The Alvarez et al. (2025) study of 47 breastfed infants provides a clinically actionable finding: all infants whose mothers had a lamotrigine dose of ≤150 mg/day had undetectable plasma concentrations and no symptoms during follow-up. Among infants whose mothers took higher doses, 14% had concentrations within the therapeutic reference range for epilepsy treatment, and one had clinical symptoms probably related to lamotrigine exposure. Three of 21 tested infants had elevated liver transaminases.

This suggests maternal dose ≤150 mg/day represents a practical threshold below which simplified monitoring is appropriate. For bipolar disorder, where typical maintenance doses are 100–200 mg/day, many patients will fall at or near this threshold.

Valproate's low RID (1–6%) is driven by its high protein binding (~90%), which traps the drug in maternal circulation regardless of the infant's metabolic capacity. Lamotrigine's moderate RID (2–21%; mean 9%) combined with the infant's profoundly immature UGT1A4 creates a "double hit": more drug reaches the infant via milk, and the infant clears it more slowly than an adult would.

This is why the MONEAD study found that the only factor significantly associated with infant plasma lamotrigine concentrations was the maternal lamotrigine concentration — the infant's own clearance capacity is too immature to serve as a reliable safety buffer. This pharmacokinetic reality is what makes the postpartum dose reset so critical.

SJS/TEN from lamotrigine is a dose-escalation phenomenon. In adults, the risk is highest during the initial titration period (first 8 weeks) and is strongly associated with rapid dose escalation, concomitant valproate use, and HLA-B*15:02 positivity. A breastfed infant is not undergoing dose titration — the exposure is continuous and relatively stable (assuming the maternal dose is stable). Furthermore, the infant has been exposed to lamotrigine in utero throughout pregnancy, meaning the immune system has already encountered the drug.

While vigilance is appropriate, the mechanistic basis for SJS risk via breast milk is fundamentally different from the therapeutic dose-titration scenario. No confirmed cases of SJS from breast milk lamotrigine exposure have been reported in the literature.

Unlike valproate (which carries a risk of thrombocytopenia), lamotrigine was associated with mild thrombocytosis (elevated platelet count) in 7 of 8 breastfed infants in the Newport et al. study. This is the opposite hematologic effect. The clinical significance is unknown, and no adverse outcomes were attributed to it. However, clinicians should be aware that if a CBC is obtained in a lamotrigine-exposed breastfed infant, an elevated platelet count may be an expected finding rather than a pathological one.

ACOG recommends "carefully but expeditiously" down-titrating lamotrigine postpartum. The MONEAD real-world data show that in practice, most clinicians tapered by ~100 mg/day increments, reaching a mean of ~116% of the conception dose by 6 weeks postpartum — not an immediate return to baseline. The Khmelev et al. (2025) study validated an empiric taper protocol (dose adjustments at days 1, 7, and 21 postpartum) and found it was not linked to increased seizure risk and was well tolerated. The PBPK model suggests reducing by 25% weekly for 3 weeks.

The key message: the dose should come down promptly but not precipitously. An abrupt return to pre-pregnancy dose on Day 1 risks breakthrough seizures or mood destabilization. A structured taper over 2–3 weeks is the evidence-based approach.

The MONEAD pharmacokinetic analysis identified that approximately 9% of pregnant women on lamotrigine experienced little to no change in clearance during pregnancy. These women did not require significant dose increases during pregnancy and therefore do not need a postpartum taper. Therapeutic drug monitoring (TDM) during pregnancy identifies this subpopulation. If a patient's lamotrigine dose was not increased during pregnancy, the postpartum dose reset is unnecessary and could precipitate subtherapeutic levels.

Among all mood stabilizers used during breastfeeding, lamotrigine has the most extensive published safety data. The Uguz & Sharma systematic review identified data on 122 women, with adverse events reported in only 4/122 infants (3.27%) — and none of the reported rashes were attributed to lamotrigine. The Alvarez et al. (2025) study added 47 more infants with a similarly reassuring profile. The Kacirova 2022 colostrum cohort added 158 mother-infant pairs. The MONEAD study provided the largest pharmacokinetic dataset with paired maternal-infant samples. This evidence base is substantially larger than that available for lithium, valproate, or any atypical antipsychotic during breastfeeding.

The NEAD study found that children breastfed by mothers taking antiepileptic drugs (including lamotrigine) had higher IQ scores at age 6 than those who were not breastfed. The MoBa study found that breastfeeding was associated with less impaired development at 6 and 18 months. While these are observational findings subject to confounding, they provide strong reassurance that the neurodevelopmental trajectory of lamotrigine-exposed breastfed infants is favorable — and that withholding breastfeeding to avoid lamotrigine exposure may deprive the infant of a net developmental benefit.

The Kacirova et al. (2022) colostrum study of 158 mothers / 143 newborns found that median newborn serum lamotrigine concentrations on postnatal days 2–5 were substantially higher than infant levels typically measured in mature-milk studies (which more often fall in the 18–30% range of maternal levels). The explanation parallels the valproate chapter: these early values reflect residual transplacental exposure, not breast milk transfer.

The newborn serum/milk concentration ratio in Kacirova 2022 approached 1.0 in colostrum samples — meaning the newborn's serum level equaled the milk level. This is consistent with passive equilibration of drug already present in the neonate from in utero exposure rather than active accumulation from breastfeeding. Clinicians should not attribute high early neonatal lamotrigine levels to breastfeeding.

The Alvarez et al. (2025) study concluded that infants whose mothers had a lamotrigine dose of ≤150 mg/day had undetectable plasma concentrations and no symptoms — and "might be offered a simplified follow-up." For these low-dose mothers, the monitoring protocol can be streamlined: daily skin inspection, weekly weight checks for the first month, and clinical assessment as needed — without mandatory infant serum drug levels. For mothers on doses >150 mg/day, the full monitoring protocol (including baseline and 4-week infant serum levels) remains appropriate.

Estrogen upregulates UGT1A4 and increases lamotrigine clearance. The MONEAD study confirmed that use of estrogen-based medication increased nonpregnant lamotrigine clearance further. If a postpartum woman on lamotrigine initiates combined oral contraceptives, her lamotrigine levels will drop — potentially causing breakthrough seizures or mood destabilization. Conversely, if she stops estrogen-containing contraception, lamotrigine levels will rise.

Progestin-only contraceptives (progestin-only pill, LNG-IUD, implant, DMPA) do not affect lamotrigine metabolism and are preferred in this population. This interaction must be documented and monitored, with TDM if estrogen-containing contraception is initiated.

For lamotrigine, the RID is highly variable (2–21%) and is less clinically useful than direct TDM. The den Besten-Bertholee et al. (2025) case series concluded that "higher plasma concentrations generally lead to higher breast milk concentrations, so aiming for the lowest possible maternal plasma levels is beneficial. These findings underscore the importance of TDM in this population."

The Sabers algorithm provides a practical framework: establish a pre-pregnancy reference concentration, monitor throughout pregnancy, and use the same reference to guide the postpartum taper. The target is to return to the pre-pregnancy reference concentration — which simultaneously optimizes maternal efficacy and minimizes infant exposure.