Antiepileptic-drug tapering and seizure recurrence: Correlation with serum drug levels and biomarkers in persons with epilepsy.

OBJECTIVES: Antiepileptic-drug (AED) serum level and inflammatory biomarkers are primarily monitored/assessed during epilepsy treatment for effective seizure control; however, their correlation with seizure recurrence (SR) following AED-tapering has not been established, and this is being investigated in this study.
MATERIALS AND METHODS: This prospective observational study enrolled persons with epilepsy (PWE) on AED monotherapy and going to start tapering after being seizure-free for ≥2 years. Data regarding seizure episodes, AED-treatment, and adverse events (using Liverpool Adverse Event profile [LAEP]-score) were recorded. Serum AED levels using high-performance liquid chromatography and biomarkers levels through enzyme-linked immunosorbent assay kits were estimated at AED-tapering commencement and at 6 months/SR time.
RESULTS: Among 129 enrolled PWE (levetiracetam [n = 52], valproate [n = 34], carbamazepine [n = 29], and phenytoin [n = 14]), SR occurred in 23.3% during follow-up (range 12-44 months). PWE with subtherapeutic serum AED level at the onset of tapering had higher SR (P = 0.004) than those with therapeutic or higher levels. Levetiracetam-treated PWEs with SR have significantly low AED levels than PWE with no-SR (P < 0.001). PWE had significantly raised inflammatory biomarkers (interleukin [IL]-1 β, tumor necrosis factor [TNF]-α, IL-6, and high-mobility group box protein 1) and decreased IL-10 than healthy control subjects. SR and no-SR groups did not differ significantly in inflammatory markers except for higher IL-1 β and TNF-α levels in SR group (P = 0.001, 0.02, respectively). Improvement in LAEP score was observed in follow-up visits without any difference between SR and no-SR groups.
CONCLUSION: Low serum AED levels (especially levetiracetam) and raised levels of TNF-α and IL-1 β during tapering commencement had a higher association with SR following AED-tapering.

Introduction

Persons with epilepsy (PWE) were considered for antiepileptic drugs (AEDs) tapering if they remain without any seizure episodes for 2 or more years[1] because of unendurable adverse drug events (AEs), unbearable treatment cost, and/or often social stigma.[23] The decision for AED tapering/stoppage of treatment is shifting from the physician perspective alone to a patient-centered shared decision-making based on the seizure-free probability even after AED tapering. According to published literature, the risk of seizure recurrence (SR) scales between 14% and 64.6% depending upon the study design and conduct. Indian studies have demonstrated the risk of SR in adult PWE as 19%–31%.[456] A meta-analysis also reported the risk of SR after drug withdrawal as about 25% in the 1st year and 29% in 2 years.[7] Thus SR incidence varies substantially, necessitating further studies to find out the predictive factors associated with SR.[8910] For optimal management of epilepsy intending to reduce the risk of SR, there is a need for the individualization of a tapering strategy.

Therapeutic drug monitoring is crucial for surveillance of drug toxicity/overdosage, compliance of drug intake, and thus abet in dose titration during AEDs treatment. Owing to complex pharmacokinetic properties and narrow therapeutic index, extensive variation in AEDs plasma concentration can result in toxicity or loss of therapeutic efficacy.[1112]

Many studies have assessed SR risk in PWE with controlled seizure undergoing tapering and ultimately stoppage of AED, but the comparative efficacy of continuation of lowered down AED doses in this population has not been established. If seizure can be controlled at lower maintenance or “sub-therapeutic” AED doses, this will lead to reduced drug toxicity and reduced treatment costs.[13] However, this does not completely abort the chances of SR, AEs, stigma to the subject being on AED treatment, and the cost of ongoing treatment. Hence, there is no clarity of being on a low dose of AED or complete AED tapering. Although AED dose has been studied concerning SR, there is a lack of evidence focussed on serum AED level and its correlation with SR during AED tapering commenced after at least 2 years of seizure freedom.

