NTP-CERHR Expert Panel Report on the Reproductive and Developmental Toxicity of Amphetamine and Methamphetamine

Published 2005 Wiley-Liss, Inc.w Birth Defects Research (Part B) 74:471–584 (2005)

NTP-CERHR Expert Panel Report on the Reproductive and Developmental Toxicity Of Amphetamine and Methamphetamine

Mari Golub,1 Lucio Costa,2 Kevin Crofton,3 Deborah Frank,4 Peter Fried,5 Beth Gladen6 Rogene Henderson,7 Erica Liebelt,8 Shari Lusskin,9 Sue Marty,10 Andrew Rowland11 John Scialli12 and Mary Vore13 1California Environment Protection Agency, Sacramento, California 2University of Washington, Seattle, Washington 3U.S. Environmental Protection Agency, Research Triangle Park, North Carolina 4Boston Medical Center, Boston, Massachusetts 5Carleton University, Ottawa, Ontario 6National Institute of Environmental Health Sciences, Research Triangle Park, North Carolina 7Lovelace Respiratory Research Institute, Albuquerque, New Mexico 8University of Alabama at Birmingham School of Medicine, Birmingham, Alabama 9New York University School of Medicine, New York, New York 10The Dow Chemical Company, Midland, Michigan 11University of New Mexico, Albuquerque, New Mexico 12Phoenix, Arizona 13University of Kentucky, Lexington, Kentucky

PREFACE studies indexed before December 31, 2004. References were also identified from databases such as REPRO- The National Toxicology Program (NTP) and the TOXs, HSDB, IRIS, and DART and from report National Institute of Environmental Health Sciences bibliographies. (NIEHS) established the NTP Center for the Evaluation This evaluation resulted from the efforts of a 13- of Risks to Human Reproduction (CERHR) in June 1998. member panel of government and non-government The purpose of the Center is to provide timely, unbiased, scientists that culminated in a public expert panel scientifically sound evaluations of human and experi- meeting held January 10–12, 2005. This report is a mental evidence for adverse effects on reproduction and product of the Expert Panel and is intended to (1) development caused by agents to which humans may be interpret the strength of scientific evidence that exposed. Amphetamines were selected for expert panel evaluation because of widespread usage in children, availability A report of the CERHR Expert Panel with the support of CERHR Staff: of studies on developmental effects in children and NTP/NIEHS, Michael Shelby, Ph.D. (Director, CERHR), Christopher Portier, Ph.D. (Associate Director, National Toxicology Program); Sciences experimental animals, and public concern about the International, Inc., Anthony Scialli, M.D. (Principal Scientist), Annette effects of these stimulants on child development. Iannucci, M.S. (Toxicologist), Gloria Jahnke, D.V.M. (Toxicologist), Jessie Amphetamines evaluated were d- and d,l-amphetamine Poulin, B.A. (Associate). and methamphetamine. d- and d,l-Amphetamine are This report is prepared according to the Guidelines for CERHR Panel Members established by NTP/NIEHS. The guidelines are available on the approved by the Food and Drug Administration (FDA) CERHR web site (http://cerhr.niehs.nih.gov/). The format for Expert Panel for the treatment of attention deficit hyperactivity Reports includes synopses of studies reviewed, followed by an evaluation disorder (ADHD) and narcolepsy. d-Methamphetamine of the Strengths/Weaknesses and Utility (Adequacy) of the study for CERHR evaluation. Statements and conclusions made under Strengths/ hydrochloride is used in pharmaceutical preparations in Weaknesses and Utility evaluations are those of the Expert Panel and are the United States and is approved for the treatment of prepared according to the NTP/NIEHS guidelines. In addition, the Panel ADHD and for short-term treatment of obesity. Metham- often makes comments or notes limitations in the synopses of the study. phetamine is also manufactured and used as an illicit Bold, square brackets are used to enclose such statements. As discussed in the guidelines, square brackets are used to enclose key items of drug. information not provided in a publication, limitations noted in the study, To obtain information about amphetamines for conclusions that differ from those of the authors, and conversions or the CERHR evaluation, the PubMed (Medline) and analyses of data conducted by the Panel. Toxline databases were searched with CAS RNs for Correspondence to: Michael D. Shelby, Ph.D., NIEHS EC-32, PO Box 12233, Research Triangle Park, NC 27709. 919-541-3455 d- and l-amphetamine (51-64-9; 156-34-3) and d-metham- E-mail: [email protected] phetamine (537-46-2) and its hydrochloride (51-57-0), Published online in Wiley InterScience (www.interscience.wiley.com) and relevant keywords. The search was limited to DOI: 10.1002/bdrb.20048 w This article is a U.S. Government work and, as such, is in the public domain in the United States of America. 472 GOLUB ET AL. amphetamines are reproductive or developmental tox- meant by the term ‘‘speed.’’ l-Methamphetamine can icants based on data from in vitro, animal, or human produce palpitations and gastrointestinal upset but does studies, (2) assess the extent of human exposures to not produce the psychological effects desired by recrea- include the general public, occupational groups, and tional users. There are no pharmaceuticals in the U.S. that other sub-populations, (3) provide objective and thor- contain the l-enantiomer. ough assessments of the scientific evidence that adverse Many of the Hazardous Substance Data Bank (HSDB) reproductive/developmental health effects may be asso- proprietary names in Table 1 were not found on the FDA ciated with such exposures, and (4) identify knowledge web site (http://www.accessdata.fda.gov/scripts/cder/ gaps to help establish research and testing priorities to drugsatfda/) and were presumed to be discontinued or reduce uncertainties and increase confidence in future foreign. Some proprietary names that were found on the assessments of risk. This report has been reviewed by FDA web site are for products listed as discontinued. The CERHR staff scientists, and by members of the Amphe- products that are currently marketed in the U.S. are listed tamines and Methylphenidate Expert Panel. Copies have in Table 2. been provided to the CERHR Core Committee, which is 1.1.2. Formula and molecular mass. The chemi- made up of representatives of NTP-participating cal formula for amphetamine is C9H13N and the agencies. molecular mass is 135.20. The chemical formula for This Expert Panel Report will be a central part of methamphetamine is C10H15N and the molecular mass is the subsequent NTP-CERHR Monograph on the 149.24. The structures are shown in Figure 1. Potential Human Reproductive and Developmental 1.1.3 Chemical and physical properties. Che- Effects of Amphetamines. This monograph will include mical and physical properties of amphetamine and the NTP-CERHR Brief, the Expert Panel Report, and all methamphetamine are listed in Table 3. public comments on the Expert Panel Report. The NTP- 1.1.4 Technical products and impurities. Table 2 CERHR Monograph will be made publicly available and summarizes active ingredient strength and lists the transmitted to appropriate health and regulatory inactive ingredients in each marketed amphetamine agencies. and methamphetamine product. The ophthalmic solu- The NTP-CERHR is headquartered at NIEHS, Re- tion (Paremyds) is marketed as a mydriatic and will not search Triangle Park, NC and is staffed and administered be further considered in this report. by scientists and support personnel at NIEHS and at Illicit amphetamines, chiefly methamphetamine hy- Sciences International, Inc., Alexandria, Virginia. drochloride, can be synthesized by different methods (Section 1.2.1) with the potential for different contaminants. According to the U.S. Drug Enforcement Admin- 1.0 CHEMISTRY, USE AND istration (DEA) (2004), chemical supply houses HUMAN EXPOSURE internationally have restricted sales of the chemicals This exposure section is initially based on secondary used to produce methamphetamine of high purity, review sources. Primary study reports are addressed by resulting in substitution of other chemicals and a the Expert Panel if they contain information that is highly decrease in methamphetamine purity. In 1994, the relevant to a CERHR evaluation of developmental or average purity of methamphetamine seized by DEA reproductive toxicity or if the studies were released was 71.9%, whereas in 1999, the average purity was subsequent to the reviews. 30.7%. Purity of seized methamphetamine increased thereafter to 35.3% in 2000 and 40.1% in 2001. The nature of the impurities was not discussed. 1.1 Chemistry 1.1.1 Nomenclature. The term ‘‘amphetamines’’ is used to denote a class of chemicals with structural 1.2 Use and Human Exposure similarity to amphetamine. The amphetamines used in 1.2.1 Production information. The methods of clinical practice include two distinct bases, amphetamine production used in the pharmaceutical manufacture of and methamphetamine, available in pharmaceutical amphetamine and methamphetamine are not available; preparations as various mixtures of enantiomers and as however, there are Internet sites that give a number of various salts. The compounds relevant to this report are different methods for the synthesis of these compounds. identified in Table 1. Many of the trade names are no Amphetamine can be synthesized by the sequential longer in use, although they remain in current lists of alkylation of methyl acetoacetate with dimethyl sulfate drug names (ChemIDplus, 2004a; HSDB, 2004). The and benzyl chloride, followed by hydrolysis and deace- most commonly encountered proprietary amphetamine tylation to give 2-phenylpropionic acid, which, through preparation is Adderalls (a mixture of amphetamine reaction with thionyl chloride and ammonia, forms 2- salts providing d-, and l-amphetamine in a 3:1 ratio). phenylpropionamide. Upon treatment with aqueous Generic equivalents of Adderall are also available. sodium hypochlorite, this amide undergoes Hofmann Generic d-amphetamine is often called dextroampheta- rearrangement to form racemic amphetamine (phenyl-2- mine and l-amphetamine is called levamfetamine [spel- aminopropane) (Rhodium, 2004). Methamphetamine can ling differences as per ChemID]. In this study, be synthesized from ephedrine via reduction of chlor- the enantiomers will be designated by the d-orl- oephedrine with hypophosphorous acid or by Birch prefixes rather than by dextroamphetamine and levam- reduction of pseudoephedrine. Pseudoephedrine is read- fetamine. Methamphetamine in U.S. pharmaceutical ily available in decongestant tablets and is the most preparations is present as d-methamphetamine hydro- common starting material for illicit methamphetamine chloride. d-Methamphetamine hydrochloride is also used (DEA, 2004). The chemical synthesis information on recreationally and is the illicit stimulant most commonly many web sites is interspersed with advice on avoiding

Birth Defects Research (Part B) 74:471–584, 2005 DOI 10.1002/bdrb AMPHETAMINE AND METHAMPHETAMINE 473 Table 1 Amphetamine Nomenclature

Names and synonyms CAS RN Trade names Street names

Amphetamine 300-62-9 Adderalla,b Isoamyne Phenedrine Speedc (1-)-Benzedrine Delcobesea,b Isomyn Profamina Bennies (1-)-a-Methylbenzeneethanamine Actedron Mecodrin Propisamine (1-)-a-Methylphenethylamine Adipan Mydrial Protioamphetamine (1-)-a-Methylphenylethylamine Allodene Norephedrane Psychedrine (Phenylisopropyl)amine Anorexide Novydrine Raphetamine Methyl-2-phenylethylamine Anorexine Obesin Rhinalator 1-Phenyl-2-aminopropane Benzebar Obesine Simpatedrin 1-Phenyl-2-aminopropane Benzedrinea Oktedrin Simpatina 1-Phenyl-2-propylamine Benzolone Ortedrine Sympamine 2-Amino-1-phenylpropane Elastonon Percomon Sympatedrine 3-Phenyl-2-propylamine Fenopromin Phenamine Weckamine Amfetamine Finam Desoxynorephedrine Amphetamine phosphate 139-10-6 Amphetamine sulfate 60-13-9 d-Amphetamine 51-64-9 Dexedrinea Dexies Dextroamphetamine Amsustain Dexamphetamine Dephadrin (1)-(S)-Amphetamine Sympamin (1)-Amphetamine Dextrostata (1)-Phenaminum Ferndexa (1)-a-Methylphenethylamine (1)-a-Methylphenylethylamine (2S)-(1)-Amphetamine (S)-(1)-Amphetamine (S)-(1)-b-Phenylisopropylamine (S)-1-Phenyl-2-aminopropane (S)-1-Phenyl-2-propylamine (S)-Amphetamine (S)-a-Methylphenethylamine (S)-a-Phenylethylamine d-(S)-Amphetamine d-1-Phenyl-2-aminopropane d-2-Amino-1-phenylpropane d-AM a-Methylphenethylamine, d-form d-(S)-Amphetamine d-1-Phenyl-2-aminopropane d-2-Amino-1-phenylpropane d-a-Methylphenethylamine d-Amphetamine sulfate 51-63-8 Acedron DAS Ephadren Dexies Dextroamphetamine sulfate Adjudets Dadex Eskatrol Fastballs Dextro-1-phenyl-2-amino- Adrizine Dams Evrodex Oranges propane sulfate Dexamphetamine sulfate Afatin Dasdel Hetamine Dextro-a-methylphenethylamine sulfate Albemap Dellipsoids Lentanet Dextro-b-phenylisopropylamine sulfate Algo-dex Dex ob Lipsoids Amdex Dex-Sule Lowedex Amphaetex Dexaline Maxiton sulfate Ampherex Dexalme Medex Amphetasul Dexalone Obesedrin Amphex Dexamed Obesonil Amptrerex Dexamine Pellcap Amsustain Dexamyl Perke Apetain Dexten Phetadex Ardex Dextenal Pomadex Betafedrine Dextroprofetamine Pro-Dexter Betaphedrine Dextrosule Psychodrine Carrtime Diocurb Recordati Cradex Diphylets Revidex Curban Ditab Robese D-Amfetasul Domafate Tempodex D-Benzedrine Dura Dex Tuphetamine

Birth Defects Research (Part B) 74:471–584, 2005 DOI 10.1002/bdrb 474 GOLUB ET AL. Table 1 (Continued)

Names and synonyms CAS RN Trade names Street names

Dynaphenyl Tydex Zamine d-Amphetamine tartrate 3994-11-4 Maxiton d-Amphetamine bitartrate l-Amphetamine 156-34-3 Levamphetamine (À)-Amphetamine (À)-Phenylisopropylamine (À)-a-Methylphenethylamine I-Amphetamine I-a-Methylbenzeneethanamine (R)-a-Methylphenethylamine l-a-Methylphenethylamine l-Amphetamine succinate 5634-40-2 Cydril Methamphetamine 537-46-2 Norodin (1)-N, a-Dimethyl-b-phenylethylamine Stimulex (1)-N, a-Dimethylphenethylamine 1-Phenyl-2-methylaminopropane Deoxyephedrine Desoxyephedrine Desyphed N-Methyl-b-phenylisopropylamine d-Methamphetamine hydrochloride 51-57-0 Adipex Dexoval Isophen Ice Wet (1)-Methylamphetamine hydrochloride Chestox Dexstim Norodin Crystal White cross (1)-N, a-Dimethylphenethylamine Deofed Dosoxy Hydrochloride Glass Wolminic hydrochloride Desepin Doxyfed Obedrin-LA Crank Nasal spray Methedrine hydrochloride Desodex Drinalfa Pervitin Meth Yellow bam Methylisomin Desoxyna Efroxine Philopon Chalk Yellow powder Semoxydrine hydrochloride Destim Eufodrinal Soxysympamine Beenies Batu N, a-dimethylbenzeneethanamine Desyfed Syndrox Blue mollies Cristy hydrochloride Tonedron Crink Hanyak Cris Hiropon Croak Hot ice Crossies Kaksonjae Crypto LA glass Desocsins LA ice Desogtion Quartz Fire Super ice Go-fast Lemon drop Granulated Soap dope orange Methlies quik Grimace Mexican crack Green monster Peanut butter Sketch Powder Stove top Quill Water Rose Shabu Speed Yaba Methamphetamine hydrochloride 300-42-5 Amdram Amedrine Normadrine (racemate) (1-)-Methylamphetamine hydrochloride Amphedroxy Fenyprin Norodrine Deoxyephedrine Amphedroxyn Kemodrin Obesin Desoxyephedrine hydrochloride Bombita Lanazine Oxydess Methylpropamine Corvitin Levetamin Oxydrene N, a-Deimethylphenethylamine Daropervamin Madrine Oxydrin hydrochloride Depoxin Mepho-D Oxyfed d,l-Desoxyephedrine hydrochloride Desamine Methampexa Phedoxe Desfedran Methamphin Phedrisox Desfedrin Methedrinal Premodrin Desoxedrin Methoxyn Psichergina Detrex Methylbenzedrin Psicopan Dexophrine Millerdrine Psychergine

Birth Defects Research (Part B) 74:471–584, 2005 DOI 10.1002/bdrb AMPHETAMINE AND METHAMPHETAMINE 475 Table 1 (Continued)

Names and synonyms CAS RN Trade names Street names

Dopidrin Neodrine Psykoton Doxephrin Neopharmadrine Semoxydrine Estimulex Noradrin Amphetamine/dextroamphetamine resin None Biphetaminea complex Hydroxyamphetamine hydrobromide 306-21-8 Paredrinea Paremyda

Registration signs omitted from trade names. From ChemIDplus, 2004b; HSDB, 2002. aIdentified in Drugs@FDA (FDA, 2004) as current US trade names (but not necessarily marketed). bAdderall and Delcobese are 3:1 mixtures of d- and l-enantiomers containing a fixed ratio (1:1:1:1) of amphetamine aspartate, amphetamine sulfate, dextroamphetamine saccharate, and dextroamphetamine sulfate. Delcobese is no longer marketed. cThe term ‘‘speed’’ is used for any stimulant.

explosion, arrest, and exploitation by professional crim- pharmacies and 120 kg were sold to hospitals. By inals, suggesting that these sites are intended for illicit contrast, fewer than 17 kg of methamphetamine were manufacturers. sold to pharmacies and hospitals combined. [The Expert Retail U.S. distribution of amphetamines for calendar Panel recognizes that therapeutic use of d-methamphe- year 2002 is shown in Table 4. The United Nations tamine in the U.S. is uncommon.] The National Institute reported that U.S. manufacture of amphetamine [as- on Drug Abuse (NIDA, 2003) states that addiction to sumed to be d,l-amphetamine] was 18,586 kg in 2000, stimulant medications does not occur when medicines 9612 kg in 2001, and 7442 kg in 2002; U.S. manufacture of are taken in the form and dosage prescribed. Ampheta- d-amphetamine was 12,306 kg in 2000, 4919 kg in 2001, mines and d-methamphetamine, however, are used and 5962 kg in 2002; U.S. manufacture of methampheta- recreationally. mine was 1306 kg in 2000, 1692 kg in 2001, and 1385 kg in The DEA reported that during the year 2000, 4% of 2002 (UN, 2002). Reported exports in 2002 were 9 kg the U.S. population reported trying methamphetamine amphetamine and 152 kg d-amphetamine; no value was at least once in their lives (DEA, 2004). Illicit use is given for methamphetamine. The DEA reported that concentrated in the Midwest, Southwest, and Pacific 8000 clandestine methamphetamine laboratories were coast regions of the country. The street price of seized in 2001, and that 298 were so-called ‘‘super labs,’’ methamphetamine ranges from $400–$3000 per ounce capable of making 410 pounds (4.5 kg) of methamphe- (138 g). Methamphetamine hydrochloride is used recrea- tamine in a 24-hr period (DEA, 2004). Most of these super tionally as a powder by nasal inhalation or is mixed with labs were in northern Mexico and they were believed to water and injected intravenously (i.v.). The pure crystal- be supplying the U.S. In 2001, 1370 kg of methampheta- line form of methamphetamine, called ice, glass, or mine was seized at the Mexico–U.S. border. crystal, is typically smoked. A pill form called Yaba is 1.2.2 Use. Amphetamine and methamphetamine made in Thailand and smuggled into the U.S.; the are central nervous system (CNS) stimulants. The methamphetamine content is 30–40 mg per pill (DEA, amphetamine preparations are indicated for the treat- 2004). ment of narcolepsy and attention/deficit-hyperactivity 1.2.3 Human exposure. The recommended start- disorder (ADHD) (PDR, 2003a,d) and methamphetamine ing doses of amphetamine for narcolepsy is 5 mg/day for is indicated for the treatment of ADHD and the short- children 6–12 years old and 10 mg for children older than term treatment of obesity (PDR, 2003c). The Expert Panel 12 years and for adults. The maximum recommended is aware of off-label uses of amphetamines to treat dose is 60 mg/day. For ADHD, the starting dose of depression, primarily as an adjunct to antidepressant amphetamine is 2.5 mg/day for children 3–6 years old medication, and to treat patients with post-stroke and 5 mg/day for children older than 6 years and for cognitive impairment (Scialli JV and Lusskin S, personal adults. The maximum dose recommended for ADHD is communication, September 22, 2004). Although depres- 40 mg/day. Amphetamine preparations are not recom- sion is common in men and women of reproductive age, mended for children younger than 3 years. Ampheta- strokes most often occur in older individuals. There is an mines are taken every 4–6 hr or, for the sustained-release increase in diagnosis and treatment of ADHD and preparations, once/day (PDR, 2003a,b,d). The recom- depression in adolescents and adults. Obesity is also mended starting dose for methamphetamine treatment common in individuals of reproductive age. More of ADHD is 5 mg/day in individuals who are at exposures in people of reproductive age can, therefore, least 6 years. The maximum recommended dose for be expected. ADHD is 25 mg/day. For obesity, the recommended The DEA estimated that the number of amphetamine methamphetamine dose is 5 mg before a meal. Metham- prescriptions written increased between 1992 and phetamine is not recommended for the treatment of 2000 from fewer than 500,000 to nearly 8 million per ADHD in children younger than 6 years old and is not year (DEA, 2004). In calendar year 2001, more than recommended for obesity treatment in children younger 4000 kg of racemic and d-amphetamine were sold to than 12 years (PDR, 2003c).

Birth Defects Research (Part B) 74:471–584, 2005 DOI 10.1002/bdrb 476 GOLUB ET AL. Table 2 Amphetamine Preparations Marketed in the U.S.

Brand name (reference)a Manufacturer Active ingredients Inactive ingredients Generic manufacturers

Adderall Shire Fixed weight ratio (1:1:1:1) of Sucrose, lactose, cornstarch, Abrika Pharmaceuticals, (PDR, 2003a) amphetamine aspartate, acacia, magnesium stea- Barr, Corepharma, amphetamine sulfate, rate, colors. Eon, Mallinckrodt, dextroamphetamine Mutual Pharmaceuti- saccharate, and dextroam- cal, Watson phetamine sulfate; total pill Laboratories doses of 5, 7.5, 10, 12.5, 15, 20, or 30 mg (equivalent to 3.13 mg free based per 5 mg total pill dose). Adderall XR Shire Extended release preparation Gelatin capsules, hydroxypro- None (PDR, 2003b) of amphetamine aspartate, pyl methylcellulose, amphetamine sulfate, methacrylic acid dextroamphetamine copolymer, Opadry beige, saccharate, and dextroam- sugar spheres, talc, triethyl phetamine sulfate; total pill citrate, colors. doses of 10, 20, or 30 mg (equivalent to 6.3 mg free based per 10 mg total pill dose). Desoxyn Abbott Methamphetamine hydro- Corn starch, lactose, sodium Able (PDR, 2003c) chloride 5 mg. paraaminobenzoate, stearic acid, talc. Dexedrine GlaxoSmithKline d-Amphetamine sulfate 5 mg Calcium sulfate, gelatin, lac- Barr, Malinkrodt (PDR, 2003d) tablets. tose, mineral oil, starch, stearic acid, sucrose, talc, colors. Dexedrine Spansule GlaxoSmithKline d-Amphetamine sulfate sus- Cetyl alcohol, dibutyl Barr, Malinkrodt (PDR, 2003d) tained-release capsule 5, 10, sebacate, ethylcellulose, or 15 mg. gelatin, hydroxypropyl methylcellulose, propylene glycol, povidone, silicon dioxide, sodium lauryl sulfate, sugar spheres, colors. DextroStat Shire d-Amphetamine sulfate 5 or Acacia, cornstarch, lactose Barr, Endo Pharmaceuti- (PDR, 2003e) 10 mg. monohydrate, magnesium cals, KV Pharmaceuti- stearate, sucrose, sodium cals, Malinkrodt starch glycolate (10 mg only). Paremyd Akorn Hydroxyamphetamine hydro- Benzalkonium chloride, ede- None (Akorn, 2001) bromide, USP 1.0%; Tropi- tate disodium, sodium camide, USP 0.25% chloride, purified water. ophthalmic solution. Hydrochloric acid and/or sodium hydroxide are added to adjust the pH. aRegistration signs omitted from trade names.

CH3 CH3

NH2 NHCH3

Amphetamine Methamphetamine

Fig. 1. Amphetamine and methamphetamine structures.

Birth Defects Research (Part B) 74:471–584, 2005 DOI 10.1002/bdrb AMPHETAMINE AND METHAMPHETAMINE 477 Table 3 Chemical and Physical Properties of Amphetamine and Methamphetamine

Property d,l-Amphetamine d-Amphetamine d-Methamphetamine

Form Colorless liquid; the salts are Colorless liquid; the salts are Colorless liquid; the hydrochloride is a white powders or crystals. white powders or crystals. white powder or clear crystal. Boiling point 200–2031C 203–2041C 2121C Density 0.913 0.949 Not located pKa 10.13 Not located 9.9 log Kow 1.76 1.76 2.07 Solubility Slight in water; soluble in diethyl ether and ethanol. Amphetamine 0.5 g/mL water, soluble in ethanol and sulfate is insoluble in ether. diethyl ether; the hydrochloride is readily soluble in water. Source: HSDB (2004).

Table 4 appliances used to manufacture methamphetamine (U.S. Retail U.S. Distribution of Amphetamines in 2002 Department of Justice, 2002). In addition to toxicity risks, children at methamphetamine production sites face risks Drug (as base) Amount (kg) of fire or explosion and neglect or abuse associated with d,l-Amphetamine 2096 the methamphetamine lifestyle (U.S. Department of d-Amphetamine 3097 Justice, 2003). In 14,260 methamphetamine lab incidents d-Methamphetamine 17 in 2003, children were present at 1442 incidents, 3419 children were affected, 1291 children were exposed to From DEA (2004). toxic chemicals, 44 children were injured, and three children were killed (White House Office of National Drug Control Policy, 2004). Methamphetamine was Illicit methamphetamine use involves doses in the detected in urine of one-third to one-half of tested drug-naı¨ve individual of about 30 mg; however, habitual children found at methamphetamine labs in Oregon use of methamphetamine to produce euphoria character- (Oregon Department of Human Services, 2003). istically results in binges during which all available Clandestine laboratories have developed so-called methamphetamine is used over a period of 3–15 days designer amphetamines that contain modifications of (Alcohol and Drug Information Clearing House, 1998). the amphetamine or methamphetamine molecular struc- Cho et al. (2001) reported a dose range of 20–250 mg or ture to produce novel stimulant drugs. One of the most more per ‘‘hit’’ in methamphetamine abusers with total popular of these new chemicals is methylenedioxy- daily doses of up to several grams. methamphetamine (MDMA; Ecstasy). These novel sti- An abstract from the University of Utah reported that mulants can be included in the class designation 0.2% of babies in a well-baby nursery and 1% of babies in amphetamines, but are not used in therapeutics and will a neonatal intensive care nursery have meconium not be discussed in detail in this report. samples positive for methamphetamine, presumably representing recent use during pregnancy by their mothers (Zone and Buchi, 2002). [This information was 1.3 Utility of Data presented to document prenatal exposures but will not There are reliable data from the DEA on the amount of be considered further since it is only an abstract.] A medicinal amphetamines available in the U.S. Estimates report from the University of California Davis Medical of the amount of amphetamines used for different Center indicated that at least 6% of pregnant women who indications by different groups of patients (children with were routinely screened for illicit drugs in urine (most of ADHD and reproductive-age women with ADHD or whom were presenting in labor) were positive for depression) are not readily available. Information on amphetamines (Gillogley et al., 1990). illicit use of amphetamines, including populations using Numerous children in the U.S. are exposed to these drugs and the amount of drug used, seems to be methamphetamine or other toxic chemicals at clandes- approximate and of uncertain reliability. tine laboratories. Toxic or hazardous chemicals used in methamphetamine production include solvents, caus- tics/irritants, and metals/salts (Oregon Department of 1.4 Summary of Human Exposure Data Human Services, 2003). Children at clandestine labs can The term ‘‘amphetamines’’ denotes a class of com- be exposed to methamphetamine through inhalation of pounds with structural similarity to amphetamine. The second hand smoke and can be exposed to both focus of this report is the amphetamines used in clinical methamphetamine and other chemicals through vapors practice, amphetamine and methamphetamine. Amphe- generated in the production process (U.S. Department of tamine is available as salts of the d- and l-enantiomers in Justice, 2003). Additional exposures can occur through a 3:1 ratio or as a salt of the d-enantiomer. Pharmaceutical dermal contact with contaminated surfaces or clothing. methamphetamine is available as a salt of the d- Oral exposure to methamphetamine or other chemicals is enantiomer. The d- and d,l-amphetamine preparations possible through ingestion of contaminated foods or are indicated for the treatment of ADHD and narcolepsy drinks that are often prepared with the same utensils and (PDR, 2003a,d), and methamphetamine is indicated for

Birth Defects Research (Part B) 74:471–584, 2005 DOI 10.1002/bdrb 478 GOLUB ET AL. the treatment of ADHD and the short-term treatment of admitted trying methamphetamine at least once in their obesity (PDR, 2003c). d-Methamphetamine hydrochlor- lives (DEA, 2004). Illicit use is concentrated in the ide is also used recreationally and is the illicit stimulant Midwest, Southwest, and Pacific coast regions of the most commonly meant by the term ‘‘speed.’’ It is country. Methamphetamine is used recreationally as a believed that all human exposures occur through powder by nasal insufflation or is mixed with water and medication use and drug abuse. No information was injected i.v. The pure crystalline form of methampheta- identified on possible environmental or occupational mine, called ice or crystal, is typically smoked. A pill exposure. form called Yaba is made in Thailand and smuggled into Recommended doses of amphetamine are 2.5–40 mg/ the U.S.; the methamphetamine content is 30–40 mg per day for treatment of ADHD in individuals 3 years of pill (DEA, 2004). age and older and 5–60 mg/day for treatment of narcolepsy in individuals 6 years of age and older. Amphetamine preparations are not recommended 2.0 GENERAL TOXICOLOGY AND for children younger than 3 years. Immediate-acting BIOLOGIC EFFECTS amphetamines are taken every 4–6 hr and sustained- release preparations once/day (PDR, 2003a,b,d). Recom- 2.1 Pharmacodynamics and Pharmacokinetics mended methamphetamine doses are 5–25 mg/day for Consistent with Section 1, Section 2 will present treatment of ADHD in individuals who are 6 years information for the amphetamine class of drugs that old or older, and 5 mg before meals for treatment of is marketed as salts of d,l-amphetamine (Adderall), obesity in individuals 12 years old and older. Metham- d-amphetamine (DextroStat or Dexedrine), and methamphetamine is not recommended for the treatment of phetamine (Desoxyn). ADHD in children younger than 6 years and is not Information in Section 2 is initially based upon recommended for obesity treatment in children younger reviews. Primary studies were reviewed if information than 12 years (PDR, 2003c). Cho et al. (2001) reported a in reviews was inadequate, if the information presented dose range of 20–250 mg or more per ‘‘hit’’ in metham- in the primary studies is highly relevant for the phetamine abusers with total daily doses of up to several evaluation of developmental or reproductive effects, or grams. if the studies were published subsequent to reviews. The DEA estimated that the number of amphetamine Some examples of highly relevant sources reviewed in prescriptions written increased between 1992 and 2000 detail are studies reporting amphetamine levels in breast from fewer than 500,000 to nearly 8 million per year milk or premature infants, studies examining species- (DEA, 2004). In calendar year 2001, more than 4000 kg of related differences, or studies comparing dose- or route- racemic and d-amphetamine were sold to pharmacies related differences in pharmacokinetics. and 120 kg were sold to hospitals. By contrast, fewer than 2.1.1 Human 17 kg of methamphetamine were sold to pharmacies and 2.1.1.1 Pharmacodynamics. According to NIDA hospitals combined. The DEA (2004) reported retail U.S. (2003), amphetamine and methamphetamine are distribution of 2096 kg d,l-amphetamine, 3097 kg d- closely related, but methamphetamine has greater CNS amphetamine, and 17 kg d-methamphetamine in 2002. effects. [The Expert Panel recognizes that therapeutic use of Amphetamines cross the blood-brain barrier and the d-methamphetamine in the U.S. is uncommon.] The major site of pharmacological action is the brain (NTP, United Nations reported that 7442 kg amphetamine 1991). Amphetamine and methamphetamine stimulate [assumed to be d,l-amphetamine], 5962 kg d-ampheta- the CNS by acting as sympathomimetic drugs (Kraemer mine, and 1385 kg methamphetamine were manufac- and Maurer, 2002; PDR, 2003a,c). The therapeutic mode tured in the U.S. in 2002 (UN, 2002). According to the of action in treatment of ADHD with amphetamine or DEA, 8000 clandestine methamphetamine laboratories, methamphetamine is not known (PDR, 2003a,c). Am- believed to be supplying the U.S., were seized in 2001 phetamine and methamphetamine are substrates of the (DEA, 2004). monoamine transporter found in cell and storage vesicle Amphetamines and d-methamphetamine are approved membranes of dopaminergic, noradrenergic, and adre- for the treatment of ADHD or narcolepsy and nergic neurons (Kraemer and Maurer, 2002). Ampheta- d-methamphetamine is approved for treatment of obe- mine is believed to block the reuptake of norepinephrine sity. The Expert Panel is aware of the off-label use of and dopamine by presynaptic neurons (PDR, 2003b). It is amphetamines to treat depression primarily as an also thought that the transporter translocates ampheta- adjunct to antidepressant medication and to treat mines into dopaminergic neurons where the drug patients with post-stroke cognitive impairment. Inas- increases the release of dopamine (Markowitz and much as depression, ADHD, and obesity are common in Patrick, 2001) and norepinephrine (Kita et al., 2003) into men and women of reproductive age, there is a potential the extraneuronal space. Amphetamine may also inhibit for amphetamine and d-methamphetamine exposure in monoamine oxidase (Solanto, 2000; Kita et al., 2003). The that population. Diagnosis and treatment of ADHD in actions of amphetamines increase the level of catechola- adults is increasing. There is no information on the mines in the synaptic space and overall catecholaminer- numbers of pregnant or lactating women prescribed gic activity of the brain (Kita et al., 2003). A review by these drugs. Kraemer and Maurer (2002) reported that amphetamines NIDA (2003) states that addiction to stimulant medica- ‘‘yhave no affinity to the adrenoceptors or dopamine tions does not occur when medicines are taken in receptors.’’ the form and dosage prescribed. Amphetamines and There is also evidence that amphetamines increase d-methamphetamine are used recreationally. The DEA release and turnover of serotonin and it has been reported that during 2000, 4% of the U.S. population suggested that many of the behavioral effects of

Birth Defects Research (Part B) 74:471–584, 2005 DOI 10.1002/bdrb AMPHETAMINE AND METHAMPHETAMINE 479 amphetamines are mediated through serotonin (Kita (Adderall IR) and sustained-release capsules (Adderall et al., 2003). XR). Both products contain a 3:1 ratio of d- and Although stimulants decrease locomotor activity in l-amphetamine salts. The sustained-released capsules children, an increase in activity is observed in experi- contain a mixture of immediate-acting pellets and mental animal studies. A review by Solanto (2000) enteric-coated delayed-released pellets. Plasma levels of discussed possible reasons for discordance between d- and l-amphetamine peak at B3 hr after oral intake of children and experimental animals. One theory is 10 mg Adderall IR by children (Table 5) or adults (Table 6) that reduced activity and increased attention in children (FDA, 2001; PDR, 2003b). The time to reach peak plasma compares to decreased activity as a secondary effect concentrations of d- and l-amphetamines is B5–7 hr after of stereotypy in animals given high doses. Several oral intake of 10–30 mg Adderall XR by adults (Table 6) studies examining divergent thinking and cognitive or children (Table 5) (FDA, 2001; PDR, 2003b). Plasma perseverance indicated no or inconsistent associations profiles of d- and l-amphetamine are similar after a single between therapeutic effects and cognitive constriction dose of 20 mg Adderall XR or two 10-mg doses of or stereotypic thinking (Greenhill, 1992; Solanto, 2000). Adderall IR administered 4 hr apart (PDR, 2003b). Peak An alternate theory of mechanism of action in children plasma d-amphetamine levels were reported at B30, 50, is that stimulation of inhibitory pre-synaptic autorecep- and 75 ng/mL in children and B15, 30, and 40 ng/mL in tors decreases dopamine activity, thus compensating adults receiving d,l-amphetamine doses of 10, 20, and for excessive dopamine activity in ADHD (Solanto, 30 mg (FDA, 2001). 2000). d-Amphetamine is also available in immediate-acting Amphetamine S-(1)(d-)enantiomers have five times (DextroStats and Dexedrines tablets) and sustained- the stimulant activity of R-(À)(l-)enantiomers (Kraemer release (Dexedrines Spansule capsules) formulations. and Maurer, 2002). The drug labels provide a limited amount of pharmaco- Metabolism of amphetamine and methamphetamine kinetics information obtained from studies in which is discussed in Section 2.1.1.4. Some metabolites of healthy volunteers were administered two or three 5-mg amphetamine or methamphetamine were reported to doses of immediate-acting formulations or one 15-mg be active. 4-Hydroxyamphetamine, which has been dose of a sustained-release formulation (PDR, 2003d,e). used as an ophthalmologic drug, releases norepinephrine The time to reach maximum blood levels was 2–3 hr for from postganglionic sympathetic nerves but has few immediate-acting formulations and 8 hr for the sus- if any CNS effects, possibly due to slowed passage tained-release formulation. Maximum blood level was through the blood-brain barrier as a result of its reported at 29.2 ng/mL in individuals administered increased water solubility (Sellers et al., 1997). Additional 10 mg Dextrostats in two divided doses. Maximum metabolites, 4-hydroxynorephedrine, norephedrine plasma levels were measured at 36.6 ng/mL in volun- (Caldwell et al., 1972; Musshoff, 2000), a-methyldopa- teers given 15 mg Dexedrines as three divided doses of mine, and a-methylnorepinephrine (Sellers et al., 1997) immediate-acting formulation and 23.5 ng/mL in volun- were reported to act as false neurotransmitters. It has teers given 15 mg Dexedrines as one dose of extended- been postulated that false neurotransmitters are involved release formulation. in habituation (Caldwell et al., 1972). a-Methylnorepi- Brown et al. (1979) measured plasma amphetamine nephrine has CNS hypotensive activity (Sellers et al., levels in 16 boys (60–144 months old) orally dosed with 1997). B0.5 mg/kg bw d-amphetamine. Blood was collected at Anggard et al. (1973) examined psychotic symptoms baseline and at 11 time points over a 30-hr time period in stimulant abusers orally administered three for determination of plasma amphetamine levels by 50-mg doses of d,l-amphetamine sulfate over 12 hr. In radioimmunoassay (RIA). Plasma amphetamine levels four subjects administered sodium bicarbonate to peaked between 3 and 4 hr with mean (7SEM) values increase urinary pH, amphetamine excretion was re- 62.773.8 ng/mL at 3 hr and 65.973.6 ng/mL at 4 hr. The duced and as a result of prolonged exposure to greatest inter-subject variations in plasma levels were amphetamine, higher levels of metabolites were excreted observed during the absorption versus elimination compared to three subjects administered ammonium phase. Apparent half-life of elimination was calculated chloride to acidify urine. There was no relationship at 6.870.5 hr. The study was repeated in six of the boys between plasma amphetamine level and intensity of who received a dose of B0.5 mg/kg bw d-amphetamine psychosis. There was a positive association between 1 week after the first study. Analyses of interclass excretion of basic metabolites (4-hydroxyamphetamine, coefficients of variation by w2 demonstrated no signifi- 4-hydroxynorephedrine, and norephedrine) and rating of cant differences in apparent half-life of elimination or psychosis (r 5 0.996 in the 4 subjects with increased peak plasma amphetamine levels during the two urinary pH). The authors interpreted the findings as different time periods. suggesting that accumulation of metabolites and not In a second study, Brown et al. (1980) reported amphetamine is responsible for psychotic symptoms. plasma-amphetamine levels in nine boys (60–144 [Methods for rating psychosis were not discussed, but a months old) orally dosed with a sustained-released reference was provided. The Expert Panel notes that formula of d-amphetamine at B0.5 mg/kg bw. caution is required in the interpretation of this study Plasma levels of amphetamine were determined by due to the small group sizes. A more detailed explana- RIA. [Blood was collected over an unspecified tion on the effects of urinary pH on amphetamine time period and data were only shown for the first excretion and metabolism is included in Section 6 hr after exposure.] Plasma amphetamine levels 2.1.1.5.] peaked between 3 and 8 hr with mean7SEM values 2.1.1.2 Absorption. The Adderall brand of d, in that time period ranging from 64.179.5 to l-amphetamines is available as immediate-release tablets 70.277.9 ng/mL.

Birth Defects Research (Part B) 74:471–584, 2005 DOI 10.1002/bdrb 480 GOLUB ET AL. Intake of Adderall XR with a high fat breakfast delayed Tmax by 2–2.5 hr for both the d- and l-enantiomers, but 1.57 2.8 1.6 3.1 did not affect extent of absorption (Table 6) (FDA, 2001; 7 7 7 7 (hr) PDR, 2003b). Rate and extent of d-amphetamine absorp- Half-life , maximum tion were reported to be similar after administration of

max s

)NS ) 10.9 Dexedrine sustained-release capsules to fasted or fed (58–75 g fat) volunteers (PDR, 2003d). 0–24 0–24 2.1.1.3 Distribution. After oral intake, amphetamines are rapidly distributed to major organ systems, including

(ng-h/mL) the brain (NTP, 1991). Linear pharmacokinetics were a 51.6 8.55 123 10.2 60 (AUC 40 (AUC 95 9.0 ported in both studies and a 7 7 7 7 7 demonstrated for Adderall XR at doses of 10–30 mg (PDR, 2003b). No unexpected accumulation occurred at steady state (PDR, 2003b). The volume of distribution for methamphetamine was 1.5 146

7 reported at 3.42 L/kg in humans (Cho et al., 2001).

(hr) AUC Steiner et al. (1984) measured blood, milk, and urine -Amphetamine l levels of amphetamine in a 36-year-old woman taking max 1.2–3.2 3.5 454 3.3 216 3.5 138 2.7 425 20 mg/day racemic amphetamine in four divided doses 7 7 7 7 7 between 10 AM and 4 PM for treatment of narcolepsy. At 10 and 42 days after delivery, two samples of blood and

3.5 2.5 milk were collected at 9:30 AM and 2 PM and a 24-hr , maximum concentration; NS, not specified; T 7 urine sample was obtained for analysis by gas chroma-

max tography (GC). Amphetamine concentrations 42 days postpartum in plasma were 20–40 ng/mL, breast milk (ng/mL) T 2.35–10.6 5.9 5.6 4.3 6.9 1.9 6.4 6.3 5.6 concentrations were 55–118 ng/mL, and milk:plasma max 7 7 7 7 7 ratios were 6.6–7.5. Maternal urinary amphetamine excretion was measured at 3.6 mg/24 hr 10 days postpartum and 8.9 mg/24 hr 42 days postpartum. Twelve-hour urine samples were collected from the 0.97 9.64 1.9 22.7 2.4 14.8 1.7 22.7 7 7 7 7 infant and urinary amphetamine excretion was reported (hr) C to be 1/1000 to 1/300 of maternal levels. No untoward Half-life

Amphetamine in Children effects on growth, neurological assessments, emotional )) NS 9.5 8.8 l- development, or motor achievements were noted in the 0–24 0–24

Table 5 infant at up to 24 months of age. a and

N Bost et al. (1989) reported methamphetamine and d- 0– amphetamine distribution in two premature twin boys born to a women who used methamphetamine by i.v. 123 (AUC 109 7.47 281 8.6 190 (AUC 231 8.0 AUC (ng-h/mL) 7 7 7 7 7 injection 5 hr before hospital admission; the infants died shortly after birth. Distribution of amphetamine and methamphetamine is reported in Table 7. Maternal blood levels were not reported. 1.3 384

7 Stewart and Meeker (1997) reported postmortem

(hr) methamphetamine and amphetamine blood concentra- Pharmacokinetics of -Amphetamine

d tions from stillbirths or infant deaths when the mother max 3.5 432 1.2–3.3 3.6 1338 3.2 704 2.7 1255 was believed to have used methamphetamine (Table 8). 7 7 7 7 7 Maternal blood concentrations were available in two cases. [The ascertainment of maternal exposure was not discussed and the timing of maternal blood sampling

11.1 2.5 was not given.] 7 2.1.1.4. Metabolism. Amphetamine is a metabolite of methamphetamine and the metabolic pathways for the

(ng/mL) T two compounds are illustrated together in Figure 2. 6.2 6.4 6.50–33.8 20.6 5.5 13.5 6.8 21.4 5.5

max The initial step of amphetamine metabolism is hydro- 7 7 7 7 7 C xylation of the a, aromatic 4-, or b carbon (Musshoff,

28.4 2000; NTP, 1991). As noted in Table 9, metabolism

b through each possible pathway varies according to unless otherwise specified. species (NTP, 1991). Oxidation of the a carbon leads to N

0– deamination and ultimately to the formation of benzoic acid, which can be conjugated with glycine to form hippuric acid (Musshoff, 2000; NTP, 1991). According to

n) the NTP review, deamination seems to be the predominant pathway of amphetamine metabolism in humans, leading to urinary excretion of primarily benzoic acid 10 mg Adderall XR (20) 73.9

subjects ( and hippuric acid (Table 9). In contrast, the NTP (1991) Â Two separate studies were conducted, at least one of which was published (Greenhill et al., 2003); a range of values is included when the endpoint was re Values are for AUC 10 mg Adderall XR (8) 28.8 From FDA (2001, 2002). Errors (e.g., SEM, SD) were not specified. AUC, area under the concentration versus time curve; C Study regimen and a b single value is included when reported in one study. 30 mg Adderall XR (20) 75.9 time. 20 mg Adderall XR (48)3 48.8 10 mg Adderall IR (9, 12) noted that the main metabolic pathway in rats is

Birth Defects Research (Part B) 74:471–584, 2005 DOI 10.1002/bdrb AMPHETAMINE AND METHAMPHETAMINE 481 Table 6 Pharmacokinetics of d- and l-Amphetamine in Adults

d-Amphetamine l-Amphetamine

Study regimen and Cmax Tmax AUC0–N Half-life Cmax Tmax AUC0–N Half-life subjects (n) (ng/mL) (hr) (ng-h/mL) (hr) (ng/mL) (hr) (ng-h/mL) (hr)

10 mg Adderall IR (8) 14.574.1 3.370.9 3007116 11.973.9 4.771.4 3.571.2 124771 15.277.8 10 mg Adderall XR (8) 13.974.2 6.271.8 2967101 12.273.3 4.471.3 6.271.8 115754 15.275.7 2 Â 10 mg Adderall IR (19) 28.377.1 6.971.3 5307114 10.972.0 9.372.4 7.171.4 203749 13.272.7 20 mg Adderall XR (19) 28.178.8 7.072.4 5677114 11.872.7 8.772.8 8.274.4 203747 13.772.8 20 mg Adderall XR (7) 29.174.9 5.071.6 521779 10.471.7 9.471.6 5.071.5 197735 12.672.5 fasting 20 mg Adderall XR (7) 28.074.4 7.172.1 557797 10.872.3 8.771.4 7.472.2 205739 13.472.7 high-fat breakfast 30 mg Adderall XR (19) 44.3711.1 5.272.0 8517214 10.472.3 13.373.7 5.672.1 289779 12.773.3 fasting 30 mg Adderall XR (19) 39.778.8 7.772.3 8237200 10.372.0 12.072.9 8.372.9 274769 12.572.6 high-fat breakfast

From FDA (2001). Errors (e.g., SEM, SD) were not specified. AUC, area under the concentration versus time curve; Cmax, maximum concentration; Tmax, maximum time.

Table 7 systemic exposure to each enantiomer has been demon- Methamphetamine and Amphetamine Distribution in strated to be equivalent to the composition of Adderall Twin Boys Exposed to Methamphetamine In Utero 5 Hr (3-d:1-l). [It is possible that the preferential b-hydro- Before Birth xylation of the d-enantiomer does not significantly affect systemic exposure because it is not expected to Organ Methamphetamine level Amphetamine level be a significant pathway with standard therapeutic dosing.] Kidney (mg/kg) 6.34–7.38 0.95–0.97 Liver (mg/kg) 9.20–11.0 0.18–1.43 The two major pathways of methamphetamine meta- Brain (mg/kg) 4.53–5.71 0.61–0.76 mg/kg bolism are N-demethylation to form amphetamine, Blood (mg/L) 6.3 [6,300 ng/mL] 0.28 [280 ng/mL] which can be metabolized through several pathways Placenta (mg/kg) 8.66 1.18 (see above), and aromatic hydroxylation to form 4-hydroxyamphetamine and then 4-hydroxynorephe- Data from Bost et al., 1989. drine (Caldwell et al., 1972; Musshoff, 2000; Kraemer and Maurer, 2002). Hydroxynorephedrine is described as aromatic hydroxylation and the main urinary metabolite a false neurotransmitter (Sellers et al., 1997) that may is p-hydroxyamphetamine (4-hydroxyamphetamine); play a role in habituation (Caldwell et al., 1972). Table 12 both metabolites generated from the aromatic hydro- compares urinary methamphetamine metabolites and xylation pathway are reported to have biological activity Table 13 compares the contribution of each metabolic (Musshoff, 2000). Hydroxylated metabolites can be pathway in humans, rats, and guinea pigs (Caldwell excreted as sulfate conjugates (Musshoff, 2000). Aliphatic et al., 1972). As noted in Table 12, humans excreted a b carbon hydroxylation accounts for only a minor larger percentage of unmetabolized methamphetamine percentage of metabolism, but is considered important than rats and guinea pigs, despite receiving lower because the resulting metabolite, norephedrine, is re- doses. The two major pathways in humans are aromatic ported to have biological activity (Musshoff, 2000). hydroxylation and demethylation, whereas deamination Additional metabolic pathways such as nitrogen hydro- and b-hydroxylation account for a smaller percentage xylation, oxidation, and conjugation are mentioned in of metabolism. In humans, aromatic 4-hydroxylation reviews (NTP, 1991; Musshoff, 2000), but do not seem is much more extensive in the metabolism of metham- to be primary pathways based on metabolites detected phetamine compared to amphetamine (Sellers et al., in urine. 1997). Racemization does not occur during amphetamine Nitrogen demethylation of racemic methamphetamine metabolism (Kraemer and Maurer, 2002). Information is reported to be stereospecific with more rapid on stereospecificity of amphetamine metabolism is biotransformation of the d-enantiomer (Musshoff, 2000). conflicting. Dopamine b-hydroxylase is the enzyme During the first 16 hr after administration of racemic that catalyzes hydroxylation and is reported to react methamphetamine, both enantiomers were excreted at stereospecifically with the d-enantiomer in b-carbon approximately equal levels, but excretion of l-metham- hydroxylation (Musshoff, 2000). Due to the preferential phetamine was increased thereafter. b-hydroxylation of d-amphetamine, less is excreted intact The role of cytochrome P450 (CYP) enzymes in compared to the l-enantiomer. An FDA (2001) review, amphetamine and methamphetamine metabolism in however, reported that metabolism and elimination do humans was discussed in a review by Kraemer and not seem to be stereoselective, because the ratio of Mauer (2002). Studies using human liver microsomes,

Birth Defects Research (Part B) 74:471–584, 2005 DOI 10.1002/bdrb 482 GOLUB ET AL. Table 8 Methamphetamine and Amphetamine Measured in Stillbirths and Infant Deaths

Blood concentration (mg/L)

Gestational age Methamphetamine Amphetamine Other measurements or comments

Stillbirths 6 months 0.34 0.05 Fetus had been stored frozen for 3–4 days. 20 weeks 0.20 0.08 30–32 weeks 1.20 0.06 Maternal blood methamphetamine 0.18 mg/L, amphetamine 0.03 mg/L. Not stated 0.36 0.01 Methamphetamine 0.54 mg/g in brain and 0.52 mg/g in liver; amphetamine 0.04 mg/g in brain and liver. Infant deaths Full term 0.40 0.06 Maternal blood methamphetamine 0.21 mg/L, amphetamine undetectable. 32 weeks 0.57 0.06 Full term 0.13 0.05 Infant died at 1 month of age. Not stated 0.03 0 Infant died at 1 month of age. 28–32 weeksa 0.355 0.080 Infant died at 4 hr of age. Mother reported taking 30–45 mg/day methamphetamine in diet pills. Infant tissue levels (mg/g): Tissue Methamphetamine Amphetamine Bile 0.384 0.050 Kidney 0.746 0.080 Lung 0.857 0.120 Brain 0.280 o0.030 Liver 0.246 o0.020 From Stewart and Meeker, 1997. aAdded from Garriott and Sprull, 1973.

including those with poor CYP2D6 metabolic capability, of samples occurred before dosing and at various demonstrated a major role of CYP2D6 in aromatic time points after dosing for up to 48 hr for blood, 4-hydroxylation of methamphetamine. Results were 72 hr for urine, and 24 hr for saliva. Methamphetamine replicated using recombinant CYP2D6 in yeast. A second and amphetamine levels in samples were analyzed investigator obtained the same results for amphetamine by GC. using microsomal preparations from cells expressing Pharmacokinetic results for oral exposure are listed in CYP2D6. Amphetamine and methamphetamine were Table 10 and for inhalation and i.v. exposure in Table 11. reported to be substrates and competitive inhibitors of The following are observations or conclusions made by CYP2D6. In their review of human and animal studies study authors based on results of these studies: (discussed in Section 2.1), Kraemer and Mauer (2002) concluded ‘‘ythere is convincing evidence about the role * No significant differences in pharmacokinetic para- of CYP2D6 in the ring hydroxylation of amphetamine meters were observed when 0.125 mg/kg bw and methamphetamine.’’ A number of original studies methamphetamine was administered orally, before were reviewed in detail. versus after the 13-day oral exposure to the slow- Two studies conducted by Cook et al. (1992, 1993) release drug. provide information on pharmacokinetics in adults * Compared to study day 1, a slight but significant exposed to methamphetamine through different routes increase in maximum plasma methamphetamine level or after single versus repeat dosing. In one study, eight was observed when 0.250 mg/kg bw was adminis- healthy men received oral doses of 0.125 or 0.250 mg/kg tered on Study Day 15, after the 13-day subchronic bw d-methamphetamine HCl orally on Study Day 1, exposure. [The Expert Panel notes this finding is not 10 mg/day of a slow-release, unlabeled d-methampheta- likely to be of pharmacological relevance.] mine medication on Study Days 2–14, and 0.125 or * Pharmacokinetic parameters were similar after inha- 0.250 mg/kg bw d-methamphetamine HCl on study day lation and i.v. exposures. 15 (Cook et al., 1992). Six volunteers were exposed to * Bioavailability was higher after inhalation (90.3%) both doses and each value obtained was based on three versus oral (67.2%) exposure. to six volunteers. Pharmacokinetic parameters were * Percentage of dose excreted in urine was greater after examined on Study Days 1 and 15. In the second study, oral exposure to 0.125 mg/kg bw compared to pharmacokinetic parameters were examined in six 0.250 mg/kg bw on Study Day 1 (statistically sig- healthy men administered methamphetamine at nificant) and Study Day 15. 21.870.3 (SEM) mg [0.26 mg/kg bw based on reported * AUC values for amphetamine (metabolite) mean body weight of 83.8 kg] through the inhalation were proportional between oral doses of 0.12 (AUC 5 route (smoking heated vapors) or 15.5 mg [0.18 mg/kg 98.3 ng-h/mL) and 0.250 mg/kg bw (AUC 5 bw] through i.v. injection (Cook et al., 1993). Mean 224 ng-h/mL); the values were 30% of parent AUC free base doses were 17.5 mg for inhalation exposures (330 ng-h/mL at low dose and 775 ng-h/mL at high and 12.42 mg for i.v. exposures. In both studies, collection dose).

Birth Defects Research (Part B) 74:471–584, 2005 DOI 10.1002/bdrb AMPHETAMINE AND METHAMPHETAMINE 483

Fig. 2. Metabolism of amphetamine and methamphetamine (Caldwell et al., 1972; NTP, 1991; Musshoff, 2000).

Table 9 Comparison of Amphetamine Urinary Metabolites in Various Species

% Dose excreted in urine (48 hr for rats, 24 hr for other species)

Dose Benzoic1hippuric 4-Hydroxy Total % of dose Species (sex) (mg/kg bw) acid Phenylacetone amphetamine Amphetamine in urine

Human (M)a 0.66c 45 2 9 37 66 Rhesus monkey (F)b 0.66d 31–38 0 0–11 3.8–31 42–73 Squirrel monkey (sex not given)a 2c 5 ND 1 23 34 Rat (Wistar, F)a,b 10e 3 0 60 13 85 Mouse (SAS/ICI, F)b 10d 31 0 14 33 78 Rabbit (New Zealand, F)a,b,f 10e 25 22 6 4 72 Dog (greyhound, F)a,b 5e 28 1 6 30 75 Guinea pig (F)b 5d 62 0 0 22 83 M, male; F, female; ND, not determined. aFrom NTP (1991). bFrom Dring et al., 1970. cEnantiomers not specified. dd-Amphetamine. ed,l-Amphetamine. fRabbits also excreted 8% 1-phenylpropanol, a metabolite not seen in most other species.

Birth Defects Research (Part B) 74:471–584, 2005 DOI 10.1002/bdrb 484 GOLUB ET AL. Table 10 Pharmacokinetic Parameters in Men Orally Administered d-Methamphetamine HCl

0.125 mg/kg bw 0.250 mg/kg bw

Parameter Study day 1 Study day 15 Study day 1 Study day 15 7 7 7 7 Tmax (hr) 3.60 0.63 3.06 0.62 3.23 0.38 2.64 0.20 7 7 7 7 Cmax (ng/mL) 19.8 2.7 20.3 3.0 37.2 1.3 41.8 1.7 Half-life (hr) 8.4670.71 9.7171.10 11.4571.57 10.9371.45 % dose in urine 54.175.8 50.0712.4 34.674.3 30.073.1 Total clearance (mL/min) 446766 404760 381766 357757 Renal clearance (mL/min) 212733 189727 138741 122723 a B B B B Cmax (ng/mL) for amphetamine (metabolite) 1.6 1.2 3.9 4.2 7 From Cook et al., 1992. All values are for methamphetamine, unless otherwise indicated; values presented as mean SEM. Tmax, maximum time; Cmax, maximum concentration. aValues estimated from a graph by CERHR.

Table 11 Pharmacokinetic Parameters in Men Administered d-Methamphetamine HCl Through Inhalation or Intravenous Exposure

Mean dose in mg (mg/kg bw)

Parameter 21.8 (0.26) by Inhalation 15.5 (0.18) Intravenously

Half-life (hr) 11.871.35 13.171.54 7 7 AUC0–N (ng-h/mL) 1013 141 787 29.7 Mean residence time (hr) 16.771.46 17.472.15 Total clearance in L/hr (mL/min) 15.970.73 (265712.2) NS Renal clearance in L/hr (mL/min) 6.6870.80 (111713.3) 6.9571.25 (116720.8) Volume of distribution (L/kg) 3.2470.36 3.7370.59 % Dose in urine 36.874.3 45.079.5 % Metabolic clearance 57.975.0 55.079.5 7 7 Cmax (ng/mL) for amphetamine (metabolite) 4.2 0.56 4.0 0.63 From Cook et al., 1993. All values are for methamphetamine, unless otherwise indicated. Values presented as mean7SE. AUC, area under the concentration versus time curve; Cmax, maximum concentration; NS, not specified.

* Renal excretion of parent drug after oral, i.v., before substantial decreases in plasma methampheta- or inhalation exposure was dependent on urine mine, suggesting development of acute tolerance. flow and pH; dose was also a factor with oral exposure. Pharmacokinetic parameters in eight male adult * Renal clearance rates after oral dosing (Table 10) volunteers who smoked 40 mg methamphetamine or exceeded the average renal filtration rate of inhaled 50 mg methamphetamine in a mist (intranasal B125 mL/min, suggesting involvement of an active exposure) were examined in a study by Harris et al. transport mechanism in renal clearance. (2003). Pharmacokinetic parameters were similar after * Dose-dependent differences in renal clearance smoking and intranasal exposure and were comparable between oral doses of 0.125 and 0.25 mg/kg bw to values reported by Cook et al. (1993). Bioavailability suggest possible saturation of a renal active transport with smoking exposure was reported at 67% by Harris, process. which was lower than the 90% value reported by Cook * A comparison of oral, inhalation, and i.v. data et al. (1993). Differences in pipe temperature and suggested that renal elimination is decreased with smoking techniques were discussed as possible reasons increased bioavailability, also suggesting saturable for the discrepancy between the two studies. Bioavail- excretion. ability was reported at 79% after intranasal exposure. * Urinary amphetamine represented B15% of the oral 2.1.1.5 Excretion. The half-life for d-amphetamine doses and B7% of the inhalation or i.v. dose. was reported at B7 hr in children taking 10 mg Adderall * Large variations in plasma to saliva ratios were IR, B8–9 hr in children taking 20–30 mg Adderall XR, observed with oral and inhalation dosing; saliva to and B10–12 hr in adults receiving 10–20 mg Adderall IR plasma ratios of methamphetamine were similar with or 10–30 mg Adderall XR. Dosing with the Adderall inhalation and i.v. dosing. formulations described above resulted in l-amphetamine * Subjective (e.g., feeling ‘‘high’’) and cardiovascular half-lives of B9–11 hr in children (Table 5) and 13–15 hr effects after inhalation and i.v. exposure subsided in adults (Table 6) (FDA, 2001).

Birth Defects Research (Part B) 74:471–584, 2005 DOI 10.1002/bdrb AMPHETAMINE AND METHAMPHETAMINE 485 Biological half-life for methamphetamine was reported metabolites excreted increased at higher urinary pH. As at 4–5 hr (PDR, 2003c). Another review reported a noted in Section 2.2.1.3, numerous drugs can increase or half-life of 12 hr for methamphetamine in humans decrease urinary pH. (Cho et al., 2001). 2.1.2 Experimental animal Amphetamine, methamphetamine, and their metabo- 2.1.2.1 Pharmacodynamics. A study using mice lites are excreted primarily in the urine. Dosing with two lacking the dopamine transporter examined the role of 5-mg d-amphetamine tablets resulted in 45% urinary amphetamine in dopaminergic neurons (Jones et al., recovery in 48 hr (PDR, 2003e). In two male volunteers 1998). The study suggested that amphetamine induced given 0.29 mg/kg bw radiolabeled methamphetamine, release of dopamine from neuronal vesicles to cytoplasm the percentages of radioactivity recovered in urine and then caused the reverse transport of dopamine from during the first 4 days after dosing were 55–69% at cytoplasm to extracellular space through the cytoplasmic 24 hr, 78–90% at 48 hr, 86–94% at 72 hr, and 88–96% at dopamine transporter. Vesicle release was the rate limit- 96 hr (Caldwell et al., 1972). In three to six male ing step, but both processes were necessary for amphe- volunteers given deuterated methamphetamine orally, tamine-induced extracellular dopamine release. urinary recoveries within 72 hr were B50–55% at a dose 2.1.2.2 Pharmacokinetics. The volume of distribution of 0.125 mg/kg bw and B30–35% at a dose of 0.25 mg/kg for methamphetamine in rats was reported at 3.95 L/kg bw (Cook et al., 1992). (Cho et al., 2001). As in humans, amphetamine metabolic On average, about 30–40% of an amphetamine or pathways in experimental animals can include hydro- methamphetamine dose is eliminated unchanged and xylation of the a carbon, the aromatic 4-carbon, the b the remainder is eliminated as metabolites (Table 12) carbon, or possibly the amine group (NTP, 1991). As (Caldwell et al., 1972; Musshoff, 2000; PDR, 2003c). noted in Table 9, metabolism through each pathway About 37–45% of a methamphetamine dose was elimi- varies according to species. Deamination can occur after nated unchanged and B7% was eliminated as ampheta- a-carbon hydroxylation. According to Table 9, the path- mine in men given 22 mg through inhalation or 15.5 mg way plays a minimal role in rats but is a major pathway by i.v. exposure (Cook et al., 1993). The percentage of in humans. Aromatic hydroxylation is the predominant parent compound excreted varies, however, according to pathway in rats and the primary urinary metabolite is urinary pH (Musshoff, 2000). Due to a pKa of B9.9, 4-hydroxyamphetamine. amphetamines are primarily ionized under normal One review reported that 4-hydroxyamphetamine can physiological pH and in acidic urine. Because the ionized be metabolized by a neuronal CYP to a-methyldopamine form is not reabsorbed significantly by the kidney, large and then to a-methylnorepinephrine, a possible false amounts are excreted unchanged. In alkaline urine, neurotransmitter (Sellers et al., 1997). Studies in whole amphetamines exist primarily in the un-ionized form, and striatal preparations of rat brain demonstrated and are readily reabsorbed by the kidney. Reabsorption d-amphetamine hydroxylation; after administration of results in prolonged half-life and increased biotransfor- d-amphetamine to experimental animals (mostly rats), mation. Excretion of unmetabolized amphetamine and 4-hydroxyamphetamine, 4-hydroxynorephedrine, a- methamphetamine was reported to range from 1–2% in methyldopamine, and a-methylnorepinephrine were highly alkaline urine to 74–76% in strongly acidic urine. found in rat brain. The half-life of hydroxyamphetamine In a study by Anggard et al. (1973), the amount was 1.5 days and the half-life of hydroxynorephedrine of amphetamine excreted unchanged in urine over 4 was 2.5 days in rat striatum. days was B70% in three subjects with acidic urine Methamphetamine metabolism in rats exposed by (pH 5 5.3–5.8) and B20–56% in four subjects with gavage and guinea pigs exposed by intraperitoneal neutral-to-basic urine (pH 5 6.7–7.5); percentage of (i.p.) injection is qualitatively similar to that for humans,

Table 12 Comparison of Urinary Methamphetamine Metabolites in Humans, Rats, and Guinea Pigs

Percentage of methamphetamine dose in urinea

Humans, 0.29 mg/kg Rats, 45 mg/kg bw Guinea pigs, Guinea pigs, Compound bw orallyb orallyc 10 mg/kg bw i.p. 45 mg/kg bw i.p.

Methamphetamine 23 11 0.5 3 Amphetamine 3 3 4 13 Norephedrine 2 0 1 19 4-Hydroxymethamphetamine 15 31 0 0 2-Hydroxyamphetamine 1 6 0 0 4-Hydroxynorephedrine 2 16 0 0 Benzoic acid 5 4 63 31 Benzyl methyl ketone precursor 1 NS 11 2 Total 52 71 79.5 68 From Caldwell et al., 1972. NS, not specified. aUrine was collected for 1 day in humans and guinea pigs and 2 days in rats. bMales, n 5 2. cn 5 3.

Birth Defects Research (Part B) 74:471–584, 2005 DOI 10.1002/bdrb 486 GOLUB ET AL. Table 13 Comparison of Methamphetamine Metabolic Pathways in Humans, Rats, and Guinea Pigs

Percentage of methamphetamine dose metabolized through each pathway

Humans, Rats, 45 mg/kg bw Guinea pigs, Guinea pigs, Pathway 0.29 mg/kg bw orallyb orallyc 10 mg/kg bw i.p.c 45 mg/kg bw i.p.c

Unmetabolized 23 11 0.5 3 Aromatic hydroxylation 18 53 0 0 Demethylation 14 28 79 64 B-hydroxylation 4 16 1 19 Deamination 6 4 74 33 From Caldwell et al., 1972. aUrine was collected for 1 day in humans and guinea pigs and 2 days in rats. bn 5 2. cn 5 3.

as described in Section 2.1.1.4 (Table 13). As noted in Kraemer and Mauer (2002) reviewed a study demon- Table 12, however, the percentage of unchanged metham- strating that quinidine, an inhibitor of CYP2D and phetamine excreted in urine is lower in rats and guinea CYP2C3, inhibited deamination by purified rabbit pigs compared to humans, despite the rats receiving CYP2C3. It was concluded that CYP2C isozymes are higher doses than humans. Aromatic hydroxylation is greatly involved in amphetamine deamination. the predominant metabolic pathway in rats, but sub- A half-life of 87 min in plasma and 62 min in brain was stantial metabolism also occurs through demethylation reported after i.v. injection of rats with 0.5 mg/kg bw d,l- and b-hydroxylation. In guinea pigs, aromatic hydro- amphetamine (NTP, 1991). Tissue half-life was reported xylation does not seem to occur and the main metabolic at 5–9 hr after i.p. injection of rats with d,l-amphetamine pathways are demethylation and deamination. Beta sulfate (NTP, 1991). hydroxylation of amines was noted as a reaction of A half-life of 70 min was reported for an unspecified interest by Caldwell et al. (1972), because resulting concentration of methamphetamine given to rats through metabolites, 4-hydroxynorephedrine and possibly nore- an unspecified route (Cho et al., 2001). Studies in rats phedrine, can act as false neurotransmitters, postulated given radiolabeled methamphetamine at 45 mg/kg bw to be involved in habituation associated with chronic orally or by i.p. injection demonstrated that B75–85% of methamphetamine intake. Urinary excretion of norephe- radioactivity was excreted in urine and B1–3% in feces drine increased and benzoic acid decreased when the i.p. over 3 days (Caldwell et al., 1972). Similar values were dose in guinea pigs was increased from 10–45 mg/kg bw. reported for guinea pigs dosed i.p. with 10 mg/kg bw Caldwell et al. (1972) suggested that deamination may be methamphetamine, but at 45 mg/kg bw, urinary excre- saturated at high doses in guinea pigs. tion of radioactivity was reported at 44–87% and fecal The role of CYP enzymes in amphetamine metabolism excretion at 5.5–29% over 4 days. in experimental animals was discussed in a review by The FDA pharmacology review for Adderall (FDA, Kraemer and Mauer (2002). By administering quinidine 2001) summarized fertility and reproductive studies that (a specific CYP2D inhibitor) to rats dosed with amphe- included pharmacokinetic data. Sprague–Dawley rats tamine, it was demonstrated that CYP2D is involved in and New Zealand White rabbits were dosed by gavage aromatic hydroxylation of amphetamine. CYP2D1/6 with free amphetamine base [the Expert Panel assumes a catalysis of methamphetamine 4-hydroxylation was 3:1 mixture of d- and l-enantiomers, as in the marketed demonstrated using liver microsomes from Sprague– product]. The results as given in the FDA summary are Dawley and Dark Agouti rats (a poor CYP2D6 metabo- shown in Tables 14 and 15. The developmental and lizer phenotype); addition of anti-P450 BTL IgG, reproductive endpoints are discussed in Sections 3.2 bufuralol (a CYP2D6 substrate), or quinine (a CYP2D6 and 4.2. inhibitor) blocked approximately 90% of the reaction [The Expert Panel noted a disproportionate increase in Sprague–Dawley rat microsomes. The reaction in AUC24 in rats dosed with 2, 6, and 20 mg/kg bw/day was mediated using reconstituted CYP2D isozymes Adderall and rabbits dosed with 2, 6, and 16 mg/kg bw/ purified from rat liver microsomes. A study measuring day Adderall, suggesting saturation or dose-dependent urinary elimination of 4-hydroxyamphetamine and elimination and drug accumulation.] amphetamine in Sprague–Dawley rats pre-dosed A number of studies were reviewed in detail because with cytochrome P450 inhibitors or inducers they examined pharmacokinetic parameters in pregnant demonstrated significantly reduced elimination of or immature animals and were highly relevant to a 4-hydroxyamphetamine in rats treated with inhibitors CERHR evaluation. 1-aminobenzotriazole, carbon tetrachloride, quinidine, Acuff-Smith et al. (1996) performed a developmental quinine, and primaquine. Based on debrisoquine meta- neurotoxicity study using pregnant Sprague–Dawley rats bolism, CYP2D1 was first thought to be the rat enzyme treated with d-methamphetamine (free base [purity not equivalent to CYP2D6 in humans, but it was determined given]) twice daily by subcutaneous (s.c.) injection from later that the equivalent enzyme in rats is CYP2D2 gestation day (GD) 7–12 or from GD 13–18 (plug 5 GD 0). (Vorhees et al., 1999). The developmental aspects of the study are discussed in

Birth Defects Research (Part B) 74:471–584, 2005 DOI 10.1002/bdrb AMPHETAMINE AND METHAMPHETAMINE 487 Section 3.2. A satellite group of pregnant animals given were measured in plasma by high performance liquid 20 mg/kg bw/injection (40 mg/kg bw/day) was used for chromatography (HPLC) to determine pharmacokinetic measurement of serum methamphetamine and amphe- parameters in ewes and fetuses. Results are listed in tamine (four dams at each of five time points over 8 hr Table 18. Methamphetamine rapidly crossed the placenta. after the 11th injection). Estimates of pharmacokinetic Though initial plasma levels were higher in ewes than parameters for methamphetamine are presented in fetuses, the longer elimination time in fetuses resulted Table 16. Amphetamine serum concentrations rose over in higher fetal AUC values. Fetal half-life was directly the first 2 hr after injection of methamphetamine, reach- correlated with maternal half-life (r 5 0.78) and inversely ing a plateau of about 400 ng/mL that was maintained throughout the remainder of the 8-hr sampling period. Plasma and brain pharmacokinetic parameters for Table 15 neonatal rats treated with s.c. methamphetamine were Pharmacokinetic Results in Pregnant Rabbits Given reported by Cappon and Vorhees (2001) and are given in Adderall Table 17.

