- Email: [email protected]
Non-epileptic seizures: delayed diagnosis in patients presenting with electroencephalographic (EEG) or clinical signs of epileptic seizures
Seizure 2002; 11: 193–197
doi:10.1053/seiz.2001.0617, available online at http://www.idealibrary.com on
Non-epileptic seizures: delayed diagnosis in patients presenting with electroencephalographic (EEG) or clinical signs of epileptic seizures † ‡ ‡ † † ‡ ` P. DE TIMARY , P. FOUCHET , M. SYLIN , J. P. INDRIETS , T. DE BARSY , A. LEFEBVRE & K. VAN † RIJCKEVORSEL †
´ de Clerlande 6, Universite´ Catholique de Louvain, Centre Neurologique Willam Lennox, allee ´ 1340 Ottignies-LLN; ‡ Universite´ Libre de Bruxelles, Service de Psychologie clinique et differentielle, 50 avenue F. Roosevelt, CP122, 1050 Bruxelles
The clinical differentiation between epileptic seizures (ES) and non-epileptic seizures (NES) is often difficult and mostly based on the presence or absence of widely recognized features of ES such as tongue biting, falling, incontinence or concomitant epileptic abnormalities in the electroencephalogram (EEG). We retrospectively analysed the records of all patients referred to our Epilepsy Centre for refractory epilepsy and finally diagnosed with NES between 1980 and 1999 (n = 103), half of them also exhibiting ES. The mean time-lapse between first attack and NES diagnosis was 8.7 ± 1.3 years and 16.5 ± 1.4 years for the NES and NES + ES groups respectively. At least one of the usual signs associated with generalized tonic–clonic seizures (tongue biting, falling or incontinence) was reported by 66% and 60% of patients with NES or NES + ES respectively. Interictal EEG abnormalities were recorded in 16% of NES patients vs. 80% of NES + ES patients. In the NES group, delay before establishing the correct diagnosis was significantly longer when the patients exhibited ≥1 symptom(s) of generalized seizures, or when patients exhibited interictal EEG abnormalities. Upon admission, 72% of NES patients and all NES + ES patients were being treated with antiepileptic drugs (AEDs). We conclude that EEG or clinical abnormalities suggestive of epileptic seizures are common in undiagnosed NES patients. Such diagnostic pitfalls, besides considerably delaying NES diagnosis, also considerably delay appropriate treatment implementation. c 2002 Published by Elsevier Science Ltd on behalf of BEA Trading Ltd.
Key words: non-epileptic seizures; fall; incontinence; tongue biting; diagnosis.
OBJECTIVES The diagnosis of epilepsy is often made in the clinician’s consulting room and is commonly based on indirect observations; i.e. a description of the ES by a patient’s relative and/or recording of abnormal interictal EEG activity. The treating physician is usually not able to observe the patient’s spells and the probability of seizures being recorded on a single standard EEG hardly exceeds 5% 1, 2 . A fraction of ES patients are referred to epilepsy centres with an alleged diagnosis of refractory epilepsy, 20% of whom 3 , presenting attacks which later prove their non-epileptic origin, hereafter referred to as NES 4 . Unnecessarily treating NES with antiepileptic medication will expose the patient to the drug’s side effects and may even exacerbate their symptoms 5 . Furthermore, NES and ES have often been observed in the same patients 6 and appropriate treatment requires recognition of both types of seizures. 1059–1311/02/$22.00/0
Such recognition can be achieved by simultaneous recording of both symptoms and EEG during an attack using EEG/Video monitoring, which is usually available in most epilepsy centres. This markedly improves the differentiation between ES and NES 7–9 . The present study was designed to evaluate how often simple EEG or clinical markers, easily within reach of the treating neurologist and generally considered as suggestive of epileptic seizures, are encountered in NES patients. We also evaluated how misinterpreting these markers may have set back the diagnosis of NES, and delayed appropriate treatment.
PATIENTS AND METHODS The medical records of 103 consecutive inpatients (males to female 35/68, mean age 36 ± 1.5 years) diagnosed with NES by EEG/video intensive monitoring
c 2002 Published by Elsevier Science Ltd on behalf of BEA Trading Ltd.
194
P. de Timary et al.
Table 1: Patient’s characteristics (n = 103).
Mean age at 1st seizure Mean age at diagnosis Male : Female Time-lapse to NES diagnosis n AEDs upon admission n subjects with 0–1–2 and 3 to 5 AEDs
(years) (years) % (years)
NES (n = 50)
NES + ES (n = 53)
P
21.0 (1.8) 33.9 (1.5) 34 : 66 8.7 (1.3) 1.46 (0.20) 14–15–11–10
14.5 (2.0) 38.0 (1.6) 34 : 66 16.5 (1.4) 2.58 (0.16) 0–12–12–29
0.02 NS NS 0.0001 0.0001
a Results are means (SEM) unless otherwise stated. NES: non-epileptic seizure; ES: epileptic seizure; AEDs: anti-epileptic drug(s); NS: non-significant.
Table 2: Clinical characteristics of patients with NES or NES and ES (n = 103).