A strong persistent inflammatory response is one of the key components of epileptogenesis. Upregulation of inflammatory mediators leads to blood–brain barrier breakdown, which facilitates neuroinflammation and subsequently synaptic changes and neuronal hyper-excitability.[14] Increased postictal serum concentrations of inflammatory mediators such as interleukin (IL)-6, IL-2, IL-1b, IL-4, and interferon-γ were described by Sinha et al.[15] Tumor necrosis factor (TNF)-α, described as a pro-inflammatory cytokine, shows a neurotropic and neurotoxic effect in the central nervous system; it upregulates AMPA receptors, augmenting glutamatergic transmission.[14] According to Lalla et al., cytokines like IL-1 β were raised with a calcified lesion in the brain and TNF-α level was higher in PWE with SR.[16] Another anti-inflammatory cytokine, IL-10, suppresses pro-inflammatory cytokine generation and was found to be significantly reduced in PWE with focal onset seizure than in healthy controls.[1718]

The high-mobility group box protein 1 (HMGB1) binds to DNA and acts as an initiator and amplifier of neuroinflammation correlated to epilepsy pathogenesis. Even HMGB1-targeted therapies against epilepsy have been implicated.[19]

Despite these shreds of evidence, there is a lack of studies focusing on serum inflammatory markers assessment during AED tapering where the major concern is the chance of SR. Therefore, this study was intended to deduce the correlation of serum AEDs level at different time points (at the onset of tapering, at 6 months/SR/stoppage of AED) and inflammatory markers level at the onset of tapering with SR during AED tapering.

Materials and Methods

Study subjects

This prospective longitudinal observational study with Institute Ethics Committee approval (IECPG/131/27.01.2016, RT-5/24.02.2016) was carried out in a tertiary care hospital from February 2016 to October 2019 according to the Declaration of Helsinki and Good Clinical Practice Guidelines of Government of India (Study CTRI reg. No. CTRI/2017/04/008392). Written informed consent/assent was acquired from subjects/guardians before the screening. The study was conducted among clinically diagnosed PWE on treatment with AEDs monotherapy (valproate [VPA], phenytoin [PHT], carbamazepine [CBZ], and levetiracetam [LEV]) coming to the neurology outpatient department, with ≥2 years of seizure freedom and who was going to start AED tapering. PWE excluded from this study include subjects who were either unable to give consent, who had Juvenile myoclonus epilepsy, with the diagnosis of other neurological or psychiatric disorder, chronic disease including inflammatory conditions, on other drugs treatment which can alter seizure threshold, with liver and renal dysfunction requiring AED dose alteration, pregnant or seeking pregnancy. Diagnosis of epilepsy was made by the treating physician according to the International League Against Epilepsy guidelines. A group of healthy control subjects was enrolled to compare the serum level of inflammatory markers with that of PWE.

Study data collection

After explaining the study details, each PWE was interviewed according to a well-designed case reporting format to collect the following data: demographic facts; seizure-related information including age at seizure onset, type, number of seizure episodes before control, and time duration of epilepsy (i.e., dates of first seizure episode to the time of tapering onset); and AED treatment information such as dose, duration, and investigational findings if any (like electroencephalography, computed tomography, and magnetic resonance imaging). AED dose at the onset of tapering was represented as a percentage of maximum recommended dose (MRD) of the concerned AED as per the Textbook of Harrison's principles of internal medicine, i.e., PHT (6 mg/kg), CBZ (35 mg/kg), VPA (60 mg/kg), and LEV (60 mg/kg).[20] Strict patient confidentiality was maintained. PWEs were followed up for a minimum of 12 months to look for any SR. This study has done AEs profiling during tapering in PWE using the Liverpool Adverse Event Profile (LAEP) score. LAEP was a patient self-reporting scale to measure the frequency of AED-related AEs. This scale has 19 items/questionnaire with a 4-point Likert scale for each item, i.e., 1 – never a problem; 2 – rarely a problem; 3 – sometimes a problem; and 4 – always a problem.[21] Although the items can be gauged individually for frequency, a global score can be derived from the sum of the ratings. This scale has a total score range from 19 to 76, with more frequent symptoms denoted by a higher score.

Among the PWE, analysis was done for risk factors like higher frequency of seizure episodes before control, longer duration of epilepsy, history of any failed tapering, and history of smoking/alcohol/tobacco intake, which have shown a positive correlation with SR as per a previous study conducted in this setup.[4]

Patients were called for follow-up visits as per existing hospital treatment protocol for PWE, i.e., at 6 months intervals for a minimum of 1 year/at the time of stoppage of AED/SR. During these visits, monitoring for occurence of SR and AEs was done. Along with follow-up visits, PWE were contacted telephonically in between for any additional information about SR.

