Keywords
Psychogenic Nonepileptic Seizures - Epilepsy - Electroencephalography
Palabras clave
Convulsiones Psicógenas no Epilépticas - Epilepsia - Electroencefalografía
INTRODUCTION
It is known that epileptic seizures (ESs) and psychogenic nonepileptic seizures (PNESs)
can occasionally be hardly distinguished based only on semiology.[1] Most of the semiological signs, such as hyperventilation, crying, resistance to
eyelid opening, and pelvic movements, are mostly specific to PNES but may be also
encountered in ESs. Video EEG monitoring (VEM) is the gold standard for the differential
diagnosis of PNES from ES.[2] Video EEG monitoring units are established mostly in tertiary health clinics, need
a skilled epilepsy team, and are additionally used for the presurgical evaluation
of drug resistant epilepsy patients. The time until seizures occur (latency) directly
affects the time of the differential diagnosis.[3] It is an easily available quantitative value, but its contribution to the differential
diagnosis of ES and PNES is not fully known. While some studies showed significant
latency differences between PNES and ES,[3]
[4]
[5] other studies found no significant difference and no contribution to diagnosis.[6]
[7]
[8]
[9] However, these studies differed in terms of the groups examined (diagnostic, classification,
preoperative evaluation),[7]
[10] recording times,[10] induction techniques used,[7]
[11]
[12]
[13] drug reduction/ discontinuation protocols,[7]
[10]
[14] PNES semiology,[10]
[15] and seizure frequency in the pre-VEM period.[6]
In this study, we aimed to investigate the difference of the seizure latencies in
ES and PNES patients and its contribution to the diagnosis.
METHODS
The data of 497 adult (> 18 years old) patients hospitalized in our VEM unit for the
differential diagnosis of ES, seizure classification, or pre-surgical evaluation over
a 7-year period were retrospectively reviewed. Forty-eight PNES patients (48/497)
who were definitively diagnosed in VEM, and 51 consecutive ES patients matched for
gender and age were included in the study. A power value of 90% and an effect size
of 0.687 was estimated with a total of 92 cases (46 ES and 46 PNES). Patients with
other non-epileptic attacks (syncope, hypoglycemia, cardiac arrhythmia, cataplexy,
and movement disorders) and 18 PNES patients with concomitant ES were excluded. A
total of 294 seizures (40.1%; n = 118) PNES and 59.9% (n = 176) ES with focal ictal
onset) were evaluated. According to their semiological features, PNES was grouped
as subjective (29/118), akinetic (46/118), minor motor (26/ 118), and hypermotor (17/118)
types. The day was divided into four equal intervals (06:01–12:00, 12:01–18:00, 18:01–24:00,
and 24:01–06:00), and the time distribution of the seizures was determined. Patients'
age, gender, medications and medications tapered during VEM, number of seizures recorded
before and during VEM, and duration of monitoring were noted. Seizure latency was
determined as the time in hours from the start of video EEG recording to the first
seizure. Interictal and ictal EEG of PNES patients was normal. Neurological examinations
and cranial imaging findings of the patients were recorded. In both groups, with the
exception of standard provocative techniques (hyperventilation and intermittent photic
stimulation), induction techniques such as saline injection and suggestion were not
used. Antiepileptic drugs (AEDs) were reduced to induce seizures, and drugs were started
again after the desired number of seizures if they were epileptic.
Our study was approved by the local Scientific Research Ethics Committee on April
24, 2021, with the decision number 09/10, and written informed consent was obtained
from all participants prior to their inclusion in the study, which was conducted in
accordance with the Declaration of Helsinki.
Statistical analysis
Results are expressed as mean ± standard deviation or percentage. The compatibility
of the qualitative data to normal distribution was examined using the Shapiro-Wilk
test. The Mann-Whitney U test was used to compare qualitative values of the ES and
PNES groups. Relationships between continuous variables were analyzed by using the
Spearman correlation coefficient. The Pearson, Yates, or Fisher Chi-square tests were
used to compare categorical data between the ES and PNES groups. Statistical analyzes
were performed using the IBM SPSS Statistics for Windows, Version 20.0 (IBM Corp.,
Armonk, NY, USA) package program. A P-value < 0.05 was accepted as statistical significance limit value.
