ArticlePDF Available

Abstract

Objective: To determine predictors of focal to bilateral tonic-clonic seizures (FBTCS) during video-electroencephalography (EEG) monitoring (VEM). Methods: All adult patients undergoing presurgical VEM from 2014 to 2015 in the department of epileptology were eligible (N = 229). Those with refractory focal epilepsy and epileptic seizures recorded during VEM were analyzed (N = 188, Group 1). To assess the effects of antiepileptic drug (AED) taper, the total AED load was calculated as the sum of the ratios of prescribed daily dose and defined daily dose of all AEDs per VEM day and was correlated with the occurrence of focal seizures without bilateral tonic-clonic seizures (FwoBTCS) and FBTCS. To validate the findings, data of patients undergoing VEM in 2004 and 2005 (Group 2, eligible N = 243, analyzed N = 203) were also investigated. Results: In Group 1, 53 patients had FBTCS and 135 patients had exclusively FwoBTCS during VEM. Reduced AED load at seizure onset was the most important modifiable risk factor for FBTCS (receiver-operating characteristic [ROC]: area under the curve [AUC] = 0.78). Furthermore, the risk of FBTCS varied with the history and frequency of FBTCS prior to VEM. For instance, patients had a 50% risk of FBTCS by reducing the AED load to ~20% when no information about history of FBTCS was taken into account, to ~30% when a positive history of FBTCS was taken into account, and to ~50% when a high frequency of FBTCS prior to VEM was taken into account. These findings were largely replicated in Group 2 (59 patients with FBTCS and 144 exclusively with FwoBTCS). Significance: The risk of FTBCS during VEM depends on the history and frequency of FTBCS prior to VEM and is particularly associated with the extent of AED reduction. Our data underscore the need for appropriate tapering regimens in VEM units.
Epilepsia. 2020;00:1–9.
|
1
wileyonlinelibrary.com/journal/epi
Received: 16 July 2019
|
Revised: 24 January 2020
|
Accepted: 30 January 2020
DOI: 10.1111/epi.16454
FULL-LENGTH ORIGINAL RESEARCH
Predictors of focal to bilateral tonic-clonic seizures during long-
term video-EEG monitoring
Max C.Pensel1,2
|
MartinSchnuerch3
|
Christian E.Elger2
|
RainerSurges2
This is an open access article under the terms of the Creative Commons Attribution-NonCommercial License, which permits use, distribution and reproduction in any medium,
provided the original work is properly cited and is not used for commercial purposes.
© 2020 The Authors. Epilepsia published by Wiley Periodicals, Inc. on behalf of International League Against Epilepsy
1Department of Psychiatry, University
Hospital of Bonn, Bonn, Germany
2Department of Epileptology, University
Hospital of Bonn, Bonn, Germany
3RTG Statistical Modeling in Psychology,
Department of Psychology, University of
Mannheim, Mannheim, Germany
Correspondence
Rainer Surges, Department of Epileptology,
University Hospital Bonn, Venusberg-
Campus 1, 53127 Bonn, Germany.
Email: rainer.surges@ukbonn.de
Abstract
Objective: To determine predictors of focal to bilateral tonic-clonic seizures
(FBTCS) during video–electroencephalography (EEG) monitoring (VEM).
Methods: All adult patients undergoing presurgical VEM from 2014 to 2015 in the
department of epileptology were eligible (N=229). Those with refractory focal epi-
lepsy and epileptic seizures recorded during VEM were analyzed (N=188, Group
1). To assess the effects of antiepileptic drug (AED) taper, the total AED load was
calculated as the sum of the ratios of prescribed daily dose and defined daily dose
of all AEDs per VEM day and was correlated with the occurrence of focal seizures
without bilateral tonic-clonic seizures (FwoBTCS) and FBTCS. To validate the find-
ings, data of patients undergoing VEM in 2004 and 2005 (Group 2, eligible N=243,
analyzed N=203) were also investigated.
Results: In Group 1, 53 patients had FBTCS and 135 patients had exclusively
FwoBTCS during VEM. Reduced AED load at seizure onset was the most impor-
tant modifiable risk factor for FBTCS (receiver-operating characteristic [ROC]: area
under the curve [AUC] = 0.78). Furthermore, the risk of FBTCS varied with the
history and frequency of FBTCS prior to VEM. For instance, patients had a 50% risk
of FBTCS by reducing the AED load to ~20% when no information about history
of FBTCS was taken into account, to ~30% when a positive history of FBTCS was
taken into account, and to ~50% when a high frequency of FBTCS prior to VEM was
taken into account. These findings were largely replicated in Group 2 (59 patients
with FBTCS and 144 exclusively with FwoBTCS).
Significance: The risk of FTBCS during VEM depends on the history and frequency
of FTBCS prior to VEM and is particularly associated with the extent of AED reduc-
tion. Our data underscore the need for appropriate tapering regimens in VEM units.
KEYWORDS
antiepileptic drugs, assessment for epilepsy surgery, patient safety
2
|
PENSEL Et aL.
1
|
INTRODUCTION
Video–electroencephalography (EEG) monitoring (VEM) is
a widely used method in the diagnostic assessment for epi-
lepsy surgery. To facilitate the occurrence of focal seizures,
it is common practice in epilepsy-monitoring units (EMUs)
to taper off antiepileptic drugs (AEDs). At present, specific
protocols for AED load reductions are lacking in most epi-
lepsy centers.1,2 During VEM, a proportion of 24%-57%
of patients develop focal to bilateral tonic-clonic seizures
(FBTCS).3 Some semiologic signs (eg, forced deviation of
head and eyes contralateral to the putative hemisphere of
seizure onset) during FBTCS may provide additional infor-
mation in the presurgical assessment. However, FBTCS are
usually not intended, as they are associated with significant
risks to patients’ safety. Seizure-related falls occur in 1%–
20% of patients in EMUs, sometimes with severe injuries
due to FBTCS.4‒6 Furthermore, FBTCS are associated with
apnea and subsequent hypoxemia, in some cases leading to
bradycardia, asystole, and sudden unexpected death in epi-
lepsy (SUDEP).7,8 Although FBTCS in EMUs are likely to
be detected earlier than in other environments, they are none-
theless a major source of complications. Therefore, it would
be of considerable value to identify factors that modify the
risk of their occurrence.
We hypothesize that the risk of FBTCS during VEM is
influenced by the patients’ history and frequency of previous
FBTCS and, more importantly, by AED load reduction. To
investigate these hypotheses, various individual and clinical
factors as well as daily AED doses were retrospectively as-
sessed in two groups of patients who underwent VEM for the
evaluation of epilepsy surgery.
2
|
METHODS
2.1
|
Study design and population
The study investigated risk factors for FBTCS during VEM
in two cohorts of patients, assessed for possible epilepsy sur-
gery in an EMU, separated by one decade. Strengthening the
reporting of observational studies in Epidemiology criteria
for case-control studies, the RECORD statement for obser-
vational studies using routinely collected health data, and
the STARD list for reporting diagnostic accuracy studies
were applied, according to the EQUATOR reporting guide-
lines (available at http://www.equat or-netwo rk.org, accessed
12/10/2019).