Serum antiepileptic drug level estimation

Serum AED levels for VPA, CBZ, PHT, and LEV were estimated at two visits: A. onset of tapering time, and b. at the time of SR, stoppage of AED, or 6-month follow-up time, whichever is earlier. Blood samples were collected from the subject coming to the neurology outpatient department in the morning before AED intake (trough sample). Subjects with SR were advised to immediately collect a blood sample from a nearby health-care facility as soon as possible and send it to the study center for analysis. The blood sample was processed by cold (4°C) centrifuging at 3500 rpm for 10 min, later serum was collected and stored at −80°C. Serum AED level was estimated using the high-performance liquid chromatography (HPLC) instrument (Agilent 1200 series, Agilent Technologies, USA) and C18 column (Zorbax C18-Merck, Germany). Reagents used for HPLC analysis include potassium dihydrogen orthophosphate, orthophosphoric acid, acetonitrile, methanol, potassium hydroxide, and triethylamine, which are purchased from Merck India Ltd. Standards of AEDs were received from Sun Pharmaceuticals, India as gift samples. Serum was mixed with acetonitrile (for VPA, CBZ, PHT) or methanol (for LEV) and vortexed. It was later centrifuged at 15,000–18,000 rpm (specific for AED) to separate precipitated protein. Then, the supernatant was filtered and injected into the HPLC system through an automatic sampler. The mobile phase comprised the buffer of potassium dihydrogen orthophosphate, methanol, and acetonitrile.

For CBZ and PHT estimation, the mobile phase passed with a 1.2 ml/min flow rate for 15 min runtime. Column compartment with thermostat maintained the temperature at 25°C and UV detector at wavelength 210 nm was used for detection.[22] For LEV estimation, the flow rate was 1 ml/min with a runtime of 15 min by maintaining column temperature at 25°C and measuring peak area and retention time at 205 nm wavelength.[23]

VPA was estimated using a PDA detector at wavelength 210 nm. The mobile phase was passed at 1.2 ml/min through a thermostated column compartment at a temperature of 50°C. The mobile phase comprised acetonitrile and phosphate buffer with a run time of 21 min.[24]

Serum biomarkers estimation

Venous blood samples (5 ml) were collected at the onset of tapering (first visit) from all monotherapy subjects enrolled in the study for serum biomarkers level estimation. The serum was separated by centrifugation and stored in a −80°C deep freezer till analysis. The inflammatory biomarkers (TNF-α, IL-1 β, IL-6, IL-10, and HMGB1) were estimated through enzyme-linked immunosorbent assay kits (R and D Systems, Minneapolis, MN 55413, USA). The kits were run as per the instruction manual from the manufacturer and absorbance was recorded through Multi-Mode Microplate Readers (SpectraMax® M Series, Molecular Devices, California, USA). For comparison purposes, blood samples of healthy control subjects (n = 52) were also collected, and the above inflammatory markers were estimated in them.

Statistical analysis

The sample size was calculated using a previous study result where SR was observed in 19.2% PWE with low serum AED level and in 34.4% PWE with therapeutic or higher serum AED level after a seizure-free time of two or more years,[13] though this was not the primary endpoint of that study. Considering significance (5%, two-sided test for one sample comparison) and power as 90%, the sample size was calculated as 90. Adding loss to follow-up of 30%, a total sample size needed is 117. The comparison was done between PWEs with SR and no-SR (NSR) at the end of the study.

Analysis of data was conducted using Statistical Package for the Social Sciences (SPSS) software version 23.0 (Chicago, IL, USA). Parametric data were presented as mean ± standard deviation and nonparametric data were by the median and interquartile range or range (minimum–maximum). Mann–Whitney U statistical test was applied for the comparison of nonparametric data. A Chi-square test was used to find out any difference in the distribution frequency among the groups. Statistical significance is considered for P < 0.05.

