an enduring predisposition to generate epileptic seizures” and requiring “the occurrence of at least one epileptic seizure”⁹ gives the possibility in certain instances of coming to a diagnosis of epilepsy even after one seizure, but these cases will probably be the exception rather than the rule.

Definitions and Nosologic Place

Isolated seizures can be defined as solitary seizure events that may occur once in a lifetime or very rarely, at lengthy intervals. A seizure event according to this definition would be one single seizure or a short series of two or more seizures in the course of 24 hours, or an isolated episode of status epilepticus.¹⁴

The best-known type of isolated seizure, febrile convulsions of early childhood, is considered to be a separate entity and is not discussed here.

The relationship of isolated seizures to other epilepsy syndromes has rarely been given much attention, let alone systematically considered. However, it appears that the rare recurrences of seizures in later life among patients with idiopathic localization-related childhood epilepsies are usually isolated generalized tonic–clonic seizures.¹⁹ The same seems to be true for another benign idiopathic epilepsy syndrome, benign familial neonatal convulsions,²² and isolated seizures may be found in the pedigree of patients with this syndrome.³

Epidemiology

Stress convulsions, defined as epileptic attacks in adults exposed to various “stressing” exogenous influences who have not previously had unprovoked epileptic attacks, were diagnosed in 37 of 1,250 patients with convulsive disorders studied during a 13-year period by Laue Friis and Lund.¹⁸ These patients can be considered a subset of the patients discussed here.

Hauser et al.¹² reported 244 patients with a first idiopathic or remote symptomatic seizure. These were drawn from a total of 1,047 patients with newly diagnosed seizures who had been enrolled during a 4 1/2-year period. They compared them with 334 patients who had acute symptomatic seizures and 435 who had had two or more seizures at first diagnosis; 34 patients were excluded from study. After 36 months of follow-up, seizures had recurred in 27% of the persons with only one seizure—that is, in 73%, the seizure had been an isolated event.

In the National General Practice Study of Epilepsy,²³ which was a prospective, population-based cohort study of 1,195 patients with newly diagnosed or suspected epileptic seizures, 220 patients had febrile seizures, and definite epileptic seizures were diagnosed in 564. Of these, 252 were registered at the time of their first seizure, 89 of whom were followed for 3 or more years. Of the latter group, 56% had had a seizure recurrence at 3 years of follow-up—that is, in 44%, the seizure had remained isolated.¹¹

The focus of these and similar studies, however, differs slightly but significantly from the subject of this chapter. They
were aimed at establishing the risk for recurrence after a first seizure and did not ask whether a second seizure signifies epilepsy or is only the recurrence of an isolated event, or Gelegenheitsanfall.

When seizure types are considered, it turns out that in all studies of this matter, generalized tonic–clonic seizures are by far the most frequent type. Percentages ranged between 61% (27% with simple and 10% with complex partial seizures; the remainder were unclassified, although these figures apply only to the cryptogenic group, the semiology of remote symptomatic seizures not being recorded)¹² and 97.5%.¹⁵ This important difference from the usual distribution of seizure types in clinical cohorts is probably attributable to the fact that with minor and less frightening seizures, a physician is frequently not seen after a first seizure, but is seen only later. These cases are frequently lost for studies of first seizures.

However, as Loiseau et al. pointed out, partial seizures also may occur as isolated events, and these have been reported in several studies.²⁰

Etiology and Basic Mechanisms

Apart from the bias toward generalized tonic–clonic seizures, it would appear that these patients are not nosologically homogeneous. Rather, they comprise both a group with an innate risk for development of some type of idiopathic generalized epilepsy and a mixed bag of patients at risk for development of symptomatic localization-related epilepsies of various etiologies.

Of the 564 patients with first seizures in the series of Sander et al.,²³ the seizures of 346 were classified as idiopathic/cryptogenic, those of 119 as remote symptomatic, those of 83 as acute symptomatic, and those of 16 as associated with neurologic deficit.

In the extended follow-up study of Hauser et al.¹³ of 208 patients with first seizures, the seizures of 149 were diagnosed as idiopathic and those of 59 as remote symptomatic. There were no acute symptomatic cases in this study.

