Current Update

 

Reprinted from the Journal of Comparative Neurology 435:259-262 (2001). This article is a US Government work and, as such, is in the public domain in the United States of America.

Commentary

No Time for Complacency: The Fetal Brain on Drugs

BARRY E. KOSOFSKY
Department of Neurology, Massachusetts General Hospital, Harvard Medical School, Laboratory of Molecular and Developmental Neuroscience, Charlestown, Massachusetts 02129

STEVEN E. HYMAN
National Institute of Mental Health, NIH, Bethesda, Maryland 20817-9669

In the current issue of Journal of Comparative Neurology, Lidow and Song (p. 256-269) report dramatic effects of intrauterine exposure to cocaine from embryonic day 40-102 (corresponding to the human second trimester of gestation) on development of the cerebral cortex. The study examined the brains of four cocaine-exposed macaque monkeys compared with four controls, all killed in postnatal year 3. This report, which documents profound alterations in cortical development resulting from intrauterine cocaine exposure, extends the previous work of this group, which had identified similar changes in the brains of macaque monkeys studied at postnatal month 2, after being exposed to cocaine at the same dose and during the identical embryonic period as used in the current report (Lidow, 1995). By using stereologic analyses, the authors document seemingly permanent reductions in the density and number of cerebral cortical neurons and disorganization of cortical lamination.

Preclinical Models of Transplacental
Exposure To Drugs of Abuse: Necessary,
But Limited and Controversial

Convincing animal models for the direct effects of cocaine exposure in utero are sorely needed. This is especially true given the complexities of the human situation in which cocaine is frequently used alongside other drugs often in a context of poor nutrition and overall health, and incomplete or absent prenatal care. Without simpler animal models, it will be difficult to identify the specific contribution of cocaine in producing adverse neurodevelopmental outcomes, to design effective therapies, or to appropriately time and target prevention efforts. However, this very complexity has given rise to substantial debates about the relevance of any given animal model to the human situation with key controversies focused on dosing, route of administration, period of exposure, and potential interactive effects of multiple drugs and poor health status.

The model of Lidow et al. is one of three published primate models of intrauterine exposure to cocaine (Lidow, 1995; Ronnekleiv and Naylor, 1995; Morris et al., 1996). Given the relative evolutionary closeness of monkeys to humans, such primate models have been crucially important in providing preclinical evidence that exposure to cocaine in utero could alter human brain development, and in what manner. Additionally, many rat and mouse models, and several rabbit, pig, and sheep models of intrauterine cocaine exposure have been described (for re-views see Spear, 1995; Dow-Edwards, 1996). The route of drug exposure used by Lidow and Song is oral and is the only such model reported, in contrast to intravenous cocaine models (in the pig, sheep, rabbit, and rat), intramuscularly models (in the primate), and subcutaneous models (in the rat and mouse). Given that each species has a unique timing of intrauterine development (i.e., rodents are born at the equivalent of the end of the second trimester of human gestation), it is important to note that Lidow and Song limit the exposure to cocaine in their primate model from E40 to E102, the equivalent of the second trimester of human gestation. It is also important to note that their monkeys received cocaine orally, 20 nig/kg per day, divided into two daily doses. Unfortunately, pharmacokinetic analyses were not performed, making it difficult to be certain that the drug levels were within the range typically experienced by human abusers.

One important controversy in the literature focuses on the direct effects of cocaine vs. indirect effects related to impaired nutrition in the pregnant mothers or to altered placental blood flow. Lidow and Song report no body growth retardation in the cocaine exposed offspring, consistent with the notion that the findings were not due to impaired fetal nutrition, and that the drug regimen used was not sufficiently high to cause anorexia in pregnant macaques. Another controversy in the literature has focused on whether observed alterations in brain structure or function are transient or permanent. By taking the findings out to three years, this study establishes that even if, as appears increasingly unlikely, the abnormalities reverse at a later time, they last long enough to significantly affect postnatal development. An additional controversy in the literature has been the extent to which permanent changes have been extensive, as seen here, or focal, as in a rabbit model in which abnormalities are restricted to the dopamine-rich fields of anterior cingulate cortex (Levitt et aL, 1997; Murphy et al., 1997). The basis for these differences in terms of species, dosing, and other experimental design factors will be important to sort out.