In this study, SR was correlated with the presence of risk factors, serum AED level at two visits including at the onset of tapering, AED dose, and biomarkers level at the onset of tapering. Serum AED level at the onset of tapering was categorized into subtherapeutic, therapeutic, and higher range, and SR was compared between them. Similarly, serum baseline biomarkers levels were compared among the PWE with SR and NSR.

Results

A total of 140 consecutive patients enrolled after the screening of 162 PWE, who had no seizure for two or more years. Eleven patients were excluded due to noncompliance to the study protocol including refusal to provide samples for analysis, discontinuing the tapering process, and loss of follow-up. Among 129 enrolled PWE who were on AED monotherapy, 52 PWE were on LEV (40.3%), VPA: 34 (26.4%), CBZ: 29 (22.5%), and PHT: 14 (10.8%). The detailed demographics are presented in Table 1. Among all monotherapy PWE, family history of epilepsy and previous tapering history was present in 21 (16.3%) and 24 (18.6%) PWE. The seizure-free period was 2–10 years (median 3.1 years). AED was tapered using a standard procedure.

Table 1

Demographics, type of antiepileptic drug, and seizure-recurrence rate among persons with epilepsy undergoing antiepileptic drug tapering

Parameters	PWE on monotherapy (n=129), n (%)
Demographic characteristics
Age in years (mean±SD)	25.0±10.4
Gender (male)	77 (59.7)
Seizure type
Focal	76 (58.9)
Generalized	53 (41.1)
Seizure free period in years
Median (range)	3.1 (2–10)
2-4	79 (61.2)
>4	50 (38.8)
Family h/o epilepsy
No	108 (83.7)
Yes	21 (16.3)
Previous h/o of tapering
No	105 (81.4)
Yes	24 (18.6)
Education
Below 10th	33 (25.6)
Above 10th	96 (74.4)
Occupation
Unemployed/student	87 (67.4)
Employed	42 (32.6)
Smoking/alcoholic/tobacco chewing
No	118 (91.5)
Yes	11 (8.5)
AED treatment parameters
AEDs monotherapy
Levetiracetam	52 (40.3)
Valproate	34 (26.4)
Carbamazepine	29 (22.5)
Phenytoin	14 (10.8)
SR in PWE	30 (23.3)
SR with time after AED tapering  SR in PWE with time after AED tapering
Within 6 months	14 (10.8)
Within 12 months	20 (15.5)
Within 24 months	28 (21.7)
Within 44 months	30 (23.3)

AED=Antiepileptic drug, PWE=Persons with epilepsy, SD=Standard deviation, SR=Seizure recurrence, H/o=History of

AED dose was reduced by a fixed amount (LEV 250 mg, CBZ 50 or 100 or 150 mg, VPA 100 or 200 mg, and PHT 50 or 100 mg) at a time without taking into consideration the dose just before the commencement of tapering. The amount of dose tapered constitutes about 6.67%–50.0% (median 25%) of the dose just before the commencement of tapering. The median interval between dose reductions was 2 months (range 2–6 months). The proposed total duration of AED tapering was 7 months (median, range 1.5–22 months). After follow-up (range 12–44 months) in 129 PWE, SR was reported in 30 PWE (23.3%), while 99 (76.7%) PWE remained seizure-free. The number of SR with time after AED tapering (% of total PWE with SR) was 14 (10.8%) within 6 months, 20 (15.5%) within 12 months, 28 (21.7%) within 24 months, and 30 (23.3%) within 44 months of follow-up [Table 1].

Antiepileptic drug levels and risk factors associated with seizure recurrence

AED analysis was performed in 129 subjects who were on LEV (52), VPA (34), CBZ (29), and PHT (14). The daily AED doses and serum levels range at onset of AED tapering were CBZ: 400–1800 mg/day and 2.1–20.1 mcg/ml (median = 8.7), PHT: 200–400 mg/day and 1.6–23.3 mcg/ml (median = 6.55), LEV: 1000–3000 mg/day and 1.3–100.3 mcg/ml (median = 23.1), and VPA: 500–2000 mg/day and 1.0–112.8 mcg/ml (median = 43.9).

