Effect of a Declined Plasma Concentration of Valproic Acid Induced by Meropenem on the Antiepileptic Efficacy of Valproic Acid

ABSTRACT

Objective

This study aimed to indicate whether a declined plasma concentration of valproic acid (VPA) induced by co‐administration of meropenem (MEPM) could affect the antiepileptic efficacy of VPA.

Methods

We retrospectively reviewed data of hospitalized patients who were diagnosed with status epilepticus or epilepsy between 2010 and 2019. Patients co‐administered VPA and MEPM during hospitalization were screened and assigned to the exposure group, while those co‐administerd VPA and other broad‐spectrum antibiotics were allocated to the control group.

Results

The exposure group and control group included 50 and 11 patients, respectively. With a similar dosage of VPA, the plasma concentration of VPA significantly decreased during co‐administration (24.6 ± 4.3 μg/mL) compared with that before co‐administration (88.8 ± 13.6 μg/mL, p < 0.0001), and it was partly recovered with the termination of co‐administration (39.8 ± 13.2 μg/mL, p = 0.163) in the exposure group. The inverse probability of treatment weighting estimated the treatment efficacy via changes in seizure frequency, seizure duration, and concomitant use of antiepileptic drugs, which were not significantly different between the exposure and control groups. In the exposure group, there was no significant differences in seizure frequency between the periods of before‐during and before‐after (p = 0.074 and 0.153, respectively). Seizure duration during VPA–MEPM co‐administration was not significantly different from that before co‐administration (p = 0.291).

Conclusions

In this study, the reduced plasma concentration of VPA induced by the co‐administration of MEPM did not affect the antiepileptic efficacy of VPA. This conclusion should be interpreted with caution, and more research is warranted.

Trial Registration

Chinese Clinical Trial Registry: ChiCTR2000034567. Registered on 10 July 2020

We retrospectively reviewed data of hospitalized patients who were diagnosed with status epilepticus or epilepsy between 2010 and 2019. In this study, the reduced plasma concentration of VPA induced by the co‐administration of MEPM did not affect the antiepileptic efficacy of VPA. This conclusion should be interpreted with caution and more research is warranted.

1. Introduction

Valproic acid (VPA) is an effective broad‐spectrum antiepileptic drug (AED) that is widely used for the management of generalized epilepsies associated with multiple seizure types [1]. Glucuronidation is the most important metabolic pathway of VPA [2]. Meropenem (MEPM), a broad‐spectrum carbapenem antibiotic with strong antimicrobial activity against gram‐negative, gram‐positive, and anaerobic bacteria, has a strong indication when the infections are resistant to the third‐ or fourth‐generation cephalosporins [3, 4]. MEPM possesses a low seizure risk, accounting for only 0%–0.7% [5

Summary.

• This study aimed to investigate whether a reduced plasma concentration of VPA induced by co‐administration of MEPM could affect the antiepileptic efficacy of VPA.

• In this study, the declined plasma concentration of VPA induced by the co‐administration of MEPM did not affect the antiepileptic efficacy of VPA.

• This conclusion should be interpreted with cautions, and more research is warranted.

], and is highly appropriate for patients with neurologic disorders when carbapenem antibiotics are indicated.

MEPM was reported to interact with VPA, with one of the mechanisms as the inhibition of β‐glucuronidase. A previous study indicated that VPA‐glucuronidase (VPA‐G) could be hydrolyzed back to VPA by acylpeptide hydrolase (APEH) [6], and the activity of APEH was reversibly and irreversibly suppressed by MEPM [7]. A decrease in activity of APEH by MEPM may cause increased urinary excretion of VPA‐G with concomitant reduction in plasma VPA concentration [8, 9]. A reduction in VPA plasma concentration due to co‐administration of MEPM emerges apparently within 24 h [10] and may remain low for 7 days and then gradually increase to baseline until 8–14 days later [11].