Strengths and Weaknesses of the Model

This investigation isolated the effects of a single drug, cocaine. The dose administered was not associated with ap-parent malnutrition in the mothers; all of the offspring were surrogate-fostered at birth. The advantages of a primate animal model include its relative evolutionary closeness to the human. A strength of this study is the prolonged (three-year) period of postnatal development before neuropathologic analysis, which extended the finding of a prior study in which the animals were allowed to develop postnatally for two months. Experimental choices had to be made, but of necessity raise questions about the generalizability of the findings. For example, the study is restricted to males, and the cocaine exposure is limited to the period from E40-E102. Moreover, only one dose of cocaine was tested, and only one route of drug administration, the oral route, which is not commonly used by either human cocaine addicts or in animal models. (The oral route was chosen because, arguably, it might produce similar pharmacokinetics to human intranasal cocaine use). A deficit in the study was failure to report head circumference at birth and over time, as this is a measure in widespread clinical use as a crude marker of brain growth.

The use of rigorous quantitative stereologic assessments entailed enormous effort, but makes the findings more compelling. Of course, although an initial descriptive study such as this represents a significant achievement, it also raises many questions about mechanism. One of the most obvious questions of mechanism is raised by the observation that similar abnormalities were found in cortical organization and neuron number in three regions of the primate brain, areas 46, 4, and 17, that differ markedly in terms of the neuroanatomic patterns of innervation of dopamine (Lewis et al., 1987), serotonin (Kosofsky et al., 1984), and norepinephrine (Lewis and Morrison, 1989).

Despite the limitations of the model, and the unanswered questions left behind by this initial analysis, the study of Lidow and Song reach a dramatic series of conclusions. Intrauterine exposure to cocaine can yield long-lived, if not permanent, and global alterations in cytoarchitecture of all regions of primate neocortex studied as follows: (1) The surface area of the neocortex is unchanged. (2) The neocortical volume is globally decreased by 20% (0.8 normal). (3) Cortical thickness is decreased globally. (4) Neuronal density is decreased by 50% (0.5 normal). (5) Neuronal numbers are decreased by 60% (0.8 x 0.5 - 0.4 normal). (6) The number of neurons in the white matter (both granule and pyramidal cells) are increased 100% in cocaine-exposed monkeys. (7) Even when these misplaced neurons are counted, cocaine-exposed animals are still missing over half of the neurons in all cortical regions assessed. (8) This loss of neurons, which presumably results from impaired neuronal migration, results in a loss of lamination and blurring of regional cytoarchitectonic differences, with an admixture of cell types evident in all layers in all neocortical structures examined.

Scientific Implications of the
Neuropathologic Findings Reported By
Lidow And Song

Although it is easy to focus on the abnormalities, one should not lose sight of the fact that many critical aspects of corticogenesis proceeded normally; specifically, neuro-genesis, gliogenesis, and tangential cortical growth all seem essentially normal in the brains of the three-year-old cocaine-exposed primates compared with controls. The generation of appropriate neuronal cell classes seems to have proceeded on schedule, producing both pyramidal cells and granule cells at the appropriate time and place in cocaine-exposed monkeys, as measured by thymidine "birthdating." Thus, it seems that many fundamental processes underlying neurogenesis and neuronal phenotype specification are not altered by gestational cocaine exposure, at least as assessed by this primate model. The current report identifies aspects of corticogenesis that are disrupted: neuronal migration and survival of both granule cells and pyramidal cells in all cortical regions examined. This results in a dramatic and permanent decrease in final neuronal number, with a loss of cortical lamination. The cocaine-induced alteration in neuronal maturation is long-lasting in contrast to the decrease in glial cells evident at two months (see prior report), which appears to normalize by three years of age (see the current report). The authors do not identify whether radial and/or tangential migration of neurons into the cortex are specifically affected (see below), but they do establish that such changes in neuronal migration and cortical lamination are not limited to dopamine-rich areas. If in fact neurogenesis is not altered, although final neuronal number is, then the implication as identified by the authors is that "early" cell death is occurring.

