Drug Testing in the Neonate

Steven W. Cotten, PhD, DABCC

INTRODUCTION

A major portion of toxicology testing deals with urine drug screening for the adultpopulation. This group can be divided into two major applications: preemploymenturine drug screening and periodic scheduled pain management screening. Painmanagement clinics usually require patients to sign opiate contracts that allow forregular testing as a means to assess compliance. The presence or absence of drugsand metabolites must match the patient’s prescribed medications, and anydiscrepant compounds found during routine screening are grounds for dismissalfrom the pain management program. In addition to these two applications,toxicology drug testing plays an important but often overlooked role in newborndrug screening. Testing this population comes with its own set of uniqueanalytical, therapeutic, and legal issues that can make screening and resultinterpretation challenging.

The author has nothing to disclose.Department of Pathology and Laboratory Medicine, University of North Carolina at Chapel Hill,101 Manning Drive, Chapel Hill, NC, 27514, USAE-mail address: scotten@unch.unc.edu

Clin Lab Med 32 (2012) 449–466http://dx.doi.org/10.1016/j.cll.2012.06.008 labmed.theclinics.com

KEYWORDS

• Neonate • Drugs of abuse • Meconium • Pregnancy • Newborn

KEY POINTS• Drug screening in the newborn population comes with a set of unique analytical,

therapeutic, and legal issues that can make testing and result interpretation challenging.

• Assessment of in utero drug exposure to cocaine, amphetamines, opiates, marijuana, andethanol may allow better intervention and management of withdrawal symptoms for theneonate.

• A range of maternal and neonatal specimens are available, but each comes with a uniqueset of limitations regarding sensitivity, invasiveness, and window of detection.

• Preanalytical issues such as specimen collection and sample extraction can influence testaccuracy, and the particular biological specimen evaluated determines the window ofdetection achieved.

• Meconium provides the longest window, but the extraction technique greatly impactssensitivity, and unique drug metabolites in meconium may lead to discrepancies betweenmaternal and neonate results.

0272-2712/12/$ – see front matter © 2012 Elsevier Inc. All rights reserved.

wlTa

450 Cotten

The rationale for newborn drug screening seeks to establish a picture of prenataldrug exposure during pregnancy. Conventional testing in adults uses both self-reporting and biological specimen testing to provide the necessary information forestablishment of compliance or abuse. In the case of newborns, self-reporting is notapplicable, and clinicians must rely on information from the mother coupled withbiological testing from both the mother and neonate. Additionally, interpretation ofdrug screening results may be left to physicians, nurses, or social services workers.As the complexity of testing and interpretation increases, it is imperative that thelaboratory be proactive in educating the necessary parties involved, particularly thosewithout extensive toxicology training, to ensure proper medical and legal decisionsare made with the information provided.

The unique analytical and legal caveats associated with newborn drug testing posea variety of challenges to laboratorians and clinicians alike when it comes to screeningthis specific population. Preanalytical issues such as specimen collection and sampleextraction can influence test accuracy. Samples can easily be adulterated because ofthe access of family members to the infant and extensive unsupervised time.Newborns frequently have dilute urine, so false-negatives are likely. Furthermore,sample volume is often low, limiting the ability for comprehensive screening andconfirmation testing. Therefore, negative urine results do not definitively rule out drugexposure.

Positive urine results cannot distinguish between intermittent and chronic use bythe mother. Often there is low agreement between specimens from the mother andnewborn as well as results from screening and confirmation. Instances of discordantresults complicates interpretation, particularly in cases with multiple specimens.Several therapeutic issues related to medical management are also unique to thispopulation. The mother may initially delay prenatal care to avoid urine drug testing.Subsequent toxicology results have limited predictive value on genetic changes that havetaken place earlier during the pregnancy. Positive drug screening results, however, canallow for proper medical management of withdrawal symptoms for certain drug classes.

Perhaps the most serious issues related to newborn drug testing are the legalimplications surrounding decisions made in the case of positive results. Positive urineor meconium drug samples in newborns trigger involvement from social services forassessment of child safety. Twelve states formally consider positive urine drug screenin an infant to be child abuse; therefore, laboratory results can potentially removenewborns from their biological parents and place them in the foster care system. Forthis reason, the caveats and limitations of drug testing in this population are of utmostimportance.

SUBSTANCE ABUSE DURING PREGNANCY

Accurate assessment of substance abuse during pregnancy is challenging forclinicians and other health care workers. Determination of abuse can come from eitherself-reporting or biological specimen testing. The US Department of Health andHuman Services in conjunction with the Substance Abuse and Mental HealthAdministration periodically conducts surveys on drug use in persons 12 years andolder in the general US population. In 2009, data estimated 22.6 million Americans(8.9%) used illicit drugs in the month before the survey.1 A total of 4.4% of pregnant

omen ages 15 to 44 admitted to substance abuse in the last month. This rate wasower than nonpregnant women (10.9%) and decreased with the age of respondent.he frequency of substance abuse was 16.2% for women ages 15 to 17, 7.4% for

ges 18 to 25, and 1.9% for ages 25 to 55. Alcohol use among pregnant women was

oactoti

451Neonatal Drug Testing

estimated at 10.8%, with 3.7% admitting to binge drinking and 1.9% admitting toheavy alcohol use.

When biological specimens from pregnant mothers or newborns are tested for thepresence of drugs, the rate of substance abuse detected increases. Using biologicalspecimens, the frequency of illicit drug use in pregnant women has been estimated at20% using maternal urine, maternal hair, newborn urine, or meconium.2–4 Tobaccoand alcohol use/abuse are also estimated around 19%.

CRITERIA FOR TESTING AND CONSENT

Clinical indication for substance abuse testing is highly variable between regions.Currently there are no federal guidelines defining criteria for the testing of newbornsor pregnant women. Therefore, it is up to individual institutions and health caresystems to draft explicit guidelines for newborn and maternal drug testing to bestidentify and manage substance abuse in their specific population. Evidence-basedclinical practice guidelines have been developed to provide recommendations forstandardized screening approaches.3,5,6 Typical institutional guidelines for toxicologyscreening for newborns may resemble the following. (Adapted from the newbornnursery guidelines from the University of North Carolina at Chapel Hill.7)

Urine or meconium drug testing is clinically indicated by the following:

● Maternal History– History of drug abuse.– Prenatal care starting after 16 weeks or less than a total of four prenatal visits.– History of child abuse, neglect, or court-ordered placement of children outside

the home.– History of domestic violence.– History of hepatitis, human immunodeficiency virus, syphilis, or prostitution.– Unexplained placental abruption.

● Infant History– Unexplained intrauterine growth restriction.– Infants with evidence of drug withdrawal (hypertonia, irritability, or tremulousness).

● Alcohol– Acute maternal alcohol intoxication is observed around the time of delivery.

Consent for collection is an additional issue requiring special attention for toxicol-gy testing in this population. Informed consent may be required for both maternalnd newborn specimens depending on institutional and state guidelines. Consent forollection of newborn urine or meconium for drug screening is often covered underhe general consent for treatment for most health care institutions. Departments ofbstetrics and gynecology within the hospital may seek separate verbal consent fromhe mother and inform the parents prior to testing if drug screening becomes clinicallyndicated.

LEGAL IMPLICATIONS

Most clinical laboratories do not routinely perform chain of custody for specimenssubmitted for toxicology analysis. This protocol differs from employment urine drugscreening where documentation of chain of custody is necessary for validity of theresults. Legal action can be taken only if chain of custody is properly documented. Inthe case of newborn drug screening, if a specimen is positive for an illicit substance,chain of custody is not required for legal action or involvement by social services. This

difference illustrates an important nuance in newborn drug testing in that both

452 Cotten

medical and legal decisions are made using the results generated by the clinicallaboratory. It is therefore imperative that all parties involved in the testing process—nurses, medical technologists, physicians, laboratory directors, substance abusecounselors, and social services workers—have a unified protocol for testing andunderstand the limitations.

A compendium of individual state laws regarding parental drug use and child abuseis available from the Child Welfare Information Gateway.8 Arkansas, Colorado,Florida, Illinois, Indiana, Minnesota, North Dakota, South Carolina, South Dakota,Texas, Virginia, Wisconsin, and the District of Columbia all consider a positivenewborn urine drug screen in their definition of child abuse or neglect (Fig. 1). Anadditional 13 states have specific reporting procedures for newborns that showevidence of exposure to drugs or alcohol. The Child Abuse Prevention and TreatmentAct requires states to develop formal policies for informing child protective servicesin cases where newborn drug exposure is documented and to develop plans forprotective care and medical management of withdrawal symptoms.9

TYPES OF SPECIMENS

The goal of toxicology drug testing in newborns is to evaluate in utero drug exposureduring the course of the pregnancy. A wide range of maternal and neonate biologicalspecimens are available to clinicians. Each specimen comes with its own unique setof limitations regarding sensitivity, invasiveness, and window of detection. Biologicalmatrices from the newborn include meconium, hair, cord blood, and neonate urine.Specimens from the mother include hair, blood, oral fluid, sweat, urine, and breastmilk. Matrices that contain drugs and metabolites from both the mother and neonateinclude the placenta and amniotic fluid. In order to develop a complete understandingof total drug exposure during pregnancy, clinicians must consider the type ofspecimen being tested, the type of drug expected, and the window of detection (Fig. 2).The rate of agreement between mother and infant can often be low because of thesecomplicating factors.

SPECIMENS FROM THE NEONATE

The three types of commonly used specimens from the newborn include neonate

Fig. 1. States that consider a positive newborn urine drug screen child abuse. Arkansas,Colorado, Florida, Illinois, Indiana, Minnesota, North Dakota, South Carolina, South Dakota,Texas, Virginia, Wisconsin, and the District of Columbia incorporate a positive urine drugscreen for a newborn in their definition of child abuse.

urine, neonate hair, and meconium. Neonate urine is the most frequently used

dbcmnh

df

�dMawmtfa

453Neonatal Drug Testing

specimen to assess in utero drug exposure. Despite its popularity, urine samplesprovide the shortest window of detection, only capturing drug use several days beforedelivery.10 Newborns also tend to have dilute urine, so false-negatives are likely.

Neonate hair forms during the last trimester, and therefore can capture drugexposure for the last 3 to 4 months during pregnancy.11 The parent compounds of

rugs typically accumulate in hair rather than the more polar metabolites found inlood, urine, and meconium through either passive diffusion from arterial bloodapillaries of the hair follicle or excretion on the surface of the head. Typically 20 to 50g of hair are needed for adequate testing, but because quantities may be limited in

ewborns and the procedure is considered partially invasive, adoption of hair testingas not gained routine usage.12

Meconium refers to the first fecal matter passed during the first days of life; it ischaracterized by a dark green/black color. Formation of meconium begins around the12th week of gestation and continues to accumulate until birth.13 Drugs of abuse are

eposited in meconium through absorbance across the placenta, metabolism by theetal liver, and swallowing of amniotic fluid.14,15 The length of accumulation imparts

the largest window of detection for drugs of abuse for any neonate specimen,theoretically capturing the entire profile of exposure over the last two trimesters.