Won et al. (2001) measured maternal and fetal brain Cmax (ng/mL) AUC (ng-h/mL) levels of methamphetamine and amphetamine after s.c. Dose st nd injection of mouse C57B1/6 dams with 40 mg/kg bw (mg/kg bw/day) 1 dose 2 dose AUC8 AUC24 d-methamphetamine hydrochloride on GD 14. In maternal striatum, methamphetamine levels peaked at 2 22.7 36.2 33.95 89.85 B510 ng/mg protein 1 hr after injection and rapidly 6 61.9 158.5 121.69 376.77 declined over the remaining 3 hr of the experiment; 16 258.9 359.9 588.23 1464.42 B amphetamine levels peaked at 77 ng/mg protein at Does were gavaged with two equal treatments separated by 8 hr 2 hr and remained at that level for up to 4 hr. Metham- on GD 6–19, and were sampled on GD 19. Data from FDA phetamine levels in fetal brain peaked at B22 ng/mg Pharmacology Review (FDA, 2001). Cmax, maximum serum protein at 1 hr and declined over the next 3 hr. Amphe- concentration; AUC, area under the time-concentration curve tamine was only detected in fetal brain at 2 and 4 hr (subscript denotes number of hr plotted after drug administra- (B18 ng/mg protein) after injection of dams. A second tion). Half-life was not studied. experiment demonstrated that levels of methamphetamine were similar in maternal striatum (B335 ng/mg protein) and cortex (B294 ng/mg protein), but lower in brainstem (B236 ng/mg protein). Fetal brain levels of Table 16 methamphetamine reported for one litter were B99 ng/ Pharmacokinetic Parameters for Methamphetamine mg protein in striatum, B102 ng/mg protein in brain- Administered to Pregnant Sprague–Dawley Rats B stem, and 57 ng/mg protein in cortex. Treatment C AUC Burchfield et al. (1991) studied the pharmacokinetics max period (GD) Tmax (hr) (ng/mL) (ng-h/mL) t1/2 (hr) and pharmacodynamics of methamphetamine in pregnant sheep. On GD 125 (B85% of term), catheters were 7–12 0.75 3,600 16,200 6 inserted in Grade Western sheep and the animals were 13–18 0.75 3,100 17,100 6 given antibiotics during a 3-day recovery period. After Estimated from figure in Acuff-Smith et al., 1996, presenting the recovery period, sheep received one or more i.v. serum concentrations for 8 hr after the 11th twice/daily dose of treatments of methamphetamine that included 0.6 mg/ methamphetamine 20 mg/kg bw/dose. Area under the time- kg bw over 12.5 min, 1.2 mg/kg bw over 12.5 min, or concentration curve (AUC; 8-hr) estimated using trapezoidal 1.2 mg/kg bw over 30 sec. Blood was collected from four method. Tmax, maximum time; Cmax, maximum serum concen- or five ewes/group and methamphetamine levels tration; t1/2, serum elimination half-life.

Table 14 Pharmacokinetic Results in Rats Given Adderall

Cmax (ng/mL) AUC (ng-h/mL)

st nd Sex Gavage regimen Dose (mg/kg bw/day) t1/2 (hr) 1 dose 2 dose AUC8 AUC24

Male Daily for 3 weeks 2 ND 39.1 46.5 100 216 6 ND 233.9 203.4 648 1,187 20 1.9 880.4 976.1 2,822 5,689 Nonpregnant female Daily for 1 week 2 2.7 81.6 81.3 236 555 6 ND 212.2 236.0 888 1,599 20 ND 1,080.7 1,196.9 3,727 8,506 Pregnant female, GD 17 Divided doses GD 6–17 2 3.0 80.3 93.2 221 455 6 2.3 248.6 233.8 700 1,566

Data from FDA Pharmacology Review (FDA, 2001). t1/2, serum elimination half-life; Cmax, maximum serum concentration; AUC, area under the time-concentration curve (subscript denotes number of hr plotted after drug administration); ND, not determined.

Birth Defects Research (Part B) 74:471–584, 2005 DOI 10.1002/bdrb 488 GOLUB ET AL. Table 17 Pharmacokinetic Parameters for Methamphetamine Administered to Neonatal Sprague–Dawley Rats

Plasma Brain

Cmax AUC8 Cmax AUC8 Dose regimen (sc) PND (ng/mL) Tmax (hr) t1/2 (hr) (ng-h/mL) (ng/mL) Tmax (hr) t1/2 (hr) (ng-h/mL)

15 mg/kg bw Â 1 dose 1 5750 0.25 1.9 23,800 15.5 1.0 4.5 63 11 5150 0.17 2.0 17,860 19.8 0.5 2.3 54 20 mg/kg bw Â 2 (6 hr apart) 1 7150 0.33 1.3 17,860 21.7 1.0 3.3 100 11 6320 0.50 0.9 22,320 32.8 0.5 2.0 76 10 mg/kg bw Â 4 (2 hr apart) 1 7300 0.33 3.2 16,370 32.9 1.0 3.8 122 11 6100 0.25 2.8 14,475 34.8 0.25 2.3 94

Data from Cappon and Vorhees, 2001. Area under the time-concentration curve (AUC8) data estimated from bar graphs in the original study. Four to six pups were used for most data points, avoiding the use of littermates, and included both males and females. Cmax, maximum serum concentration; Tmax, maximum time; t1/2, serum elimination half-life.

Table 18 Pharmacokinetic Parameters of Methamphetamine in the Pregnant Sheep and Fetus

Treatment of ewe (n) Measured in ewe or fetus Cmax (ng/mL) AUC0–N (ng-min/mL) Half-life (min)

0.6 mg/kg bw, i.v. infusion (4) Ewe 300772.5 8,83472,585 22.373.4 Fetus 200721.0 8,88872,605 31.7712.4 1.2 mg/kg bw, i.v. infusion (5) Ewe 6947189 19,65177,023 29.778.0 Fetus 4577104 20,68875,870 33.879.2 1.2 mg/kg bw i.v. bolus (5) Ewe 10,19775,552 30,590710,573 38.879.4 Fetus 7217252 26,17976,728 39.5711.1 Data presented as mean7SD. From Burchfield et al., 1991.

correlated with pretreatment fetal oxyhemoglobin satura- [concentrations converted from lg/mL to ng/mL]. tion (r 5 À0.79). Methamphetamine levels were mea- Effects on the fetus are discussed in Section 3.2.1.1. sured in four fetuses from ewes killed 2 hr after treatment. Methamphetamine levels were reported to be highest in lung 4 placenta 4 kidney 4 intestine 4 liver 4 brain 4 heart. Average fetal organ to plasma 2.2 General Toxicity ratios for methamphetamine were B19 in lung and B6 2.2.1 Human in brain. Methamphetamine increased maternal and fetal 2.2.1.1 Side effects of medication therapy. Adverse blood pressure and reduced fetal oxyhemoglobin satura- events occurring in at least 1% of volunteers in a trial for tion and arterial pH (Section 3.2.1.1). sustained-release Adderall are listed in Table 19 (FDA, Stek et al. (1993) implanted catheters in seven mixed- 2001; PDR, 2003b). The events considered common and breed ewes and their fetuses at GD 107 (early third drug-related by the FDA (2001) included fever, loss of trimester). One mg/kg bw methamphetamine HCl appetite, emotional lability, insomnia, and nervousness. [purity not specified] was injected into the maternal As noted in Table 20, a dose-related pattern was noted for vena cava over 1.5 min and samples of maternal and fetal anorexia, weight loss, and insomnia. blood (n 5 5 maternal-fetal pairs) and amniotic fluid Adverse effects listed in drug labels for d,l-ampheta- (n 5 3 fetuses) were taken at timed intervals. Maternal mine, d-amphetamine, and methamphetamine include serum methamphetamine peaked at 2 min (the first heart palpitations, tachycardia, elevated blood pressure, sampling time) at 29007120 ng/mL (mean7SEM), with over-stimulation, restlessness, dizziness, insomnia, eu- half-time disappearance from maternal serum of about phoria, dyskinesia, dysphoria, tremor, exacerbation of 30 min [estimated from a graph]. Fetal serum metham- motor and phonic tics and Tourette disorder, headache, phetamine peaked at 5 min at 1900 ng/mL, after which dry mouth, unpleasant taste, diarrhea, other gastrointest- elimination from serum paralleled elimination in the inal disturbances, constipation, anorexia, weight loss, dam at concentrations 200–300 ng/mL lower than ma- urticaria, impotence, and changes in libido (PDR, ternal concentrations. Amniotic fluid methamphetamine 2003a–e). The labels report that psychotic episodes are peaked at 120 min at 10507100 ng/mL, after which rare at recommended doses, but state that behavior concentrations were similar to those in maternal serum. disturbance and thought disorder may be exacerbated in Twenty-four hr after the methamphetamine dose, mater- psychotic patients. Isolated cases of cardiomyopathy nal serum methamphetamine was 150720 ng/mL, fetal after chronic amphetamine use were reported in labels serum methamphetamine was 210730 ng/mL, and for d,l- and d-amphetamines (PDR, 2003b,d,e). The amniotic fluid methamphetamine was 3507140 ng/mL American Academy of Pediatrics (2001) also reported

Birth Defects Research (Part B) 74:471–584, 2005 DOI 10.1002/bdrb AMPHETAMINE AND METHAMPHETAMINE 489 Table 19 Adverse Events in Volunteers Taking d,l-Amphetamine

% with adverse events

Body system Effect d,l-Amphetamine (n 5 374) Placebo (n 5 210)

General Abdominal pain (stomachache) 14 10 Accidental injury 3 2 Asthenia (fatigue) 2 0 Fever 5 2 Infection 4 2 Viral infection 2 0 Digestive system Loss of appetite 22 2 Diarrhea 2 1 Dyspepsia 2 1 Nausea 5 3 Vomiting 7 4 Nervous system Dizziness 2 0 Emotional lability 9 2 Insomnia 17 2 Nervousness 6 2 Metabolic/nutritional Weight loss 4 0 From FDA (2001) and PDR (2003b).

Table 20 CNS symptoms include agitation, hallucinations, Dose-Response Relationship of Some Common psychosis, and seizures. Intracerebral hemorrhage and d,l-Amphetamine Adverse Events ischemic strokes have been reported with amphetamine compounds. Cerebral vasculitis as well as cerebral artery Incidence (%) spasm and occlusion have been implicated as causes for both stroke and hemorrhage (Kase, 1986; Perez et al., Adverse event Placebo 10 mg 20 mg 30 mg 1999; Rothrock et al., 1988). A case of transient cortical Anorexia 11.4 16.3 23.1 26.6 blindness in an infant exposed to methamphetamine has Weight loss 0 1.6 2.5 8.9 been reported (Gospe, 1995). Insomnia 1.9 11.6 19.0 19.4 Cardiovascular signs and symptoms of toxicity include chest pain, palpitations, tachycardia, hyperten- From FDA (2001) and PDR (2003b). sion, and ventricular dysrhythmias. Myocardial infarc- tions associated with amphetamine use are thought to be secondary to direct cardiac toxicity (myocarditis), vasos- pasm, and thrombus formation (Bashour, 1994; Furst jitteriness and social withdrawal as common side effects et al., 1990). Profound hypotension, bradycardia, and associated with stimulant treatment. metabolic acidosis have occurred with massive amphe- Evidence of growth suppression and development of tamine overdoses. Acute and chronic cardiomyopathy tics in amphetamine-treated children is discussed in have been associated with amphetamine use (Call et al., detail in Section 3.1. 1982; Hong et al., 1991). Direct amphetamine toxicity and According to drug labels, use of d,l-amphetamine, indirect hypertension have both been implicated as d-amphetamine, and methamphetamine is contraindi- etiologies. cated in individuals with advanced arteriosclerosis, Severe systemic toxicity reported with overdoses symptomatic cardiovascular disease, moderate-to-severe of amphetamine compounds includes a pattern of hypertension, hyperthyroidism, glaucoma, hypersensi- fulminant hyperthermia, convulsions, disseminated tivity or idiosyncrasy to sympathomimetic amines, intravascular coagulation, hepatocellular damage, rhab- agitation, and history of drug abuse (PDR, 2003a,e). domyolysis, acute renal failure, dysrhythmias, and 2.2.1.2 Overdose symptoms. Amphetamine toxicity refractory hypotension (Ginsberg et al., 1970; Kendrick after overdose in humans is an extension of pharmaco- et al., 1977). logical activity that can affect the central nervous, 2.2.1.3 Drug interactions. Drug labels for d,l-amphe- cardiovascular, pulmonary, and gastrointestinal systems tamine, d-amphetamine, and methamphetamine warn (Albertson et al., 1999). Gastrointestinal toxicity can be against taking monoamine oxygenase inhibitors (MAOI) manifested by nausea, vomiting, and diarrhea. within 14 days of amphetamine or methamphetamine Pulmonary symptoms of toxicity include dyspnea, use (PDR, 2003a–e). Metabolism of amphetamines is hemoptysis, and pleuritic chest pain (Schaiberger et al., slowed by MAOI antidepressants and a metabolite of 1993). Pulmonary sequelae include noncardiogenic furazolidone, thus potentiating the effect associated pulmonary edema and pulmonary hypertension caused with monoamine release from adrenergic nerves. Poten- by endothelial injury, direct spasm, and dysregulation of tiation can lead to potentially fatal hypertensive crisis, mediators of vascular tone. neurological toxicity, and malignant hyperpyrexia.

Birth Defects Research (Part B) 74:471–584, 2005 DOI 10.1002/bdrb 490 GOLUB ET AL. Amphetamines and methamphetamine could enhance described as likely markers of damage to the striatal tricyclic or sympathomimetic drug activity resulting in dopaminergic system, but it was noted that there were no marked and sustained increases in brain d-amphetamine obvious behavioral symptoms. Dopaminergic effects in levels and potentiation of cardiovascular effects. humans receiving therapeutic doses of amphetamines The drug label for methamphetamine states that are not known. phenothiazines could inhibit stimulatory effects of 2.2.2 Experimental animal. Symptoms of acute amphetamines (PDR, 2003c). In addition, the drug label amphetamine toxicity in rats and mice include hyper- warns that dietary regimen changes associated with activity, piloerection, salivation, and hyperpnea (NTP, methamphetamine use could alter insulin requirements 1991). Dilated pupils and convulsions have also been in diabetics. reported (NIOSH, 2002). Symptoms of acute metham- Additional drug interactions are discussed in labels phetamine toxicity in rodents were reported as excite- for d,l- and d-amphetamine. Gastrointestinal acidifying ment, convulsions, and changes in seizure threshold agents (e.g., guanethidine, reserpine, glutamic acid HCl, (NIOSH, 2001). The oral LD50 for d-amphetamine in ascorbic acid, fruit juices) inhibit absorption of amphe- rats is reported at 96.8 mg/kg bw. LD50 values for tamines and urinary acidifying agents (e.g., ammonium d,l-amphetamine and methamphetamine in various chloride, sodium acid phosphatase, methenamine) in- species are summarized in Tables 21 and 22. [The Expert crease ionization and therefore urinary excretion. Both Panel questions the accuracy of mouse LD50 values groups of acidifying agents lower blood levels and reported for d,l-amphetamine (Table 21). It does efficacy of amphetamines. The opposite situation occurs not make sense that the oral LD50 would be less with agents that alkalinize the gastrointestinal tract (e.g., than the i.v. LD50, nor that the s.c. LD50 would be lowest sodium bicarbonate, antacids) or urine (e.g., acetazola- of all. The discrepancies may be due to the use of mide, some thiazides). The effects of amphetamines may animals of different ages. In addition, although the be potentiated by propoxyphene and inhibited by LD50 for oral administration of d,l-amphetamine in chlorpromazine, haloperidol, and lithium carbonate. mice is listed as 24 mg/kg in Table 21, the NTP Amphetamines could inhibit adrenergic blockers, seda- carcinogenicity studies (Dunnick and Eustis, 1991; tive effects of antihistamines, and hypotensive effects of NTP, 1991) described in Section 2.4.2 used an estimated antihypertensive drugs. Amphetamines could potentiate 19 (female) and 30 (male) mg/kg bw/day dose for 103 analgesic effects of meperidine, the adrenergic effect of weeks and had no differences in survival of treated norepinephrine, and the anticonvulsant effects of phe- groups.] nobarbital and phenytoin. Intestinal absorption of etho- Continuous i.v. infusion of rats with amphetamines suximide could be delayed by amphetamines. 1 mg/kg bw/hr for 12 days resulted in reduced brain d-Amphetamine could interact with pimozide, which norepinephrine and cardiac catecholamine levels and diminishes its stimulant effects, and with various increased motor activity and stereotypic behaviors, antihypertensive agents (Markowitz et al., 1999). Effects manifested as grooming, scratching, rearing, limb flicks, of d-amphetamine can also be diminished by propranolol and biting (NTP, 1991). Behavioral effects were reversed and metoprolol. Based on a case report, reactions with upon cessation of dosing. Amphetamine effects clonidine can result in syncope, hypotension, bradycar- after repeated oral dosing are reported below in the dia, and sedation. d-Amphetamine can interact with description of an NTP study. Repeated parenteral guanethidine and bretylium [drugs that are no longer dosing of rodents with methamphetamine has caused clinically relevant]. excitement, weight loss or decreased weight gain, 2.2.1.4 Drug abuse. Abuse of amphetamines and alteration in classical conditioning, changes in pancreas, methamphetamine is associated with development of and hemorrhage (NIOSH, 2001). Higher doses [not tolerance, psychological dependence, and social disabil- specified] of amphetamines were reported to induce ity (PDR, 2003a,c,e). There are reports of some patients greatly increasing dosages above recommended levels. Abrupt withdrawal after prolonged intake of high doses Table 21 can result in extreme fatigue, mental depression, and LD50 Values for d,l-Amphetamine sleep electroencephalogram (EEG) changes. Toxicity associated with chronic intoxication can include derma- Species Exposure route LD50 (mg/kg bw) toses, insomnia, irritability, hyperactivity, and personal- a ity changes. Although rare with oral intake, psychosis Mouse Oral 24 Intraperitoneal 13a that is indistinguishable from schizophrenia has been a Intravenous 31.8 reported. Repetitive and stereotypic behavior that can Subcutaneous 7a advance to self-injurious behavior has been observed Rat Oral 55 with high or repeated dosing with amphetamines (Kita Intraperitoneal 125 et al., 2003). Animal models used to investigate self- Subcutaneous 160 injurious behavior in humans are discussed below in Dog Oral 23 Section 2.2.2. Intravenous 6 Studies in humans who abused methamphetamine and Guinea pig Intraperitoneal 50 possibly other substances showed persistent reductions Subcutaneous 105 in striatal dopamine transporter as observed by imaging Rabbit Intravenous 22 techniques (Madras et al., 2002; Kita et al., 2003) and Data from NIOSH (2002) for all routes and from NTP (1991) for deficits in striatal dopamine, tyrosine hydroxylase oral route. a activity, and dopamine transporters during autopsy (Kita The Expert Panel questions the accuracy of the mouse LD50 et al., 2003). The effects in autopsy subjects were data, as explained in text above.

Birth Defects Research (Part B) 74:471–584, 2005 DOI 10.1002/bdrb AMPHETAMINE AND METHAMPHETAMINE 491 Table 22 dopamine antagonists provides evidence of dopaminer- LD50 Values for Methamphetamine gic involvement. NTP toxicity studies were reviewed in detail because Species Exposure route LD50 (mg/kg bw) reproductive organs were examined in a subchronic and a carcinogenicity study and growth was measured in the Mouse Intraperitoneal 15 Intravenous 6.3 subchronic study. The carcinogenicity study is described Subcutaneous 7.56 in Section 2.4.2. Rat Subcutaneous 10.93 The NTP (1991) conducted 14-day and 13-week studies Guinea pig Oral 90 to determine the toxicity of d,l-amphetamine sulfate in F344/N rats and B6C3F1 mice. Rats and mice were Data from Registry of Toxic Effects of Chemical Substances assigned randomly to groups and administered U.S.P (RTECS) database (NIOSH, 2001). grade d,l-amphetamine sulfate (99% purity) through feed. Drug stability and concentration in feed were verified. Survival data were analyzed using Kaplan brain lesions such as hyperemia, hemorrhage, and glial Meier, Cox, or Tarone life table test methods. Continuous proliferation in monkeys and enlarged and chromatolytic data were analyzed using Dunn, Shirley, or Jonckheere medulla oblongata neurons in cats (Dunnick and Eustis, tests. 1991). In the 14-day study, 7-week-old rats (five/sex/group) A review by Kita et al. (2003) reported that parenteral were fed diets containing 0, 47, 94, 188, 375, or 750 ppm dosing with methamphetamine or amphetamine concen- d,l-amphetamine sulfate. Nine-week-old mice (five/sex/ trations exceeding therapeutic levels has resulted in group) were fed diets with 0, 125, 250, 500, 1000, or persistent regional (mainly striatal) depletion of brain 2000 ppm d,l-amphetamine sulfate. Animals were ob- dopamine levels in rats, mice, cats, guinea pigs, and served daily and weighed before, during, and after the monkeys, and decreased tyrosine hydroxylase activity in study. After killing, animals were necropsied and organ rats and cats. Decreased numbers of dopamine transport weights were measured. Major organs were fixed in 10% pumps were reported in mice and rats. Catecholamine neutral buffered formalin for a histological analysis. turnover was found to be increased in rats. Reductions in Included among organs analyzed were mammary gland, brain norepinephrine levels were observed in monkeys prostate, testis, seminal vesicle, ovary, and uterus. Doses and cats. Decreases in serotonin concentration and at which histological evaluation was conducted in each tryptophan hydroxylase activity were reported in rats species are discussed below. and cats and reduced numbers of serotonin transporter In the 14-day study, all rats survived. Final mean pumps were reported in one rat study. Some recovery of body weights in the 375 and 750 ppm groups were 7–9% neurotransmitter levels was observed at lower doses in lower for males and 5–16% lower for females compared monkeys (B2.0 mg/kg bw) and rats (r12.5 mg/kg bw), to controls. Reduction in final body weight was but at higher doses, the effects persisted for months. statistically significant for females at 750 ppm. Feed Studies examining effects of amphetamines on neuro- intake was reduced in groups exposed to Z94 ppm chemical indices in developing animals are discussed in during the first week of the study and was marginally detail in Section 3.2.3. reduced in males from the 750 ppm group during the Neuroanatomical evidence of amphetamine-induced second week of the study. Hyperactivity [determined by toxicity was also reported. Swollen nerve fibers and general observations, not standard measures] was terminals were observed in rats and mice dosed with observed in rats exposed to Z375 ppm. According to amphetamine or methamphetamine (Kita et al., 2003). the results section, absolute heart weight was reduced in Methamphetamine treatment has resulted in reductions females exposed to Z375 ppm. [The only organ weight of brain dopaminergic axons and axon terminals in rats effect data shown in Table J1 of the NTP study are and monkeys and serotonergic axons and axon terminals for liver. According to that table, relative (to body in rats (McCann and Ricaurte, 2004). Development of weight) liver weights were increased in males exposed neurotoxicity depends on dose, number of exposures, to 750 ppm and females exposed to Z188 ppm.] Histo- intervals between dosing, and duration of neuron pathologic evaluation was conducted on animals from exposure (McCann and Ricaurte, 2004). It has been the control and 750 ppm groups; no lesions were reported that administration of reuptake blockers with observed [data not shown]. amphetamines prevents neuronal damage (Kita et al., Four males in the 1000 ppm group and one male in the 2003; McCann and Ricaurte, 2004). Apoptosis was 2000 ppm group died in the 14-day mouse study. reported as a possible mechanism of neurotoxicity in According to study authors, the deaths were not clearly mice and rats (Kita et al., 2003). related to treatment. Weight loss occurred in males A review by Kita et al. (2003) discusses the animal exposed to Z500 ppm. Mean final body weights of literature and modeling that supports self-injurious females exposed to Z250 ppm were significantly lower behaviors in humans. Kita et al. (2003) state ‘‘The than controls by 12–13%. Feed intake was similar in all acute appearance of self-injurious behavior associated groups, although it was noted that the 2000 ppm males with administration of amphetamines to rodents may scattered their feed. Mice in the 1000 ppm group were accurately predict the later neurotoxicity and at the same either hyperactive or lethargic and hyporesponsive. time serve as a model for the study of this disabling According to the results section, relative liver weights symptom in humans.’’ Animal literature does not were increased in males exposed to Z250 ppm and provide a neurochemical basis for self-injurious behavior females from the 2000 ppm group. Other significant although it points to involvement of central dopamine organ weight changes listed in Table J3 of the NTP report neurons. Reversal of some self-injurious behaviors by included decreased absolute brain and thymus weight

Birth Defects Research (Part B) 74:471–584, 2005 DOI 10.1002/bdrb 492 GOLUB ET AL. and increased relative kidney weights in males of the died, and seven females and six males from the 2000 ppm 2000 ppm group; absolute heart weight was increased in group died. Mean final body weights of all treated females exposed to Z125 ppm. Histopathologic evalua- groups were lower than controls. At doses Z250 ppm, tion was conducted in controls, the 2000 ppm group, and body weights were 18–30% lower than controls in males males in the 1000 ppm group. No lesions were observed and 13–19% lower than controls in females. All treated [data not shown]. mice were hyperactive and severity increased according The 13-week studies were conducted according to FDA to dose. Fighting was observed in males exposed to Good Laboratory Practice (GLP) procedures. At 7–8 Z250 ppm. According to study authors, increases in weeks of age, 10 rats/sex/dose were fed diets containing relative organ weights were due to reduced body 0, 47, 94, 188, 375, or 750 ppm d,l-amphetamine sulfate. weights. [Decreases in absolute organs weights were At 8–9 weeks of age, 10 mice/sex/group were fed diets also consistent with reduced body weights.] A histolo- containing 0, 125, 250, 500, 1000, or 2000 ppm d,l- gical evaluation was conducted in the control and amphetamine sulfate. Doses were based upon results 2000 ppm groups, males in the 1000 ppm group, and all observed in the 14-day studies. Animals were observed mice that died during the study. No treatment-related daily and weighed before, during, and after the study. lesions were observed [data not shown]. After killing, animals were necropsied and organs were weighed. Major organs were fixed in 10% neutral buffered formalin for histological analysis. Included 2.3 Genetic Toxicity among organs analyzed were mammary gland, prostate, Details of study protocols and results for in vitro testis, ovary, and uterus. Doses at which histologic genetic toxicity testing of d,l-amphetamine sulfate are evaluation was conducted in each species are discussed listed in Table 23. Mutagenicity testing in S. typhimurium below. Thymus and spleen were also examined in some produced negative results except in strain TA98 in the rats from lower dose groups. presence of hamster but not rat S9 activation (NTP, 1991). All rats survived in the 13-week study. Final body The response was classified as equivocal by the NTP. weights were significantly reduced in both sexes of all Negative results were obtained for sister chromatid treated groups. Final mean body weights in the 188, 375, exchange and chromosomal aberration in Chinese and 750 ppm groups were 11, 18, and 38% lower in males hamster ovary cells. Although chromosomal aberrations and 15, 26, and 32% lower in females. Feed intake by the were increased in the first trial with metabolic activation, 750 ppm males was 20% lower than controls. Hyper- the result was not repeated in two additional trials and activity was observed in all dose groups and severity the NTP classified the results as negative. An unpub- increased according to dose. According to study authors, lished study reported in the FDA Pharmacology Review organ weight changes were due to body weight reduc- of Adderall (FDA, 2001) reported that Adderall was tions. [The study authors’ statement is confirmed by negative for mutagenicity in an E. coli reversion test at dose-related reductions in absolute weight and in- concentrations up to 5000 mg base/plate, with and with- creases in relative organ weights observed in many out S9 activation. organs. The exceptions that appeared to be dose-related Details of study protocols and results for in vivo were an increase in absolute brain weight (Z94 ppm genetic toxicity assays are listed in Table 24. The FDA females) and decreases in relative (to body weight) Pharmacology Review (FDA, 2001) indicated that Ad- liver (750 ppm males) and thymus (Z94 ppm males, derall was negative in the mouse micronucleus test at Z375 ppm females) weights.] A histopathological ex- doses up to 10 mg/kg bw with a mean7SD ampheta- amination was conducted in controls and the 750 ppm mine plasma concentration up to 433.4772.4 ng/mL. The group. No treatment-related lesions were observed [data FDA review, however, also cited a published study not shown]. demonstrating an increase in micronucleated polychro- In the 13-week mouse study, eight males of the matic erythrocytes in mice given 2 oral doses of 500 ppm group died, three males of the 1000 ppm group Z12.5 mg/kg bw d,l-amphetamine (Tariq et al., 1987).

Table 23 Results of In Vitro Genetic Toxicity Testing of d,l-Amphetamine

Testing with metabolic Reference Concentration activation Species/cell type/strain Endpoint Results

NTP (1991) 100–10,000 mg/plate Yes Salmonella typhimurium Mutagenicity at the m Mutagenicity in TA98, TA98, TA100, TA1535, histidine operon with rat but not hamster and TA1537 S9 metabolic activation; 2 in all other strains; results judged equivocal by NTP FDA (2001) r5000 mg/plate Yes Escherichia coli Mutagenicity 2 NTP (1991) 50–1600 mg/L Yes Chinese hamster Sister chromatid 2 ovary cells exchange NTP (1991) 300–1000 mg/L Yes Chinese hamster Chromosomal 2 ovary cells aberration 2, No change; m, statistically significant increase.

Birth Defects Research (Part B) 74:471–584, 2005 DOI 10.1002/bdrb AMPHETAMINE AND METHAMPHETAMINE 493

Table 24 Results of In Vivo Genetic Toxicity Testing of d,l-Amphetamine

Reference Species Dose in mg/kg bw (route) Cell type Endpoint Results

FDA (2001) Mouse 2.5–10 (oral) Bone marrow erythrocytes Micronucleus 2 Tariq et al. (1987) Mouse 6.25–25.0 (2 oral doses) Bone marrow erythrocytes Micronucleus m at Z12.5 mg/kg bw 2, No change; m, statistically significant increase.

2.4 Carcinogenicity development of adenoma. An increase in cataracts and 2.4.1 Human. No data were identified. retinal atrophy in females was believed to be due to 2.4.2 Experimental4animal. The NTP (Dunnick excessive light exposure resulting from cage placement. and Eustis, 1991; NTP, 1991) examined the carcinogeni- Because amphetamines can cause hyperthermia, pupil city of d,l-amphetamine sulfate in F344/N rats and dilation, and increased activity, it was also postulated by study authors that the drug could have contributed B6C3F1 mice in a study conducted according to FDA GLP procedures. Rats and mice were randomly assigned indirectly to the formation of eye lesions. An increase in bone myelofibrosis was thought to be due to loss of to groups and administered U.S.P grade d,l-ampheta- adipose tissue secondary to weight loss. The study mine sulfate (99% purity) through diet. Fifty rats (7–8 authors concluded that none of these lesions were weeks old) and mice (8–9 weeks old)/sex/group were directly caused by drug exposure. Dose-related reduc- given diets containing 0, 20, or 100 ppm d,l-amphetamine tions were observed for neoplasms in adrenal gland, sulfate for 103 weeks. Study authors estimated doses of pituitary gland, mammary gland, and uterus. B0, 1, and 5 mg/kg bw/day in rats. Doses were In the mouse study, there were no differences in estimated at 0, 4, or 30 mg/kg bw/day in male mice survival of treated groups. Mean body weights of males and 0, 3, or 19 mg/kg bw/day in female mice. The dose were reduced by 10–36% in the 100 ppm group and 10– estimate in high-dose male mice may have been affected 19% in the 20 ppm group. Mean female body weights by feed scatter. Dose selection was based upon results of were reduced by 10–34% in the 100 ppm group and 10– the 13-week studies described in Section 2.2.2. Drug 19% in the 20 ppm group. Final body weights in the stability and concentrations in feed were verified. 100 ppm groups were 60–70% of control values. [Based Animals were observed daily and weighed before, on information presented in NTP Tables F3 and F4, during, and after the study. After killing, all animals CERHR estimated that compared to controls, body were necropsied and major organs from animals in all weights of animals in the 100 ppm group were reduced dose groups were fixed in 10% neutral buffered formalin by B25%; this suggests that maximum tolerated dose for a histological analysis. Included among organs was exceeded.] Average feed consumption was not analyzed were mammary gland, prostate, testes, seminal reduced in treatment groups, but the authors stated that vesicles, ovaries, and uterus. Survival data were ana- the values were not corrected for scatter, which may have lyzed using Kaplan Meier, Cox, or Tarone life table test led to overestimation in males. Seventeen females in the methods. Continuous data were analyzed using the 100 ppm group were mistakenly bypassed during sacri- Dunn, Shirley, or Jonckheere test. Tumor incidence data fice at the end of the study. They received water but no were assessed using logistic regression analysis, Cox life feed for 5 days, at which time they were killed, table test, Fisher Exact test, or Cochran-Armitage trend necropsied, and apparently included in histological test. analyses. A slight increase in thyroid follicular cell No difference in survival was observed in the rat neoplasms in male mice was not considered treatment- carcinogenicity study. Body weights were 10–34% lower related by study authors due to lack of precursor in female rats exposed to 20 or 100 ppm and male rats follicular cell hyperplasia, a lack of increase in follicular exposed to 100 ppm. [Compared to control animals, cell adenomas in female mice, and a reduced incidence of CERHR calculated body weight reductions of 11% in follicular cell hyperplasia in female mice. Treated mice 20 ppm males, 10% in 20 ppm females, and 20% in displayed dose-related reductions in neoplasms of the 100 ppm females, based on body weights presented in liver, pituitary, lung, and harderian gland. Ovarian Table F1 and F2 of the NTP report. The Panel notes that atrophy was increased in females of the 100 ppm group, body weight reductions 410% indicate that maximum but it was not stated if the effect was statistically tolerated dose was exceeded.] Feed intake of the significant. The incidences of ovarian atrophy were 14/ 100 ppm females was 84% of control levels. An increase 49 in the control group, 12/48 in the 20 ppm group, and in testicular interstitial cell adenomas in male rats was 25/46 in the 100 ppm group [P 5 0.0054, v2 performed by not considered to be treatment-related by study authors CERHR]. Study authors postulated that ovarian atrophy due to the common occurrence of that tumor. [The may have been due to reduced body weight rather than a results section listed the interstitial cell adenoma direct drug effect. incidence as 34/50 in controls, 43/50 in the 20 ppm The study authors concluded that there was no group, and 48/50 in the 100 ppm group. Table A3 of the evidence of carcinogenic activity of d,l-amphetamine NTP reported listed the incidence as 43/50 in controls, sulfate in male or female F344/N rats or B6C3F1 mice. 43/50 in the 20 ppm group, and 48/50 in the 100 ppm In a review of the NTP studies, the FDA noted that the group.] Thyroid follicular cell adenomas were not dose- high dose in rats (5 mg/kg bw 5 30 mg/m2) is equivalent related and hyperplasia was not observed before to the maximum advised dose for children, ‘‘30 mg/

Birth Defects Research (Part B) 74:471–584, 2005 DOI 10.1002/bdrb 494 GOLUB ET AL. 25 kg Â 25 5 30 mg/m2,’’ and that the high doses in male susceptibility). The authors questioned whether metabo- (30 mg/kg bw 5 90 mg/m2) and female (19 mg/kg lizer status could explain the mortality differences bw 5 57 mg/m2) mice are only three and two times the between the strains inasmuch as the inter-strain differ- maximum advised children’s dose (FDA, 2001). ence in mortality was also seen among treated males, which do not manifest the poor-metabolizer phenotype. A second study suggested a possible increased 2.5. Potentially Susceptible Populations susceptibility in rats with poor or intermediate metham- 2.5.1 Pharmacogenetics. CYP2D6 is involved in phetamine metabolizer phenotypes (Vorhees et al., 1999). the aromatic hydroxylation of amphetamines in humans Dark Agouti female rats (poor metabolizers) that were and rats (Kraemer and Maurer, 2002). The gene for s.c. injected with 15 mg/kg bw methamphetamine on CYP2D6 is located on the long arm of human chromo- PND 11–20 performed less effectively on the Morris some 22. Polymorphisms for CYP2D6 are associated with Maze test of spatial learning and memory and had at least 12 variants that alter enzyme activity (DeVane, greater acoustic startle amplitudes compared to female 1994; Bertilsson et al., 1997). People with the usual Sprague–Dawley rats receiving the same treatments. CYP2D6 activity are called extensive metabolizers and Treated male Dark Agouti rats (intermediate metaboli- people with lower levels of activity are called poor zers) also displayed increased startle amplitudes com- metabolizers. Poor metabolizer phenotypes occur in 5– pared to treated male Sprague–Dawley rats. Details of 8% of Caucasians and 2–10% of blacks and Asians. There this study are included in Section 3.2.3. is considerable variation within racial groups. For Inhibition of methamphetamine 4-hydroxylation in example, there is a higher incidence in African-Amer- rats was reported to enhance stereotyped behavior in icans (8.5%) than in Zimbabweans (1.8%) of one of the rats, but reviewers noted that metabolic differences in inactive CYP2D6 alleles and up to 29% of Ethiopians humans compared to rodents complicate the application carry duplicated or multiduplicated CYP2D6 alleles. of the study findings to humans with inborn CYP2D6 CYP2C3 is involved in amphetamine deamination deficiencies (Sellers et al., 1997). reactions in rabbits (Kraemer and Maurer, 2002). [The 2.5.2 Sex-related differences. Amphetamine enzyme is not a member of the of the CYP2C human pharmacokinetic parameters in men and women were gene family (Hines and McCarver, 2002), but the Expert compared in an FDA review (2001). Higher systemic Panel finds it reasonable to assume that one or more exposure of women was attributed to weight differences. members of the CYP2C family (CYP2C8, CYP2C9, According to the FDA, the difference was reduced when CYP2C18, or CYP2C19) may be involved in deamina- the doses were normalized for body weight. Table 25 tion reactions in humans.] The gene family is highly outlines the dose-normalized values for d-amphetamine polymorphic with functional variants identified at in men and women administered d,l-amphetamine. CYP2C9 and CYP2C19 loci. Two of three known Some studies suggest that male mice are more defective alleles of CYP2C9 result in decreased activity; susceptible than female mice to methamphetamine- CYP2C19 polymorphisms resulting in slow metabolizer induced depletion of striatal dopamine, and some but phenotypes are present in 3% of Caucasians and nearly not all studies suggest that estrogen may act as a 20% of Asians (Anzenbacher and Anzenbacherova, neuroprotectant (Fukumura et al., 1998; Gao and Dluzen, 2001). 2001; Kita et al., 2003). Slight effects were replicated in Vorhees et al. (1998) performed a study, sponsored by rats (Kita et al., 2003, but one study demonstrated that an NIH grant, to explore the effects of methamphetamine methamphetamine-induced reductions in striatal dopa- metabolism on developmental neurotoxicity, summar- mine and serotonin were similar in male and female rats ized in Section 3.2.3. Two strains of rat were used: ACI when hyperthermia was maintained at a consistent level Black (Dark) Agouti, the females of which are poor in both sexes (Fukumura et al., 1998). metabolizers of debrisoquine, and Sprague–Dawley, 2.5.3 Age-Related differences. Ontogeny of he- which are extensive metabolizers of debrisoquine. In patic CYP2D6 in humans was reviewed by Hines and rats, debrisoquine oxidation differences are due to McCarver (2002). One study found no evidence of polymorphisms of CYP2D1, believed also to metabolize hepatic CY2D6 expression in 11–13-week-old fetuses. methamphetamine. The authors tested the hypothesis Another study reported that CYP2D6 protein and mRNA that there would be differences between these strains in were detected in 30% of liver samples from fetuses o30 behavioral testing of adult females after administration weeks old at 5% of adult levels. In fetuses 430 weeks of methamphetamine 30 mg/kg bw s.c. twice daily on old, CYP2D6 protein and mRNA were detected in 50% of postnatal day (PND) 11–20. The test battery consisted of liver samples and activity was 15% of adult levels. an acoustic startle test, straight-channel swimming, and a CYP2D6 protein was reported to increase at birth, Morris maze, which is a test of spatial memory. The regardless of gestational age (i.e., time period of gestation behavioral test results did not differ by strain, arguing completed at time of delivery), and reach 50–75% of against the hypothesis. There was, however, a large adult levels during the neonatal period. difference in mortality during the treatment period: Hines and McCarver (2002) also reviewed the onto- 59.4% of ACI Black Agouti and 18.8% of Sprague– geny of hepatic CYP2C enzymes. CYP2C8, CYP2C9, and Dawley female offspring died (Po0.01, Fisher test). CYP2C18 transcripts were detected in human fetal livers The authors postulated that there may have been an at 10% of adult levels, but there was no evidence of increased sensitivity to methamphetamine toxicity in the CYP2C activity in 16–40-week-old human fetuses. ACI Black Agouti strain, but that a difference in CYP2C expression seems to be activated at birth and is sensitivity would have been obscured by testing only independent of gestational age. Activity in human the survivors of the treatment period (i.e., that the cohort neonates is B30% of adult levels and remains constant of survivors were not representative of overall strain up to 1 year of age.

Birth Defects Research (Part B) 74:471–584, 2005 DOI 10.1002/bdrb AMPHETAMINE AND METHAMPHETAMINE 495 Table 25 Dose-Normalized Comparison of d-Amphetamine Pharmacokinetic Parameters in Men and Women Given d,l-Amphetamine

Cmax/dose AUC0–N/dose Dosing regimen Sex Dose (mg/kg bw) (ng/mL)/(mg/kg bw) (ng-hr/mL)/(mg/kg bw)

Single dose of 30 mg Male 0.354 106.6 2237 Female 0.486 104.8 1811 30 mg/day for 7 days Male 0.398 145.0 2109 Female 0.493 154.2 2063 Data from 10 subjects/sex; from FDA review (FDA, 2001).

Table 26 Dose-Normalized Comparison of d-Amphetamine Pharmacokinetic Parameters in Children and Adults Given d,l-Amphetamine Repeated Dosing at 30 mg/day

Cmax/dose AUC0–N/dose Population Dose (mg/kg bw) (ng/mL)/(mg/kg bw) (ng-hr/mL)/(mg/kg bw)

Adults (n 5 19) 0.45 148.7 2064 Children (n 5 20) 0.88 101.1 1550 From FDA review (FDA, 2001).

Table 5 and Table 6 in Section 2.1.1.2 illustrate some d-amphetamine, whereas men on the high dose of d- differences in d,l-pharmacokinetic parameters in adults amphetamine were rated as more fidgety. Other tests of and children. At equivalent doses, the half-life is shorter, vigilance, learning, and speech communication showed but systemic bioavailability is greater in children similar effects in boys and men, although the hyperactive compared to adults. The FDA attributed lower body boys had more exaggerated responses to d-amphetamine weight to the increased bioavailability in children (FDA, than normal boys or men in many domains. 2001). When normalized to body weight, children had Alhava and Mattila (1974) examined heart and brain lower Cmax and AUC values than adults. Table 26 distribution of 50 mg/kg bw d,l-amphetamine adminis- outlines the dose-normalized values for d-amphetamine tered by i.p. injection to adult male or developing (either in children and adults administered d,l-amphetamine. sex) NMRI mice at the ages of 3–5, 13–15, or 32–35 days. Rapoport et al. (1980) performed a study at NIH that In adults and developing mice Z13 days of age, evaluated differences between boys and men in response amphetamine peaked within 30 min in both brain and to a single dose of d-amphetamine. The study included heart, and levels were higher in brain than in heart. In 14 normal boys and 15 hyperactive boys (some of whom infant mice (3–5 days old), amphetamine levels peaked in had been on medication) with a mean age of about 10 brain within 2 hr, were initially lower than heart levels, years in both groups (SD 5 2.1). The boys were given which peaked within 30 min, and were much lower than B16 mg d-amphetamine. The adult group consisted of 15 adult brain levels. When doses were adjusted according men given a mean of 34 mg d-amphetamine and 16 men to surface area (i.e., twice the dose given to infant vs. given a mean of 17 mg d-amphetamine (SD 5 1–2 mg to adult mice), peak brain levels were similar in adults and adjust for weight differences; the high dose was about infants; the time to reach peak ranged from 30 min to 2 hr. the same as the boys’ dose on a mg/kg bw basis). The Heart levels of amphetamines were higher in infants than mean age7SD of the men was 2273 years. Baseline in adults when doses were adjusted to surface area. In measures were performed on a Monday. On Wednesday, vivo results were not replicated in in vitro experiments subjects were randomized to receive either amphetamine that measured distribution in heart and brain slices. The or placebo followed by a battery of tests. On Friday, each study authors suggested that the increased time to reach subject received the opposite treatment (amphetamine or peak in infant brains may have been a result of placebo) followed by the same tests. Motor activity, circulatory factors and brain immaturity. evaluated using a sensor attached to the trunk and A number of studies compared neurological endpoints expressed by comparing the d-amphetamine to the in adult versus developing experimental animals ex- placebo sessions, was decreased 44% in hyperactive posed to amphetamine or methamphetamine. boys, 24% in normal boys, and 9% in men given the low Laviola et al. (2002) i.p. injected male and female adult dose of d-amphetamine. There was no effect of the high and peri-adolescent (30–45-day old) mice (eight or nine/ dose of d-amphetamine on motor activity in men. Adults sex/group; strain not specified) with saline or 2 mg/kg reported feeling euphoria and feeling less tired on d- bw amphetamine [purity and enantiomer not specified], amphetamine whereas boys reported feeling cranky and returned them to their cages for 100 min, subjected more tired. Interviewer behavior evaluations (blind to them to mild stress by removal of sawdust from the treatment condition) of hyperactive boys rated them cage, and then killed them 20 min later for measurement less hyperactive, less fidgety, and less silly on of plasma corticosterone levels. Amphetamine treatment

Birth Defects Research (Part B) 74:471–584, 2005 DOI 10.1002/bdrb 496 GOLUB ET AL. significantly increased plasma corticosterone levels in [Although the authors present values as percentage male and female peri-adolescents but had no effect on of neurons examined, it actually seems that the data adult mice. are presented as percentage of animals examined. Lanier and Isaacson (1977) treated Long–Evans Considering that extracellular recordings were taken, hooded rats with d,l-amphetamine 0, 2, 5, or 10 mg/kg the Panel noted that firing would represent a sum bw i.p. during three different juvenile age periods (PND of all surrounding neurons and data therefore reflect 18–22, PND 34–38, PND 45–49) and during adulthood. information from each animal.] To determine route of An increase in locomotor activity in open field testing exposure effects, d-amphetamine was given by i.v. occurred at all ages except PND 34–38. Lesioning the injection to 12 pups and 4 adult males at doses of hippocampus on PND 21 or 26 resulted in greater 0.25, 0.5, 1.0, and 2.0 mg/kg bw administered sequen- sensitivity to the lower doses of amphetamine. The tially at 1-min intervals. Consistent with i.p. data, i.v. authors postulated that the hippocampus becomes administration of d-amphetamine resulted in atypical functionally capable of suppressing locomotor activity responses that differed significantly from adult during the PND 34–39 interval, accounting for the lack of responses during the first 2 weeks of life. [Results were inhibition when amphetamine was given during the only shown for an 8-day-old rat and an 18-day-old rat.] PND 18–22 period. By PND 45, they argued, the Injection i.p. or i.v. with apomorphine significantly catecholamine system had matured sufficiently so that inhibited dopaminergic neuron activity in all age groups, stimulatory effects of amphetamine overrode hippocam- thus demonstrating that dopamine autoreceptors are pal suppression. functional in rats at birth. Haloperidol, a D1/D2 Gazzara et al. (1986) evaluated a possible dopamine- antagonist, consistently reversed apomorphine-induced mediated age-related difference in behavioral response inhibition at all ages. to amphetamine treatment in rats. They noted that adult Other authors evaluated the age-related differences in rats demonstrate increased locomotor activity when amphetamine and methamphetamine sensitization (Fuji- treated with low doses (0.5–1.5 mg/kg bw amphetamine) wara et al., 1987; Kolta et al., 1990), noting that adult rats and stereotyped behavior (e.g., sniffing, licking) when pretreated with amphetamines display an augmentation given high doses (10 mg/kg bw amphetamine), but that of locomotor response when subsequently rechallenged before 35 days of age, rats exhibit only increased with an amphetamine dose. This sensitization response locomotion, regardless of dose. The increased locomotor does not occur until 3–4 weeks of age. The authors behavior is attributed to dopamine release from neurons suggested that the appearance of mature presynaptic in the nucleus accumbens, whereas the stereotyped dopamine autoreceptors may be necessary for sensitiza- behavior is due to dopamine release in the caudate- tion (Fujiwara et al., 1987) or that maturation of putamen. These authors used electrochemical recordings dopamine reuptake sites is the limiting factor in the from electrodes stereotactically placed in the caudate- development of sensitization (Kolta et al., 1990). [Since putamen of PND 21–22, PND 35–36, and adult Sprague– the publication of these studies, amphetamine sensiti- Dawley rats. After baseline recordings were made, zation in preweanling rats has been demonstrated animals were treated with 0.1 or 1.0 mg/kg bw d- (reviewed in Section 3.2.3); however, the interval amphetamine sulfate (HPLC or reagent grade) s.c. and between sensitizing treatments and subsequent chal- voltametric recordings were made every 10 min. Adult lenge dose must be much shorter than in mature and 35–36-day-old rats showed an increase in caudate- animals; therefore, this model continues to be an putamen dopamine release after treatment with 1.0 mg/ example of age-related differences in response to kg bw amphetamine, whereas 21–22-day-old animals amphetamines.] showed a decrease in dopamine release after treatment Tsuchida et al. (1994, 1996) examined the ontogeny of with either dose (except for a transient elevation after the amphetamine-induced effects on dopamine and its high dose). The authors proposed that amphetamine may metabolites in the striatum of Sprague–Dawley rats. On decrease the firing rate of nigrostriatal neurons in young PND 7 (earlier study only), 14, 21, 28, and 56, striatal rats due to local release of dendrite dopamine from perfusate levels of dopamine, 3,4-dihydroxyphenylacetic neurons in the substantia nigra with inhibition of acid (DOPAC), and homovanillic acid were measured neuronal firing through autoreceptors in these neurons. using an in vivo microdialysis technique and HPLC for This mechanism may be more sensitive in immature up to 180 min in 4–6 male rats/group treated with 4 mg/ animals. kg bw methamphetamine HCl [purity not specified]. Trent et al. (1991) examined age-related differences in Data were analyzed by Kruskal–Wallis test, Mann– nigral dopaminergic neuronal firing after treatment Whitney U-test, ANOVA, and Fisher protected least of Sprague–Dawley rats with d-amphetamine sulfate significant difference test. Baseline levels of dopamine [purity not specified]. In the first set of experiments, did not differ by age but DOPAC and homovanillic acid 5 mg/kg bw d-amphetamine was given by i.p. injection levels increased with age. After amphetamine treatment, to 6 adult male rats or immature rats on PND 1–6, 7–15, dopamine levels were increased in all age groups. At or 16–28 (n 5 11–20 rats/age group). A paradoxical 20 min, the dopamine increase in PND 14 rats was response was noted in 1–6-day-old rats with increased smaller compared to the other age groups, whereas at firing in 45%, reduced firing in 25%, and no effect in 30% 40–60 min, dopamine increases in the PND 56 group of dopaminergic neurons examined. Either inhibition or were larger compared to the other groups. DOPAC levels no effect on neuronal firing was observed in the older were reduced at all ages, but the reduction in PND 14 groups of rats after d-amphetamine treatment. Percent rats was significantly smaller than in the other groups at dopaminergic neurons with inhibited firing in response 40–100 min after treatment. Homovanillic acid levels to amphetamine treatment was 50% at 7–15 days of age, were increased in the PND 14 group but reduced in all 82% at 16–28 days of age, and 100% in adulthood. other age groups.

Birth Defects Research (Part B) 74:471–584, 2005 DOI 10.1002/bdrb AMPHETAMINE AND METHAMPHETAMINE 497 Ehrlich et al. (2002) used Western blot techniques to eight animals. Astrogliosis was observed in three examine DFosB and dopamine-related protein levels in methamphetamine-treated animals. The somatosensory response to i.p. injection of saline, 5 mg/kg bw/day cortex of 20- and 40-day-old animals was unaffected by amphetamine, or 20 mg/kg bw/day cocaine for 7 days in methamphetamine treatment. The study authors postu- adult (60-day-old), peri-adolescent (33-day-old), or post- lated that increased resistance of immature animals may weanling (24-day-old) male CD-1 mouse brain. Each be related to dopaminergic and glutamatergic system dose group contained 10–12 animals. Amphetamine ontogeny. Evidence that these systems are still develop- significantly increased DFosB protein expression in ing in the 20- and 40-day old rats was suggested by cited nucleus accumbens [B225% of control levels] and studies demonstrating that N-methyl-D-aspartate chan- caudate putamen [B175% of control levels] of peri- nels in the frontal cortex reach adult levels after 45 days adolescent, but not adult or preweanling mice. Amphe- of age, in vitro glutamate-stimulated striatal dopamine tamine had no effect on dopamine transporter or release remains below adult levels at 45 days of age, and dopamine and cAMP regulated phosphoprotein striatal dopamine release in response to electrical (DARPP-32) expression. Findings in cocaine-treated stimulation of dopaminergic tracts of the medial fore- groups were similar except that cocaine treatment brain bundle increases three-fold between 33 and 114 significantly increased DFosB levels in caudate putamen days of age. of all groups. Age-related effects of DFosB and dopamine The ontogeny of methamphetamine-induced neuro- transporter expression were also examined. Expression toxicity in the rat was reviewed by Vorhees and Pu of DFosB was similar in preweanling and peri-adolescent (1995). They observed that adults lack behavioral mice but was significantly higher in caudate putamen manifestations of methamphetamine toxicity, despite [B3 times higher] and nucleus accumbens [B2 times decreases in striatal dopamine and forebrain serotonin higher] of adult mice. Dopamine transporter expression and despite gliosis of the striatum, whereas develop- did not vary by age. The study authors concluded that mental exposure produces smaller neurochemical changes in DFosB induction in adolescent mice may be changes and an absence of anatomic disruption while related to increased tendency for addiction in adolescents at the same time being associated with clear behavioral versus adults. alterations when these rats are evaluated as adults. The Pu and Vorhees (1993) examined the effects of authors believed the age-related differences in neuro- methamphetamine exposure on brains of developing toxicity might be related to a shift in sensitivity of and adult Sprague–Dawley rats. Male rats were adminis- tyrosine hydroxylase between PND 40 and 60 and a shift tered methamphetamine HCl at 0 (saline control), 10, or in glial fibrillary acid protein responsiveness between 20 mg/kg bw/dose (expressed as free base) by i.p. PND 20 and 40. Since that review was written, additional injection, four times daily at 20, 40, 60, and 80 days work has suggested that alterations in dorsal striatal of age. Each age group contained 10 rats, two of which protein kinase A activity may underlie the sensitivity of were treated with saline and 4/group of which preweanling rats to amphetamine-induced behavioral were treated with the low and high methamphetamine alterations (discussed in Section 3.2.3). dose. Three days after the last treatment, the animals Fukui et al. (2003) conducted an in vitro study to were killed and brains were sectioned and stained investigate methamphetamine effects on dopamine with antibodies to glial fibrillary acid protein signaling in neostriatal slices from young (14–15- or 21– and tyrosine hydroxylase. Rats treated with 10 and 22-day-old) or adult (6–8-week-old) male C57Bl/6 mice. 20 mg/kg bw/dose methamphetamine displayed The slices were incubated in 100 mM methamphetamine agitation, fine body tremors, and piloerection. One adult [14.9 lg/mL assuming value is for the free base and not in the 20 mg/kg bw/dose group died. In all 60- and the salt] for up to 5 min and an immunoblotting 80-day-old rats treated with either dose of methamphe- technique was used to measure dopamine and cAMP- tamine, tyrosine hydroxylase-positive cells were severely regulated phosphoprotein Mr 32 kDA (DARPP-32) depleted and glial cell hypertrophy and proliferation phosphorylation at the Thr34 and Thr75 sites. In all were dramatically increased in the caudate-putamen. three age groups, methamphetamine increased Thr34- Glial cell proliferation was increased in the ventral lateral DARPP-32 phosphorylation but reduced Thr75-DARPP- region of the striatum in 40-day-old rats treated with 32 phosphorylation. Similar effects were noted by either methamphetamine dose, but tyrosine hydroxylase- incubating slices in SKF81297, a dopamine D1 receptor positive terminals were unaffected. Neither endpoint agonist, but methylphenidate and cocaine only increased was affected by methamphetamine treatment in 20-day Thr34-DAPP-32 phosphorylation in adult animals. old rats. Pu et al. (1996) examined the effects of methamphetamine exposure on glutamate-positive neurons in the 2.6 Summary of General Toxicology and somatosensory cortex of adult and developing Sprague– Biological Effects Dawley rats. Male 20-day-old, 40-day-old, and adult rats 2.6.1 Pharmacodynamics and pharmacokinetics were given 4 i.p. injections of saline or 10 mg/kg bw/ 2.6.1.1 Pharmacodynamics. Amphetamine and dose d-methamphetamine HCl every 2 hr. Eight adult methamphetamine stimulate the CNS by acting as and five immature animals were used in each dose sympathomimetic drugs (Kraemer and Maurer, 2002; group. Three days after the last treatment, the animals PDR, 2003a). The therapeutic mode of action in treatment were killed and brains were sectioned and stained with of ADHD with amphetamine or methamphetamine is not antibody to glutamate-glutaraldehyde and glial fibrillary known. It is believed that amphetamines increase levels acid protein. Methamphetamine treatment of adult rats of catecholamine in the synaptic space by blocking resulted in marked-to-moderate depletion of glutamate- reuptake of norepinephrine and dopamine by presy- positive neurons in the somatosensory cortex in five of naptic neurons (PDR, 2003b), by releasing dopamine

Birth Defects Research (Part B) 74:471–584, 2005 DOI 10.1002/bdrb 498 GOLUB ET AL. (Markowitz and Patrick, 2001) and norepinephrine (Kita Methamphetamine and amphetamine were distributed et al., 2003) from dopaminergic neurons, and possibly by to infant blood, brain, liver, kidney, lung, and bile. inhibiting monoamine oxidase (Solanto, 2000; Kita et al., Respective blood levels of methamphetamine and am- 2003). Dopamine receptors and adrenoceptors reportedly phetamine were measured at 0.2–1.2 and 0.01–0.08 mg/L have no affinity for amphetamines (Kraemer and in four stillborn infants, 0.355–6.3 and 0.08–0.28 mg/L in Maurer, 2002). There is also evidence that amphetamines three infants dying shortly after birth, and 0.03–0.13 and increase release and turnover of serotonin and it has been 0–0.05 mg/L in two infants dying at 1 month of suggested that many of the behavioral effects of age. Methamphetamine brain levels were reported amphetamines are mediated through serotonin (Kita at 0.28–5.71 mg/g and amphetamine brain levels at et al., 2003). Amphetamine d-enantiomers have five o0.03–0.76 mg/g in three infants dying shortly after times the stimulant activity of l-enantiomers (Kraemer birth. A s.c. exposure study in mice demonstrated and Maurer, 2002). transfer of methamphetamine from dam to fetus on GD Although stimulants decrease locomotor activity in 14, with fetal brain levels of methamphetamine that were children, an increase in activity is observed in experi- B2–5 times lower than concentrations in dam brain mental animal studies. A review by Solanto (2000) (Won et al., 2001). Amphetamine was also detected in discussed possible theories for discordance between brains of dams and fetuses at a concentration at least four children and experimental animals. times higher in dams. Studies in pregnant sheep also 2.6.1.2 Pharmacokinetics. d,l-Amphetamine and d- demonstrated transfer of methamphetamine from the amphetamine are available as immediate- and sustained- ewe to the fetus after dosing of the ewe. Fetal elimination release formulations. Pharmacokinetic parameters re- paralleled maternal elimination (Burchfield et al., 1991; ported to FDA are summarized in Tables 5 and 6. In Stek et al., 1993). One study reported a methampheta- studies by Brown et al. (1979, 1980), administration of mine plasma:brain ratio of 6 in the sheep fetus (Burch- 0.5 mg/kg bw d-amphetamine to boys resulted in peak field et al., 1991). blood levels of B63–70 ng/mL within 3–8 hr. Bioavail- The initial step of amphetamine metabolism in hu- ability for methamphetamine has been reported at 67.2% mans is hydroxylation of the a, aromatic 4-, or b carbon after oral dosing of adults with 0.125–0.250 mg/kg bw (NTP, 1991; Musshoff, 2000). Oxidation of the a carbon d-methamphetamine (Cook et al., 1992) and 67–90.3% leads to deamination and ultimately to the formation of after inhalation of 15.5–40 mg vapors (Cook et al., 1993; benzoic acid, which can be conjugated with glycine to Harris et al., 2003). form hippuric acid (NTP, 1991; Musshoff, 2000). Accord- Absorption of amphetamine and methamphetamine ing to the NTP review, deamination seems to be the was also demonstrated in animal studies. In gavage predominant pathway of amphetamine metabolism in studies in rats and rabbits, maximum blood levels of humans. Smaller amounts of amphetamine are metabo- amphetamines were dependent on dose and ranged from lized through aromatic hydroxylation. Aliphatic b carbon B39–1080 ng/mL in rats dosed with 2–20 mg/kg bw and hydroxylation accounts for only a minor percentage of 23–259 ng/mL in rabbits dosed with 2–16 mg/kg bw d,l- metabolism in humans, but can be important due to the amphetamine (FDA, 2001). Similar blood levels were generation of norephedrine, a metabolite with possible observed after dosing of pregnant and non-pregnant rats. biological activity (Musshoff, 2000). An FDA review Methamphetamine dosing of rats through the s.c. route (2001) reported that metabolism and elimination of resulted in maximum blood levels of 3100–3600 ng/mL amphetamine do not seem to be stereoselective because in pregnant rats given 20 mg/kg bw (Acuff-Smith et al., the ratio of systemic exposure to each enantiomer has 1996) and 5150–7300 ng/mL in neonates given 14–40 mg/ been demonstrated to be equivalent to racemic amphe- kg bw (Cappon and Vorhees, 2001). tamine formulations (3-d:1-l). After oral intake by humans, amphetamines are As in humans, amphetamine metabolic pathways rapidly distributed to major organ systems, including in animals can include hydroxylation of the a carbon, brain (NTP, 1991). Linear pharmacokinetics were demon- the aromatic 4-carbon, the b carbon, or possibly the strated for extended-release d,l-amphetamine at doses of amine group (NTP, 1991). Deamination can occur 10–30 mg (PDR, 2003b). No unexpected accumulation after a-carbon hydroxylation. Whereas deamination is a occurred at steady state (PDR, 2003b). The volume of major metabolic pathway in humans, it plays a minimal distribution for methamphetamine was reported at role in amphetamine metabolism in rats (Table 9). 3.42 L/kg in humans (Cho et al., 2001). Aromatic hydroxylation is the predominant pathway In a lactating 36-year-old woman taking 20 mg/ in rats and the primary urinary metabolite is 4-hydro- day racemic amphetamine in four divided doses, xyamphetamine (NTP, 1991). One review reported amphetamine concentrations at 42 days postpartum that 4-hydroxyamphetamine can be metabolized by were 20–40 ng/mL in plasma and 55–118 ng/mL in a neuronal CYP to a-methyldopamine and then to breast milk; milk:plasma ratios were 6.6–7.5. Maternal a-methylnorepinephrine, a possible false neurotransmit- urinary amphetamine excretion was measured at ter (Sellers et al., 1997). Studies in rats admini- 3.6 mg/24 hr 10 days postpartum and 8.9 mg/24 hr stered d-amphetamine demonstrated the presence 42 days postpartum. Urinary amphetamine excretion in of 4-hydroxyamphetamine, 4-hydroxynorephedrine, the infant was reported to be 1/1000 to 1/300 of maternal a-methyldopamine, and a-methylnorepinephrine in rat levels. brain. Three studies reported methamphetamine and amphe- The two major pathways of methamphetamine meta- tamine levels in blood and or tissues of infants who had bolism in humans are N-demethylation to form amphe- been exposed to methamphetamine in utero and were tamine, which can be metabolized through several stillborn or died after birth (Garriott and Spruill, 1973; pathways, and aromatic hydroxylation to form 4-hydro- Bost et al., 1989; Stewart and Meeker, 1997). xyamphetamine and then 4-hydroxynorephedrine

Birth Defects Research (Part B) 74:471–584, 2005 DOI 10.1002/bdrb AMPHETAMINE AND METHAMPHETAMINE 499 (Caldwell et al., 1972; Musshoff, 2000; Kraemer and have been noted at 30 mg but doses of 400–500 mg do not Maurer, 2002). Humans excreted a larger percentage of necessarily result in fatality. unmetabolized methamphetamine than rats or guinea Amphetamine and methamphetamine abuse is asso- pigs, even at lower doses. The two major pathways in ciated with development of tolerance, psychological humans are aromatic hydroxylation and demethylation, dependence, and social disability (PDR, 2003a,c,e). whereas deamination and b-hydroxylation account for a Abrupt withdrawal after prolonged intake of high doses smaller percentage of metabolism. In humans, aromatic can result in extreme fatigue, mental depression, and 4-hydroxylation is much more extensive in the metabo- sleep EEG changes. Toxicity associated with chronic lism of methamphetamine compared to amphetamine intoxication can include dermatoses, insomnia, irritabil- (Cook et al., 1992; Sellers et al., 1997). ity, hyperactivity, and personality changes. Although rare Half-lives for d-amphetamine ranged from B7–9 hr in with oral intake, psychosis that is indistinguishable from children and from 10–12 hr in adults taking 10–30 mg schizophrenia has been reported. Repetitive and stereo- extended- or sustained-release d,l-amphetamine formu- typic behavior that can advance to self-injurious behavior lations. Half-lives for the l-enantiomer were slightly has been observed with high or repeated dosing with longer: B9–11 hr in children and 13–15 hr in adults amphetamines (Kita et al., 2003). (Tables 5 and 6) (FDA, 2001). Biological half-life for 2.6.2.2 Experimental animals. As noted in Section methamphetamine was reported at 4–5 hr (PDR, 2003c). 2.2.2, symptoms of acute amphetamine toxicity in rats Another review reported a half-life of 12 hr for metham- and mice include hyperactivity, piloerection, salivation, phetamine in humans (Cho et al., 2001). In humans, hyperpnea, dilated pupils, and convulsions. Symptoms amphetamine, methamphetamine, and their metabolites of acute methamphetamine toxicity in rodents were are excreted primarily in urine. reported as excitement, convulsions, and changes in Gavage studies in rats dosed with up to 20 mg/kg bw seizure threshold. The oral LD50 for d-amphetamine in B d,l-amphetamine reported half-lives of 2–3 hr (FDA, rats is reported at 96.8 mg/kg bw. LD50 values for d,l- 2001). Half-lives of 87 min in plasma and 62 min in brain amphetamine and methamphetamine in various species were reported after i.v. injection of rats with 0.5 mg/kg are summarized in Tables 21 and 22. bw d,l-amphetamine (NTP, 1991). Studies in rats given Neuroanatomical evidence of toxicity has also radiolabeled methamphetamine at 45 mg/kg bw orally been reported after amphetamine or methamphetamine or by i.p. injection demonstrated that B75–85% of exposure. Findings have included swollen nerve fibers radioactivity was excreted in urine and B1–3% in feces and terminals in rats and mice (Kita et al., 2003), over 3 days (Caldwell et al., 1972). reduction of brain dopaminergic axons and axon 2.6.2 General toxicity terminals in rats and monkeys, and reduction of 2.6.2.1 Humans. Side effects that the FDA considers serotonergic axons and axon terminals in rats (McCann common and related to treatment with d,l-amphetamine and Ricaurte, 2004). Continuous i.v. infusion of rats with include fever, loss of appetite, emotional lability, insom- amphetamines at 1 mg/kg bw/hr for 12 days resulted in nia, and nervousness (FDA, 2001). Dose-related patterns reduced brain norepinephrine and cardiac catecholamine were noted for anorexia, weight loss, and insomnia. levels and increased motor activity and stereotypic Additional side effects reported in drug labels for or behaviors (NTP, 1991). Repeated parenteral dosing of reviews of d,l- and d-amphetamine and methampheta- rodents with methamphetamine has caused excitement, mine are heart palpitations, tachycardia, elevated blood weight loss or decreased weight gain, alteration in pressure, over-stimulation, restlessness, dizziness, in- classical conditioning, changes in pancreas, and hemor- somnia, euphoria, dyskinesia, dysphoria, tremor, exacer- rhage (NIOSH, 2001). bation of motor and phonic tics and Tourette disorder, Effects of repeated dietary exposure of rats and mice to headache, dry mouth, unpleasant taste, diarrhea, other d,l-amphetamine were reported in an NTP study (1991). gastrointestinal disturbances, constipation, anorexia, Hyperactivity, reduced feed intake, and decreased body weight loss, urticaria, impotence, changes in libido, weight were commonly observed in rats and mice. After jitteriness, and social withdrawal. Psychotic episodes 13 weeks of exposure, final body weights were 410% and cardiomyopathy were reported as rare and isolated lower compared to controls at dietary amphetamine events. Drug labels warn that behavior disturbance and concentrations of Z188 ppm in rats and at 125 (females) thought disorders may be exacerbated in psychotic and 250 (males) ppm in mice. patients. 2.6.3 Genetic toxicity. With the exception of an Amphetamine toxicity after overdose in humans is equivocal mutagenicity response in S. typhimurium strain thought to be an extension of pharmacological effects on TA98, d,l-amphetamine produced negative results in the central nervous, cardiovascular, and gastrointestinal mutagenicity testing of additional S. typhimurium strains systems (NTP, 1991). As noted in Section 2.2.1.2, overdose and E. coli and in sister chromatid exchange and symptoms include restlessness, irritability, tension, chromosomal aberration assays in Chinese hamster weakness, insomnia, chills, fever, angina, dry mouth, ovary cells (NTP, 1991; FDA, 2001). In in vivo micro- gastrointestinal cramps and diarrhea, stimulation fol- nucleus assays in mice, micronucleus frequency was not lowed by fatigue and depression, tremor, hyperreflexia, increased in one study with dosing up to 10 mg/kg bw rapid respiration, confusion, hallucinations, panic, rhab- (FDA, 2001), but was increased in a second study at domyolysis, arrhythmias, hypertension or hypotension, doses Z12.5 mg/kg bw (Tariq et al., 1987). circulatory collapse, nausea, and vomiting. Overdose can 2.6.4 Carcinogenicity. In a 2-year GLP dietary lead to convulsions, coma, and death. There is great carcinogenicity study, there was no evidence of neoplasia variation in toxic response to amphetamine (PDR, 2003a). at d,l-amphetamine doses up to 100 ppm (B5 mg/kg Idiosyncratic toxicity has been reported at doses as low bw/day in rats and 19–30 mg/kg bw/day in mice) as 2 mg but is rare at doses below 15 mg. Severe reactions (Dunnick and Eustis, 1991; NTP, 1991). An increase in