The antiepileptic efficacy of VPA is dose‐ and plasma concentration‐dependent [1], with a therapeutic range of 50–100 mg/L. Therefore, it is recommended that the concomitant use of VPA with MEPM should be avoided [12, 13]. However, clinical experience from patients suffered from the VPA–MEPM interaction is contradictory, and a number of scholars have argued that there may be no need to change VPA to another antiepileptic agent during MEPM co‐administration [10, 14, 15, 16, 17].

In China, the guiding principles for clinical application of antibacterial drugs have forbidden the co‐administration of VPA and MEPM since 2015. According to our clinical practice before 2015, it seemed that such co‐administration did not affect the antiepileptic effect of VPA. Thus, we did not withdraw VPA only because of the indication for MEPM even after 2015. In the present study, we conducted a retrospective observational analysis on patients with concomitant administration of VPA and MEPM via a self‐contrasted method and compared our findings with those of other broad‐spectrum antibiotics, in order to investigate the effects of such co‐administration on the antiepileptic efficacy of VPA.

2. Methods

2.1. Patients and Data Collection

We retrospectively reviewed the data of patients who were admitted to the Department of Neurology, Nanfang Hospital, Southern Medical University (Guangzhou, China), from June 2010 to January 2019. Patients who had received both VPA (tablet: 0.5 g, and/or injection: 0.4 g; Sanofi‐Aventis, Paris, France) and MEPM (0.5 g; Sumitomo Dainippon Pharma Co. Ltd., Oita, Japan) during hospitalization were screened as potential candidates. The inclusion criteria were as follows: (1) patients who aged ≥18 years old; (2) patients who were diagnosed with epilepsy or status epilepticus (SE). The exclusion criteria were as follows: (1) VPA was not concomitantly used with MEPM or in the form of co‐administration for <24 h; (2) VPA used for seizure prophylaxis; (3) incomplete data. Patients with co‐administration of VPA and other broad‐spectrum antibiotics (e.g., ceftazidime, cefepime, piperacillin‐tazobactam, and cefoperazone‐sulbactam) during the same study period were set as controls. The study period was divided into three timeframes: before co‐administration, during co‐administration, and within 14 days after the withdrawal of co‐administration [18].

Patients' demographic and clinical data were collected retrospectively, including gender, age, diagnosis (epilepsy or SE), etiology of epilepsy, hepatic function (alanine aminotransferase [ALT], aspartate aminotransferase [AST], and albumin [ALB]), renal function (creatinine clearance rate [Ccr] [Cockcroft–Gault formula]), infection index (white blood cell [WBC], and C‐reactive protein [CRP]), seizure attacks (based on medical records and results of electroencephalogram [EEG]), withdrawal of VPA and the causes, and time and etiology of death. Data related to indication for MEPM, and duration of co‐administration of VPA with MEPM were collected as well.

The following information was collected based on the abovementioned timeframes: dosage and plasma concentration of VPA; records of seizures, including the seizure frequency, seizure duration, and type of AEDs; and safety assessment (i.e., hepatic and renal function, plasma ammonia, and the incidence of new‐onset SE).

An estimated Ccr < 60 mL/min [19] was defined as renal insufficiency. The plasma concentration of VPA was determined by enzyme‐multiplied immunoassay technique (EMIT; Via‐ProE system, Siemens Healthineers, Erlangen, Germany), with a therapeutic range of 50–100 mg/L. When the plasma concentration was tested at the three timeframes (i.e., 3 days [five half‐time] after the administration of VPA, 24 h after the co‐administration, and within 14 days after the termination of co‐administration), the plasma concentration was recorded to the corresponding study period.

The study protocol was approved by the Ethics Committee of Nanfang Hospital. Informed consent was waived by the institutional review board owning to the retrospective nature of the study.