Collection of meconium is considered noninvasive, with 99% of full-term infantspassing their entire meconium after 48 hours. Drug-exposed neonates show a slowerevacuation of meconium, with the median first stool passing at day 3 and 90% by day12.16 For analysis, a minimum of 0.5 grams is collected and stored at �20°C to

80°C prior to organic solvent extraction and measurement. The choice of solventepends on the compound measured and the subsequent method used for analysis.ethanol is the most common extraction solvent for cocaine, opiates, cannabinoids,

nd amphetamines when liquid chromatography–mass spectrometry (LCMS) is used,ith extraction efficiencies between 40% and 80% depending on the individualetabolite assayed. Extraction under acidic conditions may increase recovery

hrough hydrolysis of glucuronide metabolites at a low pH. The extraction procedureor recovery of drug metabolites from meconium samples has a much greater effect on

Fig. 2. Window of detection for biological specimens: The window of detection variesdepending on the sample chosen for drugs of abuse screening. (From Lozano J, García-AlgarO, Vall O, et al. Biological matrices for the evaluation of in utero exposure to drugs of abuse.Ther Drug Monit 2007;29:711–34, Figure 2; with permission.)

ssay sensitivity than the screening or confirmation procedure itself.

Fmet

m

phh

454 Cotten

SPECIMENS FROM THE MOTHER

Biological specimens available for toxicology testing from the mother include mater-nal hair, urine, blood, amniotic fluid, placenta, cord blood, and oral fluid. Of thesespecimens, maternal hair and urine are used most frequently during pregnancy andimmediately after delivery to assess drug use by the mother. Maternal blood onlycaptures acute consumption occurring a few days or hours prior to collection and hasfound limited value in testing for drugs of abuse. Oral fluid is noninvasive comparedwith blood, and drugs that are weak bases such as cocaine, opiates, benzodiaz-epines, and nicotine tend to accumulate in oral fluid relative to serum. The distribution,detection window, and analyte stability in oral fluid are still unknown for many drugsof abuse.17,18 Generally only parent compounds are the major component found inoral fluid, and the specimen only detects acute consumption hours before collection.

The placenta is the maternal organ that forms after the fourth week of pregnancyand acts as the interface between maternal and fetal blood, allowing for exchange ofnutrients and waste during development. Metabolism and degradation of xenobioticsby the placenta can also occur and impact drug transfer to fetus. Biomarkers ofethanol use by the mother, namely fatty acid ethyl esters (FAEEs), are not transferredto the fetus but are instead absorbed and degraded by the placenta.19 Therefore, any

AEEs detected in the neonate are a result of direct exposure to ethanol andetabolism by the fetus in utero. Despite its important role in drug metabolism and

xposure during pregnancy, placenta testing remains an esoteric specimen used inoxicology testing.

Amniotic fluid accumulates throughout pregnancy, having a volume of approxi-ately 30 mL at week 10 and up to 1000 mL by week 37.20 The highly aqueous nature

of the specimen selectively enriches for water-soluble drugs. Collection of amnioticfluid, however, can be dangerous for the fetus, and its use for drug testing is notroutine. Despite this obstacle, several studies have used amniotic fluid to confirmcocaine exposure, but its popularity as a biological fluid for prenatal drug useremains low.21–24

The two most popular biological specimens from the mother for assessing prenataldrug exposure are maternal hair and urine. Maternal hair captures chronic drug useover the longest period of time. Collection is noninvasive and provides a directestimate of maternal drug exposure throughout the entire pregnancy or before. Drugdeposition in the hair may, however, be affected by hair care products or cosmetichair treatments that damage the follicle or hair proteins.25 Maternal hair analysis

rovides only an indirect estimation of drug exposure for the fetus. Several studiesave attempted to correlate cocaine concentrations between maternal and neonateair with unpredictable results.26 Maternal urine in conjunction with neonate urine is

probably the most popular specimen to assess drug exposure several days beforedelivery. Despite its short window of detection, collection is easy, extraction isefficient, and sensitivities in the ng/mL can be achieved for most drugs tested usingLC or gas chromatography–mass spectrometry (GCMS).

AGREEMENT BETWEEN MATERNAL AND NEONATE SPECIMENS

Immunoassay screening and mass spectrometry (MS) confirmation results for bothmaternal and neonate specimens, along with maternal self-reporting, may showvarying degrees of agreement. Several factors contribute to these discrepancies.Maternal self-reporting is estimated to capture only 4% of illicit drug users from ageneral population. Urine specimens from both the mother and neonate have a

narrow window of detection for most drugs and are greatly affected by hydration

dstaap

tf(

455Neonatal Drug Testing

status. In the case of meconium, extraction efficiency and completeness of collectionmay negatively select for drug detection. Furthermore, if GCMS is used for confirma-tion, extraction, derivatization, and volatilization will not be equal for different drugclasses and their respectively glucuronidated metabolites. Comparing results fromdifferent matrices (maternal urine vs neonate meconium) often shows the presence ofcertain drugs in one specimen that are absent in another.

SPECIMEN COLLECTION AS A SOURCE OF PREANALYTICAL VARIABILITY

In addition to distinctive specimens used in newborn testing, collection methods forurine and meconium are unique. These newborn-specific factors may contribute topreanalytical variability that can affect assay results. The physical collection of urinespecimens from newborns is often overlooked as a source of preanalytical variability.A recent study illustrated a wide range of collection procedures in place for neonateurine collection at one institution.27 Nursing staff at this institution used specially

esigned urine bags, diapers turned inside-out, cotton balls, or gauze followed byyringe extraction from the textile matrix to collect the urine specimen prior to sendinghe sample to the laboratory. In some instances the infant was given a bath prior topplication of the collection device, whereas other times baby wipes were used. Asresult a variety of textiles including diapers, baby wipes, gauze, and lotions could

otentially come in contact with the sample prior to analysis.It is therefore important to validate the collection procedure to identify interferences

hat may generate false-positives or false-negatives. A false-positive interferencerom baby wash soaps has recently been reported for several tetrahydrocannabinolTHC) immunoassays for cannabinoids.27 The interferent was identified after anincrease in positive specimens was observed specifically from the institution’snewborn nursery. This result can be problematic because confirmatory testing is notmandated in the clinical setting. The interferent was found to be analyte- andmanufacturer-dependent. A variety of other household chemicals, namely liquid soapand Visine, have previously been reported as positive interferents in drugs of abuseassays in the context of adulterants for preemployment screening but not in thenewborn population.28,29

METHODS FOR SCREENING AND CONFIRMATION

Conventional urine drug toxicology consists first of a broad screening method thatdetects a class of related compounds followed by confirmation of positive results witha second method that detects a defined list of compounds within that class.Regardless of the matrix, urine drug screening almost exclusively uses immunoassay-based platforms that target cocaine, amphetamines, cannabinoids, opiates, orphencyclidine. Clinical laboratories confirm presumptive positive results using MScoupled with LC or GC that imparts increased sensitivity and specificity for quanti-tating a specific set of compounds. Given the medical and legal decisions thatpotentially impact patient care and the unique specimens associated with neonatedrug screening, it is recommended that all toxicology screening results for drugs ofabuse be confirmed.

SPECIFIC DRUGS OF ABUSECocaine

Illicit drug use including cocaine among pregnant women is estimated at 4% basedon national studies.1 After ingestion via smoking or insufflation, cocaine enters the

blood where it can be metabolized by the liver or rapidly cross the placenta via

bcmam

456 Cotten

passive diffusion. Several important metabolic pathways for cocaine are recognized(Fig. 3). The various metabolites have a diverse range of polarities and therefore arenot uniformly enriched in meconium.30,31 Cocaine, norcocaine, p-hydroxycocaine,

enzoylecgonine, ecgonine methyl ester, ecgonine, anhydroecgonine methyl ester,ocaethylene, and norcocaethylene are the major metabolites frequently found ineconium. Interestingly, ecgonine has been shown to occur at the highest frequency

nd median level for cocaine-exposure infants using solid phase extraction of theeconium.30 Evaluation of cocaine metabolism for either meta or para hydroxylation

showed para hydroxylation to occur at a higher frequency in samples. Median valuesfor meta hydroxylation (m-hydroxycocaine or m-hydroxybenzoylecgonine), however,exceeded median values for their corresponding para metabolites when present.30

Concomitant use of ethanol and cocaine generates the ethyl derivatives norcoca-ethylene, cocaethylene, and ecgonine ethyl ester via transesterification with ethyl

Fig. 3. Cocaine metabolism: The cocaine metabolites present in a sample are dependent onconcomitant ethanol use, route of administration, and the biological specimen evaluated.

alcohol instead of water. Two specific metabolites can be used to differentiate

h5

mflucpbc

Aut1tsc

457Neonatal Drug Testing

insufflated cocaine from smoked (crack) cocaine. The pyrolytic products anhy-droecgonine methyl ester and anhydroecgonine are both generated directly from theloss of benzoic acid by cocaine during volatilization. The frequency of detection forboth ethanol- and crack cocaine–associated metabolites is greater than 95% in themeconium of infants whose mother’s urinalysis was positive for benzoylecgonine.30

Conventional methods for both immunoassay and MS confirmation target benzo-ylecgonine, the major ionic metabolite of cocaine found in urine. Current immunoas-says effectively recognize both benzoylecgonine and m-hydroxybenzoylecgonine butshow little reactivity toward cocaine, ecgonine methyl ester, ecgonine, or cocaethyl-ene.32 This result may account for some discrepant values between neonate andmother dyads in regard to positive or negative result agreement.

Methanol extraction of meconium is frequently used for isolation of cocainemetabolites. This extraction can be coupled with solid phase extraction to furtherenrich additional compounds other than benzoylecgonine through reconstitutionin acidic buffer followed by additional washing and elution in methylene chloride/2-propanol-ammonium hydroxide (78/20/2).30 This inclusive extraction procedure

as reported recoveries of 15 unique cocaine metabolites between 38.9% and9.1%.Immunoassay-based screening methods including radioimmunoassay, enzyme-ultiplied immunoassay technique (EMIT), enzyme-linked immunosorbent assay, and

uorescence polarization have all been used for cocaine determination with neonaterine and meconium.12 GC and LCMS methodologies have been adopted foronfirmation testing of meconium samples allowing for quantitation of comprehensiveanels of cocaine metabolites. LCMS methods achieve greater sensitivity than GCMSecause of elimination of the derivatization and volatilization steps required by gashromatography.

Effects of in utero cocaine exposureA large epidemiologic study of 11,811 mother-infant pairs compared meconium andmaternal self-reporting to assess drug exposure.33 Of the analyzed samples, 9.5%were positive for cocaine or its metabolites using EMIT immunoassay followed byGCMS. At birth, neonate exposure to cocaine was significantly associated with lowbirth weight and altered length and head circumference from normal ranges.34

Additionally, prenatal cocaine exposure has been linked to delayed language devel-opment, decreased total language performance, behavior problems in school, andincreased risk of obesity through early adolescence.35,36

Cannabinoids

Cannabinoids are a class of compounds produced by Cannabis sativa and Cannabisindica with �9-THC being the most prevalent and the major psychoactive component.

ssessment of prenatal cannabis exposure relies on maternal self-reporting, maternalrine, neonate urine, and meconium. Metabolism of THC generates 11-nor-�9-etrahydrocannabinol-9-carboxylic acid (THC-COOH) through the active intermediate1-hydroxy-�9-tetrahydrocannabinol (11-OH-THC). The majority of immunoassaysarget THC-COOH because it is the major constituent in cannabinoid-positive urineamples.37 Meconium does, however, have a greater likelihood of detecting sporadicannabis use during pregnancy; it is often preferred over newborn urine (Fig. 4).