Birth Defects Research (Part B) 74:471–584, 2005 DOI 10.1002/bdrb 500 GOLUB ET AL. lesions such as cataracts, retinal atrophy, and bone dopamine release, firing of dopaminergic neurons, and myelofibrosis in rats and ovarian atrophy in mice did sensitization after challenge dosing, whereas mouse not seem to have been directly caused by drug exposure. studies suggested alterations in corticosterone release The study authors concluded that there was no evidence and DFosB expression. Methamphetamine studies in rats of carcinogenic activity of d,l-amphetamine sulfate in demonstrated age-related alterations in dopamine re- male or female F344/N rats or B6C3F1 mice. lease or turnover, tyrosine hydroxylase activity, suscept- 2.6.5 Potentially susceptible populations. Poly- ibility to glial cell proliferation, toxicity to glutamate morphisms for CYP2D6, the enzyme involved in the positive neurons, or alterations in dorsal protein kinase A aromatic hydroxylation of amphetamines in humans and activity. rats, are associated with at least 12 variants that alter enzyme activity (DeVane, 1994; Bertilsson et al., 1997). Poor metabolizer phenotypes occur in 5–8% of whites 3.0 DEVELOPMENTAL TOXICITY DATA and 2–10% of blacks and Asians. CYP2C3 is involved in amphetamine deamination reactions in rabbits (Kraemer 3.1 Human Data and Maurer, 2002). The enzyme is not a member of the 3.1.1 Exposure during pregnancy CYP2C human gene family (Hines and McCarver, 2002), 3.1.1.1 Case reports and case series. Case reports and but the Expert Panel finds it reasonable to assume that series are presented in this section in the order of one or more members of the CYP2C family may be publication. involved in deamination reactions in humans. The gene Briggs et al. (1975) described a normal child born to a family is highly polymorphic with functional variants woman who used 100–180 mg/day d-amphetamine for identified at CYP2C9 and CYP2C19 loci. Two of three narcolepsy. The child was followed to the age of 18 known defective alleles of CYP2C9 result in decreased months. activity. CYP2C19 polymorphisms resulting in slow Strengths/Weaknesses: A strength is the report of a metabolizer phenotypes are present in 3% of Caucasians high-dose exposure in a woman who seemed to be at low and nearly 20% of Asians (Anzenbacher and Anzenba- risk for other drug exposures. No information was cherova, 2001). provided, however, on length of use of the medication Two studies examined the effects of methampheta- or on specific developmental and neurologic mine on Dark Agouti rats, which have a CYP2D1 assessments. polymorphism that results in poor methamphetamine Utility (Adequacy) for CERHR Evaluation Process: metabolism in females and intermediate metabolism in This single case report is of limited utility. males (Vorhees et al., 1998, 1999). Administration of A group from the Department of Pediatrics at the methamphetamine 30 mg/kg bw s.c. twice daily on PND Karolinska Institute (St. Go¨ran’s Hospital) in Stockholm 11–20 did not affect behavioral testing results in off- followed a cohort of amphetamine-addicted pregnant spring of Dark Agouti versus Sprague–Dawley rats, but women and their newborn children for the first year of mortality was higher in male and female Dark Agouti rat the children’s life. Several publications described aspects offspring compared to Sprague–Dawley rats. In a second of this project; a summary description was published by study, s.c. injection with 15 mg/kg bw methamphetamine Larsson (1980). on PND 11–20 resulted in less effective performance on The first report in the series was a retrospective review the Morris Maze test of spatial learning and memory in of the records of 23 infants whose mothers were female Dark Agouti rats and greater acoustic startle identified as having been addicted to amphetamine amplitudes in male and female Dark Agouti rats (Eriksson et al., 1978). Six of the mothers discontinued compared to Sprague–Dawley rats. Inhibition of amphetamine use by the beginning of the second methamphetamine 4-hydroxylation in rats was reported trimester and 17 women continued to use amphetamine to enhance stereotyped behavior in rats, but reviewers throughout pregnancy. The authors noted that the noted that metabolic differences in humans versus mothers made fewer prenatal visits than is typical of animals complicate the application of the study findings pregnant women in Sweden. There were three pregnan- to humans with inborn CYP2D6 deficiencies (Sellers cies complicated by preeclampsia and six preterm births et al., 1997). (o37 weeks gestation). Three children were Z2SD Comparison of d,l-amphetamine pharmacokinetics below the mean birth weight for gestational age. Two parameters in adults versus children indicated that children were drowsy and required tube feeding, half-life is shorter in children and bioavailability is lower suggesting amphetamine withdrawal. These complicated in children when adjusted for weight differences (FDA, pregnancies were among women who continued to use 2001). A comparison of d-amphetamine responses in amphetamine. hyperactive boys and men demonstrated that activity Prospective evaluation and follow-up of pregnancies was decreased to a greater extent in hyperactive boys, in amphetamine-addicted women began in 1976 and and that although hyperactive boys reported feeling included deliveries in the Stockholm metropolitan area cranky and more tired, the men reported feeling between June 1, 1976 and December 31, 1977. There were euphoric and less tired (Rapoport et al., 1980). 71 children born to 69 women; one woman delivered Eleven studies compared neurological effects after twice during the study period and one woman delivered amphetamine or methamphetamine exposure in devel- twins. Psychosocial and medical characteristics of the oping versus adult animals in an attempt identify mothers (Larsson et al., 1979; Eriksson et al., 1981) and mechanisms responsible for discordant responses in outcomes of the newborns (Eriksson et al., 1981) were adults versus children. Summaries of those studies and described. Of the 69 women, 17 stopped using amphe- their references are included in Section 2.5.3. Ampheta- tamine upon learning of their pregnancy during the first mine studies in rats suggested age-related alterations in trimester, and 52 continued to use amphetamine

Birth Defects Research (Part B) 74:471–584, 2005 DOI 10.1002/bdrb AMPHETAMINE AND METHAMPHETAMINE 501 throughout pregnancy. It was noted that many of the strangers). There were two children with abnormal mothers came from ‘‘multiproblem families,’’ and many emotional development in the group exposed to amphe- of them had experienced foster care as children. Women tamine only during the first trimester, and in both cases, who stopped amphetamine use early in pregnancy came the mothers had resumed amphetamine abuse postpar- for prenatal care earlier and more reliably than women tum. The authors believed that prompt fostering of who continued to use amphetamine and 14 of 17 women children born to women who continued to use amphe- who gave up amphetamine use identified the pregnancy tamine was important in childhood development. They as a life-changing experience that contributed to their questioned whether health and psychological problems decision to discontinue illicit drug use. Three (18%) of 17 in children exposed prenatally to amphetamine were due women who stopped using amphetamine continued to to effects of the drug or due to the poor environment use alcohol. Of the 53 women who continued to use provided by the drug-abusing mother. amphetamine during pregnancy, 17 (32%) also used A 4-year follow-up on 65 of these children born to 63 of alcohol. the women (Billing et al., 1985, 1988) included evaluation One child with malformations was born to a woman by a single psychologist using the Terman-Merrill Test, who had stopped using amphetamine in the first an IQ test, and observation of 2 hr of free play. The trimester. This child died at 3 weeks of age. Two infants psychologist reviewed a parental questionnaire on were stillborn; both mothers had continued to use psychomotor and emotional adjustment. Based on the amphetamine. One quarter of the amphetamine-using psychologist’s observations and his review of the women delivered preterm and two of the preterm infants questionnaires, he rated adjustment in each child as very died in the first week of life. Among the infants born poor, poor, fair, good, or very good. Sixty percent of the to women on amphetamine, there were nine who children were judged fair, good, or very good. A number were transferred to a ward for observation due to of potential predictors of adjustment were examined one maternal drug abuse; five of nine developed symptoms at a time; adjustment was significantly correlated with that were interpreted as amphetamine withdrawal (e.g., length of maternal ethanol abuse during pregnancy, somnolence). whereas correlations with length of drug abuse before or Billing et al. (1980), supported by the Swedish Medical after delivery of the child were lower and did not achieve Research Council and the Fo¨rsta Majblomman Founda- statistical significance [multiple regression results were tion, published a 1-year follow-up of children born to not presented]. Physical characteristics of the children at Swedish women who used amphetamine during preg- 4 and 8 years of age were published separately (Eriksson nancy. Of 69 infants born alive, 66 were alive at the time et al., 1985, 1989) and as a part of the 10-year follow-up of follow-up. Sixteen children were born to women who report, discussed below. stopped their drug use in the first trimester (two of A 10-year follow-up of the 69 children born to 71 whom restarted drug use after pregnancy). The remain- women who had abused amphetamines during preg- ing 50 children were born to women who continued to nancy tested and examined the children at ages 1, 4–5, use amphetamine throughout pregnancy; 37 of these and 8–9 (Eriksson and Zetterstro¨m, 1994). The tests used children were discharged with their mothers after birth to examine the children are summarized in Table 28. and 13 were discharged to foster care. This report was Data obtained in these subjects were compared to based on interviews with the mothers or foster parents, Swedish general population statistics because the reviews of medical and social welfare records, and authors stated, ‘‘From the start we were aware of the evaluations by a psychologist who was unaware of the fact that a control group could not be found and followed children’s background. in Sweden.’’ In the amphetamine-exposed group, peri- The health and psychological status of the children is natal mortality was 5.6%, infant mortality was 5.8%, summarized in Table 27. The authors noted that post- malformations were diagnosed in 3% of children, 18% of natal maternal amphetamine abuse seemed to be an children had a birth weight lower than 2500 g, 20% were important predictor of abnormal ‘‘emotional’’ develop- born at a gestational age o37 weeks, and 38% had to be ment (autism, delayed speech, or lack of wariness of transported to a neonatal unit. There were no significant

Table 27 Twelve-Month Health and Psychological Status of Children Born to Amphetamine-Using Women

Amphetamine exposure/initial caregiver

1st trimester Entire pregnancy

Number of children (%) Mother Mother Fostered

Total sample 16 37 13 Neonatal hospital admissions 2 (12.5) 12 (32.4) 10 (76.9) Children admitted in first year 3 (18.8) 15 (40.5) 3 (23.1) Abnormal motor developmenta 2 (12.5) 6 (16.7)b 1 (7.7) Abnormal emotional development 2 (12.5) 10 (27.8)b 1 (7.7) In foster care at the end of year 1 1 (6.2) 12 (33.3)b 13 (100) From Billing et al., 1980. aDelay of Z2 months. bThirty-six children evaluated.

Birth Defects Research (Part B) 74:471–584, 2005 DOI 10.1002/bdrb 502 GOLUB ET AL. Table 28 Evaluations of the Offspring of Amphetamine-Abusing Women Over 10 Years of Life

Age (year) Evaluations

1 Maternal interview, Observation, Gesell, Growth, Health 4 Maternal interview, Observation, Terman Merril, Goodenough, Bender, Koh Block Design, Growth, Health 8 Maternal interview, Observation, Terman Merril, Goodenough, Bender, Koh Block Design, School achievement, Growth, Health 10 School achievement, Growth, Health From Eriksson and Zetterstro¨m, 1994. Details of assessment methods were not provided. differences in weight and length at birth or at 1, 4, 8, or 10 the development of children up to age 14. [The Expert years of age in boys born to mothers who used Panel notes the apparent sexually dimorphic response, amphetamine during pregnancy compared to general which is not reported elsewhere.] population figures for Swedish boys. Girls born to Strengths/Weaknesses: Although these reports are mothers who had abused amphetamine were found to presented together because they deal with the same have lower weights and lengths at birth and at ages 1 and cohort of children, the quality of the articles is highly 10 years and increased weights at age 4 years. [The variable. The strengths and weaknesses of the individual Expert Panel notes the apparent sexually dimorphic articles are summarized in Table 29. response, which is not reported elsewhere.] Utility (Adequacy) In the CERHR Evaluation Process: The authors stated that all of the children were found This set of reports is generally useful in the evaluation to have a normal intellectual capacity, except for one process, although individual reports taken by themselves child (IQ of 60–80 at 4 years of age) who they classified as may be of limited or no utility. The most useful of these mentally retarded. Mean IQ of the amphetamine-exposed reports is the study of Billing et al. (1980). The utility of children was 103 at 4 years of age, which was said to be the studies in the evaluation process extends only to an lower than the Swedish mean of 110. At 10 years of age, 8 evaluation of the amphetamine-using lifestyle. Because children (12%) attended a class a year below the norm for of the psychosocial factors and the exposure to ethanol their age level. A negative correlation at 8 years of age that accompanied prenatal amphetamine exposure in was reported between an index [not otherwise ex- this cohort, conclusions relevant to the amphetamine plained] of psychometric test results (Bender, Koh Block exposures per se are not possible. Design, and Goodenough) and increased exposure to Dominguez et al. (1991), support not indicated, amphetamine during fetal life [data not presented]. presented 10 cases of brain and ocular abnormalities in Behavioral problems at 4 and 8 years of age were children born to women who had been exposed during found to be frequent in the study group [compared to a pregnancy to amphetamine, cocaine, phenylpropanola- control group of Swedish children evaluated 25 years mine, or heroin. The authors considered these cases earlier]. Among amphetamine-exposed children, 35% together due to the vasoconstricting properties of the demonstrated aggressive behavior and 59% had peer- drugs of exposure and due to similarities among related problems at age 4. At age 8, 23% of the children the outcomes. There was a single child exposed to showed aggressive behavior and 13% show peer-related amphetamine. This infant, whose mother received no problems. There was a significant correlation between prenatal care, was born at 27 weeks gestation, with a aggressive behavior at ages 4 and 8 and amphetamine birth weight of 1070 g. The child was said to have had a exposure during fetal life [methods not explained, data long nursery stay due to prematurity. Examination at not presented]. A psychologist found that 40% of the birth showed motor delay with truncal hypotonia, and children at ages 4 and 8 were poorly adjusted. The neuroimaging at an unspecified gestational age showed authors noted that by 10 years of age, only 30% of the dysgenesis of the corpus callosum with enlarged amphetamine-exposed children were with their biologi- occipital horns. cal mothers, the rest having been placed in foster homes Strengths/Weaknesses: The listing of brain and ocular or adopted. abnormalities may help other health care workers to Subsequent reports presented 14-year follow-up on 65 identify similar outcomes in their patients; however, children from the original cohort (Eriksson et al., 1994; polydrug users are not useful in evaluating possible Cernerud et al., 1996). By age 14, 80% of the children had effects of a single agent. This report represents a single been placed in foster homes or adopted. By age 14–15, 10 retrospectively ascertained case report of amphetamine (15%) of the children were a class below their age level. use during pregnancy and adverse perinatal outcome. There was a significant correlation between IQ level at The maternal interviews were characterized as extensive age 8 and grades at age 14 [data not presented]. but were not necessarily structured. Because the neonates Compared to 494 Stockholm children born in 1976, were identified based on urine drug screening, the amphetamine-exposed boys were an average of 5 cm mothers may have been interviewed while under the taller and girls an average of 2 cm shorter. Boys were also influence of the abused drugs. There was no evidence of an average of about 10 kg heavier. The authors specu- screening for tobacco, ethanol, or other drugs of abuse. lated that alterations in timing of puberty might be an Utility (Adequacy) for CERHR Evaluation Process: explanation for height and weight differences, but did This report is not useful for the evaluation process. not have data on puberty events. The authors concluded Bays (1991) published a letter-to-the-editor describing that amphetamine exposure during the fetal life affects seven children with limb reduction defects, two of

Birth Defects Research (Part B) 74:471–584, 2005 DOI 10.1002/bdrb AMPHETAMINE AND METHAMPHETAMINE 503 Table 29 Strengths and Weaknesses of the Papers on the Karolinska Institute Cohort of Amphetamine-Exposed Children

Study Strengths Weaknesses

Eriksson et al., 1978 Trimester data presented (negative consequences Risk factors other than amphetamine addiction were appeared only when amphetamine use was not mentioned; differences between women who throughout pregnancy). stopped and women who continued amphetamines during pregnancy were not indicated; there was no reference to general population rates of the negative outcomes seen in the sample. Larsson et al., 1979 Had access to social data; recognized that addicted Role (or absence) of father not considered; ethanol use pregnant women come from ‘‘problem families;’’ identified as differentiating women who did and differences recognized between women who did did not stop amphetamines, but patterns of drink- and did not stop amphetamine use during ing not described. pregnancy. Eriksson et al., 1981 Lists outcomes found in neonates of amphetamine- No detail on other risk factors, making a causal link using women; mentions possible effects of mater- with amphetamines problematic; inadequate at- nal-child separation. tention to tobacco use and clinic attendance. Billing et al., 1980 Considered pre- and postnatal amphetamine use; Lassitude may have been due to problems with foster blinded evaluations; contrasting of fostered and homes or problems encountered before fostering, non-fostered children; recognized the effects of even during the first 2 months of life; the multiple separations; included data on parent/ hospitalization rate in fostered children may partner and other people in the home. depend on unreported factors such as how many other children are in the home and who provides transportation for the child; the determinants used to diagnose ‘‘emotional’’ problems are an odd mixture, raising questions about their reliability. Billing et al., 1985, Blinded evaluations; data from 4-year-olds; prenatal Not clear how fostered children differed from non- 1988 ethanol identified as an important contributor to fostered children; determination of alcohol as a outcomes; identification of father criminality as contributing factor was not clear; pattern of alcohol important is in agreement with other studies. use (e.g., binging, chronic) not explored; collection of prenatal data may have been retrospective and unreliable; the general assessment of the children’s emotional well-being seems subjective and unverifiable. Eriksson et al., 1985, Presentation of outcome by child’s sex; useful out- Comparison of exposed children with 25-year-old 1989; Eriksson and come measures (growth, cognitive, behavioral). general population statistics; pre- and postnatal Zetterstro¨m, 1994 risk factors other than amphetamines were not mentioned; it is not clear how fostered children (13/69 at birth and 70% by age 10) were treated analytically; the timing of being placed in foster care was likely to be a surrogate for continued amphetamine use and therefore important; an index describing the mothers’ emotional and abuse problems was not explained; the implication that longer amphetamine exposures were causally associated with aggression was not appropriate because there are too many intervening psychosocial variables by age 8. Cernerud et al., 1996; Comparison of amphetamine cohort to a similar-age Although the authors reported that 80% of the cohort Eriksson et al., 1994; population; evaluation by child’s sex; introduces was in foster care by age 14, there was no apparent Eriksson et al., 2000 the concept of exposure possibly leading to consideration of fostering in the analysis or vulnerability to psychosocial problems. interpretation of the data. It must be assumed that 52 of the 65 children were removed from their homes because of continuing maternal drug addiction, which would have an important impact on the outcomes considered (performance in school, growth); school performance was evaluated using grades from different teachers; no investigation was reported of the children’s possible substance abuse.

whom were born to women who admitted illicit methamphetamine and cocaine may have been respon- methamphetamine use. One of these two mothers plus sible for the limb defects in these cases. an additional mother in the series used cocaine. The Strengths/Weaknesses: This letter could alert clini- author believed the vasoconstrictive effects of cians to look for other instances of limb defects in

Birth Defects Research (Part B) 74:471–584, 2005 DOI 10.1002/bdrb 504 GOLUB ET AL. exposed populations; however, this case series involves and lack of consideration of non-amphetamine risk only two methamphetamine-exposed pregnancies, one of factors for adverse perinatal outcome. which was also exposed to cocaine. There is no indication Utility (Adequacy) In the CERHR Evaluation Process: of evaluation for other drugs and ethanol. This report is not useful in the evaluation process. Utility (Adequacy) In the CERHR Evaluation Process: Sherman and Wheeler-Sherman (2000) published an This report is not useful in the evaluation process. abstract describing the offspring of 202 women who Joffe and Kasnic (1994) presented a case report of a admitted to the illicit use of amphetamine, methamphe- pregnant methamphetamine addict who was prescribed tamine, or similar drugs during pregnancy. There were d-amphetamine beginning at 28 weeks gestation as part 33 infants (16.3%) with major congenital anomalies of an addiction treatment program. The initial dose of among which were 13 cardiac defects (chiefly septal), d-amphetamine was 100 mg once or twice/day; this dose 12 gastrointestinal defects (especially gastroschisis and was decreased and by 32 weeks gestation, the woman imperforate anus), 11 genitourinary malformations (most was not taking any prescribed or illicit amphetamines. At commonly hydronephrosis), and five CNS malforma- 40 weeks gestation, she delivered a healthy infant who tions. There were no details on the ascertainment of did not have withdrawal symptoms or evidence of exposure or outcome information. intracranial bleeding. Strengths/Weaknesses: Strengths include the exclu- Strengths/Weaknesses: The strength of this report is sion of tobacco use. Weaknesses include the possible the novel clinical exposure. A weakness is the lack of influence of ethanol use on the results, the lack of detail information on assessment beyond signs of withdrawal on methods, and the lack of explanation for a rate of and the presence of intracranial bleeding. anomalies that is much higher than reported elsewhere Utility (Adequacy) In the CERHR Evaluation Process: in the literature. This report is of minimal utility in the evaluation process. Utility (Adequacy) In the CERHR Evaluation Process: Finnegan and Ehrlich (1990) described neonatal This abstract is not useful in the evaluation process. abstinence symptoms and treatment among the offspring van Tonningen-van Driel et al. (1999) and McElhatton of 300 drug-abusing women. The drugs of abuse were et al. (1999) represent two reports from teratology opioids and non-opioids, including methylphenidate and information services providing follow-up information methamphetamine. The statement was made in this after recreational pregnancy exposure to methylenediox- article that fewer infants required treatment for absti- ymethamphetamine (MDMA; Ecstasy). The first is a nence among those exposed to non-opioids [with no data study from Holland (van Tonningen-van Driel et al., presented on infant condition by individual drug 1999) [published in Dutch; reviewed in abstract] that exposure]. identified 49 pregnancies about which a physician or Strengths/Weaknesses: The strength of this report is midwife had called with concerns about a patient’s the large number of women; however, the weaknesses exposure to MDMA [the authors note that these women include the evaluation of polydrug using women and the often have additional drug exposures as well]. Follow- lumping of diverse substances as non-opioids. up was available by questionnaire for 43 pregnancies. Utility (Adequacy) In the CERHR Evaluation Process: There were three elective terminations of pregnancy and This report is not useful in the evaluation process. two spontaneous abortions. There was one set of triplets. Furara et al. (1999) reported pregnancy outcome of 28 Of the 40 live-born babies, one had a congenital cardiac of 33 women who said they used amphetamines when malformation. The second report, from the UK National they booked for prenatal care over a 23-month period at a Teratology Information Service (McElhatton et al., 1999), UK hospital. Fifteen of these women admitted to use of was a follow-up on inquiries about 302 pregnancies multiple illicit drugs. Amphetamine and other illicit drug exposed to MDMA, of which 31 were ongoing and 135 use was ascertained purely by maternal report; testing of (45%) were lost to follow-up. Information on the biologic samples was not employed. Fourteen women pregnancies with follow-up was obtained by question- said they stopped amphetamine use during pregnancy. naire from health care providers. There were 11 Five women had previous obstetric complications. Mean spontaneous abortions and 48 elective abortions (1 gestational age was 37.8 weeks (range 5 23–42 weeks), elective abortion for fetal anomalies). The spontaneous with 28.6% of the pregnancies ending before 37 com- abortion rate (8%) was within the expected range, but the pleted weeks of gestation. The mean birth weight was elective termination rate (35%) was higher than the UK 2920 g (range 5 1680–3997 g), with 25% of the babies average. Delivery before 37 weeks gestation occurred in weighing o2500 g. There was one stillbirth and one seven pregnancies, one of which was a twin pregnancy. neonatal death, both in women who continued amphe- There was one neonatal death of a child who seemed to tamine use during pregnancy. Seven children were be morphologically normal (although no autopsy was admitted to the neonatal intensive care unit [it is not reported). The mother had taken MDMA, heroin, and stated whether illicit drug exposure was a criterion for methadone. Birth weight in the sample was described as admission]. The authors did not identify an effect of self- normal with only three term infants weighing o2500 g. reported amphetamine cessation during pregnancy on Of the 78 live born infants, 12 had congenital malforma- birth weight or the incidence of preterm labor. tions, for an incidence of 15.4% (95% CI 5 8.2–25.4%), Strengths/Weaknesses: The authors’ impression that compared to an expected rate of 2–3%. The authors noted stopping amphetamines during pregnancy was asso- an incidence of talipes equinovarus (clubfoot) that was ciated with a decrease in perinatal death was not greater than expected based on national statistics. There statistically evaluated. Weaknesses include use of poly- were three infants with this malformation for a rate of drug abusing women, lack of objective determination of 38/1000 births (95% CI 5 8–109) compared to an ex- abstinence, lack of consideration of ethanol, lack of pected rate of 1/1000 births. In addition, all three determination of the timing of amphetamine cessation, affected infants were girls whereas idiopathic talipes

Birth Defects Research (Part B) 74:471–584, 2005 DOI 10.1002/bdrb AMPHETAMINE AND METHAMPHETAMINE 505 equinovarus has a 3:1 male predominance in the UK. The the evaluation process, but this abstract is useful as authors also noted that two infants had ventricular septal supplemental information. defects (one of whom may also have had an atrial septal 3.1.1.2 Controlled studies. defect) for a rate of 26/1000 births (95% CI 5 3–90/1000). Nora et al. (1970), supported by March of Dimes and The expected incidence of congenital heart disease was NIH, reported in a letter-to-the-editor a comparison of 5–10/1000 births, with one-fourth to one-third of these 184 children with congenital heart disease and 108 cases including ventricular septal defect. Due to the small children without congenital heart disease [children seen size of the case series, the authors declined to draw in the well-baby clinic of the same hospital]. Exposure conclusions regarding a possible causal connection to d-amphetamine was determined by asking the mother between MDMA exposure in pregnancy and congenital within 1 year of birth whether she had been exposed. Of anomalies. the 184 children with congenital heart disease, 33 (18%) Strengths/Weaknesses: Both reports are limited by had maternal exposure to d-amphetamine, compared to being derived from a self-selected population of indivi- nine (9%) of the children without congenital heart duals who inquired about MDMA exposures. In the van disease (Po0.05 [OR 5 2.40, 95% CI 5 1.06–5.95, calcu- Tonningen-van Driel et al. (1999) study, no indication was lated by CERHR using v2 feature of CDC SABER given on frequency and timing of exposure and it is not program]). When exposure to d-amphetamine was clear what assessments were performed. The McElhatton restricted to the ‘‘vulnerable period’’ (not otherwise et al. (1999) study gave a useful comparison to general specified) 20 (11%) of the children with congenital heart population statistics, including sex-related rates, and the disease and three (3%) of children without congenital authors were appropriately reserved about the conclu- heart disease were identified (P 5 0.025 [OR 5 4.27, 95% sion that MDMA exposure was causally involved in the CI 5 1.22–22.89]). A family history of congenital heart adverse outcomes. disease was found in 49 (27%) of the children with Utility (Adequacy) In the CERHR Evaluation Process: congenital heart disease and six (6%) of the control The McElhatton et al. (1999) study is useful in the children (Po0.001 [OR 5 6.17, 95% CI 5 2.49–18.23]). evaluation process to the extent that data on MDMA can Maternal exposure during the vulnerable period plus be applied to amphetamine and methamphetamine; the family history of congenital heart disease was identified report by van Tonningen-van Driel et al. (1999) study is in nine (5%) of the children with congenital heart disease less useful based on lack of detail. and one (1%) of the children without congenital heart McElhatton et al. (2000) presented an abstract from the disease (P-value not calculated by authors [CERHR UK National Teratology Information Service in which calculation: P 5 0.07; OR 5 5.50, 95% CI 5 0.74–243]). they reported follow-up information on 281 pregnancies The authors indicated that they wished to modify the for which inquiries had been made about pregnancy conclusion expressed in a previous letter (Nora et al., exposures to amphetamine-like drugs. Of 281 women, 41 1967), which stated that the probability of an association abused amphetamines alone, 18 abused amphetamines between d-amphetamine exposure during pregnancy and and other drugs (including MDMA), and 136 abused congenital heart disease seemed to be low. [This MDMA with or without other drugs (except other communication is described in this section because its amphetamines). [The Expert Panel assumes that design seems to be that of a case-control study, these women were also reported in McElhatton et al. although the letter format lacks the level of detail (1999)]. There were 86 women with exposure to anorectic typically included in a study report.] drugs, 77 of whom used phentermine. There were 21 Strengths/Weaknesses: This letter has the strengths of children with congenital anomalies born to the 195 considering family history as a risk factor for cardiovas- women who abused amphetamines with or without cular malformation and of including a large group of other drugs. There were 64 elective abortions, 14 affected children. Weaknesses include failure to indicate spontaneous abortions, and two stillbirths/neonatal the reasons for maternal use of d-amphetamine, lack of deaths in this group. The anomaly rate was 10.8% of information on other exposures such as ethanol, and lack live-born children, which the authors described as of detail on study design or on the study population. It elevated compared to an expected rate of 2–3%. does not seem that estimates were adjusted for potential Musculoskeletal defects (7.5% of live births) and con- confounders. genital heart disease (2.5% of live births) were also said to Utility (Adequacy) In the CERHR Evaluation Process: be elevated compared to an expected rate of 0.5–1%. [The This letter is not useful in the evaluation process. methods of exposure and outcome ascertainment were Nelson and Forfar (1971), supported by Ciba Ltd. and not described.] the Distillers Company Ltd., interviewed 458 mothers Strengths/Weaknesses: The breakdown of usage of who gave birth to children with congenital malforma- amphetamine, MDMA, and other drugs is a strength, tions and mothers in each of two control groups. One although it is not clear what the other drugs were. The control group consisted of 500 women who gave birth to comparison of outcomes to general population statistics normal babies immediately after a case child and a is also a strength, but it is not clear what outcome second group consisted of an additional 411 women who assessments were performed, other than anomaly deter- gave birth to normal children and who were matched to mination, or how elective and spontaneous abortions case mothers on age, parity, and infant sex. Women were were distributed among drug-using subgroups. Among interviewed before discharge from the maternity unit women exposed to licit drugs, most used phentermine, and were asked about medication use. Reported medica- which is not likely to produce relevant information on tion exposure was confirmed with the general practi- amphetamine exposure. tioner, hospital records, or prescription records. A Utility (Adequacy) In the CERHR Evaluation Process: discrepancy in records occurred for 11.3% of the 4731 Abstracts are not used directly to develop conclusions in reported prescriptions [the authors state that these

Birth Defects Research (Part B) 74:471–584, 2005 DOI 10.1002/bdrb 506 GOLUB ET AL. prescriptions were ‘‘rejected’’ but do not indicate if the affected infants had used amphetamines during the mothers were considered unexposed or were removed second and third month of pregnancy, and a fifth mother from consideration]. Pregnancy dating relied on last used amphetamines during the last trimester. Other menstrual period and associations were tested by medications were also used by some of these women. specific exposure periods. These exposure periods were Three of the 50 control mothers had used amphetamines. described as the prefertilization period (first 14 days), the The authors reported that the difference between period of ‘‘maximum organogenesis’’ (first 56 days), and exposure rates in affected and unaffected pregnancies the period of ‘‘total organogenesis’’ (first trimester). A was significant at Po0.05 [OR 5 8.95, 95% CI 5 1.20– category called ‘‘whole of pregnancy’’ was designated for 70.92 calculated by CERHR using CDC SABER exposure at any time of pregnancy, including during one program]. of the more specific periods. The associations between Strengths/Weaknesses: The focus on a specific defect prescriptions for d-amphetamine products and congeni- and the inclusion of information on timing of exposure tal anomalies are shown in Table 30. A significant are strengths; however, weaknesses include the retro- association with d-amphetamine was shown when all spective nature of the report, the lack of consideration of children with malformations were compared to children covariates, and the lack of detail on aspects of the without malformations, regardless of the time period methods, including the selection of controls. studied. For children with minor malformations, a Utility (Adequacy) for CERHR Evaluation Process: significant association was shown only for exposures This report is not useful in the evaluation process. during the first 15 days (before fertilization) and the first The National Collaborative Perinatal Project (Heino- 56 days. Details were given for the 10 children with nen, 1977) reported on 50,282 mother-child pairs in malformations after maternal exposure in the first 56 which pregnancy had lasted at least 5 lunar months. days. Four children had urogenital anomalies; other Information on medication exposure during pregnancy anomalies [assumed to be one per child] included was collected at the time of the first prenatal visit and congenital heart disease, cleft lip, limb deformity, ear recorded prospectively thereafter. Outcome information abnormality, dislocated hip, and pilonidal sinus. was based on physical examination of the child up to the Strengths/Weaknesses: Strengths include the attempt age of 1 year in 91% of the sample and for up to 4 years of to match mothers for age and parity, efforts to confirm age in an unspecified proportion of the sample. There maternal reports using a second source such as phar- were 29 malformed children from 367 exposures to d- macy records, analysis by different periods of exposure, amphetamine during the first 4 lunar months, 17 and division of malformations into major and minor, malformed children from 215 exposures to ‘‘ampheta- inasmuch as the latter may be more prone to under- mines’’ [not otherwise specified], and 5 malformed ascertainment. Weaknesses include the lack of considera- children from 89 exposures to methamphetamine. Re- tion of confounds such as maternal socioeconomic status, lative risks were calculated using the entire sample as a nutritional status, and tobacco and ethanol use. No reference group (3248 malformed children from 50,282 information is provided on why women were taking the pregnancies, less the specific exposed individuals). The medication. If these women were using the medication crude relative risk [95% CI, calculated by CERHR using for weight control, it is possible that excess maternal the CDC SABER program] was 1.23 [0.82–1.82] for d- weight was an independent risk factor for malforma- amphetamine, 1.23 [0.72–2.05] for amphetamines, and tions. It is also a weakness that the information on 0.87 [0.31–2.22] for methamphetamine. Standardized amphetamines is based on only 13 cases. The apparent relative risks were calculated using a multiple-logistic- inattention to a multiple comparison problem is also a risk function analysis procedure (with confounders weakness. tailored to each exposure group). Evaluation of 12 Utility (Adequacy) for CERHR Evaluation Process: different malformation categories for all sympathomi- This report is of limited utility in the evaluation process. metic drug exposures together showed no significant Levin (1971) presented 11 cases of biliary atresia difference of the standardized relative risk from unity for presenting in infants between January 1, 1969–June 1, any category. For the 367 d-amphetamine exposures 1970. Prenatal drug use histories were solicited from the during the first 4 lunar months, the standardized relative infants’ mothers and compared to histories obtained risk for any malformation was 1.08 (95% CI 5 0.65–1.68), from the mothers of 50 control infants of the same age for major malformations 1.29 (95% CI 5 0.73–2.10), and [details of the methods not given]. Four mothers of for minor malformations 1.46 (95% CI 5 0.59–2.98). For

Table 30 Number of Exposures to d-Amphetamine Among Women Delivering Malformed and Normal Babies

Malformations

Exposure period All (n 5 458) Major (n 5 175) Minor (n 5 283) Control (n 5 911) OR (95% CI) for total malformation

Whole of pregnancy 13a 5 8 10 2.63 (1.07–6.52) First trimester 11a 4 7 8 2.78 (1.03–2.61) First 56 days 10a 37a 5 4.04 (1.27–13.65) First 14 days 8a 26a 2 8.08 (1.59–55.25) Data from Nelson and Forfar (1971). aPo0.05 according to the authors; statistical method not indicated.

Birth Defects Research (Part B) 74:471–584, 2005 DOI 10.1002/bdrb AMPHETAMINE AND METHAMPHETAMINE 507 exposures anytime during pregnancy and any malforma- magnitude of the birth weight decrement [estimated tion, standardized relative risks were statistically not from a graph] was 100–400 g, depending on the pre- different from unity for d-amphetamine (1069 expo- pregnancy weight and pregnancy weight gain category. sures), amphetamines (509 exposures), and methamphe- The author indicated that a decrease in birth weight tamine (320 exposures). without a change in birth length or head circumference Strengths/Weaknesses: This study included a large was consistent with a drug-associated reduction in multi-site sample with prenatal recruitment and the placental blood flow due to vasoconstriction. report specified the background characteristics of the Strengths/Weaknesses: Strengths include the large women who were recruited, not recruited, and who sample of women who took amphetamines for the refused participation. There was also a well-defined same reason (weight control); the clear exclusion criteria; methodology for identification of the substances used the matching for ethnic origin parity, and smoking; the and at least two medical record sources were used to attempt to analyze results by gestational age of exposure determine malformations. The authors thoughtfully (leading to the biologically plausible weight effects with addressed the issue of subjective differences in third trimester exposure); and the inclusion of informa- the analysis of certain malformations and analyzed tion about birth weight and head circumference. Weak- these malformations separately from those that are nesses include the lack of information on medication more ‘‘uniform.’’ Another strength is sensitivity to dose, and the lack of inclusion of covariates such as potential ethnic differences in malformations of interest. maternal age in the analysis. There was a well-specified multivariate model and Utility (Adequacy) for CERHR Evaluation Process: a tight targeting of the period of possible exposure to This study is useful in the evaluation process although the first trimester of pregnancy. Major weaknesses of limited by the identified weaknesses. this study include failure in the Collaborative Perinatal Milkovich and van den Berg (1977), supported by the Project to identify prenatal alcohol exposure and a lack National Institute of Child Health and Human Develop- of information about dose of exposure to amphetamines ment, evaluated pregnancy outcome of nearly all the or indications for use. There were multiple comparisons white women insured by Kaiser Health Plan who and lack of adjustment for covariates. A family history delivered infants in the San Francisco East Bay Area of birth defects was not considered. A final weakness between 1959 and 1966. The authors evaluated associa- is conflation of the drugs of interest in the current tions between anorectic drugs used by the mother during review (methamphetamine and d-amphetamine) with pregnancy (from treatment records) and congenital a range of exposures to related but not identical anomalies in the offspring. Congenital anomalies were substances commonly found in patent cold medicines assessed based on diagnosis at birth or at a Kaiser clinic that contribute the largest number of observations to the through 61 months of age. A diagnosis of ‘‘severe analysis. congenital anomaly’’ was made in 3.4% of children of Utility (Adequacy) for CERHR Evaluation Process: 8989 women who did not use anorectic drugs and 3.4% of This study is useful in the evaluation process with a 1694 women who reportedly used amphetamines during recognition of the limitations in interpretation associated pregnancy [crude RR 5 1.01, 95% CI 5 0.76–1.32]. The with the weaknesses of the study. authors noted an apparent increase in oral clefts, Naeye (1983), supported by the U.S. Public Health reporting three affected children of 175 women who Service (PHS), presented an analysis of birth weight, were believed to have used amphetamines during the length, and head circumference stratified by maternal first 56 days of pregnancy and five of 1694 women prepregnancy weight and pregnancy weight gain from reported to have used amphetamines anytime during the National Collaborative Perinatal Project. Only term pregnancy. The unexposed group gave birth to 21 infants without congenital anomalies were considered. affected children of 10,213 total children. [The authors Infants whose mothers took d-amphetamine for weight do not present a statistical analysis. The Expert Panel control (n 5 237) were compared to infants whose calculated for five affected children of 1694 compared to mothers did not take d-amphetamine (n E41,774 [an 21 affected children of 10,213, the point estimate is 1.43 unspecified number of women who took d-ampheta- (95% CI 5 0.54–3.8), P 5 0.41.] There was no increase in mine for psychotropic purposes was excluded]). Drug- congenital heart disease incidence in children born to using women were matched to all non-users who had the exposed women (14 affected children/1694 exposed) same ethnic origin, parity, and smoking habits during compared to unexposed women (rate given only as 0.9%). pregnancy [the number of these matched controls was Strengths/Weaknesses: Strengths include the use of a not indicated]. There was no difference in birth weight Kaiser population, which indicates good access to health among infants born to women who discontinued d- care, good ascertainment of exposure, and a broad amphetamine by 28 weeks of pregnancy compared to socioeconomic base. The sample size was large and control women. Among infants born to women who 90% of the exposures were for appetite suppression. It is continued d-amphetamine use into the third trimester, a strength that specific defects were evaluated and that birth weights were decreased compared to controls if the the mother’s weight gain during pregnancy was con- mother gained 412 kg during pregnancy and had a pre- sidered. Weaknesses include the lack of detail identified gravid weight o45 kg. Among infants born to women in the summary above, the lack of information about with a pre-pregnancy weight Z45 kg, birth weight was family history and about syndrome diagnosis, and lower than in controls if the woman gained at least 8 kg failure to control for tobacco use, which may be a risk during pregnancy. Birth length, head circumference, factor for oral clefts. The identified increase in cleft palate perinatal mortality, and incidence of maternal diastolic is a weak finding based on few cases. blood pressure 485 mm Hg did not show a significant Utility (Adequacy) for CERHR Evaluation Process: association with d-amphetamine exposure. The This study is useful in the evaluation process, but its

Birth Defects Research (Part B) 74:471–584, 2005 DOI 10.1002/bdrb 508 GOLUB ET AL. utility is decreased by the lack of details and the small California, San Diego. After excluding preterm infants number of children with oral clefts. and infants with other possible causes of neurologic Oro and Dixon (1987), supported by NIH, presented abnormalities, a stimulant-exposed group of 81 infants 104 infants enrolled over an 18-month period due to was identified based on infant urine toxicology screen- maternal or infant urine toxicology screening positive for ing. This group included 27 infants who had been cocaine, methamphetamine, or opioids. Medical records exposed to methamphetamine. A neurologically-at-risk were reviewed and comparisons made between a pooled comparison group consisted of 87 drug-free infants group of 46 infants exposed to cocaine or methamphe- suspected on clinical grounds of having hypoxic- tamine (28 of whom were exposed to methamphetamine ischemic encephalopathy. An additional comparison and five of whom were exposed to both drugs), a group group consisted of 19 normal, drug-free infants. The exposed to opioids (n 5 49), and a comparison group of cranial ultrasound findings were made with knowledge drug-free mother-infant pairs (n 5 45). Gestational age, of group membership. birth weight, length, and head circumference were lower Stimulant-exposed infants were smaller than the in the groups exposed to cocaine or methamphetamine. infants in the other groups. Mean birth weight was Mean gestational age in the cocaine/methamphetamine 29047475 g in the stimulant-exposed group, 33467687 g group was 37.973.0 weeks compared to 39.471.4 week in the neurologically-at-risk group, and 33517420 g in in the drug-free control group (Po0.05). Mean birth the normal control group. Mean birth length was weight was 29017711 g in the cocaine/methampheta- 48.973.1 cm in the stimulant-exposed group, 50.57 mine group and 32467552 g in the drug-free control 4.8 cm in the neurologically-at-risk group, and 51.872.5 group (Po0.05). Mean length was 48.075.1 cm in the in the normal control group. Mean head circumference cocaine/methamphetamine group and 50.772.8 in the was 33.471.5 cm in the stimulant-exposed group, 34.47 drug-free control group (Po0.05). Mean head circumfer- 1.8 in the neurologically-at-risk group, and 34.371.1 in ence was 33.272.7 cm in the cocaine/methamphetamine the normal control group (all differences between group and 34.471.5 cm in the control group (Po0.05) stimulant-exposed and other groups significant at [the errors are assumed to be SD]. Po0.05 or greater, ANOVA followed by corrected t-test). There were more perinatal complications in the [The errors are assumed by the Expert Panel to be SD. cocaine/methamphetamine group than in the drug-free The stimulant-exposed group included cocaine-ex- control group (28% compared to 9%). Complications in posed babies and anthropomorphic data were not the cocaine/methamphetamine group consisted of pre- separated by individual stimulant exposures. Some of maturity, decreased head circumference, smallness for the infants in this study were previously presented by gestation age, fetal bradycardia, seizure, premature Oro and Dixon (1987); the reported measurements are rupture of membranes, hyperbilirubinemia, placental similar although not identical in the two studies.] hemorrhage, and anemia. Of these complications, only Mean gestational ages were 4–5 days lower than in the the last two were identified as occurring in a statistically comparison groups (38.971.3 weeks in children exposed larger proportion of drug-exposed than control infants. to stimulants, 39.671.3 weeks in the neurologically-at- Neurologic and behavioral abnormalities were re- risk children, and 39.470.7 weeks in control children); ported to occur in a majority of cocaine/methampheta- the effect in stimulant-exposed children was statistically mine-exposed infants and consisted of abnormal sleep significant compared to both groups. The authors stated patterns (81%), tremors (71%), poor feeding (58%), that the gestational age difference did not explain the hypertonia (52%), vomiting (51%), sneezing (45%), anthropomorphic differences. Stimulant-exposed infants high-pitched cry (42%), frantic fist sucking (42%), were more likely than infants in the other groups to be tachypnea (19%), loose stools (16%), fever (19%), yawn- black and growth-restricted (19.8% intrauterine growth ing (12.9%), hyperreflexia (15%), and excoriation (6%). restriction among stimulant-exposed babies compared to Based on comparisons with the group of infants 8.0% of neurologically-at-risk babies and 5.3% of normal exposed to opioids, the authors concluded that adverse controls, Po0.01 for comparison of stimulant group to effects on neonatal condition were due to cocaine or either of the other groups). methamphetamine rather than maternal socioeconomic Abnormal cranial ultrasound findings were identified characteristics or poor attendance at prenatal visits. in 26 (35.1%) of the stimulant-exposed infants, 24 (27.6%) Strengths/Weaknesses: It is a strength that illicit drug of the neurologically-at-risk infants, and one (5.3%) of the use was documented by testing maternal and infant normal control infants. The abnormality in the normal urine. This study includes important endpoints such as control infant was subependymal hemorrhage. Nine gestational age, birth weight, head circumference, fetal (37.5%) of the 24 infants exposed to methamphetamine growth, and behavior. Weaknesses include: poor expla- had abnormal results consisting of one (2.4%) with white nation for the selection of controls; mixing of ampheta- matter cavities, three (12.5%) with white matter densities, mine and cocaine exposure; lack of consideration of four (16.7%) with intraventricular hemorrhage, four ethanol exposure, malnutrition, or living situation; lack (16.7%) with subarachnoid hemorrhage, three (12.5%) of consideration of maternal disease; and poor descrip- with subependymal hemorrhage, and two (8.3%) with tion of the statistical analysis with failure to show the ventricular enlargement. Of stimulant-exposed infants, results of the regression analysis. Dose information was 22% had multiple findings (not further broken down by considered unreliable. specific stimulant). There was no association between the Utility (Adequacy) In the CERHR Evaluation Process: cranial ultrasound findings and the birth weight or This study is not useful in the evaluation process. length by [unspecified] univariate analysis. Dixon and Bejar (1989), supported by March of Dimes, The authors found the ultrasound findings consistent collected anthropomorphic and cranial ultrasound data with a presumed vasoconstriction mechanism of stimu- on three groups of newborns at the University of lant toxicity. They believed maternal lifestyle factors to be

Birth Defects Research (Part B) 74:471–584, 2005 DOI 10.1002/bdrb AMPHETAMINE AND METHAMPHETAMINE 509 less likely explanations because a cocaine1opioids between the groups in estimated gestational age subgroup of the stimulant-exposed infants had fewer (39.171.5 weeks in the exposed group and 39.372.0 in lesions than a cocaine-only subgroup, but would have the control group). been expected to have less favorable maternal social The authors indicated that they found no separate conditions. effect of cigarette smoking on anthropomorphic para- Strengths/Weaknesses: This study has strengths and meters [analysis not shown]. There was said to be no weaknesses similar to that of Oro and Dixon (1987), with difference in the proportion of infants in each group with the added strength that a single radiologist evaluated all Apgar scores r6 [data not shown]. ultrasound examinations and the additional weaknesses The authors identified as limitations in their study the of failure to match for gestational age (which was shorter reliance on self-reported illicit drug use and the in the exposed group) and failure to consider ethanol, difficulty of separating effects of methamphetamine trauma history, and medical conditions. from lifestyle factors such as nutritional status. They Utility (Adequacy) In the CERHR Evaluation Process: concluded, however, that the restricted fetal growth was This study is not useful in the evaluation process. associated with methamphetamine abuse during Little et al. (1988), supported by NIH, over a 10-month pregnancy. period identified 52 pregnant women who reported that Strengths/Weaknesses: A strength is the ability to they had abused i.v. methamphetamine. Pregnancy identify 52 i.v. methamphetamine users; however, a outcome in these women was compared using medical weakness is that this number is too small to perform record review with that of a control group of 52 women multivariate analysis to control for pregnancy exposure not known to have abused drugs, selected by their to other substances. The selection of controls to represent deliveries having occurred next after those of the the population served by the hospital is a strength. Some methamphetamine-abusers. Thirty-eight of the 52 of the stated abnormalities (e.g., Mongolian spots) are methamphetamine abusers admitted to other recrea- normal ethnic variations and should not have been tional drug use during pregnancy. The most common counted as abnormalities. Other weaknesses were iden- other drugs used by methamphetamine-exposed women tified by the authors of the article: the reliance on self- during pregnancy were tobacco (46%), marijuana (39%), reported illicit drug use and the difficulty of separating and cocaine (27%). Among the control women, tobacco effects of methamphetamine from lifestyle factors such use was identified in 11% and marijuana use in 2%. as nutritional status. Comparisons were made between the groups using Utility (Adequacy) In the CERHR Evaluation Process: unspecified nonparametric tests and ANOVA. There was This study is not useful in the evaluation process. no difference between the groups in pregnancy compli- Gillogley et al. (1990), support not indicated, eval- cations. In the methamphetamine group, there were six uated pregnancy outcome among women with positive reports of pregnancy complications consisting of preg- urine drug screens delivering at the University of nancy-induced hypertension, peripartum hemorrhage, California, Davis Medical Center, compared to a group and syphilis. In the control group there were 14 reports of of drug screen-negative women matched for ethnic pregnancy complications including pregnancy-induced origin and nearest discharge date. Women were excluded hypertension, peripartum hemorrhage, syphilis, chor- from the control group if they gave a history of illicit ioamnionitis, and hepatitis. No differences were found in drug use or if screening of their babies identified illicit neonatal complications. Eleven methamphetamine-ex- drug exposure. Screening was conducted for opioids, posed neonates had records of complications consisting cocaine, and amphetamines. Information on exposure of meconium-stained fluid, fetal heart rate decelerations, and pregnancy outcome was obtained by medical record tachypnea, fetal tachycardia, and withdrawal symptoms. review. Most of the urine toxicology screens were carried Among control infants, there were nine complications out on women presenting in labor as part of the consisting of meconium-stained fluid, tachypnea, and hospital’s routine. Infant screening was performed only congenital syphilis. There was no difference between the if the mother had a positive screen or if the neonate was two groups in the proportions with congenital abnorm- clinically suspected to have been exposed to illicit drugs. alities detected at birth. The methamphetamine group Among 1643 women with known toxicology screen had six reports of infant abnormalities consisting of cleft results, 379 (23%) were positive for one or more of the lip/palate, low body fat, Mongolian spots, systolic tested substances and 106 (6%) were positive only for murmur, undescended testes, and ventricular septal amphetamines. An additional 35 women had positive defect. There were seven infants with abnormalities in screens for multiple substances (not specified), and 39 the control group including one case each of undes- were positive for something (details not given). Prenatal cended testes, hairy nevus, hip/knee click, natal teeth, care had not been received by 40% of women who were supernumerary nipple, umbilical hernia, and vaginal tag. positive for amphetamines compared to 11% of the 293 The authors indicated that their sample size had a 39% control women. A history of smoking was present in 80% power to detect a 4-fold increase in malformations. of amphetamine-positive women and 32% of control Infants in the methamphetamine group had a lower women, and a history of alcohol use was present in 22% mean (7SD) birth weight (2957.07574.0 g) compared to of amphetamine-positive women and 11% of control control infants (3295.87433.3 g, Po0.001). Mean birth women (both comparisons Po0.05 by Fisher Exact or w2 length (7SD) was lower in the methamphetamine- test). Pregnancy complications and mode of delivery exposed infants (48.172.0 cm) compared to the control were not different between amphetamine-positive and group (49.872.3 cm, Po0.001). Mean head circumference control women. There were no differences between the (7SD) was also reduced in methamphetamine-exposed amphetamine and control groups in gestational age, infants (33.271.0 cm) compared to control infants proportion born prematurely (o37 weeks), and propor- (33.971.2 cm, Po0.001). There was no difference tion with low birth weight (o2500 g), but mean birth

Birth Defects Research (Part B) 74:471–584, 2005 DOI 10.1002/bdrb 510 GOLUB ET AL. weight and head circumference were reduced in the years old at the time of the study. The exposed children amphetamine group [the data table indicates the mean were identified through a state-funded drug-treatment birth weight in the amphetamine group was 294.7 g, program in which their mothers had been enrolled. which is assumed to be a misprint]. Mean head Control children came from the same population of circumference was 32.8 cm in the amphetamine group lower- and middle-income urban residents [recruitment and 33.5 cm in the control [variances not given]. There methods not indicated]. Parents completed the Child was only one child in the amphetamine-positive group Behavior Checklist and children underwent magnetic with a congenital anomaly (hypospadias) and the resonance imaging studies of the brain and localized 1H- congenital anomaly rate was not different from the magnetic resonance spectroscopy of the right frontal control sample. Regression analysis on the entire sample white matter and basal ganglia. The Child Behavior of illicit drug-positive women (exposed to ampheta- Checklist did not show any problems in the control mines, opioids, cocaine, or multiple chemicals) showed group. The methamphetamine-exposed group included illicit drug use to make an independent contribution to two children with social problems and delinquent birth weight and head circumference when smoking and behavior and one child with thought problems, aggres- gestational age were controlled [there was no regression sive behavior, and anxiety. There were no anatomic on the individual drugs of abuse]. The authors drew abnormalities in the brains of any of the children by conclusions about illicit drug use in general, but did not magnetic resonance imaging. Magnetic resonance spec- offer conclusions specifically on amphetamine use. troscopy showed increased concentrations of creatine Strengths/Weaknesses: The identification of women and glutamate/glutamine in the basal ganglia of by substance of abuse and separate analysis of the methamphetamine-exposed children without a signifi- women positive only for amphetamines is a strength; cant difference in N-acetyl containing compounds, which there was, however, an insufficient number of ampheta- are markers of neuronal loss. The authors believed the mine-only women for use in the regression analysis. The increased creatine concentration could represent gliosis, use of record review and computerized discharge although there was no increase in the concentration of information for exposure and outcome determination is myoinositol, which would be expected with gliosis. an important weakness, in part because all information Alternatively, the increased creatine concentration could on cigarette smoking and alcohol consumption would have represented an alteration in energy metabolism. The have come from these sources. Medical records are authors noted that previous studies had shown similar known to be inconsistent in reliably providing this kind increases in creatine associated with past cocaine use in of information. The study noted that information on adults and with prenatal cocaine exposure in children. quantity or frequency of alcohol consumption was not They suggested that multiple episodes of decreased collected, and it is not clear whether information was placental blood flow associated with the vasoconstrictive collected on the amount of cigarette smoking as opposed effects of methamphetamine could have resulted in long- to a dichotomous determination of use (smoker/non- lasting alterations in energy metabolism in the brain. smoker). Information on past drug use was also proble- Strengths/Weaknesses: The use of DSM-IV criteria for matic; of women with a positive drug screen, 38% denied methamphetamine dependence is a strength, as is the illicit drug use at any time. In addition, drug screening exclusion of women with other illicit drug dependence was performed at only one point in time. No information and women medicated for chronic illness. The sample on nutrition during pregnancy was collected. Another size is too small to control for other risk factors and there weakness is the screening of newborns based on a was no prenatal ethanol exposure in the controls. As the criterion (suspicion for illicit drug exposure) that may not authors point out, the magnetic resonance findings have been uniformly applied. The typographical error in cannot be considered specific to methamphetamine the table is a problem because the magnitude of the exposure because other potential exposures, such as deficit in birth weight can only be guessed. The statistical lead, were not evaluated. analysis was inadequate in that the regression analysis Utility (Adequacy) In the CERHR Evaluation Process: on birth weight, head circumference, and length did not This study is correctly described by its authors as pre- adjust for any sociodemographic factors and the multi- liminary and provides only supplemental information. variate approach was used for one set of analyses that Smith et al. (2003), sponsored by NIH, retrospectively combined all drug exposure groups. As a result, evaluated neonatal outcome in pregnancies identified as conclusions about amphetamine exposure were based methamphetamine-exposed by urine toxicology screen of on data without consideration of the possible role of the mother or a ‘‘history of drug use’’ in the mother. other exposures (e.g., smoking or alcohol) or demo- Toxicology screens were triggered by a record of fewer graphic factors. than four prenatal visits, first prenatal visit after 30 Utility (Adequacy) In the CERHR Evaluation Process: weeks gestation, maternal history of drug abuse, symp- This study is of minimal utility in the evaluation process toms or behaviors suspicious for drug abuse, placental in adding evidence for an effect of prenatal amphetamine abruption, and out-of-hospital birth. A comparison abuse on fetal growth. Independence from effects of group of unexposed infants was identified using new- tobacco, ethanol, or other lifestyle factors cannot be born logbooks and matching for maternal parity and assumed. month of birth. Unexposed status was based on negative Smith et al. (2001), funded by NIH, evaluated 12 urine toxicology screen or negative maternal history. children with a history of intrauterine methamphetamine Comparisons were made between infants using t-tests exposure and 14 control children without a history of and Fisher exact test; however, because more mothers in illicit drug exposure. Six of the methamphetamine- the exposed group smoked cigarettes, comparisons exposed children had also been exposed to cigarette were repeated using two-way ANOVA with smoking as smoking [presumably in utero]. The children were 6–9 the second factor. Results were presented for 135

Birth Defects Research (Part B) 74:471–584, 2005 DOI 10.1002/bdrb AMPHETAMINE AND METHAMPHETAMINE 511 methamphetamine-exposed infants and 160 controls [the selected from a well-baby clinic and matched for discrepancy in numbers was not explained]. Metham- ethnicity. All children were born at 37–41 weeks phetamine-exposed mothers were older, more likely to gestation. Maternal drug history was ascertained by smoke cigarettes, and more likely to be Caucasian. medical record review, interview, or urine toxicology Ethanol was used by 44% and marijuana by 59% of screen at the time of delivery. Of the eight babies with methamphetamine-exposed mothers, compared to none prenatal stimulant exposure, four were exposed to of the control mothers. Mean gestational age was lower cocaine, two were exposed to cocaine and ampheta- in methamphetamine-exposed pregnancies than controls mines, and two were exposed only to amphetamines. (mean7SD 5 37.571.0 compared to 39.771.5 weeks, Three mothers in the drug-exposed group and three Po0.001). [The Expert Panel notes that by design, mothers in the control group had used ethanol during neither group included babies born before 37 weeks pregnancy. Six drug-exposed and two control mothers gestation.] There were no differences between groups in smoked cigarettes. [Results were presented as a group mean birth weight, length, head circumference, or for stimulant-exposed infants and for controls without incidence of obstetrical complications. There were 13 regard to which stimulant was used and without methamphetamine-exposed infants identified as small regard to cigarette or ethanol use.] The mothers using for gestational age compared to two control infants stimulants were significantly older (25.575.1 years, (Po0.001 [statistical test and control for smoking not mean7SD) than control mothers (19.373.3 years), and indicated; crude RR 5 7.70, 95% CI 5 2.08–46.20 calcu- birth weight was significantly lower among drug- lated by CERHR using CDC SABER program]). Mean exposed infants than controls (31007454.9 g compared birth weight and head circumference were significantly to 36007331.3 g, mean7SD). Visual-evoked potentials lower in infants exposed to methamphetamine in all to an alternating checkerboard pattern were recorded trimesters compared to those exposed for only the first or between 4–5 months of age. There was no difference second trimesters. Methamphetamine exposure and between drug-exposed and control infants in the mean smoking were associated with deficits in birth weight latency, amplitude, or duration of evoked waves. Visual and head circumference compared to methamphetamine recognition memory was tested between 6–12 months of exposure alone [although it is not clear whether the age using the Fagan Test of Infant Intelligence. In this effect of smoking explains the decrease in birth weight test, a picture was shown to the infant for a period of and head circumference in infants exposed to metham- time to permit familiarization. The familiar picture was phetamine for three trimesters of pregnancy compared then paired on a screen with a novel picture. The to control]. Withdrawal symptoms were recorded in 49% proportion of time the infant spent looking at the novel of methamphetamine-exposed infants, but only 4% picture was recorded. The full test used the mean of 10 required pharmacologic treatment for withdrawal. The such pairings to derive a novelty preference score. authors concluded that methamphetamine exposure Children with low scores [cut-off not given] were during pregnancy is associated with intrauterine growth considered to be ‘‘at risk.’’ The mean7SD Fagan score restriction in infants born at term, but that other factors in the drug-exposed group (59.1576.2) was lower than such as alcohol, tobacco, and marijuana ‘‘likely contrib- the score in the control group (64.6473.6). Four children uted’’ to the findings. in the exposed group were considered at risk, compared Strengths/Weaknesses: The identification of exposure to none in the control group (Po0.05). Because the by trimester and the inclusion of smoking status in the assessment of visual-evoked potentials showed no drug- comparisons are strengths; however, it is a weakness that associated deficits, the authors concluded that the exposure to substances (including tobacco) were not impaired performance on the Fagan test may indicate controlled in a dose-related manner. Differences in ethnic cognitive impairment in drug-exposed children and origin/ethnicity between groups and failure to interpret could not be attributed to delayed maturation of visual smallness for gestational age in a ethnic-normed manner pathways. The authors could not exclude altered visual are also weaknesses. The restriction to children at term processing or attention as explanations for the poorer has the advantage of providing a group for which performance on the Fagan test. gestational age effects on weight should have been Strengths/Weaknesses: The strength of this study is reduced; however, even within term infants, weight the masking of the neurologist to exposure status. should have been adjusted for gestational age. The Weaknesses include the small size of the convenience finding that 44% of methamphetamine-using women sample, failure to match for gestational age, use of a drank alcohol and 84% smoked make the weight one-tailed P-value, and conflation of cocaine and findings difficult to interpret, although impairments in amphetamines. growth with amphetamines are biologically plausible. Utility (Adequacy) for CERHR Evaluation Process: The behavioral findings are uninterpretable because This report is not useful for the evaluation process. there is no indication that the raters were masked to Boe et al. (1999) presented an abstract from the exposure history. University of California, Davis in which 774 pregnancies Utility (Adequacy) for CERHR Evaluation Process: were identified as methamphetamine-exposed and This report is not useful in the evaluation process with 562,799 pregnancies were identified as unexposed. the limitations noted. Pregnancy outcome information was linked to exposure Hansen et al. (1993), supported by the University of information and relative risks were calculated. [Exposure California at Davis, performed visual-evoked potential and outcome ascertainment were based on hospital measurements on eight infants with a history of prenatal discharge data and birth certificates; no other detail stimulant exposure and eight control infants. Stimulant- was provided.] Statistically significant increases in exposed infants were recruited from a special clinic for relative risk were shown for associations between prenatally drug-exposed infants and controls were methamphetamine exposure and placental abruption,

Birth Defects Research (Part B) 74:471–584, 2005 DOI 10.1002/bdrb 512 GOLUB ET AL. chorioamnionitis, premature rupture of the membranes, hyperactive African American children to an 8-week delivery before 36 weeks, transient tachypnea of the trial of placebo (n 5 10), chlorpromazine (n 5 17), hydro- term newborn, respiratory distress syndrome at term, xyzine (n 5 17), or d-amphetamine (n 5 17). The mean meconium aspiration, intraventricular hemorrhage, and age of the children was 8.7 years (range 5 6.5–11 years), neonatal death. and the mean Wechsler Intelligence Scale for Children Strengths/Weaknesses: The strength of this report is (WISC) full-scale IQ was 85. Side effects were assessed the sample size. Weaknesses include the reliance on birth [when and how assessed is not indicated] and efficacy certificates for history of other exposures and failure to evaluations were undertaken. Side effects that were control for maternal demographic factors and amount of increased by d-amphetamine compared to placebo (at prenatal care. The most important weakness is absence of Po0.1) included decreased appetite (76% compared to a full report [confirmed with authors (Gilbert WM, 20%), insomnia (53% compared to 10%), increased personal communication, October 26, 2004)]. depression (49% compared to 10%), irritability (29% Utility (Adequacy) for CERHR Evaluation Process: compared to 0), headache (41% compared to 10%), and Abstracts are not carried forward into the evaluation stomachache (41% compared to 0). [Fisher exact test process. performed by CERHR shows a significant difference at Felix et al. (2000) presented an abstract from the Pr0.05 between d-amphetamine and control propor- California Teratogen Information Service in which tions for decreased appetite and insomnia.] Palpitations follow-up was reported on 228 pregnant women who and elevated blood pressure did not occur more often had abused amphetamines between the years 1979–1999. with d-amphetamine than placebo. Mean WISC IQ These women were compared to a group of 337 women increased with d-amphetamine therapy to 96.2, whereas who had only been exposed to low grade fever or other there was no change in the placebo group. The individual exposures not considered to be teratogenic. [The abstract dose of d-amphetamine was adjusted based on each does not identify the source of these women; however, child’s response; the mean daily dose was 25 mg. The the Expert Panel assumes that these women telephoned authors expressed surprise that depression worsened in the information service seeking advice about their more children in the d-amphetamine group than the exposures. No information was given on whether any control group (d-amphetamine 8/17, control 1/10 of the amphetamine-exposed women used prescribed [P 5 0.09, Fisher Exact test by CERHR]). [The Expert as opposed to illicit amphetamines.] The authors Panel notes that side effects seemed dose-dependent.] reported a high rate of tobacco, alcohol, and other illicit Strengths/Weaknesses: Strengths of this study include drug use in the amphetamine groups as well as a 24% double-blinding, random assignment, use of multiple loss to follow-up in this group compared to 3% loss to raters, and the 8-week time period. Issues of dose were follow-up in the control group. considered, but the sample is too small for ready There were three stillbirths (1.3%) in the exposed interpretation. The sample was limited to a single group and none in the control group. There was no ethnicity, African-American, which is underrepresented difference in the incidence of major structural defects in other samples. Results were valid and consistent with between the groups (amphetamine-exposed 5%, control those of other studies. Use of a control psychoactive group 3.4%). The defects in the amphetamine-exposed substance (chlorpromazine) that increases appetite and group included tracheal-esophageal fistula, pyloric ste- weight gain, made the d-amphetamine effect more nosis, transposition of the great vessels, cystic adenoma- dramatic. A weakness is the low pretreatment IQ of toid malformation, vascular ring constricting the trachea, subjects. In addition, the study was conducted in 1972 unilateral inguinal hernia, and multiple malformations and used what are now considered antiquated forms of associated with a chromosomal anomaly. Among the psychiatric diagnoses (e.g., ‘‘neurosis’’). It is therefore amphetamine-exposed children, there was a 21% inci- also unclear if any of the children in the ‘‘hyperactive’’ dence of minor anomalies compared to 11% in the control group, including the one who became psychotic, was in group. Of 80 amphetamine-exposed infants examined fact suffering from juvenile bipolar disorder, a diagnosis blindly after 1 week of life, five (6.2%) had neurologic that did not exist at that period. signs including abnormal tone and irritability. None of Utility (Adequacy) for CERHR Evaluation Process: the control infants demonstrated neurologic abnormal- This study is of limited-to-moderate adequacy for the ities [denominator not given for control group]. evaluation process. Strengths/Weaknesses: Strengths include use of a Efron et al. (1997), supported by a hospital research masked neurologic examiner and consideration of fund, randomized 114 boys and 11 girls with ADHD to a tobacco use. Weaknesses include failure to specify 2-week trial of d-amphetamine or methylphenidate, maternal ethnicity and lack of definition of minor followed by a 24-hr wash-out period, followed by a malformations. The most important weakness is the 2-week trial of the other stimulant. The mean age of the absence of a full report [confirmed with authors subjects was 104.8 months (range 5 60–179 months). The (Chambers C, personal communication June 24, 2004)]. d-amphetamine dose was 0.15 mg/kg bw/dose and the Utility (Adequacy) In the CERHR Evaluation Process: methylphenidate dose was 0.30 mg/kg bw/dose, both Abstracts are not carried forward into the evaluation given twice/day, rounded to the nearest tablet size. process. Investigators, subjects, teachers, and family members 3.1.2 Adverse effects in children were blinded to the identity of the medication. Evalua- 3.1.2.1 General side effects. Adverse effects of tions included effectiveness endpoints [not discussed amphetamine preparations listed in the product labels here] and side effects, derived from a behavior ques- are given in Section 2.2.1.1. tionnaire completed by parents at the end of each 2-week Greenberg et al. (1972) supported by NIH, PHS, and treatment period. A list of 17 common side effects was Pfizer, randomized 61 of an initial group of 76 presented on the questionnaire with a ranking scale for