Neurological examination Epileptiform clinical signs Tongue biting Fall Incontinence Tongue biting or fall or incontinence
NES (n = 50)
NES + ES (n = 53)
P
Normal/abnormal
48/2
46/7
NS
Absent/present Absent/present Absent/present 0/at least 1
43/7 20/30 37/13 17/33
44/9 29/24 36/17 22/31
NS NS NS NS
a NES: non-epileptic seizure; ES: epileptic seizure; NS: non-significant.
between January 1980 and May 1999 in our Epilepsy Centre were retrospectively analysed. Upon admission, patients were submitted to careful general and neurological examination, and questioned about their medications, their epileptic past and about the presence of incontinence, tongue biting or falls concomitant with the attacks. The following criteria were used to establish presence of NES: (1) the recording of >1 attacks clinically similar to those usually presented by the patient, (2) good quality EEG recordings, not fully obscured by muscle artefacts and not displaying epileptiform activity during the attack, nor displaying pre or postictal changes. (3) Video recordings allowing us to exclude ES when (i) the symptoms failed to follow a physiologic pattern, or (ii) the epileptiform behaviour was not simultaneous with EEG changes. In dubious cases, a seizure provocative test (using saline as placebo) was performed 10–12 . Additional evidence included a 2.5–10.0 fold increase in basal prolactin levels, 20 min after the seizure 13, 14 . Finally, in most cases, synchronized cardiac Holter monitoring was obtained simultaneously to EEG-video recordings to exclude a cardiogenic origin to the seizure like events.
Statistical analysis Differences between group means were analysed with unpaired Student’s t for measurements at the interval
level and differences between subject proportions with the χ 2 test for measurements at the categorical level. Two sided P values were used for all comparisons, with significance defined at P = 0.05.
RESULTS The cohort (n = 103) was divided into two groups of patients according to the presence or absence of concomitant ES. Group characteristics are indicated in Table 1. The group presenting with NES + ES had their first (epileptic or non-epileptic) seizure at a younger age than the NES group, and the mean age at NES diagnosis was only slightly higher in the NES + ES group. 72% of patients in the NES group and all patients in the NES + ES group were receiving antiepileptic drugs (AEDs) upon admission (Table 1). Most patients had normal standard neurological examination (Table 2). The proportion of patients presenting with tongue biting, falling or incontinence during seizures did not significantly differ between groups (Table 2). In 66% of NES patients and 58% of NES + ES patients, at least one of these clinical signs of epileptic seizures had previously been reported to occur during fits. Table 2 shows that 80% of patients presenting with NES + ES and only 16% of patients with NES had an abnormal interictal EEG (nonspecific slowing, mild asymmetry). Unfortunately, variants of physiologic patterns such as phantom spike and wave are often considered as an abnormal
Non-epileptic seizures: delayed diagnosis
195
pattern. Among patients with recent and available imaging data (n = 56), including brain computed tomography (CT), and/or magnetic resonance imaging (MRI), data were abnormal in 35% of NES and 44% of NES + ES patients. However this number is too low to perform a statistical analysis between both groups. In the population of patients presenting with NES alone (Fig. 1), time lapses between age at first attack and diagnosis of NES were compared according to the presence or absence of (i) at least one ‘specific’ clinical sign of epileptic seizures (tongue biting, fall or incontinence), (ii) anomalies in the interictal EEG. The existence of clinical signs of epileptic seizures or of abnormalities in the interictal EEG was associated with significantly longer delay before diagnosis. 20
NS
0,03
0,05
Abn N MRI or CT scan
0 1 or more "Specific" Signs of Seizure
N Abn Interictal EEG
Time (Years)
16 12 8 4 0
Fig. 1: Time lapses between first attack and appropriate diagnosis were compared in ES patients presenting with normal (N) or abnormal (Abn) cerebral imaging data, with no or ≥1 clinical sign of epileptic seizures (tongue biting, fall or incontinence) or with a normal or abnormal interictal EEG. MRI: Magnetic resonance imaging; CT: computed tomography.