Differences in cannabinoid metabolite content exist between urine and meconiumin the neonate that can lead to high rates of false-negatives. Studies evaluatingmeconium samples that screen positive by EMIT assay but confirm negative for

THC-COOH revealed a 40% false-positive rate.38 Further analysis of meconium

gua

cedc

rbs

o

cct(T

458 Cotten

samples that screened negative for THC-COOH detected 11-OH-THC and 8�-11-dihydroxy-�9-tetrahydrocannabinol (8�-di-OH-THC) instead.38,39 These data sug-

est that meconium may selectively enrich distinct cannabinoid metabolites fromrine, and that confirmation rates can be increased by incorporation of thesedditional compounds (see Fig. 4).

A dose-response relationship between maternal drug use and neonate meconiumannabinoid concentration has not been established, making estimation of maternalpisodic use difficult. Serial collection of meconium samples over the first severalays has demonstrated a 60% likelihood of a subsequent positive result if the firstollected sample is positive.14 This issue highlights the heterogeneous nature of

meconium collection and how the dynamics of gastrointestinal motility can affect testresults.

Sample preparation for cannabinoid extraction from meconium can be achievedusing normal saline, methanol, acetic acid/diphenylamine, or hexane/ethyl ace-tate.40–43 Methanol extraction and subsequent analysis of THC-COOH can yieldecoveries between 50% and 72%.39 THC and its metabolites are glucuronidated inoth urine and meconium; therefore, enzymatic, acid or base hydrolysis improvesensitivities for GC or LCMS.

Interferences in THC-COOH immunoassay assays have been reported on a varietyf manufacturer platforms.27,29,44 Newborn-specific factors related to sample collec-

tion (urine) may also impart preanalytical variation for THC immunoassay results.27 Toomplicate immunoassay interpretation there are three prescription drugs thatontain cannabinoid-based molecules. Medicinal marijuana is available by prescrip-ion in one out of three states in the United States, the pain management drug MarinolTHC) is given to specific pain populations, and the new neurologic agent (Sativex,

Fig. 4. THC metabolism: In addition to the major urinary metabolite THC-COOH, meconiummay selectively accumulate 11-OH-THC and 8�-11-dihydroxy-�9-tetrahydrocannabinol as aresult of maternal THC use during pregnancy.

HC/cannabidiol) is used for the treatment of multiple sclerosis. Although there are no

plelceaw

ogmp

459Neonatal Drug Testing

recommendations for use of the medications during pregnancy, drug screeningresults should be critically evaluated if the mother has a medical history involving anyof these prescription drugs.

Effects of in utero cannabinoid exposureNumerous studies have sought to measure quantitatively the effects of maternalcannabis use on neonate development. Several studies have reported negativephysical effects on birth weight, length, and gestational age.45,46 Below-normal

erformance on intelligence tests, increased frequency of depression, and greaterikelihood to use cannabis in adolescence have also been associated with prenatalxposure.47-49 Yet other studies find no adverse effects at birth or demonstrable

ong-term differences between exposed and nonexposed neonates.50–52 Such dis-repant findings have been suggested to arise from inability to properly stratifyxposed from nonexposed infants.53 Despite a lack of conclusive evidence for orgainst harmful effects of prenatal THC exposure, substance abuse often correlatesith other environmental and social factors that can negatively affect development.

Opiates

The frequency and amount of opiate consumption has increased dramatically in thepast decade. Drug screening has become increasingly complex as the need todiscriminate between over-the-counter, prescription, and illicit opiates grows in theclinical setting. The rapid emergence of pain management clinics now requiresspecific quantitation of multiple prescription opiates to assess patient compliance. Inthe setting of pregnancy, opiate testing is important for both maternal and neonatehealth, particularly for opiate-dependent mothers and the proper management ofneonate abstinence syndrome at birth.

The type of opiate molecule in question will determine its immunoassay reactivity,transport across the placenta, and tissue distribution in the neonate. Morphine-basedopiates such as morphine, codeine, oxycodone, hydrocodone, buprenorphine, andheroin are lipophilic and will be transferred readily across the placenta to the fetus.Their corresponding glucuronidated metabolites are more hydrophilic and will crossthe placenta via diffusion-limited transfer.54 The rate of diffusion is gated by the size

f the molecule as well as the permeability of the placenta, which changes withestational age. Based on animal studies, synthetic opiates such as fentanyl,ethadone, tramadol, and loperamide may exhibit greater transfer across thelacenta than morphine-based drugs and subsequent increased fetal exposure.55

Opiate metabolism generates glucuronidated compounds that are found in bothurine and meconium. Neonate urine samples provide a narrow window of detection ofmaternal drug use in the past 3 to 7 days. For heroin, the generation of 6-mono-acetylmorphine is only detectable hours after use before the intermediate is convertedentirely to morphine. The major cyclic metabolite of methadone, 2-ethylidene-1,5-dimethyl-3,3-diphenyl-pyrrolidine (EDDP), has been detected in meconium usingfluorescence polarization and high-performance LC with diode array detection.56 Acorrelation was seen with the maternal methadone dose and the amount of meconiummethadone measured but not with the methadone metabolite EDDP.

Immunoassay reactivity toward opiates varies depending on manufacturer, plat-form, and target compound. Semisynthetic morphine–based opiates such as oxy-codone and buprenorphine may show partial reactivity, whereas nonmorphine–basedopiates such as methadone and fentanyl may not react at all. Methanolic extraction

of meconium is frequently used for isolation of opiates from the specimen prior to

radtdifstrc

pepAai(cetc

460 Cotten

analysis. Extensive coverage of the methodologies and detection limits for a variety ofopiate assays in the setting of neonate drug testing has been reviewed elsewhere.12

Neonate abstinence syndromeMethadone is currently the only approved medication for use during pregnancy byopioid-dependent women.57 Maintenance therapy for pregnant women is thought toeduce illicit drug–seeking behavior, minimize withdrawal-induced stress on the fetus,nd improve prenatal care. Maternal methadone use exposes the neonate to opiatesuring gestation, which can present central and autonomic nervous system dysfunc-ion at birth, frequently called neonate abstinence syndrome (NAS). Assessmentepends on observed behaviors such as excessive crying, poor sleep, tremors,

ncreased muscle tone, generalized seizure, hyperthermia, tachypnea, poor feeding,ailure to thrive, and irritability. Pharmacologic management depends on degree ofeverity, with 52% of NAS patients requiring only opiates for treatment.58 Second lineherapies include phenobarbital and opiates in 32% of cases. Buprenorphine hasecently been reported as a safe and effective treatment for NAS, along withlonidine.59,60

Amphetamines

Amphetamine drug testing in the neonate is similar to opiate testing in that it mustdiscriminate between over-the-counter, prescription, and illicit compounds within thesame class. Immunoassay screening serves only to capture the maximum number oftrue-positives for reflex confirmation testing while minimizing false-positives. Speci-ficity of reagents and platforms varies for amphetamine drugs of abuse detection. Itis therefore up to laboratories to educate health care professionals in proper test utilityand institution-specific limitations.

Methamphetamine abuse has increased sharply in specific regions of the countryover the last 10 years, with 11,239 clandestine laboratories discovered in 2010.61

Tennessee, Kentucky, and Illinois account for 30% of the total manufacturing ofmethamphetamine in the United States. Approximately 17% to 44% of laboratorieshave children living in the same building. In the last several years, a shift in thesynthetic route used for the manufacturing of methamphetamine has occurred thatallows for small-scale mobile manufacture in 2 L bottles. The method known as“shake and bake” has increased the number of amateur producers and subsequentlyburn victims admitted to hospitals.

Maternal use of methamphetamine increased from 8% in 1994 to 24% in 2006 forpregnant women positive for an illicit substance.62 Methamphetamine crosses the

lacenta within 30 seconds of injection in animal studies and exhibits slowerlimination in the fetus resulting in longer exposure. Heavy prenatal use of metham-hetamine is associated with high concentrations (200–1000 ng/g) in meconium.mphetamines, including methamphetamine, MDMA (methylenedioxymethamphet-mine), and MDA (methylenedioxyamphetamine) can be deaminated or hydroxylated

n the liver to inactive metabolites. Unlike other drug classes, a significant portion30%–40%) is excreted unchanged in the urine, making detection of the parentompound acceptable for assessing exposure. Detection in neonate urine capturesxposure several days before birth. False-positive results from labetalol, a drug usedo treat hypertension during pregnancy, have been reported, highlighting the need foronfirmation of positive immunoassay screening results from maternal urine.63

Methanolic extraction often coupled with solid phase purification is used forextraction of amphetamines from meconium.42 Confirmation by GC and LCMS offers

sensitivities in the ng range per gram of meconium. Extraction efficiencies between

mccn

461Neonatal Drug Testing

40% and 87% have been reported across a wide range of concentrations. Addition-ally, specific byproducts of methamphetamine production have been used in theforensic setting to determine which route was used for synthesis.61,64

Side effects of prenatal methamphetamine and amphetamine exposure includedecreased gestational age, birth weight, length, and occipitofrontal circumference.62

Additionally, infants are more likely to be admitted to neonate intensive care units afterbirth. Beyond initial clinical effects, decreases in mother-infant bonding, decreases inbreast-feeding, higher adoption rates, neglect, and court-ordered placement outsidethe home are associated with mothers who use methamphetamine.65

Ethanol

Assessment of ethanol abuse during pregnancy is challenging because of rapidmetabolism and elimination by the body.66 The pseudo-zero order kinetics and rapiddistribution throughout tissues make detection in the blood and urine difficult.Diagnostic tests to assay neonate exposure to alcohol are currently evaluating theemerging biomarkers of FAEEs, which are produced through nonoxidative breakdownof ethanol (Fig. 5).67 Normal reaction of fatty acids with glycerol in the body produces

onoglycerides, diglycerides, and triglycerides that act as secondary messengers andellular components. In the presence of ethanol, fatty acids form FAEEs through aondensation reaction. Once formed, FAEEs are deposited in fatty tissues and meco-

Fig. 5. Fatty acid ethyl ester biosynthesis: FAEEs are generated in the nonoxidative break-down of ethanol. Reaction of ethanol with endogenous fatty acids generating FAEEs andtheir accumulation in meconium may be a good indicator of maternal ethanol use duringpregnancy.

ium, thereby serving as markers for long-term alcohol exposure during pregnancy.

fiai

siflsmb(ed

462 Cotten

Studies evaluating transfer of FAEEs across the placenta demonstrate that FAEEsgenerated by maternal metabolism are broken down and degraded in the placenta.19

Accumulation of FAEEs in meconium, therefore, represents ethanol that has beentransferred across the placenta and metabolized by the fetus in utero. The majorFAEEs currently under evaluation include ethyl laurate, ethyl myristate, ethyl palmi-tate, ethyl palmitoleate, ethyl stearate, ethyl oleate, ethyl linoleate, ethyl alpha-linoleate, ethyl arachidonate, and ethyl heptadecanoate.68

Correlation between maternal alcohol consumption and FAEEs levels in meconiumshow a positive relationship, but the specific analytes and cutoff levels definingexposure remain unclear. Total FAEEs levels are suggested to provide betterpredictive power for maternal alcohol abuse rather than select individual metabolites.Furthermore, the cutoff levels for positive prenatal alcohol exposure vary by study,with reported cutoffs ranging from 50 to 600 ng/g.66

Extraction of FAEEs from meconium has been achieved with hexane/acetone orsolid phase extraction with recoveries ranging from 20% to 93% depending on thespecific metabolite tested. Testing methods have used GC–flame ion detector, fullscan GC-MS, selected ion monitoring GC-MS, and GC-MS/MS for screening andconfirmation of various analytes.