Birth Defects Research (Part B) 74:471–584, 2005 DOI 10.1002/bdrb AMPHETAMINE AND METHAMPHETAMINE 513 the evaluation of severity ranging from 0 (not present) to epileptiform EEGs. The authors concluded that a normal 9 (severe). Side effects were evaluated with regard to EEG before stimulant therapy was reassuring that whether they were present or absent at baseline and on seizures would not occur on therapy. They were not treatment and with regard to mean of the severity ranks. convinced that the stimulant therapy was the cause of the Trouble sleeping and poor appetite were more commonly seizures that occurred based on the timing of the seizures present after 2 weeks of d-amphetamine than at baseline with respect to the start of stimulant therapy, and based (poor sleeping: 54% at baseline and 70% on d-ampheta- on the low overall incidence of seizures (2%) in the mine; poor appetite: 34% at baseline, 69% on d- stimulant-treated population. [The specific stimulants amphetamine). Fingernail biting and unusual happiness were not named except in the four cases of seizure. The were reported to be more common at baseline than on d- stimulants used in these cases included methylpheni- amphetamine therapy (fingernail biting: 50% at baseline, date and d-amphetamine.] 40% on d-amphetamine; unusual happiness 42% at Strengths/Weaknesses: Strengths include use of DSM- baseline, 26% on d-amphetamine). The remainder of the III and IV diagnoses as well as parent and teacher symptoms were identified as present in similar propor- reports. Several weaknesses preclude this study from tions of children at baseline and on d-amphetamine. The determining if stimulants are associated with seizures in authors concluded that many side effects identified on children with epileptiform EEGs. Authors state the stimulant medication may be side effects associated with obvious, if the EEG is normal, the child is at low risk the underlying disorder rather than due to the medica- for a seizure; if the EEG is abnormal, the child with tion therapy. ADHD is at considerable risk for the eventual occurrence Strengths/Weaknesses: A strength of this study is the of a seizure, without attributing the effect to medication. double-blind crossover design, with children used as The EEGs were performed anywhere from before to 8 their own controls. The pre-medication period served as weeks after initiation of stimulant therapy, thus making it the non-stimulant comparison and d-amphetamine and difficult to ascertain whether the EEG findings in fact methylphenidate were compared to each other. Other reflected neurological issues preceding the treatment. strengths include use of DSM-IV criteria to diagnose Major weaknesses also include lack of information ADHD, as well as a standardized child behavioral regarding masking of EEG readers, current treatment of measure. Side effects were rated not only in a yes/no child, and family history of epilepsy. The incidence of manner but also by level of severity. Statistical methods epileptiform EEG in this study (15.4%) is much higher were clearly specified. A weakness, as noted by study than the estimated incidence of EEG abnormalities in an authors, is potential bias introduced by parental report- unselected population of children (2%). ing. For example, many of the potential ‘‘side effects’’ Utility (Adequacy) for CERHR Evaluation Process: (behaviors on the Side Effects Rating Scale) decreased This study is of marginal adequacy for the evaluation significantly while on one or both stimulants, possibly process. indicating that parents could have been ‘‘motivated’’ to 3.1.2.2 Onset or worsening of tics. Tourette disorder have their child get better on the trial. Also noted by is a chronic neurologic disorder characterized by study authors is that the moderate doses administered repeated and involuntary body movements (tics) and twice daily in this study cannot be generalized to uncontrollable vocal sounds. Tics can include eye regimens of different dose levels and frequencies. blinking, repeated throat clearing or sniffing, arm Another weakness is that each phase of drug exposure thrusting, kicking movements, shoulder shrugging, or was 2 weeks followed by ‘‘wash-out’’ and then exposure jumping. A large proportion of children with Tourette to the next drug for 2 weeks. It would have been better to disorder have comorbid ADHD (Leckman, 2003). In 1974, test one drug for a longer period of time. In addition, a case report was published describing a 9-year-old boy children were of a wide age range, spanning from treated with methylphenidate for hyperactivity who preschool to post-pubertal (5–15 years). Because the developed Tourette disorder on therapy (Golden, 1974) study was done in Australia, the children may have been and since that report, additional articles have described facing somewhat different environmental challenges tics or Tourette disorder in association with stimulant than those in the United States. therapy (Table 31). There is only one controlled study of Utility (Adequacy) for CERHR Evaluation Process: amphetamine exposure and development of tics (Efron This study is adequate for use in the evaluation process. et al., 1997). Hemmer et al. (2001), supported by the Crown Family, Strengths/Weaknesses: The overall data set is weak. performed EEGs on 179 males (3–20 years old) and 55 The one controlled study (Efron et al., 1997) has serious females (3–19 years old) with ADHD. The mean age of weaknesses (see Table 31). the subjects was 9–10 years. Epileptiform EEGs were Utility (Adequacy) for CERHR Evaluation Process: obtained in 36 subjects before stimulant therapy or up to The data set is not useful in the evaluation process. 8 weeks after the initiation of therapy. The decision to 3.1.2.3 Substance abuse disorder. accept or decline stimulant therapy was made by parents Biederman et al. (1999), in a study supported by the and did not seem to be influenced by the EEG results. NIMH and NIDA, evaluated the risk of substance abuse There were 175 subjects treated with stimulants of whom disorders associated with psychotropic medication for 30 (17%) had had an epileptiform EEG. Of the 29 subjects treatment of ADHD. Data were obtained and reanalyzed who declined stimulant therapy, six (21%) had had an from an ADHD longitudinal genetics study conducted in epileptiform EEG. Seizures occurred in four subjects 260 families. Females were not evaluated because most [follow-up period not specified]. All of the subjects with medicated subjects were male and subjects younger than seizures were in the stimulant group, although one child 15 were excluded due to significantly younger age of the had a seizure after being off stimulant medication for 2 medicated versus unmedicated subjects. Subject groups months. Three of four children with seizures had prior consisted of Caucasian males who were Z15 old years

Birth Defects Research (Part B) 74:471–584, 2005 DOI 10.1002/bdrb 1 OU TAL. ET GOLUB 514 Table 31 Studys of Tics in Children Treated with Stimulant Medication

Stimulant dose (mg/day Medication except where indicated) Characteristics of the children Outcome Commentsa Reference

Methylphenidate, Not reported 32 Tourette disorder patients 17/32 experienced worsened This study is descriptive, with no doses Golden, 1977 d-amphetamine, who had been exposed to symptoms when on stimulants or ages given. Other medications in methamphetamine stimulants (unspecified ages) the mix were not considered. Omitted from consideration were 39/45 subjects with preexisting Tourette disorder that did not worsen. d-Amphetamine, methyl- Not reported 4 boys, 8–11 year old Tourette disorder developed This study indicates the independence Lowe et al., phenidate, d-ampheta- and continued after of Tourette disorder from medication, 1982 mine/pemoline, medication but includes only 4 children who methylphenidate/ used 3 medications alone and in pemoline combination with no indication of dose. Methylphenidate, Not reported 200 children, 48 with Tourette 8/48 tics worsened Strengths: Inclusion of a fairly homo- Erenberg et al., d-amphetamine, disorder geneous population of 48 children 1985 pemoline with preexisting Tourette disorder. Weaknesses: Retrospective review, dose range not given; there was no independent diagnosis; follow-up duration was variable.

it eet eerh(atB 44154 2005 74:471–584, B) (Part Research Defects Birth Methylphenidate, Not reported 170 twins and individuals 50% with worsening of tics, Strengths: Twin study (in 6 monozygotic Price et al., pemoline, some developed tics twins, discordant for medications; 1986 d-amphetamine other twin developed Tourette disorder suggesting genetic basis). Weaknesses: Doses were not given; Tourette disorder was preexisting; the relationship of tic worsening to therapy was vague. Methylphenidate, Up to 90 mg/day methyl- 45 hyperactive boys age 6–12 10/45 had increase in tics or Strengths: Compared methylphenidate Borcherding d-amphetamine phenidate; up to 45 mg/ years development of tics only on and d-amphetamine; high doses used. et al., 1990 day d-amphetamine methylphenidate; 6/45 only Weaknesses: Tics were not clearly dis- on d-amphetamine, and 11 on tinguished from other behavioral both changes and disorders (e.g., obses- sive-compulsive disorder) that developed over the course of the study. Methylphenidate, Methylphenidate dose 122 children with ADHD age Tic/dyskinesias occurred in Strengths: Large study with standar- Lipkin et al., d-amphetamine, (mean7SD): 21.1711.7 3.6–15.8 years 8.2% of children treated with dized diagnosis of ADHD (DSM- 1994

O 10.1002/bdrb DOI pemoline with tics, 24.4717.2 medication IIIR). without tics; d-amphe- Weaknesses include retrospective tamine dose design, unknown weight-adjusted (mean7SD): 14.275.2 doses, and the reliance on chart with tics, 15.876.8 review for parental assessments of without tics unusual tic-like movements such as ‘‘eye bugging.’’ Prior medication history not detailed. AMPHETAMINE AND METHAMPHETAMINE 515 and had previously received medication for ADHD (n 5 56), had ADHD but were not medicated (n 5 19), or did not have ADHD (n 5 137). The average duration of treatment was 4.4 years. [The types of medications used et al., 1997 2001 1997 were not specified.] Multiple logistic regression was Castellanos Varley et al., Efron et al., used to correct confounding by age, socioeconomic status, lifetime risk of conduct disorder, and substance use disorders in parents. Substance use disorders were examined for alcohol, marijuana, hallucinogens, cocaine/ stimulants, and tobacco. ADHD subjects who had been medicated had a significantly reduced risk of any substance use disorder compared to unmedicated subjects with ADHD (OR 5 0.15, 95% CI 5 0.04–0.6). Un- medicated subjects with ADHD had a significantly increased risk of any substance abuse disorder compared

ve.’’ to controls without ADHD (OR 5 6.3, 95% CI 5 1.8–21.4). ¨ With the exception of tobacco use, the medicated group had reduced risk of all other individual substance use disorders compared to the unmedicated ADHD group, wide range of doses; double blind. drugs were evaluated. Usedscale Barkley as in Ahmann et al., 1993. item scored was ‘‘ticsmovements,’’ or (in nervous this casescale the may rating overestimate prevalencetics of at baseline, confoundingsis); analy- the duration ofonly the 2 study weeks was onnotation each of drug; whether no subjects‘‘drug were naı but the sample size was too small to evaluate statistical Strengths: Two drugs were compared; Weaknesses: Small sample size. Strengths: Moderately high doses of two Weaknesses: Lack of placebo control; the Strengths: Comparison of 3 medications. Weaknesses: Retrospective chart review. significance for individual substances. The study authors concluded that pharmacotherapy is associated with an 85% reduction in risk for substance abuse disorders in youths with ADHD. Strengths/Weaknesses: Strengths of this study include well articulated competing hypotheses and longitudinal amphetamine. design, as well as masked assessment and careful d- definition of the sample restricted to white males older than 15 years. Authors pay considerable attention to quality control and structured DSM-IIIR interviews were used to establish the diagnosis of substance use disorders cularly with Methylphenidate-associated tics returned to placeboline base- with continued methylphenidate treatment. present in 35% ofbaseline subjects and at 26–28% ofjects sub- after 2 weeksdrug on either not more frequent inmedication any category (because of small numbers, abuse and dependence were Reversible increase in tics, parti- Tics or ‘‘nervous movements’’ Tics present in 7.8% of children; analyzed as a single category as substance abuse disorder). Important variables including comparisons of treated and untreated ADHD children were considered in analyses. The study addresses, implicitly, the issue of SD self-medication by determining that treated groups had a 7

ampheta- diminished odds ratio of substance abuse. Other 27.6 months; d-

7 strengths include control of parental substance use disorder and presentation of outcomes as an aggregate of any substance use disorder and disaggregated by

SD 104.8 substance. Limitations include use of an exclusively 2.0 years 7

7 tertiary-referred white sample, so findings may not 9.4 age range 60–179 months (5–15 years) 126 children on mine, 13 on pemoline apply to less privileged or lower income risk groups. 20 children, mean age 125 children with ADHD, mean Other weaknesses include lack of specification of the drugs that the treated members of the cohort received. The authors correctly identified lack of power as diminishing confidence in null findings. Other weak-

ampheta- nesses are that the age ranges of subjects and the time d- period they had been off medication were not specified. Utility (Adequacy) for CERHR Evaluation Process: This study is useful for the evaluation process. Barkley et al. (2003) (also reported in Fischer and amphetamine 10–45 d- mine 0.3 mg/kg bw/ day kg bw/day, Barkley (2003)), in a study supported by the NIMH, Methylphenidate 35–90; Methylphenidate 0.6 mg/ Not reported 374 children on methylphenidate, examined possible associations between stimulant medication therapy during childhood or adolescence and substance use during adolescence and young adulthood. During each evaluation period, subjects or their parents were asked about stimulant therapy, behavior, mental health, illicit drug use, and education history; psychological tests were conducted and subjects were rated according to scales. Groups consisted of 91% males and 9% females. Racial distribution was 94% Caucasian, 5% amphetamine amphetamine amphetamine, d- d- d- pemoline black, and 1% Hispanic. Subject ages were 12–20 years at None of the studies screened for substance abuse. Methylphenidate, a Methylphenidate, Methylphenidate, adolescent evaluation and 19–25 years at adult

Birth Defects Research (Part B) 74:471–584, 2005 DOI 10.1002/bdrb 516 GOLUB ET AL. evaluation. During the adolescent evaluation, 119 hyper- (25% in untreated compared to 47% in treated, P 5 0.018). active subjects were available for interview and parents Additional analyses indicated that risk of cocaine use were questioned about stimulant therapy during child- was primarily mediated by severity of conduct disorder hood. Ninety-eight were treated with stimulants during and not by use of stimulant medication. Increased childhood, 21 were not. Percentages treated with each duration of stimulant treatment was not found to type of stimulant during childhood were 80% methyl- adversely affect risk of substance use. No significant phenidate, 3% d-amphetamine, and 20% pemoline. Some differences in adult substance abuse/dependence rates subjects received more than one type of stimulant; d- (diagnosed by DSM-III-R criteria) were noted in hyper- amphetamine was given to 2% and pemoline to 22% of active subjects who were or were not treated with the children in the methylphenidate group. All children stimulants in childhood or during high school. [There in the pemoline group had also received d-amphetamine. was no statistical analysis for abuse/dependency in Mean durations of treatment during childhood were 44.8 adults.] The study authors concluded that there was no months for methylphenidate, 32.8 months for ampheta- compelling evidence that stimulant treatment of children mine, 13.3 months for pemoline, and 40.2 months for or adolescents with ADHD leads to increased risk of stimulants in general. During the adult evaluation, 147 substance experimentation, use, dependence, or abuse by hyperactive subjects were questioned about stimulant adulthood. treatment during high school, but were not asked to Strengths/Weaknesses: A strength of this study is that identify the specific stimulant medication taken. Thirty- substance abuse was defined by DSM-IIIR criteria. This two subjects were treated with stimulants and 115 were study considered not only substance use but also not treated with stimulants during high school. Mean frequency/quantity and distinguished experimentation duration of stimulant treatment during high school was from problem use. Initiation and experimentation did not 26.6 months. Seven of the subjects were receiving differ by stimulant medication exposure status. Another stimulant treatment at the time of the interview. [Severity strength is consideration of duration of treatment, with of ADHD symptoms and conduct disorders were the considerable detail provided on length of time subjects only potentially confounding factors considered.] received different medications. Two time frames of At the adolescent evaluation, subjects were asked if stimulant medication use and drug use were examined; they had ever tried cigarettes, alcohol, marijuana, uniquely, drug use was examined while a few subjects hashish, cocaine, heroin, hallucinogens, unprescribed were still receiving medication. Importantly, the study stimulants, sedatives, or tranquilizers. [Information showed that cocaine use was related to adolescent about frequency of use was not obtained and substance treatment, but that this relationship was lost when abuse/dependency was not considered.] The propor- severity of ADHD was statistically controlled; this tions of hyperactive subjects who had ever tried any of finding emphasizes the need for such control in other the substances were similar in the stimulant-treated and studies. In addition to the paucity of control variables untreated groups by chi-square analysis. No significant (including family history), a major weakness noted by differences were found when all stimulants (cocaine, authors on page 100 of the Pediatrics article is that the amphetamines) were combined or when duration of assessor was not masked to stimulant exposure history. stimulant therapy was considered. It is a strength and a weakness that the authors specify Subjects were questioned in adulthood about their use the medications to which the children were exposed, of alcohol, marijuana, cocaine, amphetamines/speed, but because of small cell sizes and a predominance of any stimulant, hallucinogens, narcotics, sedatives, or methylphenidate, stimulants were only evaluated as a other drugs. Frequencies of substance use were log single generic exposure. The authors did use standard transformed, due to high standard deviations, and instruments. Weaknesses include the fact that tobacco compared by ANOVA. Stimulant treatment in childhood use was not adequately evaluated. The authors were did not significantly increase the frequency of any type correct in noting that it is difficult to ascertain whether of substance use in early adulthood. The frequency of the weak association between high school stimulant cocaine use was significantly higher (P 5 0.043) in treatment and cocaine use was an artifact of multiple subjects who were treated with stimulant medications comparisons. Another conceptual weakness they did not in high school, but the results were no longer significant consider, however, is that perhaps children who are more when corrected for severity of ADHD and conduct deviant and therefore, with or without treatment, more disorder. [Table 3 of the study, which presents effects prone to substance use disorders are more likely to of high school stimulant treatment, lists group num- continue to be treated into high school. Although bers for childhood treatment (n 5 21 untreated, 98 important, it was not stated whether subjects treated in treated) instead of high school treatment (n 5 115 high school received stimulants at both ages, especially treated, 32 untreated).] The proportion of subjects who as findings were mediated by severity of ADHD. The ever used each of the substances was analyzed by chi- authors themselves point out that their study design did square. If statistically significant findings were observed, not permit them to identify the temporal sequences of a binary logistic analysis was conducted to adjust for conduct disorder and substance use disorder, leading to severity of ADHD symptoms and conduct disorders. difficulties in interpretation particularly of this worri- A greater percentage of adults who were treated some finding of possible connection of stimulant treat- with stimulants in childhood and in high school used ment to cocaine use. cocaine at least one time (5% untreated compared to 26% Utility (Adequacy) for CERHR Evaluation Process: treated in childhood, P 5 0.037 and 20% untreated This study is useful for the evaluation process. compared to 40% treated in high school, P 5 0.016). Lambert and Hartsough (1998) and Lambert (2002), Due to increased cocaine use, the use of any stimulant supported by the Tobacco-Related Disease Research was also increased in adults treated during high school Program, examined the effects of ADHD and stimulant

Birth Defects Research (Part B) 74:471–584, 2005 DOI 10.1002/bdrb AMPHETAMINE AND METHAMPHETAMINE 517 treatment on tobacco use and substance dependency in a were not significant for either tobacco or cocaine longitudinal sample of 492 adults. According to informa- dependency.] tion provided in the Lambert (2002) study, subjects were A comparison of subjects who had ADHD as children born in the San Francisco area between 1962–1968. About with subjects who did not have ADHD as children 22% of the subjects were female and 23% represented showed that subjects with ADHD began smoking ethnic minority groups. The authors reported that among regularly at a younger age, had a higher rate of smoking subjects using stimulant medications, 69% used only as adults, and had higher cocaine dependency rates. The methylphenidate, 16% used combinations of methylphe- study authors concluded that there is a possible link nidate and other stimulants, and 15% used other CNS between stimulant medication and rates of smoking and stimulants (amphetamines, pemoline). At various stages tobacco and cocaine dependency in adulthood. throughout their lives, the subjects were questioned In the Lambert (2002) study, subjects were divided into about their use of tobacco, alcohol, marijuana, stimulants, groups of 268 who received no CNS stimulant treatment and cocaine. A total of 399 subjects were said to be and a group of 131 who received stimulant treatment. available for interview. [These studies seem to have [According to Table 18.2 in the study, the group with numerous discrepancies or mathematical errors in text no stimulant treatment was comprised of 162 subjects compared to tables or in different parts of tables. In without ADHD and 106 with ADHD (41 severe, 25 adding numbers presented in some study tables, it moderate, and 40 mild). The stimulant treatment group seems that either mathematical errors were made or was comprised of 10 subjects without ADHD and 121 more than 399 subjects were evaluated for some subjects with ADHD (62 severe, 48 moderate, and 11 endpoints (i.e., Table 3 in Lambert and Hartsough, mild.] The percentage of subjects who had not yet 1998). In other cases, fewer than 399 subjects were become regular smokers was significantly higher included in analyses and it is not clear if or why some (Pr0.05 by Lee Desu statistic) in the untreated group subjects were excluded (Table 5 in Lambert and (B60%) compared to the treated group (45%). The same Hartsough, 1998).] subjects were evaluated according to the age when In the Lambert and Hartsough (1998) study, subjects stimulant treatment was ended: age 10, age 11–13, or were placed into hyperactive or control groups. Accord- after age 14. Stimulant treatment seemed to protect ing to information presented in Lambert (2002), there against smoking during childhood. However in adult- were 217 hyperactive subjects (136 with primary hyper- hood, smoking rates were significantly higher (Po0.001 activity with no causal explanation, 31 with secondary by chi-square) in treated groups (41%) compared to the hyperactivity possibly due to organic factors, and 50 with untreated group (19%). Adjusted OR were calculated. untreated hyperactivity). There were 182 controls (141 [The confounding factors considered in the analyses age controls and 41 with non-ADHD behavioral pro- are not clearly identified, but it seems that childhood blems). Information in Lambert (2002) indicates that only conduct disorders were considered in addition to 80% of the primary hyperactive group and 66% of the socioeconomic status, cognitive ability, and ethnicity. secondary hyperactive group received stimulant treat- It is not clear how many subjects were included ment. [It is not clear why untreated subjects in the and how the subjects were classified in calculating primary and secondary hyperactive group were not put the OR. It is assumed that, as in previous analyses, into the untreated hyperactive control group.] It seems subjects with and without ADHD were collapsed into that about 3% of controls received stimulant treatment. the same groups based on stimulant exposure.] In the Subgroups of individuals were grouped together based group treated with stimulants for 41 year, OR were on similarity of health, familial, educational, and social described as significant for daily smoking (2.817) and background factors. [There was no discussion of cocaine dependency (2.251) in adulthood. In subjects adjustment for additional confounding factors such as exposed for o1 year, a significant OR (3.951) was severity of ADHD.] The rate of smoking in adults who obtained for daily smoking in adulthood [95% CI were had ADHD as children and who never used stimulant not given]. ADHD severity was found to be significantly medication (n 5 47) was 37.0%; for adults who had used related to tobacco, cocaine, and stimulant dependency in stimulant medication for up to 1 year (n 5 28), the rate of adulthood. smoking was 22.0%; and for adults who had used Strengths/Weaknesses: A strength of these studies is stimulant medication for Z1 year (n 5 52), the rate of the emphasis on cigarette consumption, which possibly smoking was 40.9% (Po0.03 for never-used compared to indicated self-medication, as higher rates of smoking use Z1 year, by w2). The Mantel-Haenszel test for linear were found in untreated ADHD subjects. However, the trend was also significant for duration of stimulant use inconsistencies in sample sizes and inaccuracies in study (Po0.01). Significant linear trends (Po0.03) were noted tables make conclusions tenuous. Other weaknesses for rates of tobacco dependency in adults who had include the very inadequate description of sample in ADHD as children and who never used stimulant terms of ethnicity, social class, parental substance use, medication (n 5 81; 32.1% rate) or had used stimulant severity of ADHD, and many other potential confounds. medication for up to 1 year (n 5 9; 38.5% rate) or Z1 year In addition, the authors tended to make sweeping (n 5 84; 48.8% rate). Significant linear trends (Po0.05) conclusions on the basis of univariate analyses. All of were also noted for cocaine dependency in adults these weaknesses make interpretation of reported find- who had ADHD as children and who never used ings problematic. stimulant medication (15.0%) or had used stimulant Utility (Adequacy) for CERHR Evaluation Process: medication for up to 1 year (17.9%) or Z1 year (27.4%). These studies are not useful for the evaluation process. [The text states that statistical significance by v2 was Paternite et al. (1999) and Loney et al. (2002), from the obtained for cocaine dependency, but the legend of same group, examined the effects of stimulant medica- Table 7 in the study indicates that results of v2 analyses tion in childhood on substance use in adulthood. One of

Birth Defects Research (Part B) 74:471–584, 2005 DOI 10.1002/bdrb 518 GOLUB ET AL. the reports (Paternite et al., 1999), partially supported by attempts). Only one adverse significant association the NIMH, indicates that the medicated subjects were with medication was reported and it is surprising: treated with methylphenidate, but the other report ‘‘better response to treatment and reduced likelihood of (Loney et al., 2002) refers generically to ‘‘CNS stimulants high school graduation.’’] such as [methylphenidate].’’ [The Expert Panel recog- In the study by Loney et al. (2002), rates of involve- nizes in reviewing these reports that few, if any, of the ment (experimentation, continuation, or escalation of subjects may have been treated with amphetamines.] use) with alcohol, tobacco, barbiturates, tranquilizers, Subjects were selected from 219 [listed as 285 in one stimulants, marijuana, glue, cocaine, LSD, and opioids study, but this figure seems to be an error] boys (98% were compared between ADHD subjects who either were white) who were referred to the University of Iowa child or were not treated with methylphenidate. The analyses psychiatric clinic at 4–12 years of age. Boys were controlled for year of birth and inattention, overactivity, diagnosed as having hyperkinetic reaction (70%) or or aggressive defiance symptoms. In unmedicated minimal brain dysfunction (30%). By more current compared to medicated subjects, adult involvement standards, B70% of the boys would have been diag- was significantly increased (Po0.05) for tobacco, stimu- nosed with ADHD and the term ADHD is used in the lants, glue, and opioids. [The unit on the Y axis of later study for convenience. Aggressiveness was noted in involvement graphs (Figures 17.1 and 17.2 of the study) 7% of the boys who would have likely received a is not specified and it is not clear what kind of analysis diagnosis of oppositional defiant disorder according to was conducted.] According to the study authors’ inter- more recent terminology. Based on treatment preferences pretation of the data, medicated subjects progressed less of 3 different physicians, 182 of the boys received far along the path from experimentation to continued stimulant medication and 37 were not given medication. use. Significantly fewer (Po0.05) medicated versus At follow-up during adulthood (21–23 years old), 97 of unmedicated subjects (respective percentages) had ex- 121 subjects medicated with methylphenidate in child- perimented with glue (B22 vs. 38%), stimulants (38 vs. hood were available for evaluation. [It seems that the 121 58%), LSD (B30 vs. 49%), and opioids (B23 vs. 42%). medicated subjects were selected from the group of 182 Medicated versus unmedicated subjects (respective medicated subjects. The number of untreated subjects percentages) had significantly lower rates of alcoholism available for evaluation in adulthood was not speci- (27 vs. 56%, P 5 0.002) and antisocial personality disorder fied.] The medicated subjects were treated between (24 vs. 44%, P 5 0.004). Drug abuse rates were similar 1967–1972 at a mean age of 8.8 years. Mean methylphe- between the two groups (17 vs. 19%). Loney et al. (2002) nidate dose was 32 mg/day (range 8–80 mg/day) and concluded that their studies did not indicate a negative mean duration of treatment was B30 months [reported effect of childhood methylphenidate treatment on future as 30.4 and 36 months in the two articles] with a range drug use, but suggested that further research is needed. of 1–76 months. [It was not stated how many untreated [The Expert Panel notes evidence of self-medication, as subjects were included in analyses.] non-treated subjects were more likely to be ‘involved’ In the Paternite et al. (1999) study, regression analyses with tobacco and stimulants.] were conducted to determine associations between Strengths/Weaknesses: Strengths include a relatively methylphenidate dose, response, or treatment duration lucid exposition of the technical problems in this field and alcoholism, drug abuse disorder, psychiatric condi- and an ethnically homogenous sample that consisted of tions, and measurements of social function and IQ. Child all preadolescent subjects at the time of intake. Other age, symptom dimensions, and the two other medication strengths were that treated and untreated subjects had variables were held constant in each analysis. Neither ADHD and that inattention/hyperactivity and aggres- alcoholism nor drug abuse disorders were significantly sion were explored separately. In the Paternite et al. associated with methylphenidate treatment, although the (1999) study, regression analyses were applied to authors concluded that there was a trend between consider many putative associations, including some increased dose and fewer diagnoses of alcoholism that were unique (e.g., social function). Weaknesses (r 5 À0.2, Po0.10). [Most data were not shown; only include the need to reclassify now outdated clinical values approaching or reaching statistical significance measures to fit modern criteria and the use of other were listed in tables.] The only negative finding related outdated measures for outcomes, as well as lack of to methylphenidate treatment was an association be- consideration of family risk factors, both genetic and tween better response to treatment and reduced like- environmental. The small size of the unmedicated lihood of high school graduation (r 5 À0.34, Po0.01). subgroup (n 5 37) would tend to bias the evaluation Additional findings included associations between in- against finding a negative effect in the unmedicated creased dosage and reduced suicide attempts; better group. It is not clear how the follow-up medicated medication response with improved psychiatric out- subjects were selected or how many untreated subjects comes and social functioning; and longer treatment were followed to adulthood. For example, in these two duration with improved psychiatric outcomes, higher articles that are considered together, authors fail to IQ, and better reading scores. Significant associations or describe clearly how an initial sample of 182 treated trends were noted between inattention-overactivity and subjects became 121 and then 97. unemployment and adverse outcomes on some psychia- Some weaknesses in the interpretation of the Loney tric or behavioral measures. Associations or trends noted et al. (2002) study were noted. The main finding was that for aggression were drug abuse disorder, antisocial medicated subjects were less likely to go from ‘experi- personality disorder, and adverse outcomes on some mentation to continued use’ (terms not defined). Drug psychiatric or behavioral measurements. [The Expert abuse (not defined) was reported to be similar among Panel notes that a number of unique positive associa- treated and non-treated groups, but medicated subjects tions with medication were noted (e.g., reduced suicide were less likely to ‘experiment’ with most drugs.

Birth Defects Research (Part B) 74:471–584, 2005 DOI 10.1002/bdrb AMPHETAMINE AND METHAMPHETAMINE 519 Therefore, the conclusion that drug abuse rates are not The study authors concluded that results suggested an impacted by medication is problematic. As fewer association between stimulant treatment in childhood medicated subjects experimented, it seems that the and reduced risk of subsequent drug and alcohol proportion of medication subjects who experimented disorders. and went on to continuous drug use was higher than the Strengths/Weaknesses: This study reviewed numer- proportion of non-medicated subjects. A statistical ous studies, some of which were not published. Strengths control is needed for this finding. of this study include statistical analyses (albeit of data of Utility (Adequacy) for CERHR Evaluation Process: heterogeneous quality and composition) and care in The Paternite et al. (1999) study is of limited utility; the checking that no one study heavily influenced the Loney et al. (2002) study is not useful for the evaluation combined estimates, as well as attention to publication process. bias. Other strengths were largely conceptual. The Wilens et al. (2003) conducted a meta-analysis of authors raised an important issue about baseline severity studies examining possible associations between long- of ADHD in moderating impact of stimulant treatment; term medication for treatment of ADHD and substance unfortunately, part of this analysis was based on use disorders. The studies reviewed in the analysis are unpublished observations (Barkley et al., 2003). Another listed in Table 32 and included published reports interesting point is that children from families with a identified in a PubMed search, data presented at history of substance use may be more resistant to scientific meetings, and unpublished findings. [Pub- stimulant treatment. Children with more severe opposi- lished studies are reviewed in detail above.] Included in tional and aggressive disorders (and thus at greater risk the analysis were prospective studies examining subjects of later substance use disorder whether treated or during adolescence (Biederman et al., 1999; Molina and untreated) are more likely to receive stimulant treatment Pelham, 1999 abstract cited in Wilens et al., 2003) and than children at lower baseline risk. It can either be young adulthood (Lambert and Hartsough, 1998; Pater- regarded as a strength or a weakness that samples were nite et al., 1999; Loney et al., 2002). One retrospective heterogeneous in the age of follow-up with two examin- study examined subjects during adulthood (Huss, 1999 ing adolescents who were presumably quite early in the abstract cited in Wilens et al., 2003). A total of 674 substance use disorder trajectory and the remainder medicated and 360 unmedicated subjects with ADHD examining adults. Another weakness is that the studies were included in the meta-analysis. The analysis did not reviewed used differing measures of varying validity to examine nicotine use. ORs for drug and alcohol document substance use disorder. Problems also include substance abuse disorders are listed in Table 32. An conflation of prospective and retrospective studies and OR41 indicates a protective effect of medication, exclusion of cigarette/tobacco use as an outcome when it whereas an ORo1 indicates an adverse effect of was a primary outcome of a limited study that found an medication. [According to the ORs and CIs listed in adverse effect of childhood stimulant treatment (Lambert Table 32, none of the studies demonstrated a significant and Hartsough, 1998). adverse effect of medication.] The pooled OR of 1.9 Utility (Adequacy) for CERHR Evaluation Process: (95% CI 5 1.1–3.6) was consistent with a nearly two-fold This study is of marginal utility for the evaluation reduction in risk of substance abuse disorders in youths process. medicated versus unmedicated for treatment of ADHD. As noted above, some studies examining the effects of Additional analyses indicated that no one study heavily ADHD medications also found associations between influenced outcome. Studies that controlled for baseline ADHD or conduct disorders and substance use (Lambert severity of ADHD were found to have larger ORs and Hartsough, 1998; Chilcoat and Breslau, 1999; [statement not consistently supported by drug data in Paternite et al., 1999). Numerous studies examined Table 32]. A greater protective effect of medication was possible associations between ADHD, independent of found in studies examining adolescent (OR 5 5.8) versus treatment, and substance abuse [not considered here].A adult subjects (OR 5 1.4) [95% CIs were not presented]. review by Wilens (2002) concluded, ‘‘There is a robust

Table 32 Meta-Analyses for Studies Examining Substance Abuse in Subjects Who Were or Were Not Medicated for ADHD

Number of ADHD subjects OR (95% CI) Similar baseline Reference severity? Medicated Unmedicated Drugs Alcohol

Lambert and Hartsough, 1998 No 93 81 0.47 (0.22–1.0) 0.6 (0.32–1.1) Biederman et al., 1999 Yes 145b 45b 3.9 (1.8–8.1) 8.1 (3.9–17.2) Huss, 1999 abstract cited in Wilens et al., No 98 21 2.2 (0.99–5.1) No data 2003 Loney et al., 2002 Yes 182 37 1.1 (0.46–2.8) 3.6 (1.7–7.4) Molina and Pelham, 1999 abstract cited Yes 53 73 4.6 (1.5–14.5) 6.6 (1.4–30.2) in Wilens et al., 2003 Barkley unpublished data cited in Yes NS NS 0.83 (0.29–2.3) 0.98 (0.36–2.7) Wilens et al., 2003a From Wilens et al. (2003). NS, not specified. aThis study may have been published later as Barkley et al. (2003). bAccording to CERHR review of this study, there were 56 medicated and 19 unmedicated subjects with ADHD.

Birth Defects Research (Part B) 74:471–584, 2005 DOI 10.1002/bdrb 520 GOLUB ET AL. literature supporting a relationship between ADHD and * Different durations of medication treatment. SUD [substance use disorders]. Non-comorbid ADHD * Stimulant exposure before study intervention and seems to confer an intermediate risk factor for SUD, measurements. although conduct and bipolar disorder seem to heighten the risk of early onset of SUD. Both family-genetic and It is also important to note that statistically significant self medication influences seem to be operational in the differences in height and weight parameters are not development and continuation of SUD in ADHD necessarily clinically significant differences. subjects.’’ [The Expert Panel noted that in general the Overall Assessment of Height and Weight Data: studies examining substance use disorders are compli- Although the observations in this area are consistent with cated by the well known association in pedigrees of most clinical experience, the quality of data in the older substance use disorders, ADHD, and other psychiatric articles is suboptimal. These articles have variable but disorders and by the studies’ varying sophistication in generally marginal-to-moderate utility, with incomplete measuring true substance use disorder compared to documentation of compliance or actual dosing regimens simple experimentation or initiation. A weakness of all and with failure to consider (in most cases) basic factors the studies is the use of self-report only to measure that are usually assessed in growth studies, such as mid- substance use without any confirmation by biologic parent height and parent BMI, family history of timing of markers such as urine or hair, which might enhance puberty onset, the child’s actual physical or endocrino- accurate identification of users.] logic level of puberty at start of treatment (some of the 3.1.2.4 Effects on height and weight. youngsters were as old as 15 when the studies were Aarskog et al. (1977) examined the acute effects of d- conducted), and measurement of skeletal maturity (bone amphetamine treatment on growth hormone levels in age), which particularly in school-aged children is children before and after 6–8 months of treatment with 5– considered a useful indication of expected growth 35 mg/day methylphenidate. Mean 7SEM serum potential. The seasonal differences in expected growth growth hormone levels in 7 children given 15 mg d- (in the northern hemisphere, children grow faster in amphetamine were 3.170.9 ng/mL at baseline and summer) are not accounted for by designs that compare 14.175.0 ng/mL at peak before methylphenidate treat- children whose families chose to leave them on stimu- ment and 8.671.5 and 10.073.6 ng/mL after methylphe- lants through the summer and children whose families nidate treatment. The mean time for achieving maximum did not leave them on medication during the summer. It growth hormone levels after d-amphetamine administra- cannot be ruled out that those who remained on the tion was 60 min before methylphenidate treatment and medicines also had other conditions or behavioral 90 min after treatment. patterns (like fetal alcohol effects) that motivated their Studies on growth in children treated with ampheta- parents to continue the medication and might also mines are summarized in Table 33. The reviewed studies decrease growth. It is not known whether final adult investigated amphetamine alone, compared ampheta- height and weight are affected by current treatment mine to a control group, compared amphetamine and regimens, which frequently include continuous use and methylphenidate, or compared amphetamine to another use beyond childhood. medication. Some of these studies lacked information on In addition, assessments of growth do not seem to be the magnitude and timing of effects, thereby contribut- masked to stimulant exposure history. For example, in ing to inconsistent results. There are several methodolo- the reports of Safer et al. (1972, 1975) and Safer and Allen gic problems and limitations in all of the studies making (1973), the nurse who obtained the measurements was it difficult to compare results among studies and not masked to the children’s drug histories, and in many determine definitive conclusions. Some of the limitations cases actually administered the drugs herself. The and problems include: studies did not control for potential confounders such as intrauterine exposure to tobacco, ethanol, and illicit * Imprecise height measurements with measurement drugs, or parental mental health. errors comparable in magnitude to the effect being Findings overall seem to suggest that appetite and studied. growth suppression are less with methylphenidate * Use of cross-sectional rather than longitudinal analy- than with amphetamines, but these findings are not sis of growth data. conclusive. There are interesting and clinically relevant * Height deficits in ADHD children could be related issues of mechanism that have not been fully elucidated. to genetic influences, i.e., parental stature (no It is unclear whether the growth alterations that are studies take into account the contribution of parental noted are primarily related to appetite suppression height). (as might be expected given the widespread use of * Comorbid conditions that may influence growth. amphetamines by dieters) or by endocrine alterations as * Use of old reference standards inappropriate to age well. If the issue is only appetite suppression, it is contemporaneous standards. possible to test a number of useful clinical interventions, * Imprecise or absent description of inclusion criteria of such as feeding the child a high-calorie supplement children both on and off medication. before the first daily dose and monitoring whether * Inclusion of children from different socioeconomic this intervention alters the patterns of growth. The groups. possible role of stimulant-associated endocrine changes * Inability to document compliance with medication. cannot be addressed with the current data set because * Use of different age groups within the same study the endocrinologic data are outdated and use compar- including pubertal ages, which could significantly ison drugs that increase the release of prolactin, creating influence both height and weight. a possible artifact of lower hormone levels with * Absence of control groups of unmedicated children. stimulants.

Birth Defects Research (Part B) 74:471–584, 2005 DOI 10.1002/bdrb it eet eerh(atB 44154 2005 74:471–584, B) (Part Research Defects Birth Table 33 Studies on Growth in Children Treated with Amphetamines Parametersa Results and author conclusions Comments Reference

n 5 29 (sex unspecified) Methylphenidate and d-amphetamine combined results: Strength: Control group Safer et al., 1972 Control: Unmedicated ADHD controls -The group that discontinued medication after 9 months (n 5 13) gained twice Weaknesses: First data set: (n 5 7); Anthropometric scale (His- as much weight in the summer as the group staying on medication (n 5 7). Most were in special classes for torical control) However, this weight rebound was not sufficient to compensate for the initial learning and behavior problems; Age: Elementary school-aged weight suppression. measurements only 3 times in 9 Z Duration: Group 1 9 months; Group -Over 2 years, medicated group (n 5 9) had a percentile weight change of months; 13/20 taken off medication Z 2 2 years À20.38, compared to 16.79 in the control group (n 5 7). For height, the during summer, making exposures Dose: 10 or 15 mg/day medicated group percentile change was À13.45 compared to 11.29 in non-uniform. Second data set: Data controls. obtained retrospectively from school -Tolerance did not develop for weight-gain suppression in the 2-year group; nurse records; medication informa- METHAMPHETAMINE AND AMPHETAMINE these children had a mean weight gain of 1.8 kg, compared to the expected tion obtained from parent or nurse 3.1 kg. reports; different medication doses O 10.1002/bdrb DOI -Percentile height decrease correlated with percentile weight decrease, but for children; small number for sub- was not significant compared to baseline. jects (9) and controls (7); use of percentiles. d-Amphetamine results: It is difficult to draw conclusions -Significant difference in weight gain between group discontinuing medica- from these data. tion in the summer (0.47 kg/mo) and group continuing medication (0.14 kg/mo). -Suppressive effect on weight gain was significant and not dose dependent. n 5 29 (sex unspecified for d-amphe- Strength: Use of a control group Safer and Allen, tamine group, however 44 of 49 in Weakness: Evaluation performed once/ 1973 study were male) year; growth was evaluated by total Control: 14 non-medicated ADHD change in the normative percentiles males for weight and height. Age (years): 6.7 X 5 Duration 2 years (mean 2.9) -Summer continuance not significant for weight; marginally significant for Dose: unspecified height (t 5 1.20, n 5 28, Po0.2). -Significant decrease in average percentile weight loss each year indicates development of tolerance for weight suppression. -Most important factors were duration of treatment and frequency of medication.

n 5 34 (sex unspecified; 4:1 male: Strength: 3-year study period Safer et al., 1975 female ratio for study as a whole) Weaknesses: Measurements twice/year; Control: Anthropometric scale (histor- measurement with a yardstick is ical control) unlikely to be sensitive to 0.01 cm, as Age (years): 10.3 (8–13) the data are expressed; potential DurationZ3 years inclusion of puberty in age group; use Dose: 12 (5–20) mg/day of historical control. General comment: Results suggest growth rebound. 521 2 OU TAL. ET GOLUB 522 Table 33 (Continued)

Parametersa Results and author conclusions Comments Reference

-Growth rate differences are statistically significant regardless of dose, which indicates rebound for both height and weight. n 5 24 (17 male) Strength: Longer treatment period (at Gross, 1976 Control: Pre-treatment percentiles least 2 years). Age (years): 9.0 (6–12.4) Weaknesses: Data partly retrospective; Duration: 5.5 (2–9) years measurements converted to percen- Dose: 16.5 (5–32.5) mg/day tiles for comparison. Comments: The utility of this study is in showing that height and weight deficits are compensated in long-term treatment.

-Data shows some weight loss in the first 3 years compared to expected norms (statistically significant only in the first year), which recovered in later years. Both height and weight exhibited a statistically significant percentile increase at final follow-up. Patients who had discontinued medication at final follow-up showed a ‘‘larger’’ height (not statistically significant) and weight (Po.06) rebound than those still on medication. it eet eerh(atB 44154 2005 74:471–584, B) (Part Research Defects Birth n 5 7 (all male) on d-amphetamine Strengths: Prospective study; no prior Greenhill et al., Control: 8 boys on phenothiazines stimulant treatment; compliance was 1977; Puig- Age (years): 6–9.5 monitored by prescription renewals; Antich et al., Duration: 1 year height velocity was evaluated; mea- 1978 Mean dose: 0.88 mg/kg bw/day surements every 3 months. Weaknesses: Small numbers of subjects; control group exposure to phenothiazines; study period of only 1 year.

-At 1 year, significant decline in weight percentile (Po0.02). Also a significant decline in height percentile (Po0.02) and drop in growth velocity (Po0.05). - Noted that the control, phenothiazine, may promote growth in children, which could have increased the apparent growth deficit for the d-amphetamine group. n 5 13 (all male) -After 1 year, height and weight both decreased significantly. Weight Strengths: Prospective design; findings Greenhill et al., Control: Pre-treatment percentiles decreased by 16 percentile points and height decreased by 10 percentile consistent with previous studies; rig- 1981 O 10.1002/bdrb DOI Age (years): 6–9.5 points. Height and weight velocities both decreased significantly. The trend orous growth-measurement techni- Duration: 12–21 months also occurred in patients on medication for 21 months, with weight ques with measurements performed Dose: 21 (10–30) mg/day decreasing by 12.1 percentile points and height decreasing by 14.8. every 3 months (agreement to within -Dose negatively correlated with height and weight velocity. 1 mm for height and 0.1 kg for -Weight deficits were greater than height deficits. weight); used weight adjusted for -Results did not suggest rebound for height or weight. height, height and weight velocities, it eet eerh(atB 44154 2005 74:471–584, B) (Part Research Defects Birth

-Noted that growth deficits were not severe, while behavioral benefits were and comparisons with expected substantial. velocities. Weaknesses: Small number of subjects; control exposure to phenothiazines. n 5 28 (25 male) Height and weight measurements represented as z-scores: Strengths: No prior drug treatment; used Schertz et al., Age (years): 9.3 (4.0–15.3) Initial weight 0.5 weight-for-height curves (BMI). 1996 DurationX5 months; mean duration of Initial height 0.1 Weaknesses: Retrospective study; data follow-up 5 10.8 months Weight change at follow-up À0.6 on weekend treatment only partially Mean dose: 14.9 mg/day (weight- Height change at follow-up À0.4 available; BMI curves were from adjusted dose 5 0.5 mg/kg bw/day) -Children were divided into 2 groups to analyze effects of pre-treatment a Caucasian sample and not weight; 84% of the heavy group (n 5 19) experienced decreased BMI from generalizable to non-white children. expected compared to 56% of the thinner group (n 5 9). -BMI slope analysis showed significant difference in growth deficit between MHTMN N METHAMPHETAMINE AND AMPHETAMINE heavy and thin groups. -Major predictor of decreased BMI was pre-treatment weight. No significant effect of dose, duration of follow-up, or age on degree of weight loss. O 10.1002/bdrb DOI -Retrospective study; did not account for drug holidays. n 5 124 (all male) -Of the 124 ADHD children, 110 were medicated at some time in their lives Strengths: Normal controls; used z Spencer et al., Control: 109 normal male controls with either methylphenidate, d-amphetamine, or pemoline. At the time of scores; corrected for parental heights; 1996 Age (years): 14.5 (6À17) the study, 53 had been treated in the preceding 2 years with a mean performed pubertal assessments. Dose: methylphenidate equivalent methylphenidate equivalent dose of 38 mg/day. Weaknesses: Measurements only at dose (twice the d-amphetamine and -ADHD children were 4 kg lighter and 3 cm shorter than controls, neither of 4-year follow-up; missing data. half the pemoline dose) 5 38 (5–120) which was statistically significant. However, height deficit was statistically mg/day significant (P 5 0.03) when height was converted to a z score. Significantly more ADHD children were at least 2 SDs shorter than the mean (P 5 0.02) when corrected for age and parental heights. -No significant difference was found for height or weight between unmedicated and stimulant-treated children with ADHD. -Modest height deficits were unrelated to weight deficits or stimulant treatment, and only evident in ADHD children in early adolescence. This result indicates that ADHD children may experience delayed growth, rather than permanent stunting of growth. -No evidence of weight suppression or delayed pubertal development. -No association between height and drug treatment, drug class, duration of treatment, or dose regimen was identified. n 5 68 (all male) -Compared 23 boys on methylphenidate to 68 boys on a racemic mixture of l- Strengths: Large number of subjects Sund and Control: Norwegian population sample and d-amphetamine. followed for an extended time period; Zeiner, 2002 Age (years): 3–13 -No statistically significant difference between methylphenidate- and am- results and conclusions are consistent Duration: 1–5 years phetamine-treated children in height or weight except for a lower weight with previous studies. Dose: Range 2.5–32.5 mg/day; mean gain in amphetamine-treated children during the first year (Po0.05). Weaknesses: Large age range of subjects; dose (mg/day): -21 boys (31%) on amphetamine and 4 on methylphenidate (17%) either lost or inclusion of pubertal subjects; once- did not gain weight during the first year, with losses ranging from 0 to yearly height and weight data, col- 9.5 kg. There was no significant dose difference between those who lost lected retrospectively; included de- weight in the first year and those who did not. All boys who had lost velopmentally-delayed children weight in the first year experienced subsequent sufficient weight gains. without reference to whether they -Children above the 50th percentile in weight prior to treatment had had a growth-retarding syndrome. significantly increased weight loss compared to those below the 50th General comments: Standard Norwe-

percentile (Po0.05), suggesting a slimming effect. gian population sample comparison. 523 2 OU TAL. ET GOLUB 524 Table 33 (Continued)

Parametersa Results and author conclusions Comments Reference

-No effect on height was observed. Multiple regression showed that -No effect of cumulative dose or age. neither cumulative dose nor age had -Concluded that methylphenidate and amphetamine do not have adverse a significant effect on growth when growth effects for most children. initial weight and height were -Retrospective study, some missing data, broad age range, measurement controlled. reliability uncertain. n 5 51 (44 male) Data reported as mean7SD of Standard Deviation Score (a z score). Strengths: Rigorous measurements, Poulton and Control: National Centre for Health -Initial height 5 0.2471.05; initial weight 5 0.6171.13. obtained every 6 months; height, Cowell, 2003 Statistics -After 6 months, weight was 1.7 kg less than expected; 76% lost weight. After weight, and height velocity corrected Age (years): 7.2 (3.1–11.4) 30 months there was a 3.0-kg deficit. for age and sex using SD scores. Duration: 6–42 months -During the first 2 years, the height deficit was approximately 1 cm/year; the Weaknesses: Retrospective study; Dose: 0.570.13 (mean7SD) mg/kg average height deficit after 42 months was 2.4 cm. no control group; no long-term bw/day; 12.5 (7.5–25) mg/day -In the first 6 months, 86% had a decreased height velocity. After 30 follow-up. months, the deficit was attenuated and most children had normal General comments: Height velocity was height velocity. lowest during the first 6 months, but -31% experienced weight deficit even without reported appetite in most cases normalized after 3 suppression. years; results consistent with other -Average weight deficit was 2.4 times the height deficit after 30 months. findings. -Significant decrease in height and weight after 6 and 18 months (Po0.001) and after 30 months (Po0.01).

it eet eerh(atB 44154 2005 74:471–584, B) (Part Research Defects Birth -Retrospective study; results do not separate methylphenidate (n 5 19) from d-amphetamine (n 5 32) treatment; 10 patients were also on clonidine; did not account for drug holidays. aData presented as mean (range) unless otherwise specified. O 10.1002/bdrb DOI AMPHETAMINE AND METHAMPHETAMINE 525 3.2 Experimental Animal Data for the rat study, the Expert Panel assumes a 3:1 ratio of 3.2.1 Prenatal toxicity endpoints. Studies exam- d-,l-amphetamine and a single dose of the test article on ining endpoints in prenatally-exposed animals or models GD 19.] Does were killed on GD 29 and uterine contents of prenatal development are discussed in this section. examined. Fetuses were evaluated for external, visceral, The section is divided into studies conducted in and skeletal abnormalities. mammalian species in vivo, mammalian species in vitro, Clinical signs were noted in the high-dose group. and in chickens. Amphetamine studies are presented There were no treatment-related effects on weight gain before methamphetamine studies. Oral exposure studies or feed consumption. There were no effects on litter are presented before parenteral exposure studies. parameters including implantations, resorptions, live 3.2.1.1 In vivo mammalian studies. young, sex ratio, and pre-implantation loss. Post-im- The FDA Pharmacology Review for Adderall (FDA, plantation loss was described as slightly increased in the 2001) summarized a developmental toxicology study in high-dose group; statistical evaluation was not provided. Sprague–Dawley rats, submitted to the agency as part of The post-implantation loss rate was 17.2% in the high- the approval process. Pregnant rats (22/dose group) dose group compared to 12.0% in the control group [SD were gavaged with amphetamine free base in water at or SEM not given]. There were no alterations in total daily dose levels of 0, 2, 6, or 20 mg/kg bw/day in placental, fetal, or litter weights. There was no increase two doses divided equally 8 hr apart [the Panel assumes in malformations in any dose group. There was reported that the test article was d- and l-amphetamine in a 3:1 to be an increase in fetuses with 13 ribs (a variation in ratio, as in the marketed product]. Treatments were this species), which appeared only on a per fetus given on GD 6–17 [blood for pharmacokinetic evalua- evaluation (at least one fetus was affected in nearly all tion was taken at intervals for 24 hr after a treatment on litters in all dose groups). Litter incidence of 13th ribs GD 17; the Panel assumes that only a single treatment was high in all dose groups with 100% incidence in was given on this day]. Dams were killed on GD 20 and control litters. [Statistical analysis was not provided. uterine contents evaluated. Fetuses were given external Evaluation by CERHR does not show a significant v2 examinations. Half the fetuses were sectioned for difference between groups in fetuses with 13 ribs by evaluation of visceral abnormalities and the other half testing; group percentages of fetuses with 13 ribs were processed for skeletal evaluation. ranged from 57–69%. The Expert Panel notes that litter Dose-related clinical signs occurred in the middle- and analysis is preferred; however, data given in the FDA high-dose groups, with a low incidence of clinical signs review do not permit such an analysis.] in the low-dose group. The severity of clinical signs in The FDA reviewer concluded that there was no clear evidence of developmental toxicity but noted that based the high-dose group led to termination of this group on on pharmacokinetic data (presented in Section 2), GD 9. Dam body weight gain was decreased by 5 and exposures in these experimental animal studies were 13% in the low- and middle-dose groups over the dosing lower than exposures in children (based on AUC) or only period. There were no treatment-related effects on litter 2–3 times higher (based on C ). The FDA reviewer also parameters (implantations, resorptions, live young, pre- max noted that although congenital malformations were not or post-implantation loss, or sex ratio) or on placental, increased in these studies, other studies in the literature fetus, or litter weights. There was no increase in identified neurobehavioral abnormalities associated with malformations or in variations except for a possible developmental exposures of rodents to amphetamines. increase in delayed cranial ossification sites. When [The FDA reviewer specifically cited three references analyzed by number of litters containing at least one (Pu and Vorhees, 1993; Vorhees et al., 1994b; Acuff- fetus with cranial ossification delay, there were no Smith et al., 1996); the Expert Panel reviewed these and differences between groups. There was an increase in other studies, discussed in Section 3.2.3.] the number of litters in which there were three or more Strengths/Weaknesses: Several strengths were noted, fetuses with delayed cranial ossification, one, four, and including the standard design under GLP using multiple six of 22 litters in the control, low-dose, and mid-dose doses administered orally. The dosing paradigm (two 1 groups, respectively. [Benchmark dose calculations doses divided equally 8 hr apart) was designed to better carried out by CERHR for this endpoint gave a mimic clinical exposure patterns. Blood samples were BMD10 of 3 mg/kg bw/day and a BMDL of 2 mg/kg collected for toxicokinetic evaluations. Group sizes were bw/day.] adequate, appropriate controls were used, and the study The same review (FDA, 2001) included a summary of a included external, visceral, and skeletal examinations of developmental toxicity study performed using pregnant term fetuses. A weakness is that dosing was not New Zealand White rabbits (22/dose group) given continued into the latter part of pregnancy, which is a divided amphetamine gavage doses totaling 0, 2, 6, or more recently introduced regulatory design. Statistical 16 mg/kg bw/day on GD 6–19. Periodic blood samples analyses were not described nor were statistical results were taken on GD 19 for pharmacokinetic studies. [As presented. It is not clear that the Staples technique for evaluating fetal morphology would have detected 1 The BMD10 is the benchmark dose associated with a 10% effect, cardiovascular problems. In rats, the high dose exceeded estimated from a curve fit to the experimental data. The BMDL represents the maximum tolerated dose (MTD), so animals in this the dose associated with the lower 95% confidence interval around this estimate. Benchmark doses are used commonly in a regulatory setting; group had to be removed from the study on GD 9. Only however, they are used in this report when the underlying data permit two dose groups were available for evaluation. In rabbits, their calculation, and are only supplied to provide one kind of description the significance of increased fetal incidence of 13th ribs of the dose–response relationship in the underlying study. Calculation of is difficult to determine given the high litter incidence a benchmark dose in this report does not mean that regulation based on the underlying data is recommended, or even that the underlying data are (100%) of 13th ribs in control animals. As stated in the suitable for regulatory decision-making. report summary, the safety margins provided by this

Birth Defects Research (Part B) 74:471–584, 2005 DOI 10.1002/bdrb 526 GOLUB ET AL. study were minimal-to-nonexistent with mean total Toxicokinetics were not determined. Apparently, the exposures lower than those seen clinically, and peak control mice were not gavaged. Statistical methods were levels approximately two-fold higher in rats and three- not identified. There were 20 litters in the control group fold higher in rabbits than levels measured clinically. and 47 litters in the amphetamine-treated group, yet only Utility (Adequacy) for CERHR Evaluation Process: half of these litters (10 and 24) were evaluated for total This study is potentially useful for the evaluation implants, live/dead fetuses, fetal body weights, and process. The panel was not able to use the study because malformations. a complete report was not available. Utility (Adequacy) for CERHR Evaluation Process: Summaries of unpublished studies on Adderall were This study is not very useful for the evaluation process. provided to the Expert Panel by Shire Pharmaceuticals Ramirez et al. (1979) examined the effects of prenatal Group (Secker, 2004). A pre- and postnatal GLP study, amphetamine exposure on estrous cyclicity, sexual performed in accordance with ICH guidelines (similar behavior, and hypothalamic monoamine levels in rats to ICH 4.1.2), used gavage administration to Sprague– [strain and number treated not specified]. Animals Dawley rats from GD 6–Lactation Day 20 (mating 5 GD were s.c. injected with saline or 0.5 mg/kg bw/day d, 0). Dose levels of amphetamine base equivalents were 0, l-amphetamine [purity not specified] during the entire 2, 6, and 10 mg/kg bw/day, n 5 25/group. Clinical signs gestation period. Litters were culled to four females and and a significant decrease in gestational feed consump- four males at birth. At 3 months of age, estrous cycles tion and body weight gain during specific periods were were monitored by taking vaginal smears on 5 days/ seen in all amphetamine-treated groups. Postnatal week for at least 4 cycles in 30 control and 20 treated survival to PND 4 was reduced at 6 and 10 mg/kg bw/ offspring. [The number of litters from which the day and postnatal survival to weaning was decreased at offspring were obtained was not specified.] On the 10 mg/kg bw/day. Offspring preputial separation and last day of estrus, the offspring were ovariectomized and vaginal opening were delayed in the 6 and 10 mg/kg ova in tubes were counted. Three weeks later, sexual bw/day groups. There were no apparent alterations in behavior in response to a male rat was monitored after F1 mating success or pregnancy length. Dams from the priming with various doses of estradiol (25–200 mg/kg) high dose group had a decrease in implantations, pups/ or estradiol (10 mg/kg) in combination with progesterone litter, and live pups/litter. [These summaries were noted (2 mg/kg bw) (n 5 10–24/group). Four weeks after the but were not used by the Expert Panel in the Evaluation sexual behavior studies, rats were primed with estradiol Process]. (100 mg) or estradiol (10 mg) in combination with pro- Yasuda et al. (1965), in a study funded by NIH, gesterone (2 mg) and killed for measurement of examined maternal and fetal toxicity in ICR-JCL mice hypothalamic dopamine, noradrenaline, serotonin, and treated with amphetamine sulfate [enantiomers and 5-hydroxyindolacetic acid levels (n 5 6–9/group). Ovula- purity not specified]. Beginning on GD 0 (day plug tion and cyclicity data were analyzed by Fisher exact detected) and continuing through pregnancy, dams were probability test, and catecholamine and behavioral data gavaged with 50 mg/kg bw/day amphetamine sulfate or were evaluated by Student t-test. with distilled water. [It is not clear if control dams Prenatal d,l-amphetamine treatment had no effect on were treated and examined concurrently with the cyclicity or ovulation. Sexual receptivity, as measured dosed animals.] The dams were killed on GD 18.5 for by lordosis response, was significantly higher in the examination of fetotoxicity. [Methods of statistical amphetamine group primed with estradiol and proges- evaluation were not discussed and it does not seem terone. In the estradiol dose–response experiment, sexual that the litter was considered in those evaluations.] receptivity in the amphetamine group was significantly Amphetamine-treated dams showed excitement for higher than controls at estradiol doses of 100 and 200 mg/ 6–7 hr after dosing. Four of 69 treated dams died during kg. The only significant effect on monoamines was a the study. Maternal body weight of treated dams was reduction in medial hypothalamic serotonin levels in the significantly lower than controls after GD 16. Feed and amphetamine group [67% of control levels] primed with water intake were reduced only during the first 2 days of estradiol and progesterone, but not estradiol alone. The dosing. There were no differences in implantation rates study authors concluded that prenatal exposure to between amphetamine-treated and control animals. amphetamines has long-lasting effects on sexual beha- Litters from 24 treated and 10 control dams were vior in female rats. available for an examination of total implantations, Strengths/Weaknesses: The evaluation of unique and resorptions, fetal mortality, weight, and external mal- important endpoints is a strength. Pregnant rats were formations. The only statistically significant effect was dosed with amphetamine from the beginning of gesta- an increase in dead fetuses (17.5% in treated vs. 9.5% in tion until term and aspects of reproduction were control). There were no significant effects on fetal body examined in female offspring, including ovulatory weight or gross malformations. function and sexual receptivity. The authors tried to Strengths/Weaknesses: The use of a high oral dose correlate behavioral changes with hypothalamic mono- with adequate group sizes is a strength. Initiation of amine levels. A weakness is the use of a single dose level dosing the day after mating demonstrated that 50 mg/kg given by s.c. injection, whereas humans typically are bw/day amphetamine did not affect implantation. exposed by oral or i.v. routes. With only one dose level, Several weaknesses were noted, such as the assessment dose–response relationships cannot be evaluated. Par- only of gross malformations. In addition, this study used ental animals were selected based on their performance only a single dose level, so dose–response comparisons in an active avoidance test, and the potential influence of were not possible. Pregnant mice were given a single such selection on study outcome is unclear. It is also a daily gavage dose of amphetamine sulfate, not two doses weakness that housing of adult females singly or as a a day as is more relevant to clinical use patterns. group is not specified. There is no information on the

Birth Defects Research (Part B) 74:471–584, 2005 DOI 10.1002/bdrb AMPHETAMINE AND METHAMPHETAMINE 527 number of exposed dams and the number of litters from primarily the neurobehavioral effects of prenatal am- which the offspring originated; consequently, there is no phetamine exposure (not developmental toxicity); thus, indication as to whether the authors controlled for litter not all intermediate dose levels were included for fetal effects. There is insufficient experimental detail to track analyses. animals throughout this study (e.g., Table 1 of the study Utility (Adequacy) for CERHR Evaluation Process: references 30 controls and 20 in utero amphetamine- This study is adequate for the evaluation process. treated animals, which were subsequently ovariecto- Nora et al. (1965, 1968), in two studies funded by NIH, mized for the sexual receptivity experiments. Figure 1 of the National Research Council of Canada, or the March the study [sexual receptivity with estradiol and proges- of Dimes, examined teratogenicity in A/Jax mice and terone] references 18 controls and 24 in utero ampheta- C57Bl/6J mice dosed with d-amphetamine sulfate [purity mine-treated animals). Furthermore, the authors’ not specified].OnGD8[plug day not specified] statistical methods were inadequate in some cases (i.e., the mice received a single i.p. dose of 50 mg/kg bw using a Student t-test to compare sexual receptivity d-amphetamine sulfate in water, a dose that was stated to across different doses of estradiol as shown in Fig. 2). be 200 times higher than the usual human dose. Controls Utility (Adequacy) for CERHR Evaluation Process: were i.p. injected with physiological saline. Fetuses were This study is of limited usefulness for the evaluation removed 1–2 days before term, fixed in Bouin solution, process. and the thorax was sectioned and examined. [There was Adams et al. (1982) at the National Center for no mention of methods used for statistical analyses and Toxicological Research (NCTR) gave d-amphetamine the litter was not considered in statistical evaluations.] sulfate 0 or 2.0 mg/kg bw/day s.c. to Sprague–Dawley In the first study (Nora et al., 1965) conducted in A/Jax rats on GD 12–15 (n 5 9/dose group; plug 5 GD 0). mice, there were 21 treated and 10 control dams. There Dosing solution concentrations were verified by GC. were 130 living fetuses in the treated group and 70 in the Dams were killed on GD 20 and uterine contents control group. The findings reported to affect more evaluated. Live fetuses were decapitated and heads treated compared to control fetuses were (percent fixed in Bouin solution for free-hand razor-blade incidence in treated compared to controls) resorptions sectioning. Visceral evaluation was performed by dissec- (29 vs. 8%), malformations (38 vs. 7%), cardiac mal- tion under a stereomicroscope. Carcasses were cleared formations (12 vs. 0%), cleft lip (18 vs. 4%), and eye and stained for skeletal examination with Alizarin Red S. abnormalities (8 vs. 3%). Cardiac malformations included A companion behavior study (summarized in Section defects of ventricular and atrial septum, defects of the 3.2.3) contributed culled pups on PND 1 for teratologic right aortic arch, and coarctation of the aorta and right evaluation by the same methods. There was a 32% aortic arch. [Fisher exact test performed by CERHR decrease in body weight gain during the dosing period, on the fetal data shows statistical significance for but the difference was not statistically significant; total malformations (Po0.001), cardiac malformations differences between the amphetamine and control (P 5 0.0015), and cleft lip (P 5 0.0074). The Expert Panel groups in dam weight gain, implantation sites/litter, live notes that litter analysis is preferred; however, data fetuses/litter, resorptions, sex ratio, and fetal weight given in the study do not permit such an analysis.] were also not statistically significant. There were no In the second study, there were 34 treated and 32 malformations in any of the amphetamine-exposed control A/Jax dams and 26 treated and 27 control C57Bl/ fetuses or pups. The authors recognized that a larger 6J dams (Nora et al., 1968). For the A/Jax mice, there number of litters may have shown a low malformation were 283 living treated fetuses and 233 living control rate, but they concluded that the 2.0 mg/kg bw/day fetuses. There were 95 resorbed embryos in the treated amphetamine dose at mid-gestation had a low likelihood group and 26 resorbed embryos in the control group. of producing this kind of developmental toxicity. Cleft palate was the most frequently occurring mal- Strengths/Weaknesses: Strengths include the rela- formation in the A/Jax strain [incidence not stated]. tively low dose and the good postnatal data sets. Findings that were significantly increased in fetuses from Concentration verification was conducted on dose solu- the treated versus control group were (percent incidence tions using GC and the groups were counterbalanced in treated compared to controls) total malformations [22 during dosing. The high-dose group (2.0 mg/kg bw/day compared to 8.6%] and cardiac anomalies (13% com- on GD 12–15) was evaluated for effects on uterine pared to 1.3%. [The percentage should be 8% if the weight, live/dead fetuses, resorption sites, and fetal authors’ figures are correctly printed (24/283); however, structures (external, visceral, and skeletal examinations). it is possible that the denominator should have been The major weakness is the small sample size (n 5 9/ printed 183, which would give 13%]). The most common group) leading to reduced sensitivity. In addition, only cardiac anomalies in treated and control A/Jax fetuses one dose group was available, so that dose–response were atrial septal defects. For the C57Bl/6 mice, there assessments could not be made. Malformations were were 161 living treated and 202 living control fetuses. examined in the culled pups from litters used in the The number of resorbed fetuses was 63 in the treated postnatal study (examined on the day of birth), reducing group and 19 in the control group. Microphthalmia was the likelihood that a low incidence of malformations was the most frequently occurring malformation in C57Bl/6J missed. For neurobehavioral assessments, nested litter fetus [incidence not stated]. Findings that were signifi- analyses show more sophisticated statistical analyses. cantly increased in C57Bl/6J fetuses for the treated The weakness of this study is the summarizing of the compared to the control group were (percent incidence teratology data rather than more detailed reporting. In in treated compared to controls) total malformations [32 addition, rats were dosed with amphetamine by s.c. compared to 7.9%] and cardiac anomalies (11 compared injection, whereas humans typically are exposed by oral to 1%). The most common cardiac anomalies in treated or i.v. routes. The objective of this study was to examine and control C57Bl/6J fetuses were ventricular septal

Birth Defects Research (Part B) 74:471–584, 2005 DOI 10.1002/bdrb 528 GOLUB ET AL. defects. The study authors concluded that amphetamine injected on 1 or 2 days between GD 9 and 11. The saline treatment increased the frequency of spontaneously controls were injected on GD 9 or 10 or 11.] Three to 15 occurring cardiac malformations in each strain of mouse. dams/group treated during GD 9–11 were killed on GD Strengths/Weaknesses: The use of a high dose is a 15 and electrocardiograms (EKGs) were recorded in 17– strength in increasing sensitivity, but the use of multiple 39 fetuses/group. Other surviving dams (n 5 5–19/ doses would have been superior. The use of two strains is group) were killed on GD 19. Implantation and resorp- a strength. In addition, fetal examiners were blind with tion sites and fetal survival and external malformations respect to treatment group in the original study (Nora were examined. EKG recordings were obtained from et al., 1965). These studies evaluate a reasonable margin fetuses without malformations from the GD 9–11 group of safety at 50 mg/kg bw d-amphetamine, which is (n 5 5–15 dams/group and 40–71 fetuses/group). After approximately 200 times the usual clinical dose. Several the EKG recordings on GD 19, fetuses were weighed, weaknesses were noted. Pregnant mice were dosed with killed, and examined for internal malformations [method d-amphetamine at a single dose level by i.p. injection, not specified]. Hearts were fixed in 4% formaldehyde, which is not considered a relevant route of exposure due sectioned, and stained with hematoxylin and eosin. Data to (1) possible direct adverse effects of the drug or were evaluated by Student t-test. solvent on the uterus and secondarily on embryo-fetal The discussion of results is limited to mice treated on development, and (2) the possibility that the drug might GD 9, 10, and 11. As noted in Table 34, amphetamine be physically transported directly to the embryo/fetus, treatment reduced fetal weight and increased maternal bypassing maternal absorption, metabolism and distri- and embryonic mortality and fetal malformations. The bution. With only one dose level, dose–response relation- types of malformations observed were microphthalmia, ships cannot be evaluated. The dosing period was amelia, exencephaly, and cleft lip. There were no internal unusually long (10 AM to 4 PM to inject 21 mice), which malformations observed. No differences in EKG patterns could result in confounding due to diurnal variations. were observed on GD 15. On GD 19, mean Q–T intervals Furthermore, there is insufficient experimental detail to in the 50 mg/kg bw group (445 msec) and 100 mg/kg bw interpret these studies. For example, there was no group (467 msec) were significantly longer than the mean information given on randomization method, maternal Q–T interval in the control group (215 msec). EKG and fetal body weight during the study, or litter data patterns in GD 19 treated embryos were similar to those (number of litters/group, average litter size, litter observed in control embryos around GD 14–16. Histo- weights, etc.). Although there were only two additional differentiation of myocardial tissue was delayed in GD dams in the d-amphetamine-treated versus control A/J 19 embryos with prolonged Q–T intervals. The study mice (34 treated females vs. 32 controls) in the 1968 study authors concluded that high doses of d-amphetamine (Nora et al., 1968), there were 50 more live fetuses and 69 retard general embryonic development and histodiffer- more resorbed fetuses than were seen in the control entiation of cardiac tissue. group, which may be further evidence that the total Strengths/Weaknesses: The unique focus on cardio- number of fetuses should have been 183 rather than 283, vascular effects and the evaluation of sensitive periods as printed. There was no mention of the statistical are strengths of this report. EKG and histology were used methods employed and litter-based analyses were not to demonstrate a delay in development of the heart on conducted. GD 19 in mouse fetuses exposed to 50 or 100 mg/kg bw/ Utility (Adequacy) for CERHR Evaluation Process: day d-amphetamine on GD 9, 10, and 11. Prolonged Q–T These two reports are of limited usefulness for the intervals and undifferentiated cardiac myocytes sup- evaluation process. ported the conclusion of incomplete maturation of Fein et al. (1987), in a study conducted at an Israeli cardiac muscles at term. Although not specifically stated Medical Center, examined teratogenicity and cardiac in the text, the tables list n values as the ‘‘number of effects of d-amphetamine in ICR mice. During GD 9–11 dams/litters evaluated,’’ suggesting that there was some (plug day 5 GD 0), mice were i.p. injected with 50 or control for litter effect. Adequate sample sizes were used 100 mg/kg bw/day d-amphetamine sulfate [purity not during most segments of this study. Several weaknesses specified]. Controls were i.p. injected with saline. [It is were noted. Pregnant mice were dosed by i.p. injection, clear that one group of mice was injected with which is not considered a relevant route of exposure due amphetamine on GD 9, 10, and 11. Other groups were to (1) possible direct adverse effects of the drug or

Table 34 Mortality and Malformations in Offspring of Mice Treated with d-Amphetamine Sulfate on GD 9–11

Dose (mg/kg bw/day)

Endpoint 0 (GD 9, 10, or 11) 50 (GD 9–11) 100 (GD 9–11)

Surviving dams, n (%) 10 (100) 5a (62) 15a (60) Live fetuses, n (%) 96 (95) 44 (88) 60a (42) Resorption sites, n (%) 6 (5) 6 (12) 84a (58) Malformed embryos, n (%) 0 4a (9) 9a (15) Fetal weight, g (mean7SD) 1.325670.084 (n 5 102) Not specified 1.152970.104a (n 5 51) From Fein et al. (1987). aPo0.01 compared to control (original states t-test, but the Expert Panel assumes that a test of proportions was used for comparisons other than fetal weight).

