Background

Between 2010 and 2017, opioid use during pregnancy increased a staggering 131%, with rates of neonatal opioid withdrawal syndrome (NOWS) increasing by 82%.1 Short-term symptoms of NOWS include irritability, agitation, fever, tremors, problems with feeding and sleeping, increased muscle tone and, in severe cases, respiratory problems and seizures.2 Interestingly, emerging preclinical evidence has posited potential neuroinflammatory effects as a result of in utero opioid exposure, indicative of molecular consequences.3 Although the aforementioned phenotypic symptoms of NOWS are well defined, molecular implications are not well understood. Even less understood are the long-term effects of in utero opioid exposure or postpartum opioid treatment for subsequent NOWS symptoms. Although non-pharmacological treatment methods are the first line of treatment, roughly 50–80% of infants born with NOWS need pharmacological intervention.4 5 Complicating matters further, there are no official guidelines for the treatment of NOWS and, with very few Food and Drug Administration approved drugs for neonates, many pharmacological treatments entail off-label use.6 Drugs commonly used for NOWS treatment include low doses of morphine, methadone or buprenorphine, weaned over time, and in some cases, adjunctive therapy with clonidine or phenobarbital is also needed for symptom management.6 7 With preclinical evidence pointing to inflammation in the brain and subsequent cognitive and memory deficits as a result of in utero opioid exposure,3 8–10 as well as clinical observations of intellectual impairment and increased attention disorders in children previously diagnosed with NOWS,11 it seems counterintuitive to treat NOWS infants with more opioids. However, it should be noted that such observations in clinical populations are inconclusive due to limited methodologies and sample size.7 11 12 Moreover, opioid treatment for NOWS is of particular interest given that the studies evaluating the pharmacokinetics and pharmacodynamics of such drugs in neonates have used very small samples, employed simulation methods to generate recommended dosing strategies, reported high intra-individual variability and/or may not be generalisable due to exclusion criteria for study enrolment.13–16 Moreover, long-term developmental effects of such treatments are unknown. However, as stated by Liu and colleagues,17 providers are ‘stuck between the devil and the deep blue sea… as there are no real alternatives to pharmacological treatment with opiates and other drugs for neonatal opiate withdrawal…’ (p.3). There is a pressing need to summarise current non-pharmacological and pharmacological interventions for NOWS, as well as potential non-opioid pharmacological interventions that may warrant future consideration and research for NOWS treatment.

Long-term outcomes of NOWS

Long-term effects of early life opioid exposure, including in utero opioid exposure and subsequent opioid treatment for NOWS, are largely unknown. Preclinical studies consistently show an association between in utero opioid exposure and cognitive effects including memory deficits18–20 and increased learning and memory errors.3 21–23 Clinical findings, on the other hand, have been inconsistent. For example, Kaltenbach and Finnegan24 did not find cognitive developmental differences in 4-year-olds who were or were not exposed to methadone in utero. Additionally, recent work using the Bayley Scale of Infant Development—3rd edition (BSID-III)25 also did not find developmental differences in infants exposed to opioids in utero when compared with unexposed, healthy controls.26 Taken together, these studies suggest environmental effects, secondary to NOWS, as the culprit for any observed cognitive deficits. Indeed, children born to mothers dependent on any substance are often exposed to other detrimental conditions including poverty, lack of resources and unstable home environments.27

However, as evidenced by Hunt et al’s28 and Maguire et al’s11 reviews, much of the literature has reported observations of neurodevelopmental deficits in children previously diagnosed with NOWS, including cognitive and social functioning deficits, as well as intellectual impairments. In a longitudinal study by Nygaard and colleagues,29 children previously diagnosed with NOWS had lower cognitive scores when compared with healthy controls from age 1 to 8.5 years. Marked sex differences were also noted as NOWS males had significantly lower cognitive scores at all time points when compared with healthy controls, whereas NOWS females only demonstrated significantly lower cognitive scores when compared with healthy controls at the last assessment (8.5 years).29 Additionally, recent work by Merhar et al30 found significantly lower cognitive, language and motor scores as measured by the BSID-III25 in infants previously diagnosed with NOWS when compared with normative BSID-III data. Similar results have been observed in older children previously diagnosed with NOWS. In high school students previously diagnosed with NOWS, literacy and numeracy scores were significantly lower when compared with non-NOWS matched controls, as well as when compared with those in the general population.27 Another study by Ornoy et al31 showed that children born to heroin-dependent mothers but adopted early in life had significantly improved intellectual scores on the Wechsler Intelligence Scales for Children-Revised (WISC-R)32 when compared with children born to and raised by heroin-dependent mothers.31 Interestingly, the intellectual scores of the same adopted children were lower than those of the healthy control children; however, with the exception of one subscale of cognitive function (WISC-R performance),32 the difference in scores was insignificant.31 These results suggest that environmental conditions may certainly play a role in developmental outcomes; but other factors, such as molecular implications that may or may not be driven by in utero opioid exposure and subsequent NOWS may also explain such developmental outcomes. While these studies echo the results of several other studies,11 28 results must be interpreted with caution due to limited methodologies and small sample sizes.

