Ionizing radiation and the developing brain

Neurotoxicology and Teratology, Vol. 12, pp. 249-260. © Pergamon Press plc, 1990. Printed in the U.S.A. 0892-0362/90 $3.00 + .00

Ionizing Radiation and the Developing Brain

W I L L I A M J. S C H U L L , * S T A T A N O R T O N t A N D R O N A L D P. JENSH$1

*Center for Demographic and Population Genetics Graduate School of Biomedical Sciences

University of Texas Health Science Center, Houston, TX 77225 ?Department of Pharmacology, Toxicology, and Therapeutics University of Kansas Medical Center, Kansas City, KS 66103

~Department of Anatomy, Jefferson Medical College of Thomas Jefferson University Philadelphia, PA 19107-6731

SCHULL, W. J., S. NORTON AND R. P. JENSH. Ionizing radiation and the developing brain. NEUROTOXICOL TERATOL 12(3) 249-260, 1990.--The unique susceptibility of the central nervous system to radiation exposure is attributable to its extensive period of development, the vulnerability of its neuronal ceils, the migratory activity of many of its cells, its inability to replace mature neurons, and the complexity of the system itself. Radiation effects may be due to glial or neuronal cell death, interruption of migratory activity, impaired capacity to establish correct connections among cells, and/or alterations in dendritic development. These structural changes are often manifested as behavioral alterations later in life. Sensitivity to radiation (dose-response) is markedly similar among all mammalian species when developmental periods are compared. This review compares and contrasts human and animal behavioral data. Neonatal and postnatal adult behavioral tests have been shown to be sensitive, noninvasive measures of prenatal radiation exposure, although currently their predictive validity for humans is uncertain. Additional research is needed to determine the presence and significance of postnatal morphologic and functional alterations due to prenatal exposure to low levels of ionizing radiation.

Ionizing radiation Prenatal Brain In utero Behavior development Fetal Perinatal

DIFFERENT embryonic and fetal structures differ in their susceptibility to teratogenic insults. A distinct feature of the developing human brain is its long period of sensitivity. This undoubtedly reflects its architectural complexity, its long developmental (and hence sensitive) period, the vulnerability of the undifferentiated neural cell as compared to the developed neuron, the fact that neuronal function is contingent upon position and that neuronal cells do not proliferate in the cortex, but must migrate there, and the inability of the brain to replace lost neurons. Patently, the brain is the culmination of an interrelated sequence of molecular, cellular and tissue events. However, it is sufficient here to note that the development of the human brain and its adnexa differs from that of most other organs or organ systems in that:

a) the brain is one of the most complex organs of the body, with an involved architecture in which different functions are localized in different structures. Differentiation of the latter takes place at different times and for different durations. This is particularly true of the development of the neocortex, which proceeds over a long time.

b) Brain function critically depends on the disposition and interconnection of structures and cells, and normal structure and function hinge on an orderly sequence of events, each of which must occur correctly in time and space.

c) Neurons are not self-renewing. The capacity of neuronal precursors to divide is exhausted during cortical histogenesis and

culminates in differentiated neurons which do not undergo further division.

Malformations of the central nervous system occur in the course of major organogenesis, or during the differentiation and growth of the brain mantle. Recent advances in developmental neurobiology have shown that many of these stem from failures in the normal interactions of cells (neural and nonneural) during the development of the primate brain. Normal interactions hinge upon (a) production of a sufficient number of neurons, (b) their appropriate positioning, (c) establishment of the requisite cell shapes, and (d) formation of synaptic connections (53, 54, 60). The sensitivity of the differentiating neural tissue to damage changes with age, but so too does its capacity to replace damaged cells. At any given time in development the probability of causing abnormalities, and their severity, changes as the dose of the teratogenic agent changes. Moreover, since for the same dose damage can vary as a function of the time in development at which the insult occurs, malformations can also differ substantially in severity.

The cells of the different structures of the brain are produced at different times. For the rat, structures have been identified in Fig. 1, based on gestational day 1 designated as the time when sperm were identified in the vaginal canal. If the proliferating ventricular and subventricular cells are damaged during periods when a certain cell type is being produced, the loss may be permanent. A

1Requests for reprints should be addressed to Dr. R. P. Jensh, Department of Anatomy, JAH 561, Jefferson Medical College, 1020 Locust Street, Philadelphia, PA 19107-6799.