2.2. Assessment of Efficacy

The antiepileptic efficacy of VPA was evaluated by comparing the doses of VPA, the seizure frequency, seizure duration, and type of concomitant use of AEDs based on the three timeframes. Seizure frequency was defined as the total episode of seizures (any type) divided by the number of weeks for each patient [20], and presented as episodes/week/person [21]. The seizure duration was calculated by dividing total duration by the number of episodes. For each episode, the duration was defined as the period between seizure onset and termination [22]. The seizure frequency and seizure duration were formulated as follows:

$$\begin{array}{c}\text{Seizure frequency}\left(\text{episodes}/\text{week}/\text{person}\right)\hfill \\ =\mathrm{sum}\text{of seizure frequency of each patient}/\hfill \\ \text{number of}\mathrm{all}\text{the patients},\hfill \end{array}$$

$$\begin{array}{c}\text{Seizure duration}\left(\min /\text{episode}\right)\hfill \\ =\text{total duration}/\text{number of episodes}.\hfill \end{array}$$

Type of concomitant administration of AEDs for each patient was recorded for making comparison as well.

The clinical endpoints included: (1) change in seizure frequency from baseline to the start of co‐administration; (2) change in seizure frequency from baseline to the end of co‐administration; (3) change in seizure duration from baseline to the start of co‐administration; (4) change in seizure duration from baseline to the end of co‐administration; (5) change in concomitant use of AEDs from baseline to the start of co‐administration; (6) change in concomitant use of AEDs from baseline to the end of co‐administration.

2.3. Statistical Analysis

Baseline and demographic characteristics were summarized as mean ± standard deviation or median (interquartile range [IQR]) for continuous variables and percentage for categorical variables. Differences in baseline characteristics between patients in the exposure group and the control group were statistically analyzed. To figure out the potential differences in baseline characteristics between groups, the method of inverse probability of treatment weighting (IPTW) [23] was utilized. The patients' probability of receiving the treatment was estimated via propensity score. A propensity score model was fitted that could incorporate baseline covariates based on both experts' opinion and results of univariate analysis. In the IPTW, the predicted probabilities from the propensity score model were used to calculate the inverse probability weights. The Mann–Whitney U test was employed to compare changes in the clinical endpoints between the exposure group and the control group. The Wilcoxon signed‐rank test and McNemar's test without IPTW were employed to compare clinical endpoints before and after treatment in the exposure group. Multiple imputation was used to handle missing data. All the statistical analyses were performed by SPSS 22.0 software (IBM Corp., Armonk, NY, USA). The level of significance was set to 0.05.

3. Results

Among 102 patients who were screened for eligibility, 50 were included in the exposure group (Figure 1). Besides, 39 patients were excluded because VPA and MEPM were not used concomitantly or co‐administration lasted for ≤24 h; two cases were excluded because VPA was used for the prevention of epilepsy; 11 patients were excluded due to incomplete data. In addition, 11 patients who were co‐administered VPA and other broad‐spectrum antibiotics were assigned to the control group (Figure 1).

FIGURE 1.

Flowchart for patient selection. MEPM, meropenem; VPA, valproic acid.

3.1. Baseline Demographic and Clinical Characteristics of the Patients

Table 1 summarizes the demographic and clinical characteristics of the patients in exposure and control groups. There were no significant differences in gender, age, type of epilepsy, diagnosis, ALT, AST, WBC, CRP, baseline dosages of VPA, seizure frequency, seizure duration, and concomitant administration of AEDs between the exposure and control groups. Ccr level was higher in the exposure group (p = 0.024), and both groups possessed a sufficient renal function. ALB level was lower in the exposure group than that in the control group (p = 0.034), suggesting that the inflammatory status was more serious in the exposure group than that in the control group. Although the infection index was not significantly different between the two groups, the exposure group tended to have a higher CRP level (p = 0.059).

TABLE 1.

Baseline characteristics of the patients.

