- Email: [email protected]
Diagnostic yield of standard-wake and sleep EEG recordings
Accepted Manuscript Diagnostic yield of standard-wake and sleep EEG recordings Pirgit Meritam, Elena Gardella, Jørgen Alving, Daniella Terney, Melita Cacic Hribljan, Sándor Beniczky PII: DOI: Reference:
S1388-2457(18)30072-5 https://doi.org/10.1016/j.clinph.2018.01.056 CLINPH 2008413
To appear in:
Clinical Neurophysiology
Accepted Date:
26 January 2018
Please cite this article as: Meritam, P., Gardella, E., Alving, J., Terney, D., Cacic Hribljan, M., Beniczky, S., Diagnostic yield of standard-wake and sleep EEG recordings, Clinical Neurophysiology (2018), doi: https://doi.org/ 10.1016/j.clinph.2018.01.056
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
1
Diagnostic yield of standard-wake and sleep EEG recordings
Pirgit Meritam1, Elena Gardella1,2, Jørgen Alving1, 3, Daniella Terney1, Melita Cacic Hribljan3, Sándor Beniczky1,4
1. Department of Clinical Neurophysiology, Danish Epilepsy Centre, Dianalund, Denmark 2. University of Southern Denmark, Odense, Denmark 3. Department of Clinical Neurophysiology, Rigshospitalet, Copenhagen, Denmark 4. Department of Clinical Neurophysiology, Aarhus University Hospital, Aarhus, Denmark
Corresponding author: Professor Sándor Beniczky, Danish Epilepsy Centre and Aarhus University Hospital. Postal address: Visbys Allé 5, 4293 Dianalund, Denmark. Telephone: +45 26981536. Email: [email protected]
2
Abstract Objective: To investigate whether Posterior Dominant Rhythm (PDR) can be reliably assessed in sleep-EEG recordings and to investigate the diagnostic yield of standard-wake and sleep-recordings. Methods: EEG recordings of 303 consecutive patients aged 18-88 years were analyzed. All patients had both standard-wake and sleep-recordings, including patients who had abnormal standard recordings. Melatonin was used in 6% of sleep EEGs, and sleep deprivation in 94%. The mean duration of sleep was 41 minutes. We measured the PDR frequency in standard and sleep-recordings, both before and after sleep. We compared the diagnostic yield of standard-wake and sleep EEG recordings. Results: Compared to standard EEG, sleep-recordings showed a significantly lower PDR frequency, both when measured before and after sleep (p<0.001). One-hundred-fifty-six patients (51%) had normal standard recordings, and 35 of them (22%) had abnormal findings in the sleep-recording. One-hundredforty-seven patients had abnormal standard recordings and in 16 of them (11%) these abnormalities were not present in sleep-recording. Conclusions: PDR is significantly slower in the wake periods of sleep-recordings, compared to standard wake recordings. Significance: Sleep and standard wake recordings are complementary.
Keywords: diagnostic yield, EEG, posterior dominant rhythm, sleep deprivation, standard recording.
3
Highlights
•
The frequency of the Posterior Dominant Rhythm (PDR) is lower in the wake periods of sleeprecordings compared to standard wake EEG.
•
Not all abnormalities identified in standard wake recordings are present in sleep.
•
Sleep and standard wake recordings are complementary.
4
1. Introduction Sleep influences neuronal excitability and thereby occurrence of epileptic discharges (Burr et al., 1991; Declerck, 1986; Janz, 1962; Marinig et al., 2000; Martins da Silva et al., 1984). Several studies showed that in patients suspected for epilepsy, following normal or inconclusive standard-wake EEG, recording sleep EEG increases the diagnostic yield (Bennett et al., 1964, 1969; Gunderson et al. 1973; Jovanovic, 1991; Niedermeyer, 1993; Roupakiotis et al., 2000) by activating epileptiform discharges (Degen, 1980; Geller et al., 1969; Klinger et al., 1991; Logothetis et al. 1986; Mattson et al., 1965; Molaie and Cruz, 1988; Pratt et al., 1968; Rowan et al., 1982; Tartara et al., 1980). Leach et al. (2006) suggested that using sleep deprived EEG as the preferred protocol may reduce the number of EEGs carried out by 45%. On the contrary, two large pediatric studies concluded that sleep deprivation before the recording, and sleep during the recording did not increase the diagnostic yield of EEGs (Gilbert et al., 2004; DeRoos et al., 2009). Furthermore, in a subgroup of patients, sleep had inhibitory effect on the occurrence of epileptiform discharges (Beniczky et al., 2012). Provocation methods, like hyperventilation and photic stimulation, are typically done in standard wake recordings. Sleep deprivation can induce seizures in patients with epilepsy (Bennett et al., 1963, 1964; Gunderson et al., 1973; Rodin et al., 1962) and it can even provoke seizures in 3-5% of patients without epilepsy (Degen and Degen, 1991), concluding that it would be prudent to limit sleep deprivation to patients in whom epilepsy is clinically suspected but who have normal or inconclusive standard EEG recording (Marinig et al., 2000). In healthcare systems that are under continuous pressure to increase efficiency, there is a growing number of requests from referring physicians to skip the standard EEG, and only record sleep EEG, in spite of the conflicting evidence and the possible risks. From a practical point of view, combining wake standard EEG (including provocation methods) with sleep EEG in a single recording will either increase recording time
5
significantly or alternatively diminish the total sleep time in the recording. As the sensitivity of a sleep EEG depends on the duration of sleep, this shortening may be of clinical importance (Craciun et al., 2014). Besides the possible inhibition of epileptiform discharges in a subgroup of patients, and the risk of inducing seizures due to sleep deprivation, skipping standard-wake EEGs raises the question about the reliability of measuring the frequency of the posterior dominant rhythm (PDR) in sleep-recordings, affected by lingering drowsiness. Ferreira et al 2006 showed significant decrease in alfa frequency band after total sleep deprivation with increase in theta frequency in temporal and occipital regions (Ferreira et al., 2006). The clinical significance of these changes however, is unclear. We determined the diagnostic yield of standard-wake and sleep EEG recordings, and we assessed whether sleep deprivation affects the frequency of the PDR, measured in the wake periods of sleep EEG recordings.
