How is Levetiracetam Monotherapy Currently Monitored in Pregnancy? A Systematic Review

A search up to 17 April 2025 was performed using the Ovid-MEDLINE, Ovid-Embase, and Ovid-Cochrane Central Register of Controlled Trials databases. The search strategy was constructed using the following keywords and their combinations: levetiracetam, drug monitoring, concentration monitoring, dose monitoring, drug monitoring, dose adjustment, dose optimisation, dose individualisation, and pregnancy. No limits were set on the publication dates. Results were imported into Covidence (Veritas Health Innovation, Australia), and duplicates were removed. Titles and abstracts of the remaining articles were independently screened by two investigators according to the inclusion criteria.

Studies were eligible for inclusion if they described levetiracetam dosing, concentrations, monitoring, efficacy, or safety during pregnancy. The following exclusion criteria were defined: (1) data other than human; (2) not focused on epilepsy; (3) not focused on pregnancy; (4) written in a language other than English; (5) sample size fewer than five patients (i.e. case reports); (6) reviews, letters to editors, commentaries, or conference abstracts; (7) lacking relevant data; (8) no information on levetiracetam monotherapy.

The risk of bias was assessed in accordance with version 2 of the Cochrane risk-of-bias tool for randomised trials (RoB 2) [19]. Two investigators independently assessed the included articles for risk of bias, and a third author resolved any discrepancies.

Relevant data were extracted from all included studies using a standardised form considering study design; number of pregnant individuals and pregnancies; mean age; mean daily dose pre-pregnancy, during pregnancy, and post pregnancy; serum concentrations pre-pregnancy, during pregnancy, and post pregnancy; TDM strategies; and efficacy and safety of levetiracetam during pregnancy. If data were sufficient, statistical techniques such as meta-analyses and subgroup analyses were applied. When data were insufficient, fragmented, or too heterogeneous to allow for robust statistical analysis, a qualitative and descriptive analysis was conducted.

To better compare drug concentrations between different dosages, dose normalised concentrations were used. The dose normalised concentration is defined as the plasma or serum concentration adjusted by the administered dose. It is calculated as plasma or serum concentration divided by total daily dose and allows comparison of plasma or serum concentrations when different doses are administered. Change in clearance across different trimesters was assessed. Clearance refers to how efficiently the body eliminates a drug from the bloodstream. It can be calculated as daily dose corrected for bodyweight (mg/kg)/serum level (μg/mL) or by dividing the daily dose by the serum level without correcting for bodyweight.

The risk-of-bias assessment (Table 1) showed that eight studies [7, 20‐24, 26, 27] had a low risk of bias. Two studies [8, 25] had a high risk of bias. The high risk of bias was exclusively the result of missing outcome data. One of the studies [25] also had an unclear risk of bias as it did not specify whether any confounders were considered during analysis. Details on the risk-of-bias assessment are available in Supplemental Data 1 and 2.

Five of the included papers [7, 22, 24, 26, 27] described information on levetiracetam dose before pregnancy, during pregnancy, and/or after pregnancy. Li et al. [27] reported the mean daily dose for the first, second, and third trimester for all samples and the mean daily dose of the samples collected in the morning. Pennell et al. [24] described the median total daily dose for the first, second, and third trimester; the post-partum period; and the non-pregnant (control) group. Reisinger et al. [7] observed that levetiracetam dose increased progressively over the different trimesters. The mean peak daily dose during the non-pregnant period was 2395.8 mg ± standard deviation (SD) 734.4 (range 1000–3500). During the first, second, and third trimesters, the mean peak dose increased to 2466.7 mg ± 766.9 (range 1000–4000), 3083.3 mg ± 811.0 (range 1500–4500), and 3383.3 mg ± 949.0 (range 1500–5000), respectively. Thangaratinam et al. [22] made no distinction between the different trimesters and reported the mean total daily dose before and after randomisation to the TDM or clinical features monitoring group (Table 2). Yin et al. [26] observed a gradual increase in levetiracetam dose throughout the different trimesters. The median total daily dose for the period before conception was 1000 mg (interquartile range [IQR] 500–1690) for the period before conception, 1000 mg (IQR 500–1930) during the first trimester, 1500 mg (IQR 1000–1750) during the second trimester, and 1750 mg (IQR 1500–2250) during the third trimester. All levetiracetam dose and serum concentration values are shown in Table 2.

Data on levetiracetam TDM strategies are shown in Table 3. Most participants had at least one clinical visit per trimester [7, 8, 20‐24, 26, 27], some had monthly visits [7, 22, 23], and others had more frequent visits if seizure control was poor or depending on the individual clinician [7, 22]. Four studies [21, 23, 26, 27] collected blood samples just before the next levetiracetam dose (trough levels), whereas some sample collections took place 3–7 h after levetiracetam ingestion (peak levels) [23, 24]. Li et al. [27] collected samples both before the morning dose and 1 h after. Voinescu et al. [8] documented the time since last dose before blood collection. Three studies did not specify standardised timing of blood draws [7, 20, 22]. Schelhaas et al. [23] gave the participants the option to choose between venous puncture or a dried blood sampling technique. Samples were analysed using high-performance liquid chromatography [8, 21, 26], liquid chromatography–mass spectrometry [24], or ultra-performance liquid chromatography with double mass spectrophotometer [23, 27]. Reisinger et al. [7] analysed all samples using routine clinical laboratory methods.