RESULTS
There was no significant difference between the age and gender of PNES and ES patients.
Age at the onset of seizure was lower in the ES group (20.7 ± 16.1 years vs. 27.5 ± 11.9
years, p < 0.001). Demographic and clinical features of the cases are summarized in [Table 1].
Table 1
Demographic and clinical features of all cases
|
PNES group (n = 48)
|
ES group (n = 51)
|
p
|
|
Sex (n)
|
Female
|
32 (66.7%)
|
28 (54.9%)
|
0.304
|
|
Male
|
16 (33.3%)
|
23 (45.1%)
|
|
Age (year)
|
32.8 ± 11.7
|
36.9 ± 15.1
|
0.09
|
|
Age at the onset of seizures (year)
|
27.5 ± 11.9
|
20.7 ± 16.1
|
< 0.001
|
|
Family history of epilepsy (yes/no)
|
11/ 37
|
12/ 39
|
0.1
|
|
Number of AEDs before VEM (n)
|
1.6 ± 0.9
|
2.2 ± 1.2
|
< 0.001
|
|
Duration of AED use (year)
|
3.88 ± 4.99
|
11.37 ± 9.83
|
< 0.001
|
|
Duration of VEM (day)
|
4.5 ± 2.4
|
5.6 ± 2.8
|
< 0.001
|
|
Number of seizures before VEM/year
|
122.6 ± 164.7
|
123.5 ± 154.4
|
0.218
|
|
Number of seizures in VEM/day
|
1.0 ± 1.3
|
1.5 ± 3.5
|
< 0.001
|
|
Onset time of the AED reduction (day)
|
1.1 ± 1.3
|
1.8 ± 1.4
|
0.006
|
|
Seizure latency (hour)
|
20.4 ± 24.2
|
45.8 ± 54.9
|
< 0.001
|
|
Seizure in the first 24 hours (yes/no)
|
35/13
|
25/26
|
0.023
|
|
Seizure in the first 48 hours (yes/no)
|
42/6
|
32/19
|
0.006
|
|
Seizure after 48 hours (yes/no)
|
6/42
|
19/32
|
0.006
|
Abbreviations: AEDs, antiepileptic drugs; ES, epileptic seizure; PNES, psychogenic
nonepileptic seizure; VEM, video-EEG monitoring.
Patients with PNES had a lower mean number of AEDs than ES patients and a shorter
mean duration of AED use (p < 0.001). The duration of VEM unit was shorter in PNES patients than in ES patients
(4.5 ± 2.4 days vs. 5.6 ± 2.8 days, p < 0.001). There was no difference in the number of seizures per year before VEM among
groups, but the mean number of seizures per day during VEM (seizures/ day) was higher
in the ES group (1.5 ± 3.5 vs. 1.0 ± 1.3, p < 0.001).
Seizure latency was shorter in PNES than in ES patients (20.4 ± 24.2 hour vs. 45.8 ± 54.9 hours,
p < 0.001). The percentage of cases who had their first seizure within ≤ 24 hours of
video-EEG recording was 72.9% in PNES and 49.1% in ES group (p = 0.023). The difference in first recorded seizure persisted on the 2nd day in favor
of PNES (87.5% of PNES patients vs. 62.7% of ES patients, p = 0.006), and recording longer than 48 hours for the 1st seizure was required in
12.5% of PNES vs. 37.3% of ES patients (p = 0.006) ([Figure 1]).
Figure 1 Cumulative probability plots of time to occurrence (in hours, following complete
electrode placement) of first recorded event during admission, for patients with epileptic
seizures and for patients with psychogenic nonepileptic seizures.
There was no difference between subjective, akinetic, minor motor, and hypermotor
types of PNES in terms of seizure latencies ([Table 2]).