We primarily investigated consecutive adult (18years)
patients with refractory focal epilepsy who had seizures
during VEM in the EMU of the University Hospital Bonn,
Germany, Department of Epileptology in the years 2014 and
2015. In a second step, patients of 2004 and 2005 were also
analyzed, to test the reliability of our results. All patients who
underwent VEM for presurgical assessment in the respective
periods were examined for eligibility. The standard battery
of presurgical assessment includes cerebral magnetic reso-
nance imaging (MRI), VEM using noninvasive scalp EEG
or invasive EEG, and neuropsychological testing. In VEM,
live video is constantly recorded to correlate patient's be-
havior and semiological signs with the regional onset and
propagation of seizure activity. Patients with idiopathic gen-
eralized epilepsy syndromes or psychogenic nonepileptic
seizures only were excluded. We also excluded patients with
incomplete data records, patients without any seizure during
VEM, and multiple cases of the same patients to guarantee
meaningful statistical analysis. Furthermore, patients were
excluded who did not take any AED at all or AED without
available defined daily dose (DDD) values. In addition, pa-
tients whose seizures could not be categorized as FwoBTCS
or FBTCS were excluded.
2.2
|
Outcome measures
Patients with focal without bilateral tonic-clonic seizures
(FwoBTCS) only during VEM served as controls for patients
with at least one FBTCS during VEM. The AED load for each
day during VEM was assessed as the sum of the ratios of pre-
scribed daily dose (PDD) per DDD for all AEDs of the pa-
tient, allowing for interindividual comparisons of total drug
loads, irrespective of specific combinations of substances.9
The DDD for individual AEDs are published by the World
Health Organization (WHO) and are accessible online.10‒13
Age, sex, type of EEG recording, history of FBTCS, frequency
of FBTCS (with “high frequency” being defined as occurrence
of at least one FBTCS within the last 3 months prior to VEM,
or an average of at least one FBTCS per 3 months over the
last 12months prior to VEM and clusters over one day being
Key points
• Potential risk factors of focal to bilateral tonic-
clonic seizures (FBTCS) during video–electro-
encephalography (EEG) monitoring (VEM) were
investigated.
• Important predisposing factors were history and
frequency of FBTCS prior to VEM.
• The antiepileptic drug (AED) load reduction was
the most important modifiable risk factor for
FBTCS during VEM.
• The risk of potentially deleterious FBTCS might
be reduced by appropriate AED tapering regimens
according to the individual seizure history.
|
3
PENSEL Et aL.
treated as one seizure), MRI lesions, localization of seizure
onset, state of vigilance at seizure onset, the AED load at VEM
start and at the time point of the first FwoBTCS or first FBTCS
(absolute and relative to VEM start), as well as the velocity of
AED reduction to first seizure (relative change of AED load
per day) were assessed (Table S1). Status epilepticus was not
an outcome measure and we did not assess history of status
epilepticus as a possible risk factor for FBTCS.
2.3
|
Data and statistical methods
Medical records were reviewed and relevant data were ex-
tracted and pseudonymized. The investigators had full access
to the data of all patients. Possible bias might arise from taper-
ing regimens that were not standardized but depended on the
individual decisions of the treating physicians. Furthermore,
tapering regimens might have been different for specific types
of AED. To circumvent these sources of bias, we analyzed data
from two patient groups separated by one decade with other
treating physicians and an expectedly different distribution of
AED. The study size was determined by the number of patients
investigated with the method of VEM during the respective
time intervals. Analyses were performed using IBM SPSS 25.
Group statistics compared patients who had FwoBTCS only (a)
with those who developed at least one FBTCS (b) during VEM,
using Student's t tests and χ2 tests (or Fischer's exact tests) at a
significance level of α=0.05. Furthermore, receiver-operating
characteristic (ROC) and hierarchical logistic regressions were
calculated for specific predictors. In ROC, the area under the
curve (AUC) determines the accuracy of the respective predic-
tors. In the logistic regression, the expected value of the binary
variable “Occurrence of FBTCS” is regressed on the logistic
function of a linear combination of predictors. Thereby, the lin-
ear combination of predictors is mapped to the interval from 0
to 1, and thus the predicted value denotes the probability that an
FBTCS occurs:
P(“Occurrence of FBTCS”)=exp(β0+β1X1+ … +βkXk)/
[1+exp (β0+β1X1+ … +βkXk)].
The regression model's fit is indicated by the deviance,
which under the null hypothesis follows a χ2 distribution with k
(number of predictors) degrees of freedom. In the same vein, a
significant difference in deviance between nested models indi-
cates that the additional predictors in the more complex model
critically contribute to the prediction of the dependent vari-
able.Averaged group data are given as mean ± SD.
This retrospective audit of data collected during standard
clinical care was approved by the local medical ethics com-
mittee (Ethikkommission an der Medizinischen Fakultät der
Rheinischen Friedrich-Wilhelms-Universität Bonn, No. 352/12).
3
|
RESULTS
3.1
|
Study population
In a first step, 229 adult patients undergoing VEM for
pre-surgical assessment in the years 2014 and 2015 were
examined for eligibility(Figure 1). A total of 188 patients
were included in the final analysis (Group 1), of whom
53 had FBTCS (subgroup 1b, absolute AED load at VEM
start=3.03±1.10) and 135 exclusively FwoBTCS (sub-
group 1a, absolute AED load at VEM start=3.58±1.58)
during VEM, resulting in 2.55 controls per case. In a sec-
ond step, 243 adult patients who underwent presurgical
assessment 10years before (years 2004 and 2005) in the
same EMU were also analyzed (Group 2). A total of 203
individuals were included in the final analyses of Group
2, comprising 59 patients with FBTCS (subgroup 2b, ab-
solute AED load at VEM start = 2.49 ± 1.38) and 144
with only FwoBTCS during VEM (subgroup 2a, absolute
AED load at VEM start = 2.79 ± 1.63), which equals
2.44 controls per case. Group 1 and Group 2 did not dif-
fer significantly in age (t= −1.50, df=389, ns), or sex
FIGURE 1 Inclusion and exclusion criteria
4
|
PENSEL Et aL.
2=0.82, df =1, ns), yet distribution of specific AED
was significantly different, Fisher's exact test = 99.86,
df=25, P≤.001 (Figure S1).
3.2
|
Predictors of FBTCS during VEM:
Receiver-operating characteristics
Predictors were only considered reliable if their values
were significantly different in the seizure-type–related sub-
groups (a and b) of both patient groups (1 and 2). Univariate
analyses revealed that “Positive history of FBTCS,” “High
frequency of previous FBTCS,” “Velocity of AED load
reduction,” and “AED load at first seizure” (absolute and
relative to VEM start) were significantly associated with
the occurrence of FBTCS during VEM in both groups of
patients (Table S2), whereas the other predictors were not
(Table S3).
The reliable predictors include non-modifiable and mod-
ifiable factors. By comparing the predictive power of the re-
liable predictors using ROC analyses, “Relative AED load at
first seizure” emerged as the best modifiable predictor (Table
1, Figure S2).
3.3
|
Predictors of FBTCS during VEM:
Logistic regressions
To further evaluate the impact of the aforementioned
modifiable factors, the association between occurrence of
FBTCS during VEM and extent (given in steps of 10%) as
well as velocity (given in steps of 10% per day) of AED
load reduction were stratified according to the history and
frequency of FBTCS prior to VEM, using hierarchical
logistic regression analyses for three different scenarios
(Tables 2‒4). The chosen scenarios are common in every-
day practice, where the clinician is challenged to define an
AED tapering regimen based on the anamnestic informa-
tion given by the patient.
In Scenario 1 (Table 2), history and frequency of FBTCS prior
to VEM are neglected, and only modifiable predictors (“AED
load reduction” and “Velocity of reduction”) are accounted for.