Birth Defects Research (Part B) 74:471–584, 2005 DOI 10.1002/bdrb AMPHETAMINE AND METHAMPHETAMINE 529 solvent on the uterus and secondarily on embryo-fetal Table 35 development and (2) the possibility that the drug might Benchmark Doses Calculated from the Mouse Study of be physically transported directly to the embryo/fetus, Kasirsky and Tansy (1971) bypassing maternal absorption, metabolism, and distribution. This study lacks the proper controls for some Benchmark dose groups as controls were dosed on GD 9, 10, or 11 dose (mg/kg and some d-amphetamine-treated animals were treated bw/day, GD 9–15) on GD 9 and 10, or GD 9–11. The MTD was exceeded as Endpoint BMD BMDL evidenced by maternal mortality (37–42%) in all dose 10 groups and the high resorption rates (20–58%) in the Maternal weight 7.8 7.2 high-dose mice. Data on maternal body weights before, Fetal weight 2.1 1.5 during, and after dosing were not presented. Statistical Resorptions (per implant) 0.6 Not meaningful methods were unclear and possibly inappropriate. It is Malformations (per live implant) 9.2 8.4 not clear why the n-values varied so much in the experiments measuring cardiac Q–T interval (n 5 3–15) and whether this impacted the outcome of this 3% in controls) and the most common malformations measurement. (number of fetuses affected) were cyclopia (5) and Utility (Adequacy) for CERHR Evaluation Process: exencephaly (2). A significant increase in malformation This study is of limited usefulness for the evaluation rate also occurred in fetuses from rabbits treated with process. 1.5 mg/kg bw/day methamphetamine on GD 12–30 Kasirsky and Tansy (1971), in a study supported by (15.5% compared to 3% in controls). The most common NIH, evaluated the teratogenicity of methamphetamine malformation was exencephaly, which affected four in mice and rabbits. Fifty CF1 mice/group were i.v. fetuses. Heart malformations were not observed in either dosed with 5 or 10 mg/kg bw/day methamphetamine species. The study authors concluded that high i.v. doses HCl [purity not specified] on GD 9–11, 9–12, 12–15, or 9– of methamphetamine can induce congenital anomalies in 15 (plug day 5 GD 1). One control group of 50 mice was mice and rabbits. Benchmark doses (Table 35) and dose– not treated and a second group of controls was given response curves (Fig. 3) for the mouse study were saline on GD 9–15. Mice were killed on GD 19 and derived using EPA Benchmark Dose Software. [Calcula- implantation sites were examined. Fetuses were weighed tion of the benchmark dose and estimation of the dose– and examined by Wilson procedure. Every third fetus response curves is particularly limited given the use of was cleared for skeletal examination. Twelve New only 3 dose levels (0, 5, and 10 mg/kg bw/day) in this Zealand White rabbits/group were i.v. dosed with study.] 1.5 mg/kg bw/day methamphetamine HCl on GD 12– Strengths/Weaknesses: The high-dose i.v. administra- 15, 15–20, or 12–30. One control group of rabbits (n 5 12) tion protocol provides unique information that is realistic was not treated and a second control group was treated for drugs of abuse. The use of multiple doses and a with saline [day of treatment not specified]. Rabbits sensitive period approach are strengths. This study were killed on GD 30. Fetuses were examined for gross identified the most critical periods for methampheta- abnormalities, weighed, and sexed. [The method for mine-induced teratogenesis in the mouse (GD 9–15) and examination of visceral and skeletal malformations was the rabbit (GD 12–15/12–30). Although there were only not discussed.] Data for weight and malformation rate three dose levels, mouse data were presented in were analyzed by w2. [The litter was not considered in sufficient detail to permit benchmark dose calculations. the statistical evaluations. Chi-square tests are not The incomplete reporting in this older study is a appropriate for analysis of weight.] weakness. In addition, there was no indication that the Mice and rabbits showed signs of excitement from statistical analyses were litter-based. Although metham- 5 min to 6–7 hr after methamphetamine administration. phetamine-treated mice and rabbits had marked effects Maternal weights were significantly reduced in every on maternal and fetal body weights, which may have treatment group in both species. Feed and water intake contributed to developmental toxicity, there were no were reduced on the first 2–3 days of exposure [data not pair-fed controls. shown]. Fetal weights were significantly decreased in all Utility (Adequacy) for CERHR Evaluation Process: treatment groups in mice and rabbits. There were no This study is adequate for the evaluation process. significant effects on resorptions in either species. [The Yamamoto et al. (1992), support not indicated, treated Expert Panel noted that the resorption rate in mouse Jcl:ICR mice with a single dose of i.p. methamphetamine controls was B4% and in treated groups, the resorption HCl [purity not indicated] in saline on GD 8 (plug 5 GD rate was B15–22%.] 0). Methamphetamine doses and number of dams used In mice, the only significant increase in malformations were 0 (n 5 13), 11 (n 5 10), 13 (n 5 11), 14 (n 5 10), 15 occurred in fetuses from the group treated with 10 mg/ (n 5 16), 17 (n 5 19), 19 (n 5 17), or 21 (n 5 26) mg/kg bw. kg bw/day methamphetamine on GD 9–15; the mal- More animals were used for the higher dose groups formation rate in that group was 13.6% compared to 1% in anticipation of treatment-induced maternal death. in either control group. Malformations (numbers of Maternal death occurred in 3 of 16 dams at 15 mg/kg fetuses affected) included exencephaly (25), cleft palate bw, 5 of 14 dams at 17 mg/kg bw, 6 of 17 dams at 19 mg/ (10), microphthalmia (8), and anophthalmia (6). kg bw, and 13 of 26 dams at 21 mg/kg bw. Restlessness The malformation rate in both groups of control rabbits and agitation were described after dosing [presumably was 3%. A significant increase in malformation rate in all dose groups]. Maternal feed consumption and occurred in fetuses of rabbits dosed with 1.5 mg/kg bw/ weight gain were not reported. Fetal body weights at day methamphetamine on GD 12–15 (12% compared to hysterotomy on GD 18 were not affected by treatment.

Birth Defects Research (Part B) 74:471–584, 2005 DOI 10.1002/bdrb 530 GOLUB ET AL.

Maternal weight Fetal malformations

Fetal weight

Resorptions Mean Response

0 2 4 6 8 10 Dose (mg/kg/day)

Fig. 3. Dose–response curves for the mouse study reported by Kasirsky and Tansy (1971). Data points and variances removed. Y-axis is an arbitrary scale with non-zero origin. Tested doses were 0, 5, and 10 mg/kg bw/day on GD 9–15. Both doses of methamphetamine were effect levels by pairwise comparison except for malformations, for which 5 mg/kg bw/day was a NOAEL. Graphs were drawn using the EPA Benchmark Dose Software. Maternal and fetal weight were plotted as distributions (mean7SD). Resorptions and fetal malformations were plotted as proportions of total and live implants, respectively.

Fetuses were examined for external malformations and not considered a relevant route of exposure due to (1) about one-third of randomly selected fetuses were possible direct adverse effects of the drug or solvent on cleared, stained with Alizarin Red S, and evaluated for the uterus and secondarily on embryo-fetal development, skeletal abnormalities. [Visceral abnormalities were not and (2) the possibility that the drug might be physically mentioned.] Maternal mortality data and litter data for transported directly to the embryo/fetus, bypassing fetal death and external and skeletal malformations are maternal absorption, metabolism and distribution. The shown in Figure 4. The litter percent affected at 21 mg/ MTD was exceeded as evidenced by maternal mortality kg bw methamphetamine was lower than at 19 mg/kg (19–50%) at doses Z15 mg/kg. A pair-fed control group bw, and these high-dose data were ignored in graphing was not included. Maternal body weights and body the dose–response data in Figure 4. [The authors weight gains were not reported. Fetuses were not proposed that the 50% maternal mortality at the evaluated for visceral alterations. Animals were group 21 mg/kg bw dose would have led to a decrease in housed during pregnancy, which may confound preg- crowding of the dams, which were initially housed nancy outcomes, particularly given unequal numbers of three to a cage during pregnancy. This decrease in animals co-housed in the different treatment groups. crowding was believed to have improved pregnancy Utility (Adequacy) for CERHR Evaluation Process: outcome in the high-dose group. The Expert Panel The utility of the study is decreased by the potentially notes that group housing of pregnant animals is not important effects of housing. The lack of effect consis- conventional and finds it difficult to evaluate these tency at the top dose makes the results of this study data given the likelihood that there were different difficult to interpret. This study is of limited utility for numbers of dams per cage in all the dose groups (none dose–response modeling. of the group sizes are evenly divisible by 3). NOAEL, Burchfield et al. (1991) studied the pharmacokinetics LOAEL, and benchmark doses are provided in Figure 4 and pharmacodynamics of methamphetamine in preg- to facilitate comparison of the data produced in this nant sheep. On GD 125 (B85% of term), catheters were study. The Expert Panel does not mean to imply that inserted in Grade Western sheep and the animals were this study should be used as the basis for regulation.] given antibiotics during a 3-day recovery period. After Strengths/Weaknesses: The use of multiple doses and the recovery period, sheep received one or more i.v. standard methodology is a strength. Litter-based statis- treatments of methamphetamine that included 0.6 mg/ tics were used to evaluate fetal mortality and malforma- kg bw over 12.5 min, 1.2 mg/kg bw over 12.5 min, or tion rates, whereas fetal-based statistics were used for the 1.2 mg over 30 sec. Blood was collected from 4–5 ewes/ percentage of malformed fetuses. Study data were given group and methamphetamine levels were measured in in sufficient detail to calculate benchmark doses. The plasma by HPLC to determine pharmacokinetics in ewes authors concluded that methamphetamine has a steep and fetuses (discussed in Section 2.1.2.2). Arterial blood dose–response for teratogenesis and that it does not seem gas samples and maternal and fetal blood pressure and to be a selective fetal toxicant, because the difference heart rate were evaluated periodically. Methampheta- between the single teratogenic dose (19 mg/kg bw) and mine rapidly crosses the placenta in sheep, and fetuses the maternal LD50 dose (21 mg/kg bw) was very small. ultimately achieve higher methamphetamine concentra- Several weaknesses were noted. Pregnant mice were tions than adults due to a longer elimination half- dosed with methamphetamine by i.p. injection, which is life. Maternal and fetal blood pressure were increased

Birth Defects Research (Part B) 74:471–584, 2005 DOI 10.1002/bdrb AMPHETAMINE AND METHAMPHETAMINE 531

Level (mg/kg ip on GD 8) 0.7 End point NOAEL LOAEL BMD10 BMDL 50 0.6 Maternal mortality 17 19 14.9 12.7 Litter % Fetal death ≥21 none 14.6 10.3 40 0.5 External malformation 17 19 16.1 12.2 Skeletal malformation 13 14 15.5 12.4 0.4 30 0.3 20

Fraction Affected 0.2 10 0.1 Mean Litter Response (%) 0 0

0 5 10 15 20 Dose (mg/kg ip on GD 8) Maternal mortality (left axis) Dead fetuses External malformations litter %, right axis, top dose omitted Skeletal malformations

Fig. 4. Dose–response curves from the data of Yamamoto et al. (1992). The curves in the figure were drawn using EPA Benchmark Dose software. Data points and variances have been omitted for clarity. The NOAEL and LOAEL were based on author assessment of significance on pair-wise comparison with control, except for maternal mortality, which is based on pair-wise comparison by CERHR. The benchmark dose corresponding to a 10% effect level (BMD10) and the benchmark dose corresponding to the lower bound of the 95% CI at this effect level (BMDL) were calculated by CERHR using EPA Benchmark Dose software version 1.3.2. by maternal methamphetamine exposure and the mag- (Anggard et al., 1970). A weakness is the possibility that nitude of the increase was independent of dose regimen. instrumentation could modify the treatment effects. In Fetal heart rate was decreased in response to the 1.2 mg/ addition, sheep were used for multiple samples, meaning kg bw maternal bolus dose of methamphetamine. Fetal that data across samples are not independent trials. oxyhemoglobin saturation was decreased by all maternal Aside from maternal methamphetamine elimination treatments and fetal arterial pH was decreased by half-life and fetal pre-exposure oxyhemoglobin satura- maternal methamphetamine at 1.2 mg/kg bw. There tion levels, other factors affecting fetal elimination half- was a greater decrease in fetal arterial pH after bolus life are unknown. Limited data were available with than slow administration of methamphetamine to the which to compare levels observed in maternal and fetal mother. The authors believed their results to be consis- sheep with humans. There are a number of factors that tent with a decrease in placental perfusion after maternal can affect cross-species extrapolation of these data methamphetamine administration. [The Expert Panel (differences in placental structure, cross-species differ- noted that the most interesting elements of this study ences in plasma protein binding or fetal metabolic were the correlation between fetal methamphetamine capacity, etc.). There are no corresponding developmen- disposition and fetal oxyhemoglobin levels, as well as tal toxicity studies in sheep to demonstrate that these the maternal/fetal physiological changes (blood pres- fetal changes result in adverse developmental outcome. sure, fetal oxyhemoglobin saturation, fetal pH, etc.) that Utility (Adequacy) for CERHR Evaluation Process: accompany methamphetamine exposure. The authors This study is adequate for the evaluation process. hypothesized that decreased maternal vasodilation may Dickinson et al. (1994), support not indicated, eval- account for fetal hypoxia and maternal and/or fetal uated the effects of methamphetamine [purity not hypertension could lead to poor fetal outcome.] specified] on pregnant sheep [breed unspecified] and Strengths/Weaknesses: Strengths of this study include their fetuses. Catheters were surgically placed in mater- the evaluation of possible mechanisms of toxicity, the nal and fetal blood vessels on GD 125–128 (third real-time evaluation of the fetal sheep, and documenta- trimester). After a 5–7-day recovery period, methamphe- tion of some important pharmacokinetic principles for tamine was given i.v. (maternal femoral vein) over 2– methamphetamine administered i.v. to pregnant sheep. 3 min at 1.25 mg/kg bw. Fetal blood was sampled 15, 30, Doses used in this study were reportedly at or below 60, 90, 120, and 180 min after methamphetamine admin- those used recreationally [60–100 mg; (Hall et al., 1988)], istration. Mean arterial blood pressure and heart rate yet maternal ovine blood concentrations were reported to increased in mother and fetus after methamphetamine. be approximately equivalent to those achieved in hu- There were increases in maternal plasma glucose, mans 1 hr after receiving 160–200 mg amphetamine i.v. insulin, epinephrine, and norepinephrine after treatment.

Birth Defects Research (Part B) 74:471–584, 2005 DOI 10.1002/bdrb 532 GOLUB ET AL. Maternal blood gases were not altered. Fetal plasma number of factors that can affect cross-species extrapola- glucose, insulin, lactate, epinephrine, and norepinephr- tion of these data (differences in placental structure, ine increased after maternal methamphetamine admin- cross-species differences in plasma protein binding, or 7 istration. Fetal arterial pO2 decreased from a mean SEM fetal metabolic capacity, etc.). There are no corresponding of 21.471.9 mm Hg to 15.371.3 mm Hg by 60 min after developmental toxicity studies in sheep to demonstrate maternal methamphetamine. There were no significant that these fetal changes result in adverse developmental changes in fetal arterial pH or pCO2 by pair-wise outcome. comparison with the baseline values [trend testing by Utility (Adequacy) for CERHR Evaluation Process: CERHR showed a significant linear trend (P 5 0.016) This study is adequate for the evaluation process. for pH, which declined from a baseline mean7SEM of Stek et al. (1993) implanted catheters in the maternal 7.3570.01 to a 180-min value of 7.2970.02]. The authors and fetal vessels of 7 mixed-breed sheep at 107 days concluded that changes in release of catecholamines after gestation (early third trimester). After a 1-week recovery methamphetamine treatment may have caused glucose period, methamphetamine HCl [purity not specified] changes and alterations in placental perfusion leading to 1 mg/kg bw was administered into the maternal vena fetal hypoxemia. cava over 1.5 min. This dose was characterized as similar Strengths/Weaknesses: Strengths include the collec- to a recreational dose of methamphetamine in humans. tion of data that bear on mechanism, the use of real-time Maternal and fetal serum and amniotic fluid samples cardiovascular data, the use of a standard experimental were drawn periodically for measurement of metham- model, and the collection of third trimester information. phetamine concentration (discussed in Section 2), and The authors present useful gestational pharmacokinetic measurements were made of cardiovascular parameters data and confirmed many of the observations reported periodically over 3 hr. Results are shown in Table 36. The by Burchfield et al. (1991) (i.e., increased blood pressure, authors concluded that methamphetamine crosses the heart rate, decreased fetal pO2, etc.). Maternal and fetal ovine placenta and produces significant and prolonged catecholamines (epinephrine and norepinephrine) were maternal and fetal cardiovascular changes that they also measured, along with increased fetal glucose, presumed would be potentially detrimental to maternal insulin, and lactic acid levels. The resulting profile was and fetal health. consistent with alterations in fetal sympathoadrenal Strengths/Weaknesses: Strengths include the collec- activity and fetal hypoxemia. There is concern that tion of data that bear on mechanism, the use of real-time instrumentation may have altered the response to cardiovascular data, the distinction between direct and treatment. A weakness of the study is that multiple indirect effects, the use of a standard experimental samples were collected from each sheep, increasing the model, and the collection of third trimester information. likelihood of confounding results (i.e., data across Time course data for maternal and fetal methampheta- samples are not independent trials). Limited data were mine are useful. To ensure the integrity of placental available with which to compare levels observed in transfer before methamphetamine dosing, fetuses were maternal and fetal sheep with humans. There are a required to meet minimum standard values for pO2 and

Table 36 Cardiovascular Parameters in Early Third-Trimester Sheep after Administration of 1 mg/kg bw over 1.5 Min to the Vena Cava of the Ewe

Maximum Min to maximum response/ Parameter Baseline response resolutiona Pb

Maternal Mean arterial blood pressure (mm Hg) 6974 148792/4180 0.03 Heart rate (beats/min) 10275 161719 5/4180 0.012 Cardiac output (L/min) 5.971.1 8.071.0 5/4180 0.09 Systemic vascular resistance, calculated 13.673.6 18.972.6 2/60 NS (mm Hg/L/min) Uterine blood flow (mL/min) 672748 983765 2/5 0.009 Uterine vascular resistance, calculated 0.1070.01 0.3970.17 5/120 0.05 (mm Hg/L/min) Fetal blood pressure (mm Hg) 49736473 5/180 0.04 Heart rate (beats/min) 15775 213711 120/4180 0.03 Umbilical blood flow (mL/min) 503740 412744 5/10, with overshoot Not given (0.15 thereafter to by CERHR) B550 mL/min Umbilical vascular resistance, calculated 0.1070.01 0.1770.02 5/60 0.02 (mm Hg/L/min) Arterial pH 7.33470.017 7.28670.016 10/NS NS 7 7 Arterial pO2 (mm Hg) 22.4 0.6 17.4 1.5 5/NS 0.06 From Stek et al., 1993. Mean7SEM unless otherwise given; n 5 7 ewes/fetuses except for maternal cardiac output and systemic vascular resistance where n 5 4. aResolution estimated from graph and not based on statistical testing. bRepeated-measures ANOVA for change over time, as per authors.

Birth Defects Research (Part B) 74:471–584, 2005 DOI 10.1002/bdrb AMPHETAMINE AND METHAMPHETAMINE 533 pH. The dose level used (1 mg/kg) approximates typical decreased in a dose-dependent manner from a mean7 human doses of methamphetamine users. The authors SEM of 21.272 mm Hg at baseline to 16.372.2 mm Hg used appropriate statistical models (repeated-measures after the maternal 1.0 mg/kg bw dose of methampheta- analysis of variance) where possible. When not possible mine. There were no statistically significant changes in due to limited degrees of freedom, they decreased the P- maternal or fetal pH [presumably arterial],orin value for paired t-tests to reduce the possibility of Type I maternal or fetal arterial pCO2. After fetal administration error. Results from this study are consistent with those of methamphetamine, there was no significant alteration reported by Burchfield et al. (1991). There is concern that in fetal arterial pO2, although there was a significant instrumentation may have altered the response to increase in fetal mean arterial blood pressure and a treatment. A weakness is that sheep were used for decrease in fetal pH. The authors concluded that the multiple samples with at least 2 days of recovery between alteration in fetal arterial pO2 after maternal metham- experiments; thus, data across samples are not indepen- phetamine exposure was likely due to effects of the drug dent trials. Amniotic fluid samples for the determination on placental or uterine vasculature because no alteration of methamphetamine levels were not available from all occurred in fetal arterial pO2 after direct administration fetuses at each time point. Limited data were available to the fetus. They also concluded that alterations in fetal with which to compare methamphetamine concentrations blood pressure and pH after maternal administration of and kinetics in humans to those observed in maternal and methamphetamine were likely due to direct effects on fetal sheep. There are a number of factors that can affect the fetus after placental transfer. [The Expert Panel cross-species extrapolation of these data (differences in notes, however, that there were no significant effects uteroplacental structure, cross-species differences in on fetal pH after maternal administration of metham- plasma protein binding or fetal metabolic capacity, etc.). phetamine in this experiment; the authors concluded a There are no corresponding developmental toxicity decrease in fetal pH despite of the lack of statistical studies in sheep to demonstrate that these fetal changes significance. The Expert Panel also notes that signifi- result in adverse developmental outcome. cant decreases in ovine fetal arterial pH were noted in Utility (Adequacy) for CERHR Evaluation Process: another study (Burchfield et al., 1991).] This study is adequate for the evaluation process. Strengths/Weaknesses: Strengths include detailed ex- Stek et al. (1995), supported by NIH, in a follow-up to amination of the pH issue, the use of a standard model, their previous study (Stek et al., 1993), evaluated whether the addressing of maternal-fetal influences, and the use a methamphetamine-associated decrease in fetal arterial of multiple low doses. Another strength is that integrity pO2 could be attributed to a reduction in uterine blood of placental transfer before methamphetamine dosing flow. Nine mixed-breed ewes underwent surgical place- was ensured by requiring fetuses to meet designated pO2 ment of maternal and fetal catheters and uterine artery and pH values. A weakness is that sheep were used for flow probes on GD 110–115. Animals were permitted to multiple samples with doses administered at 30- to 35- recover for at least 5 days before experimentation. min intervals, meaning that data across samples are not Methamphetamine HCl [purity not specified] was given independent trials. Limited data were available with in increasing doses of 0.03, 0.1, 0.3, and 1.0 mg/kg bw. which to compare levels observed in maternal and fetal Each dose was given by i.v. infusion [possibly through a sheep with humans. There are a number of factors that vena cava catheter] over 60–90 seconds. Infusions were can affect cross-species extrapolation of these data separated by 30–35 min and recordings were made 5 min (differences in uteroplacental structure, cross-species after each infusion. On a different day, fetal infusions differences in plasma protein binding or fetal metabolic were given using methamphetamine doses of 0.03, 0.1, capacity, etc.). There are no corresponding developmen- 0.3, 1.0, and 3.0 mg/kg estimated bw at 30–35-min tal toxicity studies in sheep to demonstrate that these intervals [the fetal vessel used for the infusion was fetal changes result in adverse developmental outcome. not specified; both the aorta and the vena cava were Utility (Adequacy) for CERHR Evaluation Process: cannulated]. Maternal mean arterial blood pressure This study is adequate for the evaluation process. increased in a dose-dependent manner, peaking at a 3.2.1.2 In vitro studies. mean of 76% over baseline after the 1.0 mg/kg bw dose. Yamamoto et al. (1995) examined developmental effects Total uterine blood flow decreased in four animals by a in Wistar rat embryos exposed to d-methamphetamine mean of 33% at the 1.0 mg/kg bw dose and was not HCl in vitro. GD 10.5 embryos (plug day 5 GD 0; n 5 8– statistically changed in two animals, with the seventh 26/group) were incubated for 24 hr in media containing 0, animal excluded due to probe malfunction. The net effect 0.1, 0.2, 0.4, 0.6, or 0.8 mM methamphetamine (0, 15, 30, 60, in the six animals was not significantly different from 90, or 120 mg/mL, according to author calculations) and baseline. Uterine vascular resistance increased in a dose- examined for mortality, growth, and malformations. dependent manner in the six evaluable animals to a Exposure to methamphetamine did not affect embryo maximum of 140% of control after methamphetamine viability. Methamphetamine treatment (concentration of 1.0 mg/kg bw. Fetal mean arterial pressure increased in a effect) significantly reduced crown-rump length, number dose-dependent manner after maternal methampheta- of somites, and protein content (Z0.4 mM), developmen- mine administration, with a mean 28% increase after the tal score (Z0.6 mM), and yolk sac diameter (0.8 mM). The 1.0 mg/kg bw dose. Umbilical blood flow increased by a number of malformed embryos was significantly in- mean of 15% from baseline after the 0.3 mg/kg bw dose creased at Z0.6 mM. The most common malformations to the mother, without additional increase at 1.0 mg/kg included microcephaly, neural tube closure defects, bw. Umbilical vascular resistance was described by the incomplete rotation of body axis, and tortuous spinal authors as having increased after the 1.0 mg/kg bw cord. Derangement and necrosis in neuroepithelial tissues maternal dose, but statistical testing showed no signifi- were observed at 0.8 mM. Based on the observation that a cant change from baseline. Mean fetal arterial pO2 70 mg/kg bw i.p. dose results in a peak blood level of

Birth Defects Research (Part B) 74:471–584, 2005 DOI 10.1002/bdrb 534 GOLUB ET AL. o50 mg/mL in rats, the study authors suggested that the methamphetamine plasma concentration seen in fatal methamphetamine concentrations inducing malforma- cases of human overdose. tions in this in vitro study are likely to be much higher Utility (Adequacy) for CERHR Evaluation Process: than the plasma level obtained with a single teratogenic This study is of limited utility. In vitro studies can be dose of 19 mg/kg bw in a mouse study. [It was not used as supplemental information in the evaluation explained how values were extrapolated from rats to process. mice. According to a study by Acuff-Smith et al. (1996), Won et al. (1992), in a study funded by several methamphetamine serum levels peaked at 3.1–3.6 lg/mL government and industry grants, examined the effects in rats s.c. exposed to 40 mg/kg bw/day methampheta- of methamphetamine in a three-dimensional, rotation- mine (Section 2.1.2.2), a dose that caused increased fetal mediated reaggregate tissue culture system. In the assay, mortality but no increase in external malformations. single cells from specific fetal brain regions are incubated Methamphetamine blood levels were reported at 0.03– in a rotation culture system where they could develop 6.3 lg/mL in infants exposed to methamphetamine in and interact with each other. For this study, cells were utero (see Section 2.1.1.3).] obtained from the rostral mesencephalic tegmentum and Yamamoto et al. (1998), supported by the Japanese corpus striatum of embryonic mice and cultured in government, cultured 9–10 somite Wistar rat embryos monoamine-free media for 15 days. Methamphetamine (about GD 10) with d-amphetamine [purity not speci- was added to the culture at concentrations of 0 or fied] at concentrations of 0, 0.1, 0.4, 0.8, 1.2, and 1.6 mM 10À7–10À4 M [14.9–14,900 ng/mL assuming the values [14, 54, 108, 162, and 216 lg/mL]. There were 20 control given are for methamphetamine free base and not the embryos and 10–13 embryos in each of the amphetamine- salt] for 7 days (culture days 15–22) and then levels of exposed groups. Embryos were evaluated after 24 hr in endogenous dopamine and serotonin were measured. culture. There were no apparent adverse effects on Methamphetamine treatment resulted in significant embryo development at amphetamine concentrations of dose-related reductions of dopamine at Z10À6 M and 0.8 mM or lower. At 1.2 mM and higher, there was serotonin at Z10À5 M. A time-course experiment demon- a significant decrease in yolk-sac diameter, crown-rump strated that 10À4 M methamphetamine reduced dopa- length, somite number, and protein content, and an mine and serotonin levels after 3 days of exposure, but increase in the proportion of abnormal embryos. One that no further reduction followed at 5 and 7 days of embryo each in the 1.2- and 1.6-mM groups died during exposure. A recovery experiment using 15-day-old the incubation period. The authors concluded that aggregates exposed to 0 or 10À4 M methamphetamine amphetamine was less potent in producing abnormal for 7 days demonstrated that dopamine levels rose from embryo development than methamphetamine based on a 28.8 to 73.7% of control levels and serotonin levels rose comparison of the lowest effective concentration from 15.1 to 42.5% of control levels during the first 9 days (1.2 mM) in this study and the lowest effective metham- of recovery. During the next 11 days of recovery, phetamine concentration (0.6 mM) in their prior study dopamine levels remained at 75% of control values and (Yamamoto et al., 1995), and based on histologic evidence serotonin levels at B50% of control values, indicating of necrosis in the neuroepithelium at 1.2 mM ampheta- that complete recovery did not occur. [Thus, longer term mine in this study, which was comparable in effect to effects on axonal growth or cell maturation during the 0.8 mM methamphetamine in the previous study. recovery experiments are possible.] Strengths/Weaknesses: Strengths of the in vitro An earlier study from the same group (Kontur et al., studies by Yamamoto et al. (1995) are that they examine 1991) used a tyrosine hydroxylase immunochemistry the direct effects of amphetamine and methamphetamine method to demonstrate that methamphetamine-induced on embryonic development during a limited develop- (10À4 M) reduction of serotonin in a reaggregate tissue mental stage and allow comparisons between in vitro culture was not due to loss of dopamine-containing cell and in vivo dosing. The whole embryo culture techni- bodies. ques seemed to be sound and the investigators controlled Strengths/Weaknesses: A strength of this study is its for initial developmental stage by limiting the study to applicability to mechanism considerations. Other embryos with 9–10 pairs of somites. A weakness is that strengths are that before culturing neurons, the authors a limited period of development can be studied with dialyzed the culture sera to remove exogenous serotonin embryo culture. There is no indication that the authors and that experiments were adequately replicated. A controlled for litter effect by dividing embryos from each weakness is that the embryonic age of the donor animals litter across the various treatment groups. It is unclear is not specified. The in vitro cell culture system why the authors used twice as many embryos in the circumvents maternal and fetal absorption, distribution, control group as in the treatment groups (20–26 control metabolism, and excretion. The authors mentioned that embryos vs. 8–14 embryos in the treated groups). the concentrations used in this study were similar to Embryos were not scored blind with respect to treatment peak brain concentrations (1.65 Â 10À4 M) achieved dur- group. This whole embryo culture system circumvents ing systemic administration of a methamphetamine maternal absorption, distribution, and metabolism, there- congener; however, there is no indication that this value fore potentially exposing embryos for a longer period of is representative of embryonic brain concentrations, the time. Furthermore, the concentrations used in these system examined in these experiments. studies were excessive. The authors estimate that the in Utility (Adequacy) for CERHR Evaluation Process: vitro teratogenic concentration of methamphetamine was This study is of limited utility. In vitro studies can be used much higher than the plasma concentrations that could as supplemental information in the evaluation process. be achieved with a single teratogenic dose given to ICR 3.2.1.3 Chicken studies. mice in a previous study. Furthermore, the in vitro Cameron et al. (1983), in a study funded in part by the teratogenic concentration was greater than the American Heart Association, examined mechanisms of

Birth Defects Research (Part B) 74:471–584, 2005 DOI 10.1002/bdrb AMPHETAMINE AND METHAMPHETAMINE 535 d-amphetamine-induced malformations in White Leg- and may have been inadequate to control for the higher horn chicken embryos. The first stage of the study d-amphetamine exposures (0.5, 0.75, 1.0 mg). Saline- involved dose range-finding studies for d-amphetamine treated controls for single experiments had only a 72% sulfate [purity not specified], a-methyl-p-tyrosine (a survival rate and those used in double experiments catecholamine synthesis blocker), metoprolol (a b-adre- (saline–saline) had only a 64–67% survival rate. Aside nergic blocker), and phentolamine (an a-adrenergic from a-methyl-p-tyrosine (the catecholamine synthesis blocker). [The rationale for selecting doses for further inhibitor), both the a and b blockers (phentolamine and study was not clear.] In the second stage of the study, metoprolol, respectively) were effective at decreasing d- embryos were exposed to d-amphetamine in combination amphetamine-induced malformations. This lack of spe- with one of the other drugs. Normal embryos were cificity is unusual. The authors suggest that adrenergic exposed to the drugs at 96 hr, incubated for 15 days, and receptors may not be as specific in 4-day-old chick then autopsied under a dissecting microscope. Embryos embryos, but there are no data to support this claim. It is were exposed to a-methyl-p-tyrosine 6 hr before d- not always clear why the selected doses were designated amphetamine treatment, and were exposed to the other as effective (e.g., the lowest dose of phentolamine was two drugs 1 hr after d-amphetamine treatment. Controls deemed effective at decreasing mortality and malforma- consisted of eggs that were opened but untreated or tions; however, the next dose level (0.010 mg), did not embryos treated with saline or a hyperosmotic solution affect survival (67 vs. 68% in controls) or malformation to match the osmolarity of d-amphetamine. In the rates (10%, a non-significant change). These inconsisten- mechanistic studies, controls treated with only d-amphe- cies make the selection of effective doses seem somewhat tamine or d-amphetamine and saline were included. arbitrary. None of these agents improved embryonic Eighteen to 96 embryos were used per group. Data were survival. evaluated by w2. Utility (Adequacy) for CERHR Evaluation Process: Malformations were increased at d-amphetamine This study is of limited utility but can add supplemental doses of Z0.25 mg/embryo and survival was decreased information in the evaluation process. at d-amphetamine doses of Z0.50 mg/embryo. All Cameron et al. (1984), in a study supported in part by malformations involved the heart and great vessels and the American Heart Association, measured heart rate included abnormal persistence of the left fourth aortic and mean ventricular blood pressure in 4-day-old White arch and ventricular septal defects. [The Expert Panel Leghorn chicken embryos exposed to a teratogenic dose notes that cardiac malformations were consistent with (0.25 mg) of d-amphetamine sulfate [purity not speci- those in the study by Nora et al. (1968).] The 0.25 mg fied]. Eggs containing control embryos were either dose was selected for the mechanistic studies; that dose unopened or injected with saline. A total of 16–34 resulted in a malformation rate of B30%, a level embryos/group were examined hourly at 1–7 hr after significantly greater than the rate observed in controls exposure. Heart rate was significantly lower in the (o10%). The incidence of d-amphetamine-induced mal- amphetamine compared to the control groups at 2, 3, formation was significantly reduced to rates of r10% by and 5 hr after exposure. Mean ventricular blood pressure co-treatment with a-methyl-p-tyrosine (0.50 mg, the high was significantly increased by 30% in the amphetamine dose), metoprolol (0.1 mg, the mid dose), and phentola- group compared to the control groups at 4 hr after mine (0.005 mg, the low dose). None of the drugs exposure. The study authors concluded that increased improved survival after d-amphetamine treatment. The embryonic blood pressure during a critical period of study authors postulated that chicken embryos may be aortic arch and ventricular septum development may be able to respond to the a- and b-adrenergic properties of a causal factor for the malformations observed in these d-amphetamine through release of endogenous catecho- structures after amphetamine exposure. [The Panel lamines and that the process may have a causal notes that increases in mean ventricular blood pressure relationship to malformations. recorded in chick embryos are consistent with the Strengths/Weaknesses: A strength is the evaluation of findings of other investigators, who have shown mechanisms of cardiovascular effects. All embryos were increases in blood pressure in d-amphetamine-exposed examined to verify that they were at normal stage 24 fetuses (e.g., sheep studies discussed previously in before use in these experiments, and adequate controls Section 3.2.1.1).] were included (i.e., in single-treated experiments: Strengths/Weaknesses: A strength of this study is the opened-untreated controls, saline-treated controls, and addition of a functional dimension to the study of controls treated with Na2SO4 to control for the hyper- amphetamine effects on the embryonic chicken heart. osmotic environment created by 0.25 mg d-amphetamine; All embryos were examined to verify that they were at in double-treatment experiments: untreated, saline1 normal Stage 24 before use in these experiments and saline, d-amphetamine alone, and d-amphetamine1sal- adequate controls were used in the experiments (i.e., ine). Treatment periods for a-methyl-p-tyrosine, phento- opened-untreated controls, saline-treated controls, and lamine and metoprolol were designed so that the controls treated with Na2SO4 to control for the hyper- maximum effectiveness of these treatments coincided osmotic environment created by 0.25 mg d-ampheta- with d-amphetamine treatment. Doses used in this study mine). The d-amphetamine concentrations used in this (0.1–1.00 mg 5 2.7 Â 10À7 to 2.7 Â 10À6 mol) were reported study (0.25 mg 5 6.8 Â 10À7 mol) were reported to be to be within the concentration range seen in humans in similar to human concentrations seen in other studies. other studies. A weakness is the lack of a maternal A weakness is the lack of a maternal system inherent in system inherent in this experimental model. There is no this experimental model. Blood pressure analyses were indication that fetal examiners were blind with respect conducted using a one-way ANOVA (not two-way to treatment group. The hyperosmotic control was ANOVA). Graph lines in Figure 2 of the study seem to designed to mimic the 0.25 mg d-amphetamine group be mislabeled. Furthermore, the authors state that ‘‘a

Birth Defects Research (Part B) 74:471–584, 2005 DOI 10.1002/bdrb 536 GOLUB ET AL. significant increase in blood pressure was seen 4 hr after endpoints in experimental animals exposed during d-amphetamine treatment, despite a lower heart rate;’’ prenatal or postnatal development. Testing was con- in fact, there was no difference in heart rate in ducted to examine neurological endpoints in many of the d-amphetamine-treated chicks at 4 hr. Although the studies; the neurological findings are discussed in investigators show that blood pressure increased after Section 3.2.3. With the exception of one oral exposure d-amphetamine treatment, they do not establish that this study conducted in monkeys, the studies summarized in 1-hr increase in ventricular blood pressure is sufficient in this section examined rats exposed parenterally. The magnitude or duration to cause persistent left fourth order of presentation is amphetamine studies followed aortic arch and ventricular septal defects. by methamphetamine studies. Utility (Adequacy) for CERHR Evaluation Process: Summaries of unpublished studies on Adderall were This study is of limited utility but can add supplemental provided to the Expert Panel by Shire Pharmaceuticals information in the evaluation process. Group (Secker, 2004). A GLP-compliant juvenile toxicity Kolesari and Kaplan (1979) examined amphetamine study in Sprague–Dawley rats was performed using and methamphetamine effects on development of White amphetamine base-equivalent doses of 0, 2, 6, or 20 mg/ Leghorn chicken embryos. Eggs containing embryos at kg bw/dose. Animals were given a dose by gavage 45–49 hr of development (Hamilton-Hamburger Stage 12) once/day from PND 7–13 and twice/day from PND 14 were injected subgerminally with 0.5 mg d-amphetamine to 59 (20 animals/group; 10 animals/group were sulfate (n 5 17–21/final stage) or 1.0 mg methampheta- continued for up to an additional 6 days). Body weight mine HCl (n 5 19/final stage). Control eggs were either gain was decreased in the highest two dose groups and, left unopened (n 5 20–21/final stage) or injected with in females, in the low dose group from PND 7–24. saline (n 5 7–12/final stage). At 24 hr after injection Clinical signs were present in all amphetamine-treated (Stage 17 or 18), embryos were examined for hematomas groups. Preputial separation and vaginal opening were and growth. Statistical significance of effects was delayed in the highest dose group. Motor activity was determined by Student t-test. Compared to either control decreased at all amphetamine doses with testing on PND group, d-amphetamine and methamphetamine treatment 22 and 47. Mating and fertility were normal. There were significantly reduced crown–rump length and cross no effects of treatment on the proportion of females with sectional area of aorta, notochord, neural tube, and regular estrous cycles or on numbers of corpora lutea, whole body. Fewer structures were affected at Stage 18 implantations, or conceptuses, or on post-implantation versus Stage 17. Embryos treated with d-amphetamine loss. Behavioral effects were not described. [These and methamphetamine had increased numbers of caudal summaries are noted but were not used by the Expert hematomas. Panel in the Evaluation Process.] Strengths/Weaknesses: A strength of this study is the Ching and Tang (1986), supported by PHS and the provision of detailed data on growth beyond overall Human Growth Foundation, treated pregnant Sprague– weight. Eggs were selected at Stage 12, the maximum Dawley rats by gavage with d-amphetamine sulfate time of susceptibility to hematoma formation. Data were [purity not specified] in water at 0- (n 5 6), 1- (n 5 7), presented for both untreated (unwindowed) and saline- 2- (n 5 3), or 5- (n 5 7) mg/kg bw/day from 3 days after a treated controls. Treatment with amphetamine and sperm-positive smear until pups were weaned on PND methamphetamine resulted in reductions in embryonic 25. [The Expert Panel assumes that pups were reared by size, a finding reported in other studies as well. A their own dams and that culling was not employed.] weakness is the lack of a maternal system inherent in this Pup weight and length were measured on the day of experimental model. This study used only a single dose birth. Sexual development was assessed in female level, so dose–response comparisons were not possible. offspring by recording the day of vaginal opening. Embryos were not scored blind with respect to treatment Offspring were decapitated on PND 25 and hypothalami group. There were no controls for osmotic changes that dissected and homogenized. A methanol extract of two were shown to occur in chick embryos in the study by hypothalami per sample was evaporated and stored Cameron et al. (1983) (10 mL volume in isotonic saline; under refrigeration (61C) until tested for growth hor- 494 mOsm/L for a 0.25 mg dose (6.8 Â 10À7 mol) vs. mone releasing-activity. [The authors do not indicate 20 mL volume of sterile saline with 0.5 mg d-amphetamine whether the two hypothalami that were combined were or 1.0 mg methamphetamine in the present study). For from littermates or from animals of the same sex.] crown-rump length, windowing the eggs had a greater Growth hormone-releasing activity was evaluated by impact on size than amphetamine treatment. The authors incubation of rat anterior pituitary cell cultures with suggest that proximity of the caudal dorsal aorta to the reconstituted hypothalamus extract and subsequent underlying yolk containing d-amphetamine/metham- radioimmunoassay for rat growth hormone. phetamine may contribute to hematoma formation. The 5 mg/kg bw dose of amphetamine caused a There are interspecies differences in the proximity of reduction in feed and water consumption over an aorta and yolk sac. Caudal hematomas in chick embryos unspecified 2 days [not specified whether during are a precursor to rumplessness in chick embryos and are pregnancy or lactation; no data were provided for the considered a model of human caudal dysplasia syn- 2 mg/kg bw dose group]. There was no treatment effect drome; however, this syndrome has not been associated on litter size or pup mortality at birth. Newborn body with human prenatal amphetamine exposure. weight was increased compared to control in offspring in Utility (Adequacy) for CERHR Evaluation Process: the 1- and 2 mg/kg bw groups, and decreased in the This study is of limited utility but can add supplemental 5 mg/kg bw group. Body weights in the lower two information in the evaluation process. amphetamine groups were 106–116% of control, and 3.2.2 Postnatal development endpoints (non- body weights in the high dose group were 92–97% of neurological). This section presents non-neurological control [estimated from a graph; the graph shows males

Birth Defects Research (Part B) 74:471–584, 2005 DOI 10.1002/bdrb AMPHETAMINE AND METHAMPHETAMINE 537 to weigh less than females in all groups, including the females at the time of vaginal opening were not reported. control group; the Expert Panel assumes the graph was Additionally, there are no data for the 1 and 5 mg/kg bw mislabeled]. At 1 week of age, pup mortality was d-amphetamine doses, thus there was no opportunity to increased in the 5 mg/kg bw group (27.1% of pups in look for dose-related trends in age at vaginal opening. five of seven litters compared to the control rate of 1.5% For experiments examining growth hormone-releasing of pups in one of five litters). [Benchmark dose was activity, there is insufficient experimental detail to calculated based on per-offspring analysis: BMD10 determine how pups were sampled. The fact that there 3.2 mg/kg bw/day, BMDL 2.2 mg/kg bw/day.] Vaginal are eight animals per group suggests that there was no opening was reported to have been delayed in 15% of control for litter effect, as the largest number of litters in offspring whose dams were treated with 2 mg/kg bw/ the d-amphetamine groups was seven. There is no day amphetamine [no data shown for other groups]. indication as to whether hypothalamic samples were Growth hormone release by cultured anterior pituitary from males or females. Growth hormone-release data cells was decreased by hypothalamic extracts from the were similar for controls in both the presence and the 1 and 5 mg/kg bw amphetamine groups [data were absence of hypothalamic extract. The authors speculate not presented for the 2 mg/kg bw group]. The control that there is a growth hormone inhibitory activity in the value (mg growth hormone/mL/2.5 hr; mean7SEM) was hypothalamic extracts, but offer no supporting evidence. 0.5970.06. The value associated with the low-dose The authors should have provided some context for these amphetamine treatment was 0.3270.07 and the value control values (e.g., previous growth hormone release associated with the high-dose amphetamine treatment values). The fact that growth hormone-releasing activity was 0.2570.03 (n 5 8 cultures per treatment group, was decreased in the 1 mg/kg bw d-amphetamine group Po0.01, Duncan test). The authors concluded that chronic and 5 mg/kg bw d-amphetamine group to nearly the administration of ‘‘relatively high doses’’ of amphetamine same degree when there were marked differences in pup during pregnancy and lactation can alter the survival, body weights/sizes on PND 1 merits careful considera- growth, and development, and the hypothalamic growth tion. There were no data for the 2 mg/kg bw d- hormone releasing activity of offspring. amphetamine group. Body weight data from these Strengths/Weaknesses: A strength of this study is that groups on PND 25 would have been useful to assist in rats were exposed orally by gavage (a relevant route) to the interpretation of these data (i.e., Would body weights multiple doses (0, 1, 2, or 5 mg/kg bw/day d-ampheta- support decreased growth hormone-releasing activity?). mine) from GD 3 to PND 25. Unfortunately, the short- The authors also show that food and water consumption comings of the study design and data analyses make this were decreased in 5 mg/kg bw d-amphetamine dams study uninterpretable. There are no maternal data on over a 2-day period; however, they do not specify which body weight, body weight gains, or gestational clinical 2 days (gestation? lactation?). The lack of experimental observations with which to assess maternal toxicity. detail make the feed and water consumption data There is a reference to altered maternal behavior in the uninterpretable. Again, 2 mg/kg bw d-amphetamine discussion that suggests that decreased postnatal survi- group data are missing. val in the high-dose d-amphetamine group was, at least Utility (Adequacy) for CERHR Evaluation Process: partially, related to altered maternal caregiving. There This study is not useful for the evaluation process. is no explanation given for the small number of litters (3) Hitzemann et al. (1976), supported by U.S. PHS, in the 2 mg/kg bw d-amphetamine group. There is no treated pregnant Sprague–Dawley rats with d-ampheta- indication that litter-based statistical analyses were mine sulfate [purity not specified] 0, 1, or 3 mg/kg bw conducted. This is critical as pups from the same litter s.c. twice daily from GD 5 through parturition on GD 19– are not independent samples. The authors mention that 22. Litters were culled randomly to eight at birth. Culled data were analyzed using a Student’s t-test or modified pups were ‘‘examined for gross pathological abnormal- t-test, which is inappropriate when there are four dose ities’’ [method of examination not otherwise specified]. groups. Figure 1 of the study shows the body weights The number of treated dams was not given. [There were and lengths of newborn offspring ranging in age from 80 control, 88 1 mg/kg bw, and 56 3 mg/kg bw pups 0–1 day. It is unclear why body weights were mixed followed from PND 4, implying that 10, 11, and 7 dams, across these 2 days when pups gain weight quickly respectively, contributed litters to the experiment.] postnatally. Also, it is not clear why there is a relatively Litter size and weight were said to be similar among large deviation between male and female body weights groups and there were no gross malformations [data not and lengths in the water control group, whereas these shown]. The authors noted 25% pup mortality in the values are similar in the saline control group. The body 3 mg/kg bw amphetamine group, compared to 15% in weight and length values for the water controls should the 1 mg/kg bw group and 6% in the control group have been discussed. For vaginal opening, the authors [percent mortality estimated from a figure]. None of the state that 15% of female offspring in the 2 mg/kg bw dead pups was abnormal on necropsy. To evaluate the d-amphetamine group showed delayed vaginal opening, possibility of lactation abnormality, 24 pups with but the authors do not define ‘‘delayed.’’ Also, females antenatal exposure to amphetamine [dose not specified] within the same litter are not independent samples. Their and 24 control pups were cross-fostered to dams of the ages at vaginal opening should be averaged to yield a opposite treatment group [number of dams and number litter mean for statistical analyses. With only three litters of pups/litter not specified]. Mortality was 25% among in the 2 mg/kg bw d-amphetamine group, it is not clear pups with antenatal amphetamine exposure and 8.3% how the authors determined that 15% had delayed among pups without antenatal amphetamine exposure. vaginal opening. The age of vaginal opening was not [The data suggest that altered maternal caregiving was presented for the water controls. The body weights in the not the primary factor in decreased neonatal survival.] water controls versus the 2 mg/kg bw d-amphetamine The authors concluded that because neonatal mortality is

Birth Defects Research (Part B) 74:471–584, 2005 DOI 10.1002/bdrb 538 GOLUB ET AL. not noted among humans with antenatal exposure to were no significant drug effects on incisor eruption or amphetamine, the effect may be peculiar to rats. vaginal patency. Eye opening was delayed a mean of 0.5 Strengths/Weaknesses: Strengths include use of rela- days in the low-dose amphetamine group. The 2.0 mg/ tively low doses (1 and 3 mg/kg given twice daily) kg bw dose had no effect on eye opening. targeted to be pharmacologically relevant and relatively Strengths/Weaknesses: This study is well conducted non-toxic and the cross-fostering design to determine with a strong design and litter-based analysis. Other whether altered maternal caregiving contributed to strengths are the low doses and the provision of some neonatal mortality. A weakness is that rats were dosed physical maturation data. Sample sizes were adequate with d-amphetamine by s.c. injection, whereas humans and proper statistical analyses were employed. A typically are exposed by oral or i.v. routes. It is unclear weakness is that rats were dosed with d-amphetamine whether dams were dosed from GD 5 to GD 19–22 (as by s.c. injection, whereas humans typically are exposed stated in the text) or from GD 5 to GD 19–20 (as stated in by oral or i.v. routes. the legend for Fig. 1). There were no data on gestation Utility (Adequacy) for CERHR Evaluation Process: length or whether altered gestation length contributed to This study is very useful in the evaluation process. increased neonatal mortality. Although some summary Martin (1975), supported by NIH, treated pregnant statements were made, data on maternal body weights, Sprague–Dawley rats with methamphetamine HCl [pur- body weight gains, clinical signs, litter sizes, and litter ity not specified] 0, 1, 3, or 5 mg/kg bw s.c. twice/day weights were not presented. Litters were randomly from GD 1 (the morning after insemination) to GD 21. adjusted to eight pups, but there is no indication that [Total daily doses were 2, 6, or 10 mg/kg bw.] There were there were attempts to balance sex within the litters (e.g., six dams/group except in the high-dose group, which four males and four females). Presenting pup body included seven dams. Litters were culled to eight [sex weights as ‘‘percent change in pup weight from control’’ not specified] on PND 7. Behavioral effects, which were is unusual and difficult to interpret, particularly given the focus of the study, are summarized in Section 3.2.3 in that the amphetamine-exposed pup body weights fluc- Table 39. Four of seven dams in the 5 mg/kg bw group tuate above and below the control mean and there is no failed to deliver and one dam in this dose group information on variance. For biogenic amines and motor delivered only one pup. One 3 mg/kg bw dam canniba- activity, there is no information about sample selection lized her litter. Maternal weight gain and litter size were (i.e., one male and one female/litter/dose group at the decreased by treatment. Gestation length was decreased appropriate ages?). There is no indication that the by a mean 0.5 days by all doses of methamphetamine. authors controlled for either litter or sex bias. Biogenic The percentage of pups with eyes open on PND 14 was amines and motor activity were only examined in decreased by methamphetamine treatment [analyzed on animals from the 0 and 3 mg/kg bw/day dose levels, a per fetus basis]. The percentages with eyes open were so dose–response relationships could not be evaluated. 71% in the control group, 38.8% in the 1 mg/kg bw Statistical analyses were not well defined. Student t-test group, 17.5% in the 3 mg/kg bw group, and 25% in the is mentioned, but this would not be appropriate for 5 mg/kg bw group (which contained only two litters). variables collected at all three dose levels. There is no [Many of the effects observed in this study have been indication that analyses were litter based. reported above for methamphetamine (e.g., decreased Utility (Adequacy) for CERHR Evaluation Process: maternal weight gain, decreased litter size, delayed This study is of limited adequacy for use in the eye opening and decreased number of litters evaluation process. at high doses). There were no gross malformations in Vorhees et al. (1985), of NCTR and the Cincinnati surviving offspring in this study. Other findings (e.g., Children’s Hospital Research Foundation, administered decreased gestation length) were novel. Pair-wise d-amphetamine sulfate [purity not specified] to preg- comparisons with the control group were not reported nant Sprague–Dawley rats at 0 (n 5 12), 0.5 (n 5 11), or 2.0 for fetal effects.] (n 5 10) mg/kg bw/day s.c. on GD 12–15 (plug 5 GD 0). Strengths/Weaknesses: A strength is the provision of There was also an uninjected control group (n 5 13). physical maturation data. Another strength is that the Litters were culled to eight (four/sex) on PND 1 and dosing interval (twice daily) was selected based on the weaned on PND 21. The primary purpose of the study half-life of amphetamines in humans, which is 14 hr. A was to evaluate offspring for neurodevelopmental end- weakness is the small group size (n 5 6–7 dams/dose points, reported in Section 3.2.3 (Table 39). Litter level) and limited statistical analysis. Statistics were used parameters were analyzed by ANOVA, except for to identify differences between saline-treated control frequency data, which were analyzed by Fisher test. animals and methamphetamine-treated animals; how- There were no significant effects of treatment on ever, post-hoc comparisons were not conducted to gestation length, dam body weight during pregnancy or determine which dose levels of methamphetamine were lactation, or offspring weight pre- or post-weaning. The significantly different from the controls. It is not clear number of pups born per litter was significantly reduced whether litter-based analyses were used when needed in the amphetamine groups at P 5 0.06 (0 mg/kg bw: (although some variables such as birth weights were 13.870.4; 0.5 mg/kg bw: 12.470.6; 2.0 mg/kg bw: analyzed by litter). Rats were dosed by s.c. injection, 12.270.5; mean7SEM [trend test by CERHR gave whereas humans typically are exposed by oral or i.v. P 5 0.03; BMD10 5 2.0 mg/kg bw/day, BMDL 1.1 mg/kg routes. The doses of methamphetamine were based on bw/day]). Post-weaning pup mortality was increased in GD 1 body weights and were not adjusted for increasing the 0 mg/kg bw group (5.2%) and the 2.0 mg/kg bw body weights during pregnancy. group (3.4%); this finding was not considered drug- Utility (Adequacy) for CERHR Evaluation Process: related. Sex ratio was decreased to 0.52 in the low-dose This study is somewhat adequate for use in the amphetamine group as a result of fewer males. There evaluation process.

Birth Defects Research (Part B) 74:471–584, 2005 DOI 10.1002/bdrb AMPHETAMINE AND METHAMPHETAMINE 539 Martin et al. (1976), supported by NIH and March of adrenals when an animal from the same group died. Dimes, treated pregnant Sprague–Dawley rats with Sixteen animals were necropsied in each group. Incidence methamphetamine HCl [purity not specified] 0or and severity of pathology and lifespan were similar in 5 mg/kg bw given s.c. twice/day from GD 1 (the methamphetamine and control groups. [The study morning after insemination) to GD 21 and PND 2–21. authors made some observations about tumors in the [CERHR assumes that the total daily dose was 10 mg/kg autopsy studies. CERHR notes that the study was not bw.] There were 25 dams treated with methampheta- designed to examine tumorigenesis and the observa- mine, 13 dams injected with saline, and 13 uninjected tions were complicated by the killing of animals at controls. Litters were culled to eight (males preferred) on different time periods, as also noted by study authors.] PND 7, and weaned on PND 21. Behavioral tests were Strengths/Weaknesses: Strengths include the ex- performed using one male/litter (summarized in Table 39 tended evaluation period and the use of standard in Section 3.2.3). Methamphetamine-exposed dams experimental methods. This is the only study to examine gained less weight during pregnancy than either control, the long-term effects of in utero and neonatal metham- and neonates were lighter on PND 7, 14, 21, and 28, and phetamine exposures. In many cases (necropsy, pathol- through the end of the experiment at 16 months of age. ogy reports, etc.), animals were coded so that evaluations Mean birth weight was numerically lower [statistical were conducted blind to treatment group. A weakness is significance could not be evaluated from the informa- the limited evaluations (e.g., mortality and some organ tion given]. Gestation length was decreased. The weights). Rats were dosed with methamphetamine at a percentage of pups with ear opening on PND 4 and single dose level (5.0 mg/kg bw/day) by s.c. injection, incisor eruption on PND 7 was not affected by treatment. whereas humans typically are exposed by oral or i.v. There was a delay in eye opening assessed on PND 14. routes. With only one dose level, dose–response relation- Strengths/Weaknesses: A strength is the extended ships cannot be evaluated. The doses of methampheta- treatment and evaluation period, with dosing interval mine were based on GD 1 body weights and were not (twice daily) selected as a realistic exposure scenario. The adjusted for increasing body weights during pregnancy. litter was the unit for statistical analyses and adequate This study was somewhat difficult to interpret because sample sizes were used. This study monitored rats into of the variety of ages at which the animals died or were adulthood. A weakness is the use of a high single dose necropsied. Despite this, no alterations in lifespan, organ administered through s.c. injection, whereas humans pathology, or tumor incidence were definitively attrib- typically are exposed by oral or i.v. routes. With only one uted to methamphetamine treatment. dose level, dose–response relationships cannot be eval- Utility (Adequacy) for CERHR Evaluation Process: uated. The doses of methamphetamine were based on This study is of limited adequacy for use in the GD 1 body weights and were not adjusted for increasing evaluation process. body weights during pregnancy. Cho et al. (1991), of the Korean National Institute of Utility (Adequacy) for CERHR Evaluation Process: Safety Research, treated pregnant Wistar rats with This study is of limited adequacy for use in the methamphetamine HCl [purity unspecified] at 0, 1, 2, evaluation process. 3, or 4.5 mg/kg bw/day s.c. on GD 7–20 (n 5 13 or Martin et al. (1979), in a study supported by NIA/NIH, 14/dose group). Caffeine 90 mg/kg bw/day was admi- examined lifespan and pathology of rats exposed to either nistered as a positive control. At birth, pups were methamphetamine or nicotine in utero. Twenty-five examined for external malformations [not further Sprague–Dawley rats were s.c. injected twice daily with discussed and results not given]. On PND 4, four 5.0 mg/kg bw methamphetamine HCl [purity not speci- pups/sex/litter were selected for evaluation of the fied] during 21 days of gestation and the last 19 days of acquisition of physical developmental landmarks and lactation. [CERHR assumes that the total daily dose was for behavioral testing (summarized in Table 39 in Section 10 mg/kg bw.] Thirteen control dams were injected with 3.2.3). [It is not stated whether the unselected pups the saline vehicle and another group of 13 dams was not were retained in the litter or culled; the Expert Panel injected. Rats receiving 3 mg/kg bw nicotine were also assumes they were culled.] The male and female examined. The dams were allowed to deliver their litters. offspring were mated after Week 14 postpartum and On PND 4, litters were culled to eight pups, maintaining evaluated for reproductive performance [evaluation as many males as possible. Statistical analyses included methods unstated]. Kruskal-Wallis rank test for lifespan and Kendall Tau Maternal body weight gain during gestation was Correlations and ANOVA for organ histopathology. suppressed at Z2 mg/kg bw methamphetamine. Only Control and treated offspring were placed into subgroups one high-dose dam delivered live pups. Male offspring for an examination of lifespan and pathology endpoints body weight gain during lactation and post-weaning was (Martin et al., 1979). Mortality was unaffected by decreased in the 3 and 4.5 mg/kg bw groups. There was methamphetamine treatment in one set of offspring no methamphetamine treatment effect on age at pinna (n 5 12/treatment group), the body weights of which detachment or abdominal hair appearance. Incisor were reduced to 80% of their normal weight over a 2- eruption and eye opening were delayed in the 3 mg/kg week period. No treatment-related differences in lifespan bw group; the magnitude of the delay was a mean of were observed in a group (n 5 15) that was fed ad lib, but 0.1–0.2 days. Testicular descent was delayed a mean of methamphetamine treatment resulted in the steepest 1.2 days in this group. Day of vaginal opening was not death curve. Methamphetamine-treated rats remained affected. Results are summarized in Table 37. [The lighter throughout most of the adult period until 16 Expert Panel notes that only eyelid opening was months of age. In an ‘‘autopsy’’ study, four rats/group affected in the 4.5 mg/kg bw group; however, because were randomly selected for necropsy with histological only one litter was produced in this dose group, the evaluation of brain, heart, lung, liver, spleen, kidney, and findings in this group are not useful.] Mating

Birth Defects Research (Part B) 74:471–584, 2005 DOI 10.1002/bdrb 540 GOLUB ET AL. Table 37 Dose Levels Affecting Physical Development in Rats Exposed Prenatally to Methamphetamine

Dam dose (mg/kg bw/day s.c.) on GD 7–20

Response description Incisor eruption Eye opening Testicular descent

NOAEL 2 2 2 LOAEL 3 3 3 BMD10 5.1 15.7 3.8 BMDL 3.1 3.2 3.3

Data from Cho et al., 1991. NOAEL, no observed adverse effect level; LOAEL, lowest observed adverse effect level; BMD10, Dose corresponding to a 10% increase in response over control, calculated from fitted dose-response curve (power model); BMDL, dose corresponding to lower bound of the 95% CI around the 10% response. BMD calculations by CERHR using EPA Benchmark Dose software v1.3.2. [Calculation of a benchmark dose in this study does not mean that regulation based on the underlying data is recommended, or even that the underlying data are suitable for regulatory decision making.] performance of male and female offspring was not were no affected litters among the four dams treated affected by treatment [data not shown]. The authors with this dose regimen on GD 13–18. A subsequent, more concluded that prenatal exposure to methamphetamine complete report of this observation was published as was associated with growth impairment and develop- Acuff-Smith et al. (1992), supported by PHS. According mental delay. [Delays in eye opening, negative geotaxis to the full report, pregnant Sprague–Dawley rats were reflex, mid-air righting reflex, incisor eruption, and treated with d,l-methamphetamine hydrochloride [purity testis descent are consistent with the decreased body not given] 0(n 5 6) or 50 (n 5 15) mg/kg bw twice daily weight gain in methamphetamine-treated dams and by s.c. injection from GD 7–12 or from GD 13–18 subsequent decreases in F1 male body weights at these (plug 5 GD 0). The 50-mg dose was expressed as the free- dose levels. Many of the effects observed in this study base equivalent. The control animals were pair-fed to the have been reported previously by Martin (1975) and methamphetamine animals. With methamphetamine Martin et al. (1976). Transient, but significant, decreases treatment, maternal weight gain was reduced, two of in motor activity and latency time for development of 15 dams died during dosing, two litters were completely avoidance responses were seen at 2.0 and 4.5 mg/kg bw/ resorbed, one litter died on the day of delivery, and 8.6% day methamphetamine; however, dose–response rela- of live-born pups died during the lactation period. There tionships for these parameters were not maintained at were no control dam or offspring deaths. There were no 3.0 mg/kg bw/day and effects were not significant at alterations in gestation length, litter size, or sex ratio. this dose level. Total distances traveled were decreased Live-born pups were raised by their dams and weaned at 2.0, 3.0, and 4.5 mg/kg bw/day.] on PND 28. Two males and two females from each litter Strengths/Weaknesses: Strengths include good physi- were killed on PND 28 and their brains removed. cal maturation data, including data on puberty. Sample Striatum and hippocampus were dissected and frozen sizes were adequate (n 5 13 or 14 dams/dose level). A for later determination of monoamines as summarized in weakness of this study is the lack of detail on mating Section 3.2.3, Table 40. The remainder of the pups were evaluation. In contrast to other studies, rats were dosed used for behavioral testing, as summarized in Table 39. once per day with methamphetamine. Rats were dosed Offspring were evaluated for eye abnormalities at the by s.c. injection, whereas humans typically are exposed time of unscheduled death, at killing on PND 28, or at the by oral or i.v. routes. Breeding was somewhat con- end of the experiment (PND 98). Eye defects were founded by pairing one male with two study females. reported in 19/114 offspring (16.7%) with prenatal The authors do not provide any information on the types methamphetamine exposure and in 0/85 control off- of statistics used to analyze their data, nor do they spring. Eight of the defects were anophthalmia or describe any methods to control for litter effects. The microphthalmia and 11 of the defects were folded retinas 4.5 mg/kg bw/day methamphetamine dose level ex- in eyes not affected by anophthalmia or microphthalmia. ceeded the MTD as there was only one live litter from the [Possible litter effects were not addressed, nor were eye 13 dams in this group. The limited group size (n 5 1 defects analyzed with regard to the degree of dam litter) at the highest dose likely accounts for the lack of toxicity. This study is presented here because it seems dose–response with some of the endpoints (e.g., maternal preliminary to the study that immediately follows. body weights, age at incisor eruption, age at testis Anophthalmia seen in offspring of dams exposed on descent, and the acquisition of avoidance response). GD 7–12 but not of dams exposed on GD 13–18 is Utility (Adequacy) for CERHR Evaluation Process: consistent with the critical period for eye development This study is of limited adequacy for use in the (Michael et al., 1987; Palmowski and Tulsi, 1987).] evaluation process. Strengths/Weaknesses: A pair-fed control was in- Vorhees and Acuff-Smith (1990) reported in a Letter cluded in the full study report (Acuff-Smith et al., to the Editor that preliminary experiments identified 1992) and proper statistical analyses were used. Blood anophthalmia associated with prenatal methampheta- samples were collected from half the dams per group on mine exposure on GD 7–12 at a maternal dose of 50 mg/ GD 12 to assess methamphetamine and amphetamine kg bw s.c. twice/day. Among the three litters treated on levels. A weakness is that rats were dosed with d, these days, 45.5–88.9% of pups had eye abnormalities l-methamphetamine at a single dose level (50 mg/kg (almost all anophthalmia). The authors stated that there bw/day) by s.c. injection, whereas humans typically are

Birth Defects Research (Part B) 74:471–584, 2005 DOI 10.1002/bdrb AMPHETAMINE AND METHAMPHETAMINE 541 exposed by oral or i.v. routes. With only one dose level, remaining offspring (discussed in Section 3.2.3, Table 39). dose–response relationships cannot be evaluated. The The adult offspring and culled pups from the 20 mg/kg 50 mg/kg dose exceeded the MTD as evidenced by the bw and both control groups were evaluated for eye loss of five of 15 litters (two maternal deaths, two totally defects. Data were evaluated using fixed-effect factorial resorbed litters, and one total litter loss that occurred analysis of variance with post-hoc Duncan multiple postnatally). The methamphetamine blood levels were range testing. A satellite group of pregnant animals at bimodally distributed and are discussed in the context of the high dose was evaluated for serum methampheta- litter outcome (i.e., footnotes in Table 2 of the study). mine and amphetamine concentrations (discussed in Although the levels on GD 12 may be relevant to litter Section 2.1.2.2). resorptions, this may not be the most critical gestation All dams had a decrease in body weight compared to stage for methamphetamine-induced fetal loss. Further- ad lib-fed and watered control animals during the more, there is no indication that methamphetamine treatment period. Maternal mortality was increased and levels on GD 12 are representative for other stages of gestation length increased in the 15 and 20 mg/kg bw gestation. Inclusion of ad lib controls would have been groups treated on GD 13–18. Gestation length was also useful, particularly for neurobehavioral assessments. increased in the 10 mg/kg bw group treated on GD 13– Utility (Adequacy) for CERHR Evaluation Process: 18. Post-weaning offspring body weights showed no The full study (Acuff-Smith et al., 1992) is of limited consistent difference by group after GD 7–12 exposure, usefulness in the evaluation process. but showed continued decreases in the two high-dose Acuff-Smith et al. (1996), supported by NIH, treated groups after GD 13–18 exposure. Litter size was pregnant Sprague–Dawley rats with d-methampheta- decreased at all dose levels of methamphetamine given mine (free base [purity not given]) 0, 5, 10, 15, or on GD 13–18 in comparison with both control groups. On 20 mg/kg bw twice daily by s.c. injection from GD 7– a per fetus basis, stillbirth and postnatal mortality (PND 12 or from GD 13–18 (plug 5 GD 0). [Following the 1–3) were increased in 20 mg/kg bw group after authors’ practice, the treatment groups are identified exposure on GD 7–12 and were increased in the 10, 15, here by the amount of methamphetamine in each s.c. and 20 mg/kg bw groups after treatment on GD 13–18. dose rather than by the daily dose, which was twice the The proportion of offspring with anophthalmia or individual s.c. dose.] Rats in the 15 and 20 mg/kg bw microphthalmia was numerically higher on a per fetus groups and one 0 mg/kg bw group were fed and watered basis after GD 7–12 exposure in the 15 and 20 mg/kg bw ad lib. Another 0 mg/kg bw group and the 5 and 10 mg/ groups, but there were no statistically significant kg bw groups were pair-fed and watered using the rats in differences. The frequency of folded retina was increased the 15 mg/kg bw group as a reference. There were 14–16 in the 20 mg/kg bw groups after exposure on GD 7–12 litters/group, except for 12 litters in the ad lib control and 13–18, but the frequency in the earlier exposure group. Litters were standardized on PND 3 to eight pups, period was similar to that in the pair-fed and watered balanced for sex. Litters were weighed weekly and control. Neurochemical data showed no significant weaned on PND 28. One male and female from eight differences in the 20 mg/kg bw group animals compared litters each in the 20 mg/kg bw group and both control to the pair-fed and watered controls, although there were groups were decapitated on PND 70. Serial sections several significant differences compared to the ad lib 2 mm apart were taken from the brain and specific brain control group. [The Expert Panel notes that the authors regions dissected (medial frontal cortex, caudate-puta- conclude that dopamine was reduced in the caudate men, nucleus accumbens, hippocampus). Sections were nucleus in female offspring exposed on GD 13–18, but frozen until assayed by HPLC for dopamine, serotonin, that the data table does not support this conclusion.] 3,4-dihydroxyphenylacetic acid, 5-hydroxyindole-3- The developmental endpoints for which data were acetic acid, and homovanillic acid (see Section 3.2.3, suitable for benchmark dose analysis are shown in Table 40). Behavioral testing was performed using the Table 38 [benchmark dose calculations performed by

Table 38 Benchmark Dose Estimates from Acuff-Smith et al. (1996)

Benchmark dose (mg/kg bw/day)

Treatment period (GD) Endpoint BMD10 BMDL

7–12 Stillborn per pups born 91 58 Postnatal mortality per live-born pup (PND 1–3) 48 40 Anophthalmia/microphthalmia per live pup 48 42 13–18 Dam mortality 27 20 Gestation length 112 75 Litter size 38 20 Stillborn per pups born 36 31 Postnatal mortality per live-born pup (PND 1–3) 53 40 Maternal body weight, GD 19 (estimated from graph) 91 56 Pair-fed and watered group was used to represent 0 mg/kg bw/day exposure. Data are expressed as total daily dose, which represents twice the amount of methamphetamine in each of the twice-daily subcutaneous injections. BMD10, exposure level associated with a 10% response, estimated from a mathematical dose-response model; BMDL, exposure level associated with the lower bound of the 95% CI around the BMD10. Calculated using EPA software v1.3.2.