According to serum AED levels at the onset of tapering, PWE were classified into three categories of AED levels, i.e., subtherapeutic, therapeutic, and higher. Among monotherapy subjects (n = 129), 32 PWE had subtherapeutic concentrations at the onset of AED tapering and 97 PWE were in the therapeutic range or higher. Among the NSR group (n = 99), 18 PWEs were at subtherapeutic concentration, 65 PWEs were at therapeutic concentration, and 16 PWE were at high concentration. Among the SR group (n = 30), 14 PWEs had subtherapeutic concentrations, 11 PWEs were at therapeutic concentration, and five PWE were at high concentration. This showed that there is more SR among those with subtherapeutic concentration at the onset of tapering (P = 0.001). In the LEV group (n = 52), 12 PWEs experienced SR, and it was more commonly encountered among those with subtherapeutic concentration at onset of tapering (n = 5) (P ≤ 0.001) [Table 2].

Table 2

Persons with epilepsy with Risk factors and their serum antiepileptic drug levels

All PWE
	Total (n=129)				NSR (n=99)				SR (n=30)				P SR vs. NSR
	High	Therapeutic range	Sub therapeutic	P	High	Therapeutic range	Sub therapeutic	P	High	Therapeutic range	Sub therapeutic	P
PWE AED level	21 (16.3)	76 (58.9)	32 (24.8)		16 (16.2)	65 (65.6)	18 (18.2)		5 (16.6)	11 (36.7)	14 (46.7)		0.001
PWE Risk factors	15 (11.6)	32 (24.8)	14 (10.8)	0.052	13 (13.1)	26 (26.2)	8 (8.1)	0.012	2 (6.6)	6 (20)	6 (20)	0.800
LEV
	Total (n=52)				NSR (n=40)				SR (n=12)				P SR vs. NSR
	High	Therapeutic range	Sub therapeutic	P	High	Therapeutic range	Sub therapeutic	P	High	Therapeutic range	Sub therapeutic	P
PWE AED level	10 (19.2)	36 (69.2)	6 (11.6)		9 (22.5)	30 (75)	1 (2.5)		1 (8.3)	6 (50)	5 (41.7)		<0.001
PWE Risk factors	8 (15.4)	14 (26.9)	2 (3.8)	0.055	7 (17.5)	12 (30)	1 (2.5)	--	1 (8.3)	2 (16.7)	1 (8.3)	--
VPA
	Total (n=34)				NSR (n=22)				SR (n=12)				P SR vs. NSR
	High	Therapeutic range	Sub therapeutic	P	High	Therapeutic range	Sub therapeutic	P	High	Therapeutic range	Sub therapeutic	P
PWE AED level	3 (8.8)	10 (29.4)	21 (61.7)		1 (4.5)	7 (31.8)	14 (63.6)		2 (16.6)	3 (25)	7 (58.3)		0.761
PWE Risk factors	2 (5.8)	5 (14.7)	9 (26.4)	0.723	1 (4.5)	3 (13.6)	5 (22.7)	--	1 (8.3)	2 (16.6)	4 (33.3)	0.929
CBZ
	Total (n=29)				NSR (n=26)				SR (n=3)				P SR vs. NSR
	High	Therapeutic range	Sub therapeutic	P	High	Therapeutic range	Sub therapeutic	P	High	Therapeutic range	Sub therapeutic	P
PWE AED level	6 (20.7)	21 (72.4)	2 (6.9)		5 (19.2)	20 (76.9)	1 (3.8)		1 (33.3)	1 (33.3)	1 (33.3)		0.056
PWE Risk factors	4 (13.8)	9 (31.0)	1 (3.4)	0.588	4 (15.4)	8 (30.7)	1 (3.8)	--	0 (0)	1 (33.3)	0 (0)	--
PHT
	Total (n=14)				NSR (n=11)				SR (n=3)				P SR vs. NSR
	High	Therapeutic range	Sub therapeutic	P	High	Therapeutic range	Sub therapeutic	P	High	Therapeutic range	Sub therapeutic	P
PWE AED level	2 (14.3)	9 (64.3)	3 (21.4)		1 (9.1)	8 (72.7)	2 (18.2)		1 (33.3)	1 (33.3)	1 (33.3)		0.570
PWE Risk factors	1 (7.1)	4 (28.6)	2 (14.3)	0.800	1 (9.1)	3 (27.3)	1 (9.1)	--	0 (0)	1 (33.3)	1 (33.3)	--