2. Methods Standard-wake EEG and sleep EEG recordings of 303 consecutive patients (140 female), aged 18-88 years (mean 42 years) were retrospectively analyzed. Inclusion criteria were existence of both standard-wake and sleep-recording in the database (Beniczky et al., 2017). The patients either had a known epilepsy diagnosis or were suspected of having epilepsy. Most importantly, sleep EEGs were recorded also in patients who had an abnormal standard-wake recording. Standard-wake EEG was always the first recording. EEGs were assessed by certified clinical neurophysiologists who were blinded to all clinical data. Each recording was reviewed by two experts and disagreements were resolved via consensus discussions in the whole group. EEGs were recorded using the standard electrode array of the IFCN (Seeck et al., 2017). In 19 sleeprecordings (6%) melatonin was given prior to the recording. In all other cases (284=94%), the patients were sleep deprived. Duration was 30 minutes for the standard recordings and 60 minutes for sleep-recordings.
6
The mean duration of sleep in sleep-recordings was 41 minutes (range: 3-66 minutes). Provocation methods such as hyperventilation and photostimulation were used in standard recordings. For each patient we measured the frequency of the PDR separately in a standard EEG and a sleep EEG recording (in the latter, during the wake period both before and after sleep). The time interval between the standard and the sleep-recording was no more than 1 year. The PDR frequency was measured in a time epoch that best fulfilled the criteria for PDR: rhythmic activity in the posterior leads, reactivity to eye opening, at least two seconds after eye closure with at least one second of stable activity. The frequency was measured using Fast Fourier Transform (linear detrending; frequency resolution: 0.25 Hz; spectral edge: 95%; Hamming window). In sleep-recordings, the PDR frequency was measured both before and after sleep, choosing the latest possible time epoch for measurement after sleep. The PDR frequency was noted as a number with one decimal point. For each recording the diagnostic yield was noted: normal vs. unequivocally abnormal findings. The abnormal findings were further classified as epileptiform discharges (ED) and abnormal slowing (AS). Statistics: normality of PDR-value distribution was assessed using Kolmogorov-Smirnov test. PDR values measured in the wake and sleep-recordings for each patient, were compared using paired t-test. Difference in age between the sub-groups was assessed using t-test for independent samples. Difference in gender between the sub-groups was assessed using Chi-square test. To assess the possible correlation between the change in PDR frequency and age we used Pearson correlation coefficient r.
3. Results The frequency of the PDR (Supplementary Document 1) was significantly lower when measured in sleeprecordings (mean before sleep: 9.96 Hz, after sleep: 9.7 Hz) compared to standard recordings (mean: 10.35Hz) (p<0.001 for all comparisons).
7
Out of 303 standard recordings, 156 (51%) were normal and 147 (49%) had abnormal findings (Supplementary Documents 2 and 3). EDs were observed in 96 recordings and AS in 106 recordings (55 patients had both EDs and AS). Out of the 156 patients with normal standard EEG, 35 patients (22%) had abnormal findings in the sleeprecording: EDs in 29 patients and AS in 13 patients (7 patients had both slowing and EDs). Out of the 147 patients with abnormal standard recordings, in 16 patients (11%) the abnormalities were not present in sleep-recordings (Supplementary Documents 2 and 3). In 41 standard recordings only IEDs were seen; in six of them (15%) the epileptiform activity was not present in sleep (in 2 of those cases the epileptiform activity was a photoparoxysmal response in the standard EEG). In 51 standard EEGs only AS was seen and in 55 cases both slowing and EDs were seen. In seven (14%) and respectively three (5%) cases, these abnormalities were not identified in the sleep EEG. Out of the 51 patients who had only slowing in their standard recordings, 43% (n=22) had EDs in their sleep-recordings (either EDs-only (n=2) or both EDs and slowing (n=20)). Twenty-nine out of the 156 patients with normal standard EEGs (19%) had EDs in their sleep-recordings (EDs-only in 22 patients, both EDs and slowing in seven patients). There was no significant correlation between change in PDR frequency and age. There was no difference in age and gender between the sub-groups in whom abnormalities appeared during sleep recordings and those in whom the abnormalities observed in standard wake recordings were not present in sleep recordings.
4. Discussion We found that the frequency of the PDR measured in the wake period before and after sleep, in sleepdeprived patients is significantly lower than the PDR measured in standard-wake recordings. This can potentially lead to erroneous clinical interpretation of the PDR, when standard-wake recordings are not
8
available. The reduced frequency is probably due to slight drowsiness that is present in the sleep-deprived patients both before the recorded (relatively short) period of sleep, and after it. Due to the large interindividual variability of the PDR frequency values, there was an overlap between the two recording conditions (standard-wake and sleep). However, the clinically relevant aspect in this study is the difference between the PDR values measured in standard-wake and in sleep recordings of the same patients. Therefore, we used paired t-test, for statistical comparison between the two conditions, to compare PDRs of the same patients in the two conditions. The difference was statistically significant (p<0.001). In patients with normal standard recordings, sleep-EEG increased the diagnostic yield by 22% in our series. However, this can merely be a sampling effect (i.e. adding a new recording to the diagnostic workup). The activation of EDs by sleep-deprivation/ sleep-recordings varies between 7-83% in different studies (Bubien and Rudkowska, 1984; Degen, 1980; Degen and Degen, 1981; Fountain et al., 1998; Klingler and Tragner, 1982; Roupakiotis et al., 2000; Rowan et al., 1982) depending on the study population and the duration of the recordings. The activation rates include the sampling effect, which varies between 9-22% in different studies (Roupakiotis et al., 2000; Pratt et al., 1986; Salinsky et al., 1987). It is still disputed whether this activation is merely due to drowsiness and sleep (Degen et al., 1987; Degen and Degen, 1991; Drake et al., 1990; Ellingson et al., 1984; Marinig et al., 2000; Roupakiotis et al., 2000; Scollo-Lavizzari et al., 1977; Tartara et al., 1980; Veldhuizen et al., 1983) or whether sleep deprivation per se has an activating effect (Marinig et al., 2000; Degen, 1980; Klinger et al., 1991; Mattson et al., 1965; Pratt et al., 1968; Rodin et al., 1962; Fountain et al., 1998; Ellingson et al., 1984; Naitoh and Dement, 1976). Our study also showed that in 11% of patients, the abnormalities observed in standard-wake recordings were not present during sleep; for EDs this was 13%. Sleep is a powerful modulator of EDs: while in some patients it activates them, in other patients sleep has an inhibitory effect on the occurrence of EDs (Beniczky et al, 2012). If standard EEG would be omitted from the diagnostic work-up, these abnormalities might go unnoticed. However, we cannot exclude that these changes are also attributed to sampling effect.