Birth Defects Research (Part B) 74:471–584, 2005 DOI 10.1002/bdrb 542 GOLUB ET AL. CERHR using U.S. EPA benchmark dose software]. The among females. The authors postulated that the adrenal authors noted that the 15 and 20 mg/kg bw groups gland may have been sensitive to small changes in ACTH experienced significant maternal toxicity and associated (resulting in large changes in corticosterone in the effects on offspring viability. Their principle conclusions absence of statistically detectable changes in ACTH) or were directed to the behavioral test results, summarized the corticosterone levels may have suppressed the ACTH in Table 39. plasma levels. The authors concluded that changes in the Strengths/Weaknesses: This is a well conducted and hypothalamic-pituitary-adrenal axis may represent a comprehensive study. Strengths of this study include use mechanism for the long-term changes in spatial learning of multiple doses within the range of realistic human and memory associated with neonatal methampheta- exposures, a design that evaluated sensitive windows for mine exposure. specific effects (e.g., eye development and dopamine and Strengths/Weaknesses: A strength is that the study serotonin neurotransmitters in the brain), use of pair-fed included two control groups: a non-injected handheld and watered controls, and the uniqueness of the folded- group and a saline-injected control. The litter was the retina data. Sample sizes (12 in ad lib controls, 14–16 unit of analyses for statistical purposes. A weakness of dams in other groups) were adequate. Pups from each this study is the small number of endpoints evaluated. litter were assigned to the various test methods and Rats were dosed with d-methamphetamine at a single proper statistical analyses were employed. Pharmacoki- dose level by s.c. injection, whereas humans typically are netics (time course at 0.5, 0.75, 1, 2, and 8 hr post- exposed by oral or i.v. routes. With only one dose level, treatment and peak concentrations) of d-methampheta- dose–response relationships cannot be evaluated. Some mine and amphetamine were determined. To facilitate cross-fostering of pups was performed, making it data interpretation, the 5 and 10 mg/kg bw methamphe- difficult to control for litter effects. Elevations in ACTH tamine groups were matched to the pair-fed controls. do not always mirror increases in corticosterone (e.g., Aside from some developmental parameters, a compre- PND 20 at 60 min). Authors targeted postnatal hippo- hensive assessment of neurobehavioral endpoints and a campal development in rats as a period comparable to pharmacological challenge with methamphetamine or hippocampal development during the third trimester in fluoxetine were conducted. The data support the authors’ humans; however, there is no evidence that the pharma- conclusions. Weaknesses are that rats were dosed by s.c. cokinetics of third trimester maternal/fetal exposures are injection, and only the control and high-dose metham- the same as s.c. exposures in rat pups at 2-hr intervals. phetamine groups were tested on the water maze. Furthermore, the authors adopted the procedure of Utility (Adequacy) for CERHR Evaluation Process: dosing rats 4 Â /day instead of 2 Â /day to induce This study is adequate for use in the evaluation process. greater neurotoxicity for the same overall dose; however, Williams et al. (2000), supported by NIH, investigated the critical element is the exposure scenario that most a possible role for the hypothalamic-pituitary-adrenal mimics the human situation. axis in mediating the developmental neurotoxicity of Utility (Adequacy) for CERHR Evaluation Process: methamphetamine in neonatal Sprague–Dawley rats. This study is of limited adequacy for use in the Animals were derived from litters standardized on evaluation process. PND 1 to 10 pups, with 4–6 pups of each sex. One or Courtney and Valerio (1968) examined developmental two pups/litter were fostered if necessary to give at least toxicity of methamphetamine HCl in Macaca mulatta four pups/sex/litter. Treatments consisted of four daily monkeys in a study supported by the A.H. Robins Co. s.c. injections of 0 or 15 mg/kg bw methamphetamine Additional drugs were examined and sponsored by other (free base [purity not specified]) on PND 11, 11–15, or companies. Monkeys were mated during Days 10–14 of 11–20, ages during and after the stress hyporesponsive the menstrual cycle, as determined by daily vaginal period, which lasts from PND 0–14. A handled but swabs. When pregnancy was obtained, Day 12 of the uninjected control was also used. On the last day of each menstrual cycle was considered GD 1. Five monkeys treatment period, one animal/sex/litter was decapitated were given 0.5 mg/kg bw/day methamphetamine HCl 15 min before or 15, 30, or 60 min after the fourth injection [purity not specified] from implantation to term (from of the day, and trunk blood was collected for measure- about GD 11–167). [Medications were administered in ment of corticosterone and adrenocorticotropic hormone honey, a sugar cube, or by gavage, but the particular (ACTH). Brains were dissected and bilateral hippocam- method used for methamphetamine was not specified. pus weights were recorded. Ten litters were used for each There was no mention of control colony treatment.] The of the nine conditions (methamphetamine, saline-control, monkeys were allowed to give birth, although some and uninjected control for each of the three treatment required cesarean section [not specified by treatment]. periods: PND 11, PND 11–15, or PND 11–20). Metham- Upon birth, infants were examined for general state of phetamine-treated pups were lighter than either control health, growth, and development. beginning on PND 13. Relative but not absolute With the exception of one monkey that lost 0.4 kg, all hippocampus weight was increased by methampheta- pregnant monkeys gained weight during the study. mine treatment [statistical analysis by CERHR showed Periodic hematologic evaluations during pregnancy a difference in relative hippocampus weight only on showed all values to be within normal ranges. Four PND 20]. Plasma corticosterone was increased by normal births and one autolyzed stillbirth occurred. methamphetamine treatment at all ages, although the Noting a 12% rate of abortion and stillbirth in their PND 15 and 20 pups did not show this increase until control colony, the authors concluded that the stillbirth 30 min after the injection. ACTH was also increased by may not have been related to methamphetamine treat- methamphetamine treatment at all ages; however, there ment. Gestation lengths and birth weights were within were several time points on PND 15 and 20 when ACTH normal ranges. Growth and development was normal in and corticosterone results were discordant, particularly the four surviving infants. At 2–6 months of age, infants

Birth Defects Research (Part B) 74:471–584, 2005 DOI 10.1002/bdrb AMPHETAMINE AND METHAMPHETAMINE 543 were killed, and macroscopic and microscopic examina- vehicle-treated controls. The decrement in pup weight tions showed normal tissues and organs. [Methods were persisted until at least PND 42 and in some cases until not specified. With the exception of birth weight, no PND 70. Pup mortality during the treatment period was data were presented. There was no evidence of also increased by neonatal methamphetamine treatment teratogenicity in the surviving offspring.] in many of the studies (Vorhees et al., 1994a,b, 1996, 1998, Strengths/Weaknesses: Weaknesses include small 1999, 2000). Treatment of neonates was associated with sample sizes and no mention of a concurrent control increased reactivity on acoustic startle testing and with group. Monkeys were given a single dose level of deficits in associative processes and memory when methamphetamine, so dose–response relationships can- animals were tested as adults. The lowest effective dose not be evaluated. One monkey’s offspring was stillborn, reported for neonatal treatment with d-methampheta- but it cannot be determine whether this is related to mine was 30 mg/kg bw/day (Vorhees et al., 1999). [The methamphetamine treatment or part of the spontaneous Expert Panel notes that studies involving direct treat- background of abortions and stillbirths in this laboratory ment of neonates typically used a single methamphe- (B12%). The dose level used in this study, 0.5 mg/kg tamine dose level and a vehicle-treated control. No bw/day, is low compared to estimated human attempt was made to ascertain a NOAEL in these exposures. studies.] The memory deficits in adult rats after Utility (Adequacy) for CERHR Evaluation Process: treatment of PND 11–20 pups involve spatial tasks such This study is of limited usefulness for the CERHR as locating a submerged escape platform in a swimming process. tank (Vorhees et al., 1994a, 1998, 1999, 2000; Acuff-Smith 3.2.3 Developmental neurotoxicity. Ampheta- et al., 1996; Williams et al., 2003c). The deficits have been mine developmental rat studies with neurobehavioral shown to be due to impairments in reference (long-term) outcomes are summarized in Table 39. Studies in which memory and not to general motor or cognitive impair- treatments were given to the dam during gestation ment or to deficits in working (short-term) memory, and involved s.c. administration of amphetamine or metham- have been attributed to alterations in hippocampus phetamine on GD 12–15 (Clark et al., 1970; Adams et al., development that would be expected to correspond to 1982; Buelke-Sam et al., 1985; Holson et al., 1985; Vorhees, developmental events in the third trimester of human 1985), GD 7–12 (Acuff-Smith et al., 1992, 1996), GD 13–18 pregnancy. Treatment of young rats with amphetamine (Acuff-Smith et al., 1996), or during all or most of the or methamphetamine is also associated with behavioral gestation period (Nasello et al., 1974; Martin, 1975; sensitization to subsequent challenge with amphetamine Hitzemann et al., 1976; Nasello and Ramirez, 1978; or methamphetamine. This sensitization can persist for Satinder and Sterling, 1983; Sato and Fujiwara, 1986; months after the last treatment (reviewed in Section Cho et al., 1991; Tan, 2003). In one study (Martin et al., 2.5.3). 1983), the treatment period during gestation was not There were three reports in which methamphetamine specified, and the study is included only for complete- (Martin et al., 1976; Martin and Martin, 1981) or d- ness. The treatments in many of the studies resulted in amphetamine (Monder, 1981) were given during gesta- significant decreases in maternal weight gain, associated tion and continued through the lactation period. The two with a decrease in maternal feed intake. One study methamphetamine reports, which present different end- reported loss of maternal weight during the treatment points from the same experiment, involved s.c. treatment period among methamphetamine-treated animals and of dams. The amphetamine study used administration in pair-fed controls (Acuff-Smith et al., 1996). Some studies drinking water [probably resulting in direct treatment reported adverse effects of treatment on pup birth of pups during the last week of the lactation period]. weight (Monder, 1981) or viability (Hitzemann et al., Adverse effects on dam and pup body weight were 1976; Acuff-Smith et al., 1992, 1996). One study was noted in these studies, and both studies showed reviewed but not included in the table due to very low developmental delay and increased motor activity in delivery rates among the animals, including the controls offspring, consistent with the studies that used only (Seliger, 1973). gestational exposures. Effective doses were 2 mg/kg Results of behavioral studies after prenatal exposures bw/day d-amphetamine [the lower of two dose levels generally suggest developmental delay (Martin, 1975; used] and 10 mg/kg bw/day methamphetamine [the Vorhees, 1985; Cho et al., 1991; Acuff-Smith et al., 1992) only dose level used]. and increased motor activity (Hitzemann et al., 1976; Several studies evaluated persistent anatomical or Nasello and Ramirez, 1978; Adams et al., 1982; Acuff- biochemical changes in the brains of rodents after Smith et al., 1992, 1996). One study (Acuff-Smith et al., prenatal or juvenile exposure to amphetamine or 1996) found prenatal exposure to methamphetamine to methamphetamine and those studies are summarized be associated with impairment of learning and memory in Table 40. With the exception of a drinking water in adult offspring. The lowest effect level reported with exposure study (Tonge, 1973), exposures occurred gestational treatment of the dam was 0.5 mg/kg bw/day through parenteral routes, mainly sc. Chemical purity for both d-amphetamine (Adams et al., 1982) and for d,l- was not reported in any of the studies. amphetamine (Nasello et al., 1974; Nasello and Ramirez, As discussed in Section 2.5.3, treatment of young 1978). (preweanling) animals is associated with persistent Studies involving direct treatment of rat neonates have changes in behavior, with little if any anatomic alteration used methamphetamine administered s.c. on PND 1–10 in the brain. Other studies summarized in Table 40 (Vorhees et al., 1994a,b, 1996) or PND 11–20 (Vorhees reported a transient increase in prefrontal cortex neuro- et al., 1994a,b, 1998, 1999, 2000; Williams et al., 2002, nal density in 14-day-old rats exposed to 10 mg/kg bw/ 2003a–c). Both administration periods have been asso- day amphetamine on GD 8–22 (Tavares et al., 1996) and ciated with a decrease in pup weight compared to reduced hippocampal volume formation in 30-day-old

Birth Defects Research (Part B) 74:471–584, 2005 DOI 10.1002/bdrb 4 OU TAL. ET GOLUB 544 Table 39 Developmental Neurotoxicity Testing of Amphetamines in Rats

Strain and Dose Regimen General Toxicity and Neurobehavioral Findings Author Conclusions Comments Reference

Gestational Exposure of Dams Long-Evans hooded. Pregnant animals Body weight was said not to vary by treatment ‘‘The results indicate modified Utility limited by lack of ex- Bell et al., 1965 (n 5 20) treated once i.p. with epi- [possibly referring to offspring body weight on emotionality as a function perimental detail, use of nephrine, 3.0 mg/kg bw d-ampheta- PND 45]. of the prenatal obsolete design, and sub- mine sulfate [purity not given], or Open-field activity (PND 45): Amphetamine-exposed treatmentsy’’ optimal exposure route. saline on either one of GD 6–9 or animals were said to enter fewer squares when 12–15 [number of animals in each tested before and after a 48-hr immobilization treatment and time group not given]. stress. [Date presented as mean number of Dams and offspring left undisturbed squares without indication of variance or n and until PND 45. without statistical analysis.] Sprague–Dawley. d-Amphetamine sul- Dam and litter parameters not given. Offspring body ‘‘Amphetamines may have a Utility limited by lack of ex- Clark et al., 1970 fate [purity not specified] 0 or 1 mg/ weight not affected by treatment at any age. different influence on be- perimental detail, use of kg bw/day s.c. on GD 12–15; n 5 4/ Locomotor activity (PND 13, 15, 18, 21, 46, 60): havior in developing or- obsolete design, inadequate dose group. Litters culled at birth to 3 Amphetamine associated with decreased activity ganisms than in adults.’’ sample size, and subopti- of each sex, weaned on PND 21. on PND 21. mal exposure route. Maze learning (PND 151): Amphetamine-exposed offspring reached goal more quickly than controls. Operant conditioning (PND 371): No effect of treatment. [No control for multiple comparison/repeated it eet eerh(atB 44154 2005 74:471–584, B) (Part Research Defects Birth measures.] Strain not given. d,l-Amphetamine given No information on maternal or pup weights. No Adrenergic system might be Utility limited by lack of ex- Nasello et al., s.c. once/daily at 0 or 0.5 mg/kg bw/ alteration in day of eye opening or vaginal opening. affected by amphetamine; perimental detail, use of 1974 day from the day of sperm positivity Conditioned avoidance (PND 45 and 90): Amphetamine- full development of this obsolete design, and sub- until delivery. Pups redistributed to exposed animals had better acquisition and reten- system not affected by am- optimal exposure route. produce litters of 8, equal sex when tion than controls at PND 90. phetamine until PND 90. possible. Offspring weaned at ‘‘one No treatment effect at PND 45. month.’’ Seizure activity to direct hippocampal application of potassium (PND 901): Amphetamine-treated animals had seizure activity after a shorter duration of exposure to potassium. Sprague–Dawley. Methamphetamine 4 of 7 dams in 5 mg/kg bw group failed to deliver and ‘‘The [5 mg/kg bw] dose level Utility limited by lack of ex- Martin, 1975 HCl [purity not specified] 0, 1, 3, or 1 dam delivered only 1 pup. One 3-mg/kg bw dam was somewhat effective in perimental detail, use of 5 mg/kg bw was given s.c. twice/day cannibalized her litter. Maternal weight gain and increasing shuttle-box per- obsolete design, and sub- from GD 1 (the morning after inse- litter size were decreased by treatment. Gestation formance. These results are optimal exposure route. mination) to GD 21. There were 6 length was decreased by a mean 0.5 days by all only tentative in that the dams/group except 7 rats in the doses of methamphetamine. numbers were small and

O 10.1002/bdrb DOI high-dose group. Litters culled to 8 Eye opening (PND 14): Delayed in methamphetamine- the study not replicated.’’ [sex not specified] on PND 7, exposed groups. weaned on PND 21. Conditional avoidance response (PND 100–120):No statistically significant treatment effect; mean avoidance response number was increased in the 5-mg/kg bw group at p 5 0.09. Osborne-Mendell. d-Amphetamine No information given on clinical signs, feed con- Intrauterine exposure to am- Utility limited by lack of ex- Seliger, 1975 [purity not specified] s.c. at 0, 5, or sumption, weight gain, litter size, viability, or other phetamine was probably perimental detail, use of 10 mg/kg bw/day on GD 5–9 or 12– possible indicators of general toxicity. stressful to the offspring, it eet eerh(atB 44154 2005 74:471–584, B) (Part Research Defects Birth 16 (GD 0 5 a 24-hr cohabitation peri- Water wading (PND 74): There was no treatment effect and this early exposure to obsolete design, and sub- od). Neonates weaned on PND 28 on the number of fecal boli dropped when animals stress resulted in habitua- optimal exposure route. [culling not mentioned]. Testing was were placed in 3 cm deep water. tion, making these animals carried out on 1 pup/litter/replicate Audiogenic seizure (PND 84): There was no treatment less emotional in later life. [the number of dams or pups is not effect on the number of defecations in the seizure indicated, except to say that 165 box. ‘‘Wild running’’ and seizures in response to dams were initially mated]. bell ringing were increased among animals whose mothers were given placebo during early pregnancy, compared to the other groups. Sprague–Dawley. d-Amphetamine sul- Increased mortality among amphetamine-exposed There was a time-dependent Utility limited by lack of Hitzemann fate [purity not specified] 0, 1, or pups. increase in motor activity in experimental detail, use of et al., 1976 3 mg/kg bw s.c. twice daily from GD Activity (PND 36 and 84): Increased by 3 mg/kg bw amphetamine-exposed ani- obsolete design, and sub- 5 through parturition on GD 19–22. amphetamine exposure during intervals in the first mals unfamiliar with the optimal exposure route. Litters culled randomly to 8 at birth. 30 min in the activity cage. No overall alteration testing cage. The signifi- Weaning not specified. [Number of over 2-hr period. cance of this effect is not METHAMPHETAMINE AND AMPHETAMINE dams not given.] Brain amines (PND 84): Norepinephrine was decreased clear. in the diencephalon and brainstem and dopamine

O 10.1002/bdrb DOI was decreased in the brainstem [The study seems to use per fetus analysis; methods were not detailed.] Wistar. d,l-Amphetamine [purity not No assessments of general toxicity were given. Prenatal amphetamine treat- The experimental conditions Nasello and specified] 0 or 0.5 mg/kg bw/day s.c. Open field testing (PND 65–85): Motor activity of males ment increased the excit- were better controlled than Ramirez, 1978 from first day of pregnancy until was increased by amphetamine treatment. ability of adult offspring. in previous studies but the parturition. Litters culled to 8 (equal Lashley III mazed (PND 90–100): Amphetamine-treated This increased excitability single s.c. dose is a weak- sexes when possible) on day of animals made more errors than did controls. may be due to an increased ness and the hypothesis is delivery and weaned at 1 month of release of potassium by speculative. age. Number of dams not given. stimuli. Eleven offspring of each sex tested per treatment except 10 males in the amphetamine group. Sprague–Dawley.d-Amphetamine sul- Control and high-dose group did not differ in dam The animals failing to learn to The use of multiple doses is a Adams et al., fate 0, 0.5, 1.0, or 2.0 mg/kg bw/day weight gain, implantation sites/litter, live fetuses/ escape demonstrated pro- strength, as is the con- 1982 (concentrations verified by GC) s.c. litter, resorptions, sex ratio, or fetal weight. There longed floating, suggesting trolled exposure and the on GD 12–15 (n 5 9/dose group). were no differences across all dose groups in an immobilization re- control for toxicity. Litters culled to 8 (3–5/sex) on PND postnatal pup body weights. sponse. The mechanism of 1, weaned on PND 21. Y-water maze (PND 38–41): A smaller proportion of prenatal amphetamine in- amphetamine-exposed than control animals failed creasing responsiveness to to learn to escape; there was no difference by postnatal amphetamine is amphetamine dose. There were no treatment- not known. related differences in retention or reversal. Spontaneous and amphetamine-induced activity (PND 45– 50): No statistical difference in baseline activity. Amphetamine-stimulated activity was accentuated in rats previously exposed to amphetamine in utero. Albino, strain not specified. Ampheta- Amphetamine-exposed offspring were heavier than The changes in locomotor ac- This study is limited by Bigl et al., 1982 mine [purity, salt, enantiomer not controls on PND 15 and 21. There was no influence tivity may reflect altered inadequate experimental specified] 0 or 2 mg/kg bw/day s.c. of treatment on litter size. activity of the GABA detail and by use of a single from GD 7 until delivery, n 5 5/dose Open field activity (PND 8, 15, 21): Amphetamine system. s.c. dose level. 545 group. Litters redistributed to exposure was associated with increased locomotor 4 OU TAL. ET GOLUB 546 Table 39 (Continued)

Strain and Dose Regimen General Toxicity and Neurobehavioral Findings Author Conclusions Comments Reference

equalize number of offspring [day of activity on PND 15 and decreased locomotor redistribution, final litter size, and activity on PND 21. sex ratio not given]. Sprague–Dawley. Methamphetamine Preference for saccharin-containing solution over tap Treatment with methamphe- There is inadequate detail in Martin et al., HCl [purity not specified] 0 or 5 mg/ water was lost at an earlier age when tested from 6 tamine during pregnancy this study. The saccharin 1983 kg bw s.c. twice/day at unspecified to 36 months of age; however, differences from alters patterns of taste pre- preference test is not indi- time of gestation. Culling and wean- placebo group were only apparent during some of ference, but ‘‘effects are cative of developmental ing of litters not mentioned. Animals the trials. neither simple nor readily neurotoxicity. maintained at 80% of free-feeding predictable.’’ weight from PND 72. Roman high- (RHA) and low-avoidance Amphetamine-treated dams of both genetic lines These results are consistent The use of genetically selected Satinder and (RLA) rats treated with d-ampheta- weaned 6/8 litters born. RHA saline group weaned with amphetamine aug- rats is not helpful. There is Sterling, 1983 mine [purity not specified] 0or 4/7 litters; RLA saline group weaned 4/4 litters. mentation of basal arousal inadequate detail in the 3 mg/kg bw/day s.c. (1 mg/kg bw in Litter size at birth and weaning seemed reduced in level in RHA rats and de- study. the morning and 2 mg/kg bw in the the RHA saline group, but variances were not creased basal arousal level afternoon), GD 7–20. Culling not provided. Pup birth and weaning weight also may in RLA rats. mentioned. Weaned on PND 28. RHA have been reduced in this group, but again, amphetamine n 5 8 litters, saline variances were not provided. The authors stated n 5 7 litters; RLA amphetamine n 5 8 that there were no statistical differences ‘‘in any litters, saline n 5 4 litters. aspect of the data.’’ it eet eerh(atB 44154 2005 74:471–584, B) (Part Research Defects Birth Avoidance training (PND 701): RHA animals showed more improvement over time if exposed prenatally to amphetamine; RLA animals showed better performance over time if not exposed to amphetamine. Sprague–Dawley. d-Amphetamine sul- Pup mortality slightly increased in vehicle control and Little evidence of ampheta- The use of multiple doses and Vorhees, 1985 fate [purity not specified] 0(n 5 12), 2.0 mg/kg bw groups, not considered drug-related. mine-induced behavioral control for general toxicity 0.5 (n 5 11), or 2.0 (n 5 10) mg/kg Decrease in litter size in amphetamine groups was teratogenesis was found. are strengths of this study. bw/day s.c. on GD 12–15. There was significant at po0.1. The few decrements in also an uninjected control (n 5 13). Surface-righting (PND 3–12): No treatment effects. swimming performance Litters culled to 8 (4/sex) on PND 1, Negative geotaxis (PND 6, 8, 10, 12): No treatment probably represent transi- weaned on PND 21. effects. ent delays in development. Pivoting locomotion (PND 7, 9, 11): No treatment effects. Auditory startle onset: No treatment effects. Olfactory orientation (PND 9, 11, 13): No treatment effects.

O 10.1002/bdrb DOI Swimming ontogeny (PND 6–24): Amphetamine groups had impaired performance on PND 6. On PND 10, the 2.0 mg/kg bw group had impaired performance. Figure-8 activity (PND 15–17; 38–43): no treatment effects. Straight channel swimming and Biel water mazeb (PND 48–53): No treatment effects. it eet eerh(atB 44154 2005 74:471–584, B) (Part Research Defects Birth Sprague–Dawley. d-Amphetamine sul- Maternal weight gain was decreased during the The purpose of the Colla- The overview nature of the Adams et al., fate [purity not specified] 0, 0.5, or dosing period. Offspring body weight affected at borative Behavioral Tera- overview reports (without 1985a; Adams 2.0 mg/kg bw/day s.c. GD 12–15, two labs on PND 1. tology Study was not to detailed experimental data) et al., 1985b; plus an untreated control. There were Eye opening (PND 12–16): No treatment effect. generate more toxicologic reduces their utility. The Buelke-Sam 7 dams/dose group/replicate to gen- Incisor eruption (PND 7–13): No treatment effect. data on d-amphetamine, report that includes the et al., 1985; erate at least 4 litters/dose group/ Testes descent (PND 21–26): No treatment effect. but to evaluate the use of a results of the study is use- Kimmel and replicate. There were 4 replicates per Vaginal opening (PND 30–40): No treatment effect. standardized, partly auto- ful. The study was very Buelke-Sam, lab and 5 labs, so at least 80 litters per Negative geotaxis (PND 7–10): No treatment effect. mated test scheme across well controlled with multi- 1985; Kimmel dose group would have been gener- Olfactory discrimination (PND 9–11): No treatment different laboratories. ple behavioral endpoints. It et al., 1985; ated. Culled on PND 1 to 8/litter effect except in one lab. There were few adverse is a weakness that only two Nelson et al., (3–5/sex). Weaned on PND 21. Auditory startle habituation (PND 1–19, 57–58):No behavioral effects at the dose levels of amphetamine 1985 treatment effect except in males in one lab. low amphetamine dose le- were used. 1-Hr Activity (figure-8 maze; PND 21, 60): No treatment vels used in these studies. effect. 23-Hr activity (figure-8 maze; PND100–108): No treat- METHAMPHETAMINE AND AMPHETAMINE ment effect. Activity after amphetamine challenge (PND 120–131):No

O 10.1002/bdrb DOI treatment effect. Operant visual discrimination task (PND 75–89):No treatment effect. Sprague–Dawley. d-Amphetamine sul- Transient decrease in maternal weight gain during There was no evidence of an The use of multiple doses with Holson et al., fate [purity not specified]. Experi- dosing in 2.0 and 3.0 mg/kg bw/day groups. No effect of prenatal ampheta- several endpoints and 1985 ment 1: 0, 0.5, 2.0, or 3.0 mg/kg bw/ difference in pregnancy weight gain or offspring mine exposure on long- male/female data are day s.c. GD 12–15 (n 5 12 dams per weights. term activity measures, but strengths. dose group). Experiment 2: 0 (n 5 9) Auditory startle habituation (Experiment 1: PND 47–48, there was ‘‘a suggestion of or 3.0 (n 5 12) mg/kg bw/day. 57–58, 120–121; Experiments 2 and 3: PND 19): an effect at higher doses on Experiment 3: 0 (n 5 21) or 3.0 (n 5 23) Increased startle in females of the 3.0 mg/kg bw/ short-term reactivity to no- mg/kg bw/day. Litters culled to 8 day group on PND 47–48 and 120–121. Response in vel fields.’’ Behavioral ef- (3–5/sex) on PND 1; weaned on PND males suppressed by 2.0 but not 3.0 mg/kg bw/day fects of treatment occur at 21. Two-day response testing in- treatment when tested on PND 57–58. 3 mg/kg bw/day, but are cluded administration of metham- Figure 8 maze activity (Experiment 1: PND 47–48, 120– sex-specific and without a phetamine challenge on the second 121, 135–137; Experiment 2: PND 47–48): No treat- discernible pattern. The in- day. ment effects except possible inconsistent effects consistency in results with short-term exposure to the maze (novel among studies may not be stimulus effects). accidental: the behavioral Activity in single photo-cell chamber (Experiment 1: PND test results may depend on 47–48, 120–121, 133): No treatment effects. the psychological state of Emergence (Experiment 1: PND 128): Amphetamine- the dam during dosing and exposed females in top two dose groups were more the offspring at testing. likely to emerge. Male effects were said to follow similar pattern although not statistically significant. Open-field activity (Experiment 2: PND 70–72):No treatment effects. Intake of sweet solution (Experiment 2: PND 85–105, 215– 217): Preference for sweet solution was increased in 3.0 mg/kg bw/day group. Sprague–Dawley. Methamphetamine General toxicity information not provided. Prenatal methamphetamine The attempt to correlate be- Sato and [purity unspecified] 0 or 2 mg/kg bw Total and vertical activity (PND 5–34): total motor exposure impairs reactivity havior and neurochemistry Fujiwara, ‘‘for 3 weeks during pregnancy.’’ activity decreased and vertical activity increased by to environmental stimuli are strengths. The lack of 1986 547 [interval not stated; presumed daily]. prenatal methamphetamine treatment. but does not sensitize the experimental detail and the 4 OU TAL. ET GOLUB 548 Table 39 (Continued)

Strain and Dose Regimen General Toxicity and Neurobehavioral Findings Author Conclusions Comments Reference Control dams reared 8 methamphe- Activity in response to methamphetamine challenge (5th offspring to postnatal use of a single s.c. dose tamine-exposed and 8 vehicle-ex- week of life): No effect of treatment. methamphetamine level are weaknesses. posed pups [number of litters not Activity in response to sound (3–4 weeks after birth): exposure. specified]. Weaned PND 21. Activity in controls was suppressed after noise. Vertical activity in prenatally methamphetamine- exposed offspring was stimulated by noise during the dark period. Brain catecholamines: frontal cortex dopamine decreased in methamphetamine-exposed pups on PND 35 but not PND 21. Decreased spiperone binding in frontal cortex (reflecting decreased serotonin receptors) on PND 35 in methamphetamine group. Wistar. Methamphetamine HCl [purity Maternal body weight gain during gestation was ‘‘[M]ethamphetamine might The use of multiple dose Cho et al., 1991 unspecified] 0, 1, 2, 3, or 4.5 mg/kg suppressed at 2 mg/kg bw methamphetamine and induce some behavioral levels and several beha- bw/day s.c. on GD 7–20 (n 5 13 or higher. Only 1 high-dose dam delivered live pups. teratogenicity in rats.’’ vioral endpoints are 14/dose group). On PND 4, 4 pups/ Male offspring body weight gain during lactation There was an increase in strengths. The small num- sex/litter were selected for behavioral and post-weaning was decreased in the 3 and conditioned avoidance re- ber of animals is a testing. [Culling and time of wean- 4.5 mg/kg bw groups. sponse, especially at the weakness. ing not specified.] Surface righting (from PND 4): No treatment effect. 3 mg/kg bw dose. Cliff avoidance (from PND 4): No treatment effect. it eet eerh(atB 44154 2005 74:471–584, B) (Part Research Defects Birth Negative geotaxis (PND 6–12): Delayed in the 3 and 4.5 mg/kg bw groups. Mid-air righting (from PND 12): Delayed in the 3 and 4.5 mg/kg bw groups. Spontaneous motor activity (from 3 weeks of age): 3–5-week-old methamphetamine-exposed pups (Z2 mg/kg bw/day) had a decrease in distance traveled. Conditioned avoidance (from 7 weeks of age):No consistent effects of treatment, although authors state that all methamphetamine-exposed offspring ‘‘tended to have higher avoidance response rates.’’ Conditional avoidance response was increased on day 2 in the 2 and 3 mg/kg bw groups but not on days 3–5. Sprague–Dawley. d,l-Methamphetamine With methamphetamine treatment, maternal weight Offspring exhibit long-term This study was well controlled Acuff-Smith HCl [purity not given] 0(n 5 6) or 50 gain was reduced, 2/15 dams died during dosing, 2 consequences of prenatal and included multiple et al., 1992

O 10.1002/bdrb DOI (n 5 15) mg/kg bw (free base litters were completely resorbed, 1 litter died on the exposure to high doses of endpoints and male/fe- equivalent) twice daily by s.c. injec- day of delivery, and 8.6% of live-born pups died methamphetamine, includ- male analysis. The single tion on GD 7–12. Controls were pair- during the lactation period. There were no control ing a decrease in early s.c. dose level and the lack fed. Litters raised by dam, not culled, dam or offspring deaths. locomotion and olfactory of information on com- weaned on PND 28. Olfactory orientation (PND 9, 11, 13): Impaired by orientation and an increase pound purity are methamphetamine. in adult locomotion and weaknesses. Early locomotion (PND 10, 12, 14): Impaired by acoustic startle. methamphetamine at Po0.1. it eet eerh(atB 44154 2005 74:471–584, B) (Part Research Defects Birth Acoustic startle (PND 27): Increased by methamphetamine. Adult locomotor activity (PND 63–64, 70–71): Increased by methamphetamine. Straight channel swimming (PND 84): Impaired by methamphetamine. Cincinnati mazec (PND 851): No effect. Sprague–Dawley. d,l-Methamphetamine No difference by treatment in number of days to Results may indicate a The attempt to correlate be- Weissman and HCl [purity not given] 0, 2, or 10 mg/ righting reflex or eye opening. No other informa- depressed response to no- havioral and biochemical Caldecott- kg bw s.c. twice/day. Treatment tion on general toxicity. Open field (PND 30): Fewer velty or reduced change is a strength; how- Hazard, 1993 started 1–3 weeks before mating, and squares crossed by high-dose methamphetamine locomotion. ever, the use of only two mating occurred only after feed in- group. Less rearing by both methamphetamine- dose levels and the limited take returned to control levels. Treat- exposed groups. behavioral testing are ment was continued throughout Morris water mazee (PND 44): No treatment difference weaknesses. pregnancy. There were 2 low-dose, 2 in latency to find platform. METHAMPHETAMINE AND AMPHETAMINE high-dose, and 3 control litters. All Brain monoamine uptake: Variable by brain region and pups were fostered to different sal- methamphetamine dose.

O 10.1002/bdrb DOI ine-treated dams. Weaned on PND 26. The pup was taken as the statistical unit (29 control pups, 17 low-dose pups, 9 high-dose pups). Sprague–Dawley. d-Methamphetamine All dams had a decrease in body weight compared to Prenatal exposure to metham- The use of multiple dose Acuff-Smith (free base [purity not given])0,5,10, ad lib fed and watered controls during the phetamine early in gesta- levels and endpoints are et al., 1996 15, or 20 mg/kg bw twice daily by s.c. treatment period. Maternal mortality was increased tion resulted in delayed strengths of this well-con- injection from GD 7–12 or from GD and gestation length increased in the 15 and 20 mg/ development of early loco- trolled study. 13–18 (plug 5 GD 0). Rats in the 15 kg bw groups treated on GD 13–18. Gestation motion and impaired per- and 20 mg/kg bw groups and one length was also increased in the 10-mg/kg bw formance on tests of 0 mg/kg bw group were fed and group treated on GD 13–18. Post-weaning offspring learning and memory. The watered ad lib. Another 0 mg/kg bw body weights showed no consistent difference by Morris water maze results group and the 5 and 10 mg/kg bw group after GD 7–12 exposure, but showed suggested impairment of groups were pair-fed and watered continued decreases in the 2 highest dose groups spatial memory and perse- using the rats in the 15 mg/kg bw after GD 13–18 exposure. veration (inability to extin- group as a reference. Litters standar- Olfactory orientation (PND 9, 11, 13): No treatment guish an old response and dized on PND 3 to 8, balanced for sex. effects. begin a new search). There were 14–16 litters/group Early locomotion (PND 10, 12, 14): Decreased in 15 and (except for 12 litters in the ad lib 20 mg/kg bw groups after exposure on GD 7–12, control group). but not GD 13–18. Locomotion (PND 15, 30, 45, 60): No pattern of effects. Spontaneous alternation (T-maze; PND 16, 31, 46, 61): Female 20 mg/kg bw offspring exposed on GD 7–12 alternated less often than pair-fed/watered group when tested on PND 31. More 15 mg/kg bw female offspring than either control failed to perform the task in the allotted time when tested on PND 31 and 46. Passive avoidance (PND 17, 20, 32, 35, 47. 50, 62, 65): Training latency not affected by treatment. Reten- tion was decreased in 20 mg/kg bw male offspring. Straight channel swimming trials (within a week following 549 PND 50): No treatment effect. 5 OU TAL. ET GOLUB 550 Table 39 (Continued)

Strain and Dose Regimen General Toxicity and Neurobehavioral Findings Author Conclusions Comments Reference

Cincinnati water maze (within a week after PND 50):15 and 20 mg/kg bw groups took longer to solve maze than did ad lib controls (but not compared to pair- fed/watered controls). Morris maze (within 1 week after PND 57): Some effects were found in the 15 and 20 mg/kg bw groups, especially with treatment on GD 7–12. Wistar. d-Amphetamine sulfate [purity Birth weight lower in amphetamine 10 mg/kg bw/ Prenatal amphetamine expo- This study addressed beha- Tan, 2003 not specified] given s.c. at 0, 5, or day group than other groups. Litter size not sure causes behavioral sen- vioral sensitization, but its 10 mg/kg bw/day from GD 8–22 affected by treatment. sitization in rats, with utility is reduced by lack of (plug 5 GD 1), n 5 10 dams/dose Stereotyped behavior after amphetamine challenge (PND different profiles of beha- specification of compound group. Dams in 0 and 5 mg/kg bw/ 23, 60): Prenatal amphetamine at either dose vioral reactivity depending purity. day groups were pair-fed and wa- increased stereotyped behavior on PND 23 after on maternal exposure level, tered to the 10 mg/kg bw/day rats. challenge dose of 2.5, 5.0 or 7.5 mg/kg bw amphetamine challenge On the day of birth, 4 males and amphetamine. Prenatal amphetamine increased dose, and age at testing. females per litter were retained and stereotyped behavior on PND 60 only after a fostered to surrogate mothers. Pups challenge dose of 60 mg/kg bw. were weaned on PND 22. Acoustic startle (PND 90): Prenatal exposure at maternal dose of 10 mg/kg bw/day resulted in greater startle amplitude after saline challenge and after

it eet eerh(atB 44154 2005 74:471–584, B) (Part Research Defects Birth amphetamine 1 mg/kg bw challenge. Prepulse inhibition was impaired by both prenatal amphetamine doses. Postnatal Exposure of Pups Sprague–Dawley. d-Amphetamine sul- Body weight and feed consumption not evaluated. As early as 2 days of age, there The use of a single dose level, Lal and Sourkes, fate [purity not specified] 10 mg/kg Animals were observed for forward and reverse is functional integrity of the the absence of experimental 1973 bw i.p. on one of PND 1, 2, 4, 7, 10, 12, locomotion and stereotyped behavior, such as neurologic system subser- detail, and the lack of con- 14, 16, 18, 20–30, 35, or in adulthood licking, tongue protrusion, and sniffing. Bursts of ving stereotyped behavior. trolled testing limit the uti- (7–9 weeks of age). Pups were kept locomotion were noted on PND 1. Stereotyped lity of this paper. with their dams [weaning and cul- behavior began on PND 2 and was similar to that in ling not addressed]. adults by PND 18. Sprague–Dawley.d-Methamphetamine Pup body weight was decreased during treatment Methamphetamine exposure This well-controlled study Vorhees et al., [purity not given] 30 mg/kg bw s.c. period and after weaning. Animals given is associated with increased used adequate numbers of 1994a,b twice daily on PND1–10 (early) or methamphetamine on PND 1–10 remained lighter reactivity, deficits in asso- animals and multiple end- PND 11–20 (late). Animals were given than controls through the end of the experiment on ciative or memory pro- points. The single dose le- s.c. water on the days during PND PND 70; animals given methamphetamine on PND cesses. Findings in late- vel is a weakness. 1–20 when they did not receive 11–20 were similar in weight to controls by PND 49. exposed animals suggest O 10.1002/bdrb DOI methamphetamine and controls re- Mortality was increased in the group given multiple forebrain effects ceived s.c. water twice daily during methamphetamine on PND 1–10: 20% of the pups on cognitive function. PND 1–2; n 5 15 litters/dose group. died compared to 1.7% of the control pups and Early-exposed animals de- Litters culled to 4/sex on PND 1; 0.9% of pups given methamphetamine on PND monstrated predominantly weaned on PND 28. One pup/sex/ 11–20. effects on startle. The de- litter was used for each behavioral Spontaneous alternation (PND 31, 46, 61): More pups crease in activity on PND testing session. failed to make a choice after late treatment when 30 associated with neonatal tested on PND 31 only. methamphetamine it eet eerh(atB 44154 2005 74:471–584, B) (Part Research Defects Birth Passive avoidance (PND 32, 47, 62): No treatment effect treatment is in contrast to on learning or retention. the increase in activity as- Acoustic startle (PND 36, 51, 66): Methamphetamine sociated with prenatal treatment increased reactivity in females averaged methamphetamine expo- across ages and decreased prepulse inhibition in sure in other studies. both sexes (only with early treatment in males). Straight channel swimming (age not given): No treatment effect. Cincinnati water maze (age not given): Increased errors in late-treated rats was significant at po0.1. There were more trial failures among early-treated males. Morris water maze (age not given): Increased latency to find platform in late-treated animals. Locomotor activity (PND 30, 45, 60): Activity was

decreased with methamphetamine treatment, par- METHAMPHETAMINE AND AMPHETAMINE ticularly on PND 30 and particularly after PND 11–20 treatments. Fluoxetine challenge resulted in a

O 10.1002/bdrb DOI decrease in activity that was accentuated in males by PND 1–10 neonatal methamphetamine exposure. Methamphetamine challenge increased activity in all groups, which was accentuated (according to the authors’ visual inspection of the data) by neonatal methamphetamine. Sprague–Dawley. d-Methamphetamine Of 119 offspring in methamphetamine group, 62 (52%) Neonatal methamphetamine The single dose level with Vorhees et al., [purity not given] 0 or 20 mg/kg bw died, compared to 2/45 control offspring. Off- facilitates startle probably very high mortality limits 1996 twice daily by s.c. injection to female spring body weight was decreased in the by augmentation of the the utility of this study. rat neonates on PND 1–10. On PND 1, methamphetamine group during treatment and basic reflex rather than an litters were culled to 8, preferentially thereafter (to PND 49). effect on other processes retaining females. Weaned on PND Acoustic startle (PND 50): Methamphetamine-treated such as habituation or pre- 28. animals showed a decrease in startle habituation at pulse inhibition. po0.1 and a decrease in startle inhibition by a prepulse stimulus (po0.05 for average response across all prepulse intensities). Sprague–Dawley. d-Amphetamine sul- No general toxicity information given. Sensitization occurs in pre- This study addresses sensiti- McDougall et al., fate [purity not given] 0 (saline), 1.0, Challenge after 2 days: weanling rats, but only zation. The multiple dose 1994 2.5, or 5 mg/kg bw/day i.p. Â 4 days Line crosses in open-field (5 min after challenge dose of short-term, suggesting a levels and endpoints are a or R-propylnorapomorphine (a do- saline, amphetamine, or R-propylnorapomorphine): In- difference in mechanism strength. The lack of infor- pamine agonist) 1 mg/kg bw/day i.p. creased by amphetamine, but sensitization shown between short- and long- mation on compound Â 4 days. Treatments began on PND only for 1 mg/kg bw dose. term behavioral purity is a weakness. 11 or PND 17. Animals were chal- Stereotyped sniffing (same timing): Increased at highest sensitization. lenged 2 or 8 days later with the same amphetamine dose; sensitization occurred only in treatment or, in the saline-treated the younger animals. animals, with saline or 1 of the Vertical activity (same timing): Increased by ampheta- amphetamine doses (the 8-day chal- mine, with sensitization in both age groups. lenge experiments used only 5 mg/kg Challenge after 8 days: bw for 11-day-olds and 2.5 mg/kg bw Line crosses in open-field (5 min after challenge dose of for 17-day-olds); n 5 8 rats/treatment saline, amphetamine, or R-propylnorapomorphine):Rats group. given amphetamine were more active than were rats given only saline. There was no difference 551 between saline and amphetamine pretreatment groups at either age. 5 OU TAL. ET GOLUB 552 Table 39 (Continued)

Strain and Dose Regimen General Toxicity and Neurobehavioral Findings Author Conclusions Comments Reference Stereotyped sniffing (same timing): Not increased by amphetamine, although increased by R-propylnorapomorphine. Vertical activity (same timing): Increased by amphetamine, but not more so in animals pretreated with amphetamine. Sprague–Dawley. Experiment 1: On No general toxicity results were given. The N-methyl-d-aspartate re- The complex experimental Duke et al., 1997 PND 17–20, pups received daily i.p. Experiment 1: ceptor modulates sensitiza- design detracts from the injections of saline or 0.3 mg/kg bw Open field testing (PND 17–20): Rats pretreated with tion of the locomotor ability to generalize the dizolcipine (a noncompetitor antago- dizolcipine followed by amphetamine had an response to amphetamine findings of this study. nist at the N-methyl-d-aspartate re- increase in line-crosses compared to pretreatment in preweanling rats. ceptor), followed 30 min later by i.p. with saline or amphetamine alone. Amphetamine saline, 2.5 mg/kg bw d-amphetamine alone increased stereotyped sniffing. sulfate [purity not given] or 1 mg/kg Open field testing (PND 22): Preexposure to ampheta- bw R-propylnorapomorphine (a do- mine on PND 17–20 was associated with increased pamine agonist). On PND 22, rats that line-crosses, sniffing, and rearing compared to had been pretreated with only saline pretreatment with saline or dizolcipine. received a challenge of one of the Experiment 2: other treatments i.p. at the same dose. Open field testing (PND 17–20): Amphetamine in- Rats that had been given a nonsaline creased sniffing and rearing but not line-crosses treatment on PND 17–20 received a compared to saline. it eet eerh(atB 44154 2005 74:471–584, B) (Part Research Defects Birth challenge of the same treatment on Open field testing (PND 22): Pretreatment with PND 22. amphetamine caused an increase in line-crosses Experiment 2: PND 17–20 treatments and rearing but not sniffing in animals were saline, 0.3 mg/kg bw dizolci- challenged with amphetamine. Pretreatment pine, 2.5 mg/kg bw d-amphetamine with amphetamine increased line crosses, sulfate, or 1 mg/kg bw R-propylnor- sniffing, and rearing in response to challenge apomorphine. PND 22 challenge was with R-propylnorapomorphine. with amphetamine or R- propylnorapomorphine. ACI Black Agouti (presumed to be a 59.4% of ACI Black Agouti and 18.8% of Sprague– The lack of effect of treatment None Vorhees et al., poor metabolizer of methampheta- Dawley offspring died during the treatment peri- on acoustic startle was not 1998 mine) and Sprague–Dawley (pre- od. Body weight was suppressed during the dosing consistent with earlier ex- sumed to be an extensive metabolizer period with persistent body weight deficits periments and was unex- of methamphetamine). On PND 1, through PND 49 (ACI Black Agouti) and PND 42 plained. The effect of litters were culled to 8, preferentially (Sprague–Dawley). methamphetamine on spa- retaining females. d-Methampheta- Acoustic startle(PND 50): No treatment effect. tial memory (tested by the mine (free base [purity not given])0 Straight-channel swimming (PND 54): No treatment Morris maze) was consis-

O 10.1002/bdrb DOI or 30 mg/kg bw was given s.c. twice/ effect. tent with previous results. day on PND 11–20. Litters were Morris maze (unstated age): Poorer performance in The lack of influence of weaned on PND 28. Only females methamphetamine groups of both strains; no strain may mean that me- were used for behavioral tests be- strain-dependent difference in performance. tabolizer status is not rele- cause only females manifest the poor vant; however, the high metabolizer phenotype. ACI Black mortality rate among ACI Agouti: 25–26 females from 25–26 Black Agouti may have led it eet eerh(atB 44154 2005 74:471–584, B) (Part Research Defects Birth litters in each group; Sprague–Daw- to a survivor population ley: 65–69 females from 10–12 litters that did not represent the in each group. behavioral susceptibility of this strain to methamphetamine. Dark Agouti (a poor metabolizer of One-third of methamphetamine-treated Dark Agouti The increased susceptibility of This study was well planned Vorhees et al., methamphetamine) and Sprague– and 13.3% of methamphetamine-treated Sprague Dark Agouti females to with multiple endpoints 1999 Dawley (presumed to be an extensive Dawley offspring died. Methamphetamine-treated methamphetamine toxicity and addresses the issue of metabolizer of methamphetamine). offspring weighed less than controls did and the in the Morris maze is con- possible genetic suscept- On PND 1, litters were culled to 8 weight difference persisted after the treatment sistent with the unmetabo- ibility. A weakness is the (balanced for sex). d-Methampheta- period. lized drug being the active use of a single dose level. mine (free base [purity not given])0 Acoustic startle (PND 50): Methamphetamine increased toxicant. The acoustic star- or 15 mg/kg bw was given s.c. twice/ startle amplitude in Dark Agouti but not Sprague– tle results were unexpected day on PND 11–20. Litters were Dawley rats. because males were also weaned on PND 28. Dark Agouti: 20 Straight channel swimming (PND 54): Methampheta- affected; male Dark Agouti METHAMPHETAMINE AND AMPHETAMINE litters given methamphetamine, 15 mine-treated Sprague–Dawley males took longer rats have an intermediate litters given vehicle; Sprague–Daw- than controls to swim in 2 of 4 trials. There were no metabolizer status. Differ-

O 10.1002/bdrb DOI ley: 15 litters given methampheta- treatment-related differences among females or in ent brain regions are re- mine, 13 litters given vehicle. The either sex in the Dark Agouti strain. sponsible for spatial additional Dark Agouti litters had Morris maze (PND 57): Dark Agouti males and learning (hippocampus) been added to the planned n 5 15 due females and Sprague–Dawley males had an and acoustic startle (brain- to mortality. increased latency to finding the hidden stem) and these different platform after methamphetamine treatment. regions may have different In memory testing, methamphetamine-treated sensitivity to females of both strains spent less time in methamphetamine. target quadrant than did controls; the Dark Agouti females were more affected than the Sprague–Dawley females. Reacquisition followed the same sex and strain pattern as memory. Dark Agouti females treated with methamphetamine showed less competence in searching for a new platform location in reversal trials. Sprague–Dawley. On PND 1, litters were Mortality was increased by methamphetamine Developmental exposure to This well-controlled study fo- Vorhees et al., culled to 8, balanced for sex. d- administration, with 15 and 21% of pups dying methamphetamine pro- cused on cognitive effects. 2000 Methamphetamine (free base [purity during the treatment period in the 2 methamphe- duces spatial learning and The single dose level and not given]) 0 or 40 mg/kg bw/day tamine groups. Body weight was decreased memory defects. high mortality are was given s.c. on PND 11–20. Ani- in the methamphetamine groups from PND weaknesses. mals received 4 daily s.c. injections at 15 to 42. 2-hr intervals: Controls received 4 Straight-channel swimming (PND 50–56): No treatment- saline injections, 1 methamphetamine related effects. group received 4 injections of 10 mg/ Morris maze (age not specified): Methamphetamine kg bw each, and a second metham- treatment was associated with poorer performance. phetamine group received metham- The results of cued learning tests (in which a visible phetamine 20 mg/kg bw followed by marker was mounted above the underwater plat- 2 saline injections followed by a form of the Morris maze) did not explain the poorer second methamphetamine injection performance on the uncued Morris maze when of 20 mg/kg bw. There were 15–18 cued learning was used as a covariate in an litters per dose group. ANCOVA. Sprague–Dawley. Experiment 1: On No information given on general toxicity. The identified changes in The use of only two dose Crawford et al., 553 PND 11–15, 32 male and 32 female Experiment 1 and 2: protein kinase A activity levels and the limited 2000a 5 OU TAL. ET GOLUB 554 Table 39 (Continued)

Strain and Dose Regimen General Toxicity and Neurobehavioral Findings Author Conclusions Comments Reference rats were given d-amphetamine Open field testing (PND 23 and 90): Amphetamine may not be relevant to behavioral testing decrease [purity not given] 0 or 2.5 mg/kg challenge produced an increase in line crossing locomotor behavior. the utility of this study. bw/day i.p. On PND 23 or PND 90, without regard to treatment regimen on PND Whether these changes im- rats were given a challenge of am- 11–15. pact learning, addiction, or phetamine 0 or 2.5 mg/kg bw i.p. Protein kinase A activity: There were changes in dorsal memory is not known. Experiment 2: On PND 11–15, 24 male striatal and accumbal protein kinase A activity and 24 female rats were given d- (described in the text). amphetamine [purity not given] 0, 2.5, or 5.0 mg/kg bw/day i.p. On PND 23 or PND 90, rats were given a challenge of amphetamine 2.5 mg/kg bw i.p. Sprague–Dawley. On PND 1, litters were Methamphetamine treatment was associated with a Spatial learning is impaired by The single dose level and Williams et al., culled to 8, with at least 6 males/ decrease in body weight gain from PND 12 to 35. neonatal methampheta- limited behavioral testing 2000 litter. d-Methamphetamine (free base By PND 42, there was no significant body weight mine exposure. No such detract from the utility of [purity not given]) 0 or 40 mg/kg difference between the treatment groups. impairment occurs in ani- this study. bw/day was given s.c. on PND 11–20 Nonspatial Morris water maze pretraining (PND 501): mals with test experience. to male offspring. Animals received 4 There was no effect of treatment on learning to These results are consistent daily s.c. 10 mg/kg bw injections at 2- swim and to find a submerged escape platform. with hippocampal response hr intervals. There were 12 litters in Spatial learning (PND 501): Methamphetamine-treated to stressors as a mediator of each treatment arm, each of which animals that were naı¨ve to the test had impaired impaired learning and it eet eerh(atB 44154 2005 74:471–584, B) (Part Research Defects Birth contributed 2 males for testing. spatial learning. Pretrained animals did not differ memory. by methamphetamine exposure. Sprague–Dawley. On the day after de- Pup weight gain was decreased during methamphe- The sensitivity of this early The use of a single dose level Williams et al., livery, litters culled to 4/sex. d- tamine treatment. In the group treated late, body time period may be due to and the pup weight effects 2003b Methamphetamine HCl [purity not weight reached control levels by PND 42. In the treatment during the stress limit the utility of this specified] given s.c. at 0 or 10 mg/kg group treated early, body weight reached control hyporesponsive period. study. bw every 2 hr for a total of 4 doses/ levels by PND 63. Methamphetamine-asso- day on PND 11–15 or on PND 16–20 Zero mazef (BPND 50): No significant effect of ciated corticosterone release (with saline vehicle given on the other treatment. Straight-channel swimming (1 day after zero during this period may in- days in the PND 11–20 treatment maze): No treatment effect. jure neurons in the hippo- period). Litters (n 5 16) were split so Morris water maze (BPND 54): Learning impaired with campus. This explanation is that 1 male and 1 female per litter was methamphetamine exposure on PND 11–15 but not not entirely satisfactory be- treated with methamphetamine for PND 16–20. cause PND 1–10 treatment is each time period and 2 animals/sex/ also during the stress hy- litter received only saline. Day of poresponsive period but has weaning not specified. not been shown to produce the same learning deficits.

O 10.1002/bdrb DOI Sprague–Dawley. On the day of delivery, Two litters discontinued due to mortality in Methamphetamine treatment This study used multiple dose Williams et al., litters culled to 5/sex. d-Methamphe- 3 of 4 high-dose pups. Pup body weight was on PND 11–20 results in levels and focused on cog- 2003c tamine HCl [purity not specified] decreased by all doses of amphetamine on adult spatial learning and nitive effects. The high given s.c. at 0, 5, 10, or 15 mg/kg bw PND 16 and 21, and was at control levels memory deficits despite mortality is a weakness. every 2 hr for a total of 4 doses/day on PND 49. prior swimming training on PND 11–20. Litters (n 5 19) were Straight channel swimming (PND 50): No effects of and maze experience. split so that 1 male and 1 female per treatment. Reference memory was it eet eerh(atB 44154 2005 74:471–584, B) (Part Research Defects Birth litter received each dose, except 2 Cincinnati water maze (PND 53–57): No effects of specifically impaired while animals/sex/litter received the high treatment. working memory was un- dose. Day of weaning not specified. Morris water maze, cued (platform visible above water; affected. The effects are PND 60–65): No effects of treatment. selective and not due to Morris water maze, hidden platform (PND 67–145): generalized cognitive or Acquisition impaired in a dose-dependent manner motor impairment. in all amphetamine-exposed groups. When platform position was shifted, performance was impaired in high-dose group. Working memory was not adversely affected by treatment. Sprague–Dawley. On the day after de- Preweaning but not postweaning body weight gain The Barnes maze results with This study used a good ex- Williams et al., livery, litters culled to 4/sex. d- was reduced by methamphetamine treatment. the aversive stimulus sup- perimental design to focus 2003a Methamphetamine HCl [purity not Barnes mazea (PND 60–65): Methamphetamine im- port the deficit in spatial on spatial learning. The specified] given s.c. at 0 or 5 mg/kg paired performance when aversion to light was the learning seen with the single dose level is a bw every 2 hr for a total of 4 doses/ motivator but not when a food reward was the Morris maze. The lack of weakness. METHAMPHETAMINE AND AMPHETAMINE day on PND 11–20. Litters (n 5 16) motivator. methamphetamine effect were split so that 2 males and 2 Forced swim (PND 90–97): Methamphetamine asso- on Barnes maze perfor-

O 10.1002/bdrb DOI females per litter were treated with ciated with less elevation of plasma corticosterone mance with a food reward methamphetamine or vehicle. Day of after forced swim. may reflect the relatively weaning not specified. Cliff-avoidance(BPND 90): After 20 mg/kg bw dose of low dose of methampheta- methamphetamine, neonatally amphetamine-trea- mine used for the neonatal ted rats avoided cliff more than controls. treatments.

Exposure During Gestation and Lactation Periods

Sprague–Dawley. Methamphetamine Methamphetamine-exposed dams gained less weight Methamphetamine resulted in The lack of experimental de- Martin et al., HCl [purity not specified] 0 or 5 mg/ during pregnancy than either control and neonates decreased gestation length tail and the use of a single 1976 kg bw was given s.c. twice/day from were lighter on PND 7, 14, 21, and 28, and through and decreased maternal dose level detract from the GD 1 (the morning after insemina- 16 months of age. Mean birth weight was numeri- weight. Methamphetamine utility of this study. tion) to GD 21 and PND 2–21. There cally lower, but statistical significance cannot be reduced offspring weight were 25 dams treated with metham- evaluated from the information given. Gestation and the reductions per- phetamine, 13 dams injected with length was decreased by methamphetamine. sisted through 16 months saline, and 13 uninjected controls. Ear opening (PND 4), Incisor eruption (PND 7), distance of life. Methamphetamine Litters culled to 8 (males preferred) traveled quadripedally (PND 7, 11): No significant treatment was associated on PND 7, weaned on PND 21. effect of treatment. with developmental delay. Behavioral tests performed using Eye opening (PND 14, 18): Delayed by methampheta- 1 male/litter. mine treatment. Wheel-running activity (PND 901): Increased in methamphetamine group compared to injected control but no difference from uninjected control. Sprague–Dawley. Methamphetamine Offspring of methamphetamine group were lighter Maternal methamphetamine This study is somewhat useful Martin and HCl [purity not specified] 0(n 5 13) than control offspring (both groups were feed- treatment is responsible for in addressing long-term ef- Martin, 1981 or 5 (n 5 25) mg/kg bw s.c. twice/day restricted, but free-feeding littermates were said increased locomotor activ- fects of developmental GD 1–21 and lactation day 3–21 also to be lighter in the methamphetamine group, ity from adulthood through exposures. (weaning). Culling not mentioned. with citation to Martin et al. (1976). death. One male/litter used for testing, Wheel-running activity (PND 90 until death): Metham- n 5 12/dose group. Rats were feed- phetamine exposure was associated with increased restricted from PND 72 to 80% of activity. Correction of activity for body weight free-feeding weight. was said to attenuate differences, but the degree and significance of the attenuation was not 555 discussed. 5 OU TAL. ET GOLUB 556 Table 39 (Continued)

Strain and Dose Regimen General Toxicity and Neurobehavioral Findings Author Conclusions Comments Reference

Sprague Dawley. d-Amphetamine sul- Maternal body weight gain was decreased by both Prenatal amphetamine treat- None. Monder, 1981 fate [purity not specified] 0, 2, or doses of amphetamine. Litters were produced by ment is associated with a 5 mg/kg bw/day was administered only 6 control dams, 13 low-dose dams, and 8 high- delay in maturation and in drinking water. Water restriction dose dams. Litter size was not affected by treat- altered preweanling beha- was used to ensure consumption of ment, but pup body weight was decreased by both vior. Water restriction may entire dose. There were 16 dams per amphetamine doses on PND 1. have had effects on these treatment group. Mating occurred Righting reflex (PND 1, 3, 6): Significant delay in parameters, and there may after 30 days of treatment, permitting turning over in both amphetamine groups on have been an ampheta- dam body weights to return to nor- PND 3. mine-water restriction mal after an initial decrease. Pups Eye opening (PND 14): Delayed by both doses of interaction. were culled to a maximum of 8 at amphetamine. birth, weaned on PND 22. Untreated Vaginal opening (PND 37): Delayed by both doses of ad lib water supply was initiated at amphetamine. weaning of pups. Open field (PND 7, 14, 21): Amphetamine-exposed animals were more active than controls on PND 7. Bridge crossing (PND 7, 14, 21): Both amphetamine- exposed groups showed impaired ability on PND 14; only high-dose animals were impaired on PND 21. aBarnes maze: A circular surface with regular holes every 121. One of the holes leads to the goal box. it eet eerh(atB 44154 2005 74:471–584, B) (Part Research Defects Birth bBiel maze: A network of pathways, one of which leads to the goal and the others of which are blind alleys. The original Biel maze was a water maze, but dry versions have been described. cCincinnati maze: A series of nine interlocking T-mazes, can be run with water or dry. dLashley III maze: A box divided into four alleys with doors permitting entry into one alley from another. The subject must exit the start box, run down alley one to the door, turn right into alley two, left into alley three, right into alley four, and left into a goal box. eMorris maze: A circular pool of water with a submerged platform that permits escape. The subject must find the platform and remember its location. fZero maze: An elevated ring with alternating quadrants of high dark walls and low transparent walls (the open area). Avoidance of the open areas represents fearfulness or anxiety. O 10.1002/bdrb DOI it eet eerh(atB 44154 2005 74:471–584, B) (Part Research Defects Birth Table 40 Results of Non-Behavioral Developmental Neurotoxicity Testing

Strain and Dose Regimen Amphetamine/Methamphetamine Findings Author Conclusions Comments Reference

Prenatal Exposure Albino rat [strain not specified]. Ani- Serotonin and 5-HIAA: Increased on PND 1 with Development of different This study lacks experimental Bigl et al., 1982 mals s.c. injected with 0 or 2.0 mg/kg amphetamine treatment. transmitter systems ap- detail and is not useful. bw/day amphetamine [salt and en- GABA: Amphetamine treatment resulted in small but pears to be affected by antiomer not specified] from GD 7 significant decrease on PND 15 and increase on indirect drug effects. until delivery. Offspring killed on PND 21. Levels inversely related to the activity of PND 1, 8, 15, and 21 for measurement glutamate decarboxylase [methods of measure- of brain monoamine and metabolite ment not detailed].

levels [litter representation not dis- Norepinephrine: No change. METHAMPHETAMINE AND AMPHETAMINE cussed]. Results of locomotor testing Dopamine: No change. Locomotive activity results are presented in Table 39. summarized in Table 39.