The mechanisms underlying cocaine-exposed alterations in the migration of cortical neurons are not investigated in this report. This underscores the importance of pursuing additional mechanistic experiments. Mechanistic investigations are currently under way in several laboratories using several different approaches, methods, and animal models. That the neuropathologic changes reported are not limited to dopamine (or serotonin, or norepinephrine) rich areas does not preclude the possibility that aminergic mechanisms may contribute to the neuropathologic abnormalities reported. Errors in neuronal migration are suggested by the neuroanatomic alterations seen in the brains of mice with genetic inactivation of the dopamine transporter (Giros et al., 1996), dopamine D1a receptor (Drago et al., 1994; Xu et al., 1994), and monoamine oxidase (Cases et al., 1996). However, the global and homogeneous changes reported by Lidow and Song argues against focal vascular pathophysiologic mechanisms being the critical mechanism underlying the abnormalities.

If the findings of the current report are replicated, it would be important to investigate whether there is selective vulnerability for particular classes of neurons. Lidow and Song report that pyramidal projection neurons as well as granule cells are arrested in their migration into the cortex. However, it is not known whether there is alteration in the tangential migration of GABAergic interneurons that originate in the ganglionic eminence and that provide most of the inhibitory GABAergic interneurons within the cortical circuitry (Parnavelas, 1992; Anderson et al., 1999). It will also be critical to determine whether cells that are in the wrong place are necessarily functionally abnormal and what the functional implications of having reduced neuronal numbers is. Should the findings of global alterations in neuronal number and position replicate, it will be important to determine whether the effects on cognition and social behavior are in some sense "global" or whether specific deficits can be identified. Inherent in these questions are considerations of the nature and extent of plasticity and ability to compensate for such neuropathologic changes, at the local circuit, large systems, and behavioral levels.

Clinical Implications of the Neuropathologic
Findings Reported By Lidow and Song

At first glance, it is not necessarily surprising that drugs with powerful neurochemical effects, that are administered repetitively, may alter brain development in the fetus. Cocaine is a psychostimulant and, like other drugs of abuse, has been implicated as a behavioral teratogen (drugs capable of altering fetal brain development, with lasting alterations in brain function). However, unlike alcohol, a teratogen that can produce an external morphologic phenotype, as is evident in the fetal alcohol syndrome CFAS), the identification of more subtle neurodevelopmental phenotypes after intrauterine exposure of a fetus to certain other drugs of abuse such as nicotine, cocaine, marijuana, or heroin has been more elusive. The effects of drugs that alter brain development without a physical phenotype or an easily and reliably identified neurodevelopmental behavioral phenotype such as seizures or retardation may well go underdetected and underreported, despite significant impact on brain development and behavior. Even when a substantial exposure history is known, there is no "gold standard" by which clinicians can identify and distinguish one drug-induced phenotype from another or from other conditions. For these reasons, the convergence of preclinical models and clinical findings are critically important. In preclinical studies, such as the one currently reported, it is possible to control pre- and postnatal factors and examine the independent contribution of a single drug such as cocaine to adverse neurodevelopmental outcomes. However, given the limitations inherent in all preclinical animal models, including the one used by Lidow and Song, one is left to wonder whether, "even if the findings are convincing, are they relevant?" In contrast, in clinical studies, which face the extreme complexities of interacting genetic, intrauterine, and postnatal environmental factors that contribute to the drug-exposed phenotype, one is apt to conclude "these findings are clearly relevant, but can they be understood?" If the goal is to understand the impact of intrauterine exposure to drugs of abuse to produce effective early interventions, it is critical to stimulate a dialogue between preclinical and clinical investigators. One wants findings that are both replicable and relevant.