9
Quantitative analysis (histograms) of occurrence of epileptiform discharges during long-term video EEG monitoring demonstrated considerable spontaneous fluctuation of the EEG abnormalities (Scherg et al., 2012). Therefore, at least a part of the difference in diagnostic yield between the standard-wake and the sleep recordings can be attributed to the sampling effect. This is an important limitation of this study. Another clinically relevant observation is, that observing AS (without EDs) in standard EEG recordings, increased the likelihood of finding EDs in the sleep-recording, as opposed to after a normal standard EEG (43% versus 19%). This is consistent with previous studies (Herigstad et al., 2001; Carpay et al., 1997). The findings in our study were not related to age and gender. However, we emphasise that we only included adult patients into this study. In conclusion: PDR is significantly slower in the wake periods of sleep-recordings, compared to standard wake recordings. There is a need for both standard-wake and sleep EEGs to optimize the accuracy and the diagnostic yield of EEGs. Last, but not least, it must be remembered that hyperventilation and photic stimulation are procedures which require an alert patient. Active eye closure after the onset of the train of flashes is most provocative (Kasteleijn-Nolst Trenité et al., 2012), and this requires a fully cooperating patient.
Disclosure of conflicts of interest None of the authors has conflicts of interests related to this work.
10
References Beniczky S, Guaranha MS, Conradsen I, Singh MB, Rutar V, Lorber B et al. Modulation of epileptiform EEG discharges in juvenile myoclonic epilepsy: an investigation of reflex epileptic traits. Epilepsia 2012;53:832-9. doi: 10.1111/j.1528-1167.2012.03454. Beniczky S, Aurlien H, Brøgger JC, Hirsch LJ, Schomer DL, Trinka E et al. Standardized computerbased organized reporting of EEG: SCORE - Second version. Clin Neurophysiol 2017; pii: S13882457(17)30906-9. doi: 10.1016/j.clinph.2017.07.418. Bennett DR. Sleep deprivation and major motor colvulsions. Neurology 1963;13:953-8. Bennett DR, Mattson RH, Ziter FA, Calverley JR, Liske EA, Pratt KL. Sleep deprivation: neurological and electroencephalographic effects. Aerospace Medicine 1964;35:888–890. Bennett DR, Ziter FA, Liske EA. Electroencephalographic study of sleep deprivation in flying personnel. Neurology (Minneapolis) 1969;19:375–377. Bubien M, Rudkowska A. 24-hour sleep deprivation as a method of activating EEG signs in epilepsy. Neurologia i Neurochirurgia Polska 1984;18:517–522. Burr W, Körner E, Stefan H. Circadian distribution of generalized spike-wave activity in relation to sleep. In: Degen R, Rodin EA, editors. Epilepsy, sleep and sleep deprivation. 2nd ed. (Epilepsy Res, Suppl.2). Amsterdam: Elsevier Science Publishers BV; 1991. p. 121-135. Carpay JA, de Weerd AW, Schimsheimer RJ, Stroink H, Brouwer OF, Peters AC et al. The Diagnostic Yield of a Second EEG After Partial Sleep Deprivation: A Prospective Study in Children with Newly Diagnosed Seizures. Epilepsia 1997;38:595-599. Craciun L, Gardella E, Alving J, Terney D, Mindruta I, Zarubova J, et al. How long shall we record electroencephalography? Acta Neurol Scand. 2014;129:e9-e11.doi: 10.1111/ane.12186. Declerck AC. Interaction sleep and epilepsy. Eur Neurol 1986;25:117-127.
11
Degen R. A study of the diagnostic value of waking and sleep EEGs after sleep deprivation in epileptic patients on anticonvulsive therapy. Electroencephalogr Clin Neurophysiol 1980;49:577-84. Degen R, Degen HE. A comparative study of the diagnosis value of drug-induced sleep EEGs and sleep EEGs following sleep deprivation in patients with complex partial seizures. J Neurol 1981;225:85-93. Degen R, Degen HE, Reker M. Sleep EEG with or without sleep deprivation? Does sleep deprivation activate more epileptic activity in patients suffering from different types of epilepsy? Eur Neurol 1987;26:51-9. Degen R, Degen HE. Sleep and sleep deprivation in epileptology. In: Degen R, Rodin EA, editors. Epilepsy, sleep and sleep deprivation, 2nd ed.(Epilepsy Res, Suppl. 2). Amsterdam: Elsevier Science Publishers BV; 1991. p. 235-260. DeRoos S, Chillag K, Keeler M, Gilbert DL. Effects of Sleep Deprivation on the Pediatric Electroencephalogram. Pediatrics 2009;123:703-8. doi: 10.1542/peds.2008-0357. Drake ME, Pakalnis A, Phillips BB, Denio LS. Sleep and sleep deprived EEG in partial and generalized epilepsy. Acta Neurol Belg 1990;90:11-9. Ellingson RJ, Wilken K, Bennett DR. Efficacy of sleep deprivation as an activation procedure in epilepsy patients. J Clin Neurophysiol 1984;1:83-101. Ferreira C, Deslandes A, Moraes H, Cagy M, Pompeu F, Basile LF et al. Electroencephalographic changes after one night of sleep deprivation. Arq Neuropsiquiatr 2006;64:388-393. Fountain NB, Kim JS, Lee SI. Sleep Deprivation Activates Epileptiform Discharges Independent of the Activating Effects of Sleep. J Clin Neurophysiol 1998;15:69-75. Gilbert DL, DeRoos S, Bare MA. Does Sleep or Sleep Deprivation Increase Epileptiform Discharges in Pediatric Electroencephalograms? Pediatrics 2004;114:658-62.