O 10.1002/bdrb DOI Wistar rats. Animals were s.c. injected Maternal weight gain in amphetamine-treated dams ‘‘These changes, whether per- The experimental design is Tavares et al., with 10 mg/kg bw/day d-ampheta- was lower than saline group. No effect on gestation manent or transitory, raise confusing. The maternal 1996 mine sulfate on GD 8–22. Saline, pair- length, litter size, male:female ratio, mean birth the possibility that some of toxicity is a weakness. fed, and unhandled controls were weight of males, or postnatal growth of males. the effects of prenatal ex- included. [Number of dams treated Forebrain:body weight relationship to PND 90: No effect. posure to amphetamine was listed as 6 but it was later stated Prefrontal cortex volume: No effect on PND 14 or 30. may be due to modifica- that there were 12 litters/group.] Morphometric analyses of prefrontal cortex: Ampheta- tions in the neurotransmit- Litters were culled to 4 male and 4 mine increased neuronal density on PND 14 but ter levels of serotonin.’’ female pups. Effects on prefrontal not PND 30. cortex were evaluated on PND 14, 30, Levels of dopamine, serotonin, and their metabolites in and/or 90. [It seems that only male prefrontal cortex: Amphetamine decreased serotonin offspring were examined, and that level and increased serotonin:5-HIAA ratio on PND 2–6 offspring/group were evaluated, 30, but not PND 14. Dopamine and its metabolites although it is not clearly stated for were unaffected. each endpoint.] Wistar rats. Animals s.c. injected with In utero amphetamine treatment resulted in reduced ‘‘This change may be a con- The receptor measurements Ramirez et al., saline or 0.5 mg/kg bw/day d,l-am- density of brain a adrenergic receptors (22% less sequence of long lasting were acceptable for the 1983 phetamine sulfate throughout entire than controls) but not b receptors. There was no alterations in the metabo- time but are not current. pregnancy [number of animals trea- effect on a or b receptor affinity. lism of brain catechola- The use of a single dose ted not specified]. Litters were culled mines produced by level is a weakness. to 8 pups at birth. Adult male off- amphetamine administra- spring [7 months old] were evalu- tion at fetal age, and may ated. [Numbers of treated litters account for the behavioral represented not specified.] alterations described in these animals.’’ [See Nasello et al., 1974 and Nasello and Ramirez, 1978 in Table 39 for description of behavior effects]. Albino rats [strain not specified]. Ani- After application of conditioning trains of pulses to ‘‘This effect may be due to the None. Ramirez and mals s.c. injected with saline or the locus coeruleus, potentiation of population changes in catecholamine Carrer, 1983 0.5 mg/kg bw/day d,l-amphetamine spike amplitude in dentate gyrus was 2–3 times metabolism and (or) cate- sulfate throughout entire pregnancy greater in rats exposed to amphetamine vs. saline cholamine receptors ob- 557 [number of animals treated not in utero. served in these animals and 5 OU TAL. ET GOLUB 558 Table 40 (Continued)

Strain and Dose Regimen Amphetamine/Methamphetamine Findings Author Conclusions Comments Reference

specified]. Adult female offspring could help explain the be- (8–9/group) were evaluated. [Num- havioral alterations caused bers of treated litters represented not by amphetamine exposure specified.] in utero.’’ [See Nasello et al., 1974 and Nasello and Ramirez, 1978 in Table 39 for description of behavior effects.] C57Bl/6 mice. Animals were unhandled Birth: Amphetamine-exposed pups had lower brain Prenatal exposure to amphe- There is inadequate detail in Middaugh et al., or i.p. injected with d-amphetamine norepinephrine than did either control. tamine could alter brain this study and the experi- 1974 sulfate 0 or 5 mg/kg bw/day for the PND 21: Amphetamine-exposed pups had lower brain catecholamines in young mental design is confusing. last 6 or 7 days of pregnancy. [Num- dopamine levels than did vehicle-treated controls. animals and produce long- ber treated not specified.] Pups were PND 30: Amphetamine-exposed pups had higher lasting effects on behavior. killed at PND 0, 3, 7, 14, 21, and 30 for brain norepinephrine levels than did either control. evaluation of dopamine and norepi- There were no other significant differences be- nephrine in brain. [Number of pups tween vehicle- and amphetamine-treated groups. A and distribution by sex and litter not second experiment with same treatment paradigm stated except that a minimum of 3 but with evaluation on PND 75 found no differ- litters was represented for each ences in brain catecholamines [data not shown]. mean. It is not clear whether the n Results of behavioral testing are listed in Table 39. values given in the data table re- it eet eerh(atB 44154 2005 74:471–584, B) (Part Research Defects Birth present individual brains or pooled brain samples. Different numbers of brains were pooled at different ages and a different n is used for different treatment groups on each day of evaluation.] Wistar rats. Animals given drinking There were no behavioral disturbances but treated rats Methamphetamine changes It is a strength that metham- Tonge, 1973 water with ascorbic acid (control), were more difficult to handle. catecholamines, primarily phetamine was given oral- 80 mg/L methamphetamine HCl Norepinephrine: Methamphetamine increased levels in norepinephrine, concentra- ly, but the exposure [chirality not specified],or80mg/L hippocampus, hypothalamus, and corpora quad- tions in brain. ‘‘ythe fact through drinking water re- methamphetamine HCL plus rigemina and decreased level in thalamus. that normetanephrine con- sulted in imprecise expo- 200 mg/L chlorpromazine HCl for 4 Normetanephrine: Methamphetamine increased levels centrations are also altered sure information. weeks before mating and throughout in cortex and hippocampus and decreased level in suggests that the whole gestation and lactation [number of pons medulla. metabolism of [norepi- dams treated not specified]. Male Dopamine: Amphetamine increased level in cortex. nephrine] may be offspring killed at 3, 6, and 9 months Methamphetamine plus chlorpromazine decreased disturbed.’’ after weaning for measurement of norepinephrine in hypothalamus and increased

O 10.1002/bdrb DOI neurotransmitters and their metabo- norepinephrine in pons medulla. lite levels in brain [results only reported for 9 months, although authors stated there were no age- related effects; it is not clear how many litters were represented]. it eet eerh(atB 44154 2005 74:471–584, B) (Part Research Defects Birth Sprague–Dawley rats. Animals s.c. in- Methamphetamine treatment did not affect striatum ‘‘Monoamine metabolism in This well-controlled study Acuff-Smith jected with d,l-methamphetamine levels of dopamine, 3,4-dihydroxyphenylacetic the hippocampus and used multiple endpoints, et al., 1992 HCl at 0 (pair-fed, n 5 6), or 50 acid, homovanillic acid, 5-hydroxytryptamine (ser- neostriatum was not af- male-female analysis, and (expressed as free base, n 5 15) mg/ otonin), or 5-HIAA or hippocampus levels of fected following in utero evaluation of neurotrans- kg bw twice daily from GD 7–12 or norepinephrine, 3-methoxy-4-hydroxyphenyl gly- exposure to [d,l-metham- mitter metabolites. The GD 13–18 (plug 5 GD 0). Two pups/ col, 5-hydroxytryptamine, or 5-HIAA. phetamine] when mea- single s.c. dose level and sex/dose killed on PND 28 and sured on [PND 28].’’ lack of information on striatum and hippocampus removed compound purity are for monoamine measurement. Re- weaknesses. mainder of pups used for behavioral testing as summarized in Table 39. Nonneurological developmental effects discussed in Section 3.2.2. Sprague–Dawley rats. Animals (n 5 9) Methamphetamine-treated and pair-fed dams gained ‘‘These data, which demon- None. Cabrera et al.,

s.c. injected with 5 mg/kg bw/day less weight during treatment, but there was no strate long-term postnatal 1993 METHAMPHETAMINE AND AMPHETAMINE d,l-methamphetamine HCl in 2 di- effect on maternal cortical serotonin uptake site deficits in [serotonin] vided doses on GD 13–20. Two con- density. Methamphetamine treatment had no effect mediated renin secretion,

O 10.1002/bdrb DOI trol groups were saline injected: 1 fed on litter size, male/female ratio, pup length, suggest selective functional ad lib (n 5 5) and 1 pair-fed (n 5 8). At anogenital distance, or birth weight. alterations of brain [seroto- birth, litters culled to 5 males and 4 p-Chloroamphetamine response in PND 30 females: nin] systems in male and females, then fostered to untreated Methamphetamine attenuated p-chloroampheta- female progeny exposed in dams. On PND 30 (female offspring) mine-induced increases in plasma renin concen- utero to and PND 70 (male offspring), 1–2 tration, but had no effect on ACTH or methamphetamine.’’ pups/litter/group i.p. injected with corticosterone. saline or 8 mg/kg bw p-chloroam- p-Chloroamphetamine response in PND 70 males: phetamine (a serotonin releaser). Methamphetamine attenuated p-chloroamphetamine-induced increases in plasma renin activity (also observed in pair-fed group) and plasma renin concentration, but had no effect on ACTH, corticosterone, or prolactin. Serotonin uptake sites in cortex and hypothalamus and 5- HT1 and 5-HT2 receptors in cortex in PND 70 males: No effect on density. Wistar rat. Animals s.c. injected with Substantia nigra: Nonsignificant trend for decreased ‘‘Collectively, the present re- This study demonstrates the Gomes-da-Silva 5 mg/kg bw/day d,l-methampheta- tyrosine hydroxylase mRNA expression in female sults indicated that gesta- transient nature of treat- et al., 2002 mine HCl in 2 divided doses on GD offspring on PND 14. tional [methamphetamine] ment effects; however, uti- 8–22. Controls were pair-fed and Ventral tegmental area: Decreased tyrosine hydroxylase exposure affects [tyrosine lity is limited by the lack of saline injected. [Numbers treated not mRNA expression in female offspring on PND 7 hydroxylase] gene expres- experimental detail and the specified.] Litters culled to 4 male and 14. sion in the postnatal life, a single s.c. dose level. and 4 female pups at birth. Offspring No effects in male offspring or in female offspring at phenomenon that appears evaluated on PND 7, 14, and 30. [Two the other time points. to be transient, since it is no to eight determinations/group but longer evident by the end the number of pups examined and of the first month of life in litters represented not specified.] the rat.’’ C57B1/6 mouse. Animals (6–7/group) Methamphetamine treatment during both gestational ‘‘This increase in fetal dopa- This study is useful in de- Heller et al., s.c. injected twice daily with saline or periods increased dopamine levels in fetal corpus mine is consistent with our monstrating prenatal ef- 2001a 40 mg/kg bw d-methamphetamine striatum (60% for GD 7–13 and 32% for GD 7–15 findings that exposure to fects. The addition of HCl on GD 7–13 or 7–15. Dams killed compared to controls) and rostral mesencephalic methamphetamine in utero postnatal follow-up would on GD 16 for examination of dopa- tegmentum (48% for GD 7–13 and 25% for GD 7–15 results in adult dopami- have been helpful. 559 mine levels in fetal brain. compared to controls). nergic neurons which are more responsive in terms of 6 OU TAL. ET GOLUB 560 Table 40 (Continued)

Strain and Dose Regimen Amphetamine/Methamphetamine Findings Author Conclusions Comments Reference methamphetamine induced release of the neurotransmitter and more sensitive to the neurotoxic effects of the drug.’’ [Based on previous study.] C57B1/6 mouse. Animals (8/group) s.c. Compared to mice exposed to saline in utero and Prenatal exposure to metham- This study is useful in inves- Heller et al., injected twice daily with saline or challenged with methamphetamine in adulthood, phetamine produces re- tigating the effects of adult 2001b 40 mg/kg bw d-methamphetamine in utero methamphetamine exposure and adult gion- and sex-specific methamphetamine chal- HCl from GD 7–18 (plug 5 GD 1). methamphetamine challenge resulted in decreased alterations in the adult lenge after prenatal expo- Adult offspring (4/sex from different striatal levels of dopamine, DOPAC, homovanillic brain that may signal in- sure. The single s.c. dose litters) received 2 s.c. injections sepa- acid, and ventral brainstem dopamine level in creased susceptibility to level is a weakness. rated by 2 hr of methamphetamine males, and reduction of striatal 3-methoxytyramine neurotoxicity of male HCl 0, 5, 10, 15, or 20 mg/kg bw. in both sexes. methamphetamine abusers Seven days later monoamines and Temperature response (increase correlated with de- who were also exposed metabolites were measured in brain. creased striatal dopamine) after methamphetamine prenatally to this drug. challenge of adults did not vary with prenatal treatment. Postnatal Exposure

it eet eerh(atB 44154 2005 74:471–584, B) (Part Research Defects Birth Wistar rat. Males were s.c. injected with Amphetamine treatment inhibited body weight gain, ‘‘The functional repercussions This study is not useful based Tavares and saline or 25 mg/kg bw/day d-am- especially on PND 6–12 and 24–30, and resulted in of the body weight gain on the grossness of the Silva, 1993 phetamine sulfate administered in 2 lower brain weight. Amphetamine treatment re- alliedtothedecreasedvo- brain measurements. The divided doses on PND 1 (day after duced mean volume of hippocampal formation lume of the dentate gyrus single s.c. dose level is a birth) to PND 30. Growth was eval- with significant difference at the level of the molecular layer after early weakness. The conclusions uated from PND 6–30 and brain dentate gyrus. psychostimulant exposure of the authors are weight on PND 30 (in males selected arelikelytobenoteworthy speculative. from 9 different litters containing 4 since, whether permanent males/litter). Hippocampal volu- or simply representing a metric parameters (n 5 6/group) developmental delay, they were measured on PND 30. will induce a reduction in the input information to- wards the hippocampusy’’ Sprague–Dawley rats. Experiment 1: Compared to rats pretreated with saline, rats pre- Behavioral relevance of The single i.p. dose level is a Crawford et al., Males and females (16/sex/group) treated with d-amphetamine on PND 11–15 had reduced phosphokinase weakness and the relevance 2000a i.p. injected with saline or 2.5 mg/kg lower dorsal striatal protein kinase A activity after A activity is uncertain since of the findings is unclear. bw/day d-amphetamine sulfate on challenge with saline (80.3% of saline pretreatment locomotor sensitization did PND 11–15 (pretreatment period), group) or d-amphetamine (66.7% of saline pre- not occur after ampheta-

O 10.1002/bdrb DOI then received challenge with either treatment group) on PND 23 and lower dorsal mine challenge (see saline or d-amphetamine on PND 23 striatal protein kinase A activity after challenge Table 39). or 90 (n 5 4/sex/group). After chal- with saline (67.3% of saline pretreatment) or d- lenge, locomotor activity was tested amphetamine (80.5% of saline pretreatment group) (see Table 39) and dorsal striatum on PND 90. (caudate-putamen) removed for measurement of striatal protein kinase A activity. it eet eerh(atB 44154 2005 74:471–584, B) (Part Research Defects Birth Experiment 2: Animals (male and female, PND 23: Protein kinase A activity was lower in the 8/sex/group) i.p. injected with saline dorsal striatum of the 2.5 mg/kg bw/day d- or 2.5 or 5.0 mg/kg bw/day d-am- amphetamine pretreatment group (59.3% of saline phetamine sulfate on PND 11–15, pretreatment group). then challenged with 2.5 mg/kg bw PND 90: In both the 2.5 and 5.0 mg/kg bw/day d-amphetamine on PND 23 or 90. Rats amphetamine pretreatment groups, protein kinase were tested for motor activity (Table A activity was lower in dorsal striatum (76.2% of 39) and killed to measure protein saline control levels at high dose and 80.6% of kinase A activity in dorsal striatum saline control levels at the low dose) and nucleus and nucleus accumbens. accumbens (77.2 and 70.3% of saline control levels at the high and low dose). Sprague–Dawley rats. Males and Protein kinase A: Activity in amphetamine group was It was speculated that reduced The evaluation of dopamine Crawford et al., females i.p. injected with saline or decreased to 77% of control values in males and dopamine levels may have receptor changes is a 2000b 2.5 mg/kg bw/day d-amphetamine 88% of control values in females. caused the upregulation of strength. The single i.p. MHTMN N METHAMPHETAMINE AND AMPHETAMINE sulfate on PND 11–17. Rats killed on Dopamine: Content in amphetamine group was de- striatal D2-receptors that dose level is a weakness. PND 90 for evaluation of protein creased to 43% of control values in males and 75% may then have inhibited kinase A activity and dopaminergic of control values in females. protein kinase A activity

O 10.1002/bdrb DOI endpoints in dorsal striata (5–6/ D2-like receptors: Density significantly increased with through the cAMP signal group) [not clear if n was per sex or amphetamine treatment (176% of control values in transduction pathway. total]. males and 129% of control values in females), but [Statement later retracted. there was no effect on affinity. See (Crawford et al., 2003) D1-like receptors: Density and affinity not significantly below.] affected. Sprague–Dawley rats. Males and fe- Methamphetamine significantly reduced protein Based on the decrease in D2- This study was better de- Crawford et al., males s.c. injected with saline or kinase A activity in males [73.4% of control value] like receptor binding sites, signed than previous stu- 2003 10 mg/kg bw methamphetamine HCl and in males and females combined [82.7% of the authors retracted their dies, and evaluated long- (as free base) [chirality not specified] control value]; decreased density of D2-like recep- previous suggestion that term effects. 4 times daily, every 2 hr, on PND 11– tors in males and females combined [87.7% of up-regulation of striatal D2 20 (day of birth 5 PND 0). Rats killed control value] with no effect on receptor affinity; receptors may be responsi- on PND 90 for evaluation of protein and reduced dopamine (86.8% of control value) ble for the inhibition of kinase A activity and dopaminergic and DOPAC (82.8% of control values) in both sexes. protein kinase A activity endpoints in dorsal striatum (6/ Males were more sensitive to effects on protein (see (Crawford et al., group, although not clear if the kinase A activity and DOPAC content. 2000b), above). As an alter- number was per sex or total). native, they suggested that desensitization of D1-like receptors may be responsible for the decline in protein kinase A activity. They also noted the comparabil- ity of the response in this study to methamphetamine 40 mg/kg bw/day and the response in their prior studies to amphetamine 2.5 or 5 mg/kg bw/day. Given the greater neurotoxicity potency of methamphetamine, the authors concluded that the decrease in 561 protein kinase A activity 6 OU TAL. ET GOLUB 562 Table 40 (Continued)

Strain and Dose Regimen Amphetamine/Methamphetamine Findings Author Conclusions Comments Reference was not due to a neurotoxicity mechanism. Sprague–Dawley rat. Animals s.c. in- Caudate dopamine was reduced by d-amphetamine Findings were consistent with The multiple dose levels per- Wagner et al., jected twice daily on PND 10–40 with 50 mg/kg bw/day and by methamphetamine 25 effects of the ampheta- mit determination of a 1981 total doses of 0, 12.5, 25, or 50 mg/kg and 50 mg/kg bw/day. There were no alterations mines on adult rat brains. NOEL. bw/day methamphetamine HCl in norepinephrine levels in caudate, midbrain, [chirality not specified] or d-amphe- hypothalamus, pons-medulla, or telencephalon. tamine sulfate. Each drug administered to at least 2 litters of 10 that had been constructed from pooled and redistributed PND 3 pups (without regard to sex). Two weeks after treatment, neurotransmitter levels were measured in brain. Sprague–Dawley rats. Animals s.c. in- Methamphetamine-treated animals demonstrated Caudate dopamine was re- The use of only two dose Lucot et al., 1982 jected with methamphetamine HCl significant reductions in caudate dopamine con- duced after neonatal levels is a weakness. [chirality not specified] from PND 7 centrations, ranging from 67–75% of control values. methamphetamine to a les- to PND 10 (0, 50, or 100 mg/kg bw/ There were no alterations in norepinephrine or ser extent than in adults day) or PND 17 to PND 20 (0 or serotonin levels in caudate, pons-medulla, or [based on comparisons 100 mg/kg bw/day). Treatments gi- telencephalon. with previously published ven in 2 equal doses per day sepa- experiments; no adult data it eet eerh(atB 44154 2005 74:471–584, B) (Part Research Defects Birth rated by 12 hr. Pups (sex not were presented in this specified) killed 2 weeks after the last study]. dose and neurotransmitter levels in brain measured. Male gerbil. Animals received a single Methamphetamine treatment increased total dendritic ‘‘The present results show that This study used a novel spe- Blaesing et al., i.p. injection of saline or 50 mg/kg bw length, slightly increased dendritic branching, and early postnatal treatment cies without a clear ratio- 2001 methamphetamine on PND 14. Brains increased spine density in pyramidal cells of the with methamphetamine nale. The use of a single i.p. from 7–8 gerbils/group were exam- prefrontal cortex. significantly interferes with injection is a weakness. The ined on PND 90. morphogenesis of prefron- use of more detailed mortal pyramidal cells.’’ phologic evaluation is a strength. O 10.1002/bdrb DOI AMPHETAMINE AND METHAMPHETAMINE 563 male rats exposed to 25 mg/kg bw/day amphetamine on litters and number of animals tested from each litter. PND 1–30 (Tavares and Silva, 1993). Increases in Many studies used statistical procedures (e.g., t-tests) dendritic length and branching and spine density of that can bias toward false positives. pyramidal cells of the prefrontal cortex were reported in Utility (Adequacy) for CERHR Evaluation Process: 90-day-old male gerbils that received 50 mg/kg bw Overall, the data provide useful information for the methamphetamine on PND 14 (Blaesing et al., 2001). evaluation process. Exposures are prenatal and postnatal. Behavioral effects were not reported in these studies. Confidence is moderate due to unresolved questions Most of the studies examining biochemical effects about the relationship between maternal or pup focused on measurement of monoamine levels in the toxicity and the behavioral effects. This level of con- brain. Three of the studies were conducted in mice fidence could be increased with more detailed analyses (Middaugh et al., 1974; Heller et al., 2001a,b) and the rest of the data. in rats (Tonge, 1973; Wagner et al., 1981; Bigl et al., 1982; Lucot et al., 1982; Acuff-Smith et al., 1992; Tavares et al., 1996; Crawford et al., 2000b, 2003). Studies of ampheta- 3.2 Utility of Developmental Toxicity Data mine and methamphetamine in rats and mice demon- There are data from humans and experimental animals strated variable effects on brain levels of dopamine, to evaluate the potential developmental toxicity of norepinephrine, and serotonin. One study demonstrated amphetamine. The experimental animal data set can be a reduction in serotonin-mediated renin secretion after evaluated for methamphetamine. Most of the publica- prenatal methamphetamine exposure (Cabrera et al., tions on human developmental toxicity involve illicit use 1993) and other studies demonstrated inconsistent for which the identity and purity of the drug being used changes in dopamine receptor density after postnatal is uncertain, making this literature of no utility for the exposure to amphetamine (Crawford et al., 2000b) or evaluation of methamphetamine developmental toxicity methamphetamine (Crawford et al., 2003). Direct com- in humans. parison of studies examining monoaminergic endpoints There are several case reports describing a variety of is precluded by differences in species, dose level and adverse outcomes among infants exposed antenatally to duration, developmental period of exposure, age of various amphetamines. These case reports are not useful evaluation, and brain region examined. No obvious in assessing the potential developmental effects of the patterns were noted. amphetamines. There is a case series from Sweden A series of studies from Crawford et al. (2000a,b, 2003) involving 14 years of follow-up of children born to examined effects on protein kinase A, a critical enzyme amphetamine-abusing women; the utility of this study is in signal transduction cascades of several different limited by the presence of other prenatal and postnatal receptors, including dopamine receptors. Prenatal ex- liabilities for abnormal development. Three case-control posure to amphetamine was shown to reduce protein and eight cohort studies of varying utility on ampheta- kinase A activity in immature and adult rats, and mine exposure during human pregnancy are available. prenatal methamphetamine exposure was shown to Some of the exposures in these studies were to reduce protein kinase A activity in adults. Other studies pharmaceutical preparations for weight loss and some demonstrated changes in brain GABA levels (Bigl et al., were recreational exposures to methamphetamine. None 1982), reduced density of brain a adrenergic receptors of these studies involved amphetamines prescribed for (Ramirez et al., 1983), and changes in nerve conduction ADHD. (Ramirez and Carrer, 1983) after prenatal amphetamine There are five experimental animal developmental exposure. Changes in brain tyrosine hydroxylase mRNA studies from which assessment of dose–response is expression after prenatal methamphetamine exposure possible, though limitations were noted in each of the were also reported (Gomes-da-Silva et al., 2002). These studies. Two of these studies (one rat, one rabbit) were findings are not known to have been replicated by other available only as FDA summaries (FDA, 2001), preclud- investigators. ing detailed review of the data by the Expert Panel. Three Strengths/Weaknesses: Strengths of the data set taken of the multidose-level studies were conducted in mice as a whole include the large number of studies (Kasirsky and Tansy, 1971; Fein et al., 1987; Yamamoto conducted over more than a 40-year time span, the good et al., 1992). Additional developmental neurotoxicity experimental design and excellent statistics in some of studies featured treatment of pregnant or neonatal the studies (particularly important is controlling the rodents. Of the 37 studies with behavioral endpoints number of litters used), and the similarities noted in a (Table 39), 23 included a single dose level plus a control, large number of these studies. Summarizing across seven included two dose levels plus a control, and seven studies suggests that developmental exposure will included three or more dose levels plus a control. produce developmental delays, alter motor function Quantitative evaluation of the behavioral endpoints (activity), and yield deficits in cognitive functions. was possible using this data set, but limited by the Biochemical studies, although not entirely confirmatory presence in many instances of high administered doses of one another, do support the hypothesis that amphe- with significant uncontrolled effects on body weight. tamine and methamphetamine alter catecholamines. There were 19 studies with non-behavioral neurodeve- Weaknesses include the inability to tell from many of lopmental endpoints (Table 40), some of which also had the reports whether body weight or lethality confounded behavioral endpoints. Of these 19 studies, 15 included a interpretation of behavioral changes and lack of informa- single dose level plus the control, and two each tion on how the high doses used in early postnatal contained two dose levels and three or more dose levels exposures relate to early human exposures. Many plus a control. Quantitative evaluation of neurodevelop- studies, especially early reports, did not contain ade- mental endpoints was possible using this data set, but quate statistical information concerning the number of limited by the presence in many instances of high

Birth Defects Research (Part B) 74:471–584, 2005 DOI 10.1002/bdrb 564 GOLUB ET AL. administered doses with significant uncontrolled effects reported reduced appetite and insomnia as the most on body weight. common side effects. Headache, stomachache, and irritability were also reported in one study (Greenberg et al., 1972). Although one study reported that an 3.4 Summary of Developmental Toxicity Data increase in depression was a surprising finding (Green- 3.4.1 Human data. An uncontrolled case series berg et al., 1972), the other study reported that ‘‘unusual from Sweden involved 71 children born to amphetamine happiness’’ was reduced by d-amphetamine therapy addicts, 65 of whom were followed to age 14 (Cernerud (Efron et al., 1997). One author noted that some side et al., 1996; Eriksson et al., 1994, 2000). Abnormalities of effects reported on stimulant therapy may actually be growth and school performance were noted in these related to the underlying disorder (Efron et al., 1997). children; however, prenatal exposure to ethanol, a high There are no useful data with which to evaluate the rate of foster care, and other socioeconomic factors were risk of tics or movement disorders in association with not excluded as causes or contributors to the outcomes in amphetamine or d-methamphetamine treatment. There is these children. Another case series (Furara et al., 1999) only one controlled study (Efron et al., 1997), which has reported that 28.6% of pregnancies in women with self- weaknesses that preclude its use in the evaluation reported amphetamine use ended before 37 weeks process. gestation and 25% had newborns weighing o2500 g. Concerns have been raised that stimulant treatment in Controlled studies on pregnancy outcome in women childhood can increase the risk for developing substance exposed to amphetamines, or exposed to stimulants abuse disorders later in life. Numerous studies examin- including amphetamines, are summarized in Table 41 ing possible associations between ADHD, independent and Table 42. Of the case-control studies in Table 41, one of treatment, and substance abuse were not considered report (Nelson and Forfar, 1971) was evaluated by the by the Expert Panel. The Panel notes a review by Wilens Expert Panel as being of limited utility and two reports (2002) that concluded, ‘‘There is a robust literature (Nora et al., 1970; Levin, 1971) were evaluated as not supporting a relationship between ADHD and SUD being useful based on methodologic problems or lack of [substance use disorders]. Noncomorbid ADHD seems detail. Of the cohort studies reported in Table 42, those to confer an intermediate risk factor for SUD, although concerning illicit drug abuse were not considered useful conduct and bipolar disorder seem to heighten the risk of in the evaluation process due to uncertainty about the early onset of SUDy.’’ In studies found to be useful or to constituents of the drug being used and for some studies, have limited usefulness for evaluating risks of substance other methodological problems. abuse, the type of stimulant treatment was not specified General side effects in children treated with d- in one study (Biederman et al., 1999); d-amphetamine amphetamine were discussed in two controlled studies was given to 3% of subjects and to 2% of subjects who (Greenberg et al., 1972; Efron et al., 1997). Both studies also received treatment with methylphenidate in the

Table 41 Case-Control Studies on Human Pregnancy Outcome after Maternal Exposure to Amphetamines

Cases and controls Exposure assessment Risk estimate Reference

Cases 5 184 children with Mothers asked within a year of 18% of case mothers and 9% of control mothers Nora et al., congenital heart the birth about d-amphetamine were exposed (Po0.05) [OR 5 2.40, 95% 1970a disease. use. CI 5 1.06–5.95]. Controls 5 108 children without congenital heart disease. Cases 5 458 mothers of Mothers were asked about d- For total malformations, exposure during all of Nelson and children with congenital amphetamine use, with con- pregnancy occurred in 13/458 cases and 10/ Forfar, malformations. firmation of use by general 911 controls [OR 5 2.63, 95% CI 5 1.07–6.52], 1971 Control 1 5 500 mothers of practitioner, hospital records, exposure in the first trimester occurred in normal children born or prescription records. 11/458 pregnancies and 8/911 controls immediately after each [OR 5 2.78, 95% CI 5 1.03–2.61], exposure case. during the first 56 days occurred in 10/458 Control 2 5 mothers of 411 pregnancies and 5/911 controls [OR 5 4.04, mothers of normal chil- 95% CI 5 1.27–13.65], and exposure during dren matched to cases the first 14 days occurred in 8/458 preg- on maternal age and nancies and 2/911 controls [OR 5 8.08, 95% parity and infant sex. CI 5 1.59–55.25]. Cases 5 11 infants with Drug use histories were solicited 4/11 case mothers used amphetamines during Levin, 1971a biliary atresia. from mothers for the first trimester and 3/50 control mothers Controls 5 50 normal in- amphetamines. used amphetamines in pregnancy fants of the same age. [OR 5 8.95, 95% CI 5 1.20–70.92].

OR and 95% CI were calculated by CERHR using the CDC SABER program. aThis study was judged not to be useful in the evaluation process due to methodologic problems or lack of detail and is presented for completeness only.

Birth Defects Research (Part B) 74:471–584, 2005 DOI 10.1002/bdrb it eet eerh(atB 44154 2005 74:471–584, B) (Part Research Defects Birth

Table 42 Cohort Studies on Human Pregnancy Outcome after Maternal Exposure to Amphetamines

Exposed and Unexposed Outcome Assessment Risk Estimate/Comparisons Reference

Therapeutic use 50,282 mother-child pairs, Physical examination of children Comparison of children with specific exposure of interest to all children without Heinonen, 1977 retrospective and prospective up to 1 year of age in 91% of exposure of interest during first 4 lunar months: record of pregnancy the sample. d-Amphetamine crude RR 5 1.23 [95% CI 5 0.82–1.82]; standardized RR for any exposures; exposures in first 4 malformation 1.08 (95% CI 5 0.65–1.68), for major malformation 1.29 (95% METHAMPHETAMINE AND AMPHETAMINE lunar months: CI 5 0.73–2.10), for minor malformation 1.46 (95% CI 5 0.59–2.98). d-Amphetamine n 5 367; Amphetamines crude RR 5 1.23 [95% CI 5 0.72–2.05]. Naeye, 1983

O 10.1002/bdrb DOI ‘‘Amphetamines’’ n 5 215; Methamphetamine crude RR 0.87 [95% CI 5 0.21–2.22]. Methamphetamine n 5 89. In a separate analysis, birth weight was 100–400 g lower among d-amphetamine- exposed nonmalformed babies if the mother took the medication after 28 weeks of gestation and gained 412 kg during pregnancy or had a prepregnancy weight Z45 kg. Caucasian women insured by Diagnosis of a congenital malfor- For exposures during pregnancy, ‘‘severe congenital anomaly’’ in 3.4% of each Milkovich and Kaiser Health Plan who deliv- mation at birth or at a Kaiser exposure group [crude RR 5 1.01 95% CI 5 0.76–1.32]. Also, cleft palate in 5/1694 van den Berg, ered in the San Francisco East clinic through 61 months of exposed, 21/10,213 unexposed [crude RR 5 1.43, 95% CI 5 0.54–3.8]. 1977 Bay area; 1694 used ampheta- age. For exposures during the first 56 days of pregnancy: Cleft palate 3/175 exposed. mines for weight loss, 10,213 did not use anorectant drugs.

Illicit Usea 46 infants exposed antenatally to Gestational age, birth weight, Oro and Dixon, cocaine or methamphetamine length, head circumference, 1987 identified by maternal or in- perinatal complications. [SD fant toxicology screens com- assumed in interpreting pared to 45 infants not exposed errors.] Cocaine and ampheta- to tested illicit drugs. mine exposure not separated. 565 6 OU TAL. ET GOLUB 566 Table 42 (Continued)

Exposed and Unexposed Outcome Assessment Risk Estimate/Comparisons Reference 81 stimulant-exposed infants (27 Presumably record review. Co- Dixon and Bejar, infants exposed to metham- caine and amphetamine expo- 1989 phetamine) based on infant sure not separated. [SD urine toxicology screening, 87 assumed in interpreting drug-free infants suspected of errors.] having hypoxic-ischemic encephalopathy (HIE), 19 normal drug-free infants. All born at term.

52 methamphetamine-abusing Record review. No difference between groups in rate of pregnancy complications or congenital Little et al., 1988 pregnant women compared anomalies. using record review with 52 it eet eerh(atB 44154 2005 74:471–584, B) (Part Research Defects Birth women not known to have abused drugs whose babies were born next in the delivery unit.

135 methamphetamine-exposed Record review. Gestational age reduced a mean 2.2 weeks (po0.001) in methamphetamine group. Smith et al., 2003 infants identified by maternal Small for gestational age increased in methamphetamine group to 13/135 compared urine toxicology or history, 160 to 2/160 (po0.001) [crude RR 5 7.70, 95% CI 5 2.08–46.20]. matched controls, all Z37 weeks gestation. 8 infants with history of maternal Fagan Test of Infant Intelligence, Birth weight 500 g lower in exposed group. Visual-evoked potentials not affected by Hansen et al.,

O 10.1002/bdrb DOI methamphetamine use during visual-evoked potentials. exposure status. Fagan performance poorer in exposed infants with 4/8 ‘‘at risk’’ 1993 pregnancy, 8 unexposed in- based on their scores, compared to 0/8 unexposed infants. fants matched for ethnicity. aEvaluated by the Expert Panel as not useful in the evaluation process due to uncertainty about the constituents of the drug being used, and, for some studies, other methodologic problems. AMPHETAMINE AND METHAMPHETAMINE 567 second study (Barkley et al., 2003); and methylphenidate base in water at total daily dose levels of 0, 2, 6, or 20 mg/ was probably used in 100% of subjects in the third study kg bw in two equal divided doses 8 hr apart. [The Panel (Paternite et al., 1999). None of the studies found assumes that the test article was d-,l-amphetamine in a evidence that prolonged treatment of ADHD with 3:1 ratio, as in the marketed product.] The high-dose stimulants in childhood increased the risk of tobacco or animals were killed on GD 9 due to excessive toxicity. cigarette use in adolescence (Biederman et al., 1999; There were clinical observations and a decrease in dam Barkley et al., 2003) or alcohol or substance abuse in body weight gain in the low and middle dose. There adolescence (Biederman et al., 1999; Barkley et al., 2003) were no treatment-related effects on implantations, or adulthood (Paternite et al., 1999; Barkley et al., 2003). resorptions, live young, pre- or post-implantation loss, One of the studies found associations between ADHD or sex ratio, or placental, fetus, or litter weights. There was conduct disorders and substance use (Paternite et al., an increase in the number of litters with three 1999). Two studies reported associations between stimu- or more fetuses exhibiting delayed cranial ossification lant treatment (15% of subjects treated with ampheta- (one, four, and six of 22 litters in the control, low-dose, mines or pemoline) and tobacco or cocaine dependency and mid-dose groups, respectively [Benchmark dose in adulthood (Lambert and Hartsough, 1998; Lambert, calculations performed by CERHR for this endpoint 2002). The Expert Panel found the studies to be gave a BMD10 of 3 mg/kg bw/day and a BMDL of unreliable for drawing conclusions due to limitations 2 mg/kg bw/day]). such as inadequate consideration of severity of ADHD, In the same FDA review (FDA, 2001), New Zealand apparent errors in the reporting of data, and questionable White rabbits (22/dose group) were given divided statistical procedures. amphetamine [assumed 3:1 ratio of d- and l-enantio- The effects of d-ord,l-amphetamine on growth of mers] gavage doses totaling 0, 2, 6, or 16 mg/kg bw/day children were evaluated in ten studies summarized in on GD 6–19. Clinical signs were noted in the high-dose Table 33. Studies have reported variable findings; dams. There were no treatment-related effects on weight however, the weight of evidence suggests that ampheta- gain or feed consumption. There were no effects on mine treatment is associated with an initial decrease in implantations, resorptions, live young, sex ratio, pre- height and weight gain in children. It is not known implantation loss, or alterations in placental, fetal, or whether final adult height and weight are affected by litter weights. A slight increase in post-implantation loss current treatment regimens, which frequently include in the high-dose group (17.2 vs. 12% in controls) was not continuous use and use beyond childhood. The quality of statistically analyzed and could not be done so by data in the older articles is suboptimal. These articles CERHR due to lack of SD or SEM. A reported increase have variable but generally marginal-to-moderate utility in fetuses with 13 ribs was statistically analyzed by with lack of masked assessments; incomplete documen- CERHR and found not to be significant. The FDA tation of compliance or actual dosing regimens; and reviewer concluded that there was no clear evidence of failure to consider (in most cases) basic factors that are developmental toxicity but noted that pharmacokinetic usually assessed in growth studies, such as mid-parent data (presented in Section 2) indicated that exposures in height and parent BMI; family history of timing of these experimental animal studies were lower than puberty onset; the child’s actual physical or endocrino- exposures in children (based on AUC) or only two to logic level of puberty at start of treatment (some of the three times higher (based on Cmax). youngsters were as old as 15 when the studies were Fein et al. (1987) i.p. injected ICR mice with 50 or conducted); and measurement of skeletal maturity (bone 100 mg/kg bw/day d-amphetamine on GD 9 and GD 11 age), which particularly in school-aged children is and control mice were injected with saline on one or considered a useful indication of expected growth more days between GD 9–11. EKGs were recorded in potential. The seasonal differences in expected growth fetuses on GD 15 and 19, whereas other litters were (in the northern hemisphere, children grow faster in evaluated for implantation and resorption sites, fetal summer) are not accounted for by designs that compare survival, and external and internal malformations on children whose families chose to leave them on GD 19. The number of surviving dams was reduced in stimulants through the summer and children whose both dose groups to 58–63% versus 100% in controls. families did not leave them on medication during the Malformations were increased in fetuses from both dose summer. Thus, it cannot be ruled out that those who groups and included microphthalmia, amelia, exence- remained on the medicines also had other conditions or phaly, and cleft lip. EKG patterns in both dose groups behavioral patterns (like fetal alcohol effects) that suggested delayed histodifferentiation of cardiac tissue. motivated their parents to continue the medication and At the high dose, resorptions were increased and fetal might also decrease growth. The studies did not control body weights (only reported for high dose) were for potential confounders such as intrauterine exposure reduced. [The Expert Panel noted that this study is to tobacco, ethanol, and illicit drugs or parental mental limited by the exceeding of MTD in dams and health. improper statistical procedures.] 3.4.2 Experimental animal data. Developmental Kasirsky and Tansy (1971) i.v. dosed 50 mice/group toxicity studies providing dose–response information are with 5 or 10 mg/kg bw/day methamphetamine HCl summarized in Table 43. [purity not specified] on GD 9–11, 9–12, 12–15, or 9–15 A GLP rat and rabbit developmental toxicity study (plug day 5 GD 1). One control group of 50 mice was not reviewed in the FDA Pharmacology Report (FDA, 2001) treated and a second group of controls was given saline was not available to the Expert Panel, but is summarized on GD 9–15. Maternal weights were significantly here because it is the only study conducted with oral reduced in every treatment group and feed and water exposure at multiple dose levels. Pregnant rats (22/dose intake were decreased on the first 2–3 days of exposure group) were gavaged on GD 6–17 with free amphetamine [data not shown]. Fetal weights were significantly

Birth Defects Research (Part B) 74:471–584, 2005 DOI 10.1002/bdrb 6 OU TAL. ET GOLUB 568 Table 43 Summary of Multiple-Dose Experimental Animal Prenatal Developmental Toxicity Studies

Maternal Effect Level (mg/kg Critical Developmental Developmental Effect Level Species/Strain Exposure bw or mg/kg bw/day) Effects (mg/kg bw or mg/kg bw/day) Reference

b Sprague–Dawley d,l-Amphetamine. Gavage 0, LOAEL 5 2 mg/kg bw/day Litters containing fetuses [BMD10 5 3; BMDL 5 2] FDA, 2001 rat 2, 6, or 20 mg/kg bw/daya (decreased body weight with delayed cranial on GD 6–17. gain) ossification. New Zealand White d,l-Amphetamine. Gavage 0, LOAEL 5 16 (clinical signs) Increased post-implantation LOAEL 5 16 FDA, 2001 rabbit 2, 6, or 16 mg/kg bw/daya loss. on GD 6–19. ICR mouse d-Amphetamine. LOAEL 5 50 (decreased Increased malformations and LOAEL 5 50 Fein et al., 1987 i.p. 50 or 100 mg/kg bw/ survival) EKG patterns suggesting day GD 9–11. delayed myocardial differentiation. Increased resorptions and LOAEL 5 100 decreased fetal weights. Jcl:ICR mouse Methamphetamine. i.p. 11, 13, LOAEL 19 (mortality) Fetal death NOAEL 21 Yamamoto et al., b b 14, 15, 17, 19, or 21 mg/kg [BMD10 5 14.9, [BMD10 5 14.6; BMDL 5 10.3] 1992 bw on GD 8. BMDL 5 12.7] External malformations LOAEL 19 b [BMD10 5 16.1; BMDL 5 12.2] Skeletal malformations LOAEL 14 b [BMD10 5 15.5; BMDL 5 12.4] 5 5 b 5 it eet eerh(atB 44154 2005 74:471–584, B) (Part Research Defects Birth CF1 mouse Methamphetamine i.v. 5.0 or LOAEL 5.0 (decreased body Resorptions LOAEL 5 [BMD10 0.6] Kasirsky and Tansy, 10.0 mg/kg bw/day on GD weight for all treatment 1971 9–11, 9–12, 12–15, or 9–15. days) Decreased fetal weight (all LOAEL 5 5 b treatment days) [BMD10 5 2.1; BMDL 5 1.5] b Increased malformations (GD LOAEL 5 10 [BMD10 5 9.2; 9–15 treatment only) BMDL 5 8.4] aDoses were given in two divided doses and the values are presented as total daily dose. b Benchmark dose estimates performed using EPA software v1.3.2. The BMD10 is the benchmark dose associated with a 10% effect, estimated from a curve fit to the experimental data. The BMDL represents the dose associated with the lower 95% CI around this estimate. Benchmark doses are used commonly in a regulatory setting; however, they are used in this study when the underlying data permit their calculation, and are only supplied to provide one kind of description of the dose-response relationship in the underlying study. Calculation of a benchmark dose in this study does not mean that regulation based on the underlying data is recommended, or even that the underlying data are suitable for regulatory decision making. O 10.1002/bdrb DOI AMPHETAMINE AND METHAMPHETAMINE 569 decreased in all treatment groups, but there were no (Ramirez et al., 1979). Malformations (cyclopia and significant effects on resorptions. [The Expert Panel exencephaly) were increased in rabbits i.v. dosed with calculated a BMD10 of 2.1 mg/kg bw/day and BMDL of 1.5 mg/kg bw/day methamphetamine on GD 12–15 1.5 mg/kg bw/day for fetal body weights on GD 9–15.] (Kasirsky and Tansy, 1971). The only significant increases in malformations (exence- Sheep in the third trimester of pregnancy have been phaly, cleft palate, microphthalmia, and anophthalmia) given i.v. doses of methamphetamine to evaluate effects occurred in fetuses from the group treated with 10.0 mg/ on maternal and fetal hemodynamics, oxygenation, and kg bw/day methamphetamine on GD 9–15 (rate was 13.6 acid-base balance. Treatment with 1.0–1.25 mg/kg bw in vs. 1% in either control group). [The Expert Panel these studies has been associated with elevations of estimated a BMD10 of 9.2 mg/kg bw/day and BMDL of maternal and fetal blood pressure and decrements in 8.4 mg/kg bw/day for malformations/live implant.] The fetal oxygenation consistent with impaired uteroplacen- Expert Panel noted that this study is relevant for tal perfusion (Burchfield et al., 1991; Stek et al., 1993, evaluation of abuse scenarios but is limited by incom- 1995; Dickinson et al., 1994). The Expert Panel deter- plete reporting and lack of litter-based analyses. mined that the studies in sheep were well conducted, but Yamamoto et al. (1992), administered mice (n 5 10–26/ their application for evaluation of human toxicity is group) a single i.p. dose of methamphetamine HCl uncertain. [purity not indicated] in saline on GD 8 (plug day 5 GD 0) A number of studies reported non-neurological post- at dose levels of 0, 11, 13, 14, 15, 17, 19, or 21 mg/kg bw. natal effects after gestational exposure to d-amphetamine Larger numbers of animals were used for the higher dose or methamphetamine in rats. The only study with oral groups in anticipation of treatment-induced maternal exposure was judged to be uninterpretable by the Expert death, which occurred in 3 of 16 dams at 15 mg/kg bw, 5 Panel due to numerous design and analyses limitations of 14 dams at 17 mg/kg bw, 6 of 17 dams at 19 mg/kg bw, (Ching and Tang, 1986). The remaining studies examined and 13 of 26 dams at 21 mg/kg bw. Maternal effects in offspring of rats dosed s.c. during gestation feed consumption and weight gain were not reported. with multiple dose levels. Although the route differs Fetal body weights and mortality were not affected from human exposures that typically occur orally or i.v. by treatment. External and skeletal malformations and limitations were noted in most studies (e.g., analysis (visceral malformations not examined) occurred at on a per fetus vs. per litter basis), the studies do provide doses Z14 mg/kg bw. [The Expert Panel calculated qualitative information and allow a dose–response BMD10 of 14.6 for fetal death, 16.1 for external assessment. Multiple-dose studies examining postnatal malformations, and 15.5 for skeletal malformations; development in rats are summarized in Table 44. Studies BMDLs were estimated at 10.3 for fetal death, 12.2 with d-amphetamine provided evidence of decreased for external malformations, and 12.4 for skeletal litter size at a dose Z0.5 mg/kg bw/day and increased malformations.] The Expert Panel noted that the i.p. pup mortality at doses Z2 mg/kg bw/day. Methamphe- dose during pregnancy is not relevant for human tamine studies suggested that types and magnitude of therapeutic exposures due to (1) possible direct adverse effects can vary according to period of gestational effects of the drug or solvent on the uterus and exposure, so a direct comparison of studies is difficult. secondarily on embryo-fetal development, and (2) the The studies suggest, however, that prenatal methamphe- possibility that the drug might be physically transported tamine exposure can result in decreased litter size directly to the embryo/fetus, bypassing maternal absorp- (Z10 mg/kg bw/day), delayed eye opening (Z3 mg/kg tion, metabolism, and distribution. The result interpreta- bw/day), reduced postnatal body weight gain (Z3 mg/ tion is very limited by unconventional group housing kg bw/day), and increased stillbirth or postnatal of dams. mortality (Z20 mg/kg bw/day). One study demon- Other developmental toxicity studies included only a strated that increases in stillbirths and postnatal mortal- single dose level. Many of the studies dosed the animals ity are greater with late- (GD 13–18) versus mid- (GD 7– by s.c. or i.p. injection, whereas humans typically are 12) gestational exposures (Acuff-Smith et al., 1996). exposed by oral or i.v. routes, and most studies Reductions in the number of dams delivering litters had significant methodological limitations, such as was observed at Z4.5 mg/kg bw/day, and increases in inadequate statistical analysis. The studies are briefly eye defects were noted at Z30 mg/kg bw/day. A delay described, however, because they provided qualitative in testicular descent was noted in one study where pups information. No increase in malformations, prenatal had other evidence of development delays (Cho et al., mortality, or adverse effects on body weight gain were 1991), but none of the reliable studies reported delays in observed in offspring of rats s.c. dosed with 2 mg/kg vaginal opening. A more detailed explanation of study bw/day d-amphetamine on GD 12–15 (Adams et al., details and results, LOAEL, benchmark dose values, and 1982). An increase in dead fetuses was observed in mice complete references are included in Table 44. gavage dosed with 50 mg/kg bw/day amphetamine The Expert Panel reviewed 37 rat studies with throughout pregnancy, but there was no increase in developmental neurotoxicity endpoints, which are sum- malformations (Yasuda et al., 1965). Mortality and marized in Table 39. Prenatal studies involved s.c. decreased body weight were observed in amphetamine- administration of amphetamine or methamphetamine treated dams in this study. An increase in malformations to dams during various time periods in gestation. (cardiac, cleft lip, eye) was observed in mice i.p. dosed The treatments in these studies resulted in signifi- with 50 mg/kg bw d-amphetamine on GD 8 (Nora et al., cant decreases in maternal weight gain associated 1965, 1968). There were no effects on estrous cyclicity with a decrease in maternal feed intake. The use of or ovulation, but there was increased sexual receptivity pair-fed control animals was unusual (Acuff-Smith in offspring of rats s.c. injected with 0.5 mg/kg bw/day et al., 1996). Some studies reported adverse effects of d,l-amphetamine during the entire gestation period treatment on pup birth weight (Monder, 1981) or viability

Birth Defects Research (Part B) 74:471–584, 2005 DOI 10.1002/bdrb 7 OU TAL. ET GOLUB 570 Table 44 Summary of Multiple-Dose Experimental Animal Pre- and Postnatal Developmental Toxicity Studies

Maternal Effect Level, mg/kg Critical Developmental Developmental Effect Level, Species/Strain Exposure bw or mg/kg bw/day Effects mg/kg bw/day Reference

Sprague–Dawley d-Amphetamine. s.c. 0, 2, or Not studied Increased pup mortality LOAEL 5 2 Hitzemann et al., rat 6 mg/kg bw/daya on GD 5 1976 through parturition b Sprague–Dawley d-Amphetamine. s.c. 0, 0.5, or NOAEL 5 2 Decreased pups/litter at birth LOAEL 5 0.5 [BMD10 5 2.0; Vorhees, 1985 rat 2.0 mg/kg bw/day on GD BMDL 5 1.1] 12–15 Sprague–Dawley Methamphetamine s.c. 0, 2, 6, LOAEL 5 2 (decreased gesta- Decreased litter size and Pair-wise comparison with the Martin, 1975 rat or 10 mg/kg bw/daya GD tion length and body delayed eye opening control group was not 1–21 weight gain). No litters reported. delivered in 4 of 7 dams at high dose Wistar rat Methamphetamine. sc 0, 1, 2, LOAEL 5 2 (decreased body Decreased male pup body LOAEL 5 3 [Testicular des- Cho et al., 1991 3, or 4.5 mg/kg bw/day on weight gain). Only 1 litter weight gain during lacta- cent: BMD10 5 3.8, GD 7–20 delivered at 4.5 mg/kg bw/ tion and post-weaning; BMDL 5 3.3; incisor erup- day delayed testicular descent, tion: BMD10 5 5.1, incisor eruption, and eye BMDL 5 3.1; eye opening: opening BMD10 5 15.7, BMDL 5 3.2]a Sprague–Dawley Methamphetamine. sc 0, 10, LOAEL 5 10 (decrease in Decreased litter size (GD LOAEL 5 10 [BMD10 5 38; Acuff-Smith et al.,

it eet eerh(atB 44154 2005 74:471–584, B) (Part Research Defects Birth rat 20, 30, or 40 mg/kg bw/ body weight) [BMD10 5 91; 13–18 exposure) BMDL 5 20] 1996 daya on GD 7–12 or GD 13– BMDL 5 56 for GD 13–18 Increased stillbirth and post- LOAEL 5 20 [Stillborn/pup: 18. exposure]; dam mortality natal mortality on PND 1–3 BMD10 5 36, BMDL 5 31; increased at 30 (GD 13–18 exposure) postnatal mortality/live- [BMD10 5 27; BMDL 5 20 born pup: BMD10 5 53, for GD 8–13 exposure] BMDL 5 40] Increased stillbirth and post- LOAEL 5 40 [Stillborn/pup: natal mortality on PND 1–3 BMD10 5 91, BMDL 5 58; (GD 7–12 exposure) postnatal mortality/live born pup: BMD10 5 48, BMDL 5 40] Decreased post-weaning off- LOAEL 5 30 spring body weights (GD 13–18 exposure) Increased pups with LOAEL 5 30 (not statistically anophthalmia or micro- significant); [BMD10 5 48, phthalmia (GD 7–12 BMDL 5 42] exposure). O 10.1002/bdrb DOI aDoses were given in two divided doses and the values are presented as total daily dose. b Benchmark dose estimates performed using EPA software v1.3.2. The BMD10 is the benchmark dose associated with a 10% effect, estimated from a curve fit to the experimental data. The BMDL represents the dose associated with the lower 95% CI around this estimate. Benchmark doses are used commonly in a regulatory setting; however, they are used in this study when the underlying data permit their calculation, and are only supplied to provide one kind of description of the dose-response relationship in the underlying study. Calculation of a benchmark dose in this study does not mean that regulation based on the underlying data is recommended, or even that the underlying data are suitable for regulatory decision making. AMPHETAMINE AND METHAMPHETAMINE 571 (Hitzemann et al., 1976; Acuff-Smith et al., 1992, 1996). exposure to amphetamine or methamphetamine. The Results of behavioral studies after prenatal exposures studies are summarized in Table 40. With the exception have not been consistent among studies, but generally of a drinking water exposure study (Tonge, 1973), suggest developmental delay (Martin, 1975; Vorhees, exposures occurred through parenteral routes, mainly 1985; Cho et al., 1991; Acuff-Smith et al., 1992) and sc. Treatment of preweanling animals is associated with increased motor activity (Hitzemann et al., 1976; Nasello persistent changes in behavior, with little anatomic and Ramirez, 1978; Adams et al., 1982; Acuff-Smith et al., alteration in the brain. Some studies reported a transient 1992, 1996). One study (Acuff-Smith et al., 1996) found increase in prefrontal cortex neuronal density in 14-day- prenatal exposure to methamphetamine to be associated old rats exposed to 10 mg/kg bw/day amphetamine on with impairment of learning and memory in adult GD 8–22 (Tavares et al., 1996) and reduced hippocampal offspring. These studies did not use ranges of doses volume formation in 30-day-old male rats exposed to and were not suitable for dose–response modeling. The 25 mg/kg bw/day amphetamine on PND 1–30 (Tavares lowest effect level reported with gestational treatment of and Silva, 1993). Increases in dendritic length and the dam was 0.5 mg/kg bw/day for both d-amphetamine branching and spine density of pyramidal cells of the (Adams et al., 1982) and d,l-amphetamine (Nasello et al., prefrontal cortex were reported in 90-day-old male 1974; Nasello and Ramirez, 1978). gerbils that received 50 mg/kg bw methamphetamine Studies involving direct treatment of rat neonates used on PND 14 (Blaesing et al., 2001). Behavioral effects were methamphetamine administered s.c. on PND 1–10 or not reported in these studies. PND 11–20. Both administration periods have been Most of the studies examining biochemical effects associated with a decrease in pup weight compared to focused on measurement of monoamine levels in the vehicle treated controls. The decrement in pup weight brain. Studies of both amphetamine and methampheta- persisted until at least PND 42 and in some cases until mine in rats and mice demonstrated variable effects PND 70. Pup mortality during the treatment period was on brain levels of dopamine, norepinephrine, and also increased by neonatal methamphetamine treatment serotonin. Direct comparison of studies examining in many of the studies. Treatment of neonates was monoaminergic endpoints is precluded by differences associated with increased reactivity on acoustic startle in species, dose level and duration, developmental testing and with deficits in associative processes and period of exposure, age of evaluation, and brain region memory when animals were tested as adults. Single-dose examined. No obvious patterns were noted. A series of studies were the rule and dose–response modeling studies from Crawford et al. (2000a,b, 2003) examined was not possible. The lowest effective dose reported amphetamine and methamphetamine effects on protein for behavioral alterations after neonatal treatment with kinase A, a critical enzyme in signal transduction d-methamphetamine was 30 mg/kg bw/day (Vorhees cascades of several different receptors, including dopa- et al., 1999). Behavioral deficits in adults after neonatal mine receptors. treatment are due to impairments in reference (long- Cardiovascular defects, particularly ventricular septal term) memory and not to general motor or cognitive defects (VSD), have been reported in a variety of study impairment or to deficits in working (short-term) types reviewed by the Expert Panel. In these studies, memory, and have been attributed to alterations in amphetamine was administered during organogenesis. hippocampus development that would be expected to Early single-dose studies (Nora et al., 1965, 1968) correspond to developmental events in the third trime- administering d-amphetamine i.p. to mice on GD 8 ster of human pregnancy. Treatment of young rats with found a statistically significant increase in VSD in amphetamine or methamphetamine is also associated fetuses. This effect was demonstrated in two strains of with behavioral sensitization to subsequent challenge mice. Another single-dose study (Fein et al., 1987) with amphetamine or methamphetamine. This sensitiza- administered 50 or 100 mg/kg i.p. on GD 9–11 to mice. tion can persist for months after the last treatment This study identified amphetamine effects on fetal EKG (reviewed in Section 2.5.3). (longer Q-T intervals) on GD 19 and delayed develop- There were three reports in which methamphetamine ment of cardiomyocytes as determined histologically on (Martin et al., 1976; Martin and Martin, 1981) or GD 19. Studies in chick embryos (Cameron et al., d-amphetamine (Monder, 1981) were given during 1983) also demonstrated structural (including VSD) and gestation and continued through the lactation period. functional (blood pressure) effects of amphetamine The two methamphetamine reports, which present exposure during organogenesis. FDA summaries (FDA, different endpoints from the same experiment, involved 2001) of data from developmental toxicity studies using s.c. treatment of dams. The amphetamine study used gavage administration of 2 and 6 mg/kg amphetamine administration in drinking water [probably resulting in (3:1 d-tol-amphetamine ratio) in rats and rabbits direct treatment of pups during the last week of the reported several individual fetuses with VSD defects, lactation period]. Adverse effects on dam and pup body one rat in the 6 mg/kg dose group, three rabbits in the weight were noted in these studies, and both studies 2 mg/kg dose group, and three rabbits in the 6 mg/kg showed developmental delay and increased motor dose group. Although all of these individual studies have activity in offspring, consistent with the studies that limitations, together they suggest that the heart is a target used only gestational exposures. Effective doses were organ of concern for embryonic exposure to ampheta- 2 mg/kg bw/day d-amphetamine [the lower of 2 dose mine. Data presented in a Letter to the Editor addressed levels used] and 10 mg/kg bw/day methamphetamine cardiovascular effects of prenatal amphetamine exposure [the only dose level used]. in humans (Nora et al., 1970) and reported increased risk The Expert Panel reviewed 17 experimental animal for congenital heart disease. Limitations of this report studies evaluating persistent anatomical or biochemical included limited detail on study design and population, changes in the brains of rodents after prenatal or juvenile lack of information on other exposures, such as ethanol,

Birth Defects Research (Part B) 74:471–584, 2005 DOI 10.1002/bdrb 572 GOLUB ET AL. and lack of information on the reason for the amphetamine exposure. Expert Panel Conclusions-Methamphetamine There are no interpretable human data on methamphetamine developmental toxicity. Expert Panel Conclusions-Amphetamine Data are sufficient to demonstrate developmental Human data are insufficient for an evaluation of toxicity of d,l-methamphetamine manifested as the developmental toxicity of amphetamine after abnormal neurobehavioral testing in rats exposed prenatal exposure. There are two studies that in- during gestation to maternal doses of 2 mg/kg bw/ dicated no statistically significant increase in the day throughout pregnancy or GD 7–20 for metham- overall rate of major or minor congenital malforma- phetamine (unspecified enantiomer) and 5 mg/kg tions, although the data do not allow an evaluation of bw/day for d-methamphetamine. There is insuffi- the risk of individual specific malformations. A study cient evidence to conclude that there are morpholo- on birth weight reported an effect of amphetamine gical effects of methamphetamine treatment on the therapy but did not control for alcohol use and other brains of these animals. confounds. There is sufficient evidence that methampheta- Data are insufficient for an evaluation of amphe- mine produces developmental toxicity in rats man- tamine effects on growth in children and adoles- ifested as behavioral alterations after treatment of cents. Growth studies in these children demonstrate neonates with 30 mg/kg bw/day. an association of reduced growth and amphetamine There is sufficient evidence that methampheta- treatment; however, a causal association with the mine produces developmental toxicity in rats man- medication is not possible due to a lack of control of ifest as decreased pup weight at maternal doses of potential confounding factors. These potential con- 3 mg/kg bw/day s.c. on GD 7–20 and decreased litter founders could be causing the observed growth size at maternal doses of 10 mg/kg bw/day s.c. on GD effects. 13–18. A number of studies examining fetal effects of Data are insufficient to evaluate whether amphe- gestational methamphetamine administration by the tamine therapy alters the risk of tobacco use, s.c. route in rats reported effects on litter size and problematic alcohol consumption, and illicit sub- postnatal pup weight. stance abuse in adolescents and adults. The experimental animal data are assumed relevant Data are insufficient to conclude whether amphe- to humans. tamine treatment of children at standard therapeutic Note: The definitions of the term sufficient and the doses increases the risk of tics or movement terms assumed relevant, relevant, and not relevant are disorders. in the CERHR guidelines at http://cerhr.niehs.nih.- Data are sufficient to demonstrate developmental gov/news/guidelines.html. toxicity of d and d,l-amphetamine manifested as abnormal neurobehavioral testing in rats exposed during gestation to maternal doses of 0.5 mg/kg bw/ day throughout pregnancy or 2 mg/kg bw/day on GD 4.0 REPRODUCTIVE TOXICITY DATA 12–15. Thereisinsufficientevidencetoconcludethat there are morphological effects of amphetamine treat- 4.1 Human Data ment on the brains of these animals. The experimental No human data were located on reproductive effects animal data are assumed relevant to humans. of amphetamines. The Expert Panel is aware of reports There are sufficient data to conclude that i.p. of sexual dysfunction in amphetamine addicts injection of amphetamine in pregnant mice at doses (Greaves, 1972; Angrist and Gershon, 1976). These of 50 mg/kg bw/day GD 9–11 increases the incidence reports are largely of an anecdotal nature and not of malformations in the offspring. Malformation suitable for an evaluation of amphetamine-related data on other species or other routes were not reproductive effects. available for the Expert Panel to review. The mouse data were not considered relevant to humans because 4.2 Experimental Animal Data of the use of i.p. dosing during pregnancy. Drugs 4.2.1 Female reproduction. Ramirez et al. (1979), administered i.p. in pregnancy can (1) have a possible funding not indicated, examined the effects of prenatal direct adverse effect on the uterus and secondarily on amphetamine exposure on estrous cyclicity, sexual embryo-fetal development, and (2) possibly be phy- behavior, and hypothalamic monoamine levels in rats. sically transported directly to the embryo/fetus, Because the exposure was prenatal, this study assesses bypassing maternal absorption, metabolism and developmental toxicity and is discussed in Section distribution. 3.2.1.1. Several rat and mouse studies reported effects on 4.2.2 Male reproduction. Cates and Jozeforicz fetal or neonatal viability, but the evidence is not (1970), funding not indicated, reported that 200 mg/mL sufficient to permit conclusions due to limitations of d-amphetamine had no effect on in vitro motility of of the studies. frog (Rana pipiens) sperm. [Data were not shown.] The experimental animal data are assumed relevant Strengths/Weaknesses: Weaknesses of this study to humans. include the use of a single dose, a non-mammalian Note: The definitions of the term sufficient and the species, and an in vitro design. No rational was given for terms assumed relevant, relevant, and not relevant are selection of test concentration. in the CERHR guidelines at http://cerhr.niehs.nih.- Utility (Adequacy) for CERHR Evaluation Process: gov/news/guidelines.html. This study is not useful in the evaluation process.

Birth Defects Research (Part B) 74:471–584, 2005 DOI 10.1002/bdrb AMPHETAMINE AND METHAMPHETAMINE 573 Schneiden (1974), funding not indicated, evaluated the determine the power of the study to detect an adverse in vitro effect of d-amphetamine exposure on rabbit effect of treatment. In addition, insufficient experimental sperm motility as compared to a glycine acetate control. detail is given to adequately assess outcomes (e.g., no Ten control samples and five or six samples per d- data on clinical signs, body weight, etc.). amphetamine concentration were evaluated. Sperm Utility (Adequacy) for CERHR Evaluation Process: motility was inhibited at a d-amphetamine concentration The utility of this study is limited. of 500 mg/mL, but there was no effect at 25 mg/mL. Yamamoto et al. (1999), funding not indicated, [Although it was stated that sperm motility was examined the effects of methamphetamine exposure on evaluated according to the method of Emmens, the reproductive toxicity in male mice. Eight-week-old male results presentation was somewhat unclear, making it ICR mice were given a single i.p. injection of difficult to verify the author’s conclusion.] d-methamphetamine HCl [purity not specified] in saline Strengths/Weaknesses: Strengths of this study include at 0 (n 5 30), 3.75 (n 5 20), 7.5 (n 5 20), or 15 (n 5 60) mg/ the identification of a dose–response relationship and the kg bw. Authors mention that 15 mg/kg bw is about 10 use of standard protocols. Weaknesses include no times the maximal dose in Japanese abusers of metham- rationale given for selection of test concentrations, small phetamine. Twenty-four hours after injection, mice were sample size, and insufficient experimental detail (e.g., paired 1:1 with untreated female mice until a plug was age and body weights, order of sample collection vs. detected or for 14 days. Dams were allowed to litter and treatment and analysis, maintenance of samples at a at birth, litter size was noted, and pups were weighed constant temperature, etc.). and examined for external malformations. The same Utility (Adequacy) for CERHR Evaluation Process: mating procedure was conducted in half the mice from This in vitro study can be used as supplemental the 15 mg/kg bw/day group 48 hr after injection. information in the evaluation process. Additional mice (5–7/group) treated with saline or Larez et al. (1979) examined dominant lethality in rats 15 mg/kg bw d-methamphetamine were evaluated for treated with d-amphetamine. Male Sprague–Dawley rats testicular and epididymal weight, serum testosterone were gavage dosed with saline (n 5 10) or 2 mg/kg bw/ level, sperm motility, sperm morphology, testicular and day d-amphetamine sulfate (n 5 8) [purity not specified] epididymal histology, and serum methamphetamine for 5 days. Males were mated to three virgin rats per and amphetamine levels. The number of matings result- week over a period of 5 weeks. On GD 14, the females ing in vaginal plugs was evaluated by w2 test. Data were killed for an examination of corpora lutea, resorp- on litter-based mortality was evaluated by Wilcoxon– tions, and implants. Mutagenic index was calculated by Mann–Whitney test. Other data were evaluated by determining the percentage of early fetal deaths relative Student t-test. to total implants. Statistical significance was determined Methamphetamine serum levels peaked at 7.5 mg/mL using the method of binomial distribution of proportions [7500 ng/mL] at 5 min and amphetamine levels peaked at and percentages. Mutagenic index was significantly 0.5 mg/mL [500 ng/mL] at 15 min. Serum levels of both increased at week 3 (Po0.05), and borderline signifi- compounds were below the detection limit 24 hr later, the cance (P E 0.05) was obtained for increases occurring time period between treatment and test initiation. during Weeks 2 and 4. The study authors concluded that Clinical signs in the 15 mg/kg bw/day group included d-amphetamine was mutagenic under the conditions of hyperlocomotion and salivation, which peaked 15 min this study. after injection. About 37% of the animals in the 15 mg/kg Strengths/Weaknesses: Although typically performed bw/day group died within 20 hr of treatment. The as a mutagenesis assay, this study provides information numbers of vaginal plugs and births were significantly on male reproductive toxicity by looking at pregnancy reduced in the 15 mg/kg bw/day group mated 24 hr outcome when males are treated before mating. Strengths after treatment, but the effects were not observed 48 hr of the study include an oral route of administration with after treatment. [In Table 1 of the study, it is not clear if vehicle control. A weakness is use of a mating period the numbers of plugs and births are expressed in terms shorter than the sperm cycle, limited pregnancy outcome of the number of animals mated and evaluated in each endpoints (only resorptions), limited detail on methods group. The number of animals listed in Table 1 is the and results, and lack of litter-based statistics. same as the number of animals that were said to be Utility (Adequacy) for CERHR Evaluation Process: treated in the methods section. The death of 37% of This study has limited utility for the evaluation process. males in the 15 mg/kg bw/day group suggests that the Kasirsky and Tansy (1971), supported by NIH, numbers mated would be well below the numbers evaluated the offspring of six male rabbits i.v. treated treated, unless the numbers given in the methods with methamphetamine HCl in saline at doses of 0, 1.5, section were for total evaluated instead of treated.] 3.0, or 5.0 mg/kg bw/day for 3 months before mating. Weights of testes and epididymides were described as [Very limited protocol details were provided, but it is ‘‘slightly lower’’ in the 15 mg/kg bw/day group 24 hr assumed that procedures were similar to those con- after treatment, but it is not clear that statistical testing ducted in female rabbits discussed in Section 3.2.1.1.] was performed [t-test by CERHR showed no significant After mating with treated males, untreated female difference in weights of the caudae epididymides or rabbits were killed on GD 30 for examination of fetuses. left testes. There was a significant 7% decrease in right There were no significant effects on whole litter resorp- testis weight in the 24-hr 15 mg/kg bw group. Organ tions, offspring survival, malformations, or fetal weight. weights were not measured at lower methamphetamine Strengths/Weaknesses: Strengths of this study are the doses. Variances were not specified, but were assumed route and duration of exposure, which are realistic for to be SEM.] Sperm motility was lower in males treated drugs of abuse. A weakness is the lack of clear with 15 mg/kg bw/day and examined 24 and 48 hr later. information on sample size. Sample size is necessary to The text identifies sperm motility as 57 and 62% of

Birth Defects Research (Part B) 74:471–584, 2005 DOI 10.1002/bdrb 574 GOLUB ET AL. control values at 24 and 48 hr post-treatment with 15 mg/ fifth test. In treated rats, number of mounts was kg methamphetamine. Serum testosterone level was increased only during the third testing and mount higher in the 15 mg/kg bw/day males examined at latency was increased only during the fourth testing. 24 hr, but lower in males examined at 48 hr. There were Body weight gain was not affected. Based on these no significant effects on litter size, pup body weight, findings, the study authors concluded that methamphe- sex ratio, postnatal mortality, sperm morphology, or tamine inhibits intromission and ejaculation in testicular or cauda epididymal histology at any dose rats. [Based on proportion of males ejaculating, the level. NOAEL is 2 mg/kg bw and the LOAEL is 4 mg/kg bw, Strengths/Weaknesses: This study provides a good according to the authors’ results. Benchmark dose spectrum of information on male reproductive effects. calculations by CERHR using the EPA Benchmark The authors rightly concluded that there was very little Dose Software give a BMD10 of 2.0 mg/kg bw, and a difference between the dose that affected reproductive BMDL of 1.1 mg/kg bw.] behavior and the dose that was toxic to the animal. Strengths/Weaknesses: A strength is that authors used Weaknesses are the limitations in endpoint information estradiol- and progesterone-treated females to avoid and analysis, use of the i.p. route, exceeding of MTD differences in estrous cyclicity and/or receptivity. Sam- at the high concentration, no data on male body weights, ple sizes were sufficient in Experiment 1, but small for use of inexperienced breeders, no indication that female behavioral assessments in Experiment 2. A weakness is mice were examined to verify normal estrous cyclicity, that rats were dosed with amphetamine by i.p. injection, no verification of female fertility, no confirmation of whereas humans typically are exposed by oral or i.v. mating by vaginal lavage samples, no measurement of routes. The study would have been stronger if observa- sperm motility in animals mated 48 hr after treatment, tional data were collected without knowledge of the and no definition of sperm motility parameters (e.g., treatment groups. In Experiment 2, it is not clear when number evaluated, number of microscopic fields). copulatory behavior was tested in relation to the weekly Weaknesses associated with testosterone analyses are doses of methamphetamine. Motor activity was not no indication that diurnal variations were considered, examined in the rats dosed for 8 weeks (1 Â /week) with insufficient sample size (n 5 5) considering variations in methamphetamine. This seems like an omission given this measurement, and analysis only in high-dose that stereotypic behavior was observed at the fifth animals. testing. Utility (Adequacy) for CERHR Evaluation Process: Adequacy/Utility for CERHR Evaluation Process: This study is of limited adequacy for use in the This study is of limited adequacy for the evaluation evaluation process. process. Saito et al. (1991), funding not indicated, examined 4.2.3 Mating studies in concurrently treated copulatory behavior in 10-week-old male Wistar–Imami- males and females. The FDA Pharmacology Review chi rats exposed to methamphetamine HCl [chirality and for Adderall (FDA, 2001) summarized an oral fertility purity not indicated]. In the first of two experiments, and early embryo development study in Sprague– male rats were given a single i.p. dose of methamphe- Dawley rats, submitted to the agency as part of the tamine in distilled water at 0, 1, 2, or 4 mg/kg bw and approval process. Male and female rats (22/sex/dose immediately tested for copulatory behavior with a group) were gavaged with free amphetamine base in sexually receptive female (n 5 6/group) or spontaneous water at total daily dose levels of 0, 2, 6, or 20 mg/kg bw motor activity (n 5 7–10/group). Copulatory behavior in 2 equal divided doses 8 hr apart [the Panel assumes data were analyzed by Fisher exact probability test and that the test article was d-,l-amphetamine in a 3:1 ratio, the Mann–Whitney U-test, whereas motor activity data as in the marketed product]. Males were treated for 29 were analyzed by the Duncan multiple t-test. In the high- days before mating and females were treated for 15 days dose group (4 mg/kg bw), the number of mounts, before mating. Males and females were paired for up to 3 intromissions, and ejaculations over a 90-min period weeks with continuation of treatment throughout the was significantly reduced compared to the control group. mating period and until GD 7 in females. Males were Compared to a 100% rate in controls, three males in the treated for a total of 8 weeks after which they were killed 4 mg/kg bw group did not ejaculate and two of those and reproductive organs weighed and examined macro- males had no intromissions. Frequency of spontaneous scopically. Females were killed on GD 14 and uterine motion was significantly increased at 2 and 4 mg/kg bw, contents evaluated. and stereotypic behaviors (e.g., compulsive gnawing and Dose-related clinical signs occurred in the middle- and sniffing) were observed in the 4 mg/kg bw group. In the high-dose groups, with few and transient clinical signs in second experiment, rats were i.p. administered 0 (n 5 7) the low-dose group. Body weight was dose-dependently or 1 (n 5 5) mg/kg bw methamphetamine HCl once/ decreased in the middle-and high-dose group as well. No week for 8 weeks. Copulatory behavior during a 90-min effects of treatment were seen on estrous cyclicity, period was observed five times at 2-week intervals. mating, fertility, live young, resorptions, or pre- or A significant reduction in percentage of rats ejaculating post-implantation loss. Absolute but not relative male over a 90-min period reached statistical significance reproductive organ weight changes occurred in high- during the fourth and fifth testing. None of the treated dose males. [Absolute epididymal weight, which is rats ejaculated during the 90-min period of the typically conserved in the presence of moderate body fifth testing, whereas all of the control rats ejaculated. weight decrements (O’Connor et al., 2000), was Rats were housed overnight and a vaginal plug decreased by 6% in the high-dose males; it is not was found for all females the next morning, thus known whether this value was statistically identified.] indicating delayed ejaculation. Percentage of treated rats In an accompanying pharmacokinetic study after B21 intromitting was also significantly reduced during the days of dosing (males) or B7 days of dosing (females)

Birth Defects Research (Part B) 74:471–584, 2005 DOI 10.1002/bdrb AMPHETAMINE AND METHAMPHETAMINE 575 with 20 mg/kg bw/day in a twice daily gavage regimen, In a developmental toxicity study that is described in blood samples were collected after the first daily dosing. detail in Section 3, rats were s.c. injected with 0 or The Cmax in males was 880–976 ng/mL and Cmax in 0.5 mg/kg bw/day d,l-amphetamine over the entire females was 1081–1197 ng/mL. The mean AUC24 value gestation period (Ramirez et al., 1979). Female offspring in males was 2139 ng-h/mL and in females was 4909 ng- exposed to amphetamine in utero were more sexually h/mL. At the 20 mg/kg bw/day dose, the mean AUC24 receptive, as determined by lordosis response, but in males was 5689 ng-h/mL and in females was 8506 ng- exhibited no effects on estrous cyclicity or ovulation. h/mL. The FDA reviewer noted for comparison pur- In a rat dominant lethality test, percent early fetal poses that children receiving 30 mg/day amphetamine deaths relative to total implants was significantly therapy have a mean AUC24 of 1800 ng-h/mL. The total increased in fetuses sired by males treated with 2 mg/ exposure at the high dose in the fertility study was only kg bw/day d-amphetamine sulfate and mated 3 weeks three times that of children on therapy for male rats and after dosing; borderline significance was obtained at 2 only five times that of children on therapy for female and 4 weeks after dosing (Larez et al., 1979). rats. The FDA reviewer called attention to the use of a The FDA (2001) reviewed a fertility study conducted different dosing regimen in the range-finding study with d,l-amphetamine in rats. The study was not (once daily dosing) versus the definitive study (two available to the Expert Panel, but is described in this divided doses/day), and the use of non-pregnant section, due to the paucity of amphetamine reproductive animals for the range-finding and the pharmacokinetic toxicity data. In that study, male and female rats (22/sex/ studies. dose group) were gavaged with free amphetamine base Strengths/Weaknesses: This is a full fertility study in water at total daily dose levels of 0, 2, 6, or 20 mg/kg using standard methods. Strengths are sufficient group bw administered in two doses divided equally 8 hr apart. sizes, appropriate controls, and sufficient duration of Males were treated for 29 days before mating and dosing before breeding. The dosing paradigm (oral throughout a 3-week mating period. After treatment, administration in two doses divided equally 8 hr apart) males were killed for weighing and macroscopic evalua- was designed to better mimic clinical exposure patterns. tion of reproductive organs. Females were treated for 15 Blood samples were collected before mating for toxico- days before mating, during the 3-week mating period, kinetic evaluations. The lack of detailed data is a and through GD 7. Females were killed on GD 14 for weakness (i.e., no data or mean values for estrous cycles, evaluation of uterine contents. Reduction in body weight time to mating, organ weights, etc.). The high dose and clinical signs occurred in the middle- and high-dose exceeded the MTD as determined by body weights and groups. Amphetamine treatment had no effect on estrous body weight gains. Pair-fed controls were not included. cyclicity, mating, fertility, live young, resorptions, or pre- Statistical analyses were not described nor were statis- or post-implantation loss. Absolute but not relative male tical results presented. Male reproductive organs were reproductive organ weight changes occurred in the high- weighed and examined macroscopically rather than by dose males. Based on preliminary pharmacokinetic histopathological examination. As stated in the report studies in rats, the FDA estimated that respective total summary, the safety margins provided by this study exposures in high-dose male and female rats were three were minimal, with mean total exposures in rats equal to and five times that of children on therapy. 3–5 times the levels measured in children given 30 mg, Saito et al. (1991) dosed male Wistar–Imamichi rats the maximum clinical dose. with a single i.p. dose of methamphetamine at 0, 1, 2, or Utility (Adequacy) for CERHR Evaluation Process: 4 mg/kg bw and immediately tested copulatory behavior This study could not be used in the evaluation process with a sexually receptive female. At 4 mg/kg bw, the because a full report of the study was not available for number of mounts, intromissions, and ejaculations over a review. 90-min period was significantly reduced compared to the control group. Frequency of spontaneous motion increased at 2 and 4 mg/kg bw, and stereotypic behavior 4.3 Utility of Data increased at 4 mg/kg bw group. In a second experiment, There are no data from controlled human studies with rats were i.p. administered 0 (n 5 7) or 1 (n 5 5) mg/kg which to evaluate possible reproductive toxicity of bw methamphetamine HCl once/week for 8 weeks. amphetamine or methamphetamine. The animal data Copulatory behavior was observed five times at 2-week are insufficient for an evaluation of possible reproduc- intervals. In the 1 mg/kg bw group, a decrease in tive toxicity after exposure to amphetamine or metham- percentage of rats ejaculating reached statistical signifi- phetamine. A one-generation fertility study of d,l- cance during the fourth and fifth testing and percentage amphetamine in rats was reviewed by the FDA, but the of treated rats intromitting was reduced during the fifth original study was not available to the Expert Panel. test. [Based on proportion of males ejaculating, the There are no methamphetamine studies that examine NOAEL is 2 mg/kg bw and the LOAEL is 4 mg/kg bw, fertility in female rats. Limited studies in male rats according to the authors results. Benchmark dose exposed to methamphetamine examine sexual function calculations by CERHR using the EPA Benchmark and fertility parameters. Dose Software version 1.3.2 gave a BMD10 of 2.0 mg/kg bw and a BMDL of 1.1 mg/kg bw.] Yamamoto et al. (1999), treated mice with a single i.p. 4.4 Summary of Reproductive Toxicity Data injection of d-methamphetamine HCl in saline at 0 4.4.1 Human data. No human data from controlled (n 5 30), 3.75 (n 5 20), 7.5 (n 5 20), or 15 (n 5 60) mg/kg studies were located. bw. [It is not clear if numbers are for total numbers 4.4.2 Experimental animal data. Repeat dosing treated or evaluated.] Twenty-four hours after injection, studies are summarized in Table 45. mice were paired 1:1 with untreated female mice until a