Fetal alcohol spectrum disorderPrenatal exposure to alcohol can result in a broad array of detrimental effects to thedeveloping fetus. Fetal alcohol spectrum disorder (FASD) is the umbrella term used toclassify neonates that exhibit one or more of the features associated with ethanolexposure including growth retardation, abnormal facial features, and central nervoussystem impairment.69 Height and weight below the 10th percentile, short palpebralssures, thin vermillion border, smooth philtrum, and impaired cognitive function arell used as criteria for diagnosis of FASD. It is estimated that the occurrence of FASD

s between 0.2 and 2 cases per 1000 births in the United States.69 Secondary effectsthat may present later in life include legal trouble, mental health issues, behavioralproblems, and low social adaptability. The cost of FASD has been estimated at $3.6billion annually, illustrating its severity as a major public health issue.

MANAGEMENT OF WITHDRAWAL FROM DRUGS OF ABUSE

Guidelines for medical management for chronic prenatal exposure to opiates,benzodiazepine, and alcohol have been proposed by the SOGC5 (Table 1). Neonateshowing abstinence syndrome from opiates can benefit from symptomatic therapy,ncluding Gravol for nausea and vomiting and acetaminophen/nonsteroidal antiin-ammatory drugs for myalgias. Methadone and buprenorphine can be initiated totep-down opiate exposure over time. If methadone is unavailable, morphine 5 to 10g by mouth every 4 to 6 hours as needed can be substituted. For chronicenzodiazepine exposure, two-thirds to three-fourths of the equivalent adult dose

mg/kg) can be administered, tapering 10% of the dose per day. Chronic alcoholxposure withdrawal can benefit from thiamine, folic acid, diazepam, or lorazepamepending on severity, with close monitoring of electrolytes.

FUTURE PERSPECTIVES

The neonate population holds unique challenges in relation to assessment of drugexposure. The specimens available show distinct differences from adult samplesregarding windows of detection and analyte deposition. Accurate results are imper-

ative given the legal and medical management issues related to newborn care.

463Neonatal Drug Testing

Current research in the area of FAEEs may provide better tools to evaluate prenatalalcohol exposure in the future. Several additional facets of neonate drug testingdeserve more research to understand exposure rates more clearly and detrimentaleffects of drugs during development. To date, little information exists regarding thefrequency and level of the cocaine adulterant levamisole in meconium. Currentestimates for levamisole prevalence in the adult population are approaching 100% forpatients who screen positive for benzoylecgonine. Additionally, the synthetic canna-binoids and novel ketone amphetamines currently emerging in adult populations havenot been evaluated in the neonate population.

REFERENCES

1. US Department of Health and Human Services. Results from the 2010 NationalSurvey on Drug Use and Health: summary of national findings. Substance Abuse andMental Health Services Administration report No. SMA 11-4658. 2011. Available at:http://oas.samhsa.gov/NSDUH/2k10NSDUH/2k10Results.htm. Accessed April 10,2012.

2. Havens JR, Simmons LA, Shannon LM, et al. Factors associated with substanceuse during pregnancy: results from a national sample. Drug Alcohol Depend2009;99:89 –95.

3. Yonkers KA, Gotman N, Kershaw T, et al. Screening for prenatal substance use:development of the Substance Use Risk Profile-Pregnancy scale. Obstet Gynecol2010;116:827–33.

4. Substance abuse in pregnancy. The global library of women’s medicine. Published 2009.Available at: http://www.glowm.com/?p�glowm.cml/section_view&articleid�115. Ac-cessed April 10, 2012.

5. Wong S, Ordean A, Kahan M. SOGC clinical practice guidelines: substance use inpregnancy: no. 256. Int J Gynaecol Obstet 2011;114:190–202.

6. Murphy-Oikonen J, Montelpare WJ, Southon S, et al. Identifying infants at risk forneonatal abstinence syndrome: a retrospective cohort comparison study of 3 screen-

Table 1Management of withdrawal symptoms

Substance Recommended Management

Alcohol Thiamine 100 mg po od � 3 d, folic acid 5 mg po od.Diazepam 20 mg po q 1–2 h until minimal symptoms.Lorazepam 2–4 mg sl/po q 2–4 h prn during labor.Monitor hydration status.

Benzodiazepines Start at two-thirds to three-fourths of diazepam equivalent dose.Taper by 10% per day.

Opiates Offer symptomatic therapy including Gravol for nausea and vomiting,acetaminophen/NSAIDs for myalgias.

Consider methadone or buprenorphine initiation.Can use morphine 5–10 mg po q 4–6 h prn if methadone is not available.

Abbreviations: NSAIDs, nonsteroidal antiinflammatory drugs; od, once daily; po, by mouth; prn, asneeded; q, every; sl, sublingual.Data from Wong S, Ordean A, Kahan M. SOGC clinical practice guidelines: substance use inpregnancy: no. 256. Int J Gynaecol Obstet 2011;114:190–202.

ing approaches. J Perinat Neonatal Nurs 2010;24:366–72.

464 Cotten

7. University of North Carolina Department of Labor and Delivery. Clinical guidelines forurine toxicology screening. Chapel Hill (NC): University of North Carolina Departmentof Labor and Delivery; 2011.

8. Child Welfare Information Gateway. Parental drug use as child abuse. Published2009. Available at: http://www.childwelfare.gov/systemwide/laws_policies/statutes/drugexposed.cfm. Accessed April 10, 2012.

9. US Department of Health and Human Services. Child Abuse Prevention and Treat-ment Act. Available at: http://www.acf.hhs.gov/programs/cb/laws_policies/cblaws/capta/. Accessed April 10, 2012.

10. Farst KJ, Valentine JL, Hall RW. Drug testing for newborn exposure to illicit sub-stances in pregnancy: pitfalls and pearls. Int J Pediatr 2011;2011:951616.

11. Gallardo E, Queiroz JA. The role of alternative specimens in toxicological analysis.Biomed Chromatogr 2008;22:795–821.

12. Lozano J, García-Algar O, Vall O, et al. Biological matrices for the evaluation of in uteroexposure to drugs of abuse. Ther Drug Monit 2007;29:711–34.

13. Browne SP, Tebbett IR, Moore CM, et al. Analysis of meconium for cocaine inneonates. J Chromatogr 1992;575:158–61.

14. Ostrea EM Jr, Romero A, Knapp DK, et al. Postmortem drug analysis of meconium inearly-gestation human fetuses exposed to cocaine: clinical implications. J Pediatr1994;124:477–9.

15. Kwong TC, Ryan RM. Detection of intrauterine illicit drug exposure by newborn drugtesting. National Academy of Clinical Biochemistry. Clin Chem 1997;43:235–42.

16. Verma A, Dhanireddy R. Time of first stool in extremely low birth weight (� or � 1000grams) infants. J Pediatr 1993;122:626–9.

17. Cone EJ, Clarke J, Tsanaclis L. Prevalence and disposition of drugs of abuse andopioid treatment drugs in oral fluid. J Anal Toxicol 2007;31:424–33.

18. Bosker WM, Huestis MA. Oral fluid testing for drugs of abuse. Clin Chem 2009;55:1910–31.

19. Chan D, Knie B, Boskovic R, et al. Placental handling of fatty acid ethyl esters:perfusion and subcellular studies. J Pharmacol Exp Ther 2004;310:75–82.

20. Sadler T. Langman’s Medical Embryology. 12th edition. Philadelphia: LippincottWilliams & Wilkins; 2011. p. 106.

21. Winecker RE, Goldberger BA, Tebbett I, et al. Detection of cocaine and its metabolitesin amniotic fluid and umbilical cord tissue. J Anal Toxicol 1997;21:97–104.

22. Ripple MG, Goldberger BA, Caplan YH, et al. Detection of cocaine and its metabolitesin human amniotic fluid. J Anal Toxicol 1992;16:328–31.

23. Jain L, Meyer W, Moore C, et al. Detection of fetal cocaine exposure by analysis ofamniotic fluid. Obstet Gynecol 1993;81:787–90.

24. Casanova OQ, Lombardero N, Behnke M, et al. Detection of cocaine exposure in theneonate. Analyses of urine, meconium, and amniotic fluid from mothers and infantsexposed to cocaine. Arch Pathol Lab Med 1994;118:988–93.

25. Marques PR, Tippetts AS, Branch DG. Cocaine in the hair of mother-infant pairs:quantitative analysis and correlations with urine measures and self-report. Am J DrugAlcohol Abuse 1993;19:159–75.

26. Potter S, Klein J, Valiante G, et al. Maternal cocaine use without evidence of fetalexposure. J Pediatr 1994;125:652–4.

27. Cotten SW, Duncan DL, Burch EA, et al. Unexpected interference of baby washproducts with a cannabinoid (THC) immunoassay. Clin Biochem 2012;45(9):605–9.

28. Uebel RA, Wium CA. Toxicological screening for drugs of abuse in samples adulter-

ated with household chemicals. S Afr Med J 2002;92:547–9.

465Neonatal Drug Testing

29. Mikkelsen SL, Ash KO. Adulterants causing false negatives in illicit drug testing. ClinChem 1988;34:2333–6.

30. Xia Y, Wang P, Bartlett MG, et al. An LC-MS-MS method for the comprehensiveanalysis of cocaine and cocaine metabolites in meconium. Anal Chem 2000;72:764 –71.

31. Murphey LJ, Olsen GD, Konkol RJ. Quantitation of benzoylnorecgonine and othercocaine metabolites in meconium by high-performance liquid chromatography. JChromatogr 1993;613:330–5.

32. Vitros Chemistry. Cocaine metabolite immunoassay reagent. Rochester (NY): OrthoClinical Diagnostics; 2010.

33. Bada HS, Das A, Bauer CR, et al. Gestational cocaine exposure and intrauterinegrowth: maternal lifestyle study. Obstet Gynecol 2002;100:916–24.

34. Bauer CR, Shankaran S, Bada HS, et al. The Maternal Lifestyle Study: drug exposureduring pregnancy and short-term maternal outcomes. Am J Obstet Gynecol 2002;186:487–95.