12
Geller MR, Gourdji N, Christoff N, Fox E. The effects of sleep deprivation on the EEGs of epileptic children. Dev Med Child Neurol 1969;11:771–6. Gunderson CH, Dunne PB, Feyer TL. Sleep deprivation seizures. Neurology (Minneapolis) 1973;23:678–686. Herigstad A, Michler RP, Sand T, Trodnem K. EEG etter søvndeprivering hos pasienter med epilepsisuspekte anfall. Tiddskr Nor Lægefor 2001;29:3387-90. Janz D. The grand mal epilepsies and the sleeping-waking cycle. Epilepsia 1962;3:69-109. Jovanovic, UJ. General considerations of sleep and sleep deprivation. In: Degen R, Rodin EA, editors. Epilepsy, Sleep and Sleep Deprivation. 2nd ed. Amsterdam: Elsevier; 1991. p. 205–215. Kasteleijn-Nolst Trenité D, Rubboli G, Hirsch E, Martins da Silva A, Seri S, Wilkins A et al. Methodology of photic stimulation revisited: updated European algorithm for visual stimulation in the EEG laboratory. Epilepsia. 2012;53:16-24.Klingler D, Trägner H. Sleep deprivation as activating procedure in EEG of patients with and without epileptic seizures I. Generalized paroxysmal discharges II. Focal discharges. Wiener Medizinische Wochenschrift 1982;132:255–261,263–267. Klinger D, Trägner H, Deisenhammer E. The nature of influence of sleep deprivation on the EEG. In: Degen R, Rodin EA, editors. Epilepsy, sleep and sleep deprivation. 2nd ed. Amsterdam: Elsevier;1991:231–3. Leach JP, Stephen LJ, Salveta C, Brodie MJ. Which electroencephalography (EEG) for epilepsy? The relative usefulness of different EEG protocols in patients with possible epilepsy. J Neurol Neurosurg Psychiatry 2006;77:1040–1042. Logothetis J, Milonas I, Bostantzopoulou S. Sleep deprivation as a method of EEG activation. Observation on the incidence of positive responses. Eur Neurol 1986;25:134 –40. Marinig R, Valente M, Bergonzi P. Sleep and sleep deprivation as EEG activating methods. Clin Neurophysiol 2000;111 Suppl 2:S47-53.
13
Martins da Silva A, Aarts JHP, Binnie CD, Laxminarayan R, Lopes da Silva FH, Meijer JW, et al. Circadian distribution of interictal epileptiform EEG activity. Electroenceph Clin Neurophysiol 1984;58:1-13. Mattson RH, Pratt KL, Calverley JR. Electroencephalograms of epileptics following sleep deprivation. Arch Neurol 1965;13:310 –5. Molaie M, Cruz A. The effect of sleep deprivation on the rate of focal interictal epileptiform discharges. Electroencephalogr Clin Neurophysiol 1988;70:288 –92. Naitoh P, Dement W. Sleep deprivation in humans. In: Remond A, editor in chief. Handbook of Electroencephalography and Clinical Neurophysiology. Amsterdam: Elsevier;1976:7A. p.146–151. Niedermeyer, E. Sleep and EEG. In: Electroencephalography. Basic Principles, Clinical Applications and Related Fields. 3rd ed. (Eds E. Niedermeyer and F. Lopez da Silva). Williams & Wilkins; 1993:153–165. Pratt KL, Mattson RH, Weikers NJ, Williams R. EEG activation of epileptics following sleep deprivation: a prospective study of 114 cases. Electroencephalogr Clin Neurophysiol 1968;24:11-5. Rodin EA, Luby ED, Gottlieb JS. The electroencephalogram during prolonged experimental sleep deprivation. Electroencephalogr Clin Neurophysiol 1962;14:544-51. Roupakiotis SC, Gatzonis SD, Triantafyllou N, Mantouvalos V, Chioni A, Zournas C et al. The usefulness of sleep and sleep deprivation as activating methods in electroencephalographic recording. Contribution to a long-standing discussion. Seizure 2000;9:580–584. Rowan AJ, Veldhuizen RJ, Nagelkerke NJD. Comparative evaluation of sleep deprivation and sedated sleep EEGs as diagnostic aids in epilepsy. Electroencephalogr Clin Neurophysiol 1982;54: 357 –64. Salinsky M, Kanter R, Dasheiff RM. Effectiveness of multiple EEGs in supporting the diagnosis of epilepsy: an operational curve. Epilepsia 1987;28:331-4.
14
Scherg M, Ille N, Weckesser D, Ebert A, Ostendorf A, Boppel T, et al. Fast evaluation of interictal spikes in long-term EEG by hyper-clustering. Epilepsia 2012;53:1196-204.Scollo-Lavizzari G, Pralle W, Radue EW. Comparative study of efficacy of waking and sleep-recordings following sleep deprivation as an activation method in the diagnosis of epilepsy. Eur Neurol 1977;15:121-3. Seeck M, Koessler L, Bast T, Leijten F, Michel C, Baumgartner C et al. The standardized EEG electrode array of the IFCN. Clin Neurophysiol 2017;128:2070-2077. doi:10.1016/j.clinph.2017.06.254. Tartara A, Moglia A, Manni R, Corbellini C. EEG findings and sleep deprivation. Eur Neurol 1980;19: 330 –4. Veldhuizen R, Binnie CD, Beintema DJ. The effect of sleep deprivation on the EEG in epilepsy. Electroencephalogr Clin Neurophysiol 1983;55:505-12.