In most articles, the target concentration was described as a plasma level available from the period before pregnancy [7, 8, 22, 23, 26] or a level after delivery [7, 8, 24] (Table 3). Decisions on dose adjustments based on the target concentrations, seizure type, seizure frequency, levetiracetam-related side effects, and patient’s history were also considered [7, 8, 26].

Data on dose adjustments were reported as the average number of dose adjustments throughout pregnancy [20] or the dose increase compared with baseline or other trimesters [7, 26]. None of the articles published any data on dose adjustment postpartum, although some did collect this information [20, 22].

Table 2 provides information on levetiracetam plasma levels. Only one study did not collect information on plasma levels as most centres in India do not have the facilities to check serum concentrations [25]. Three studies [20, 23, 24] described the dose normalised concentration of levetiracetam calculated as plasma or serum concentration divided by total daily dose (Table 2). Pennell et al. [24] reported the concentrations for separate trimesters and the postpartum period and found that the dose normalised concentrations decreased by up to 36.8% (P < 0.001) compared with the postpartum period. Moreover, significantly lower dose normalised concentrations with gestational age were found (median − 0.06 μg/L/mg per week; standard error [SE] 0.03; P = 0.01) in pregnancy compared with controls [24]. Pregnant individuals also had significantly lower baseline dose normalised concentrations than controls (median − 2.37 μg/L/mg; SE 0.7; P < 0.001) [24]. Freund et al. [20] described the difference in dose normalised concentrations between pre- and post-conception. Schelhaas et al. [23] described monthly dose normalised concentrations, but no information on levetiracetam monotherapy was available. Li et al. [27] reported both mean concentrations and dose normalised concentrations for the first, second, and third trimester for all samples. Additionally, they reported the mean concentrations of the samples that were collected in the morning. Three studies [7, 8, 26] reported on clearance, with Voinescu et al. [8] describing an overall mean relative levetiracetam clearance of 1.710 (95% confidence interval [CI] 1.42–2.06) in the first trimester, 1.421 (95% CI 1.68–1.73) in the second trimester, and 1.367 (95% CI 1.15–1.63) in the third trimester, indicating that peak clearance occurred in the first trimester. Reisinger et al. [7] reported a clearance of 1.09 ± SD 0.30 in the nonpregnant baseline. Clearance was 2.16 ± 1.72 in the first trimester, 3.35 ± 2.60 in the second trimester, and 2.15 ± 1.11 in the third trimester. Yin et al. [26] reported that peak clearance occurred in the second trimester. The average peak clearance was 2.78-fold (P < 0.001) higher during pregnancy than at baseline [26]. Both Reisinger et al. [7] and Yin et al. [26] reported peak clearance in the second trimester.

Seven articles assessed efficacy [7, 8, 20, 22, 25‐27], but only five articles [7, 20, 25‐27] reported these data for levetiracetam monotherapy (Table 4). Bansal et al. [25] reported that, even though 21 of 28 patients experienced good seizure control pre-conception (defined as no seizures in the 9 months before conception), 12 subjects had one or more seizures during pregnancy. Two other studies [7, 26] reported an increase in seizure frequency during pregnancy in 46.7% [7] and 53.3% [26] of the pregnancies. However, Li et al. [27] found no statistically significant difference in seizure frequency between non-pregnant and pregnant individuals. Freund et al. [20] observed three pregnancies with breakthrough seizures or an increased seizure frequency while treated with levetiracetam monotherapy. In one patient, they were able to compare pre-pregnancy levetiracetam levels and that at the time of the seizure. This showed a decrease in dose normalised concentration by 73%. Yin et al. [26] analysed the effect of the ratio-to-target concentration value on seizure frequency but found no significant effect. Only two articles [22, 25] analysed the safety of levetiracetam use during pregnancy (Table 4). Thangaratinam et al. [22] demonstrated that an increased levetiracetam dose did not affect maternal and neonatal outcomes. However, they did find higher cord blood levels in the TDM group (22.5 mg/L ± SD 17.0) than in the clinical features monitoring group (13.9 mg/L ± SD 10.5), with a mean difference of 7.8 mg/L (95% CI 0.86–14.8). Bansal et al. [25] reported maternal and neonatal outcomes and analysed whether there was a difference between levetiracetam and carbamazepine, sodium valproate, or phenytoin. They found an increased occurrence of seizures during pregnancy when comparing levetiracetam and sodium valproate (P = 0.031). Those on phenytoin were at increased risk of foetal distress than were those receiving levetiracetam (P = 0.003). No information on levels was reported.