DISCUSSION From all patients referral to epilepsy centres with the alleged diagnosis of refractory epilepsy, around 20% (9–40%) are diagnosed with NES, following long term video-EEG monitoring studies 3 . Among the latter, we found that more than half had both ES and NES in our survey (n = 103). Patients with NES In our Centre, patients with NES exhibited an 8.7 ± 1.3 years delay between first attack and diagnosis, a time-lapse in accordance with previously reported data 8, 15 but longer than that reported by Meierkort 7 . That the diagnosis was delayed is not surprising in view of the difficulty in correctly diagnosing epilepsy outside specialized Centres. King et al. 16 reported that neurologists correctly predicted NES in 50% of patients entering their epilepsy care unit, the proportion being improved (to 80%) after bedside
observation of seizures by staff members. The remaining fraction necessitated video-EEG monitoring for correct diagnosis 16 . However, referring physicians are rarely afforded the opportunity to observe the spells. Our data demonstrate how the scarcity of convincing EEG and/or clinical information in possession of the treating neurologist results in considerable postponement of the correct diagnosis of NES. Neurological examinations were rarely abnormal and when present, the abnormalities were observed in both groups, although more often in NES + ES patients. These indicators of central nervous system impairments as well as a history of brain damage 15 are therefore not sufficient to support the epileptic nature of all seizures. Besides motor events, convulsive seizures are classically characterized by the occurrence of typical signs: tongue biting, incontinence, falling and injuries. Even recent textbooks mention that these signs rarely occur during NES 17, 18 and when present, probably help the physicians in establishing diagnosis of convulsive ES. Our data show that these so-called ‘specific’ seizure-related signs are frequent in NES, and that the existence of concurrent epilepsy does not increase their frequency. Similar findings have been reported by others 7, 8, 19 . It is worth mentioning that 66% of our NES patients had at least one of these signs. In our subgroup of NES patients in whom at least one ‘specific’ sign of epileptic seizures had been reported, our delay before diagnosis was 5 years longer than in the symptom-free group. Our data therefore suggest that episodes of tongue biting, incontinence and fall might have delayed questioning by treating physicians on the possible non-epileptic character of the spells, even when fits had not responded to AED for several years. These signs should therefore no longer be regarded as ‘specific’ signs of epileptic seizures. Peguero et al. 19 rather suggested that in NES patients, they support the severity of the underlying psychiatric disease, are often associated with a history of suicide attempts and speculated they are possible equivalents of self-injury . In keeping with previous observations 16, 20 , the interictal EEG presented with abnormalities in one out of five (n = 10) NES patients, including two patients with a phantom spike-and-wave physiologic pattern 21, 22 . It is of note that the anomalies in the interictal EEG, even if non-specific, were associated with a significantly longer delay before diagnosis. Although an abnormal interictal EEG was more frequent in patients that presented with NES + ES, it should be emphasized that routine EEG may add to the confusion of the treating physician. Only a positive ictal EEG can provide convincing evidence of the epileptic origin of a seizure. Conversely, the absence of abnormalities during an ictal EEG is not
196
sufficient to exclude an epileptic origin, especially in cases of simple partial seizures or frontal lobe seizures 23 . In the latter case, patients may present with exuberant complex partial seizures that evoke psychiatric symptoms, and the ictal EEG may be normal 24–26 . However, these attacks usually occur at night-time and the clinical pattern for each patient is stereotyped, so that epilepsy can be diagnosed from repeated video clinical observation of fits 26 and confirmed if needed by intracranial monitoring. Patients with NES + ES
The concurrence of both seizure types in the same patient is commonly described, although opinion widely varies as to how frequently it occurs (10 to 80%) 6, 9, 27 . One difficulty in studying this population is that it is impossible retrospectively to distinguish symptoms that were of epileptic or of non-epileptic nature. The diagnosis of NES is even harder in patients in whom it coexists with epilepsy. It could be speculated, therefore, that coexistence of both seizure types would delay the diagnosis of NES. Delay between first attack and diagnosis was indeed longer in our NES + ES group than in the NES group. However, the correct diagnosis of both types of seizures will allow patients to recognize them and possibly seek appropriate solutions to their psychological problems.
Iatrogenic consequences of NES On admission, 72% of our NES patients and all NES + ES patients were on AEDs, a proportion closely matching that obtained by Leis et al. 28 . The referring clinician had prescribed more than one medication in 40% of NES patients, probably because their seizures were not responsive to a single medication. In keeping with these observations, several authors 19, 28, 29 observed that patients with NES seizures are more likely to be endotracheally intubated following pseudostatus epilepticus than patients with ES. Incorrect diagnosis of epilepsy may thus unnecessarily expose patients for prolonged periods, to treatment side effects and to the risks of iatrogenic injuries. This stresses the importance of obtaining an EEG-video recording during seizures in refractory cases and in every case in which doubts exist as to the epileptic nature of the spells. CONCLUSION Before NES is unravelled using video-EEG monitoring technique, its diagnosis is usually missed for
P. de Timary et al.
several years. In our NES population, the presence of at least one ‘typical’ sign of epileptic seizures (tongue biting, fall or incontinence) or abnormal (aspecific) interictal EEG has been found in 66% of patients.These pitfalls may have mislead the physicians and prevented them from reconsidering the diagnosis of epilepsy and have been associated with a longer delay for the correct diagnosis of NES. A consequence of such late diagnosis is that patients are often unnecessarily treated with AEDs and that a possible psychiatric origin is not even contemplated. Therefore a diagnosis of epilepsy should rapidly be reconsidered in patients in whom seizures do not respond to AED treatment.
ACKNOWLEDGEMENTS We thank Dr M. P. Hermans for helpful comments on this manuscript and technicians and nurses for their effective presence during EEG-video recordings.