S1388-2457(18)30072-5 https://doi.org/10.1016/j.clinph.2018.01.056 CLINPH 2008413
To appear in:
Clinical Neurophysiology
Accepted Date:
26 January 2018
Please cite this article as: Meritam, P., Gardella, E., Alving, J., Terney, D., Cacic Hribljan, M., Beniczky, S., Diagnostic yield of standard-wake and sleep EEG recordings, Clinical Neurophysiology (2018), doi: https://doi.org/ 10.1016/j.clinph.2018.01.056
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
1
Diagnostic yield of standard-wake and sleep EEG recordings
Pirgit Meritam1, Elena Gardella1,2, Jørgen Alving1, 3, Daniella Terney1, Melita Cacic Hribljan3, Sándor Beniczky1,4
1. Department of Clinical Neurophysiology, Danish Epilepsy Centre, Dianalund, Denmark 2. University of Southern Denmark, Odense, Denmark 3. Department of Clinical Neurophysiology, Rigshospitalet, Copenhagen, Denmark 4. Department of Clinical Neurophysiology, Aarhus University Hospital, Aarhus, Denmark
Corresponding author: Professor Sándor Beniczky, Danish Epilepsy Centre and Aarhus University Hospital. Postal address: Visbys Allé 5, 4293 Dianalund, Denmark. Telephone: +45 26981536. Email: [email protected]
2
Abstract Objective: To investigate whether Posterior Dominant Rhythm (PDR) can be reliably assessed in sleep-EEG recordings and to investigate the diagnostic yield of standard-wake and sleep-recordings. Methods: EEG recordings of 303 consecutive patients aged 18-88 years were analyzed. All patients had both standard-wake and sleep-recordings, including patients who had abnormal standard recordings. Melatonin was used in 6% of sleep EEGs, and sleep deprivation in 94%. The mean duration of sleep was 41 minutes. We measured the PDR frequency in standard and sleep-recordings, both before and after sleep. We compared the diagnostic yield of standard-wake and sleep EEG recordings. Results: Compared to standard EEG, sleep-recordings showed a significantly lower PDR frequency, both when measured before and after sleep (p<0.001). One-hundred-fifty-six patients (51%) had normal standard recordings, and 35 of them (22%) had abnormal findings in the sleep-recording. One-hundredforty-seven patients had abnormal standard recordings and in 16 of them (11%) these abnormalities were not present in sleep-recording. Conclusions: PDR is significantly slower in the wake periods of sleep-recordings, compared to standard wake recordings. Significance: Sleep and standard wake recordings are complementary.
Keywords: diagnostic yield, EEG, posterior dominant rhythm, sleep deprivation, standard recording.
3
Highlights
•
The frequency of the Posterior Dominant Rhythm (PDR) is lower in the wake periods of sleeprecordings compared to standard wake EEG.
•
Not all abnormalities identified in standard wake recordings are present in sleep.
•
Sleep and standard wake recordings are complementary.
4
1. Introduction Sleep influences neuronal excitability and thereby occurrence of epileptic discharges (Burr et al., 1991; Declerck, 1986; Janz, 1962; Marinig et al., 2000; Martins da Silva et al., 1984). Several studies showed that in patients suspected for epilepsy, following normal or inconclusive standard-wake EEG, recording sleep EEG increases the diagnostic yield (Bennett et al., 1964, 1969; Gunderson et al. 1973; Jovanovic, 1991; Niedermeyer, 1993; Roupakiotis et al., 2000) by activating epileptiform discharges (Degen, 1980; Geller et al., 1969; Klinger et al., 1991; Logothetis et al. 1986; Mattson et al., 1965; Molaie and Cruz, 1988; Pratt et al., 1968; Rowan et al., 1982; Tartara et al., 1980). Leach et al. (2006) suggested that using sleep deprived EEG as the preferred protocol may reduce the number of EEGs carried out by 45%. On the contrary, two large pediatric studies concluded that sleep deprivation before the recording, and sleep during the recording did not increase the diagnostic yield of EEGs (Gilbert et al., 2004; DeRoos et al., 2009). Furthermore, in a subgroup of patients, sleep had inhibitory effect on the occurrence of epileptiform discharges (Beniczky et al., 2012). Provocation methods, like hyperventilation and photic stimulation, are typically done in standard wake recordings. Sleep deprivation can induce seizures in patients with epilepsy (Bennett et al., 1963, 1964; Gunderson et al., 1973; Rodin et al., 1962) and it can even provoke seizures in 3-5% of patients without epilepsy (Degen and Degen, 1991), concluding that it would be prudent to limit sleep deprivation to patients in whom epilepsy is clinically suspected but who have normal or inconclusive standard EEG recording (Marinig et al., 2000). In healthcare systems that are under continuous pressure to increase efficiency, there is a growing number of requests from referring physicians to skip the standard EEG, and only record sleep EEG, in spite of the conflicting evidence and the possible risks. From a practical point of view, combining wake standard EEG (including provocation methods) with sleep EEG in a single recording will either increase recording time
5
significantly or alternatively diminish the total sleep time in the recording. As the sensitivity of a sleep EEG depends on the duration of sleep, this shortening may be of clinical importance (Craciun et al., 2014). Besides the possible inhibition of epileptiform discharges in a subgroup of patients, and the risk of inducing seizures due to sleep deprivation, skipping standard-wake EEGs raises the question about the reliability of measuring the frequency of the posterior dominant rhythm (PDR) in sleep-recordings, affected by lingering drowsiness. Ferreira et al 2006 showed significant decrease in alfa frequency band after total sleep deprivation with increase in theta frequency in temporal and occipital regions (Ferreira et al., 2006). The clinical significance of these changes however, is unclear. We determined the diagnostic yield of standard-wake and sleep EEG recordings, and we assessed whether sleep deprivation affects the frequency of the PDR, measured in the wake periods of sleep EEG recordings.