Data are expressed as n (%). P between SR and NSR groups compared the distribution of the number of PWE in subtherapeutic vs. high or therapeutic ranges based on the Chi-Square test. Risk factors include higher frequency of seizure episodes before control, longer duration of epilepsy, history of any failed tapering, and history of smoking/alcohol/tobacco intake. VPA=Valproate, PHT=Phenytoin, CBZ=Carbamazepine, LEV=Levetiracetam, PWE=Persons with epilepsy, SR=Seizure recurrence, NSR=Nonseizure-recurrence

The presence of risk factors (higher frequency of seizure episodes before control, longer duration of epilepsy, history of any failed tapering, and history of smoking/alcohol/tobacco intake) did not differ significantly between the SR and NSR groups among all monotherapy PWE and in individual AED groups also. Among all monotherapy subjects, 14 PWE had risk factors out of 32 PWE with subtherapeutic concentration and 47 PWE had risk factors out of 97 PWE with therapeutic or higher concentration. Among the SR group, six PWE had risk factors out of 14 PWE with subtherapeutic concentration and eight PWE had risk factors out of 16 PWE with therapeutic or higher concentration. In VPA, CBZ, and PHT groups, there was no significant difference in the number of PWE in categories of serum level or according to the presence of risk factors [Table 2].

There is no significant difference in AED levels at SR/stoppage of AED/at 6 months between SR and NSR group in any of the AED monotherapy groups [Table 3].

Table 3

Serum antiepileptic drug levels comparison between seizure-recurrence and nonseizure-recurrence persons with epilepsy

LEV (n=52)
Visits	AED serum level, median (IQR)		P
	SR (n=12)	NSR (n=40)
Onset of tapering (1st visit)	13.2 (4.1-35.5)	24.7 (13.3-43.7)	0.048
At SR/6 months (2nd visit)	4.55 (2.6-6.3)	7.1 (4.5-19.6)	0.057
VPA (n=34)
Visits	AED serum level, median (IQR)		P
	SR (n=12)	NSR (n=22)
Onset of tapering (1st visit)	59.85 (28.6-85.6)	40.55 (29.45-61.1)	0.45
At SR/6 months (2nd visit)	25.55 (18.9-49.9)	25.65 (13-33.6)	0.46
CBZ (n=29)
Visits	AED serum level, median (IQR)		P
	SR (n=3)	NSR (n=26)
Onset of tapering (1st visit)	10.4 (8.5-13.2)	8.6 (6.65-10.7)	0.40
At SR/6 months (2nd visit)	9.6 (6.2-11.2)	4.3 (2.75-6.9)	0.12
PHT (n=14)
Visits	AED serum level, median (IQR)		P
	SR (n=3)	NSR (n=11)
Onset of tapering (1st visit)	13 (07-19)	6.05 (3.85-10.6)	0.20
At SR/6 months (2nd visit)	4.05 (1.4-6.7)	1.8 (1.0-4.7)	0.62

VPA=Valproate, PHT=Phenytoin, CBZ=Carbamazepine, LEV=Levetiracetam, SR=Seizure-recurrence, NSR=Nonseizure-recurrence, IQR=Interquartile range, AED=Antiepileptic drug

Antiepileptic drug dose of persons with epilepsy as a percentage of maximum recommended dose

The dose at which tapering started corresponded to 20%–40% of MRD in maximum number of PWE (n = 72) (20%–30% dose, n = 38; 30%–40% dose, n = 34). Similarly, the maximum number of PWE (n = 16) who experienced SR was at 20%–40% of MRD at the onset of tapering (20%–30% dose, n = 7; 30%–40% dose, n = 9). Similar observations were noted in VPA, CBZ, and LEV groups. However, in the PHT group, the maximum number of PWEs was on 300 mg dose (80%–89% of MRD) and three PWEs experienced SR, who were on PHT dose 70%–80%, 60%–70%, and 30%–40% of MRD, respectively. These findings suggested that the AED dose at the commencement of tapering has no significant association with SR.