Characteristics	Control (N = 11)	Exposure (N = 50)	p Value*
Gender
Male	6 (54.5%)	34 (68.0%)	0.617
Female	5 (45.5%)	16 (32.0%)
Age	59.18 ± 19.87	54.94 ± 18.86	0.506
Type of epilepsy
Epilepsy	9 (81.8%)	42 (84.0%)	1.000
SE	2 (18.2%)	8 (16.0%)
Diagnosis
1	5 (45.5%)	17 (34.0%)	0.835
2	3 (27.3%)	13 (26.0%)
3	1 (9.1%)	6 (12.0%)
4	1 (9.1%)	2 (4.0%)
5	0 (0.0%)	1 (2.0%)
6	1 (9.1%)	11 (22.0%)
ALT	47.02 ± 55.95	33.73 ± 42.59	0.382
AST	148.70 ± 339.00	51.47 ± 68.18	0.366
Ccr	69.61 ± 28.78	100.82 ± 71.76	0.024
ALB	41.73 ± 10.26	33.57 ± 5.43	0.034
WBC	15.64 ± 6.59	12.52 ± 4.52	0.064
CRP	32.64 ± 38.44	74.89 ± 70.10	0.059
Dose of VPA	1.70 ± 0.82	1.47 ± 0.70	0.470
Seizure frequency	0.00 (0.00, 0.09)	0.00 (0.00, 0.00)	0.359
Seizure duration	0.00 (0.00, 1.00)	0.00 (0.00, 0.00)	0.763
Concomitant AEDs
No	9 (81.8%)	32 (64.0%)	0.432
Yes	2 (18.2%)	18 (36.0%)

Note: Continuous variables are represented as mean ± SD or median (P25, P75), while dichotomous variables are represented as number (%). Diagnosis: 1 cerebrovascular disease, 2 meningitis or encephalitis, 3 postcardiac arrest encephalopathy, 4 traumatic brain injury, 5 prior epilepsy, 6 others. Bold values indicate p < 0.05.

Abbreviations: AEDs, antiepileptic drugs; ALB, albumin; ALT, alanine aminotransferase; AST, aspartate aminotransferase; Ccr, creatinine clearance rate; CRP, C‐reactive protein; P25, the 25th percentile; P75, the 75th percentile; SD, standard deviation; SE, status epilepticus; VPA, valproic acid.

^*Baseline characteristics with p ≤ 0.05 will be incorporated into a propensity score estimation model to adjust for possible confounding effect.

Moreover, the indication for MEPM included: third‐ and fourth‐generation cephalosporins‐resistant pneumonia, urinary infection and pneumonia, septicemia, and intracranial infection. The median duration of VPA administration before co‐administration was 8 (range, 1–51) days. The median duration of co‐administration was 7.5 (range, 1–52) days.

3.2. The Dosages and Plasma Concentrations of VPA in the Exposure and Control Groups

The dosages and plasma concentrations of VPA in the exposure and control groups in the three timeframes are shown in Figure 2. In the exposure group, the dosage of VPA was 1.5 ± 0.1 g/day (n = 33) before co‐administration, 1.5 ± 0.08 g/day (n = 50) during co‐administration, and 1.4 ± 0.07 g/day (n = 44) after the withdrawal of co‐administration (p = 0.769). The plasma concentration of VPA significantly decreased during co‐administration (24.6 ± 4.3 μg/mL) compared with that before co‐administration (88.8 ± 13.6 μg/mL, p < 0.0001), and partly recovered with the termination of co‐administration (39.8 ± 13.2 μg/mL, p = 0.163). In the control group, both dosage and plasma concentration of VPA remained stable (p = 0.379 and 0.129, respectively).

FIGURE 2.

Concentration/dosages ratio of VPA before, during, and after co‐administration in the exposure and control groups. *p < 0.05 and **p < 0.01, respectively.

3.3. The IPTW Estimated Treatment Efficacy in Both Exposure and Control Groups

Changes in seizure frequency and seizure duration between exposure group and control group are listed in Table 2. Changes in seizure frequency for during‐before were  −0.004 ± 0.017 in the exposure group and −0.002 ± 0.080 in the control group (p = 0.328). Changes in seizure frequency for after‐before were −0.004 ± 0.020 in the exposure group and −0.026 ± 0.099 in the control group (p = 0.505).

