Dear NICU-colleagues

I would like your expertise in a patient case. We have a preterm infant born at 28+2 weeks of gestation following PPROM and chorioamnionitis, treated with SALSA for RDS, and subsequently managed on HFNC until a week ago. The infant is now at 34+0 weeks postmenstrual age and is off all respiratory support. Aside from this, the infant is very healthy, with no IVH, no PDA, and adequate growth. Enteral nutrition was initiated with donor breast milk, and the infant reached full enteral feeds of 150 mL/kg by 7 days of age.

The infant’s mother has complex epilepsy and was maintained on Tegretol (carbamazepine) 800 mg + 1200 mg/day, Briviact (brivaracetam) 200 mg + 200 mg/day, and mirtazapine 45 mg once daily throughout the entire pregnancy. She began pumping early on and expressed a strong desire to feed the baby with her own milk. Initially we were hesitant, however, given her strong wishes and our goal to promote the use of mother's own milk, we have permitted partial feeding with MOM 12 mL out of 50 mL per feed, 8 times a day. The infant currently weighs 2340 g.

We now need to decide whether to allow her to transition to full breastfeeding. While we do use Briviact to treat some of our pediatric patients, this scenario presents a complex preterm polypharmacy issue, and our experience with Briviact safety during breastfeeding is limited. The straightforward approach would be to transition directly to formula, but the mother has worked incredibly hard for this, and we want to support her wishes if it is safe to do so. I read @Dotan S framework on polypharmacy (https://www.nature.com/articles/s41390-025-04416-z) and I know that @Mariana Oliveira usually has great insights on this topic.

What would be an acceptable approach according to you? Allt toughts are welcome to help us make the most reasonable decision.

Best Gustaf

Dear @Gustaf Lernfelt

Thank you for sharing this complex and thoughtful case. When evaluating maternal medications and breastfeeding safety, I always start by consulting with a clinical pharmacist to carefully analyze the pharmacokinetics of each drug involved — particularly relative infant dose (RID), protein binding, molecular weight, oral bioavailability, and half-life.

For your specific polypharmacy situation:

Carbamazepine is generally considered compatible with breastfeeding, though monitoring for sedation and feeding difficulties is recommended.

Brivaracetam has limited lactation data, but its pharmacokinetic profile suggests meaningful transfer into milk — this is the drug I would be most cautious about.

Mirtazapine data are relatively reassuring, though again, infant sedation monitoring is warranted.

Rather than reducing milk volume as the primary strategy, our approach is to time feeds strategically — avoiding breastfeeding at peak maternal plasma concentrations (typically 1–3 hours post-dose) and encouraging feeds just before the next dose when drug levels are at their trough. Given the mother’s divided dosing schedule, this may actually be quite feasible to implement.

Since the infant is already tolerating 12 mL MOM per feed without apparent adverse effects, this is reassuring data you already have in hand. A gradual stepwise increase in MOM proportion, with close monitoring for sedation, poor feeding, or neurological signs, seems reasonable rather than an abrupt switch to formula.

I would also strongly encourage involving the mother’s neurologist to explore whether brivaracetam could be substituted or the dose adjusted — though I recognize that complex epilepsy management cannot always be easily modified.

I look forward to hearing @Dotan S perspective on the polypharmacy framework, as these cases are indeed among the most challenging we face.

Hi Gustaf,
This is indeed a complex clinical scenario that should be approached from several angles - mother, infant, drug and feeding. This approach was discussed only a few days ago in a webinar hosted by the Irish Medicines in Pregnancy Service (IMPS) (available for watching here, highly recommended). Taking these four pillars into consideration will help optimize care for both mother and infant.

First, the drugs in question - let's summarize what we know about the metabolism of these drugs:
• Carbamazepine is metabolized in the liver, mainly by the CYP3A4 enzyme, yielding the pharmacologically active metabolite carbamazepine-10,11-epoxide (as well as several other non-active metabolites). Carbamazepine also induces the activity of CYP3A4, CYP1A2 and CYP2C19 to varying degrees (primarily CYP3A4).
• Brivaracetam is primarily metabolized by a CYP-independent hydrolysis and secondarily by hydroxylation mediated by CYP2C19. It also inhibits the enzyme epoxide hydrolase, which hydrolyzes the carbamazepine-epoxide metabolite.
• Mirtazapine is metabolized via several hepatic enzymes: CYP2D6, CYP1A2 and CYP3A4. One of its metabolites is known to be pharmacologically active (N-desmethylmirtazapine).