2. Methods Standard-wake EEG and sleep EEG recordings of 303 consecutive patients (140 female), aged 18-88 years (mean 42 years) were retrospectively analyzed. Inclusion criteria were existence of both standard-wake and sleep-recording in the database (Beniczky et al., 2017). The patients either had a known epilepsy diagnosis or were suspected of having epilepsy. Most importantly, sleep EEGs were recorded also in patients who had an abnormal standard-wake recording. Standard-wake EEG was always the first recording. EEGs were assessed by certified clinical neurophysiologists who were blinded to all clinical data. Each recording was reviewed by two experts and disagreements were resolved via consensus discussions in the whole group. EEGs were recorded using the standard electrode array of the IFCN (Seeck et al., 2017). In 19 sleeprecordings (6%) melatonin was given prior to the recording. In all other cases (284=94%), the patients were sleep deprived. Duration was 30 minutes for the standard recordings and 60 minutes for sleep-recordings.
6
The mean duration of sleep in sleep-recordings was 41 minutes (range: 3-66 minutes). Provocation methods such as hyperventilation and photostimulation were used in standard recordings. For each patient we measured the frequency of the PDR separately in a standard EEG and a sleep EEG recording (in the latter, during the wake period both before and after sleep). The time interval between the standard and the sleep-recording was no more than 1 year. The PDR frequency was measured in a time epoch that best fulfilled the criteria for PDR: rhythmic activity in the posterior leads, reactivity to eye opening, at least two seconds after eye closure with at least one second of stable activity. The frequency was measured using Fast Fourier Transform (linear detrending; frequency resolution: 0.25 Hz; spectral edge: 95%; Hamming window). In sleep-recordings, the PDR frequency was measured both before and after sleep, choosing the latest possible time epoch for measurement after sleep. The PDR frequency was noted as a number with one decimal point. For each recording the diagnostic yield was noted: normal vs. unequivocally abnormal findings. The abnormal findings were further classified as epileptiform discharges (ED) and abnormal slowing (AS). Statistics: normality of PDR-value distribution was assessed using Kolmogorov-Smirnov test. PDR values measured in the wake and sleep-recordings for each patient, were compared using paired t-test. Difference in age between the sub-groups was assessed using t-test for independent samples. Difference in gender between the sub-groups was assessed using Chi-square test. To assess the possible correlation between the change in PDR frequency and age we used Pearson correlation coefficient r.
3. Results The frequency of the PDR (Supplementary Document 1) was significantly lower when measured in sleeprecordings (mean before sleep: 9.96 Hz, after sleep: 9.7 Hz) compared to standard recordings (mean: 10.35Hz) (p<0.001 for all comparisons).
7
Out of 303 standard recordings, 156 (51%) were normal and 147 (49%) had abnormal findings (Supplementary Documents 2 and 3). EDs were observed in 96 recordings and AS in 106 recordings (55 patients had both EDs and AS). Out of the 156 patients with normal standard EEG, 35 patients (22%) had abnormal findings in the sleeprecording: EDs in 29 patients and AS in 13 patients (7 patients had both slowing and EDs). Out of the 147 patients with abnormal standard recordings, in 16 patients (11%) the abnormalities were not present in sleep-recordings (Supplementary Documents 2 and 3). In 41 standard recordings only IEDs were seen; in six of them (15%) the epileptiform activity was not present in sleep (in 2 of those cases the epileptiform activity was a photoparoxysmal response in the standard EEG). In 51 standard EEGs only AS was seen and in 55 cases both slowing and EDs were seen. In seven (14%) and respectively three (5%) cases, these abnormalities were not identified in the sleep EEG. Out of the 51 patients who had only slowing in their standard recordings, 43% (n=22) had EDs in their sleep-recordings (either EDs-only (n=2) or both EDs and slowing (n=20)). Twenty-nine out of the 156 patients with normal standard EEGs (19%) had EDs in their sleep-recordings (EDs-only in 22 patients, both EDs and slowing in seven patients). There was no significant correlation between change in PDR frequency and age. There was no difference in age and gender between the sub-groups in whom abnormalities appeared during sleep recordings and those in whom the abnormalities observed in standard wake recordings were not present in sleep recordings.
4. Discussion We found that the frequency of the PDR measured in the wake period before and after sleep, in sleepdeprived patients is significantly lower than the PDR measured in standard-wake recordings. This can potentially lead to erroneous clinical interpretation of the PDR, when standard-wake recordings are not
8
available. The reduced frequency is probably due to slight drowsiness that is present in the sleep-deprived patients both before the recorded (relatively short) period of sleep, and after it. Due to the large interindividual variability of the PDR frequency values, there was an overlap between the two recording conditions (standard-wake and sleep). However, the clinically relevant aspect in this study is the difference between the PDR values measured in standard-wake and in sleep recordings of the same patients. Therefore, we used paired t-test, for statistical comparison between the two conditions, to compare PDRs of the same patients in the two conditions. The difference was statistically significant (p<0.001). In patients with normal standard recordings, sleep-EEG increased the diagnostic yield by 22% in our series. However, this can merely be a sampling effect (i.e. adding a new recording to the diagnostic workup). The activation of EDs by sleep-deprivation/ sleep-recordings varies between 7-83% in different studies (Bubien and Rudkowska, 1984; Degen, 1980; Degen and Degen, 1981; Fountain et al., 1998; Klingler and Tragner, 1982; Roupakiotis et al., 2000; Rowan et al., 1982) depending on the study population and the duration of the recordings. The activation rates include the sampling effect, which varies between 9-22% in different studies (Roupakiotis et al., 2000; Pratt et al., 1986; Salinsky et al., 1987). It is still disputed whether this activation is merely due to drowsiness and sleep (Degen et al., 1987; Degen and Degen, 1991; Drake et al., 1990; Ellingson et al., 1984; Marinig et al., 2000; Roupakiotis et al., 2000; Scollo-Lavizzari et al., 1977; Tartara et al., 1980; Veldhuizen et al., 1983) or whether sleep deprivation per se has an activating effect (Marinig et al., 2000; Degen, 1980; Klinger et al., 1991; Mattson et al., 1965; Pratt et al., 1968; Rodin et al., 1962; Fountain et al., 1998; Ellingson et al., 1984; Naitoh and Dement, 1976). Our study also showed that in 11% of patients, the abnormalities observed in standard-wake recordings were not present during sleep; for EDs this was 13%. Sleep is a powerful modulator of EDs: while in some patients it activates them, in other patients sleep has an inhibitory effect on the occurrence of EDs (Beniczky et al, 2012). If standard EEG would be omitted from the diagnostic work-up, these abnormalities might go unnoticed. However, we cannot exclude that these changes are also attributed to sampling effect.