In this light, it is critically important to consider the generalizability of the current findings: one must ask whether there is a species specificity, the extent to which route, dose, frequency, and timing contributed to the findings, and whether the factors that contributed to the magnitude and global nature of the alterations in cortical cytoarchitecture described, are operating in the human fetus exposed to cocaine in utero. These are hypotheses that can potentially be tested in the clinical setting, but improved neuroimaging tools will be needed. When post-mortem tissue is available, it must be used wisely, by investigators who are mindful of hypotheses generated by preclinical models, and of potentially confounding coexposures and variables. High-resolution magnetic resonance (MR) techniques that are currently being developed could perhaps identify alterations in cortical lamination, and application of MR spectroscopy may additionally provide in vivo data regarding alterations in neuronal and glial cells in cocaine-exposed children (Smith et al., 2001). A forthcoming report (Bookstein et al., 2001) using a sophisticated morphometric MR analytic method to identify alterations in the size, shape, and contour of the corpus callosum of adults with FAS and fetal alcohol exposure compared with controls may provide such a marker for the alcohol-exposed phenotype. Given the profound loss of cortical neurons, and presumably cortico-cortical connections in cocaine-exposed primates reported by Lidow and Song, there is an urgent need to perform a comparable analysis on brain images obtained from cocaine-exposed children, with strict attention to the potential additive or synergistic (and likely confounding!) effects of concomitant alcohol exposure, which is commonly co-abused by pregnant individuals addicted to cocaine.

If the cortical abnormalities reported by Lidow and Song are relevant to the human situation, one might expect global delays in cognitive development. Current clinical research does not seem to bear out such a prediction, however (see, for example, the longitudinal studies re-ported in Harvey and Kosofsky, 1998). Although some meta-analyses of longitudinal studies of children exposed to cocaine in utero report subtle deficits in cognition (Lester et al., 1998), others have not identified such deficits as being specifically attributable to cocaine (Frank et al., 2001). In fact, the review by Frank et al. concluded "among children aged six years or younger, there is no convincing evidence that prenatal cocaine exposure is associated with developmental toxic effects that are different in severity, scope, or kind from the sequelae of multiple other risk factors. . .including prenatal exposure to tobacco, marijuana, or alcohol". Such a statement does not exonerate cocaine, however, because all of these drugs are significant neurobehavioral teratogens; collectively, in fact, drugs of abuse may represent the largest preventable cause of developmental disability. In the case of marijuana and nicotine (see for example Fried et al., 1998), and alcohol (see for example Streissguth et al., 1996), the specific, discrete, and sometimes handicapping deficits attributed to intrauterine exposure to drugs of abuse may not become apparent until children reach their teenage years and young adult ages. This may well be the case for cocaine, as data evolving from the ongoing Maternal Life Style study sponsored by NIH suggest that if and when there are behavioral and neurodevelopmental deficits in cocaine-exposed infants and young children, such deficits may be subtle and contextual (Lester, 1998). The extent to which the types of neuropathologic changes reported by Lidow and Song contribute to such deficits in humans during the first years of life and thereafter remains to be determined.

Public Policy Implications of the
Neuropathologic Findings Reported By
Lidow and Song