Table 2
Demographic and clinical characteristics of PNES subgroups
|
Variables
|
Subjective
(n = 29)
|
Akinetic
(n = 46)
|
Minor motor (n = 26)
|
Hypermotor
(n = 17)
|
p*
|
|
Age (year)
|
26 (20–53)
|
37 (18–58)
|
38 (18–51)
|
26 (20–48)
|
0.771
|
|
Age at the onset of seizures (year)
|
21 (12–48)
|
32.5 (10–55)
|
23.5 (12–48)
|
22 (12–47)
|
0.784
|
|
Number of AEDs before VEM (n)
|
2 (0–3)
|
1 (0–3)
|
2 (1–3)
|
2 (0–2)
|
0.109
|
|
Duration of VEM (day)
|
4 (1–10)
|
3.5 (1–9)
|
5 (1–8)
|
4 (1–12)
|
0.307
|
|
Number of seizures in VEM/day
|
0.5 (0.2–1.5)
|
1 (0.2–5.0)
|
0.35 (0.1–7.0)
|
1 (0.2–1.7)
|
0.094
|
|
Onset time of the AED reduction (day)
|
1 (0–3)
|
0.5 (0–3)
|
0.5 (0–4)
|
1 (0–5)
|
0.863
|
|
Seizure latency (hour)
|
11 (1–70)
|
7.5 (1–89)
|
16.5 (1–79)
|
15.0 (1–77)
|
0.830
|
Abbreviations: AEDs, antiepileptic drugs; VEM, video-EEG monitoring.
Notes: Median (minimu –maximum); *Kruskal-Wallis test.
Most of the ESs and PNESs occurred between 06:01 and 24:00 ([Table 3]). While the seizures of ES patients were almost equally distributed among all time
intervals, most of the PNES cumulated especially between 18:01 and 24:00 (%35.6 vs.
%22.2, p < 0.01). A total of 20.5% of ESs and 10.2% of PNESs appeared between 24:01 and 06:00
(p = 0.029).
Table 3
Distribution of all seizures (n = 294) during the day
|
Time interval (hours)
|
Number of seizures
|
p
|
|
PNES (n = 118)
|
ES (n = 176)
|
|
06:01–12:00
|
26 (22.1%)
|
51 (28.9%)
|
0.184
a
|
|
12:01–18:00
|
38 (32.2%)
|
50 (28.4%)
|
0.486
a
|
|
18:01–24:00
|
42 (35.6%)
|
39 (22.2%)
|
< 0.01 a
|
|
24:01–06:00
|
12 (10.2%)
|
36 (20.5%)
|
< 0.05 b
|
Abbreviations: ES, epileptic seizure; PNES, psychogenic nonepileptic seizure.
Notes: n (%); aPearson Chi-square test; bYates Chi-square test.
DISCUSSION
In our study, the percentage of the seizure occurrence in the first 24 and 48 hours
were significantly higher in PNES patients than in ES patients, and the seizure latency
was significantly shorter in PNES. Psychogenic nonepileptic seizures occurred more
frequently between 18:01 and 24:00 and less frequently between 24:01 and 06:00 that
ES.
Thirty-eight to 89.6% of patients that were hospitalized for differential diagnosis
of ES had seizures within the first 24 hours in VEM.[3]
[4]
[6]
[7]
[8]
[9] The rate increased to 96.2% in the first 48 hours, and it was suggested that if
induction techniques were to be used as a diagnostic tool, they could be avoided for
the first 48 hours of VEM.[4] The rate of recording the seizure in the first 24 and 48 hours of VEM was higher
in PNES than in ES.[3]
[4] A relatively shorter VEM (≤ 48 hours) was found to be sufficient for the diagnosis
in patients with high clinical suspicion for PNES.[3]
Similar to these studies, we found that 72.9% of PNES patients had their 1st seizure
in the first 24 hours in contrast to 49.1% of ES patients (p = 0.023). On the 2nd day of VEM, we saw that this difference increased even more
between PNES and ES (87.5% vs. 62.7%, p = 0.006). Recording longer than 48 hours was required in only 25.3% of patients for
the record of the 1st seizure. The mean seizure latency was found to be shorter in
our PNES patients than in ES patients. This was in concordance with the study of Parra
et al., which reported a shorter latency in PNES than in ES (15 ± 16.3 hours vs. 28.6 ± 34 hours,
respectively).[4] Sagi et al.[3] found a shorter median latency in PNES patients than in ES patients (13.76 hours
vs. 22.4 hours, respectively), but the mean latency showed no significant difference.