Taken individually, both of these factors are significantly pre-
dicting the occurrence of FBTCS (see χ2 and RN
2 in models
1.1.1 and 1.1.2). Here, exp (β) indicates the multiplicative factor
by which the odds of the dependent variable P(“Occurrence of
FBTCS”) change when the respective predictor value changes
by one unit (AED load reduction of 10%, velocity of reduction
of 10% per day), that is, the odds ratio. It is important to note
that the two modifiable predictors (“AED load reduction” and
“Velocity of reduction”) are significantly correlated with each
other (Group 1: r=.79***, Group 2: r= .64***). Therefore,
we also calculated a combined model (Model 1.2), where both
predictors are included together. In comparison to “Velocity of
AED load reduction” alone, the addition of “AED load reduc-
tion” significantly improved the predictive strength (Model 1.2
vs Model 1.1.1), whereas the combined model did not increase
the predictive strength compared to “AED load reduction” alone
(Model 1.2 vs Model 1.1.2). This finding suggests that “AED
load reduction” is a much stronger predictor than “Velocity of
reduction,” of which the influence in the combined analysis is
no longer significant. In the combined analysis (Model 1.2), the
odds ratio, exp (β), for every 10% reduction per day is calcu-
lated at 0.99 in Group 2, indicating no meaningful influence of
“Velocity of reduction” on FBTCS occurrence at all. In Group
1, the estimation is clearly below 1.00, exp (β) = 0.58, sug-
gesting even a negative influence on FBTCS occurrence, al-
though not a significant one. Taken together, one can conclude
that “AED load reduction” is a better predictor for FBTCS than
“Velocity of reduction.
In Scenario 2, the same analyses were performed after
inclusion of the information on history of FBTCS as a
non-modifiable factor (Table 3). In Group 1, a positive his-
tory of FBTCS (Model 2.1) already significantly increases
the odds for FBTCS by a factor of exp (β)=8.05 (odds ratio
[OR]), corresponding to a relative risk (RR) of 5.47. Similarly,
this holds also true for Group 2 (OR = 5.41, RR = 3.82).
However, by adding the information on AED load reduction,
TABLE 1 Receiver-operating characteristic (ROC): accuracy of
reliable predictors for FBTCS during VEMa
N AUC P95% CI
Non-modifiable predictors of FBTCS during VEM
Positive history of FBTCSb
Group 1 (2014/15) 182 0.63 .004 0.55 0.72
Group 2 (2004/05) 200 0.63 .005 0.55 0.71
High frequency of previous FBTCSc
Group 1 (2014/15) 139 0.79 ≤.001 0.69 0.88
Group 2 (2004/05) 144 0.65 .006 0.54 0.76
Modifiable factors of FBTCS during VEM
Velocity of AED load reduction
Group 1 (2014/15) 188 0.67 ≤.001 0.59 0.75
Group 2 (2004/05) 203 0.69 ≤.001 0.62 0.76
Absolute AED load at first seizure
Group 1 (2014/15) 188 0.78 ≤.001 0.72 0.85
Group 2 (2004/05) 203 0.72 ≤.001 0.65 0.79
Relative AED load at first seizure
Group 1 (2014/15) 188 0.78 ≤.001 0.71 0.85
Group 2 (2004/05) 203 0.79 ≤.001 0.73 0.85
Null hypothesis 0.50
aThe null hypothesis describes a predictor at chance level.
bOnly patients with data regarding history of FBTCS prior to VEM.
cOnly patients with data regarding frequency of FBTCS prior to VEM.
|
5
PENSEL Et aL.
TABLE 2 Scenario 1. Logistic regression without information on history or frequency of FBTCS
Predictors
Group 1 (2014/15)
N=188
Group 2
(2004/05)
N=203
χ2 (df) RN
2
exp (β) [95% CI] ≙
Odds ratio (OR) χ2 (df) RN
2
exp (β) [95% CI] ≙
Odds Ratio (OR)
Model 1.1.1
Constant 14.84 (1)*** 0.11 0.16*** 13.88 (1)*** 0.09 0.22***
Velocity of reduction 2.46 [1.52, 3.99]*** 1.63 [1.22, 2.18]***
Model 1.1.2
Constant 42.04 (1)*** 0.29 0.07*** 43.09 (1)*** 0.27 0.09***
AED load reduction 1.38 [1.24, 1.55]*** 1.36 [1.23, 1.51***
Model 1.2
Constant 43.71 (2)*** 0.30 0.08*** 43.09 (2)*** 0.27 0.09
Velocity of reduction 0.58 [0.25, 1.35] ns 0.99 [0.73, 1.36] ns
AED load reduction 1.48 [1.26, 1.74]*** 1.36 [1.21, 1.54]***
Model comparisons 1.2 vs 1.1.1 1.2 vs 1.1.2 1.2 vs 1.1.1 1.2 vs 1.1.2
χ2 (df) 28.87 (1)*** 1.67 (1) ns 29.22 (1)*** 0.00 (1), ns
Δ RN
20.19 0.01 0.18 0.00
Note: RN
2=Nagelkerke (pseudo-) R2.
AED load reduction [10% reduction of PDD/DDD].
Velocity of reduction [10% reduction of (PDD/DDD)/d].
***P ≤ .001; ns=not significant.
TABLE 3 Scenario 2. Logistic regression including information on history of FBTCS
Predictors
Group 1 (2014/15) N=182 Group 2 (2004/05) N=200
χ2 (df) RN
2
exp (β) [95%CI] ≙
Odds Ratio (OR) χ2 (df) RN
2
exp (β) [95%CI] ≙
Odds Ratio (OR)
Model 2.1
Constant 17.73 (1)*** 0.13 0.07*** 15.54 (1) *** 0.11 0.10***
History of FBTCS 8.05 [2.37, 27.30]*** 5.41 [2.03, 14.43]***
Model 2.2.1
Constant 30.78 (2)*** 0.22 0.03*** 28.77 (2)*** 0.19 0.05***
History of FBTCS 8.31 [2.39, 28.94]*** 6.01 [2.08, 17.38]***
Velocity of reduction 2.47 [1.48, 4.15]*** 1.61 [1.20, 2.15]**
Model 2.2.2
Constant 54.76 (2)*** 0.37 0.01*** 53.33 (2)*** 0.33 0.02***
History of FBTCS 8.51 [2.31, 31.35]** 5.16 [1.79, 14.92]**
AED load reduction 1.38 [1.23, 1.56]*** 1.36 [1.22, 1.51]***
Model 2.3
Constant 55.71 (3)*** 0.38 0.01*** 53.43 (3)*** 0.34 0.02***
History of FBTCS 8.18 [2.22, 30.23]** 5.25 [1.80, 15.27]**
Velocity of reduction 0.66 [0.29, 1.55] ns 1.05 [0.77, 1.44] ns
AED load reduction 1.46 [1.24, 1.72]*** 1.35 [1.19, 1.52]***
Model
comparisons
2.2.1 vs 2.1 2.2.2. vs 2.1 2.3 vs 2.2.1 2.3 vs 2.2.2 2.2.1 vs 2.1 2.2.2. vs 2.1 2.3 vs 2.2.1 2.3 vs 2.2.2
χ2 (df) 13.05 (1)*** 37.03 (1)*** 24.93 (1)*** 0.95 (1) ns 13.23 (1)*** 37.78 (1)*** 24.66 (1)*** 0.10 (1) ns
Δ RN
20.09 0.24 0.15 0.01 0.08 0.23 0.14 0.00
** P ≤ .01; ***P ≤ .001; ns=not significant.
6
|
PENSEL Et aL.
the predictive power for occurrence of FBTCS even signifi-
cantly improves (Model 2.2.2 vs Model 2.1), demonstrating
that the risk of FBTCS to occur during VEM is predicted by
AED load reduction beyond the patients’ history of FBTCS.