The prospective cohort of adolescents and adults with first seizures of Wolf²⁶ comprised 104 patients as of January 31, 2002. Of these, 26 (25%) had a positive family history of epilepsy or febrile convulsions. Febrile convulsions were a previous manifestation of seizure susceptibility in 5 (4.8%), and 1 patient had had benign rolandic epilepsy in childhood. Twenty-five (24%) were photosensitive and thus revealed a trait most frequently seen in association with idiopathic generalized epilepsy syndromes.²⁴ In an additional 20 (19.2%), the electroencephalogram (EEG) showed some other type of generalized epileptiform discharge. On the other hand, indications of a remote symptomatic etiology were present in 15 (14.4%), and 41 (39.4%) displayed some focal trait (clinical, EEG, or both). In 16 patients (15.4%), no definite indicators of either a focal or a bihemispheric onset were found.

Clinical Presentation

There is good reason to assume that the manifestation of epilepsy of any type is often preceded by a stage of increased risk for development of seizures. This risk may be discovered on EEG if for some reason it happens to be performed. More often, risk becomes apparent through a first seizure. In a continuum stretching from no risk through low risk and high risk to active epilepsy, an isolated seizure would identify a person as belonging to a high-risk group for development of any type of epilepsy.

Risk for Relapse

The usual approach to this issue is to define the risk for relapse after a first seizure, which is the inverse expression of the chance of its remaining isolated. The discussion of this risk is highly controversial. The percentages in various recent studies vary from a low of 16% in the first year of follow-up and 27% in 3 years¹² to a maximum of 67% after 1 and 78% after 3 years.¹¹

The reasons for these variances are not entirely clear. They could not be explained by differences in therapy. In the study of Annegers et al.,¹ 60.6% of 424 patients received drugs. In cryptogenic (“idiopathic” in the authors’ terminology) cases, the relapse rates in patients taking drugs were considerably higher after 3 years (44% vs. 36%) and 5 years (60% vs. 41%), whereas in remote symptomatic cases there was no difference (70% for each group at 3 years and 80% vs. 76% at 5 years). In the Royal College of Physicians Study,¹⁵ 41 of 306 patients (13.4%) had had drugs prescribed by their family doctor or neurologist, and this had no influence on the relapse rate.

It has been suggested that the figures for recurrence in the optimistic studies were unrealistically low because the patients who had already had a second seizure were excluded from them at the onset of a prospective follow-up—thus excluding the group with the highest risk for relapse.⁸ This is certainly true, but when Hauser et al.¹³ in a second study included such patients who had been excluded from their previous one, only a minor increase in the relapse rate resulted. According to Berg and Shinnar,² with retrospective ascertainment, the success of follow-up may depend on the outcome because the patients who have recurrence are easier to find and follow. In addition, they pointed out that in such studies, “the investigator may have little control over the quality of the initial assessment.”

A serious methodologic weakness of the British studies showing a high relapse rate is indeed that most¹⁵ or even all¹¹ patients were not given their diagnoses by the investigators; rather, the diagnoses of other physicians were relied on, and no special precautions were taken to exclude the possibility that a patient with a first major seizure leading to a diagnosis had previously been subject to minor seizures. Some of these, such as absences, myoclonic seizures in juvenile myoclonic epilepsy, or isolated auras, can easily escape the attention of an inexperienced observer. In the study of Sander et al.,²³ which relied on information collected by general practitioners, only 3% of 564 patients with “definite epileptic seizures” were given diagnoses of “true absences or myoclonic jerks with or without generalized tonic–clonic seizures.” Obviously, these diagnoses were frequently missed, as can also be seen in another article from the same investigation.²¹ In this study of 1,195 primarily registered patients, only 9% were excluded because it was determined that they had already had seizures. In contrast, in the investigation of Hauser et al.,¹³ who are the only authors who discussed this problem and provided for it, 74% of patients “with newly identified unprovoked seizures had experienced multiple seizure episodes prior to their first medical contact”; these authors even excluded 13 patients referred by neurologists after a “first seizure” because a history of complex partial seizures had been overlooked—a diagnosis that would be thought hard to miss. It cannot be ruled out and seems even probable that the studies reporting high relapse rates comprise an unknown and potentially large number of patients with established but undiagnosed epilepsy.

Only gold members can continue reading. Log In or Register to continue