In other studies, no significant relationship was found between the mean latencies
of ES and PNES.[6]
[7]
[8]
It is known that patients with PNES are prone to suggestion. Informing patients about
the purpose of VEM may provoke PNES and may result in a shorter seizure latency in
PNES.[13]
[14] It has been shown that a seizure while waiting in the outpatient clinic or during
the examination is 75% predictive for PNES.[16] A patient who had a seizure while placing the EEG electrodes has a higher probability
of suffering from PNES.[7] One of our PNES patients had a seizure within the first minute of the video EEG
recording immediately after completing the electrode mounting procedure. The effects
of hyperventilation and intermittent photic stimulation techniques on latency have
been investigated in a few studies and their inducing effect has been shown especially
in PNES.[11]
[12]
[13] The prolongation of hyperventilation to 5 minutes is effective in accelerating the
occurrence of seizures.[12] Intravenous saline injection has been shown to induce the occurrence of PNES, but
it is assumed not to be ethical because it may harm the patient-physician relationship
and is not absolute sensitive.[17]
[18] All of our patients were informed about the purpose of VEM, but none of them underwent
seizure induction, except for standard provocative techniques (3 minutes hyperventilation
and intermittent photic stimulation). Several studies reported that motor and hypermotor
types of PNES tend to occur earlier than other PNES types.[10]
[15] Because the pathophysiological brain mechanism of PNES is unclear, no explanation
could be made for this finding. We found no difference in seizure latency among PNES
types.
It was reported that there was a significant relationship between the frequency of
seizures before VEM and the seizure latency of ES and PNES patients during the VEM.[6] In contrast, Eisenman et al.[14] did not find any significant relationship. The self-reported pre-VEM seizure numbers
of our cases were similar, and no correlation was found between the seizure latency
and the number of pre-VEM seizures.
Psychogenic nonepileptic seizures are not associated with sleep, while some ESs occur
especially during sleep.[1]
[7]
[19] Nocturnal ESs are sometimes confused with PNESs due to the peculiar movements that
may accompany them. Some PNESs seem to occur during sleep, but their EEG demonstrates
wakefulness.[20] In our study, we found that PNES was not homogeneously distributed during the day
and showed a diurnal pattern. A smaller proportion of patients with PNES had seizures
between 24:01 and 06:00 than ES (10.2% vs. 20.5%, p = 0.029). These results are consistent with previous studies and show that seizures
occurring between 24:01 and 06:00 are more likely to be ES than PNES.[19]
[21]
While ES and PNES patients mostly have seizures in the first 48 hours of video-EEG
monitoring,[3]
[4]
[5]
[7]
[8]
[9]
[10]
[15]
[22] 35% of patients require follow-up longer than three days, and 7% longer than 1 week.[8] Monitoring longer than 5 days does not contribute additionally to the occurrence
of PNES, and the inconclusive rate is higher than that of ES patients (28% vs. 12.5%).[23] In our study, the mean VEM duration of the PNES cases was 4.5 ± 2.4 days (min. 24 hours-
max. 10 days) and 5.6 ± 2.8 days (min. 24 hours - max. 12 days) of the ES cases. Reducing
or terminating the AEDs may occasionally induce seizures and may alter the length
of VEM. In our study, the drug reduction protocol was decided by considering the clinical
characteristics of cases.
Limitations
Our study was a retrospective study, and the AED reduction protocol was not uniform
among all patients and may influence the seizure latency. The positive aspects of
our study are that spontaneous seizures were assessed without using induction techniques
other than hyperventilation and photic stimulation, and the diagnosis of patients
was made with the gold standard method VEM. Some PNES patients had nocturnal seizures
and VEM documented wakefulness in these seizures.
In conclusion, our study demonstrated that seizure latency was significantly shorter
in PNES than ES, and PNES clustered during daylight hours. Although not strictly reliable,
seizure latency can also be considered in the differential diagnosis of ES and PNES.