Whereas “Velocity of reduction” is also significantly increas-
ing the predictive quality over “History of FBTCS” (Model
2.2.1), in the combined analysis with “AED load reduction”
(Model 2.3), again, its influence is no longer significant. Of
the patients with a negative history of FBTCS, the fraction
that developed FBTCS during VEM was 6.67% in Group 1 (3
of 45) and 9.43% in Group 2 (5 of 53), see Table S1. Logistic
regressions revealed that also in these patients, “AED load
reduction” was a significant predictor for FBTCS occurrence
and excelled “Velocity of reduction” (Table 5).
The analysis was further refined in Scenario 3 for patients
with a positive history of FBTCS by including “Frequency of
previous FBTCS” into the model (Table 4). In this scenario, high
frequency of previous FBTCS is associated with a significantly
greater risk of FBTCS during VEM in both groups (Group 1:
OR= 10.54, RR = 5.18; Group 2: OR =2.51, RR = 1.84).
Again, “AED load reduction” as well as “Velocity of reduction”
increase the predictive quality of the model, but in the com-
bined analysis (Model 3.3) only “AED load reduction” plays
an important role while “Velocity of reduction” no longer does.
We also aimed at quantifying the AED load reductions,
which facilitate occurrence of FBTCS. To that end, we cal-
culated the specific values for “AED load reduction” at a
50% probability of FBTCS (and a corresponding 50% prob-
ability of an FwoBTCS), based on the estimations of the
combined models in the three scenarios. Because “Velocity
of reduction” was shown to be less important than the ac-
tual AED load reduction, the following numbers are based
on the observed mean levels of velocity. According to the
estimation of Model 1.2, the probability of an FBTCS
in Group 1 amounts to 50% when the relative AED load
is reduced to 24.45% of the initial AED load at a mean
TABLE 4 Scenario 3. Logistic regression including information on frequency of previous FBTCS
Predictors
Group 1 (2014/15) N=94 Group 2 (2004/05) N=93
χ2 (df) RN
2
exp (β) [95%CI] ≙ Odds
Ratio (OR) χ2 (df) RN
2
exp (β) [95%CI] ≙
Odds Ratio (OR)
Model 3.1
Constant 23.15
(1)***
0.30 0.12*** 4.26 (1)* 0.06 0.32***
Frequency of previous
FBTCS
10.54 [3.55, 31.34]*** 2.51 [1.04, 6.08]*
Model 3.2.1
Constant 29.81
(2)***
0.38 0.04*** 8.98 (2)* 0.13 0.16***
Frequency of previous
FBTCS
11.06 [3.54, 34.59]*** 2.43 [0.98, 6.04] ns
Velocity of reduction 2.77 [1.22, 6.33]* 1.67 [1.03, 2.68]*
Model 3.2.2
Constant 36.61
(2)***
0.45 0.02*** 19.42 (2)*** 0.26 0.06***
Frequency of previous
FBTCS
11.38 [3.48, 37.19]*** 3.22 [1.19, 8.70]*
AED load reduction 1.33 [1.12, 1.57]*** 1.32 [1.13, 1.54]***
Model 3.3
Constant 36.61
(3)***
0.45 0.02*** 19.54 (3)*** 0.26 0.07***
Frequency of previous
FBTCS
11.38 [3.48, 37.20]*** 3.31 [1.20, 9.10]*
Velocity of reduction 0.99 [0.30, 3.29] ns 0.89 [0.47, 1.70] ns
AED load reduction 1.33 [1.06, 1.66]* 1.35 [1.11, 1.64]**
Model
comparisons
3.2.1 vs 3.1 3.2.2 vs 3.1 3.3
vs
3.2.1
3.3 vs 3.2.2 3.2.1 vs 3.1 3.2.2
vs
3.1
3.3 vs 3.2.1 3.3 vs 3.2.2
χ2 (df) 6.66 (1)* 13.46 (1)*** 6.80 (1)** 0.00 (1) ns 4.73 (1)* 15.16 (1)*** 10.56 (1)** 0.12 (1) ns
Δ RN
20.07 0.14 0.07 . 00 0.07 0.20 0.14 0.00
*P ≤ .05; ** P ≤ .01; ***P ≤ .001; ns=not significant.
|
7
PENSEL Et aL.
velocity of reduction of 8.69% per day. The analysis of
Group 2 yields similar results, that is, an AED load reduc-
tion to 20.34% is associated with a probability of 50% for an
FBTCS to occur during VEM at a mean velocity of 12.15%
per day. Respective calculations were also performed for
Models 2.3 and 3.3, resulting in thresholds for AED load
reduction that represent a probability of 50% for an FBTCS
under defined circumstances of history and frequency of
FBTCS (see Table 5; see also Table S4 for lower FBTCS
probabilities). No meaningful AED load reduction thresh-
olds could be assessed for patients with a known negative
history of FBTCS or a positive history but low frequency
of FBTCS prior to VEM, due to an insufficient number of
these specific cases.
4
|
DISCUSSION
4.1
|
Key findings
In two different patient groups separated by a decade, the
occurrence of FBTCS during VEM depended on the pa-
tient's history and frequency of FBTCS prior to VEM and
was strongly influenced by the extent of AED reduction
during VEM. Crucially, although the amount and the rate
of AED load reduction both mattered, the amount of reduc-
tion was shown to be a better predictor, even when taking
into account history and frequency of previous FBTCS.
Our study suggests that the risk of potentially deleterious
FBTCS in an EMU is associated with AED load reduction
and underscores the need for prospectively tested tapering
protocols.
4.2
|
Limitations
Our retrospective clinical study comes with limitations.
First, AED load reduction regimens were not standard-
ized, but depended on the individual decisions of the
treating physicians and might have been unsystematically
influenced by the patients’ anamnestic information on
prior FBTCS. Second, tapering regimens might have been
different for varying types of AEDs. Both aspects are im-
portant, as AEDs can significantly differ with respect to
pharmacokinetics, mechanisms of action, and drug-drug
interactions on the one hand and efficacy on control of
FBTCS on the other hand.14 Because polypharmacothera-
pies varied considerably between individual patients, the
effects of specific AED properties could not be system-
atically addressed. Being aware of these significant weak-
nesses, however, we have selected and analyzed two patient
groups separated by a decade, assuming that the proportion
of AEDs with distinct properties is very different (as shown
in Figure S1) and that the treating physicians were differ-
ent. It is important to note that only those factors that were
shown to correlate significantly with FBTCS occurrence
during VEM in the two patient groups were considered as
reliable. This approach appeared to us as a reasonable way
of mitigate the effects of individual tapering regimens and
AED properties. Therefore, we are confident that despite
the above-mentioned limitations, our findings are solid
and the conclusions clinically meaningful.