35. Bandstra ES, Morrow CE, Accornero VH, et al. Estimated effects of in utero cocaineexposure on language development through early adolescence. Neurotoxicol Teratol2011;33:25–35.

36. Bada HS, Bann CM, Bauer CR, et al. Preadolescent behavior problems after prenatalcocaine exposure: relationship between teacher and caretaker ratings (MaternalLifestyle Study). Neurotoxicol Teratol 2011;33:78–87.

37. Vitros Chemistry. THC immunoassay reagent. Rochester (NY): Ortho Clinical Diag-nostics; 2010.

38. ElSohly MA, Feng S. Delta 9-THC metabolites in meconium: identification of 11-OH-delta 9-THC, 8 beta,11-diOH-delta 9-THC, and 11-nor-delta 9-THC-9-COOH asmajor metabolites of delta 9-THC. J Anal Toxicol 1998;22:329–35.

39. Feng S, ElSohly MA, Salamone S, et al. Simultaneous analysis of delta9-THC and itsmajor metabolites in urine, plasma, and meconium by GC-MS using an immunoaffinityextraction procedure. J Anal Toxicol 2000;24:395–402.

40. Maynard EC, Amoruso LP, Oh W. Meconium for drug testing. Am J Dis Child1991;145:650–2.

41. Bar-Oz B, Klein J, Karaskov T, et al. Comparison of meconium and neonatal hairanalysis for detection of gestational exposure to drugs of abuse. Arch Dis Child FetalNeonatal Ed 2003;88:F98–100.

42. ElSohly MA, Stanford DF, Murphy TP, et al. Immunoassay and GC-MS procedures forthe analysis of drugs of abuse in meconium. J Anal Toxicol 1999;23:436–45.

43. Moore C, Lewis D, Becker J, et al. The determination of 11-nor-delta 9-tetrahydro-cannabinol-9-carboxylic acid (THCCOOH) in meconium. J Anal Toxicol 1996;20:50–1.

44. Warner A. Interference of common household chemicals in immunoassay methodsfor drugs of abuse. Clin Chem 1989;35:648–51.

45. Hatch EE, Bracken MB. Effect of marijuana use in pregnancy on fetal growth. Am JEpidemiol 1986;124:986–93.

46. El Marroun H, Tiemeier H, Steegers EA, et al. Intrauterine cannabis exposure affectsfetal growth trajectories: the Generation R Study. J Am Acad Child Adolesc Psychiatry2009;48:1173–81.

47. Goldschmidt L, Richardson GA, Willford J, et al. Prenatal marijuana exposure andintelligence test performance at age 6. J Am Acad Child Adolesc Psychiatry 2008;47:254–63.

48. Gray KA, Day NL, Leech S, et al. Prenatal marijuana exposure: effect on child

depressive symptoms at ten years of age. Neurotoxicol Teratol 2005;27:439–48.

466 Cotten

49. Day NL, Goldschmidt L, Thomas CA. Prenatal marijuana exposure contributes to theprediction of marijuana use at age 14. Addiction 2006;101:1313–22.

50. Dreher MC, Nugent K, Hudgins R. Prenatal marijuana exposure and neonatal out-comes in Jamaica: an ethnographic study. Pediatrics 1994;93:254–60.

51. Lester BM, Dreher M. Effects of marijuana use during pregnancy on newborn cry.Child Dev 1989;60:765–71.

52. van Gelder MM, Reefhuis J, Caton AR, et al. Characteristics of pregnant illicit drugusers and associations between cannabis use and perinatal outcome in a population-based study. Drug Alcohol Depend 2010;109:243–7.

53. Gray TR, Eiden RD, Leonard KE, et al. Identifying prenatal cannabis exposure andeffects of concurrent tobacco exposure on neonatal growth. Clin Chem 2010;56:1442–50.

54. Olsen GD. Placental permeability for drugs of abuse and their metabolites. NIDA ResMonogr 1995;154:152–62.

55. Szeto HH. Kinetics of drug transfer to the fetus. Clin Obstet Gynecol 1993;36:246 –54.

56. Stolk LM, Coenradie SM, Smit BJ, et al. Analysis of methadone and its primarymetabolite in meconium. J Anal Toxicol 1997;21:154–9.

57. Jansson LM, Velez M, Harrow C. The opioid-exposed newborn: assessment andpharmacologic management. J Opioid Manag 2009;5:47–55.

58. Sarkar S, Donn SM. Management of neonatal abstinence syndrome in neonatalintensive care units: a national survey. J Perinatol 2006;26:15–7.

59. Kraft WK, Gibson E, Dysart K, et al. Sublingual buprenorphine for treatment ofneonatal abstinence syndrome: a randomized trial. Pediatrics 2008;122:601–7.

60. Esmaeili A, Keinhorst AK, Schuster T, et al. Treatment of neonatal abstinencesyndrome with clonidine and chloral hydrate. Acta Paediatr 2010;99:209–14.

61. Vearrier D, Greenberg MI, Miller SN, et al. Methamphetamine: history, pathophysiol-ogy, adverse health effects, current trends, and hazards associated with the clandes-tine manufacture of methamphetamine. Dis Mon 2012;58:38–89.

62. Good MM, Solt I, Acuna JG, et al. Methamphetamine use during pregnancy: maternaland neonatal implications. Obstet Gynecol 2010;116:330–4.

63. Yee LM, Wu D. False-positive amphetamine toxicology screen results in threepregnant women using labetalol. Obstet Gynecol 2011;117:503–6.

64. Shakleya DM, Plumley AE, Kraner JC, et al. Trace evidence of trans-phenylpropene asa marker of smoked methamphetamine. J Anal Toxicol 2008;32:705–8.

65. Ladhani NN, Shah PS, Murphy KE. Prenatal amphetamine exposure and birthoutcomes: a systematic review and metaanalysis. Am J Obstet Gynecol 2011;205:219, e1–7.

66. Gareri J, Klein J, Koren G. Drugs of abuse testing in meconium. Clin Chim Acta2006;366:101–11.

67. Best CA, Laposata M. Fatty acid ethyl esters: toxic non-oxidative metabolites ofethanol and markers of ethanol intake. Front Biosci 2003;8:e202–17.

68. Laposata M. Fatty acid ethyl esters: ethanol metabolites which mediate ethanol-induced organ damage and serve as markers of ethanol intake. Prog Lipid Res1998;37:307–16.

69. Riley EP, Infante MA, Warren KR. Fetal alcohol spectrum disorders: an overview.Neuropsychol Rev 2011;21:73–80.

Emerging Biomarkers ofIntrauterine Neonatal andPediatric Exposures toXenobiotics

Kaitlyn Delano, BSca,b, Gideon Koren, MD, FRCPC, FACMTa,b,*

KEYWORDS

� Biomarker � Drugs � Alcohol � Pediatric � Screening

KEY POINTS

� Biomarkers of exposure can provide clinically relevant information to assist in diagnosisor in evaluation of the severity of chemical exposure, leading to optimal management.

� Using hair and meconium as matrices for biomarkers allows for a longer window of detec-tion of exposures to xenobiotics including drugs, alcohol, and environmental chemicals.

� The use of biomarkers of alcohol that provide more reliable information concerningprenatal ethanol exposure than maternal self-reports aids in the diagnosis of fetal alcoholspectrum disorder.

� Numerous biomarkers are available to detect intrauterine exposure to drugs of abuse,which is associated with many adverse outcomes for fetus, child, and mother.

� Development of environmental toxicant biomarkers provides clinicians with tools todetect long-term, low-level exposures in humans, which may have detrimental effectsto their offspring.

INTRODUCTION

There are multiple definitions available for a biomarker, specific to how the biomarkeris used. The official National Institutes of Health definition of a biomarker is “a charac-teristic that is objectively measured and evaluated as an indicator of normal biologicprocesses, pathogenic processes, or pharmacologic responses to a therapeutic inter-vention.1” In the context of detecting external toxins after exposure during intrauterinelife, biomarkers are critical, because many chemicals may not exist any more in theblood or urine of the neonate. Consequently, without the availability of appropriate

a Department of Pharmacology, University of Toronto, 1 King’s College Circle, Toronto, OntarioM5S 1A8, Canada; b Division of Clinical Pharmacology and Toxicology, The Hospital for SickChildren, 555 University Avenue, Toronto, Ontario M5G 1X8, Canada* Corresponding author.E-mail address: gkoren@sickkids.ca

Pediatr Clin N Am 59 (2012) 1059–1070http://dx.doi.org/10.1016/j.pcl.2012.07.005 pediatric.theclinics.com0031-3955/12/$ – see front matter � 2012 Elsevier Inc. All rights reserved.

Delano & Koren1060

biomarkers, even potential toxic intrauterine exposures may be missed. Therefore,biomarkers must be able to generate relevant preclinical or clinical interpretations.2

The sensitivity and specificity of a biomarker are important, because biomarkers toosensitive or nonspecific may not detect exposures or effects that are clinicallyrelevant.3

In this review, we focus on biomarkers of internal dose, a subtype of biomarkers ofexposure, which indicate the occurrence and extent of exposure to a compound or itsmetabolite(s).3 Measuring the amount of the compound or metabolite in a matrixallows for a measurement of the exposure, rather than only estimating it.3 By usingbiomarkers of both the compound and metabolite(s), more information concerningthe exposure can be gathered and more accurate interpretations can be made bythe clinician. When available, using multiple biomarkers in conjunction may providemore clinically relevant information about the exposure.Illicit substance use in pregnancy is associated with significant maternal and

neonatal morbidity and economic burdens to the health care system.4 Despite a poten-tial increase in substance abuse during pregnancy, it remains underdiagnosed orcompletely undiagnosed, putting both the fetus and the mother at risk for long-termsequelae.5 Maternal self-report has been commonly used in the past to assess poten-tial fetal exposure, but has been found to be unreliable and not correlate with expo-sure.6 Using biomarkers to detect use of drugs, alcohol, or environmental toxinscan help determine the optimal management of the child.

HAIR AND MECONIUM AS MATRICES FOR IN UTERO BIOMARKERS

Most often, blood and urine are used to test for drug and alcohol use or exposure.Although these 2 matrices are well established, they provide information on only veryrecent use or exposure, because of the short elimination half-lives of most drugs ofabuse.7,8 Longer-term, chronic exposure is not detected using blood or urine, requiringalternative matrices to capture this type of exposure. Hair and meconium, in neonates,have emerged as novel matrices that provide a wider window of detection. These 2matrices can be tested to assess prenatal exposure to chemicals, including thoseresulting from maternal usage.