REFERENCES 1. Perry, T. R., Gummit, R. J., Gates, J. R. and Leppik, I. E. Routine EEG vs intensive monitoring in the evaluation of intractable epilepsy. Public Health Reports 1983; 98: 384–389. 2. Ebersole, J. S. Ambulatory cassette EEG. Journal of Clinical Neurophysiology 1985; 2: 397–418. 3. Francis, P. and Baker, G. A. Non-epileptic attack disorders (NEAD): a comprehensive review. Seizure 1999; 8: 53–61. 4. Scull, D. A. Pseudoseizures or non-epileptic seizures (NES); 15 synonyms. Journal of Neurology, Neurosurgery and Psychiatry 1997; 62: 200. 5. Niedermeyer, E., Blumer, D., Holscher, E. and Walker, B. A. Classical hysterical seizures facilitated by anticonvulsivant toxicity. Psychiatria Clinica 1970; 3: 71–84. 6. Ramsay, R. E., Cohen, A. and Brown, M. C. Coexisting epilepsy and non-epileptic seizures. In: Non-epileptic Seizures (Eds J. A. Rowan and J. Gates ). Boston, Butterworth– Heinemann, 1993: pp. 47–55. 7. Meierkort, H., Will, B., Fish, D. and Shorvon, S. The clinical features and prognosis of pseudoseizures diagnosed using video-EEG telemetry. Neurology 1991; 41: 1643–1646. 8. Desai, B. T., Porter, R. J. and Penry, J. K. Psychogenic seizures. A study of 42 attacks in six patients, with intensive monitoring. Archives of Neurology 1982; 39: 202–209. 9. Rowan, J. A. An introduction to current practice in the diagnosis of non-epileptic seizures. In: Non-epileptic Seizures (Eds J. A. Rowan and J. R. Gates). Boston, ButterworthHeinemann, 1993: pp. 1–7. 10. Cohen, R. J. and Suter, C. Hysterical seizures: suggestion as a provocative EEG test. Annals of Neurology 1982; 11: 391–395. 11. Drake, M. E. Saline activation of pseudoepileptic seizures: clinical, EEG and neuropsychiatric observations. Clinical Electroencephalography 1985; 16: 171–176. 12. French, J. A., Kanner, A. M., Rosenbaum, D. H. and Rowan, A. J. Do techniques of suggestion aid the differential diagnosis of psychogenic vs epileptic seizures? Epilepsia 1987; 28: 612–613. 13. Trimble, M. Serum prolactin in epilepsy and hysteria. British Medical Journal 1978; 2: 1682.
Non-epileptic seizures: delayed diagnosis 14. Ohman, R., Walinder, J., Balldin, J. and Wallin, L. Prolactin response to electroconvulsive therapy. Lancet 1976; 2: 936–937. 15. Blumer, D. The paroxysmal somatoform disorder: a series of patients with non-epileptic seizures. In: Non-epileptic Seizures (Eds J. A. Rowan and J. R. Gates). Boston, Butterworth–Heinemann, 1993: pp. 165–172. 16. King, D. W., Gallagher, B. B., Murfin, A. J., Smith, D. B., Markus, D. J., Hartlage, L. C. et al. Pseudoseizures: diagnostic evaluation. Neurology 1982; 32: 18–23. 17. Porter, R. J. Epileptic and non-epileptic seizures. In: Nonepileptic Seizures (Eds J. A. Rowan and J. R. Gates). Boston, Butterworth–Heinemann, 1993: pp. 9–20. 18. Betts, T. Psychiatric aspects of non-epileptic seizures. In: Epilepsy: A Comprehensive Textbook (Eds J. Engel and T. Pedley). Philadelphia, Raven, 1997: pp. 2101–2116. 19. Peguero, E., Abou-Khalil, B., Fakhoury, T. and Mathews, G. Self-injury and incontinence in psychogenic seizures. Epilepsia 1995; 36: 586–591. 20. Luther, J. S., McNamara, J. O., Carwile, S., Miller, P. and Hope, V. Pseudoepileptic seizures: methods and video analysis to aid diagnosis. Annals of Neurology 1982; 12: 458–462. 21. White, J. C., Langston, J. W. and Pedley, T. A. Benign epileptiform transient of sleep: clarification of the small sharp spike controversy. Neurology 1977; 27: 1061–1068.
197 22. Reiher, J. and Klass, D. W. Two common EEG patterns of doubtful clinical significance. Medical Clinics of North America 1984; 933–940. 23. Williamson, P. D., Spencer, D. D. and Spencer, S. S. Complex partial seizures of frontal lobe origin. Annals of Neurology 1985; 18: 497–504. 24. Ludwig, B., Ajmone-Marsan, C. and Van Buren, J. Cerebral seizures of probable orbitofrontal origin. Epilepsia 1975; 16: 141–158. 25. Niedermeyer, E. and Walker, A. E. Mesio frontal epilepsy. Electroencephalography and Clinical Neurophysiology 1971; 31: 104–105. 26. Williamson, P. D. Psychogenic non-epileptic seizures and frontal seizures: diagnostic considerations. In: Non-epileptic Seizures (Eds J. A. Rowan and J. R. Gates). Boston, Butterworth–Heinemann, 1993: pp. 55–72. 27. Lesser, R. P., Lueders, H. and Dinner, D. S. Evidence for epilepsy is rare in patients with psychogenic seizures. Neurology 1983; 33: 502–504. 28. Leis, A. A., Ross, M. A. and Summers, A. K. Psychogenic seizures: ictal characteristics and diagnostic pittfalls. Neurology 1990; 40: 1404–1407. 29. Pakalnis, A., Drake, M. E. Jr and Philips, B. Neuropsychiatric aspects of psychogenic status epilepticus. Neurology 1991; 41: 1104–1106.