** Please see the attached diagram for better clarity (hopefully...) **

Assuming steady state for all drugs, from this metabolism overview we can appreciate that maternal levels of brivaracetam and mirtazapine are expected to be lower than reported in the absence of carbamazepine, which is reassuring for mirtazapine, whose milk levels were not predicted to be high to begin with. We can put it aside and not worry about it.
For brivaracetam, data on passage into breastmilk is limited, and the dose is very high (even supra-therapeutic per common guidelines), so milk levels would also be expected to be rather high, even if the overall AUC of the drug is somewhat decreased due to enzyme induction by carbamazepine.

So as an interim summary: from a PK point of view, we are not worried about mirtazapine or its pharmacologically active metabolite N-desmethylmirtazapine. While the levels of the latter are expected to be higher in this scenario due to induction of CYP3A4, its baseline passage into milk is still expected to be rather low (cf. Kristensen et al. (2007). Transfer of the antidepressant mirtazapine into breast milk. British journal of clinical pharmacology, 63(3), 322–327. https://doi.org/10.1111/j.1365-2125.2006.02773.x). However, we are not comfortable with brivaracetam because its levels may still be relatively high.

Now for carbamazepine:
The parent drug is considered compatible with breastfeeding, although both it and its active metabolite (carbamazepine-10,11-epoxide) are known to pass into breastmilk (cf. Kacirova et al. (2022). Carbamazepine and carbamazepine-epoxide concentrations in mothers, colostrum, and breastfed newborns: Comparison with concentrations determined during delivery and in the mature milk period. Biomedicine & pharmacotherapy = Biomedecine & pharmacotherapie, 151, 113176. https://doi.org/10.1016/j.biopha.2022.113176).

Considering the fact that the maternal dosage is rather high, this may translate into relatively high milk levels for both. Lactmed mentions that sedation, poor sucking, withdrawal reactions and 3 cases of hepatic dysfunction have been reported following exposure to carbamazepine in breastmilk.
This is complicated by the fact that brivaracetam inhibits epoxide hydrolase, effectively increasing the risk for higher levels of carbamazepine's active metabolite, which has anticonvulsant and neurotoxic effects. So we could anticipate a higher exposure to the active metabolite.

If we try to summarize pharmacokinetic interactions data, we may expect relatively high exposure to carbamazepine and brivaracetam based on high maternal doses, and to carbamazepine epoxide based on inhibition of its metabolism, alongside some unpredictable exposure to the active N-desmethylmirtazapine as mentioned above.
All of these exposures may translate into adverse CNS effects in the infant, i.e., lethargy, poor sucking, impaired weight gain and such. Additionally, neonatal hepatic function should be taken into account, which MIGHT be lower than normal due to prematurity - most CYPs function at a fraction of adult function (usually 10-30% in the first weeks of life, depending on gestational age and comorbidities, and then slowly increasing with overall development). We can assume that neonatal clearance may be somewhat lower than a term infant, which may contribute to the already high exposure we anticipated.

On the other hand, infant and maternal characteristics should be taken into consideration:
This is a premature infant that needs human milk for optimal overall development, normal development of the immune system and lowering risk for NEC. In addition, he is in the NICU, under constant monitoring. Furthermore, the mother's medical condition is controlled and she is willing to breastfeed. Under proper guidance, this would benefit both mother and infant, not only physically, but also emotionally and would consolidate bonding and wellbeing for both. Therefore, the benefits of human milk provision in this case may still outweigh the risks mentioned above, but it should be done prudently.