Hair

Hair follicle development occurs because of ectodermal and mesodermal interactionsduring epidermal development, beginning at approximately the eighth week of develop-ment.9 The base of the follicle, the dermal papilla, is derived from the mesodermalmesenchyme of the dermis, whereas the remainder of the hair follicle is derived fromthe ectoderm.9 The pigmentation of hair and skin is caused by melanocytes, whichdevelop from the neural crest cells.9 Melanin pigments, eumelanin and pheomelanin,are synthesized and stored in the melanosomes. Eumelanin produces brown and blackhair, whereas pheomelanin is responsible for red and blond hair. The proportion ofthese 2 melanin pigments is what dictates the final color of human hair.10 The appear-ance of hair follicles occurs at around the 10th week of fetal development, andcontinued differentiation results in the formation of various components of the follicle.9

Three types of glands are associated with the hair follicle: sebaceous, apocrine, andsweat glands. The sebaceous gland, responsible for the production of sebum,develops on the side of the follicle and is associated with capillary networks, similarto the hair follicle. Sebum is composed of free and combined fatty acids and unsapo-nifiable material (eg, cholesterol and waxes).9,11 Apocrine glands secrete an oily,colorless substance directly into the follicle, and are localized in the axilla, eyelids,

Emerging Biomarkers 1061

and external auditory meatus.11 Sweat glands, located on most of the body surface,produce sweat, comprising mainly water and salts.11

Hair growth occurs in a 3-phase cycle, consisting of anagen, catagen, and telogenphases.11 The anagen phase represents hair production and begins at the 15th weekof fetal development, and the scalp of the fetus is completely covered with anagenphase follicles between the 18th and 20th week of gestation.9 This phase is character-ized by a rapid proliferation of matrix cells, which fill the follicle bulb, extending throughepithelial cells.9 The matrix cells are then keratinized, forming the strand of hair.11

Growth of the hair continues until expression of epidermal growth factors, resultingin apoptosis of follicular keratinocytes and melanocytes, or the catagen phase. Duringthe catagen phase, the bottom of the hair fiber is fully keratinized. At this point, the hairfollicle is dormant and in the telogen phase. In respect to a neonate, the first full cycleof hair growth is complete between the 24th and 28th week of development, with thenext cycle starting soon after the first one is completed.9 Consequently, biomarkerspresent in hair at birth reflect exposure to toxins during the third trimester, and areable to be tested until approximately 3 months after birth. Therefore, if there is hairpresent after birth, the window of detection is relatively long.In children (and adults), the hair growth cycle continues, with each phase having

a distinct length of time. The growth of each hair follicle is independent, with approx-imately 85% of all head hair follicles in the anagen phase (growing phase) at any giventime.11 The remainder of hair follicles are not in the growing phase, and this must betaken into account when interpreting any nonneonatal hair results, because drugincorporation does not occur during the resting phase.Routes of incorporation of compounds into hair include direct incorporation of

a chemical via the capillary networks of the hair follicle, as a result of secretionsfrom the sebaceous and sweat glands, and as a result of environmental or externalexposure.12 Measuring and interpreting biomarkers in hair must properly take thesedifferent routes of incorporation into account.The many factors that determine the concentration of a drug in hair should also be

considered. These factors include hair color (melanin content), physicochemical prop-erties of the drug, and cosmetic treatment of hair.13 For example, it has been foundthat drugs preferentially incorporate into darker hair, most likely because of the relativeamounts of eumelanin.10 Also, physicochemical properties including lipophilicity,basicity, and membrane permeability affect the ability of drugs to incorporate intohair. Generally, basic, lipophilic drugs tend to accumulate more readily into hairsamples than more acidic or polar drugs.12 Cosmetic treatment of hair has been foundto decrease hair concentrations of drugs, mainly because of increased hair damageand the removal of hair color pigment.13 Hair damage causes drug molecules to belost more easily from the matrix, whereas removal of pigment reduces the amountof melanin to which the drugs are bound.13 As a result, clinical interpretations ofdrug concentrations in hair must take these factors into account.Hair samples are typically collected from the posterior vertex, because hair from this

area of the scalp shows the most constant rate of growth.11 Analysis of hair provideslong-term information on an individual’s drug use or exposure. Taking advantage ofthe uniform growth rate of human hair, approximately 1 cm/mo, it is possible to segmenthair samples to more accurately assess the time and pattern of use or exposure.7 Thewindow of detection for nonneonatal hair samples is therefore dependent on hair length.

Meconium

The first few bowel movements of a neonate are composed of meconium.14 This highlycomplex matrix begins to form approximately during the 12th week of gestation and

Delano & Koren1062

consists of water, gastrointestinal tract epithelial cells, bile acids and salts, enzymes,sugars, lipids, intestinal secretions, and swallowed amniotic fluid.15 Fetal swallowingof amniotic fluid is the mechanism believed to concentrate compounds within meco-nium as fetal urine is deposited into the amniotic fluid and is subject to swallowingagain.15 Determining factors of drug incorporation into meconium and the extent oftheir concentration are mainly determined on the ability of the drug to cross theplacenta. Most drugs are able to transfer across the placenta, and the rate of transferis then determined by molecule size, ionization state, lipophilicity, and proteinbinding.15 Because most drugs are small enough to transfer via passive diffusion,the major limiting factor, in terms of drug transport to the fetus, is placental bloodflow.16 Once meconium is formed in the fetal intestine, it is considered a physicallystatic matrix, becoming a record of fetal exposure to the drugs in question duringthe second and third trimesters of pregnancy.17 A positive meconium test indicatesintrauterine exposure during the second and third trimesters, but is unable to showtime or pattern of use.18

Dose-response relationships are difficult to determine using meconium samples,mainly because of urine contamination. If fetal exposure occurs close to term andthe compound is incorporated into the urine, contamination of meconium can occuronce urine is evacuated into a soiled diaper.15 This situation increases the sensitivityof meconium testing because of the increased compound levels in the sample.However, it could affect the ratio of drugs and metabolites in the sample, and thedevelopment of dose-response relationships.Collection of meconium specimens is easy and noninvasive. Because it is discarded

material and there is usually sufficient quantity for analysis, this matrix is practical anduseful.14 Ninety-nine percent of infants pass their first meconium within 48 hours,giving this matrix a wider window of detection than blood or urine.15 Once 48 hourshave passed, it is necessary to evaluate the texture and odor of the sample to deter-mine whether it still meconium or has changed to postnatal feces. The time allowed forsample collection, 2 days, may be seen as limited, but if the neonate is at high risk fordrug or alcohol in utero exposure and in hospital care, obtaining a viable sample is notproblematic. Collected samples should be minimally 0.5 g, to provide sufficientsample for all analyses. Storage of samples for analysis should be at –20�C or –80�C.15

BIOMARKERS FOR ALCOHOL USE

Alcohol exposure during pregnancy can have many serious consequences for theoffspring, including birth defects and deficits in cognitive performance and mentaldevelopment.19,20 The fetal alcohol spectrum disorder (FASD) is an umbrella term,which describes the consequences associated with prenatal alcohol exposure.6

Affecting approximately 1% of all live births in North America, FASD is a major socialand economic burden, which can be minimized by early diagnosis and diseasemanagement.21 The clinical presentation of the disease is inconsistent, some lackingevidence of central nervous system neurodevelopment abnormalities.15 Typically, forthe definite diagnosis of FASD, confirmation of prenatal ethanol exposure is neededbecause some of the characteristics of the disease, in particular physical, may beabsent.8

Detection of ethanol or its aldehyde has a limited window because they are bothrapidly cleared from the blood. A lack of relationship between maternal ethanolconsumption and maternal ethanol levels also limits their use to provide an objectiveanalysis.22 Ethanol itself is not readily incorporated into matrices because of its volatilenature, limiting the window of detection to the hours leading up to delivery, when many

Emerging Biomarkers 1063

alcohol-dependent mothers may not drink.23 Two biomarkers, fatty acid ethyl esters(FAEEs) and ethyl glucuronide (EtG), have been established and are currently usedto detect exposure or use of alcohol.

FAEEs

FAEEs are nonoxidative metabolites of ethanol, formed through the esterification ofethanol with endogenous fatty acids or fatty acyl-coenzyme A (CoA).23 These reac-tions are catalyzed by FAEE synthase or microsomal acyl-CoA:ethanol O-acetyltrans-ferase.19 In particular, FAEE synthase is present in almost all human tissues, withactivity being detected in the heart, liver, lungs, adipose tissue, gall bladder, andpancreas.23 Depending on the carbon chain length and the location of the doublebond, different species of FAEEs can be formed.8 Most FAEEs are transported byalbumin within the blood, and once free, are readily broken down by cellular structuresin the blood, liver, and pancreas.23 Unlike ethanol, FAEEs persist in the body for morethan a day after significant alcohol consumption, and are able to accumulate invarious matrices.22 Also in contrast to ethanol, FAEEs are readily metabolized bythe placenta and thus do not cross.8 This characteristic indicates that FAEE levelspresent in meconium represent fetal ethanol metabolism and thus fetal exposure toethanol.17,18

Several individual FAEEs are analyzed in samples to detect alcohol exposure or use.Interindividual variation in the amount of specific FAEEs formed can result fromgenetic variations in the enzymes responsible for FAEE formation, the amounts ofspecific fatty acids in different diets, the degree of alcohol exposure or consumption,and FAEE synthase enzyme kinetics.8,15,22 Ethyl palmitate, oleate, stearate, and lino-leate are the predominate FAEEs found in meconium of ethanol-exposed neonates.23

The cumulative level of select FAEEs is measured because this provides a redundancysystem, resulting in higher efficiency, sensitivity, and specificity.15,24

In several studies looking at baseline FAEE levels, infants born to women who didnot drink alcohol during pregnancy had low FAEE levels.25 Because the bodyproduces some ethanol during normal metabolism, FAEEs are detected in individualswho do not consume alcohol, thus requiring a clear cutoff value specific to the matrixfor differentiation between heavy-drinking and non drinking individuals.26

With respect to meconium, the positive cutoff of cumulative FAEE levels was estab-lished at 2 nmol/g meconium.24 This cutoff has 100% sensitivity and 98.4% specificityfor detection of heavy fetal alcohol exposure.27 However, this cutoff value does notallow differentiating between neonates born to non drinkers and social drinkers.24

An important limitation of FAEE meconium testing is that samples excreted later inthe postpartum period have higher levels of FAEEs than samples collected earlierfor the same infant because of de novo production of alcohol from carbohydrates inthe meconium.28 This situation could lead to false-positive FAEE results, and it is rec-ommended to collect meconium samples within 24 hours to ensure that FAEE resultsproperly reflect in utero ethanol exposure.The FAEE cutoff value for adult hair was established to be 50 ng/g hair, but at this

level would identify not only heavy drinkers but some social drinkers too.29 Hair FAEEresults between 20 and 50 ng/g hair indicate moderate levels (ie, social) of drinking.8

This FAEE cutoff provides optimal sensitivity and specificity, both 90%.19 Hair careproducts can result in increased FAEE levels by causing localized FAEE productionon the scalp.29 Differentiating between external and incorporated FAEEs may not benecessary in neonatal hair samples, because the use of hair care products in neonatesis not relevant. To confirm FAEE results in adult hair, a secondary confirmatory test isneeded, as discussed later.