4.3
|
Clinical implications
According to a recent survey, only about one-third of the
responding EMUs had a written policy for AED with-
drawal during VEM.2 Knowledge of the specific impact
of AED load reductions on complications during VEM is
limited, and available data focus mainly on the occurrence
of status epilepticus and seizure clusters.2,15,16 FBTCS
are, however, far more frequent than status epilepticus
during VEM3,6,16 and are significantly related to physical
injuries5,17 and sudden unexpected death in epilepsy (or
SUDEP)8. Previous studies dealing with FBTCS during
VEM included between 54 and 151 patients, with 18% to
57% of them having FBTCS during VEM.3,18‒23 The pro-
portion of people with FBTCS in our study (2014/15: 28%;
2004/05: 29%) was largely within the range reported by
TABLE 5 Threshold values for AED load reductions
Scenario
Relative AED load
Mean ~Group 1 (2014/15) Group 2 (2004/05)
1: No information on prior FBTCS 24.45% 20.34% 22.39% 20%
2: Positive history of FBTCS prior to
VEM
32.28% 28.37% 30.33% 30%
3: High frequency of FBTCS prior to
VEM
52.50% 43.79% 48.15% 50%
Note: The results provide the actual AED load in relation to the AED load at VEM start (relative AED load) at the point of reduction when the probability for FBTCS
becomes equal to the probability of FwoBTCS (0.5 or 50%), stratified by positive history and high frequency of FBTCS prior to VEM. Velocity of AED load reduction
was fixed at the respective group mean value. Results are based on the estimated logistic regression models (Tables 2‒4).
8
|
PENSEL Et aL.
other studies. In previous studies, complete withdrawal of
AED was associated with a twofold increase in FBTCS
in VEM3 and occurred more frequently in patients with a
positive history of FBTCS prior to VEM,24 yet risk factors
were mostly not analyzed in greater detail or not validated
in larger patient groups. However, a recent study investi-
gated the rate of FBTCS in a dichotomous relation to esti-
mated therapeutic vs nontherapeutic drug levels based on
the patient's drug with the longest half-life, without finding
significant effects.25
The present study provides a detailed approach and im-
plies certain levels of AED load reductions that are associated
with an elevated risk of FBTCS during VEM, considering
the patient’s history and frequency of previous FBTCS. As
a quantitative measure we calculated the AED load of each
VEM day and the relationship between the AED load at the
day of seizure onset and the total AED load at VEM start,
which allowed us to group and compare all patient data and
to detect general patterns, irrespective of specific AED com-
binations.26‒28 Without taking into account information on
FBTCS history, a 50% risk of an FBTCS during VEM was
associated with a reduction to ~20% of the initial AED load.
Taking into account a positive history of FBTCS, a 50% risk
of an FBTCS during VEM was associated with a reduction to
~30%, and with taking into account a high frequency of pre-
vious FBTCS, a 50% risk of an FBTCS during VEM was as-
sociated with a reduction to ~50% of the initial AED load (see
Table 5). The velocity of AED reduction during VEM does
not seem to contribute significantly to the risk of FBTCS,
which is in line with previous studies.3,21 However, it must
be acknowledged that due to the retrospective study design, a
more important role of the tapering velocity cannot be ruled
out, particularly concerning extremely high or extremely low
rates of reduction, and may depend on specific AEDs.
In conclusion, we believe that our study provides practical
estimates of AED load reductions that are associated with
an elevated risk of FBTCS during VEM. Prospective studies
are, however, recommended to corroborate our findings and
to further investigate AED load reduction regimens for given
drugs that specifically balance the risk of FBTCS during
VEM and the length of the stay in an EMU.
ACKNOWLEDGMENTS
The authors would like to thank Robert Schnuerch,
University of Bonn, Department of Psychology, for con-
siderable help with statistical analyses. We would also like
to thank the ‘Verein zur Förderung der Epilepsieforschung
e.V.', Bonn, Germany, for travel costs of M.C.P. to present
the results at a congress. M.S. is supported by a grant from
the Deutsche Forschungsgemeinschaft (DFG, GRK 2277)
to the Research Training Group “Statistical Modeling in
Psychology” (SMiP). This research did not receive any fur-
ther grants from the public, commercial, or not-for-profit
sector funding agencies. There was no external funding for
this study.
CONFLICTS OF INTEREST
The authors declare that the research was conducted in the
absence of any commercial or financial relationship that
could be construed as a potential conflict of interest. C.E.E.
has served as a paid consultant for Desitin, Pfizer, and UCB
Pharma. He was an employee of the Life and Brain Institute
Bonn. R.S. has received fees as speaker or consultant from
Bial, Desitin, Eisai, LivaNova, Novartis, and UCB Pharma.
M.C.P. and M.S. have no conflicts of interest to disclose. We
confirm that we have read the Journal's position on issues
involved in ethical publication and affirm that this report is
consistent with those guidelines.
ORCID
Max C. Pensel https://orcid.org/0000-0002-2760-2269
Rainer Surges https://orcid.org/0000-0002-3177-8582
REFERENCES
1. Kobulashvili T, Höfler J, Dobesberger J, Ernst F, Ryvlin P, Cross JH,
et al. Current practices in long-term video-EEG monitoring services:
a survey among partners of the E-PILEPSY pilot network of reference
for refractory epilepsy and epilepsy surgery. Seizure. 2016;38:38–45.
2. Jehi L. Antiepileptic drug management in the epilepsy monitoring
unit: any standards? Epilepsy Curr. 2016;16:116–7.
3. Guld AT, Sabers A, Kjaer TW. Drug taper during long-term vid-
eo-EEG monitoring: efficiency and safety. Acta Neurol Scand.
2017;135:302–7.
4. Rheims S, Ryvlin P. Patients' safety in the epilepsy monitoring unit:
time for revising practices. Curr Opin Neurol. 2014;27:213–8.
5. Dobesberger J, Walser G, Unterberger I, Seppi K, Kuchukhidze G,
Larch J, et al. Video-EEG monitoring: Safety and adverse events in
507 consecutive patients. Epilepsia. 2011;52:443–52.
6. Noe KH, Drazkowski JF. Safety of long-term video-electroenceph-
alographic monitoring for evaluation of epilepsy. Mayo Clin Proc.
2009;84:495–500.
7. Surges R, Thijs RD, Tan HL, Sander JW. Sudden unexpected death
in epilepsy: risk factors and potential pathomechanisms. Nat Rev
Neurol. 2009;5:492–504.
8. Ryvlin P, Nashef L, Lhatoo SD, Bateman LM, Bird J, Bleasel A,
et al. Incidence and mechanisms of cardiorespiratory arrests in
epilepsy monitoring units (MORTEMUS): a retrospective study.
Lancet Neurol. 2013;12:966–77.
9. Deckers CLP, Hekster YA, Keyser A, Meinardi H, Renier WO.
Reappraisal of polytherapy in epilepsy: a critical review of drug
load and adverse effects. Epilepsia. 1997;38:570–5.
10. World Health Organization. International Working Group for Drug
Statistics Methodology. Introduction to drug utilization research.
Geneva: World Health Organization, 2003.
11. World Healths Organization, Collaborating Centre for Drug
Statistics Methodology. ATC/DDD Index 2018. Available from
URL: https ://www.whocc.no/atc_ddd_index/ . Accessed November
28, 2018.
12. Deutsches Institut für Medizinische Dokumentation und
Information, Köln. Anatomisch-therapeutisch-chemische
|
9
PENSEL Et aL.
Klassifikation mit Tagesdosen. Amtliche Fassung des ATC-Index
mit DDD-Angaben für Deutschland im Jahre 2018. Available from
URL: https ://dimdi.de/dynam ic/de/klass i/downl oadce nter/atcdd d/
versi on201 8/. Accessed November 28, 2018.
13. Merlo J, Wessling A, Melander A. Comparison of dose stan-
dard units for drug utilisation studies. Eur J Clin Pharmacol.
1996;50:27–30.
14. Hemery C, Ryvlin P, Rheims S. Prevention of generalized ton-
ic-clonic seizures in refractory focal epilepsy: a meta-analysis.
