We identified a total of 105 published papers reporting maternal, pregnancy and/or neonatal outcomes after exposure to the alcohol pharmacotherapies of interest during pregnancy (Fig. 1). There was a total of 61% (n = 64) clinical studies and 39% (n = 41) animal studies. Studies were mostly performed in North America (32.4%), Australia (21.0%) and Europe (21.0%). The medication most commonly researched in the included articles was topiramate (Table 1).

Fig. 1

Preferred Reporting Items for Systematic reviews and Meta-Analyses (PRISMA) flow chart for the selection of studies included in the review

Table 1 Characteristics of included studies.3.1 Naltrexone3.1.1 Clinical

A small amount of clinical research (n = 8) has examined the safety of naltrexone in pregnancy; however, all studies investigate the treatment of pregnant women with opioid use disorders [19,20,21,22,23,24,25,26] and not AUD. The largest of the studies prospectively evaluated the maternal and foetal effects of naltrexone as a treatment for opioid use disorder in a cohort of 121 pregnant patients [25]. When compared with neonates exposed to methadone or buprenorphine, naltrexone exposed neonates had a significantly lower rate of neonatal abstinence syndrome (NAS) and associated hospitalisation, with all other outcomes largely comparable. A smaller study was in agreement, with no NAS diagnoses among naltrexone-exposed infants, and a shorter length of stay compared with buprenorphine-exposed infants [26]. Another study of 68 mother–child dyads found that females who conceived while on naltrexone treatment had significantly higher birth rates compared to non-opioid dependent controls [22]. Additionally, females exposed to naltrexone during pregnancy had higher rates of ectopic pregnancy and pregnancy and labour complications [22]. However, the authors suggest this increased occurrence of ectopic pregnancies reflects the higher birth rate in the naltrexone treated group. Additionally, naltrexone exposed neonates were smaller (birth weight and body length), had longer hospital stays and had higher rates of urogenital birth defects [23], while in childhood they had elevated rates of hospital and emergency department presentations, and middle ear infections (otitis media) compared with non-exposed children [24]. Lastly, two case series with a combined 18 cases of implant naltrexone exposure during pregnancy found normal obstetric and birth outcomes [19, 21], as did a brief communication reporting 17 pregnancies with implant naltrexone treatment [20]. It should be noted that some of these cases also appear in the population of Kelty et al.’s (2017) retrospective cohort study [23]. It is important to bear in mind some of these differences in results may be due to other factors associated with illicit opioid use. Patients who are opioid dependent and commenced treatment during pregnancy likely underwent opioid withdrawal while pregnant. While the impact opioid withdrawal has on pregnancy is not well understood, it may impact maternal and foetal health outcomes [27].

3.1.2 Pre-clinical

The 13 animal studies identified which investigate the health effects of naltrexone use during pregnancy have yielded inconsistent results [28,29,30,31,32,33,34,35,36,37,38,39,40]. Importantly, these animal studies almost exclusively use models of opioid use disorders (1 out of 13) rather than AUD. In these pre-clinical trials, the effects of naltrexone on maternal and neonatal outcomes appear to depend on the dose and frequency of administration. Rat models of high human equivalent daily doses [50 mg/kg, intraperitoneal (i.p.) injections] of maternal naltrexone throughout gestation resulted in elevated birth weight by 8–14% [34,35,36] and crown-rump length [35], with no effects on gestation length or litter size [34, 35]. Youngentob et al. (2012) [39] also found 15% elevated bodyweight of prenatal naltrexone exposed offspring, weighed between postnatal day 12 and postnatal day 14. In addition, relative organ weights (heart, liver, and skeletal muscle) at postnatal day 21 were increased by gestational naltrexone exposure [34, 35]. Moreover, naltrexone-exposed offspring weighed 8.2–36% more than their control counterparts at postnatal day 21 [35, 36, 40] and 16.3–24.3% at postnatal day 30 [40]. In contrast, rat offspring born to mothers treated with a clinically relevant low-dose sustained-release naltrexone implant (25 mg) during gestation displayed reduced birth weight, increased litter size, and no alterations to brain morphometry in 8-week-old offspring [30].