After integrating drugs, infant and maternal factors, I would suggest implementing mixed feeding, i.e., alternating between mother's own milk and donor milk (if available) at a ratio that seems reasonable for all involved ("start low, go slow"). @Mariana Oliveira rightly pointed out that you have already allowed him to be exposed to mother's own milk in small amounts with no adverse effects, so this is reassuring, but when talking about higher volumes, careful exposure management is still recommended. This may probably be the best course of action here until the infant's response to such potentially higher exposure would be more apparent. At any rate, mother should also avoid night feedings, as an epileptic postpartum patient must have regular, uninterrupted sleep to prevent seizures and to be able to function during the day. So the non-exclusive feeding pattern is essentially already there, and alternating between mother’s own milk and donor milk during the day may be optimal, at least until things get clearer. Then - if the infant shows no signs of adverse effects that may be attributed to the drugs - you could consider gradually increasing the volume of human milk over donor milk, hopefully up to exclusive or almost exclusive breastfeeding.

I would refer here to @Mariana Oliveira's suggestion to time feeding in accordance with drugs PK characteristics (half-life, Tmax, etc.). This is indeed a valid approach, which can aid in reducing exposure, e.g., not feeding within Tmax, which is the period when drug levels are building in serum and milk, or attempting to time feedings when drugs are at minimal levels (trough). The only caveat here is that these methods work best when feeding schedule is more flexible and controllable, i.e., in older infants. In the NICU, when infants feed every 2-3 hours, it may be difficult to time feedings in accordance with drug characteristics. But this is definitely worth considering, and even if it is not feasible now, it may be feasible in the future when the infant grows (also if something changes, i.e., a new drug is introduced).

The infant should be monitored as is customary in the NICU, also by the mother, who should learn to identify his habits, sleeping patterns, etc. Due to anticipated high exposure to carbamazepine, infant’s liver enzymes should be monitored regularly. Therapeutic drug monitoring for carbamazepine (which I assume is available in your facility) should also be considered for the infant, but this may be reserved for use only if other markers suggest high exposure, to avoid excessive blood draws, pain, etc.

Finally, while the wellbeing and seizure control for the mother are of the utmost importance, seeing how she is controlled and seizure-free, I agree with @Mariana Oliveira that you could consider involving her neurologist to discuss lowering her dosage(s) carefully, to reduce infant exposure through milk while keeping the mother well. This is delicate of course, but might be considered via a joint discourse with her doctors (primary physician, neurologist) AND the mother herself.

Please keep us posted, if possible. It would be interesting to know what transpires. If you happen to measure carbamazepine serum levels for mother and infant, we could even calculate relative infant dose (RID). Theoretically, this can also be estimated using published serum levels in, e.g., Lactmed, although serum levels under such doses as in this scenario are not available. Moreover, their warning regarding anticonvulsant medications should be heeded: "In published reports of anticonvulsant use during breastfeeding, most women were taking a combination of anticonvulsants. Some other anticonvulsants (e.g., phenytoin, carbamazepine) stimulate the metabolism of other drugs including anticonvulsants, whereas others (e.g., valproic acid) inhibit the metabolism of other drugs. Therefore, the relationship of the maternal dosage to the concentration in breastmilk can be quite variable, making calculation of the weight-adjusted percentage of maternal dosage less meaningful than for other drugs in this database." [Lactmed - Brivaracetam].

Good luck!

Hi @Mariana Oliveira , I'd like to just ask for a practical clarification. I've read this case, and now I'm wondering- this mother seems to be interested in a transition to direct breastfeeding. While pumping and bottle/tube feeding we can control amounts of MOM intake, however with direct breastfeeding, especially at home, parents have much less control of the amount of consumed milk.

What would be your practical advice for the mother while preparing them for a home discharge- aim to breastfeed before medication dose (as much as infant wants, without controlling the amount?), and offer formula if infant demands feeds at peak plasma concentration moments? Would you recommend that mother pumps-and-dumps in between the feeds to sustain the lactation (hoping when infant grows, the medication levels will become less problematic and she will be able to fully breastfeed until a desired timepoint, eg. 1 or 2 yrs of age?).

I'm just thinking of this case in terms of long term planning and supporting lactation beyond NICU and immediate post-discharge.