Public understanding of the impact of prenatal cocaine exposure has lurched from an initial "over-reaction," which characterized drug-exposed infants and children as irrevocably and irreversibly damaged to a subsequent, excessive "sigh of relief" with which many argued that drugs such as cocaine do not have lasting effects if children are raised in "good environments." In fact, there is much preclinical evidence that cocaine and other drugs of abuse are neuroteratogens that can produce serious abnormalities in brain development. What we are learning is that the behavioral impact of any such neural abnormalities that might occur in humans depends on other complex pre- and postnatal factors, which may include genetic vulnerability as well. We must learn in an appropriate scientific manner whether infants who have been exposed to cocaine or other drugs in utero have a milder phenotype if raised in healthy environments, and if so, what specific environmental factors might promote resilience.
It is anticipated that significant advances in preclinical and clinical research over the coming years will provide a much clearer picture and deeper understanding of the long-term implications of intrauterine cocaine exposure on infant and child development, academic performance, socialization, and neurologic and psychiatric well being. However, even before fully satisfactory data are in, an interim synthesis of the preclinical and clinical data available today suggests that society must take the problems of substance abuse during pregnancy very seriously. Programs that can effectively help addicted pregnant women avoid drugs deserve to be a priority. Early evidence also suggests that postnatal interventions can be helpful. By identifying drug- and alcohol-exposed children in the early postnatal period, and providing appropriate services for them, benefit is likely to accrue to the children and also to their families, schools, and communities (for example, see Karoly et al., 1998). It is not yet clear which prevention and treatment models are optimal. Especially because resources seem to be very limited (see Kaltenbach and Finnegan, 1998; Chavkin, 2001), rigorous clinical trials are warranted, and adoption of well-tested programs with fidelity makes public health sense. One public health strategy for improving the response of society to the cocaine-addicted, pregnant woman and her fetus has been published recently (Wenzel et al., 2001).
In summary, we are far from having the complete scientific picture we need at the neurobiological, behavioral, clinical, and health policy levels. What we do know, however, should be a strong antidote to complacency. We certainly have enough information to make it a priority to identify and treat addicted individuals who are pregnant and to make serious ameliorative efforts aimed at their children.

LITERATURE CITED

Anderson S, Mione M, Yun K. Rubenstein JL. 1999. Differential origins of neocortical projection and local circuit neurons: role of Dix genes in neocortical interneuronogenesis. Cereb Cortex 9:646-654.

Bookstein FL, Sampson PD, Streissguth AP, Connor PD. 2001. Geometric morphometrics of corpus callosum and subcortical structures in tlie fetal-alcohol-affected brain. Teratology (in press).

Cases 0, Vitalis T, Seifl, De Maeyer E, Sotelo C, Caspar P. 1996. Lack of barrels in the somatosensory cortex of monoamine oxidase A-deficient mice: role of a serotonin excess during the critical period. Neuron 16:297-307.

Chavkin W. 2001. Cocaine and pregnancy: time to look at the evidence. JAMA 285:1626-1628.

Dow-Edwards D. 1996. Comparability of human and animal studies of developmental cocaine exposure. NIDA Res Monogr 164:146-174.

Drago J, Gerfen CR, Lachowicz JE, Steiner H. Hollon TR, Love PE, Ooi GT, Grinberg A. Lee EJ, Huang SP, Bartlett PF. Jose PA, Sibley DB,
Westphal H. 1994. Altered striatal function in a mutant mouse lacking
D,A dopamine receptors. Proc Nati Acad Sri U S A 91:12564-12568.

Frank DA. Augustyn M, Knight WG, Pell T, Zuckerman B. 2001. Growth, development, and behavior in early childhood following prenatal co-caine exposure: a systematic review. JAMA 285:1613-1625.

Fried PA, Watkinson B, Gray R. 1998. Differential effects on cognitive functioning in 9- to 12-year olds prenatally exposed to cigarettes and marihuana. Neurotoxicol Teratol 20:293-306.

Giros B. Jaber M, Jones SR, Wightman RM. Caron MG. 1996. Hyperloco-motion and indifference to cocaine and amphetamine in mice lacking the dopamine transporter. Nature 379:606-612.

Harvey JA, Kosofsky BE. 1998. Cocaine: Effects on the developing brain. New York: Ann New York Acad of Sci. Volume 846.

Kaltenbach K, Finnegan L. 1998. Prevention and treatment issues for pregnant cocaine-dependent women and their infants. Ann N Y Acad
Sci 846:329-334.