9
Quantitative analysis (histograms) of occurrence of epileptiform discharges during long-term video EEG monitoring demonstrated considerable spontaneous fluctuation of the EEG abnormalities (Scherg et al., 2012). Therefore, at least a part of the difference in diagnostic yield between the standard-wake and the sleep recordings can be attributed to the sampling effect. This is an important limitation of this study. Another clinically relevant observation is, that observing AS (without EDs) in standard EEG recordings, increased the likelihood of finding EDs in the sleep-recording, as opposed to after a normal standard EEG (43% versus 19%). This is consistent with previous studies (Herigstad et al., 2001; Carpay et al., 1997). The findings in our study were not related to age and gender. However, we emphasise that we only included adult patients into this study. In conclusion: PDR is significantly slower in the wake periods of sleep-recordings, compared to standard wake recordings. There is a need for both standard-wake and sleep EEGs to optimize the accuracy and the diagnostic yield of EEGs. Last, but not least, it must be remembered that hyperventilation and photic stimulation are procedures which require an alert patient. Active eye closure after the onset of the train of flashes is most provocative (Kasteleijn-Nolst Trenité et al., 2012), and this requires a fully cooperating patient.
Disclosure of conflicts of interest None of the authors has conflicts of interests related to this work.
10
References Beniczky S, Guaranha MS, Conradsen I, Singh MB, Rutar V, Lorber B et al. Modulation of epileptiform EEG discharges in juvenile myoclonic epilepsy: an investigation of reflex epileptic traits. Epilepsia 2012;53:832-9. doi: 10.1111/j.1528-1167.2012.03454. Beniczky S, Aurlien H, Brøgger JC, Hirsch LJ, Schomer DL, Trinka E et al. Standardized computerbased organized reporting of EEG: SCORE - Second version. Clin Neurophysiol 2017; pii: S13882457(17)30906-9. doi: 10.1016/j.clinph.2017.07.418. Bennett DR. Sleep deprivation and major motor colvulsions. Neurology 1963;13:953-8. Bennett DR, Mattson RH, Ziter FA, Calverley JR, Liske EA, Pratt KL. Sleep deprivation: neurological and electroencephalographic effects. Aerospace Medicine 1964;35:888–890. Bennett DR, Ziter FA, Liske EA. Electroencephalographic study of sleep deprivation in flying personnel. Neurology (Minneapolis) 1969;19:375–377. Bubien M, Rudkowska A. 24-hour sleep deprivation as a method of activating EEG signs in epilepsy. Neurologia i Neurochirurgia Polska 1984;18:517–522. Burr W, Körner E, Stefan H. Circadian distribution of generalized spike-wave activity in relation to sleep. In: Degen R, Rodin EA, editors. Epilepsy, sleep and sleep deprivation. 2nd ed. (Epilepsy Res, Suppl.2). Amsterdam: Elsevier Science Publishers BV; 1991. p. 121-135. Carpay JA, de Weerd AW, Schimsheimer RJ, Stroink H, Brouwer OF, Peters AC et al. The Diagnostic Yield of a Second EEG After Partial Sleep Deprivation: A Prospective Study in Children with Newly Diagnosed Seizures. Epilepsia 1997;38:595-599. Craciun L, Gardella E, Alving J, Terney D, Mindruta I, Zarubova J, et al. How long shall we record electroencephalography? Acta Neurol Scand. 2014;129:e9-e11.doi: 10.1111/ane.12186. Declerck AC. Interaction sleep and epilepsy. Eur Neurol 1986;25:117-127.
11
Degen R. A study of the diagnostic value of waking and sleep EEGs after sleep deprivation in epileptic patients on anticonvulsive therapy. Electroencephalogr Clin Neurophysiol 1980;49:577-84. Degen R, Degen HE. A comparative study of the diagnosis value of drug-induced sleep EEGs and sleep EEGs following sleep deprivation in patients with complex partial seizures. J Neurol 1981;225:85-93. Degen R, Degen HE, Reker M. Sleep EEG with or without sleep deprivation? Does sleep deprivation activate more epileptic activity in patients suffering from different types of epilepsy? Eur Neurol 1987;26:51-9. Degen R, Degen HE. Sleep and sleep deprivation in epileptology. In: Degen R, Rodin EA, editors. Epilepsy, sleep and sleep deprivation, 2nd ed.(Epilepsy Res, Suppl. 2). Amsterdam: Elsevier Science Publishers BV; 1991. p. 235-260. DeRoos S, Chillag K, Keeler M, Gilbert DL. Effects of Sleep Deprivation on the Pediatric Electroencephalogram. Pediatrics 2009;123:703-8. doi: 10.1542/peds.2008-0357. Drake ME, Pakalnis A, Phillips BB, Denio LS. Sleep and sleep deprived EEG in partial and generalized epilepsy. Acta Neurol Belg 1990;90:11-9. Ellingson RJ, Wilken K, Bennett DR. Efficacy of sleep deprivation as an activation procedure in epilepsy patients. J Clin Neurophysiol 1984;1:83-101. Ferreira C, Deslandes A, Moraes H, Cagy M, Pompeu F, Basile LF et al. Electroencephalographic changes after one night of sleep deprivation. Arq Neuropsiquiatr 2006;64:388-393. Fountain NB, Kim JS, Lee SI. Sleep Deprivation Activates Epileptiform Discharges Independent of the Activating Effects of Sleep. J Clin Neurophysiol 1998;15:69-75. Gilbert DL, DeRoos S, Bare MA. Does Sleep or Sleep Deprivation Increase Epileptiform Discharges in Pediatric Electroencephalograms? Pediatrics 2004;114:658-62.