249

250 SCHULL, NORTON AND JENSH

"n

~ EYE ORBIT - L " % CEREBRAL CORTEX

Y WEIGHT ~--~,~.,~ORPUS CALLOSUM

~~_~CEREBELLUM~IPPOCAMPusPURKINJEpyRAMIDAL

i i ..I \ \ /

¢4

/ ==.

/

I J E RETINA

DENTATE GRANULE

CEREBELLUM GRANULE

FIG. 1. Diagram of periods of maximum sensitivity to gamma radiation in the developing rat.

brief insult may, therefore, lead to preferential damage to a particular region and consequently bring about a permanent functional or behavioral abnormality.

IONIZING RADIATION AS A CENTRAL NERVOUS SYSTEM TERATOGEN

Ionizing radiation is a unique agent for causing in utero damage to offspring. Unlike drugs, there are no considerations of placental transfer or metabolism; timing of exposure is precise; maternal effects are minimal with doses below 1.0 Gy (100 RAD). Therefore, variability in postnatal behavior of offspring has fewer sources than in chemically exposed fetuses.

Ionizing radiation is a recognized cause of mental retardation in humans exposed in utero. It has been recently suggested that mental retardation may be a more serious risk than cancer following in utero exposure (23). The gestational period during which mental retardation is caused by ionizing radiation has been defined. In spite of differences in rate of development of various brain structures, comparable periods of animal brain development can be examined. There is very good correlation with specific changes in the brain. The sequence of events is certainly the same. The primary difference may be their duration in time. The embryology of the rat brain is well known, and parallels exist in rats and humans. There are no known significant differences in sensitivity of similar cell types between the 2 species. Chronic exposure may be more detrimental to the human because of the longer period of time necessary to achieve a given end.

Ionizing radiation could interfere with the development of the brain in a variety of ways (3, 4, 20, 22). Radiation effects could arise from the death of glial or neuronal precursors or both, or the killing of postmitotic, but still immature, neurons. Such effects could stem from an intrusion on migration either through an alteration of the cell surface phenomena that are involved or through the death of the glial cells that guide the migrating neurons (52, 55-59). It is not clear whether neuronal and glial cells are

TABLE 1

TIME TABLE OF SELECTED MALFORMATIONS FOLLOWING EXPOSURE OF THE DEVELOPING RAT TO IONIZING RADIATION (l.0-2.0 Gy)

Area Affected Developmental Stage Peak Effect References

Cerebral cortex

Corpus callosum

Olfactory tubercle and Anterior commissure

Hippocampus 1) Pyramidal cells

2) Dentate granule cells

Cerebellum 1) Purkinje cells 2) Granule cells

Eye 1) Orbit 2) Retina

Body weight

gd 11-18

gd 13-20

gd 13-15

gd 14-18

pnd 2-15

gd 13-16 pnd 4-21

gd 9-12 gd 21-pnd 7

gd 13-18

gd 14-17 Hicks and D'Amato (21) Norton and Kimler (46)

gd 14-17 Hicks and D'Amato (22)

gd 13-14 Hicks et al. (19) Mullenix et al. (39)

? Tamaki and Inouye (68) Hicks and D'Amato (21) Gazarra and Altman (13)

gd 15 Altman and Bayer (1) pnd 7-15 Altman and Bayer (1)

gd 10 Hicks et al. (18) ? Hicks and D'Amato (20)

gd 15 Hicks and D'Amato (20) Norton and Kimler (46)

IONIZING RADIATION--DEVELOPING BRAIN 251

TABLE 2

PREWEANING MOTOR ACTS*

Gy gd Ou~ome

Air fighting 0.6 10 NS Air righting 0.6 18 delayed

Surface fighting 0.6 10 NS Surface fighting 0.6 18 NS Surface fighting 1.25 15 delayed

Negative geotaxis 0.6 10 NS Negative geotaxis 0.6 18 NS Negative geotaxis 1.0 11 delayed

Reflex suspension 1.0 1 l NS Reflex suspension 1.0 13 NS Reflex suspension 1.0 15 delayed Reflex suspension 1.0 17 NS

*Data from Norton (44); Jensh and Brent (25-27); Kimler and Norton (31).

NS = not significant.

equally radiosensitive; however, disturbances of myelin formation, a mature glial function, have not been described in experimental situations following irradiation. Third, abnormality might reflect an impaired capacity of the neurons to connect correctly. Development of neuronal connection, or synaptogenesis, is a multifactorial phenomenon; it involves timing, space, possibly diffusible agents, and surface-mediated competition. Irradiation could also lead to disoriented dendritic arborization, or a reduced number of dendrites or dendritic spines per cerebral cortical neuron. The establishment of cell connections is a competitive process and neuronal cells that do not establish connections die. A late migrating neuron could find its opportunities to establish connections preempted by earlier migrating cells.