Birth Defects Research (Part B) 74:471–584, 2005 DOI 10.1002/bdrb 576 GOLUB ET AL. Table 45 Summary of Multiple-Dose Amphetamine and Methamphetamine Reproductive Toxicity Studies

Sex/Species/ Strain Drug/Exposure Critical Effect Effect Level Reference

Male and female d,l-Amphetamine. Gavage No effects on estrous cyclicity, NOAEL 5 20 mg/kg FDA, 2001 Sprague–Daw- (free base) 0, 2, 6, or 20 mg/ mating, fertility, live young, bw/day ley rat kg bw in 2 equal divided resorptions, or pre- or post- doses 8 hr apart for 29 days implantation loss. before mating through a 3- week mating period in males and 15 days before mating through GD 7 in females. Male Wistar- Methamphetamine HCl. i.p. 0, Decreased number of mounts, LOAEL 5 4 mg/kg bw Saito et al., Imamichi rat 1, 2, or 4 mg/kg bw, single intromissions, and ejacula- NOAEL 5 2 mg/kg 1991 dose.a tions over a 90-min period. bw/day [BMD10 of 2.0 mg/ kg bw, BMDL of 1.1 mg/kg bw] Male ICR mouse d-Methamphetamine HCl. i.p. Decreased vaginal plugs and LOAEL 5 15 mg/kg Yamamoto 0, 3.75, 7.5, or 15 mg/kg bw, births in mice mated 24 hr bw/day et al., 1999 single injection. after treatment, but not NOAEL (for vaginal 48 hr after treatment. plugs and Decreased sperm motility births) 5 7.5 mg/kg at 24 and 48 hr after treat- bw/day ment; increased serum testosterone level at 24 hr, but lower at 48 hr after treatment [Sperm and testosterone were only examined in animals treated with 15 mg/kg bw/ day]. No significant effects on litter NOAEL 5 15 mg/kg size, pup body weight, sex bw/day ratio, or postnatal mortality. Male rabbit Methamphetamine HCl. i.v. 0, No significant effects on NOAEL 5 5 mg/kg Kasirsky and 1.5, 3.0, or 5.0 mg/kg bw/ whole litter resorptions, bw/day Tansy, 1971 day for 3 months before offspring survival, malfor- mating. mations, or fetal weight. aResults of a single-dose study with multiple exposures over a period of weeks is discussed in the summary text.

plug was detected or for 14 days. The same mating no significant difference in weights of the caudae procedure was conducted in half the mice from the epididymides or left testis. There was a significant 7% 15 mg/kg bw/day group 48 hr after injection. Dams were decrease in right testis weight in the 24-hr 15 mg/kg bw allowed to litter and at birth, litter size was noted, and group. Variances were not specified but were assumed pups were weighed and examined for external mal- to be SEM.] formations. Clinical signs were observed in the 15 mg/kg Sperm motility was lower in males treated with bw/day group and about 37% of the animals died within 15 mg/kg bw/day and examined 24 and 48 hr later. 20 hr of treatment. The number of vaginal plugs and Serum testosterone level was higher in the 15 mg/kg births were significantly reduced in the 15 mg/kg bw/ bw/day males examined at 24 hours, but lower in males day group mated 24 hr after treatment, but the effects examined at 48 hr. The Expert Panel noted that repro- were not observed 48 hr after treatment. There were no ductive competency of untreated female mice was not significant effects on litter size, pup body weight, sex verified, and measurements of testosterone levels were ratio, or postnatal mortality at any dose level. Additional inadequate because factors such as diurnal variations mice (5–7/group) were treated with 0 or 15 mg/kg bw d- were not considered. methamphetamine and evaluated for testicular and In a study that provided limited protocol details, epididymal weight, serum testosterone level, sperm six male rabbits were i.v. treated with methamphetamine motility, and serum methamphetamine and ampheta- HCl in saline at doses of 0, 1.5, 3.0, or 5.0 mg/kg bw/day mine levels. Weights of testes and epididymides were for 3 months before mating (Kasirsky and Tansy, described as ‘‘slightly lower’’ in the 15 mg/kg bw/day 1971). The rabbits were mated with untreated females group 24 hr after treatment. [t-Test by CERHR showed that were killed on GD 30 for examination of

Birth Defects Research (Part B) 74:471–584, 2005 DOI 10.1002/bdrb AMPHETAMINE AND METHAMPHETAMINE 577 fetuses. There were no significant effects on whole litter 8–11. These data were not considered relevant to humans resorptions, offspring survival, malformations, or fetal due to use of i.p. dosing during pregnancy. Drugs weight. administered i.p. in pregnancy can (1) possibly have a direct adverse effect on the uterus and secondarily on embryo-fetal development, and (2) possibly be Expert Panel Conclusions–Amphetamine physically transported directly to the embryo/fetus, There are no human reproductive data. bypassing maternal absorption, metabolism, and distri- Experimental animal data are insufficient to bution The majority of other developmental toxicity evaluate male and female reproductive toxicity of studies used s.c. exposures to examine the developmen- amphetamine. tal effects of amphetamine. Data from several studies in Only one study was found in the open literature on mice and rats suggest that decreased fetal or neonatal male reproduction and one study on the effects of viability may occur with prenatal amphetamine expo- prenatal exposure on the female reproductive system. sure; however, this finding was inconclusive based on Reproductive toxicity studies submitted to FDA were limitations in the animal studies or excessive maternal not available in sufficient detail for review. toxicity. Although limited by parenteral administration of single dose levels in most studies, rat data are sufficient Expert Panel Conclusions-Methamphetamine to demonstrate developmental toxicity of d- and d, There are no human reproductive data. l-amphetamine. Toxicity is manifested as abnormal There are no experimental animal female reproduc- neurobehavioral function in rats exposed during gesta- tive toxicity studies or male or female fertility studies. tion to maternal doses of 0.5 mg/kg bw/day throughout Data are insufficient to conclude that metham- pregnancy or 2 mg/kg bw/day on GD 12–15. The data phetamine affects male mating behavior. are insufficient for making conclusions about effects on In one study rats injected with 1 mg/kg metham- monoamine levels and morphological effects of amphe- phetamine for a 5-week period failed to ejaculate tamine treatment on the brains of these animals. during a 90-min test. 5.1.2.2 Methamphetamine. There were no methamphetamine developmental toxicity studies in experimental animals that used the oral route of exposure. Available studies employed different routes of exposure, but s.c. 5.0 SUMMARIES, CONCLUSIONS, AND injection was most common. A qualitative evaluation of CRITICAL DATA NEEDS these studies indicates that decreased litter size and postnatal body weight gains were observed in pregnant 5.1 Developmental Toxicity rats exposed to methamphetamine by s.c. injection. Four 5.1.1 Human data. The Expert Panel concluded rat studies supported the conclusion of altered litter size that there was insufficient evidence to evaluate the and/or decreased postnatal survival with methamphe- developmental toxicity of therapeutic use of ampheta- tamine treatment. These effects were seen at similar mines in humans. There were two cohort studies in doses to those altering maternal body weights, although which pregnant women were prescribed amphetamines decreased offspring survival was also seen in studies that to limit weight gain during gestation that did not included a pair-fed control. A similar number of studies show statistically significant increases in overall support alterations in pup body weights. In some cases, malformations, but their power for studying malforma- these body weight effects persisted beyond cessation of tions was low. These studies are not sufficient to treatment. Additional comparisons of methamphetamine draw conclusions about risk of specific malformations. studies were complicated by exposures during varying There is insufficient evidence for a conclusion on the periods of gestation. effect of gestational exposure to amphetamine on birth Although limited by parenteral administration of weight. single-dose levels in most studies, rat data are sufficient There are insufficient data to evaluate whether to demonstrate developmental neurotoxicity of d,l- amphetamine treatment in children is associated with methamphetamine. Toxicity is manifested as abnormal depressed growth, increased tics, or altered risk of neurobehavioral testing in rats exposed during gestation tobacco use, problematic alcohol use, or illicit substance and/or lactation to maternal doses of 2 mg/kg bw/day use in childhood, adolescence, and adulthood. throughout pregnancy or GD 7–20 for methamphetamine The Expert Panel concludes that the data regarding (enantiomer unspecified) and 5 mg/kg bw/day for d- illicit methamphetamine are insufficient to draw conclu- methamphetamine. There is insufficient evidence to sions concerning developmental toxicity in humans. conclude that there are effects on monoamine levels or 5.1.2. Experimental animal data morphological effects of methamphetamine treatment on 5.1.2.1 Amphetamine. There is insufficient informa- the brains of these animals. tion to evaluate the effects of oral amphetamine on There is sufficient evidence that methamphetamine developmental toxicity in animals. Although there are produces developmental toxicity in rats manifested as limitations in most available studies, a weight of behavioral alterations after treatment of neonates with evidence approach suggests patterns of developmental 30 mg/kg bw/day. effects after in utero amphetamine exposure. Increased malformations, which included cardiac malformations, cleft lip, eye abnormalities and exencephaly, were 5.2 Reproductive Toxicity observed in three mouse studies involving i.p. adminis- There are no human reproductive toxicity data for tration of Z50 mg/kg bw/day amphetamine on GD amphetamine or methamphetamine.

Birth Defects Research (Part B) 74:471–584, 2005 DOI 10.1002/bdrb 578 GOLUB ET AL. The Expert Panel concluded that there are insufficient * developmental toxicity in children after intrauterine experimental animal amphetamine data to evaluate exposure; reproductive toxicity. Only two relevant studies were * growth in children and adolescents receiving available in the open literature, both of which were very medication; limited in scope. Guideline reproductive toxicity studies * altered risks of tobacco use, problematic alcohol conducted for submission to FDA as preclinical toxicity consumption, or illicit substance abuse in adolescents data were not available in sufficient detail for the Expert or adults; Panel to review. * altered risks of tics or movement disorders in treated The Expert Panel concluded that data are not sufficient children; to evaluate reproductive toxicity of methamphetamine * fetal and neonatal viability in laboratory animals in animals. There are no female reproductive exposed prenatally; toxicity studies or male or female fertility studies. * morphological effects in brains of laboratory animals; Delayed ejaculations occurred in rats injected with and methamphetamine for a 5-week period, but the data * male and female reproductive toxicity in experimental were judged to be insufficient for evaluating male animals. mating behavior. There were no human reproductive data available. 5.3 Summary of Human Exposures The Expert Panel judged the data sufficient to The focus of this report is the amphetamines approved conclude that there was evidence for the following: for clinical practice: d-, and d,l-amphetamine and * neurobehavioral alterations after prenatal exposure in d-methamphetamine. The number of amphetamine pre- rats to maternal s.c. doses of 0.5 mg/kg bw/day scriptions written increased between 1992–2000 from throughout pregnancy or 2 mg/kg bw/day on GD 12– fewer than 500,000 to nearly 8 million per year (Drug 15; and Enforcement Administration, 2004). Treatment of ADHD * increased incidence of congenital malformations in in teenagers and adults is increasing. More people of mice after i.p. injections of 50 mg/kg bw/day. The reproductive age may be taking amphetamine, but there data were not considered relevant to humans because is no information on the numbers of pregnant or lactating of the route of administration during pregnancy and women who are taking the drug(s). the corresponding lack of information on pharmaco- The d and d,l-amphetamine preparations are indicated kinetics. Drugs administered i.p. in pregnancy can (1) for the treatment of ADHD and narcolepsy (PDR, possibly have a direct adverse effect on the uterus and 2003a,d), and methamphetamine is indicated for the secondarily on embryo-fetal development, and (2) treatment of ADHD and short-term treatment of obesity possibly be physically transported directly to the (PDR, 2003c). The Expert Panel recognizes that embryo/fetus, bypassing maternal absorption, meta- therapeutic use of d-methamphetamine in the U.S. is bolism, and distribution. uncommon. Recommended doses of amphetamine are 2.5–40 mg/day for treatment of ADHD in individuals The Expert Panel concluded that, based on the 3 years of age and older and 5–60 mg/day for treatment experimental animal findings, there was some concern of narcolepsy in individuals 6 years of age and older. with regard to potential neurobehavioral alterations due The Expert Panel is aware of off-label uses of to prenatal amphetamine exposure in humans both in amphetamines to treat depression, primarily as an therapeutic and non-therapeutic settings. adjunct to antidepressant medication, and to treat 5.4.2 Methamphetamine. There is a substantial patients with post-stroke cognitive impairment. published database of studies designed to investigate d-Methamphetamine hydrochloride is used as a the potential adverse reproductive and developmental recreational drug. Routes of administration include effects of methamphetamine exposure in humans inhalation, nasal, iv, and oral. There is also potential for and laboratory animals. The Expert Panel felt that the abuse of d- and d,l-amphetamine. studies that focused upon humans were uninterpretable due to such factors as a lack of control of potential 5.4 Overall Conclusions confounding factors and the issue of the purity and The Expert Panel was impressed with the paucity of contaminants of the methamphetamine used by the drug interpretable toxicity data relevant to human therapeutic abusers. use. The decision to use the medication must be made by The Expert Panel found data were available but the responsible health care provider, the patient, and the insufficient to evaluate the following: family if the patient is a minor. * morphological effects in brains of laboratory animals; 5.4.1 Amphetamine. There is a substantial pub- and lished database of studies designed to investigate the * male mating behavior in laboratory animals. There potential adverse reproductive and developmental ef- were no experimental animal female reproductive fects of amphetamine exposure in humans and labora- toxicity or male or female fertility studies. tory animals. Thorough review of these studies led the Expert Panel to judge that within all of the human The Expert Panel judged the data sufficient to domains and some of the animal domains, the data were conclude that there was evidence for the following: generally insufficient to reach valid scientific conclusions. * neurobehavioral alterations after pre- or postnatal Specifically, the Expert Panel found data were avail- exposure in laboratory animals; and able but insufficient to evaluate: * effects on litter size and postnatal pup weight.

Birth Defects Research (Part B) 74:471–584, 2005 DOI 10.1002/bdrb AMPHETAMINE AND METHAMPHETAMINE 579 The Expert Panel concluded that, based on the * Data on long-term neurobehavioral outcomes across experimental animal findings, there was concern with the life cycle after therapeutic postnatal exposure in regard to potential adverse perinatal outcomes and humans are needed. neurobehavioral alterations due to prenatal methamphe- * Studies are needed to evaluate possible effects of tamine exposure in humans both in therapeutic and non- amphetamine treatment on tobacco use, problematic therapeutic settings. use of alcohol, and illicit substance use in children, adolescents, and adults. * 5.5 Critical Data Needs Epidemiological data are needed regarding the extent of therapeutic and off-label exposure during pericon- Critical data needs are defined as research or studies ceptual period, pregnancy, and lactation among that would provide information to substantially reduce women with ADHD, narcolepsy, or depression. uncertainty and increase confidence in assessing human * Data are needed to determine the numbers of reproductive and developmental risks. Although studies amphetamine prescriptions for teenagers, adults, documenting effects of childhood exposures were avail- and children o3 years old. able, the Expert Panel found the studies were generally * Studies are needed to characterize the long-term limited due to inadequate design, use of outdated effects of amphetamine treatment on heart rate and methodsorstandards,andinsufficientnumbersof blood pressure. The studies need to consider all subjects treated only with amphetamines. Therefore, the populations, including the under-represented popula- Expert Panel concluded that better quality studies are tions denoted above. required to effectively evaluate toxicity concerns associated with amphetamines. There are a number of considerations that should be applied in the design of 5.5.1.2 Experimental animal studies. quality human studies for amphetamines. The studies * Data are needed on developmental and reproductive need to use current techniques and age-standardized toxicity of amphetamines in experimental animals. norms. Confounding factors such as prenatal or postnatal The studies should include relevant routes of expo- exposure to tobacco, alcohol, and illicit drugs, parental sure, multiple dose levels and pharmacokinetic psychiatric disorders, and care-giving environment need endpoints. to be noted and adequately controlled. Subpopulations * Data are needed on the effects of amphetamines that are susceptible to development of ADHD (e.g., during pregnancy with detailed pharmacokinetic data children born prematurely) need to be considered in study on transfer of the drugs and their metabolites to the design and interpretation. The studies need to consider fetus. currently under-represented populations such as children * Experimental animal data from studies submitted to born prematurely, non-Caucasian individuals, and fe- regulatory agencies should be made available to the males. Current trends in treatment such as infrequent public. use of drug holidays and durations of exposure that often * Studies are needed to evaluate possible effects of extend through adolescence and into adulthood need to be amphetamine on pubertal timing and quality. considered. The studies should compare endpoints in * Studies are needed to model data obtained by non- individuals within the same developmental stage (e.g., oral routes of administration so these data can be childhood vs. adolescence) and use appropriate controls. more useful in evaluating human oral exposures. * Valid animal models of ADHD need to be used in 5.5.1 Amphetamines studies of amphetamine toxicity to evaluate toxicity in 5.5.1.1 Human studies. a system that more closely approximates the human patient population. * Pharmacokinetic data related to therapeutic use * Nonhuman primate or guinea pig studies would be during pregnancy and lactation are needed. useful to evaluate effects of gestational amphetamine * Data on possible reproductive effects of ampheta- use in the second and third trimester. mines in humans are needed. * Growth data (height and weight) in both prepubertal 5.5.2 Methamphetamine. The Expert Panel recog- and pubertal children and adolescents are needed. * nizes that pure toxicology studies in humans are not Rates of progression through pubertal stages in boys possible in evaluating the developmental and reproduc- and girls with ADHD treated with amphetamines are tive impact of illicitly synthesized methamphetamine. needed. * The Panel recognizes, however, an urgent scientific and Studies are needed to characterize the effects of public health need for epidemiologic studies to evaluate, amphetamines on tic and other movement disorders. * with covariate control, the potential short- and long-term Because of the wide age spectrum in which the drug is effects of in utero exposure and passive postnatal used, studies are needed to identify possible devel- exposure to illicit methamphetamine and its contami- opmental variations that can affect toxicity. * nants. In addition, evaluation of reproductive toxicity of Toxicity data are needed for under-represented exposure to illicit methamphetamine and its contami- populations of children and adolescents including nants would be of interest. girls, non-Caucasians, children with dysmorphic and The following data needs were identified for experi- genetic syndromes and global mental retardation, and mental animal studies: children born prematurely. * Data on neurobehavioral outcomes from infancy to * Data are needed on developmental and reproductive adulthood after prenatal exposure in humans are toxicity of methamphetamine in experimental ani- needed. mals. The studies should include relevant routes of

Birth Defects Research (Part B) 74:471–584, 2005 DOI 10.1002/bdrb 580 GOLUB ET AL.

exposure, multiple dose levels, and pharmacokinetic Bertilsson L, Dahl ML, Tybring G. 1997. Pharmacogenetics of anti- endpoints. depressants: clinical aspects. Acta Psychiatr Scand Suppl 391: * 14–21. Data are needed on the effects of methamphetamine Biederman J, Wilens T, Mick E, Spencer T, Faraone SV. 1999. Pharma- during pregnancy, with detailed pharmacokinetic cotherapy of attention-deficit/hyperactivity disorder reduces risk for data on transfer of the drugs and their metabolites substance use disorder. Pediatrics 104:e20. to the fetus. Bigl V, Dalitz E, Kunert E, Biesold D, Leonard BE. 1982. The effect of D- * amphetamine and amitriptyline administered to pregnant rats on the Studies are needed to evaluate possible effects of locomotor activity and neurotransmitters of the offspring. Psycho- methamphetamine on pubertal timing and quality. pharmacology 77:371–375. * Nonhuman primate or guinea pig studies would be Billing L, Eriksson M, Larsson G, Zetterstro¨m R. 1980. Amphetamine addiction and pregnancy. 3. One year follow-up of the children.- useful to evaluate effects of gestational methamphe- psychosocial and pediatric aspects. Acta Paediatr Scand 69:675–680. tamine use in the second and third trimester. Billing L, Eriksson M, Steneroth G, Zetterstro¨m R. 1985. Pre-school children of amphetamine-addicted mothers. I. Somatic and psychomotor development. Acta Paediatr Scand 74:179–184. Billing L, Eriksson M, Steneroth G, Zetterstro¨m R. 1988. Predictive indicators for adjustment in 4-year-old children whose mothers used REFERENCES amphetamine during pregnancy. Child Abuse Negl 12:503–507. Blaesing B, Nossoll M, Teuchert-Noodt G, Dawirs RR. 2001. Postnatal AAP. 2001. Clinical practice guideline: treatment of the school-aged child maturation of prefrontal pyramidal neurones is sensitive to a single with attention-deficit/hyperactivity disorder. Pediatrics 108: early dose of methamphetamine in gerbils (Meriones unguiculatus). 1033–1044. J Neural Transm 108:101–113. Aarskog D, Ferang F, Klove H, Stoa KF, Thorsen T. 1977. Effect of the Boe NM, Eby-Wilkens E, Field NT, Hedriana HL, Gilbert WM. 1999. stimulant drugs, dextroamphetamine and methylphenidate on Methamphetamine use during pregnancy increases the risk of secretion of growth hormone in hyperactive children. J Pediatr adverse maternal and neonatal outcomes. Am J Obstet Gynecol 90:136–139. 180:S71. Acuff-Smith KD, George M, Lorens SA, Vorhees CV. 1992. Preliminary Borcherding BG, Keysor CS, Rapoport JL, Elia J, Amass J. 1990. Motor/ evidence for methamphetamine-induced behavioral and ocular vocal tics and compulsive behaviors on stimulant drugs: is there a effects in rat offspring following exposure during early organogen- common vulnerability? Psychiatry Res 33:83–94. esis. Psychopharmacology (Berl) 109:255–263. Bost RO, Kemp P, Hnilica V. 1989. Tissue distribution of methampheta- Acuff-Smith KD, Schilling MA, Fisher JE, Vorhees CV. 1996. Stage- mine and amphetamine in premature infants. J Anal Toxicol 13: specific effects of prenatal d-methamphetamine exposure on 300–302. behavioral and eye development in rats. Neurotoxicol Teratol Briggs GG, Samson JH, Crawford DJ. 1975. Lack of abnormalities in a 18:199–215. newborn exposed to amphetamine during gestation. Am J Dis Child Adams J, Buelke-Sam J, Kimmel CA, LaBorde JB. 1982. Behavioral 129:249–250. alterations in rats prenatally exposed to low doses of d-ampheta- Brown GL, Ebert MH, Mikkelsen EJ, Hunt RD. 1980. Behavior and motor mine. Neurobehav Toxicol Teratol 4:63–70. activity response in hyperactive children and plasma amphetamine Adams J, Buelke-Sam J, Kimmel CA, Nelson CJ, Reiter LW, Sobotka TJ, levels following a sustained release preparation. J Am Acad Child Tilson HA, Nelson BK. 1985a. Collaborative Behavioral Teratology Psychiatry 19:225–239. Study: protocol design and testing procedures. Neurobehav Toxicol Brown GL, Hunt RD, Ebert MH, Bunney WE Jr, Kopin IJ. 1979. Plasma Teratol 7:579–586. levels of d-amphetamine in hyperactive children. Serial behavior and Adams J, Oglesby DM, Ozemek HS, Rath J, Kimmel CA, Buelke-Sam J. motor responses. Psychopharmacology (Berl) 62:133–140. 1985b. Collaborative Behavioral Teratology Study: programmed Buelke-Sam J, Kimmel CA, Adams J, Nelson CJ, Vorhees CV, Wright DC, data entry and automated test systems. Neurobehav Toxicol Teratol St Omer V, Korol BA, Butcher RE, Geyer MA, et al. 1985. 7:547–554. Collaborative Behavioral Teratology Study: results. Neurobehav Ahmann PA, Waltonen SJ, Olson KA, Theye FW, Van Erem AJ, LaPlant RJ. Toxicol Teratol 7:591–624. 1993. Placebo-controlled evaluation of Ritalin side effects. Pediatrics Burchfield DJ, Lucas VW, Abrams RM, Miller RL, DeVane CL. 1991. 91:1101–1106. Disposition and pharmacodynamics of methamphetamine in preg- Akorn. 2001. Product label on Paremyd. Somerset, NJ: Akorn, Inc. nant sheep. JAMA 265:1968–1973. Albertson TE, Derlet RW, Van Hoozen BE. 1999. Methamphetamine and Cabrera TM, Levy AD, Li Q, van de Kar LD, Battaglia G. 1993. Prenatal the expanding complications of amphetamines. West J Med 170: methamphetamine attenuates serotonin mediated renin secretion in 214–219. male and female rat progeny: evidence for selective long-term Alcohol and Drug Information Clearing House. 1998. Methamphetamine. dysfunction of serotonin pathways in brain. Synapse 15:198–208. Available at http://www.prevlink.org/clearinghouse/catalog/ Caldwell J, Dring LG, Williams RT. 1972. Metabolism of (14 C)metham- drugs/meth/methbro.pdf. Alcohol and Drug Information Clearing- phetamine in man, the guinea pig and the rat. Biochem J 129: house and Nebraska Prevention Information Network. 11–22. Alhava E, Mattila M. 1974. Dose-dependent differences of amphetamine Call TD, Hartneck J, Dickinson WA, Hartman CW, Bartel AG. 1982. Acute levels in brain and heart of adult and developing mice. Acta cardiomyopathy secondary to intravenous amphetamine abuse. Ann Pharmacol Toxicol 34:211–221. Intern Med 97:559–560. Anggard E, Gunne LM, Jonsson LE, Niklasson F. 1970. Pharmcokinetic Cameron RH, Kolesari GL, Kalbfleisch JH. 1983. Pharmacology of and clinical studies on amphetamine-dependent subjects. Eur J Clin dextroamphetamine-induced cardiovascular malformations in the Pharmacol 3:3–11. chick embryo. Teratology 27:253–259. Anggard E, Jonsson LE, Hogmark AL, Gunne LM. 1973. Amphetamine Cameron RH, Kolesari GL, Rajala GM. 1984. Elevated blood pressure in metabolism in amphetamine psychosis. Clin Pharmacol Ther 14: the embryonic chick induced by a teratogenic dose of dextroamphe- 870–880. tamine sulfate. Teratology 29:87–92. Angrist B, Gershon S. 1976. Clinical effects of amphetamine and L-DOPA Cappon GD, Vorhees CV. 2001. Plasma and brain methamphetamine on sexuality and aggression. Compr Psychiatry 17:715–722. concentrations in neonatal rats. Neurotoxicol Teratol 23:81–88. Anzenbacher P, Anzenbacherova E. 2001. Cytochromes P450 and Castellanos FX, Giedd JN, Elia J, Marsh WL, Ritchie GF, Hamburger SD, metabolism of xenobiotics. Cell Mol Life Sci 58:737–747. Rapoport JL. 1997. Controlled stimulant treatment of ADHD and Barkley RA, Fischer M, Smallish L, Fletcher K. 2003. Does the treatment of comorbid Tourette’s syndrome: effects of stimulant and dose. J Am attention-deficit/hyperactivity disorder with stimulants contribute Acad Child Adolesc Psychiatry 36:589–596. to drug use/abuse? A 13-year prospective study. Pediatrics 111: Cates NR. 1970. Effects of psychoactive drugs on sperm motility. Res 97–109. Commun Chem Pathol Pharmacol 1:223–229. Bashour TT. 1994. Acute myocardial infarction resulting from Cernerud L, Eriksson M, Jonsson B, Steneroth G, Zetterstro¨m R. 1996. amphetamine abuse: a spasm-thrombus interplay? Am Heart J Amphetamine addiction during pregnancy: 14-year follow-up of 128:1237–1239. growth and school performance. Acta Paediatr 85:204–208. Bays J. 1991. Fetal vascular disruption with prenatal exposure to cocaine ChemIDplus. 2004a. Amphetamine. Available at http://chem2.sis.nlm. or methamphetamine. Pediatrics 87:416–418. nih.gov/chemidplus/ProxyServlet?objectHandle 5 DBMaint&action Bell RW, Drucker RR, Woodruff AB. 1965. The effects of prenatal Handle 5 default&nextPage 5 jsp/chemidlite/ResultScreen.jsp& injections of adrenalin chloride and D-amphetamine sulfate on TXTSUPERLISTID 5 000300629. National Library of Medicine. subsequent emotionality and ulcer-proneness of offspring. Psychon ChemIDplus. 2004b. Methylphenidate. Available at http://chem2. Sci 2:269–270. sis.nlm.nih.gov/chemidplus/ProxyServlet?objectHandle 5 DBMaint

Birth Defects Research (Part B) 74:471–584, 2005 DOI 10.1002/bdrb AMPHETAMINE AND METHAMPHETAMINE 581

&actionHandle 5 default&nextPage 5 jsp/chemidlite/ResultScreen. Eriksson M, Larsson G, Zetterstro¨m R. 1981. Amphetamine addiction and jsp&TXTSUPERLISTID 5 000113451. National Library of Medicine. pregnancy. II. Pregnancy, delivery and the neonatal period. Socio- Chilcoat HD, Breslau N. 1999. Pathways from ADHD to early drug use. medical aspects. Acta Obstet Gynecol Scand 60:253–259. J Am Acad Child Adolesc Psychiatry 38:1347–1354. Eriksson M, Zetterstro¨m R. 1994. Amphetamine addiction during Ching M, Tang L. 1986. Neuroleptic drug-induced alterations on neonatal pregnancy: 10-year follow-up. Acta Paediatr Suppl 404:27–31. growth and development. I. Prenatal exposure influences birth size, FDA. 2001. Adderal XR: Chemistry Review(s), Clinical Pharmacology mortality rate, and the neuroendocrine system. Biol Neonate 49: and Biopharmaceutics Reviews(s), Medical Review, Pharmacology 261–269. Review, and Statistical Review. Center for Drug Evaluation and Cho AK, Melega WP, Kuczenski R, Segal DS. 2001. Relevance of Research. Report nr 21-303. pharmacokinetic parameters in animal models of methamphetamine FDA. 2002. Adderal: Chemistry Review(s), Clinical Pharmacology and abuse. Synapse 39:161–166. Biopharmaceutics Reviews(s), and Medical Review. Center for Drug Cho DH, Lyu HM, Lee HB, Kim PY, Chin K. 1991. Behavioral Evaluation and Research. Report nr 11-522/S-030. teratogenicity of methamphetamine. J Toxicol Sci 16(Suppl):37–49. FDA. 2004. Drugs at FDA. Available at: http://www.accessdata.fda.gov/ Clark CH, Gorman D, Vernadakis A. 1970. Effects of prenatal adminis- scripts/cder/drugsatfda/index.cfm. FDA CDER. tration of psychotropic drugs on behavior of developing rats. Dev Fein A, Shviro Y, Manoach M, Nebel L. 1987. Teratogenic effect of Psychobiol 3:225–235. D-amphetamine sulfate: histodifferentiation and electrocardiogram Cook CE, Jeffcoat AR, Hill JM, Pugh DE, Patetta PK, Sadler BM, White pattern of mouse embryonic heart. Teratology 35:27–34. WR, Perez-Reyes M. 1993. Pharmacokinetics of methamphetamine Felix RJ, Chambers CD, Dick LM, Johnson KA, Jones KL. 2000. self-administered to human subjects by smoking S-(1)-methamphe- Prospective pregnancy outcome in women exposed to ampheta- tamine hydrochloride. Drug Metab Dispos 21:717–723. mines. Teratology 61:441. Cook CE, Jeffcoat AR, Sadler BM, Hill JM, Voyksner RD, Pugh DE, White Finnegan LP, Ehrlich SM. 1990. Maternal drug abuse during pregnancy: WR, Perez-Reyes M. 1992. Pharmacokinetics of oral methampheta- evaluation and pharmacotherapy for neonatal abstinence. Mod mine and effects of repeated daily dosing in humans. Drug Metab Methods Pharmacol 6:255–263. Dispos 20:856–862. Fischer M, Barkley RA. 2003. Childhood stimulant treatment and risk for Courtney KD, Valerio DA. 1968. Teratology in the Macaca mulatta. later substance abuse. J Clin Psychiatry 64(Suppl):19–23. Teratology 1:163–172. Fujiwara Y, Kazahaya Y, Nakashima M, Sato M, Otsuki S. 1987. Crawford CA, Williams MT, Newman ER, McDougall SA, Vorhees CV. Behavioral sensitization to methamphetamine in the rat: an 2003. Methamphetamine exposure during the preweanling period ontogenic study. Psychopharmacology (Berl) 91:316–319. causes prolonged changes in dorsal striatal protein kinase A activity, Fukui R, Svenningsson P, Matsuishi T, Higashi H, Nairn AC, Greengard P, dopamine D2-like binding sites, and dopamine content. Synapse Nishi A. 2003. Effect of methylphenidate on dopamine/DARPP 48:131–137. signalling in adult, but not young, mice. J Neurochem 87:1391–1401. Crawford CA, Zavala AR, Karper PE, Collins RL, Loring-Meier T, Watson Fukumura M, Cappon GD, Broening HW, Vorhees CV. 1998. Metham- JB, McDougall SA. 2000a. Amphetamine treatment during the phetamine-induced dopamine and serotonin reductions in neostria- preweanling period produces enduring changes in striatal protein tum are not gender specific in rats with comparable hyperthermic kinase A activity. Pharmacol Biochem Behav 66:835–840. responses. Neurotoxicol Teratol 20:441–448. Crawford CA, Zavala AR, Karper PE, McDougall SA. 2000b. Long-term Furara SA, Carrick P, Armstrong D, Pairaudeau P, Pullan AM, Lindow effects of postnatal amphetamine treatment on striatal protein kinase SW. 1999. The outcome of pregnancy associated with amphetamine A activity, dopamine D(1)-like and D(2)-like binding sites, and use. J Obstet Gynaecol 19:377–380. dopamine content. Neurotoxicol Teratol 22:799–804. Furst SR, Fallon SP, Reznik GN, Shah PK. 1990. Myocardial infarction DeVane CL. 1994. Pharmacogenetics and drug metabolism of newer after inhalation of methamphetamine. N Engl J Med 323:1147–1148. antidepressant agents. J Clin Psychiatry 55(Suppl):38–47. Gao X, Dluzen DE. 2001. The effect of testosterone upon methampheta- Dickinson J, Andres R, Parisi V. 1994. The ovine fetal sympathoadrenal mine neurotoxicity of the nigrostriatal dopaminergic system. Brain response to the maternal administration of methamphetamine. Am J Res 892:63–69. Obstet Gynecol 170:1452–1457. Garriott JC, Spruill FG. 1973. Detection of methamphetamine in a Dixon SD, Bejar R. 1989. Echoencephalographic findings in neonates newborn infant. J Forensic Sci 18:434–436. associated with maternal cocaine and methamphetamine use: Gazzara RA, Fisher RS, Howard SG. 1986. The ontogeny of amphetamine- incidence and clinical correlates. J Pediatr 115:770–778. induced dopamine release in the caudate-putamen of the rat. Brain Dominguez R, Vila-Coro AA, Slopis JM, Bohan TP. 1991. Brain and ocular Res 393:213–220. abnormalities in infants with in utero exposure to cocaine and other Gillogley KM, Evans AT, Hansen RL, Samuels SJ, Batra KK. 1990. The street drugs. Am J Dis Child 145:688–695. perinatal impact of cocaine, amphetamine, and opiate use detected Dring LG, Smith RL, Williams RT. 1970. The metabolic fate of by universal intrapartum screening. Am J Obstet Gynecol 163: amphetamine in man and other species. Biochem J 116:425–435. 1535–1542. DEA. 2004. DEA web site. Available at: http://www.usdoj.gov/dea/ Ginsberg MD, Hertzman M, Schmidt-Nowara WW. 1970. Amphetamine index.htm. intoxication with coagulopathy, hyperthermia, and reversible renal Duke MA, O’Neal J, McDougall SA. 1997. Ontogeny of dopamine failure. A syndrome resembling heatstroke. Ann Intern Med 73: agonist-induced sensitization: role of NMDA receptors. Psychophar- 81–85. macology (Berl) 129:153–160. Golden GS. 1974. Gilles de la Tourette’s syndrome following methylphe- Dunnick JK, Eustis SL. 1991. Decreases in spontaneous tumors in rats and nidate administration. Dev Med Child Neurol 16:76–78. mice after treatment with amphetamine. Toxicology 67:325–332. Golden GS. 1977. The effect of central nervous system stimulants on Efron D, Jarman F, Barker M. 1997. Side effects of methylphenidate Tourette syndrome. Ann Neurol 2:69–70. and dexamphetamine in children with attention deficit Gomes-da-Silva J, Perez-Rosado A, de Miguel R, Fernandez-Ruiz J, Silva hyperactivity disorder: a double-blind, crossover trial. Pediatrics MC, Tavares MA. 2002. Prenatal exposure to methamphetamine in 100:662–666. the rat: ontogeny of tyrosine hydroxylase mRNA expression in Ehrlich ME, Sommer J, Canas E, Unterwald EM. 2002. Periadolescent mesencephalic dopaminergic neurons. Ann NY Acad Sci 965:68–77. mice show enhanced DFosB upregulation in response to cocaine and Gospe SM Jr. 1995. Transient cortical blindness in an infant exposed to amphetamine. J Neurosci 22:9155–9159. methamphetamine. Ann Emerg Med 26:380–382. Erenberg G, Cruse RP, Rothner AD. 1985. Gilles de la Tourette’s Greaves G. 1972. Sexual disturbances among chronic amphetamine users. syndrome: effects of stimulant drugs. Neurology 35:1346–1348. J Nerv Ment Dis 155:363–365. Eriksson M, Billing L, Steneroth G, Zetterstro¨m R. 1985. Pre-school Greenberg LM, Deem MA, McMahon S. 1972. Effects of dextroamphe- children of amphetamine-addicted mothers. II. Environment and tamine, chlorpromazine, and hydroxyzine on behavior and perfor- supportive social welfare. Acta Paediatr Scand 74:185–190. mance in hyperactive children. Am J Psychiatry 129:532–539. Eriksson M, Billing L, Steneroth G, Zetterstro¨m R. 1989. Health and Greenhill LL. 1992. Pharmacologic treatment of attention deficit development of 8-year-old children whose mothers abused amphe- hyperactivity disorder. Psychiatr Clin North Am 15:1–27. tamine during pregnancy. Acta Paediatr Scand 78:944–949. Greenhill LL, Puig-Antich J, Chambers W, Rubinstein B, Halpern F, Eriksson M, Cernerud L, Johnson B, Steneroth G, Zetterstro¨m R. 1994. Sachar EJ. 1981. Growth hormone, prolactin, and growth responses Amphetamine abuse during pregnancy: longterm follow-up of in hyperkinetic males treated with d-amphetamine. J Am Acad Child exposed children. Teratology 50:45A. Psychiatry 20:84–103. Eriksson M, Jonsson B, Steneroth G, Zetterstro¨m R. 2000. Amphetamine Greenhill LL, Puig-Antich J, Sassin J, Sachar EJ. 1977. Hormone and abuse during pregnancy: environmental factors and outcome after growth responses in hyperkinetic children on stimulant medication 14–15 years. Scand J Public Health 28:154–157. [proceedings]. Psychopharmacol Bull 13:33–36. Eriksson M, Larsson G, Winbladh B, Zetterstro¨m R. 1978. The influence of Greenhill LL, Swanson JM, Steinhoff K, Fried J, Posner K, Lerner M, amphetamine addiction on pregnancy and the newborn infant. Acta Wigal S, Clausen SB, Zhang Y, Tulloch S. 2003. A pharmacokinetic/ Paediatr Scand 67:95–99. pharmacodynamic study comparing a single morning dose of

Birth Defects Research (Part B) 74:471–584, 2005 DOI 10.1002/bdrb 582 GOLUB ET AL.

adderall to twice-daily dosing in children with ADHD. J Am Acad Lambert NM, Hartsough CS. 1998. Prospective study of tobacco smoking Child Adolesc Psychiatry 42:1234–1241. and substance dependencies among samples of ADHD and non- Gross MD. 1976. Growth of hyperkinetic children taking methylpheni- ADHD participants. J Learn Disabil 31:533–544. date, dextroamphetamine, or imipramine/desipramine. Pediatrics Lanier LP, Isaacson RL. 1977. Early developmental changes in the 58:423–431. locomotor response to amphetamine and their relation to hippo- Hall RC, Popkin MK, Beresford TP, Hall AK. 1988. Amphetamine campal function. Brain Res 126:567–575. psychosis: clinical presentations and differential diagnosis. Psychiatr Larez A, Briceno E, Ochoa Y, Montenegro M, Aponte N. 1979. Med 6:73–79. Mutagenicity obtained experimentally by oral administration of Hansen RL, Struthers JM, Gospe SM Jr. 1993. Visual evoked potentials dextroamphetamine sulphate to the rat. Bull Narc 31:67–70. and visual processing in stimulant drug-exposed infants. Dev Med Larsson G. 1980. The amphetamine addicted mother and her child. Acta Child Neurol 35:798–805. Paediatr Scand Suppl 278:1–24. Harris DS, Boxenbaum H, Everhart ET, Sequeira G, Mendelson JE, Jones Larsson G, Eriksson M, Zetterstro¨m R. 1979. Amphetamine addiction and RT. 2003. The bioavailability of intranasal and smoked methamphe- pregnancy. Psycho-social and medical aspects. Acta Psychiatr Scand tamine. Clin Pharmacol Ther 74:475–486. 60:334–346. Heinonen OP. 1977. Birth defects and drugs in pregnancy. Littleton, MA: Laviola G, Adriani W, Morley-Fletcher S, Terranova ML. 2002. Peculiar Publishing Sciences Group Inc. response of adolescent mice to acute and chronic stress and to Heller A, Bubula N, Freeney A, Won L. 2001a. Elevation of fetal amphetamine: evidence of sex differences. Behav Brain Res 130: dopamine following exposure to methamphetamine in utero. Brain 117–125. Res Dev Brain Res 130:139–142. Leckman JF. 2003. Phenomenology of tics and natural history of tic Heller A, Bubula N, Lew R, Heller B, Won L. 2001b. Gender- disorders. Brain Dev 25(Suppl):S24–S28. dependent enhanced adult neurotoxic response to methampheta- Levin JN. 1971. Amphetamine ingestion with biliary atresia. J Pediatr mine following fetal exposure to the drug. J Pharmacol Exp Ther 79:130–131. 298:769–779. Lipkin PH, Goldstein IJ, Adesman AR. 1994. Tics and dyskinesias Hemmer SA, Pasternak JF, Zecker SG, Trommer BL. 2001. Stimulant associated with stimulant treatment in attention-deficit hyperactivity therapy and seizure risk in children with ADHD. Pediatr Neurol disorder. Arch Pediatr Adolesc Med 148:859–861. 24:99–102. Little BB, Snell LM, Gilstrap LC 3rd, 1988. Methamphetamine abuse Hines RN, McCarver DG. 2002. The ontogeny of human drug- during pregnancy: outcome and fetal effects. Obstet Gynecol 72: metabolizing enzymes: phase I oxidative enzymes. J Pharmacol 541–544. Exp Ther 300:355–360. Loney J, Kramer JR, Salisbury H. 2002. Medicated versus unmedicated Hitzemann BA, Hitzemann RJ, Brase DA, Loh HH. 1976. Influence of ADHD children: adult involvement with legal and illegal drugs. In: prenatal D-amphetamine administration on development and Jensen PS, Cooper JR, editors. Attention deficit hyperactivity behavior of rats. Life Sci 18:605–612. disorder. Kingston, NJ: Civic Research Institute. p 1–16. Holson R, Adams J, Buelke-Sam J, Gough B, Kimmel CA. 1985. d- Lowe TL, Cohen DJ, Detlor J, Kremenitzer MW, Shaywitz BA. 1982. Amphetamine as a behavioral teratogen: effects depend on dose, sex, Stimulant medications precipitate Tourette’s syndrome. JAMA age and task. Neurobehav Toxicol Teratol 7:753–758. 247:1729–1731. Hong R, Matsuyama E, Nur K. 1991. Cardiomyopathy associated with the Lucot JB, Wagner GC, Schuster CR, Seiden LS. 1982. Decreased sensitivity smoking of crystal methamphetamine. Jama 265:1152–1154. of rat pups to long-lasting dopamine and serotonin depletions HSDB. 2002. Methylphenidate. Available at http://toxnet.nlm.nih.gov/ produced by methylamphetamine. Brain Res 247:181–183. cgi-bin/sis/search/f?./temp/BPMHaIx:1. National Library of Madras BK, Miller GM, Fischman AJ. 2002. The dopamine transporter: Medicine. relevance to attention deficit hyperactivity disorder (ADHD). Behav HSDB. 2004. Amphetamine. Available at: http://toxnet.nlm.nih.gov/cgi- Brain Res 130:57–63. bin/sis/htmlgen?HSDB. Markowitz JS, Morrison SD, DeVane CL. 1999. Drug interactions with Joffe GM, Kasnic T. 1994. Medical prescription of dextroamphetamine psychostimulants. Int Clin Psychopharmacol 14:1–18. during pregnancy. J Perinatol 14:301–303. Markowitz JS, Patrick KS. 2001. Pharmacokinetic and pharmacodynamic Jones SR, Gainetdinov RR, Wightman RM, Caron MG. 1998. Mechanisms drug interactions in the treatment of attention-deficit hyperactivity of amphetamine action revealed in mice lacking the dopamine disorder. Clin Pharmacokinet 40:753–772. transporter. J Neurosci 18:1979–1986. Martin JC. 1975. Effects on offspring of chronic maternal methamphe- Kase CS. 1986. Intracerebral hemorrhage: non-hypertensive causes. tamine exposure. Dev Psychobiol 8:397–404. Stroke 17:590–595. Martin JC, Martin DC. 1981. Voluntary activity in the aging rat as a Kasirsky G, Tansy MF. 1971. Teratogenic effects of methamphetamine in function of maternal drug exposure. Neurobehav Toxicol Teratol mice and rabbits. Teratology 4:131–134. 3:261–264. Kendrick WC, Hull AR, Knochel JP. 1977. Rhabdomyolysis and shock Martin JC, Martin DC, Radow B, Sigman G. 1976. Growth, development after intravenous amphetamine administration. Ann Intern Med and activity in rat offspring following maternal drug exposure. Exp 86:381–387. Aging Res 2:235–251. Kimmel CA, Buelke-Sam J. 1985. Collaborative Behavioral Teratology Martin JC, Martin DC, Sigman G, Day-Pfeiffer H. 1983. Saccharin Study: background and overview. Neurobehav Toxicol Teratol 7: preferences in food deprived aging rats are altered as a function of 541–545. perinatal drug exposure. Physiol Behav 30:853–858. Kimmel CA, Buelke-Sam J, Adams J. 1985. Collaborative Behavioral Martin JC, Martin DD, Radow B, Day HE. 1979. Life span and pathology Teratology Study: implications, current applications and future in offspring following nicotine and methamphetamine exposure. directions. Neurobehav Toxicol Teratol 7:669–673. Exp Aging Res 5:509–522. Kita T, Wagner GC, Nakashima T. 2003. Current research on metham- McCann UD, Ricaurte GA. 2004. Amphetamine neurotoxicity: accom- phetamine-induced neurotoxicity: animal models of monoamine plishments and remaining challenges. Neurosci Biobehav Rev disruption. J Pharmacol Sci 92:178–195. 27:821–826. Kolesari GL, Kaplan S. 1979. Amphetamines reduce embryonic size and McDougall SA, Duke MA, Bolanos CA, Crawford CA. 1994. Ontogeny of produce caudal hematomas during early chick morphogenesis. behavioral sensitization in the rat: effects of direct and indirect Teratology 20:403–412. dopamine agonists. Psychopharmacology (Berl) 116:483–490. Kolta MG, Scalzo FM, Ali SF, Holson RR. 1990. Ontogeny of the enhanced McElhatton PR, Bateman DN, Evans C, Pughe KR, Thomas SH. 1999. behavioral response to amphetamine in amphetamine-pretreated Congenital anomalies after prenatal ecstasy exposure. Lancet rats. Psychopharmacology (Berl) 100:377–382. 354:1441–1442. Kontur PJ, Won LA, Hoffmann PC, Heller A. 1991. Survival of McElhatton PR, Pughe KR, Evans C, Porter K, Bateman DN, developing dopaminergic neurons in reaggregate tissue culture Thomas SH. 2000. Is exposure to amphetamine-like drugs in following treatment with methamphetamine. Neurosci Lett 129: pregnancy associated with malformations? J Toxicol Clin Toxicol 254–258. 38:195–196. Kraemer T, Maurer HH. 2002. Toxicokinetics of amphetamines: metabo- Michael MI, Khalil SH, Matta CA, Rizk TA. 1987. Normal development of lism and toxicokinetic data of designer drugs, amphetamine, the prenatal mouse eye. Folia Morphol (Praha) 35:228–236. methamphetamine, and their N-alkyl derivatives. Ther Drug Monit Middaugh LD, Blackwell LA, Santos CA 3rd, Zemp JW. 1974. Effects of d- 24:277–289. amphetamine sulfate given to pregnant mice on activity and on Lal S, Sourkes TL. 1973. Ontogeny of stereotyped behaviour induced by catecholamines in the brains of offspring. Dev Psychobiol 7:429–438. apomorphine and amphetamine in the rat. Arch Int Pharmacodyn Milkovich L, van den Berg BJ. 1977. Effects of antenatal exposure to Ther 202:171–182. anorectic drugs. Am J Obstet Gynecol 129:637–642. Lambert NM. 2002. Stimulant treatment as a risk factor for nicotine use Monder H. 1981. Effects of prenatal amphetamine exposure on the and substance abuse. In: Jensen PS, Cooper JR, editors. Attention development of behavior in rats. Psychopharmacology (Berl) 75: deficit hyperactivity disorder. Kingston, NJ: Civic Research Institute. 75–78.

Birth Defects Research (Part B) 74:471–584, 2005 DOI 10.1002/bdrb AMPHETAMINE AND METHAMPHETAMINE 583

Musshoff F. 2000. Illegal or legitimate use? Precursor compounds Ramirez OA, Carrer HF. 1983. Noradrenergic modulation of neuronal to amphetamine and methamphetamine. Drug Metab Rev 32: transmission in the offspring of amphetamine-treated rats. Can J 15–44. Physiol Pharmacol 61:766–769. Naeye RL. 1983. Maternal use of dextroamphetamine and growth of the Ramirez OA, Carrer HF, Nasello AG. 1979. Prenatal amphetamine fetus. Pharmacology 26:117–120. exposure: ovulation, sexual behavior and hypothalamic monoamine Nasello AG, Astrada CA, Ramirez OA. 1974. Effects on the acquisition of content in rats. Pharmacol Biochem Behav 11:605–609. conditioned avoidance responses and seizure threshold in the Ramirez OA, Keller EA, Orsingher OA. 1983. Prenatal amphetamine offspring of amphetamine treated gravid rats. Psychopharmacologia reduces alpha but not beta adrenergic receptor binding in brain of 40:25–31. adult rats. Life Sci 32:1835–1838. Nasello AG, Ramirez OA. 1978. Open-field and Lashley III maze Rapoport JL, Buchsbaum MS, Weingartner H, Zahn TP, Ludlow C, behaviour of the offspring of amphetamine-treated rats. Psycho- Mikkelsen EJ. 1980. Dextroamphetamine. Its cognitive and behavior- pharmacology (Berl) 58:171–173. al effects in normal and hyperactive boys and normal men. Arch Gen Nelson CJ, Felton RP, Kimmel CA, Buelke-Sam J, Adams J. 1985. Psychiatry 37:933–943. Collaborative Behavioral Teratology Study: statistical approach. Rhodium. 2004. Rhodium web site. Available at: http://www.rho- Neurobehav Toxicol Teratol 7:587–590. dium.ws/chemistry/index.html. Nelson MM, Forfar JO. 1971. Associations between drugs administered Rothrock JF, Rubenstein R, Lyden PD. 1988. Ischemic stroke associated during pregnancy and congenital abnormalities of the fetus. Br Med J with methamphetamine inhalation. Neurology 38:589–592. 1:523–527. Safer D, Allen R, Barr E. 1972. Depression of growth in hyperactive NIDA. 2003. InfoFacts: methamphetamine. Available at http:// children on stimulant drugs. N Engl J Med 287:217–220. 165.112.78.61/Infofax/methamphetamine.html. National Institute Safer DJ, Allen RP. 1973. Factors influencing the suppressant effects of on Drug Abuse. two stimulant drugs on the growth of hyperactive children. NIOSH. 2001. RTECS report for phenethylamine, N,alpha-dimethyl-, Pediatrics 51:660–667. hydrochloride, (S) - (1). Available at http://www.cdc.gov/niosh/ Safer DJ, Allen RP, Barr E. 1975. Growth rebound after termination of rtecs/sh533c98.html. NIOSH. stimulant drugs. J Pediatr 86:113–116. NIOSH. 2002. RTECS report for phenethylamine, alpha-methyl-, sulfate Saito TR, Aoki S, Saito M, Amao H, Niwa T, Terada M, Sugiyama M, (2:1), (1-)-. Available at http://www.cdc.gov/niosh/rtecs/ Takahashi KW. 1991. Effects of methamphetamine on copulatory si1ab3f0.html. NIOSH. behavior in male rats. Jikken Dobutsu 40:447–452. Nora JJ, McNamara DG, Fraser FC. 1967. Dexamphetamine Sulfate and Satinder KP, Sterling JW. 1983. Differential effects of pre- and/or post- Human Malformations. Lancet 1:570–571. natal d-amphetamine on avoidance response in genetically selected Nora JJ, Sommerville RJ, Fraser FC. 1968. Homologies for congenital heart lines of rats. Neurobehav Toxicol Teratol 5:315–320. diseases: murine models, influenced by dextroamphetamine. Ter- Sato M, Fujiwara Y. 1986. Behavioral and neurochemical changes in pups atology 1:413–416. prenatally exposed to methamphetamine. Brain Dev 8:390–396. Nora JJ, Trasler DG, Fraser FC. 1965. Malformations in mice induced by Schaiberger PH, Kennedy TC, Miller FC, Gal J, Petty TL. 1993. Pulmonary dexamphetamine sulphate. Lancet 2:1021–1022. hypertension associated with long-term inhalation of ‘‘crank’’ Nora JJ, Vargo TA, Nora AH, Love KE, McNamara DG. 1970. methamphetamine. Chest 104:614–616. Dexamphetamine: a possible environmental trigger in cardiovascular Schertz M, Adesman AR, Alfieri NE, Bienkowski RS. 1996. Predictors of malformations. Lancet 1:1290–1291. weight loss in children with attention deficit hyperactivity disorder NTP. 1991. Toxicology and carcinogenesis studies of dl-amphetamine treated with stimulant medication. Pediatrics 98:763–769. sulfate (CAS No. 60-13-9) in F344/N rats and B6C3F1 mice (feed Schnieden H. 1974. The effect of some psychoactive drugs and metallic studies). Research Triangle Park, NC: NTP, U.S. Department of salts on the motility of rabbit spermatozoa. ICRS J Int Res Commun Health and Human Services, Public Health Service, National 2:1507. Institutes of Health. Report nr NTP-TR-387. Secker R. 2004. Comments on the draft NTP-CERHR Expert Panel report O’Connor JC, Davis LG, Frame SR, Cook JC. 2000. Evaluation of a Tier I on the reproductive and developmental toxicity of amphetamine and screening battery for detecting endocrine-active compounds (EACs) methamphetamine. Hampshire, UK: Shire Pharmaceuticals Group. using the positive controls testosterone, coumestrol, progesterone, Seliger DL. 1973. Effect of prenatal maternal administration of D- and RU486. Toxicol Sci 54:338–354. amphetamine on rat offspring activity and passive avoidance Oregon Department of Human Services. 2003. Children in methamphe- learning. Physiol Psychol 1:273–280. tamine ‘‘labs’’ in Oregon. CD Summary 52(16). Seliger DL. 1975. Prenatal maternal d-amphetamine effects on emotion- Oro AS, Dixon SD. 1987. Perinatal cocaine and methamphetamine ality and audiogenic seizure susceptibility of rat offspring. Dev exposure: maternal and neonatal correlates. J Pediatr 111:571–578. Psychobiol 8:261–268. Palmowski AM, Tulsi RS. 1987. A quantitative and qualitative study of Sellers EM, Otton SV, Tyndale RF. 1997. The potential role of the craniofacial development in the rat embryo aged 302 hr to 374 hr cytochrome P-450 2D6 pharmcogenetic polymorphism in drug (days 12.5 to 15.5). J Craniofac Genet Dev Biol 7:331–340. abuse. NIDA Research Monograph Series 173:6–26. Paternite CE, Loney J, Salisbury H, Whaley MA. 1999. Childhood Sherman MP, Wheeler-Sherman J. 2000. Cranky babies: outcomes inattention-overactivity, aggression, and stimulant medication his- associated with prenatal amphetamine exposure. J Perinatol 20:478. tory as predictors of young adult outcomes. J Child Adolesc Smith L, Yonekura ML, Wallace T, Berman N, Kuo J, Berkowitz C. 2003. Psychopharmacol 9:169–184. Effects of prenatal methamphetamine exposure on fetal growth and PDR. 2003a. Adderall Tablets (Shire U.S.). Thompson Medical Economics. drug withdrawal symptoms in infants born at term. J Dev Behav PDR. 2003b. Adderall XRt Capsules (Shire U.S.). Thompson Medical Pediatr 24:17–23. Economics. Smith LM, Chang L, Yonekura ML, Grob C, Osborn D, Ernst T. 2001. PDR. 2003c. Desoxyn Tablets (Abbott). Thompson Medical Economics. Brain proton magnetic resonance spectroscopy in children exposed PDR. 2003d. Dexedrine spansule capsules, Dexedrine tablets (GlaxoS- to methamphetamine in utero. Neurology 57:255–260. mithKline). Thompson Medical Ecomonomics. Solanto MV. 2000. Clinical psychopharmacology of AD/HD: implications PDR. 2003e. DextroStat Tablets (Shire U.S.). Thompson Medical for animal models. Neurosci Biobehav Rev 24:27–30. Economics. Spencer TJ, Biederman J, Harding M, O’Donnell D, Faraone SV, Wilens Perez JA Jr, Arsura EL, Strategos S. 1999. Methamphetamine-related TE. 1996. Growth deficits in ADHD children revisited: evidence for stroke: four cases. J Emerg Med 17:469–471. disorder-associated growth delays? J Am Acad Child Adolesc Poulton A, Cowell CT. 2003. Slowing of growth in height and weight on Psychiatry 35:1460–1469. stimulants: a characteristic pattern. J Paediatr Child Health 39: Steiner E, Villen T, Hallberg M, Rane A. 1984. Amphetamine secretion in 180–185. breast milk. Eur J Clin Pharmacol 27:123–124. Price RA, Leckman JF, Pauls DL, Cohen DJ, Kidd KK. 1986. Gilles de la Stek AM, Baker RS, Fisher BK, Lang U, Clark KE. 1995. Fetal responses to Tourette’s syndrome: tics and central nervous system stimulants in maternal and fetal methamphetamine administration in sheep. Am J twins and nontwins. Neurology 36:232–237. Obstet Gynecol 173:1592–1598. Pu C, Broening HW, Vorhees CV. 1996. Effect of methamphetamine on Stek AM, Fisher BK, Baker RS, Lang U, Tseng CY, Clark KE. 1993. glutamate-positive neurons in the adult and developing rat Maternal and fetal cardiovascular responses to methamphetamine in somatosensory cortex. Synapse 23:328–334. the pregnant sheep. Am J Obstet Gynecol 169:888–897. Pu C, Vorhees CV. 1993. Developmental dissociation of methampheta- Stewart JL, Meeker JE. 1997. Fetal and infant deaths associated with mine-induced depletion of dopaminergic terminals and astrocyte maternal methamphetamine abuse. J Anal Toxicol 21:515–517. reaction in rat striatum. Brain Res Dev Brain Res 72:325–328. Sund AM, Zeiner P. 2002. Does extended medication with amphetamine Puig-Antich J, Greenhill LL, Sassin J, Sachar EJ. 1978. Growth hormone, or methylphenidate reduce growth in hyperactive children? Nord J prolactin and cortisol responses and growth patterns in hyperkinetic Psychiatry 56:53–57. children treated with dextro-amphetamine. Preliminary findings. Tan SE. 2003. Prenatal amphetamine exposure alters behavioral reactivity J Am Acad Child Psychiatry 17:457–475. to amphetamine in rats. Neurotoxicol Teratol 25:579–585.

Birth Defects Research (Part B) 74:471–584, 2005 DOI 10.1002/bdrb 584 GOLUB ET AL.

Tariq M, Parmar NS, Qureshi S, el-Feraly FS, Al-Meshal IA. 1987. startle facilitation in rats as a function of prepulse stimulus intensity. Clastogenic evaluation of cathinone and amphetamine in somatic Neurotoxicol Teratol 18:135–139. cells of mice. Mutat Res 190:153–157. Wagner GC, Schuster CR, Seiden LS. 1981. Neurochemical consequences Tavares MA, Silva MC. 1993. Body weight gain and hippocampal following administration of CNS stimulants to the neonatal rat. volumes of rats exposed neonatally to psychostimulants. Brain Res Pharmacol Biochem Behav 14:117–119. 619:137–145. Weissman AD, Caldecott-Hazard S. 1993. In utero methamphetamine Tavares MA, Silva MC, Silva-Araujo A, Xavier MR, Ali SF. 1996. Effects of effects: I. Behavior and monoamine uptake sites in adult offspring. prenatal exposure to amphetamine in the medial prefrontal cortex of Synapse 13:241–250. the rat. Int J Dev Neurosci 14:585–596. White House Office of National Drug Control Policy. 2004. Statement of Tonge SR. 1973. Permanent alterations in catecholamine concentrations in Scott Burns-Deputy Director for State and Local Affairs-White House discrete areas of brain in the offspring of rats treated with Office of National Drug Control Policy-Before the House Committee methylamphetamine and chlorpromazine. Br J Pharmacol 47:425–427. on Government Reform-Subcommittee on Criminal Justice, Drug Trent F, Nakamura S, Tepper JM. 1991. Amphetamine exerts anomalous Policy and Human Resources-‘‘Fighting Methamphetamine in the effects on dopaminergic neurons in neonatal rats in vivo. Eur J Heartland: How Can the Federal Government Assist State and Local Pharmacol 204:265–272. Efforts?’’. Washington, DC: White House Office of National Drug Tsuchida K, Akiyama K, Sakai K, Ujike H, Li X, Kuroda S. 1996. Ontogeny Control Policy. p 1–10. of striatal dopamine release in rats after acute administration of Wilens T. 2002. Attention deficit hyperactivity disorder and substance methamphetamine. Pharmacol Biochem Behav 53:575–580. abuse disorders-the nature relationship, subtypes at risk, and Tsuchida K, Ujike H, Kanzaki A, Fujiwara Y, Akiyama K. 1994. Ontogeny treatment issues. In: Jensen PS, Cooper JR, editors. Attention deficit of enhanced striatal dopamine release in rats with methampheta- hyperactivity disorder. Kingston, NJ: Civic Research Institute. p 1–17. mine-induced behavioral sensitization. Pharmacology Biochemistry Wilens TE, Faraone SV, Biederman J, Gunawardene S. 2003. Does And Behavior 47:161–169. stimulant therapy of attention-deficit/hyperactivity disorder beget U.S. Department of Justice. 2002. Children at risk. Information Bulletin: later substance abuse? A meta-analytic review of the literature. 1-7. Pediatrics 111:179–185. U.S. Department of Justice. 2003. Children at clandenstine methamphe- Williams MT, Blankenmeyer TL, Schaefer TL, Brown CA, Gudelsky GA, tamine labs: helping meth’s youngest victims. OVC Bulletin:1-11. Vorhees CV. 2003a. Long-term effects of neonatal methamphetamine UN. 2002. Psychotropic substances. Available at http://www.incb.org/ exposure in rats on spatial learning in the Barnes maze and on cliff e/index.htm?: United Nations International Narcotics Control Board. avoidance, corticosterone release, and neurotoxicity in adulthood. van Tonningen-van Driel MM, Garbis-Berkvens JM, Reuvers-Lodewijks Brain Res Dev Brain Res 147:163–175. WE. 1999. [Pregnancy outcome after ecstasy use; 43 cases followed Williams MT, Inman-Wood SL, Morford LL, McCrea AE, Ruttle AM, by the Teratology Information Service of the National Institute for Moran MS, Rock SL, Vorhees CV. 2000. Preweaning treatment with Public Health and Environment (RIVM)]. Ned Tijdschr Geneeskd methamphetamine induces increases in both corticosterone and 143:27–31. ACTH in rats. Neurotoxicol Teratol 22:751–759. Varley CK, Vincent J, Varley P, Calderon R. 2001. Emergence of tics in Williams MT, Moran MS, Vorhees CV. 2003b. Refining the critical period children with attention deficit hyperactivity disorder treated with for methamphetamine-induced spatial deficits in the Morris water stimulant medications. Compr Psychiatry 42:228–233. maze. Psychopharmacology (Berl) 168:329–338. Vorhees CV. 1985. Behavioral effects of prenatal d-amphetamine in rats: a Williams MT, Morford LL, Wood SL, Wallace TL, Fukumura M, Broening parallel trial to the Collaborative Behavioral Teratology Study. HW, Vorhees CV. 2003c. Developmental D-methamphetamine treat- Neurobehav Toxicol Teratol 7:709–716. ment selectively induces spatial navigation impairments in reference Vorhees CV, Acuff-Smith KD. 1990. Prenatal methamphetamine-induced memory in the Morris water maze while sparing working memory. anophthalmia in rats. Neurotoxicol Teratol 12:409. Synapse 48:138–148. Vorhees CV, Ahrens KG, Acuff-Smith KD, Schilling MA, Fisher JE. 1994a. Williams MT, Vorhees CV, Boon F, Saber AJ, Cain DP. 2002. Methamphe- Methamphetamine exposure during early postnatal development in tamine exposure from postnatal day 11 to 20 causes impairments in rats: I. Acoustic startle augmentation and spatial learning deficits. both behavioral strategies and spatial learning in adult rats. Brain Psychopharmacology (Berl) 114:392–401. Res 958:312–321. Vorhees CV, Ahrens KG, Acuff-Smith KD, Schilling MA, Fisher JE. 1994b. Won L, Bubula N, McCoy H, Heller A. 2001. Methamphetamine Methamphetamine exposure during early postnatal development in concentrations in fetal and maternal brain following prenatal rats: II. Hypoactivity and altered responses to pharmacological exposure. Neurotoxicol Teratol 23:349–354. challenge. Psychopharmacology (Berl) 114:402–408. Won L, Kontur PJ, Choi HK, Hoffmann PC, Heller B, Heller A. 1992. Vorhees CV, Inman-Wood SL, Morford LL, Broening HW, Fukumura M, Acute and persistent effects of methamphetamine on developing Moran MS. 2000. Adult learning deficits after neonatal exposure to monoaminergic neurons in reaggregate tissue culture. Brain Res D-methamphetamine: selective effects on spatial navigation and 575:6–12. memory. J Neurosci 20:4732–4739. Yamamoto Y, Yamamoto K, Abiru H, Fukui Y, Shiota K. 1995. Effects of Vorhees CV, Morford LL, Inman SL, Reed TM, Schilling MA, Cappon GD, methamphetamine on rat embryos cultured in vitro. Biol Neonate Moran MS, Nebert DW. 1999. Genetic differences in spatial learning 68:33–38. between Dark Agouti and Sprague–Dawley strains: possible correla- Yamamoto Y, Yamamoto K, Fukui Y, Kurishita A. 1992. Teratogenic tion with the CYP2D2 polymorphism in rats treated neonatally with effects of methamphetamine in mice. Nippon Hoigaku Zasshi methamphetamine. Pharmacogenetics 9:171–181. 46:126–131. Vorhees CV, Pu C. 1995. Ontogeny of methamphetamine-induced Yamamoto Y, Yamamoto K, Hayase T. 1999. Effect of methamphetamine neurotoxicity in the rat model. NIDA Res Monogr 158:149–171. on male mice fertility. J Obstet Gynaecol Res 25:353–358. Vorhees CV, Reed TM, Schilling MA, Fisher JE, Moran MS, Cappon GD, Yamamoto Y, Yamamoto K, Hayase T, Fukui Y, Shiota K. 1998. Effects of Nebert DW. 1998. CYP2D1 polymorphism in methamphetamine- amphetamine on rat embryos developing in vitro. Reprod Toxicol treated rats: genetic differences in neonatal mortality and effects on 12:133–137. spatial learning and acoustic startle. Neurotoxicol Teratol 20: Yasuda M, Ariyuki F, Nishimura H. 1965. Effect of amphetamine on 265–273. pregnancy in Icr-Jcl mice. Okajimas Folia Anat Jpn 41:227–231. Vorhees CV, Reed TM, Schilling MS, Acuff-Smith KD, Fisher JE, Moran Zone SE, Buchi KF. 2002. Prenatal amphetamine use in Utah’s newborn MS. 1996. Neonatal methamphetamine-induced long-term acoustic intensive care units. Pediatr Res 51:360A.

Birth Defects Research (Part B) 74:471–584, 2005 DOI 10.1002/bdrb