doi:10.1053/seiz.2001.0617, available online at http://www.idealibrary.com on
Non-epileptic seizures: delayed diagnosis in patients presenting with electroencephalographic (EEG) or clinical signs of epileptic seizures † ‡ ‡ † † ‡ ` P. DE TIMARY , P. FOUCHET , M. SYLIN , J. P. INDRIETS , T. DE BARSY , A. LEFEBVRE & K. VAN † RIJCKEVORSEL †
´ de Clerlande 6, Universite´ Catholique de Louvain, Centre Neurologique Willam Lennox, allee ´ 1340 Ottignies-LLN; ‡ Universite´ Libre de Bruxelles, Service de Psychologie clinique et differentielle, 50 avenue F. Roosevelt, CP122, 1050 Bruxelles
The clinical differentiation between epileptic seizures (ES) and non-epileptic seizures (NES) is often difficult and mostly based on the presence or absence of widely recognized features of ES such as tongue biting, falling, incontinence or concomitant epileptic abnormalities in the electroencephalogram (EEG). We retrospectively analysed the records of all patients referred to our Epilepsy Centre for refractory epilepsy and finally diagnosed with NES between 1980 and 1999 (n = 103), half of them also exhibiting ES. The mean time-lapse between first attack and NES diagnosis was 8.7 ± 1.3 years and 16.5 ± 1.4 years for the NES and NES + ES groups respectively. At least one of the usual signs associated with generalized tonic–clonic seizures (tongue biting, falling or incontinence) was reported by 66% and 60% of patients with NES or NES + ES respectively. Interictal EEG abnormalities were recorded in 16% of NES patients vs. 80% of NES + ES patients. In the NES group, delay before establishing the correct diagnosis was significantly longer when the patients exhibited ≥1 symptom(s) of generalized seizures, or when patients exhibited interictal EEG abnormalities. Upon admission, 72% of NES patients and all NES + ES patients were being treated with antiepileptic drugs (AEDs). We conclude that EEG or clinical abnormalities suggestive of epileptic seizures are common in undiagnosed NES patients. Such diagnostic pitfalls, besides considerably delaying NES diagnosis, also considerably delay appropriate treatment implementation. c 2002 Published by Elsevier Science Ltd on behalf of BEA Trading Ltd.
Key words: non-epileptic seizures; fall; incontinence; tongue biting; diagnosis.
OBJECTIVES The diagnosis of epilepsy is often made in the clinician’s consulting room and is commonly based on indirect observations; i.e. a description of the ES by a patient’s relative and/or recording of abnormal interictal EEG activity. The treating physician is usually not able to observe the patient’s spells and the probability of seizures being recorded on a single standard EEG hardly exceeds 5% 1, 2 . A fraction of ES patients are referred to epilepsy centres with an alleged diagnosis of refractory epilepsy, 20% of whom 3 , presenting attacks which later prove their non-epileptic origin, hereafter referred to as NES 4 . Unnecessarily treating NES with antiepileptic medication will expose the patient to the drug’s side effects and may even exacerbate their symptoms 5 . Furthermore, NES and ES have often been observed in the same patients 6 and appropriate treatment requires recognition of both types of seizures. 1059–1311/02/$22.00/0
Such recognition can be achieved by simultaneous recording of both symptoms and EEG during an attack using EEG/Video monitoring, which is usually available in most epilepsy centres. This markedly improves the differentiation between ES and NES 7–9 . The present study was designed to evaluate how often simple EEG or clinical markers, easily within reach of the treating neurologist and generally considered as suggestive of epileptic seizures, are encountered in NES patients. We also evaluated how misinterpreting these markers may have set back the diagnosis of NES, and delayed appropriate treatment.
PATIENTS AND METHODS The medical records of 103 consecutive inpatients (males to female 35/68, mean age 36 ± 1.5 years) diagnosed with NES by EEG/video intensive monitoring
c 2002 Published by Elsevier Science Ltd on behalf of BEA Trading Ltd.
194
P. de Timary et al.
Table 1: Patient’s characteristics (n = 103).
Mean age at 1st seizure Mean age at diagnosis Male : Female Time-lapse to NES diagnosis n AEDs upon admission n subjects with 0–1–2 and 3 to 5 AEDs
(years) (years) % (years)
NES (n = 50)
NES + ES (n = 53)
P
21.0 (1.8) 33.9 (1.5) 34 : 66 8.7 (1.3) 1.46 (0.20) 14–15–11–10
14.5 (2.0) 38.0 (1.6) 34 : 66 16.5 (1.4) 2.58 (0.16) 0–12–12–29
0.02 NS NS 0.0001 0.0001
a Results are means (SEM) unless otherwise stated. NES: non-epileptic seizure; ES: epileptic seizure; AEDs: anti-epileptic drug(s); NS: non-significant.
Table 2: Clinical characteristics of patients with NES or NES and ES (n = 103).