Long-term development of children previously diagnosed with NOWS who also received pharmacological intervention for NOWS symptom management has also been evaluated. For instance, using the BSID-III,25 Bakhireva and colleagues evaluated the development of infants aged 5–8 months (N=42) previously treated (n=16) or not treated (n=26) pharmacologically for NOWS and found no difference between the groups.26 Additionally, Merhar et al found that type of pharmacological treatment for NOWS infants (eg, methadone, buprenorphine or morphine) did not influence BSID-III25 development scores in children aged 18–28 months; however, sample sizes for buprenorphine and morphine were low (n=11 and n=14, respectively).30 Moreover, Benninger et al found that neurodevelopmental scores of infants previously diagnosed with and pharmacologically treated for NOWS were not different from normative data at 3–4 and 9–12 months; however, cognitive and language scores were significantly different from normative data at 15–18 months.33 More recently, Benninger et al found that infants previously treated pharmacologically for NOWS had significantly reduced cognitive, language and motor scores at 1 year of age when compared with population means.34 This may be of particular importance given the evidence that opioids, including those used in the postpartum period to manage NOWS symptoms, can adversely affect the developing central nervous system and may work to explain some of the long-term cognitive outcomes seen in both preclinical and clinical models.

NOWS treatment methods

The current treatment for NOWS includes both pharmacological and non-pharmacological interventions. First-line, non-pharmacological interventions that have been shown to reduce length of hospital stay and/or the use of pharmacotherapy include increased kangaroo care or skin-to-skin contact, breast feeding,35 mom’s ability to room with baby36 and decreased environmental stimulation.37 More recent studies have also started to evaluate acupuncture as a non-pharmacological treatment method for NOWS, with promising results.38 Despite these non-pharmacological interventions, 50–80% of the infants with NOWS will need pharmacological treatment for symptom management.4 5 Risk factors for NOWS severity and identifying which patients will require pharmacological treatment are not well understood but some studies have suggested neonatal genetic differences in drug-metabolising enzymes39 or mu-opioid receptors,40 41 as well as sex,42 maternal polysubstance use or maternal methadone dose.27 Pharmacological interventions primarily include opioids such as low doses of methadone, buprenorphine or morphine weaned over time, and, in extreme cases, adjunctive clonidine or phenobarbital.6 37 Interestingly, meta-analyses have shown low efficacy of morphine for NOWS when compared with buprenorphine,6 particularly as it relates to length of hospital stay. However, recent work has reported that 86% of the infants with NOWS receive morphine treatment, while 13% receive methadone and less than 1% receive buprenorphine.43 44

Nonetheless, opioids as a primary pharmacological treatment for NOWS symptom management requiring such intervention is of particular interest given the lack of knowledge regarding molecular consequences that may or may not be driving long-term outcomes. While such associations have not been made in humans, inflammation as a result of in utero opioid exposure has been associated with later cognitive and memory impairments in rat models. Therefore, it is possible that inflammatory effects may explain increased clinical implications of attention disorders and intellectual and cognitive functioning deficits observed in human children previously diagnosed with NOWS. Interestingly, other studies have suggested that an increase in exogenous opioids may affect the endogenous opioid system, which, during neurodevelopment, is responsible for regulation of brain growth. Thus, it has been postulated that an overactive endogenous opioid system, as a result of increased exogenous opioids, may restrict brain growth.45–47 Indeed, after controlling for size, term infants exposed to opioids in utero have been found to have smaller head circumference and brain volume when compared with their peers who were not exposed to opioids in utero.45 47 While it is unclear if these effects impact future cognitive development or other psychological conditions,45 continuing to expose this particular patient population to opioids in the postpartum period for NOWS symptom management seems questionable. Unfortunately, there are no current pharmacological alternatives to opioids for the treatment of NOWS, and there is a pressing need to better understand potential non-opioid pharmacologics for NOWS treatment. The purpose of this scoping review is to discuss the current non-pharmacological and pharmacological interventions for NOWS, as well as potential, novel non-opioid pharmacological treatments.

Study objectives

The objectives for this scoping review include: (1) summarize current treatment practices for NOWS including non-pharmacological and pharmacological interventions; (2) discuss potential novel, non-opioid pharmacological interventions for NOWS; and (3) establish the breadth of relevant preclinical and clinical scientific literature, including a table with study characteristics (eg, study aims; subjects; sample size; results) and discussion regarding study limitations. This will be the first study to review NOWS treatment and discuss novel pharmacological treatment options that warrant further consideration and study. Previously, neonatal abstinence syndrome (NAS) was used interchangeably with NOWS; however, in 2016, it was recommended that NOWS be the primary nomenclature for infants exposed to opioids in utero.48 Thus, both terms will be searched for exclusively in preclinical and clinical studies in the context of non-pharmacological interventions and pharmacological treatment.