Delano & Koren1064

EtG

EtG is a minor metabolite of ethanol, formed when ethanol is glucuronidated with acti-vated glucuronic acid instead of water.10 Although EtG testing lacks sensitivity indetecting moderate or social drinkers, it is highly specific for heavy alcohol use.30

EtG measurements are used to confirm hair FAEE results. As mentioned earlier,increased hair FAEE levels can reflect alcohol abuse, social drinking, or use ofethanol-containing hair care products.29 EtG can help rule out ambiguous FAEEresults caused by external contamination with alcohol hair products. If a hair sampleis also positive for EtG (cutoff of 30 pg/mg hair), this indicates excessive alcohol useduring the tested time frame, and the potential influence of hair care products iseliminated.29

BIOMARKERS FOR DRUGS OF ABUSE

Illicit substance use in women of childbearing age has increased over the past 3decades, with a parallel increase during pregnancy.5 Drug use during pregnancy isa risk factor for both maternal and fetal complications.31 As well, children exposedto drug use in their environment are susceptible to many adverse outcomes associ-ated with drug use affecting their physical and mental health, as well as their socialwell-being.32 The overall increased use of drugs of abuse has necessitated the devel-opment of biomarkers of drug use, obviating the shortcomings of maternal reports,providing clinicians with tools that can detect exposure to or use of these illicitcompounds. Biomarkers for individual drugs or xenobiotics, each with unique impactson the health of the child, are discussed.

Cocaine

Prenatal cocaine exposure is associated with low birth weight, prematurity, sponta-neous abortions, stillbirths, and microcephaly. In addition, placenta complicationshave been described, including placental abruption and increased risk of diminishedblood flow to the fetus.33 Exposure to cocaine during childhood can increase the riskfor hypertension, ventricular arrhythmia, seizures, and intracranial bleeding.34 Behav-ioral problems also present themselves during development to both prenatally andpostnatally exposed children. Infants exposed in utero are found to have attentiondeficits with an increased incidence of attention-deficit/hyperactivity disorder.32

Because of the short elimination half-life of cocaine (50 minutes), detection in bloodor urine matrices is limited to a few days after use.34 Because cocaine and its metab-olites readily accumulate in both hair and meconium, these matrices can provide infor-mation on in utero exposure for infants as well as environmental exposure in olderchildren.Cocaine crosses the placenta via passive diffusion and is almost always found with

at least 1 of its metabolites in meconium samples.15 The site of metabolite production,either fetal or maternal, is undetermined. Themost common cocaine metabolite testedand found in biologic matrices is benzoylecgonine. This metabolite is formed throughhydrolysis.15 Other metabolites, including norcocaine and cocaethylene, are also usedin sample analysis for cocaine.Using all 4 of these biomarkers allows clinicians to obtain an accurate picture of

cocaine exposure and make a correct clinical interpretation. The latter is importantbecause environmental cocaine exposure occurs in different ways, including inhala-tion of crack cocaine smoke, exposure to cocaine residues on surfaces, and throughdirect contact with a cocaine-using caregiver. The presence of cocaine in a childwithout the metabolites that are produced systemically indicates environmental

Emerging Biomarkers 1065

exposure, with low risk of systemic exposure. If benzoylecgonine is also detected atconcentrations at least 10% of cocaine hair levels, it is a strong indication of systemicexposure by the child. The presence of norcocaine indicates higher levels of exposure,because this metabolite is not so readily detected, as is benzoylecgonine.Cocaethylene is an emerging biomarker that can provide additional information

regarding exposure. This metabolite is formed when cocaine and ethanol are presentconcurrently.35 By using cocaethylene as a biomarker, alcohol use can be detectedwithout assessing alcohol-specific biomarkers. This strategy has great advantages fordetecting a risk of fetal alcohol syndrome in neonates who are positive for thismetabolite.

Opioids

Opioid use has been reported in 1% to 21% of pregnant women.5 With increasing ratesof opioid use and dependency, the incidence of neonatal abstinence syndrome (NAS)has doubled in the past 5 years in Ontario, Canada.36 Because most infants born toopioid-dependent mothers suffer from NAS, this has become a real public healthissue.37 NAS is characterized by the presence of central nervous system hyperirritability,gastrointestinal dysfunction, andmetabolic, vasomotor, and respiratory disturbances.36

The treatment required for these symptoms increases the hospital stay to an average of30 to 40 days for infants, primarily in a neonatal intensive care unit.36 The recent intro-duction of buprenorphine has resulted in a decreased hospital stay and shorter durationof treatment of infants.36 Testing women suspected of using opioids could allow forearly treatment of the baby and mother. Although the methadone-treated mother iscommonly known to the medical system and social services, expecting mothersaddicted to other opioids may go undetected, and use of meconium or hair biomarkersmay be the first clue to the poor neonatal adaptation shown by the neonate.Heroin use during pregnancy may be involved in a cycle of overdose and withdrawal

and is associated with complications, including spontaneous abortion, antepartumhemorrhage, and stillbirth.38,39 Heroin is rapidly deacetylated to the active metabolite6-monoacetylmorphine (6-MAM), readily crosses the placenta, and is incorporatedinto fetal tissues within 1 hour of administration.5 Because 6-MAM persists in thesystem for a longer period than heroin, it is used as a biomarker to detect heroinuse. Along with heroin metabolism to 6-MAM, detectable levels of morphine canalso be produced.40 In addition, illicit heroin usually contains acetylcodeine, and userscommonly top up their heroin with codeine before injection.40 This practice maycomplicate the interpretation of test results, because the source of codeine andmorphine may be unknown in 6-MAM–positive samples.Codeine andmorphine are commonly tested together because morphine is a metab-

olite of codeine through CYP2D6 metabolism.41 Morphine is widely distributed in fetaltissues and its level in meconium has been found to correlate with maternal dose, time,and duration of gestational exposure.15 Similar to codeine and morphine testing, oxy-codone and its metabolite oxymorphone are commonly tested together. Hydrocodoneis tested with one of its metabolites, hydromorphone, which is also available as a sepa-rate analgesic. It is best to determine which drug was prescribed before interpreting theresults, because each of the metabolites is also a prescription drug. Because opioidsare not contraindicated during pregnancy and can be used during labor, it is necessaryto try to identify all medications administered to the mother to properly assess opioidlevels.

Amphetamines

This group of drugs of abuse includes methamphetamine (commonly called crystalmeth), amphetamine (commonly called speed, also Adderall or Dexedrine), and

Delano & Koren1066

MDMA (3,4-methylenedioxy-N-methylamphetamine) (commonly called ecstasy). Casereports have reported congenital abnormalities, including heart defects and cleftpalate, in infants exposed to amphetamines during the first trimester.5 All of thesecompounds have been associated with similar impacts on behavior and cognitionof exposed children, including lower IQ scores, difficulties with advancement inschool, and physical fitness activities.32 In addition, children of addicted mothershave been found to show more behavioral problems.32 Each compound in this grouphas its ownmethod of detection, and results are used in conjunction to provide a moreaccurate interpretation.Methamphetamine is able to cross the placenta at a rapid rate and even at lower

levels on the fetal side; it persists because of slower elimination, resulting in prolongedfetal exposure.15,42 The fetal elimination rate of amphetamine is also reduced, but toa greater extent than that of methamphetamine.42 The prolonged elimination of bothmethamphetamine and amphetamine can cause accumulation of both compoundson the fetal side if the mother is using these compounds on a consistent basis. Envi-ronmental exposure is also a concern for children, especially if methamphetamine issmoked in the household.Each one of these amphetamine derivatives can be tested for in meconium and hair

samples to assess exposure in children. Interpretation of samples positive for amphet-amines can become complex depending on which compounds are detected. If eithermethamphetamine or amphetamine is detected alone, this indicates exposure or useof this compound. Samples positive for both methamphetamine and amphetaminecan be interpreted in 3 ways. First, the amphetamine could be a product of metham-phetamine metabolism, indicating methamphetamine use or exposure. Second, seizedillicit methamphetamine contains amphetamine as well, indicating methamphetamineuse and, unknowingly, amphetamine exposure. Third, it can suggest both metham-phetamine and amphetamine were used during the tested time frame. It is necessaryto test for all amphetamine compounds if exposure is suspected, because the interpre-tation of amphetamines can be complex.

Cannabinoids

No clear increase in pregnancy complications for users of marijuana is known.However, some studies have associated marijuana use during pregnancy with lowerbirth weights and longer gestations.5 Users of marijuana have been documented touse other illicit drugs, which increases the risk of complications. In addition, it wasrecently found that children exposed to marijuana had lower performance on stan-dardized tests, indicating that long-term behavioral and neurodevelopmental issuesmay occur in these children.43

Cannabis sativa, marijuana, produces the group of cannabinoids, with more than 50unique compounds. D9-tetrahydrocannabinol (THC) is the primary compound of thecannabinoids and is metabolized to the active compound 11-hydroxy-D9-tetrahydro-cannabinol and subsequently 11-nor-D9-tetrahydrocannabinol-9-carboxylic acid(THC-COOH).15 Analysis of samples for cannabinoid content is primarily conductedusing enzyme-linked immunosorbent assay (ELISA) techniques and yields a qualitativeresult of positive or negative depending on the cutoff value set for the method.Because of the commonality of infrequent cannabis users and their chance ofincreasing false-negative results, the sensitivity of the immunoassay analysis canvary depending on the tested population.15 Confirmation of ELISA results can be con-ducted using gas chromatography-mass spectrometry, but because of the nature ofTHC and the extraction methods used, which alter the ratio between THC-COOHand total cannabinoids, lower rates of confirmation have been found.15 Positive ELISA

Emerging Biomarkers 1067

results for cannabinoids indicate use or high exposure, because THC and its metab-olites are not readily incorporated into matrices. Implementing liquid chromatography-mass spectrometry analysis is a potential method for THC analysis, because it is ableto detect lower levels of compounds.

BIOMARKERS OF ENVIRONMENTAL EXPOSURES

Although there are numerous biomarkers for a multitude of environmental chemicalexposures, we use 1 powerful example of environmental exposure through dietprimarily to show the importance of biomarkers in this field.