EFFECTS OF RADIATION EXPOSURE

The limitations of the human data make inevitable the use of 0.25 other animal species for both descriptive and experimental studies. Although extrapolations must be made with care, it is clear that the use of experimental animals is vital to progress in understanding the effects of potentially neurotoxic substances (61). At the same time, direct evidence from human studies, especially that of a 0.5 quantitative nature, will eventually be the most convincing.

Animal Data

The most commonly used species for experimental studies of ionizing radiation is the rat. As noted in Table 1, some generalizations can be made, based on the available data, for the developmental periods during which brain structures are most sensitive to ionizing radiation. However, not all behavioral studies have attempted to link specific morphological damage with behavioral outcome. The timing of exposure to ionizing radiation has varied widely. Doses of gamma radiation have varied from a low of about 0.1 Gy (10 RAD) to a maximum of 3.0 Gy (300 RAD). The lowest dose at which significant numbers of dead cells can be found postirradiation in the sensitive period of the fetal cerebral cortex is approximately 0.25 Gy (25 RAD). Doses above 2.0 Gy (200 RAD) cause significant fetal death or early postnatal death.

Sensory tests. Brightness discrimination between 80 lux and 10 lux is not impaired in 21/2-month-old rats (M.P.I. strain) irradiated

TABLE 3

POSTWEANING MOTOR TESTS--INCLINED PLANE*

Gy gd Outcome

0.25 6 NS 11 NS 16 NS 21 NS

0.5 6 NS 11 NS 16 delayed 21 NS

1.0 6 increased 11 NS 16 delayed 21 delayed

*Data from Werboff et al. (72). NS = not significant.

with 2.0 Gy on gd 18 (68). However, Long Evans rats receiving radiation localized to the hippocampus on pnd 2, 3 (1.5 Gy), 5, 7, 9, 11, 13, and 15 (2.0 Gy) make more errors in discriminating 12 candles (cd)/m 2 from 0.8 cd/m 2. The same rats do not have more difficulty than control rats in distinguishing 12 cd/m 2 from 0.07 cd/m 2 (13). These effects of hippocampal irradiation have been interpreted as either sensory impairment or impaired attention.

Tactile sensitivity in the rats with hippocampal exposure, using a rough-smooth discrimination, is not impaired. When the difference in surfaces is less marked, rats with a deficit of hippocampal

TABLE 4

OPEN FIELD ACTIVITY*

Gy gd Outcome

1.0

2.0

3.0

6 NS 11 decr 16 NS 21 incr

6 decr 11 decr 15 decr 16 incr 21 incr

6 incr 11 decr 15 decr 16 inc r 21 incr

22 decr

15 incr 17 incr 18 incr 22 deer

*Data from Fowler and Echols Werboff et al. (72).


(10); Jensh et al. (25,26); Norton (44);


TABLE 5

EFFECTS OF IRRADIATION ON COGNITIVE PERFORMANCE IN RATS

Behavioral Test

Radiation Exposure Difference Dose Gestational Age at From (Gy) Day Testing Controls Reference

Shuttle box avoidance Shuttle box avoidance

Shuttle box avoidance



T-maze T-maze T-maze T-maze

Hebb-Williams maze problems (food reward)




Operant Conditioning Brightness disc. Brightness disc. Brightness disc. Fixed ratio DRH/DRL DRH/DRL

2.0 18 95 days more rapid Tamaki and Inouye (69) 2.0 16 5 months more rapid Furchtgott et al. (12)

2.0 16 14 months more rapid Furchtgott et al. (12) fractionated to pnd 2-15 60-180 days more rapid Wallace et al• (71)

hippocampus 0.6 10 60 days more rapid Jensh et al. (26)

0.6 18 60 days more rapid Jensh et al. (26)

0.6 10 60 days NS Jensh et al. (26) 0.6 18 60 days NS Jensh et al. (26) 2.0 16 5 months incr. reversal errors Furchtgott et al. (12) 1.0 16 14 months incr. reversal errors Furchtgott et al. (12)

1.5 14 80-120 days more errors Fowler et al. (9)

2.0 18 35 days more errors Kiyono et al. (32)

1.0 15 45 days more errors Furchtgon et al. (10)