TABLE 2.

IPTW‐estimated changes in the seizure frequency and seizure duration between control and exposure groups.

	Control	Exposure	p Value
	Mean ± SD
Seizure frequency
Δduring‐before	−0.002 ± 0.080	−0.004 ± 0.017	0.328
Δafter‐before	−0.026 ± 0.099	−0.004 ± 0.020	0.505
Seizure duration
Δduring‐before	0.228 ± 0.950	−0.186 ± 1.615	0.172
Δafter‐before	−0.125 ± 1.155	−0.397 ± 1.192	0.983

Changes in seizure duration for during‐before were −0.186 ± 1.615 in the exposure group and 0.228 ± 0.950 in the control group (p = 0.172). Changes in seizure duration for after‐before were −0.397 ± 1.192 in the exposure group and −0.125 ± 1.155 in the control group (p = 0.983).

Changes in concomitant use of AEDs in both exposure and control groups are summarized in Table 3. There were no significant differences in the concomitant administration of AEDs between the two groups during and after administration of MEPM (p = 0.507 and 0.065, respectively).

TABLE 3.

IPTW‐estimated changes in the use of concomitant AEDs between control and exposure groups.

Change from baseline	Control count (%)	Exposure count (%)	p Value
During
From yes to no	0 (0.0)	5 (8.3)	0.507
No changes	49 (90.7)	46 (76.7)
From no to yes	5 (9.3)	9 (15.0)
After
From yes to no	0 (0.0)	10 (16.7)	0.065
No changes	53 (96.4)	45 (75.0)
From no to yes	2 (3.6)	5 (8.3)

3.4. Antiepileptic Efficacy in the Exposure Group

Seizure frequency and seizure duration in the three study periods in the exposure and control groups are shown in Table 4. Before co‐administration, seizure was found in 10 (20%) patients, with an average frequency of 0.007 ± 0.017 episodes/week/person. During co‐administration, 7 (14%) patients experienced seizure, with an average frequency of 0.003 ± 0.008 episodes/week/person. After co‐administration, 3 (6%) patients experienced seizure, with an average frequency of 0.003 ± 0.012 episodes/week/person. There were no significant differences in seizure frequency between the periods of before‐during and before‐after (p = 0.074 and 0.153, respectively).

TABLE 4.

Seizure frequency and seizure duration in control and exposure groups.

	Control	Exposure	p Value*	p Value**
	Mean ± SD
Seizure frequency
Before	0.041 ± 0.085	0.007 ± 0.017	0.074E	0.153E
During	0.025 ± 0.042	0.003 ± 0.008
After	0.025 ± 0.082	0.003 ± 0.012
Seizure duration
Before	0.364 ± 0.674	0.500 ± 1.182	0.291E	0.019E
During	0.364 ± 0.674	0.300 ± 1.069
After	0.455 ± 1.508	0.080 ± 0.340

Abbreviations: E, exposure; SD, standard deviation.

^*Wilcoxon signed‐rank test for before vs. during.

^**Wilcoxon signed‐rank test for before vs. after.

The average seizure duration was 0.500 ± 1.182 min/episode before co‐administration, 0.300 ± 1.069 min/episode during co‐administration, and 0.080 ± 0.340 min/episode after co‐administration. Seizure duration during VPA–MEPM co‐administration was not significantly different from that before co‐administration (p = 0.291). However, seizure duration after co‐administration was significantly shorter than that before co‐administration (p = 0.019).

The data related to the concomitant use of AEDs in the three timeframes in the exposure group are presented in Tables 5 and 6. Before co‐administration, 17 (34%) patients were not treated with AEDs, 15 (30%) with VPA monotherapy, and 18 (36%) with concomitant administration of AEDs. During co‐administration, 31 (62%) patients were treated with VPA monotherapy and 19 (38%) with concomitant use of AEDs. After co‐administration, 4 (8%) patients were not treated with AEDs, 31 (62%) patients were treated with VPA monotherapy and 13 (26%) with concomitant administration of AEDs. There were no significant differences in the concomitant use of AEDs between the periods of before‐during and before‐after (p = 1.000 and 0.267, respectively).