Inflammatory biomarkers level and seizure recurrence

PWE, in both SR and NSR groups or individual AED monotherapy groups, had significantly high levels of inflammatory mediators (TNF-α, IL-1 β, IL-6, IL-10, and HMGB1) as compared to healthy control subjects at the onset of tapering. The SR group had significantly higher IL-1 β and TNF-α as compared to the NSR group (P = 0.001 and 0.02, respectively). The comparison of individual AED treatment groups revealed a significant difference between them concerning IL-10, TNF-α, and HMGB1. VPA group had significantly higher IL-10 and TNF-α, whereas the LEV group had higher HMGB1. Among the monotherapy groups, subgroup comparison (SR vs. NSR) was not done due to a limited sample size [Table 4].

Table 4

Comparison of inflammatory biomarkers at onset of tapering among different groups

Groups	Median (IQR)		P value (all PWE vs. control)	Median (IQR)		P			Median (IQR)				P value (each monotherapy vs. control)	P value (among monotherapy groups)
	Control (n=52)	All PWE (n=129)		PWE with SR (n=23)	PWE without SR (n=106)	Control versus SR	Control versus NSR	SR versus NSR	LEV (n=52)	VPA (n=34)	PHT (n=14)	CBZ (n=29)
IL-1β (pg/ml)	12.8 (10.4-16.0)	73.5 (49.3-93.6)	<0.001	82.0 (33.7-97.4)	72.6 (50.5-94.0)	<0.001	<0.001	0.001	73.1 (56.9-108.3)	79.2 (49.2-98.0)	76.0 (70.9-99.1)	58.3 (35.5-87.4)	<0.001	NS
IL-6 (pg/ml)	73.8 (64.5-82.8)	94.7 (70.7-119.0)	<0.001	90.6 (68.5-104.2)	94.8 (70.7-126.8)	0.001	<0.001	NS	80.5 (67.6-122.7)	97.2 (70.8-138.8)	106.7 (84.2-112.6)	99.3 (80.6-110.7)	<0.001	NS
IL-10 (pg/ml)	175.4 (140.5-210.4)	61.4 (56.4-70.5)	<0.001	62.7 (57.4-64.9)	60.6 (56.3-72.7)	<0.001	0.001	NS	64.0 (58.5-74.8)	66.7 (57.1-80.6)	56.7 (54.9-58.7)	56.6 (54.1-60.2)	<0.001	<0.001
TNF-α (pg/ml)	4.4 (2.9-11.9)	115.5 (86.4-151.3)	<0.001	125.7 (94.7-146.5)	109.5 (86.1-151.6)	<0.001	0.01	0.02	127.4 (92.0-193.6)	130.4 (88.8-176.9)	92.7 (80.1-96.7)	103.7 (82.3-141.2)	<0.001	0.011
HMGB1 (ng/ml)	5.9 (4.3-11.5)	15.7 (8.3-33.3)	<0.001	11.6 (7.5-28.0)	16.0 (8.7-34.5)	0.001	<0.001	NS	22.2 (9.4-34.8)	11.6 (7.7-30.5)	9.0 (6.9-16.8)	14.4 (7.8-34.7)	<0.001	0.026

NS=Not significant, VPA=Valproate, PHT=Phenytoin, CBZ=Carbamazepine, LEV=Levetiracetam, SR=Seizure-recurrence, NSR=Nonseizure-recurrence, IQR=Interquartile range, PWE=Persons with epilepsy, IL=Interleukin, TNF-α=Tumor necrosis factor-α, HMGB1=Highmobility group box protein 1

Liverpool adverse event profile score assessment

Most commonly encountered AEs were anger and aggression, somnolence, and memory loss. These were found to be decreased at 6 months following reduction of the AEDs dose. The mean LAEP score at onset of tapering was 25.3 ± 4.1 and after 6-month follow-up with a reduction in the AED's dose, the LAEP score also reduced to 21.2 ± 3.4 (P = 0. 001), i.e., with significant improvement in AEs profile. However, the comparison of the SR and NSR group did not have any significant difference in LAEP scores.

Discussion

AED tapering even after 2–3 years of seizure-free period is still contentious due to the possibility of SR during and after AED tapering which is about 19%–31% as per previous studies conducted in India.[456] Some of the previous studies have reported the importance of risk factor analysis to optimize the AED tapering pattern, and there is a lack of conclusive evidence for this. To the best of our knowledge, this is one of the initial studies that has correlated serum AEDs levels with the incidence of SR at different time points during AED tapering. Furthermore, there is a lack of confirmatory evidence correlating inflammatory biomarkers level at the onset of tapering with the SR following tapering.