12
Geller MR, Gourdji N, Christoff N, Fox E. The effects of sleep deprivation on the EEGs of epileptic children. Dev Med Child Neurol 1969;11:771–6. Gunderson CH, Dunne PB, Feyer TL. Sleep deprivation seizures. Neurology (Minneapolis) 1973;23:678–686. Herigstad A, Michler RP, Sand T, Trodnem K. EEG etter søvndeprivering hos pasienter med epilepsisuspekte anfall. Tiddskr Nor Lægefor 2001;29:3387-90. Janz D. The grand mal epilepsies and the sleeping-waking cycle. Epilepsia 1962;3:69-109. Jovanovic, UJ. General considerations of sleep and sleep deprivation. In: Degen R, Rodin EA, editors. Epilepsy, Sleep and Sleep Deprivation. 2nd ed. Amsterdam: Elsevier; 1991. p. 205–215. Kasteleijn-Nolst Trenité D, Rubboli G, Hirsch E, Martins da Silva A, Seri S, Wilkins A et al. Methodology of photic stimulation revisited: updated European algorithm for visual stimulation in the EEG laboratory. Epilepsia. 2012;53:16-24.Klingler D, Trägner H. Sleep deprivation as activating procedure in EEG of patients with and without epileptic seizures I. Generalized paroxysmal discharges II. Focal discharges. Wiener Medizinische Wochenschrift 1982;132:255–261,263–267. Klinger D, Trägner H, Deisenhammer E. The nature of influence of sleep deprivation on the EEG. In: Degen R, Rodin EA, editors. Epilepsy, sleep and sleep deprivation. 2nd ed. Amsterdam: Elsevier;1991:231–3. Leach JP, Stephen LJ, Salveta C, Brodie MJ. Which electroencephalography (EEG) for epilepsy? The relative usefulness of different EEG protocols in patients with possible epilepsy. J Neurol Neurosurg Psychiatry 2006;77:1040–1042. Logothetis J, Milonas I, Bostantzopoulou S. Sleep deprivation as a method of EEG activation. Observation on the incidence of positive responses. Eur Neurol 1986;25:134 –40. Marinig R, Valente M, Bergonzi P. Sleep and sleep deprivation as EEG activating methods. Clin Neurophysiol 2000;111 Suppl 2:S47-53.
13
Martins da Silva A, Aarts JHP, Binnie CD, Laxminarayan R, Lopes da Silva FH, Meijer JW, et al. Circadian distribution of interictal epileptiform EEG activity. Electroenceph Clin Neurophysiol 1984;58:1-13. Mattson RH, Pratt KL, Calverley JR. Electroencephalograms of epileptics following sleep deprivation. Arch Neurol 1965;13:310 –5. Molaie M, Cruz A. The effect of sleep deprivation on the rate of focal interictal epileptiform discharges. Electroencephalogr Clin Neurophysiol 1988;70:288 –92. Naitoh P, Dement W. Sleep deprivation in humans. In: Remond A, editor in chief. Handbook of Electroencephalography and Clinical Neurophysiology. Amsterdam: Elsevier;1976:7A. p.146–151. Niedermeyer, E. Sleep and EEG. In: Electroencephalography. Basic Principles, Clinical Applications and Related Fields. 3rd ed. (Eds E. Niedermeyer and F. Lopez da Silva). Williams & Wilkins; 1993:153–165. Pratt KL, Mattson RH, Weikers NJ, Williams R. EEG activation of epileptics following sleep deprivation: a prospective study of 114 cases. Electroencephalogr Clin Neurophysiol 1968;24:11-5. Rodin EA, Luby ED, Gottlieb JS. The electroencephalogram during prolonged experimental sleep deprivation. Electroencephalogr Clin Neurophysiol 1962;14:544-51. Roupakiotis SC, Gatzonis SD, Triantafyllou N, Mantouvalos V, Chioni A, Zournas C et al. The usefulness of sleep and sleep deprivation as activating methods in electroencephalographic recording. Contribution to a long-standing discussion. Seizure 2000;9:580–584. Rowan AJ, Veldhuizen RJ, Nagelkerke NJD. Comparative evaluation of sleep deprivation and sedated sleep EEGs as diagnostic aids in epilepsy. Electroencephalogr Clin Neurophysiol 1982;54: 357 –64. Salinsky M, Kanter R, Dasheiff RM. Effectiveness of multiple EEGs in supporting the diagnosis of epilepsy: an operational curve. Epilepsia 1987;28:331-4.
14
Scherg M, Ille N, Weckesser D, Ebert A, Ostendorf A, Boppel T, et al. Fast evaluation of interictal spikes in long-term EEG by hyper-clustering. Epilepsia 2012;53:1196-204.Scollo-Lavizzari G, Pralle W, Radue EW. Comparative study of efficacy of waking and sleep-recordings following sleep deprivation as an activation method in the diagnosis of epilepsy. Eur Neurol 1977;15:121-3. Seeck M, Koessler L, Bast T, Leijten F, Michel C, Baumgartner C et al. The standardized EEG electrode array of the IFCN. Clin Neurophysiol 2017;128:2070-2077. doi:10.1016/j.clinph.2017.06.254. Tartara A, Moglia A, Manni R, Corbellini C. EEG findings and sleep deprivation. Eur Neurol 1980;19: 330 –4. Veldhuizen R, Binnie CD, Beintema DJ. The effect of sleep deprivation on the EEG in epilepsy. Electroencephalogr Clin Neurophysiol 1983;55:505-12.