Neurological examination Epileptiform clinical signs Tongue biting Fall Incontinence Tongue biting or fall or incontinence
NES (n = 50)
NES + ES (n = 53)
P
Normal/abnormal
48/2
46/7
NS
Absent/present Absent/present Absent/present 0/at least 1
43/7 20/30 37/13 17/33
44/9 29/24 36/17 22/31
NS NS NS NS
a NES: non-epileptic seizure; ES: epileptic seizure; NS: non-significant.
between January 1980 and May 1999 in our Epilepsy Centre were retrospectively analysed. Upon admission, patients were submitted to careful general and neurological examination, and questioned about their medications, their epileptic past and about the presence of incontinence, tongue biting or falls concomitant with the attacks. The following criteria were used to establish presence of NES: (1) the recording of >1 attacks clinically similar to those usually presented by the patient, (2) good quality EEG recordings, not fully obscured by muscle artefacts and not displaying epileptiform activity during the attack, nor displaying pre or postictal changes. (3) Video recordings allowing us to exclude ES when (i) the symptoms failed to follow a physiologic pattern, or (ii) the epileptiform behaviour was not simultaneous with EEG changes. In dubious cases, a seizure provocative test (using saline as placebo) was performed 10–12 . Additional evidence included a 2.5–10.0 fold increase in basal prolactin levels, 20 min after the seizure 13, 14 . Finally, in most cases, synchronized cardiac Holter monitoring was obtained simultaneously to EEG-video recordings to exclude a cardiogenic origin to the seizure like events.
Statistical analysis Differences between group means were analysed with unpaired Student’s t for measurements at the interval
level and differences between subject proportions with the χ 2 test for measurements at the categorical level. Two sided P values were used for all comparisons, with significance defined at P = 0.05.
RESULTS The cohort (n = 103) was divided into two groups of patients according to the presence or absence of concomitant ES. Group characteristics are indicated in Table 1. The group presenting with NES + ES had their first (epileptic or non-epileptic) seizure at a younger age than the NES group, and the mean age at NES diagnosis was only slightly higher in the NES + ES group. 72% of patients in the NES group and all patients in the NES + ES group were receiving antiepileptic drugs (AEDs) upon admission (Table 1). Most patients had normal standard neurological examination (Table 2). The proportion of patients presenting with tongue biting, falling or incontinence during seizures did not significantly differ between groups (Table 2). In 66% of NES patients and 58% of NES + ES patients, at least one of these clinical signs of epileptic seizures had previously been reported to occur during fits. Table 2 shows that 80% of patients presenting with NES + ES and only 16% of patients with NES had an abnormal interictal EEG (nonspecific slowing, mild asymmetry). Unfortunately, variants of physiologic patterns such as phantom spike and wave are often considered as an abnormal
Non-epileptic seizures: delayed diagnosis
195
pattern. Among patients with recent and available imaging data (n = 56), including brain computed tomography (CT), and/or magnetic resonance imaging (MRI), data were abnormal in 35% of NES and 44% of NES + ES patients. However this number is too low to perform a statistical analysis between both groups. In the population of patients presenting with NES alone (Fig. 1), time lapses between age at first attack and diagnosis of NES were compared according to the presence or absence of (i) at least one ‘specific’ clinical sign of epileptic seizures (tongue biting, fall or incontinence), (ii) anomalies in the interictal EEG. The existence of clinical signs of epileptic seizures or of abnormalities in the interictal EEG was associated with significantly longer delay before diagnosis. 20
NS
0,03
0,05
Abn N MRI or CT scan
0 1 or more "Specific" Signs of Seizure
N Abn Interictal EEG
Time (Years)
16 12 8 4 0
Fig. 1: Time lapses between first attack and appropriate diagnosis were compared in ES patients presenting with normal (N) or abnormal (Abn) cerebral imaging data, with no or ≥1 clinical sign of epileptic seizures (tongue biting, fall or incontinence) or with a normal or abnormal interictal EEG. MRI: Magnetic resonance imaging; CT: computed tomography.