1.0 18 ? more errors Seo et al. (66)

3.0 14 6 months deficit Graham et al. (15) 1.5 14 6 months deficit Graham et al. (15) 1.5 14 80-120 days deficit Fowler et al. (9) 0.9 13 3-4 months NS Bornhausen (2) 0.6 13 3-4 months NS Bornhausen (2) 0.6 16 3-4 months deficit Bornhausen (2)

NS = not significant•

30-

o l - z O u

2O

bJ

.9. r, lO

DAY 11

D COflTEX PHD 28

--~ BODY WEIGHT pND6

,N

DAY 13

GESTATIONAL DAY OF IRRADIATION

X

N DAY 15 DAY 17

HG. 2. Relationship of gestational day of exposure (1.0 Gy) to thickness of the frontal cerebral cortex on postnatal day 28 and to body weight of rats on postnatal day 6.

granule cells have more difficulty in learning a rough-rough discrimination (13).

M o t o r t e s t s . Various tests of motor function have been used to evaluate the postnatal response to prenatal exposure. Both early (preweaning) and postweaning periods have been examined. Motor tests evaluate various sensorimotor areas including both spinal and brain pathways. One unique consequence of gestational irradiation is the "hopping ra t . " Prenatal irradiation with 1.25 to 1.5 Gy on days 13-15 results in rats which carry out mirror acts with both fore and hind paws, including walking by hopping instead of alternating gait. This has been the subject of some analysis of the pathways involved (7, 21, 39).

Preweaning tests of motor function include surface righting, air righting, visual placing, negative geotaxis and forelimb reflex suspension (25, 26, 31, 44). Doses of 0.1 to 1.25 Gy have been examined on gestational days from 10 to 18 (Table 2). The gestational period when these behaviors are most seriously affected is from exposure on days 15 to 18. This correlates well with the period showing maximum reduction in thickness of the central cortex. The tests most likely to show changes are reflex suspension and air righting. As expected, the degree of effect is related to dose. Doses below 0.2 Gy have not altered any of these behaviors.

Postweaning motor tests include forelimb suspension, swimming performance, inclined plane and locomotor gait (25, 26, 39, 46, 72). A range of doses (0.1 to 1.25 Gy) and gestational days (6 to 21) have been used. Climbing the inclined plane is delayed by


TABLE 6

BODY WEIGHT OF RATS AT DIFFERENT AGES FOLLOWING GESTATIONAL EXPOSURE TO IONIZING RADIATION

Radiation Exposure Age at Significant Dose (Gy) Day Measurement Decrease Reference

0.1 10 60days NS 0.2 18 60days Sig 0.6 10 3 days NS 0.6 18 3 days Sig 0.25 15 3 days NS 0.25 15 4 months NS 0.5 15 3 days Sig 0.5 15 4months NS 0.75 15 3 days Sig 0.75 15 4months NS 1.25 15 3 days Sig 1.25 15 4months Sig 1.0 11 6 days NS

1.0 13 6 days Sig

1.0 15 6 days Sig

1.0 17 6 days Sig

Jensh et al. (25) Jensh et al. (25) Jensh et al. (26) Jensh et al. (26) Norton (44) Norton (44) Norton (44) Norton (44) Norton (44) Norton (44) Norton (44) Norton (44) Norton, unpublished

data Norton, unpublished



data


irradiation with 0.5 Gy and above when irradiation is performed on gestational day 16 (Table 3). Gait is altered by exposure to 1.25 Gy on gestational days 14 or 15. Since not all days and doses are included in each report, closer timing of each test is not available. Forelimb suspension and swimming performance are not altered by doses up to 0.6 Gy on gestational day 18.

Activity tests. Measures of activity include both short-term exploration by the rat of a novel environment (open field, tilting cage, continuous corridor or home cage emergence) or circadian activity (activity wheel or residential maze) (10, 25, 26, 31, 47, 72). Doses of ionizing radiation have varied from 0.1 to 3.0 Gy. Timing of irradiation has varied from gestational day 6 to 2l and at parturition. Age at testing varied from 21 days postnatally to 5 months. Open field tests have shown both increased and decreased activity (Table 4). Doses of 0.5 Gy and above alter activity in this test. Short-term exploration of a novel environment, using continuous corridor apparatus or home cage emergence, is suppressed by similar doses administered on gd 15 or 17. Higher doses of radiation cause nocturnal hyperactivity in circadian tests. The hyperactivity may become more evident with age up to 5 months.