In this study, we found that the frequency of SR was 23.3%, with a maximum of 44 months of follow-up after starting AED tapering. It is in accordance with previous studies such as Medical Research Council Antiepileptic Drug Withdrawal Study Group (1993), with SR 22% and Specchio et al., with SR 18%, after AED tapering in PWE with 2 years of seizure-free period.[2526]

There is a lack of previous studies commenting upon serum AED level and its correlation with SR during tapering, so it is tough to establish an accurate sample size. One previous study by Cardoso et al. has compared lower AED dose continuation versus complete tapering of AED dose after being seizure-free for at least 2 years and opined that once PWE decided to taper down their medication, it should be complete.[13] In the same study, they have found a nonsignificant association of serum AED levels and the risk of SR in adults considering two categories, i.e., therapeutic or higher concentration and low concentration. In the Cardoso et al. study, the daily AED doses and serum levels range at the onset of tapering were similar to the current study. In the current study, the dose at the onset of tapering did not have any significant difference between SR and NSR groups. PWE with a subtherapeutic serum AED levels at the onset of tapering was with a significantly higher chance of SR as compared to those with therapeutic or higher serum AED levels among all monotherapy subjects. However, apart from LEV, other individual AEDs groups analysis did not have such a correlation. This may be due to the small sample size in this study; these preliminary findings suggest the need for a study with a larger sample size to confirm these results.

The presence of risk factors among SR and NSR groups, also in between different categories of serum level, did not yield any significant difference, though this study is not aimed/adequately powered to find out the difference in the presence of risk factors among SR and NSR groups.

As per the literature, IL-1 receptor antagonists and IL-10 have an anti-inflammatory effect and exert a protective and anticonvulsant role in seizure models.[17] Similarly, HMGB1 has been positioned as a unique and emerging target against epileptogenesis as per a recent review article.[19] Another review study has reported that TNF-α, IL-1b, and IL-6 have a significant association with seizures.[18] Significantly higher serum IL-1 βand HMGB1 concentrations were found among febrile seizure subjects than in the control group having fever only (P < 0.05).[19] These data support the inflammatory hypothesis of epileptogenesis. In this study, all PWE undergoing AED tapering at baseline had higher concentrations of inflammatory mediators (TNF-α, IL-6, IL-10, IL-1 β, and HMGB1) as compared to healthy control subjects. PWE with SR had higher IL-1 β and TNF-α as compared to NSR group. These findings suggest that higher inflammatory markers at the onset of tapering have a significant association with SR, and anti-inflammatory therapy targeting cytokines and HMGB1 may have a potential role in the prevention of seizures or subsequent epileptogenesis. There are ample restraints to the findings generated through cytokine estimation in blood as compared to cerebrospinal fluid; however, cytokine release is significant in the process of epileptogenesis, and clinical studies with immunomodulatory treatment can confirm the possible role of cytokines in seizure and their screening potential.

During AED tapering, there is a reduction of AED dose, which will lead to improvement in AED-related AEs. In contrast, precipitation of AEs causes distress to the subject leading to SR. As per previous literature, the rate of discontinuation of AEDs due to AEs varies and is reported as 0.5%–3.3%,[27] 3.9%,[28] and it can be higher as 27% for CBZ, 13% for VPA, or 10% for lamotrigine,[29] and rate of reduction of AED dose due to ADRs is about 48.4%.[3] Hence, this study has done AEs profiling and compared it between PWE with SR and NSR. Although this study found a significant reduction in AEs in due course of tapering, there was no significant difference between PWE with SR and NSR.

Conclusion

The judgment to start or stop the AED tapering entails a careful assessment of risk factors on an individual basis for reduced risk of SR. The current study result opines the possible association of subtherapeutic serum AED level (especially levetiracetam) and higher inflammatory markers at the onset of tapering with the risk of SR in PWE undergoing AED tapering. These concepts necessitate congregating further evidence regarding the role of monitoring serum AED level and inflammatory markers level along with other established risk factors before onset of AED tapering with an aim of optimal management of epilepsy and reduced risk of SR.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.