Epilepsia. 2014;55:1789–99.
15. Rose AB, McCabe PH, Gilliam FG, Smith BJ, Boggs JG, Ficker DM,
et al. Occurrence of seizure clusters and status epilepticus during in-
patient video-EEG monitoring. Neurology. 2003;60:975–8.
16. Haut SR, Swick C, Freeman K, et al. Seizure clustering during ep-
ilepsy monitoring. Epilepsia. 2002;43:711–5.
17. Asadi-Pooya AA, Nikseresht A, Yaghoubi E, Nei M. Physical in-
juries in patients with epilepsy and their associated risk factors.
Seizure. 2012;21:165–8.
18. Al Kasab S, Dawson RA, Jaramillo JL, Halford JJ. Correlation of
seizure frequency and medication down-titration rate during vid-
eo-EEG monitoring. Epilepsy Behav. 2016;64:51–6.
19. Di Gennaro G, Picardi A, Sparano A, Mascia A, Meldolesi GN,
Grammaldo LG, et al. Seizure clusters and adverse events during
pre-surgical video-EEG monitoring with a slow anti-epileptic drug
(AED) taper. Clin Neurophysiol. 2012;123:486–8.
20. Henning O, Baftiu A, Johannessen SI, Landmark CJ. Withdrawal
of antiepileptic drugs during presurgical video-EEG monitoring:
an observational study for evaluation of current practice at a refer-
ral center for epilepsy. Acta Neurol Scand. 2014;129:243–51.
21. Kumar S, Ramanujam B, Chandra PS, Dash D, Mehta S, Anubha
S, et al. Randomized controlled study comparing the efficacy of
rapid and slow withdrawal of antiepileptic drugs during long-term
video-EEG monitoring. Epilepsia. 2018;59:460–7.
22. Rizvi SAA, Hernandez-Ronquillo L, Wu A, Téllez Zenteno
JF. Is rapid withdrawal of anti-epileptic drug therapy during
video EEG monitoring safe and efficacious? Epilepsy Res.
2014;108:755–64.
23. Yen DJ, Chen C, Shih YH, et al. Antiepileptic drug withdrawal in
patients with temporal lobe epilepsy undergoing presurgical vid-
eo-EEG monitoring. Epilepsia. 2001;42:251–5.
24. Swick CT, Bouthillier A, Spencer SS. Seizure occurrence during
long-term monitoring. Epilepsia. 1996;37:927–30.
25. Hartl E, Seethaler M, Lauseker M, Rémi J, Vollmar C, Noachtar S.
Impact of withdrawal of antiepileptic medication on the duration of
focal onset seizures. Seizure. 2019;67:40–4.
26. Canevini MP, De Sarro G, Galimberti CA, Gatti G, Licchetta L,
Malerba A, et al. Relationship between adverse effects of antie-
pileptic drugs, number of coprescribed drugs, and drug load in a
large cohort of consecutive patients with drug-refractory epilepsy.
Epilepsia. 2010;51:797–804.
27. Kitazawa YU, Jin K, Kakisaka Y, Fujikawa M, Tanaka F,
Nakasato N. Predictive factors of higher drug load for seizure
freedom in idiopathic generalized epilepsy: comparison between
juvenile myoclonic epilepsy and other types. Epilepsy Res.
2018;144:20–4.
28. Hampel KG, Gómez-Ibáñez A, Garcés-Sánchez M, Hervás-Marín
D, Cano-López I, González-Bono E, et al. Antiepileptic drug re-
duction and increased risk of stimulation-evoked focal to bilateral
tonic–clonic seizure during cortical stimulation in patients with
focal epilepsy. Epilepsy Behav. 2018;80:104–8.
SUPPORTING INFORMATION
Additional supporting information may be found online in
the Supporting Information section.
How to cite this article: Pensel MC, Schnuerch M,
Elger CE, Surges R. Predictors of focal to bilateral
tonic-clonic seizures during long-term video-EEG
monitoring. Epilepsia. 2020;00:1–9. https ://doi.
org/10.1111/epi.16454
... hours to years in both intensity and spatial distribution. Specifically, while seizures often share common features in the same patient (Burns et al., 2014;Kramer et al., 2010;Schevon et al., 2012;Schindler et al., 2011;Truccolo et al., 2011;Wagner et al., 2015), electrographic seizure activity may change in terms of duration , spatial spread (Karthick, Tanaka, Khoo, & Gotman, 2018;Marciani & Gotman, 1986;Naftulin et al., 2018;Pensel, Schnuerch, Elger, & Surges, 2020), spectral properties (Alarcon, Binnie, Elwes, & Polkey, 1995) from one seizure to the next. Our recent study (Schroeder et al., 2020) has additionally shown that the seizure EEG spatio-temporal evolution from seizure start to seizure termination (or short: 'seizure evolution') also changes from one seizure to the next in the same patient. ...
... Specifically, it is wellknown that AED changes and withdrawal can change the severity and evolutions of seizures. For example, bilateral tonic-clonic seizures are more prevalent when AED levels are reduced (Pensel et al., 2020). ...
Article
Full-text available
Epilepsy is recognised as a dynamic disease, where both seizure susceptibility and seizure characteristics themselves change over time. Specifically, we recently quantified the variable electrographic spatio-temporal seizure evolutions that exist within individual patients. This variability appears to follow subject-specific circadian, or longer, timescale modulations. It is therefore important to know whether continuously recorded interictaliEEG features can capture signatures of these modulations over different timescales. In this study, we analyse continuous intracranial electroencephalographic (iEEG) recordings from video-telemetry units and find fluctuations in iEEG band power over timescales ranging from minutes up to 12 days. As expected and in agreement with previous studies, we find that all subjects show a circadian fluctuation in their iEEG band power. We additionally detect other fluctuations of similar magnitude on subject-specific timescales. Importantly, we find that a combination of these fluctuations on different timescales can explain changes in seizure evolutions in most subjects above chance level. These results suggest that subject-specific fluctuations in iEEG band power over timescales of minutes to days may serve as markers of seizure modulating processes. We hope that future study can link these detected fluctuations to their biological driver(s). There is a critical need to better understand seizure modulating processes, as this will enable the development of novel treatment strategies that could minimise the seizure spread, duration or severity and therefore the clinical impact of seizures.
... It is also well-known that AED changes and withdrawal can change the severity and dynamics of seizures. For example, bilateral tonic-clonic seizures are more prevalent when AED levels are reduced [37]. We were unable to include this information in the current study, but future studies may wish to investigate AED levels as additional potential explanatory variables for seizure dissimilarity. ...
Preprint
Epilepsy is recognised as a dynamic disease, where seizures and their features change over time. Specifically, we recently demonstrated that seizures themselves change in terms of their evolution. However, the underpinning modulators of seizure evolution remain unclear. In this work, we analyse continuous (interictal) intracranial Electroencephalographic (iEEG) recordings, and elucidate fluctuations in iEEG band power over different timescales (ranging from minutes to days). We find that all patients show an approximately circadian fluctuation in their EEG band power, but also many other fluctuations on patient-specific timescales. Importantly, we find that a combination of fluctuations on different timescales can explain the changes in seizure evolution in most patients above chance-level. We interpret these results as evidence that seizure modulating factors exist, and they vary over time (patient-specifically). These time-varying modulating factors can be captured in fluctuations of EEG band power, and future work should link them to the exact biological time-varying processes.