Karoly LA, Greenwood PW, Everingham SS, Hoube J, Kilburn MR, Rydell CP, Sanders M, Chiesa J. 1998. Investing In Our Children. Santa
Monica: RAND.

Kosofsky BE. Molliver ME. Morrison JH, Foote SL. 1984. The serotonin and norepinephrine innen'ation of primary visual cortex in the cyno-molgus monkey (Macaca fascicularis). J Comp Neurol 230:168-178.

Lester BM. 1998. The Maternal Lifestvles Study. Ann N Y Acad Sci
846:296-305.

Lester BM, LaGasse LL, Seifer R. 1998. Cocaine exposure and children: the meaning of subtle effects. Science 282:633-634.

Levitt P, Harvey JA, Friedman E. Simansky K, Murphy EH. 1997. New evidence for neurotransmitter influences on brain development. Trends Neurosci 20:269-274.

Lewis DA. Morrison JH. 1989. Noradrenergic innervation of monkey pre-frontal cortex: a dopamine-beta-hydroxylase immunohistochemical study. J Comp Neurol 282:317-330.

Lewis DA, Campbell MJ, Foote SL, Goldstein M, Morrison JH. 1987. The distribution of tyrosine hydroxylase-immunoreactive fibers in primate neocortex is widespread but regionally specific. J Neurosci 7:279-290.

Lidow MS. 1995. Prenatal cocaine exposure adversely affects development of the primate cerebral cortex. Synapse 21:332-341.

Lidow MS, Song Z-M. 2001. Primates exposed to cocaine in utero display reduced density and numbers of cerebral cortical neurons. J Comp Neurol 435:265-277.

Morris P, Binienda Z, Gillam MP, Harkey MR. Zhou C. Henderson GL.
Paule MG. 1996. The effect of chronic cocaine exposure during preg-nancy on maternal and infant outcomes in the rhesus monkey. Neuro-toxicol Teratol 18:147-154.

Murphy EH, Fischer I. Friedman E, Grayson D, Jones L. Levitt P, O'Brien-Jenkins A, Wang HY, Wang XH. 1997. Cocaine administration in pregnant rabbits alters cortical structure and function in their progeny in the absence of maternal seizures. Exp Brain Res 114:433-441.

Parnavelas JG. 1992. Development of GABA-containing neurons in the visual cortex. Prog Brain Res 90:523-537.

Ronnekleiv OK, Naylor BR. 1995. Chronic cocaine exposure in the fetal rhesus monkey: consequences for early development of dopamine neu-rons. J Neurosci 15:7330-7343.

Smith LM. Chang L, Yonekura ML, Gilbride K, Kuo J, Poland RE, Walot
I, Ernst T. 2001. Brain proton magnetic resonance spectroscopy and imaging in children exposed to cocaine in utero. Pediatrics 107:227-
231.

Spear LP. 1995. Neurobehavioral consequences of gestational cocaine ex-posure: a comparative analysis. In: Rovee-Collier C, Lipsitt LP, editors. Advances in infancy research. Norwood: Ablex. p 55-105.

Streissguth AP, Barr HM, Kogan J. Bookstein FL. 1996. Understanding the occurrence of secondary disabilities in clients with fetal alcohol syndrome and fetal alcohol effects. Seattle: University of Washington
Press.

Wenzel SL, Kosofsky BE, Harvey JA, Iguchi MI, Steinberg P. Watkins KE, Shaikh R. 2001. Prenatal cocaine exposure: scientific considerations and public policy implications. Santa Monica: RAND.

Xu M, Moratalla R, Gold LH, Hiroi N, Koob GF. Graybiel AM, Tonegawa S. 1994. Dopamine Dl receptor mutant mice are deficient in striatal expression of dynorphin and in dopamine-mediated behavioral re-sponses. Cell 79:729-742.

 

Previous Updates