Methadone prescribed to replace illicit opiate use during pregnancy stabilises lifestyle and reduces the incidence of preterm birth but is commonly associated with neonatal abstinence syndrome (NAS). In addition, there are concerns regarding visual development in infants exposed to methadone and benzodiazepines in utero.1 2 In a recent study, 26% of infants born to mothers misusing drugs failed trained health visitor vision screening between 8 and 24 weeks of age, and 42% of these infants had confirmed ophthalmology problems.1 Visual evoked potentials (VEPs) can be used to assess the integrity and maturity of the visual pathway and are adversely affected by methadone use in adult humans and animal models.3 We report VEPs recorded in the first 2 weeks of life from infants born to mothers misusing drugs and prescribed substitute methadone in pregnancy.

METHODS

This was a prospective case–control study. Eligible infants were born at term (⩾37 weeks’ gestation) to mothers misusing drugs and prescribed methadone during pregnancy. Exclusion criteria were ocular abnormality, major congenital abnormality and significant neonatal illness. Infants were recruited from the postnatal ward of the Women’s Reproductive Health Service in the Princess Royal Maternity Hospital, Glasgow between September 2006 and February 2007. Control subjects were healthy term infants born in the study hospital. The study was approved by the Glasgow Royal Infirmary research ethics committee and written informed parental consent was obtained for all infants.

The multidisciplinary care package offered to women misusing drugs included a postpartum hospital stay of up to 10 days to provide parenting support and to monitor for signs of NAS. Infants were regularly assessed for significant NAS according to the local protocol and, when required, pharmacological treatment was commenced with oral morphine solution as first line therapy. In utero drug exposure was determined from maternal history and infant urine toxicology. Infant bag urine samples were obtained after consent and before administration of any medication to the baby. Samples were stored at −20°C until they were analysed in a single batch on an Abbott Architect c8200 analyser (Abbott, Abbott Park, IL, USA) using Abbott Multigent reagents (EMIT immunoassays). Assays included opiates, methadone, benzodiazepines, amphetamines, cannabinoids and cocaine metabolites.

The first VEP recording was made within 4 days of birth; a second recording was undertaken after 1 week if the infant remained in hospital. VEPs were recorded using three silver-silver chloride electrodes in the midline occipital, midline frontal (reference) and mastoid (ground) positions. Scalp-electrode impedance was below 5 kΩ. A hand-held integrating sphere (Colorburst; Diagnosys LLC, Lowell, MA) was presented 10 cm from the infant’s eyes in the midline and delivered bright white flashes (50 cd s/m2) at 1 Hz. Two averaged VEPs with a minimum of 30 flashes each were collected to ensure reproducibility. Awake/sleep state and degree of eye opening were documented. The flash VEPs were stored and subsequently assessed by two independent observers (DLMC and HM), blinded to the infant’s clinical course. VEPs were classified as typical (predominant positivity near 200 ms, P2), atypical (more complex response with unusual peak latencies), immature (predominant negativity near 300 ms, N3) or non-detectable (fig 1). The largest peak to trough amplitude was measured for all detectable VEPs and peak latencies were measured for P2 and N3 components.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1 The four different VEP waveforms. Classifications are (A) typical, (B) atypical, (C) immature and (D) non-detectable.

RESULTS

Twenty one infants of mothers misusing drugs and prescribed methadone and 20 control infants were enrolled. Age at first VEP recording did not differ between the groups (median 28 h and 31.5 h in the drug-exposed and control groups, respectively). Birth weights and occipito-frontal circumferences were smaller in the drug-exposed group (2818 g vs 3486 g; p<0.001 and 32.9 cm vs 34.9 cm; p<0.001, respectively) and the drug-exposed infants were of slightly earlier gestation (38.6 (SD 1.4) vs 39.8 (SD 0.95) weeks; p = 0.002). VEPs were repeated after 1 week in 14 of the drug-exposed infants, seven of whom developed NAS requiring pharmacological treatment. Early hospital discharge of control infants and some drug-exposed infants made repeat VEP testing impractical in this pilot study. Interpretable data were obtained in 54/55 VEP sessions; in one session the infant was too unsettled to allow an interpretable recording, but the VEP was successfully recorded 4 days later. Eleven of the 13 infant urine samples obtained were positive for methadone; for both negative infant samples, maternal urine testing confirmed methadone use. Other substances detected in infant urine included benzodiazepines (8/13), cocaine (2/13) and cannabinoids (1/13).

At 1–4 days of age, drug-exposed infants had an abnormal distribution of VEP waveforms compared to controls (χ² = 12.0, p<0.01) with fewer typical VEPs, more immature waveforms and non-detectable VEPs in five cases (fig 2). These VEPs were also smaller in amplitude (median 10.6 μV, range 0–30) compared to controls (median 24.4 μV, range 8–69) (ANOVA F = 20.8, p<0.001). Peak latencies for P2 and N3 did not differ between the groups. Sleep and eyelid closure had no significant effect on the amplitudes of these bright flash VEPs, but P2 was prolonged in sleep compared with P2 of awake infants (215 (SD 28) ms vs 194 (SD 16) ms; p = 0.03). After 1 week, VEPs in the drug-exposed infants had an increased proportion of typical and fewer non-detectable VEPs (fig 2), but amplitudes remained low (median 11.3 μV, range 0–21). There were no significant differences in VEPs between infants who did or did not develop NAS.

• Download figure
• Open in new tab
• Download powerpoint

Figure 2 Distribution of VEP waveforms. White: control group (n = 20). Black: drug-exposed infants at a median age of 1 day (n = 21). Grey: drug-exposed infants at a median age of 8 days (n = 14). Data are expressed as a percentage of each group.

DISCUSSION

Recent studies have reported nystagmus and impaired visual development in infants born to mothers who used both methadone and benzodiazepines in pregnancy.1 2 We have demonstrated marked differences between flash VEPs recorded in the first 4 days of life from term infants born to mothers misusing drugs and those of control infants, with VEPs of drug-exposed infants being of relatively immature waveform and considerably lower amplitude. Moreover, VEPs were non-detectable in 24% of drug-exposed neonates but detectable in all control infants. After 1 week VEPs in drug-exposed infants had matured but were still more likely to be non-detectable or of lower amplitude than those of newborn controls. The infants of mothers misusing drugs in this study were smaller and marginally more immature at birth than their non drug-exposed counterparts, consistent with previous descriptions of this population4 and reflecting the confounding effects of social deprivation, smoking and maternal ill health upon fetal growth and potentially visuo-cortical development. The reduction in VEP amplitude in our drug-exposed cohort was, however, almost four times greater than that reported by Stanley et al5 in symmetrically growth restricted non drug-exposed infants at term corrected age; furthermore, all of this latter cohort had detectable VEPs of comparable waveform to appropriately grown control infants. The majority of our mothers used benzodiazepines in addition to prescribed methadone, but the study population was too small to evaluate the independent effects of these different drugs. Half the infants who had repeat VEPs measured after 1 week had commenced oral morphine solution; no differences in VEPs between treated and untreated infants were apparent, but numbers were small. In the absence of longer term follow-up, it is unclear whether changes in neonatal VEPs in these maternal drug-exposed infants are due to transient effects of circulating methadone and/or other substances of misuse, or whether they are a consequence of teratogen exposure, with long term implications for visual function.

To our knowledge, this is the first study of VEPs in neonates exposed to methadone and other substances of misuse in utero. Further study is required to clarify direct and/or indirect effects of maternal methadone and other illicit drug misuse on infant VEPs, and the relationship between neonatal VEPs and subsequent visual development.