Article
Objective The goal of this study was to identify a strategy for antiepileptic drug (AED) reduction to allow efficient recording of focal seizures (FS) in patients undergoing video-electroencephalography (EEG) in an epilepsy monitoring unit (EMU) while avoiding the risk of complications associated with more severe seizure types. Methods We retrospectively reviewed consecutive patients admitted to our institution's EMU from July 1, 2016 to December 31, 2017. We included 114 presurgical patients who had AEDs reduced and at least one seizure during the admission. We compared AED dosages at which FS versus focal to bilateral tonic–clonic seizures (f-BTCS), seizure clusters, and lorazepam administration occurred. We also examined rate of AED reduction and seizure types. We used a receiver-operating characteristic (ROC) curve to identify a dose maximizing FS and minimizing other seizure types. Results Antiepileptic drug withdrawal rates ranged from 0 to 100% in the first 24 h (mean: 20%, standard deviation: 20%). Focal to bilateral tonic–clonic seizures and lorazepam administration occurred at a lower median AED dose than did FS (0%, 7.2%, and 43.8%, respectively, expressed as a percentage of the patient's outpatient daily AED dose; p < 0.001). A daily EMU-administered dose of one-third of the patient's outpatient AED dose allowed 55.0% of FS to occur while avoiding 82.0% of more severe seizure types. The seizure types had no difference in rate of AED withdrawal in the first 24 h of EMU stay. Conclusions Focal seizures occurred at a higher AED dose than did f-BTCS. This may imply that a low minimally effective dose of AED could allow FS to be recorded while providing protection against f-BTCS. This strategy could improve efficacy and safety in the EMU.
Article
Full-text available
Purpose: The European Union-funded E-PILEPSY network aims to improve awareness of, and accessibility to, epilepsy surgery across Europe. In this study we assessed current clinical practices in epilepsy monitoring units (EMUs) in the participating centers. Method: A 60-item web-based survey was distributed to 25 centers (27 EMUs) of the E-PILEPSY network across 22 European countries. The questionnaire was designed to evaluate the characteristics of EMUs, including organizational aspects, admission, and observation of patients, procedures performed, safety issues, cost, and reimbursement. Results: Complete responses were received from all (100%) EMUs surveyed. Continuous observation of patients was performed in 22 (81%) EMUs during regular working hours, and in 17 EMUs (63%) outside of regular working hours. Fifteen (56%) EMUs requested a signed informed consent before admission. All EMUs performed tapering/withdrawal of antiepileptic drugs, 14 (52%) prior to admission to an EMU. Specific protocols on antiepileptic drugs (AED) tapering were available in four (15%) EMUs. Standardized Operating Procedures (SOP) for the treatment of seizure clusters and status epilepticus were available in 16 (59%). Safety measures implemented by EMUs were: alarm seizure buttons in 21 (78%), restricted patient's ambulation in 19 (70%), guard rails in 16 (59%), and specially designated bathrooms in 7 (26%). Average costs for one inpatient day in EMU ranged between 100 and 2200 Euros. Conclusion: This study shows a considerable diversity in the organization and practice patterns across European epilepsy monitoring units. The collected data may contribute to the development and implementation of evidence-based recommended practices in LTM services across Europe.
Article
Purpose: To systematically evaluate the duration of focal onset seizures under medication withdrawal as a function of drug half-life. Methods: Adults with drug resistant focal epilepsy and invasive electroencephalographic (iEEG) recording between 01/2006 and 06/2016 (n = 128) were identified. Patients with multifocal or unknown epileptic foci were excluded, as well as subclinical seizures, isolated auras, or status epileptic. Antiepileptic drugs (AEDs) were withdrawn upon admission. The seizure duration was determined based on the invasive EEG data, and the latency since start of the monitoring was noted in hours. A negative binomial mixed model was used to compare the seizure durations before and after a cut-off, which was set at 2.5 half-lives of the individual anticonvulsive medication as this is thought to separate therapeutic and ineffective drug levels. Results: In total, 70 patients were included in the study and the duration of 672 seizures analyzed. On average, the patients were treated with 2.36 ± 0.78 AEDs. The individual cut-off of 2.5 half-lives was on average reached after 95.02 ± 80.18 h. The seizure frequency (321 vs. 351) and the rate of generalization (15.6% vs. 16.8%) was comparable before and after the individual cut-off point. The mean seizure duration was not statistically significantly prolonged after 2.5 half-lives by a factor of 1.168 for focal onset seizures (p = 0.090) and a factor of 1.091 for secondary generalized seizures (p = 0.545). Conclusions: Although AED withdrawal increases the likelihood for epileptic seizures, it did not prolong the seizure duration, nor did it increase the rate of secondary generalization in our study.
Article
Purpose Predictive factors of higher drug load for seizure freedom were investigated in idiopathic generalized epilepsy (IGE), focusing on the difference between juvenile myoclonic epilepsy (JME) and other types of IGE (non-JME IGE). Methods Twelve patients with JME and 12 patients with non-JME IGE, who achieved seizure freedom for 1 year or longer with appropriate antiepileptic drugs (AEDs) after video electroencephalography monitoring, were reviewed retrospectively. The sum of prescribed daily dose/defined daily dose ratio of all prescribed AEDs at the final visit was defined as total AED load. Patients requiring total AED load >1 were classified into the higher AED load group. Clinical background and the presence of interictal focal epileptiform abnormalities (FEAs) were compared between the higher and lower AED load groups. Results Higher AED load group of patients with JME had interictal FEAs and family history of epilepsy more frequently than the lower AED load group (p = 0.03 and p = 0.03). Similar comparison of patients with non-JME IGE showed no significant differences. Conclusions The presence of interictal FEAs and a family history of epilepsy are significantly associated variables for higher AED load for seizure freedom in patients with JME, but not in patients with non-JME IGE.
Article
Introduction: Stimulation-evoked focal to bilateral tonic-clonic seizure (FBTCS) can be a stressful and possibly harmful adverse event for patients during cortical stimulation (CS). We evaluated if drug load reduction of antiepileptic drugs (AEDs) during CS increases the risk of stimulation-evoked FBTCS. Material and methods: In this retrospective cohort study, we searched our local database for patients with drug-resistant epilepsy who underwent invasive video-EEG monitoring and CS in the University Hospital la Fe Valencia from January 2006 to November 2016. The AED drug load was calculated with the defined daily dose. We applied a uni- and multivariate logistic regression model to estimate the risk of stimulation-evoked FBTCS and evaluate possible influencing factors. Furthermore, we compared patients whose AEDs were completely withdrawn with those whose AEDs were not. Results: Fifty-eight patients met the inclusion criteria and were included in the analysis. Stimulating 3806 electrode contact pairs, 152 seizures were evoked in 28 patients (48.3%). Ten seizures (6.6%) in seven patients (12.1%) evolved to FBTCS. In the univariate and multivariate analysis, a 10% reduction in drug load was associated with an increase of the odds ratio (OR) of stimulation-evoked FBTCS by 1.9 (95%-CI 1.2, 4.0, p-value=0.04) and 1.9 (95%-CI 1.2, 4.6, p-value=0.04), respectively. In patients, whose AEDs were completely withdrawn the OR of FBTCS increased by 9.1 (95%CI 1.7, 69.9, p-value=0.01) compared with patients whose AEDs were not completely withdrawn. No other factor (implantation type, maximum stimulus intensity, number of stimulated contacts, history of FBTCS, age, gender, or epilepsy type) appears to have a significant effect on the risk of stimulation-evoked FBTCS. Conclusions: The overall risk of stimulation-evoked FBTCS during CS is relatively low. However, a stronger reduction and, especially, a complete withdrawal of AEDs are associated with an increased risk of stimulation-evoked FBTCS.