TABLE 5.

Antiepileptic drugs in the exposure group.

	Before, n (%)	During, n (%)	After, n a (%)
None AED	17 (34.0)	0 (0.0)	4 (8.0)
VPA monotherapy	15 (30.0)	31 (62.0)	31 (62.0)
≥2 types of AEDs (including VPA)	18 (36.0)	19 (38.0)	13 (26.0)
2	13 (26.0)	14 (28.0)	9 (18.0)
3	4 (8.0)	4 (8.0)	4 (8.0)
4	1 (2.0)	1 (2.0)	0 (0.0)
LEV	8 (16.0)	10 (20.0)	5 (10.0)
CNP	1 (2.0)	1 (2.0)	1 (2.0)
Midazolam	1 (2.0)	1 (2.0)	1 (2.0)
OXC	1 (2.0)
PB		1 (2.0)	1 (2.0)
LTG	1 (2.0)
Diazepam	1 (2.0)
CBZ		1 (2.0)
TPM			1 (2.0)
LEV, midazolam	1 (2.0)	2 (4.0)	1 (2.0)
LEV, CNP			1 (2.0)
LEV, TPM		1 (2.0)	1 (2.0)
LEV, LTG		1 (2.0)	1 (2.0)
LEV, CBZ	1 (2.0)
Diazepam, midazolam	1 (2.0)
PB, midazolam	1 (2.0)
LEV, midazolam, CNP	1 (2.0)

Abbreviations: AED, antiepileptic drug; CBZ, carbamazepine; CNP, clonazepam; LEV, levetiracetam; LTG, lamotrigine; OXC, oxcarbazepine; PB, phenobarbital; TPM, topiramate; VPA, valproic acid.

^a Two patients expired.

TABLE 6.

Use of concomitant AEDs in the exposure group.

	Before		p Value*
	No a	Yes
	Count (%)	Count (%)
During
No	26 (81.3)	5 (27.8)	1.000
Yes	6 (18.7)	13 (72.2)
After
No	28 (87.5)	9 (50.0)	0.267
Yes	4 (12.5)	9 (50.0)

Abbreviation: AEDs, antiepileptic drugs.

^a Non‐AED and valproic acid monotherapy.

^* p Value from the McNemar test.

3.5. Safety Assessment in the Exposure Group

There were no significant differences in either ALT or AST levels (p = 0.221, p = 0.266, respectively) or estimated Ccr levels (p = 0.246) among the three timeframes in the exposure group. Besides, five patients withdrew VPA due to hyperammonemia (77 and 82.3 μmol/L, n = 2), ≥3 times upper limit of normal (ULN) for serum ALT level (n = 2) [24], and decreased number of WBCs (n = 1). Two (4.0%) patients died within the study period, of who the causes were acute progressive brainstem infarction and mediastinal emphysema, respectively.

4. Discussion

In the present study, we found that the reduced plasma concentration of VPA induced by the co‐administration of MEPM did not affect the antiepileptic efficacy of VPA. With a similar dosage of VPA, the plasma concentration of VPA decreased with the concomitant administration of MEPM and it was recovered with the termination of combination. However, co‐administration did not increase the seizure frequency and seizure duration.

A previous study indicated that the plasma concentration of VPA promptly decreased when it was concomitantly used with MEPM [25]. In the present research, the plasma concentration of VPA reduced with the concomitant administration of MEPM from 88.8 to 24.6 μg/mL, which is consistent with the results of previous studies. Because it has been reported that such a decrease in VPA plasma concentration cannot be reversed by increasing VPA dosage [10], we did not change VPA dosage. Haroutiunian et al. [18] suggested that the decline of plasma concentration may remain for 7 days, and then, gradually increase to baseline after 8–14 days later. In the present research, it was revealed that the plasma concentration of VPA was recovered within 14 days after the termination of combination from 24.6 to 39.8 μg/mL with no changes in VPA dosage. Due to the inconsistency of period of examination, some within 7 days and some between 7 and 14 days, the plasma concentration of VPA after co‐administration did not recover to baseline.