Two studies have investigated the effects of naltrexone exposure during pregnancy on the timing of parturition [31, 32]. In a pig model, a single intravenous injection of low-dose naltrexone ( 1mg/kg) administered in late gestation did not delay the onset of parturition relative to saline controls [31]. However, Javadi-Payder et al. (2009) reported that a single i.p. injection of naltrexone in mice at gestation day 15 had a dose dependent response on the duration of gestation and preterm delivery. A dose of 5 mg/kg produced no changes in preterm delivery incidence; however, a dose of 10 mg/kg marginally decreased gestational days in comparison with saline control mice [32].

Research has revealed intriguing insights into the impact of prenatal naltrexone exposure on physiological, behavioural and reproductive development in rat models. Mclaughlin et al. (1997) found that pre-natal naltrexone exposed rats (50 mg/kg/day, maternal i.p. injection, high human equivalent dose) had accelerated physical and behavioural maturation compared to controls [36]. Similarly, Cohen et al. (1996) reported that daily ad libitum (40 mg/kg) pre-natal naltrexone given in drinking water accelerated masculine sexual behaviour, but suppressed female receptivity [29]. Furthermore, Cajú et al. (2011) demonstrated that in utero naltrexone exposure (50 mg/kg/day, i.p. injection) from mid to late gestation impacted male testicular development, resulting in a reduction in the total length of seminiferous tubules and Sertoli cell population, but having no effect on sperm production [28].

The impact of prenatal naltrexone exposure on locomotor activity in rats presents mixed results. Rats administered a daily dose of 50 mg/kg throughout gestation exhibited reduced motor activity at postnatal day 21 [36]. In contrast, rats subjected to a slightly lower dose (40 mg/kg) for a shorter duration (from gestation day 13 until parturition) showed no changes in motor development during early prenatal life [29]. Adding further complexity, locomotor activity increased in mice prenatally exposed to 10 mg/kg naltrexone [33, 37] daily from late gestation (implant) [33] and twice daily from gestational day (GD) 1 to post-partum day 21 (subcutaneous injection) [37]. The reasons for these discrepancies remain unclear, and further investigation is needed to determine whether differences in rodent species, naltrexone dosage, route of administration, or duration of exposure coinciding to different stages of development contribute to the varied outcomes.

Examining broader neuronal effects, one rat study showed that continuous naltrexone exposure via a minipump throughout gestation and the early pre-natal period increased brain cortical thickness and reduced neuronal packing density in the offspring [38]. However, this exposure had no effect on neuronal number, suggesting that naltrexone exposure during development enhances neuronal maturation specifically in the brain cortex. Another rat study examined naltrexone’s neuroprotective effects on olfactory development after prenatal alcohol exposure [39]. Naltrexone mitigated the heightened behavioural response to the odour of ethanol observed in prenatal alcohol-exposed offspring, suggesting a potential neuroprotective effect of prenatal naltrexone exposure on olfactory development [39].

3.2 Acamprosate3.2.1 Clinical

There has been very little investigation into the safety of acamprosate in pregnancy [41,42,43], with clinical research limited to a single retrospective cohort study, and one small case series [42, 44]. In the cohort study of 52 women and their exposed neonates, acamprosate use during pregnancy was not associated with poor maternal or neonatal health outcomes when compared with pregnant women with a recent history of problematic alcohol use and those from the general community [42]. The study revealed no stillbirths or neonatal deaths, and there were no differences in rates of low birth weight, pre-term birth, or congenital abnormalities between the groups [42]. The small case series by the Teratology Information Service in France examined 18 pregnancies involving first trimester acamprosate exposure [44]. These 18 pregnancies resulted in two spontaneous abortions, three elective abortions, one therapeutic abortion (major malformations), ten unremarkable neonates (one premature, one died of sudden infant death syndrome) and two neonates with birth anomalies (one with minor facial anomalies and one with cleft lip) [44]. However, concurrent exposure to alcohol and other drugs occurred in several cases, confounding results.