Article
Objective: Antiepileptic drugs (AEDs) are routinely withdrawn during long-term video-electroencephalography (EEG) monitoring (LTM), to record sufficient number of seizures. The efficacy of rapid and slow AED taper has never been compared in a randomized control trial (RCT), which was the objective of this study. Methods: In this open-label RCT, patients aged 2-80 years with drug-resistant epilepsy (DRE) were randomly assigned (1:1) to rapid and slow AED taper groups. Outcome assessor was blinded to the allocation arms. Daily AED dose reduction was 30% to 50% and 15% to <30% in the rapid and slow taper groups, respectively. The primary outcome was difference in mean duration of LTM between the rapid and slow AED taper groups. Secondary outcomes included diagnostic yield, secondary generalized tonic-clonic seizure (GTCS), 4- and 24- hour seizure clusters, status epilepticus, and need for midazolam rescue treatment. The study was registered with Clinical Trial Registry-India (CTRI/2016/08/007207). Results: One hundred forty patients were randomly assigned to rapid (n = 70) or slow taper groups (n = 70), between June 13, 2016 and February 20, 2017. The difference in mean LTM duration between the rapid and slow taper groups was -1.8 days (95% confidence interval [CI] -2.9 to -0.8, P = .0006). Of the secondary outcome measures, time to first seizure (2.9 ± 1.7 and 4.6 ± 3.0 days in the rapid and slow taper groups respectively, P = .0002) and occurrence of 4-hour seizure clusters (11.9% and 2.9% in the rapid and slow taper groups, respectively, P = .04) were statistically significant. None of the other safety variables were different between the 2 groups. LTM diagnostic yield was 95.7% and 97.1%, in rapid and slow taper groups respectively (P = .46). Significance: Rapid AED tapering has the advantage of significantly reducing LTM duration over slow tapering, without any serious adverse events.
Article
First, most responder EMUs don't have a written policy for AED withdrawal: Only 38% of those that mainly monitor adults and 27% of pediatric EMUs have a written protocol for their AED tapers. One might expect then a significant variability in how and when AEDs are stopped in this context. Some of the Q-PULSE findings support this variation in practice. Around half of the survey responders said they have no consistent pattern to the sequence of AED reduction (i.e., stopping all medications at the same time vs one at a time, etc.), suggesting that these decisions are mainly made on a case-by-case basis. Another area of variability relates to allowing a pre-admission taper of AEDs. Forty-seven percent of adult patients expecting an admission to an EMU that exclusively monitors adults are never allowed a pre-admission AED taper, while only 24% of them face this rule if expecting an admission to an EMU that monitors both adults and children. A pre-admission AED withdrawal is favored by some as a tool to expedite the occurrence of seizures in the EMU, thus shortening the patient's length-of-stay, yet it is feared by others as it may carry risks of acutely worsening seizure control and causing injury before the patient is actually safely monitored in the hospital. For the same patient population (adults with uncontrolled spells), it is difficult to imagine a scientific reason that this risk–benefit calculation should vary across epilepsy centers. Given the significant value on either side of the argument and the seemingly discrepant practices across epilepsy centers, the safety and appropriateness of pre-admission AED taper should be studied further.
Article
Objectives: Anti-epileptic drugs (AED) are often tapered to reduce the time needed to record a sufficient number of seizure during long-term video-EEG monitoring (LTM). Fast AED reduction is considered less safe, but few studies have examined this. Our goal is to examine whether the rate of AED reduction affects efficiency and safety. Materials & methods: We performed a retrospective observational study of patients referred for presurgical evaluation. Each patient was categorized by two dichotomous parameters of AED tapering: (i) fast vs slow AED reduction the first 24 h of LTM and (ii) complete vs partial AED discontinuation during LTM. Results: Of 79 patients, 51% underwent a fast AED reduction and 58% ended up with AEDs completely discontinued. Complete AED discontinuation was associated with three times increased likelihood of receiving rescue therapy during LTM and double risk of having secondary generalized tonic-clonic seizures (sGTCS) compared to the group partially discontinued. Fast vs slow AED reduction had no effect on the safety of LTM. The fast AED reduction group and the complete AED discontinuation group had a significantly longer time to first seizure and total recording time compared to the slow AED reduction and partial discontinuation groups, respectively. Conclusions: Fast AED reduction was found safe in patients undergoing presurgical video-EEG monitoring. Patients completely discontinued from AEDs had more sGTCS than patients partially discontinued. Further studies are suggested to confirm this finding and to evaluate whether fast reduction is safe and efficient in other subgroups of patients referred for LTM.
Article
OBJECTIVE: To determine the rate of medical complications from long-term video-electroencephalographic (EEG) monitoring for epilepsy. PATIENTS AND METHODS: We reviewed the medical records of 428 consecutive adult patients with epilepsy who were admitted for diagnostic scalp video-EEG monitoring at Mayo Clinic's site in Arizona from January 1, 2005, to December 31, 2006; 149 met inclusion criteria for the study. Seizure number and type as well as timing and presence of seizure-related adverse outcomes were noted. RESULTS: Of the 149 adult patients included in the study, seizure clusters occurred in 35 (23%); 752 seizures were recorded. The mean time to first seizure was 2 days, with a mean length of stay of 5 days. Among these patients, there was 1 episode of status epilepticus, 3 potentially serious electrocardiographic abnormalities, 2 cases of postictal psychosis, and 4 vertebral compression fractures during a generalized convulsion, representing 11% of patients with a recorded generalized tonic-clonic seizure. No deaths, transfers to the intensive care unit, falls, dental injuries, or pulmonary complications were recorded. An adverse event requiring intervention or interfering with normal activity occurred in 21% of these patients. Length of stay was not affected by occurrence of adverse events. CONCLUSION: Prolonged video-EEG monitoring is an acceptably safe procedure. Adverse events occur but need not result in substantial morbidity or increase length of hospitalization. Appropriate precautions must be in place to prevent falls and promptly detect and treat seizure clusters, status epilepticus, serious electrocardiographic abnormalities, psychosis, and fractures.
Article
Objectives Secondary generalized tonic–clonic seizures (SGTCS) are among the most severe forms of seizures, and the main risk factor for sudden unexpected death in epilepsy (SUDEP). Whether some antiepileptic drugs (AEDs) might be more efficacious than others on SGTCS in patients with drug-resistant focal epilepsy thus represents an important clinical issue for which no data are currently available.Methods We performed a meta-analysis of randomized controlled trials of adjunctive AED in which information on efficacy outcomes (i.e., responder rate and/or frequency per 28 days relative to baseline) were available both for all seizure types and for SGTCS. The primary analysis evaluated the efficacy of AEDs on all types of seizure and on SGTCS by comparing the responder rates for AED and for placebo.ResultsResponder rate was available both for all seizure types and for SGTCS in 13 of the 72 eligible trials, evaluating 7 AEDs. Only three AEDs—lacosamide, perampanel and topiramate—showed greater efficacy than placebo. However, confidence intervals of relative risks overlapped for all AEDs but pregabalin, which demonstrated significantly lower efficacy than lacosamide, perampanel, and topiramate. Moreover, there was a nonsignificant trend toward a lower relative risk of responder rate for SGTCS than for all seizure types, which appeared related to a greater response to placebo for this outcome.SignificanceIndirect comparison of AEDs using randomized placebo-controlled add-on trials does not support robust differences between AEDs to prevent SGTCS. Alternative designs for evaluation of therapeutic interventions in patients at risk for SGTCS-related complications are required.