The efficacy of many antiepileptic agents, including levetiracetam, is ideally based on personalised target concentrations [28]. Therefore, it is difficult to compare levetiracetam treatments between individuals, especially during pregnancy. The ideal target should be based on pre-pregnancy levels at which no or few seizures occurred [28]. Valuable information on the changes in levetiracetam pharmacokinetics and clearance during pregnancy can be calculated from changes in their levels. However, only 3 of 10 studies provided information on levetiracetam clearance [7, 8, 26]. Two studies calculated clearance by dividing daily dose corrected for bodyweight (mg/kg)/serum level (μg/mL) [7, 8], and one calculated the ratio of the total daily dose (mg/day) to serum level (mg/L) [26]. All three studies described a significant increase in clearance, reported as ratios (1.71 [P 0.0001] [8]; 2.78 [P < 0.001] [26]) or as a percentage (224% [P = 0.014] [7]), and Voinescu et al. [8] concluded that peak clearance occurred in the first trimester. However, Reisinger et al. [7] and Yin et al. [26] found a peak clearance in the second trimester. These three studies reported peak clearance at different trimesters, substantiating the fact that more information on the pharmacokinetics of levetiracetam during pregnancy is needed. The studies did not report the point during pregnancy at which clearance starts to increase, and this is valuable information for clinicians, so further research on this topic is needed.

Even though most papers [7, 8, 20‐24, 26, 27] obtained levetiracetam plasma levels regularly, only three [7, 8, 26] clearly described that dose adjustments were based on plasma level results. Moreover, none of the articles used a standardised dose adjustment protocol. Most clinicians used the pre-pregnancy level as target concentration when making dose adjustments [7, 8, 26]. However, levetiracetam levels before pregnancy were not always available [7, 8]. When pre-pregnancy levels are available, using a ratio-to-target concentration threshold could be a useful approach for making dose adjustments. For lamotrigine, another antiepileptic drug frequently used during pregnancy, different studies have already described a ratio of 0.65, meaning that, if the lamotrigine concentration drops below 65% of the target concentration, increased seizure frequency is likely [7, 9]. Only one study [23] assessed this value for levetiracetam monotherapy. Schelhaas et al. [23] advised keeping the ratio-to-target concentration for levetiracetam above 0.65 for those with seizures in the year before conception. Nevertheless, this recommendation should be interpreted with care as the analysis included only 15 patients. The analysed sensitivity for this value was approximately 73%; specificity was around 62%, which means some individuals will have a dose increase during pregnancy without being at risk. However, early levetiracetam dose increases have a less negative impact in pregnancy than the potential harm caused by a deterioration in seizure control [23].

The efficacy of levetiracetam might decrease during pregnancy because of the more efficient renal clearance. To compensate for increased clearance, dose increases are necessary to maintain the same plasma concentration. Despite substantial dose increments, greater seizure frequency during pregnancy is still reported [7, 26]. One possible explanation is that the subjects were not yet aware of their pregnancy early in the first trimester. Therefore, seizures occurred as levetiracetam levels dropped before doses were adjusted [20].

Besides maintaining adequate plasma levels, being seizure free during the 9–12 months before conception also reduces the risk of seizures during pregnancy [23, 26]. These findings were supported by Yin et al. [26], who reported that only 13.3% (2/15) of the subjects with adequate seizure control in the 9 months before pregnancy experienced a deterioration of seizure control compared with 75% (6/8) of those who were not seizure free the 9 months before conception. Therefore, optimal seizure control pre-pregnancy leads to better seizure control during pregnancy.

None of the studies that clearly described TDM-based dose adjustments reported information on adverse maternal and neonatal effects. Therefore, it is an important area for future studies to explore.

The significant heterogeneity in study designs, outcome measurements, and available data made it challenging to analyse and interpret combined results. Additionally, only few data on adverse outcomes were available. This made it hard to address the safety of TDM-guided dosing of levetiracetam during pregnancy. Moreover, some data had to be extracted from a graph, which may have resulted in a loss of accuracy for some values. Despite these limitations, this review still identifies factors that can be improved in future levetiracetam TDM studies.

Levetiracetam dose adjustments are often necessary during pregnancy. However, other antiepileptics, such as lamotrigine and oxcarbazepine, also require dose adjustments during pregnancy, highlighting the need for TDM guidelines [29]. To address the lack of guidelines for levetiracetam TDM during pregnancy, future studies should focus on the standardization of TDM strategies. It is important that future studies include the collection of a pre-pregnancy sample so researchers can determine the ideal (personalised) target concentration or validate a ratio-to-target concentration. Monthly clinical checkups are also advised; these will give us detailed information across the different trimesters. For more robust results, future studies should be prospective in nature and include larger numbers of participants receiving levetiracetam during their pregnancy. Values should be reported as true drug concentration data rather than dose–concentration ratios. To enable better advice on levetiracetam dose adjustments during pregnancy, a population pharmacokinetic model in pregnancy needs to be developed. This should be followed by the use of dose simulations of different patient groups (e.g. different trimesters, comorbidities, concurrent medication, etc.) [30]. Finally, to guide personalised dose adjustments, the Bayesian approach can be used [30, 31]. Long-term outcomes of the offspring exposed in utero should also be assessed. The efficacy and safety of levetiracetam treatment during pregnancy can then be compared with that of other antiepileptic medications. To reduce protracted turnaround times and complex quantification methods, future studies could also look at point-of-care testing using standardised laboratory methods.