DISCUSSION From all patients referral to epilepsy centres with the alleged diagnosis of refractory epilepsy, around 20% (9–40%) are diagnosed with NES, following long term video-EEG monitoring studies 3 . Among the latter, we found that more than half had both ES and NES in our survey (n = 103). Patients with NES In our Centre, patients with NES exhibited an 8.7 ± 1.3 years delay between first attack and diagnosis, a time-lapse in accordance with previously reported data 8, 15 but longer than that reported by Meierkort 7 . That the diagnosis was delayed is not surprising in view of the difficulty in correctly diagnosing epilepsy outside specialized Centres. King et al. 16 reported that neurologists correctly predicted NES in 50% of patients entering their epilepsy care unit, the proportion being improved (to 80%) after bedside
observation of seizures by staff members. The remaining fraction necessitated video-EEG monitoring for correct diagnosis 16 . However, referring physicians are rarely afforded the opportunity to observe the spells. Our data demonstrate how the scarcity of convincing EEG and/or clinical information in possession of the treating neurologist results in considerable postponement of the correct diagnosis of NES. Neurological examinations were rarely abnormal and when present, the abnormalities were observed in both groups, although more often in NES + ES patients. These indicators of central nervous system impairments as well as a history of brain damage 15 are therefore not sufficient to support the epileptic nature of all seizures. Besides motor events, convulsive seizures are classically characterized by the occurrence of typical signs: tongue biting, incontinence, falling and injuries. Even recent textbooks mention that these signs rarely occur during NES 17, 18 and when present, probably help the physicians in establishing diagnosis of convulsive ES. Our data show that these so-called ‘specific’ seizure-related signs are frequent in NES, and that the existence of concurrent epilepsy does not increase their frequency. Similar findings have been reported by others 7, 8, 19 . It is worth mentioning that 66% of our NES patients had at least one of these signs. In our subgroup of NES patients in whom at least one ‘specific’ sign of epileptic seizures had been reported, our delay before diagnosis was 5 years longer than in the symptom-free group. Our data therefore suggest that episodes of tongue biting, incontinence and fall might have delayed questioning by treating physicians on the possible non-epileptic character of the spells, even when fits had not responded to AED for several years. These signs should therefore no longer be regarded as ‘specific’ signs of epileptic seizures. Peguero et al. 19 rather suggested that in NES patients, they support the severity of the underlying psychiatric disease, are often associated with a history of suicide attempts and speculated they are possible equivalents of self-injury . In keeping with previous observations 16, 20 , the interictal EEG presented with abnormalities in one out of five (n = 10) NES patients, including two patients with a phantom spike-and-wave physiologic pattern 21, 22 . It is of note that the anomalies in the interictal EEG, even if non-specific, were associated with a significantly longer delay before diagnosis. Although an abnormal interictal EEG was more frequent in patients that presented with NES + ES, it should be emphasized that routine EEG may add to the confusion of the treating physician. Only a positive ictal EEG can provide convincing evidence of the epileptic origin of a seizure. Conversely, the absence of abnormalities during an ictal EEG is not
196
sufficient to exclude an epileptic origin, especially in cases of simple partial seizures or frontal lobe seizures 23 . In the latter case, patients may present with exuberant complex partial seizures that evoke psychiatric symptoms, and the ictal EEG may be normal 24–26 . However, these attacks usually occur at night-time and the clinical pattern for each patient is stereotyped, so that epilepsy can be diagnosed from repeated video clinical observation of fits 26 and confirmed if needed by intracranial monitoring. Patients with NES + ES
The concurrence of both seizure types in the same patient is commonly described, although opinion widely varies as to how frequently it occurs (10 to 80%) 6, 9, 27 . One difficulty in studying this population is that it is impossible retrospectively to distinguish symptoms that were of epileptic or of non-epileptic nature. The diagnosis of NES is even harder in patients in whom it coexists with epilepsy. It could be speculated, therefore, that coexistence of both seizure types would delay the diagnosis of NES. Delay between first attack and diagnosis was indeed longer in our NES + ES group than in the NES group. However, the correct diagnosis of both types of seizures will allow patients to recognize them and possibly seek appropriate solutions to their psychological problems.
Iatrogenic consequences of NES On admission, 72% of our NES patients and all NES + ES patients were on AEDs, a proportion closely matching that obtained by Leis et al. 28 . The referring clinician had prescribed more than one medication in 40% of NES patients, probably because their seizures were not responsive to a single medication. In keeping with these observations, several authors 19, 28, 29 observed that patients with NES seizures are more likely to be endotracheally intubated following pseudostatus epilepticus than patients with ES. Incorrect diagnosis of epilepsy may thus unnecessarily expose patients for prolonged periods, to treatment side effects and to the risks of iatrogenic injuries. This stresses the importance of obtaining an EEG-video recording during seizures in refractory cases and in every case in which doubts exist as to the epileptic nature of the spells. CONCLUSION Before NES is unravelled using video-EEG monitoring technique, its diagnosis is usually missed for
P. de Timary et al.
several years. In our NES population, the presence of at least one ‘typical’ sign of epileptic seizures (tongue biting, fall or incontinence) or abnormal (aspecific) interictal EEG has been found in 66% of patients.These pitfalls may have mislead the physicians and prevented them from reconsidering the diagnosis of epilepsy and have been associated with a longer delay for the correct diagnosis of NES. A consequence of such late diagnosis is that patients are often unnecessarily treated with AEDs and that a possible psychiatric origin is not even contemplated. Therefore a diagnosis of epilepsy should rapidly be reconsidered in patients in whom seizures do not respond to AED treatment.
ACKNOWLEDGEMENTS We thank Dr M. P. Hermans for helpful comments on this manuscript and technicians and nurses for their effective presence during EEG-video recordings.