As the antiepileptic efficacy of VPA is dose‐ and plasma concentration‐dependent, the combination of VPA and MEPM should be avoided based on some experts' recommendations [12, 26, 27, 28]. However, controversial outcomes were reported [10, 15, 17]. In the present study, we found that the reduced plasma concentration of VPA induced by the co‐administration of MEPM did not affect the antiepileptic efficacy of VPA, seizure frequency, seizure duration, and the concomitant administration of AEDs. VPA is highly albumin‐bound, and only the free fraction is pharmacologically active [29]. Some studies showed that plasma binding of VPA was unchanged by carbapenem antibiotics in rats and rabbits [30, 31]. In addition, Hobara et al. [32] reported that carbapenem antibiotics increased the unbound fraction of VPA in rats and humans. Therefore, it may be assumed that the unchanged or increased unbound VPA fraction would justify why the MEPM‐induced declined plasma concentration of VPA did not affect the antiepileptic efficacy of VPA. In the current study, the serum ALB level was lower in the exposure group, which may contribute to the unchanged antiepileptic efficacy. Thus, further prospective pharmacokinetic studies should be performed to confirm such hypothesis.

To date, controversial outcomes have been reported by a number of case reports [16, 33]. They all reported cases with concomitant administration of VPA with carbapenem antibiotics, whose epilepsy was relapsed during co‐administration. They suggested that reduction of VPA concentration promoted the relapse of epileptic seizures in previously controlled patients. However, the seizures disappeared only once carbapenem was withdrawn, when the plasma concentration of VPA was still under therapeutic range and remained low for some days. Therefore, they speculated that the relapse was more likely due to the seizure susceptibility of carbapenem rather than reduced plasma concentration of VPA [34]. However, in contrast to previous researches, the present study showed that there was no significant difference in the frequency and duration of seizures in the exposure group. This may be due to the limited sample size. The other possible reason may be that the majority of patients did not have SE, and the under therapeutic level of VPA may be sufficient to control the seizures. Thus, a larger study with severe epileptic patients should be conducted to more precisely elucidate the effects of co‐administration of VPA and MEPM. AEDs should be tapered rather than halted abruptly [35]. Rapid changes (over days to a few weeks) in drug treatment increase the risk of seizures. In particular, VPA is associated with withdrawal of seizures and should be discontinued gradually, and at least approximately 2 to 4 weeks may be required to reach a steady‐state therapeutic level. This may increase expenses and duration of hospitalization. Based on our results, when MEPM was indicated, there seemed to be no need to change the AED, which could shorten the treatment duration and decrease the seizure risk. However, further prospective studies are warranted to confirm the results of the current research.

There are several limitations in the present study. First, the retrospective nature of this study made it susceptible to selection bias and information bias. Second, this retrospectively observational study was restricted by a small sample size. Third, seizure susceptibility may be changed by several factors, especially in patients with SE, which can therefore be a confusing factor. Fourth, the free VPA serum concentrations were not available in this study. Hence, further multicenter, prospective studies with a larger sample size are warranted to explore the influences of MEPM on free VPA serum concentration and to confirm the results of this study.

5. Conclusions

In this study, the co‐administration of MEPM did not influence the antiepileptic efficacy of VPA. This conclusion should be interpreted with caution, and more research is warranted.

Author Contributions

Shengnan Wang and Dongmei Wang designed the research. Chunping Gu and Yongfang Zhang drafted the manuscript. Kaibin Huang, Zhenzhou Lin, and Qiong Chen collected the data. Yan Chen and Yongming Wu analyzed the data and performed the statistical analysis. All the authors reviewed and approved the submitted version of the manuscript.

Conflicts of Interest

The authors declare no conflicts of interest.

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.