3.2.2 Pre-clinical

In a mouse model of AUD from our group, exposure to acamprosate at approximate therapeutic doses (1.6 g/L in drinking water, ad libitum) from prior to pregnancy until postpartum day 7 demonstrated no adverse effects on maternal and birth outcomes, neonatal outcomes, or neurodevelopmental markers [43]. There were also no impacts on offspring motor control, locomotion or anxiety; however, the effects on short-term memory remained uncertain. Notably, the study provided preliminary evidence to suggest that acamprosate displayed neuroprotective effects against damage caused by in utero alcohol exposure. This finding is supported by another non-clinical study whereby hamsters prenatally exposed to acamprosate (1.33 g/L or 6.0 g/L, ad libitum; GD 5/8 to postnatal day 6) and/or ethanol demonstrated that acamprosate prevented the neuronal cell damage and death (measured via lesion size) caused by alcohol exposure [41].

3.3 Disulfiram3.3.1 Clinical

There were no clinical studies reporting disulfiram exposure during pregnancy within our search criteria. This lack of recent research is likely attributable to research in the late 1970s and 80s in which the use of disulfiram in pregnancy was associated with an increased risk of poor neonatal health outcomes including congenital anomalies [45,46,47]. In these reports, as summarised by Briggs et al. (2017), there were 38 foetuses (36 pregnancies) prenatally exposed to disulfiram of which 11 (28.9%) had congenital anomalies. Additionally, there were 6 elective terminations, 1 spontaneous abortion, 1 stillbirth, 14 unremarkable newborns and 5 lost to follow-up [44]. These outcomes must be interpreted with caution due to concurrent exposure to alcohol occurring in many of these pregnancies. Briggs et al. (2017) identified a further 25 neonates with in utero disulfiram exposure (via personal communication) whereby a major birth abnormality occurred in one neonate (4.0%) [44].

3.3.2 Pre-clinical

The three pre-clinical studies on disulfiram exposure in pregnancy included in this review present conflicting findings [48,49,50]. In guinea pigs administered oral disulfiram tablets (125 mg/kg) for 3 days late in pregnancy, there was no impact on litter size, incidence of spontaneous abortion or mean body weight of offspring. However, female offspring exposed to disulfiram exhibited smaller brain weights compared with controls, but not in male offspring [48]. In pregnant mice, a single dose (75 mg/kg, i.p. injection) of disulfiram on GD9 resulted in offspring with limb deformities including forelimb defects and postaxial ectrodactyly [49]. In contrast, some benefits were noted in mice exposed to disulfiram for the first three days of pregnancy at a smaller dose (50 mg/kg) via oral gavage [50]. This exposure increased the degree of vascularisation and volume of uterine horns, potentially providing a better environment for embryo implantation. Additionally, there were no differences in litter size and birth weight compared with control offspring.

3.4 Baclofen3.4.1 Clinical

Information on the safety of baclofen in pregnancy is extremely limited, with no studies identified congruent with our inclusion criteria. However, a single case series of 3 women and 4 pregnancies exposed to an intrathecal baclofen pump was identified. These cases reported no maternal adverse effects except for one instance of pre-eclampsia [51]. There was some variation in neonatal outcomes, with one infant small for gestational age, one large for gestational age. The remaining two infants were appropriate size for gestational age; however, they were both born premature and admitted to the special care nursery. No infants were reported to have evidence of teratogenicity or neurologic complications at birth, however two of the infants were diagnosed with jaundice.

3.4.2 Pre-clinical

Three pre-clinical studies investigated baclofen exposure in pregnancy, one focussing on maternal behaviour outcomes [52], and the other two on foetal effects [53, 54]. In the maternal behaviour model, rats injected intracerebrally with a single dose of baclofen (200 ng) late in pregnancy displayed reduced maternal care behaviours as demonstrated by lower quality nest building, longer latency for pup licking, and reduced pup retrieving compared to saline controls [52]. In the two studies examining rat foetal outcomes, both found that in utero baclofen exposure (30 mg/kg via i.p. injection) altered neural tube formation, widening the vertebral arch of the foetuses [53, 54]. These findings indicate a risk of spina bifida and other neural tube defects in neonates exposed to baclofen in utero.

3.5 Nalmefene

No human or animal studies investigating the safety of nalmefene in pregnancy appeared in our literature search.

3.6 Topiramate3.6.1 Clinical

The safety of topiramate use during pregnancy has been a subject of much investigation with 44 clinical studies conducted, but none with an AUD lens. A considerable focus of this research has been the identified elevated risk of congenital anomalies among children prenatally exposed to topiramate [55,56,57,58,