Methylmercury

Methylmercury is a known neurotoxin produced from inorganic mercury by anaerobicorganisms that live in aquatic environments.44 Because methylmercury strongly bindsto cysteine-containing proteins, it is not readily eliminated from the body; its elimina-tion half-life is approximately 60 days, and it can also accumulate in the environ-ment.45 Fish consumption is the main source of methylmercury ingestion, witha dose-response relationship between the amount of fish consumed and the totalsystemic mercury burden, which can be measured in hair and blood.44 Methylmercuryreadily crosses the placenta and because the blood-brain barrier does not fullydevelop until 6 months after birth, prenatal exposure can result in adverse neurodeve-lopmental effects.45 These adverse neurodevelopment effects can include deficits inthe functional domains of language, memory, and attention.45 Adverse effects ofhigh prenatal exposure to methylmercury have been well documented because of 2serious poisoning events in Japan and Iraq.45 In addition, neurologic damage canalso occur in children and adults who consume fish containing high mercury levelson a regular basis. This damage is often characterized by ataxia, sensory distur-bances, and mental state changes.45

Total mercury levels are commonly tested in hair samples, both maternal and fetal.Data show that methylmercury makes up 80% of the total mercury detected in hair,with the remaining 20% being inorganic mercury.45 With respect to maternal hair,levels as low as 10 mg/g hair have been associated with severe adverse effects tothe fetus.44 In a recent meta-analysis, the lower observable adverse effect level ofmercury for adverse fetal outcome was defined at 0.3 mg/g maternal hair.46 Analysisof hair of women who consume more than 340g (12oz) of fish per week could bea useful tool to assess whether this consumption is associated with increasedmercurylevels. If so, these women may wish to modify their diet before becoming pregnant,decreasing the body’s burden of mercury and ensuring minimal fetal exposure tomethylmercury.44

IMPORTANCE OF BIOMARKERS

Substance abuse, prenatal or not, is an ongoing public health concern, affecting notonly the user themselves but also their families, and the health system economically.Beyond being biomarkers for physical and neurologic well-being of exposed children,hair and meconium measures identify women who continue to use drugs of abuse oralcohol despite being aware of their pregnancy. Children raised by an addicted motherhave increased risk for neglect and abuse.8 Behavioral problems associated withexposure to drugs of abuse can commence in utero, but the quality of the postnatalenvironment can also modify the child’s behavior and development.32 Hence, thesebiomarkers constitute strong predictors of environmental risk for the child, anda need for close follow-up of the well-being of the child. Identifying abuse is critical

Delano & Koren1068

to allow social workers and children’s aid societies to implement the necessary inter-ventions and provide the best environment possible for the child.Polydrug use is commonly reported in those who tested positive for certain drugs. In

terms of opioid-dependent women, benzodiazepines, cocaine, and marijuana are themost common concomitantly used drugs.36 This finding was confirmed by showingthat meconium samples positive for opioids were likely to also be positive for cocaine,benzodiazepines, methadone, and FAEEs in a previous study.37 As well, stimulant useof amphetamines and cocaine by the mother was a potential risk factor of alcohol useand subsequently fetal alcohol exposure.19 Knowing which drugs are often usedconcomitantly can allow clinicians to potentially diagnose drug exposures in childrenthat otherwise would have gone undetected.Environmental exposures to chemicals are increasingly becoming more publicly

acknowledged. Developing biomarkers for specific environmental toxicants canhelp clinicians detect both high-level and low-level exposures in humans. Detectioncan then lead to a decrease in exposure to these toxicants, and possibly discontinuingpublic use of compounds causing major adverse effects on health.

SUMMARY

The impact that drug and alcohol use or abuse during pregnancy can have onneonates and infants is serious. Because mothers may grossly underreport usebecause of fears and embarrassment, identification of biomarkers in newborns thatpoint to such use has become a major research focus and is becoming morecommonly accepted and used in clinical practice. Using such biomarkers providesclinicians with an opportunity to appropriately diagnose and, where possible, treatnewborn conditions associated with prenatal alcohol or drug abuse. It also providessocial workers with the opportunity to implement necessary interventions to providea safer environment for the child. Similarly, the development of effective biomarkersto detect environmental chemical exposures allows the health care team to imple-ment, where available, mitigating interventions to minimize their adverse effects.

REFERENCES

1. Biomarkers Definitions Working Group. Biomarkers and surrogate endpoints:preferred definitions and conceptual framework. Clin Pharmacol Ther 2001;69:89–95.

2. Goodsaid FM, Frueh FW, Mattes W. Strategic paths for biomarker qualification.Toxicology 2008;245:219–23.

3. Timbrell JA. Biomarkers in toxicology. Toxicology 1998;129:1–12.4. Goel N, Beasley D, Rajkumar V, et al. Perinatal outcome of illicit substance use in

pregnancy–comparative and contemporary socio-clinical profile in the UK. Eur JPediatr 2011;170:199–205.

5. Keegan J, Parva M, Finnegan M, et al. Addiction in pregnancy. J Addict Dis 2010;29:175–91.

6. Zelner I, Shor S, Gareri J, et al. Universal screening for prenatal alcohol exposure:a progress report of a pilot study in the region of Grey Bruce, Ontario. Ther DrugMonit 2010;32:305–10.

7. Klein J, Karaskov T, Koren G. Clinical applications of hair testing for drugs ofabuse–the Canadian experience. Forensic Sci Int 2000;107:281–8.

8. Koren G, Hutson J, Gareri J. Novel methods for the detection of drug and alcoholexposure during pregnancy: implications for maternal and child health. Clin Phar-macol Ther 2008;83:631–4.

Emerging Biomarkers 1069

9. Gareri J, Koren G. Prenatal hair development: implications for drug exposuredetermination. Forensic Sci Int 2010;196:27–31.

10. Kulaga V, Velazquez-Armenta Y, Aleksa K, et al. The effect of hair pigment on theincorporation of fatty acid ethyl esters (FAEE). Alcohol Alcohol 2009;44:287–92.

11. Harkey MR. Anatomy and physiology of hair. Forensic Sci Int 1993;63:9–18.12. Henderson GL. Mechanisms of drug incorporation into hair. Forensic Sci Int 1993;

63:19–29.13. Jurado C, Kintz P, Menendez M, et al. Influence of the cosmetic treatment of hair

on drug testing. Int J Legal Med 1997;110:159–63.14. Moller M, Koren G. Unsuspected prenatal opioid exposure: long-term detection

by alternative matrices. Ther Drug Monit 2010;32:1–2.15. Gareri J, Klein J, Koren G. Drugs of abuse testing in meconium. Clin Chim Acta

2006;366:101–11.16. SzetoHH.Kineticsofdrug transfer to the fetus.ClinObstetGynecol 1993;36:246–54.17. Gareri J, Lynn H, Handley M, et al. Prevalence of fetal ethanol exposure in

a regional population-based sample by meconium analysis of fatty acid ethylesters. Ther Drug Monit 2008;30:239–45.

18. Chan D, Caprara D, Blanchette P, et al. Recent developments in meconium andhair testing methods for the confirmation of gestational exposures to alcohol andtobacco smoke. Clin Biochem 2004;37:429–38.

19. Kulaga V, Shor S, Koren G. Correlation between drugs of abuse and alcohol byhair analysis: parents at risk for having children with fetal alcohol spectrumdisorder. Alcohol 2010;44:615–21.

20. Hutson JR, Rao C, Fulga N, et al. An improved method for rapidly quantifying fattyacid ethyl esters in meconium suitable for prenatal alcohol screening. Alcohol2011;45:193–9.

21. Zelner I, Shor S, Lynn H, et al. Clinical use of meconium fatty acid ethyl esters foridentifying children at risk for alcohol-related disabilities: the first reported case.J Popul Ther Clin Pharmacol 2012;19:e26–31.

22. Brien JF, Chan D, Green C, et al. Chronic prenatal ethanol exposure andincreased concentration of fatty acid ethyl esters in meconium of term fetalguinea pig. Ther Drug Monit 2006;28:345–50.

23. Caprara DL, Klein J, Koren G. Diagnosis of fetal alcohol spectrum disorder (FASD):fatty acid ethyl esters and neonatal hair analysis. Ann Ist Super Sanita 2006;42:39–45.

24. Chan D, Klein J, Karaskov T, et al. Fetal exposure to alcohol as evidenced by fattyacid ethyl esters in meconium in the absence of maternal drinking history in preg-nancy. Ther Drug Monit 2004;26:474–81.

25. Chan D, Bar-Oz B, Pellerin B, et al. Population baseline of meconium fatty acidethyl esters among infants of nondrinking women in Jerusalem and Toronto.Ther Drug Monit 2003;25:271–8.

26. Caprara DL, Klein J, Koren G. Baseline measures of fatty acid ethyl esters in hairof neonates born to abstaining or mild social drinking mothers. Ther Drug Monit2005;27:811–5.

27. Chan D, Klein J, Koren G. Validation of meconium fatty acid ethyl esters asbiomarkers for prenatal alcohol exposure. J Pediatr 2004;144:692.

28. Zelner I, Hutson JR, Kapur BM, et al. False-positive meconium test results for fattyacid ethyl esters secondary to delayed sample collection. Alcohol Clin Exp Res2012. [Epub ahead of print].

29. Gareri J, Appenzeller B, Walasek P, et al. Impact of hair-care products on FAEEhair concentrations in substance abuse monitoring. Anal Bioanal Chem 2011;400:183–8.

Delano & Koren1070

30. Kulaga V, Pragst F, Fulga N, et al. Hair analysis of fatty acid ethyl esters in thedetection of excessive drinking in the context of fetal alcohol spectrum disorders.Ther Drug Monit 2009;31:261–6.

31. Bandstra ES, Morrow CE, Mansoor E, et al. Prenatal drug exposure: infant andtoddler outcomes. J Addict Dis 2010;29:245–58.

32. Lester BM, Lagasse LL. Children of addicted women. J Addict Dis 2010;29:259–76.

33. Joya X, Gomez-Culebras M, Callejon A, et al. Cocaine use during pregnancy as-sessed by hair analysis in a Canary Islands cohort. BMC Pregnancy Childbirth2012;12:2.

34. Taguchi N, Mian M, Shouldice M, et al. Chronic cocaine exposure in a toddlerrevealed by hair test. Clin Pediatr (Phila) 2007;46:272–5.

35. Natekar A, Koren G. Interpretation of combined hair fatty acid ethyl esters,cocaine and cocaethylene. Ther Drug Monit 2011;33:284.

36. Abrahams R, Chase C, Desmoulin J, et al. The opioid dependent mother andnewborn–an update. The 6th Annual Ivey Symposium. J Popul Ther Clin Pharma-col 2012;19:e73–7.

37. Moller M, Karaskov T, Koren G. Opioid detection in maternal and neonatal hairand meconium: characterization of an at-risk population and implications to fetaltoxicology. Ther Drug Monit 2010;32:318–23.

38. Su PH, Chang YZ, Yang C, et al. Perinatal effects of combined use of heroin,methadone, and amphetamine during pregnancy and quantitative measurementof metabolites in hair. Pediatr Neonatol 2012;53:112–7.

39. Mactier H. The management of heroin misuse in pregnancy: time for a rethink?Arch Dis Child Fetal Neonatal Ed 2011;96:F457–60.

40. Lee S, Cordero R, Paterson S. Distribution of 6-monoacetylmorphine andmorphine in head and pubic hair from heroin-related deaths. Forensic Sci Int2009;183:74–7.

41. Kirchheiner J, Schmidt H, Tzvetkov M, et al. Pharmacokinetics of codeine and itsmetabolite morphine in ultra-rapid metabolizers due to CYP2D6 duplication.Pharmacogenomics J 2007;7:257–65.

42. White S, Laurenzana E, Hendrickson H, et al. Gestation time-dependent pharma-cokinetics of intravenous (1)-methamphetamine in rats. Drug Metab Dispos2011;39:1718–26.

43. Goldschmidt L, Richardson GA, Willford JA, et al. School achievement in 14-year-old youths prenatally exposed to marijuana. Neurotoxicol Teratol 2012;34:161–7.

44. Koren G, Bend JR. Fish consumption in pregnancy and fetal risks of methylmer-cury toxicity. Can Fam Physician 2010;56:1001–2.

45. Schoeman K, Bend JR, Koren G. Hair methylmercury: a new indication for thera-peutic monitoring. Ther Drug Monit 2010;32:289–93.

46. Schoeman K, Bend JR, Hill J, et al. Defining a lowest observable adverse effecthair concentrations of mercury for neurodevelopmental effects of prenatal meth-ylmercury exposure through maternal fish consumption: a systematic review.Ther Drug Monit 2009;31:670–82.