REFERENCES 1. Perry, T. R., Gummit, R. J., Gates, J. R. and Leppik, I. E. Routine EEG vs intensive monitoring in the evaluation of intractable epilepsy. Public Health Reports 1983; 98: 384–389. 2. Ebersole, J. S. Ambulatory cassette EEG. Journal of Clinical Neurophysiology 1985; 2: 397–418. 3. Francis, P. and Baker, G. A. Non-epileptic attack disorders (NEAD): a comprehensive review. Seizure 1999; 8: 53–61. 4. Scull, D. A. Pseudoseizures or non-epileptic seizures (NES); 15 synonyms. Journal of Neurology, Neurosurgery and Psychiatry 1997; 62: 200. 5. Niedermeyer, E., Blumer, D., Holscher, E. and Walker, B. A. Classical hysterical seizures facilitated by anticonvulsivant toxicity. Psychiatria Clinica 1970; 3: 71–84. 6. Ramsay, R. E., Cohen, A. and Brown, M. C. Coexisting epilepsy and non-epileptic seizures. In: Non-epileptic Seizures (Eds J. A. Rowan and J. Gates ). Boston, Butterworth– Heinemann, 1993: pp. 47–55. 7. Meierkort, H., Will, B., Fish, D. and Shorvon, S. The clinical features and prognosis of pseudoseizures diagnosed using video-EEG telemetry. Neurology 1991; 41: 1643–1646. 8. Desai, B. T., Porter, R. J. and Penry, J. K. Psychogenic seizures. A study of 42 attacks in six patients, with intensive monitoring. Archives of Neurology 1982; 39: 202–209. 9. Rowan, J. A. An introduction to current practice in the diagnosis of non-epileptic seizures. In: Non-epileptic Seizures (Eds J. A. Rowan and J. R. Gates). Boston, ButterworthHeinemann, 1993: pp. 1–7. 10. Cohen, R. J. and Suter, C. Hysterical seizures: suggestion as a provocative EEG test. Annals of Neurology 1982; 11: 391–395. 11. Drake, M. E. Saline activation of pseudoepileptic seizures: clinical, EEG and neuropsychiatric observations. Clinical Electroencephalography 1985; 16: 171–176. 12. French, J. A., Kanner, A. M., Rosenbaum, D. H. and Rowan, A. J. Do techniques of suggestion aid the differential diagnosis of psychogenic vs epileptic seizures? Epilepsia 1987; 28: 612–613. 13. Trimble, M. Serum prolactin in epilepsy and hysteria. British Medical Journal 1978; 2: 1682.
Non-epileptic seizures: delayed diagnosis 14. Ohman, R., Walinder, J., Balldin, J. and Wallin, L. Prolactin response to electroconvulsive therapy. Lancet 1976; 2: 936–937. 15. Blumer, D. The paroxysmal somatoform disorder: a series of patients with non-epileptic seizures. In: Non-epileptic Seizures (Eds J. A. Rowan and J. R. Gates). Boston, Butterworth–Heinemann, 1993: pp. 165–172. 16. King, D. W., Gallagher, B. B., Murfin, A. J., Smith, D. B., Markus, D. J., Hartlage, L. C. et al. Pseudoseizures: diagnostic evaluation. Neurology 1982; 32: 18–23. 17. Porter, R. J. Epileptic and non-epileptic seizures. In: Nonepileptic Seizures (Eds J. A. Rowan and J. R. Gates). Boston, Butterworth–Heinemann, 1993: pp. 9–20. 18. Betts, T. Psychiatric aspects of non-epileptic seizures. In: Epilepsy: A Comprehensive Textbook (Eds J. Engel and T. Pedley). Philadelphia, Raven, 1997: pp. 2101–2116. 19. Peguero, E., Abou-Khalil, B., Fakhoury, T. and Mathews, G. Self-injury and incontinence in psychogenic seizures. Epilepsia 1995; 36: 586–591. 20. Luther, J. S., McNamara, J. O., Carwile, S., Miller, P. and Hope, V. Pseudoepileptic seizures: methods and video analysis to aid diagnosis. Annals of Neurology 1982; 12: 458–462. 21. White, J. C., Langston, J. W. and Pedley, T. A. Benign epileptiform transient of sleep: clarification of the small sharp spike controversy. Neurology 1977; 27: 1061–1068.
197 22. Reiher, J. and Klass, D. W. Two common EEG patterns of doubtful clinical significance. Medical Clinics of North America 1984; 933–940. 23. Williamson, P. D., Spencer, D. D. and Spencer, S. S. Complex partial seizures of frontal lobe origin. Annals of Neurology 1985; 18: 497–504. 24. Ludwig, B., Ajmone-Marsan, C. and Van Buren, J. Cerebral seizures of probable orbitofrontal origin. Epilepsia 1975; 16: 141–158. 25. Niedermeyer, E. and Walker, A. E. Mesio frontal epilepsy. Electroencephalography and Clinical Neurophysiology 1971; 31: 104–105. 26. Williamson, P. D. Psychogenic non-epileptic seizures and frontal seizures: diagnostic considerations. In: Non-epileptic Seizures (Eds J. A. Rowan and J. R. Gates). Boston, Butterworth–Heinemann, 1993: pp. 55–72. 27. Lesser, R. P., Lueders, H. and Dinner, D. S. Evidence for epilepsy is rare in patients with psychogenic seizures. Neurology 1983; 33: 502–504. 28. Leis, A. A., Ross, M. A. and Summers, A. K. Psychogenic seizures: ictal characteristics and diagnostic pittfalls. Neurology 1990; 40: 1404–1407. 29. Pakalnis, A., Drake, M. E. Jr and Philips, B. Neuropsychiatric aspects of psychogenic status epilepticus. Neurology 1991; 41: 1104–1106.