Cognitive tests. Tests of learning have used three types of conditioning: shuttle box avoidance, maze problems, and operant conditioning. A summary of the findings is in Table 5. In addition to the generalizations regarding dose and time of irradiation, two pattems emerge. Firstly, the irradiated rats learn simple condi- tioned avoidance more rapidly than controls. Several investigators have associated this performance with the hyperactivity measured in other conditions, noted above. Secondly, performance of irradiated rats is poorer than controls in complex problems and the difference from controls is more marked with difficult learning problems than with simpler tests.

Other measurements. Several tests which have been used are difficult to categorize. For example, focal neonatal irradiation of the hippocampus has been employed to alter several complex nonconditioned behaviors including dominance behavior, defense of home cage, competition for food and competition for a

receptive female. As adults, the irradiated male rats show less dominance or aggression in all tests except in competition for the receptive female where they spend more time in mounting than control males (71).

One measurement reported very commonly is body weight. Diminished body weight associated with decreased size of various brain structures has been repeatedly documented (Table 6). Ges- tational time of irradiation is significantly related to body weight reduction (Fig. 2) as well as to dose of radiation. With high doses of radiation, body weight is permanently reduced. With doses below 0.75 Gy some investigators report that rats catch up as adults (44).

Underlying neurobiological events. At exposures of cells of 0.3 to 0.4 Gy of gamma radiation, death of dividing cell is possible. However, postmitotic migrating neuroblasts of the form- ing cortical plate are more sensitive to radiation than the dividing cells of the germinal matrix (20,34). Radiation also acts by altering differentiation of cells and by changing proliferation rates (20). The net effect in the brain is the formation of structures which have fewer neurons and in which orientation of cortical layer V neurons is disturbed (34). In addition, dendritic branching may be reduced (63).

One unique characteristic of ionizing radiation for comparing behavioral and morphological changes is the fact that there are no maternal distributional or metabolic factors in exposing fetuses to radiation. Furthermore, there are no known differences in sensitivity of developing neurons of different species to ionizing radiation. Although rat data have been emphasized here, similar effects on brains of other species have been reported at similar doses. An effort has been made to give quantitative data in the sections on behavioral tests and these will not be repeated here.

It is possible that there is an effect of radiation on behavior below the currently detectable threshold doses of behavioral changes (23). The lowest dose of radiation on any gestational day to achieve a significant change is 0.2 Gy. However, this dose does not necessarily reflect the lowest dose which could alter behavior in a smaller-than-significant percentage of the animals. Further- more, there are very few data on the possible interaction of very low dosage damage from ionizing radiation with other insults. For example, it has been shown that rats exposed to ionizing radiation on gestational day 15 are more hyperactive when exposed to morphine as adults than nonirradiated rats (63).

Human Data

Hiroshima and Nagasaki. Few population-based studies of the effects of in utero exposure on the developing human embryo and fetus exist. Among these, however, the size, length of study, variability in dose, and postfertilization age at exposure make the experiences in Hiroshima and Nagasaki most important. These populations were exposed at a variety of developmental phases and, therefore, presumably a variety of sensitivities. Dose estimates. Recently published analyses of the effects have used the estimated absorbed dose to the mother's uterus based on the DS86 dosimetry (60). Absorbed doses to the fetus itself are not available, and may not be for some time. However, phantom studies have shown that the correspondence between the dose in the uterus and in fetal tissues is high in the second half of pregnancy, and that uterus dose may overestimate the energy absorbed by the developing tissues in the first half when more fluid surrounds the embryo or fetus (15,30). Developmental ages. Developmental age is the most important single factor in determining the nature of the insult to the embryo or fetus resulting from exposure to ionizing radiation. Accord- ingly, since different functions in the human brain are localized into different stages of gestation and over different periods of time, gestational ages (here taken to be synonymous with developmental


TABLE 7

SEVERELY MENTALLY RETARDED CASES WITH PROBABLE NONRADIATION RELATED ETIOLOGIES*

Dose Category Gy

8-15-Week Period 16-25-Week Period

Retarded Rate % Retarded Rate %

T65DR Sample

Control 2 0.8 2 ( - 1 b) 0.6 (0.3) 0.01-0.09 3 4.3 1 ( - 1 c) 1.2 (0.0) 0.10-0.49 4 ( - 2 a) 8.0 (4.2) 0 0.0 0.50--0.99 4 30.8 2 13.3

1.00-1.99 5 } 75.0 0 } 20.0 2.00+ 1 1 Total 19 ( - 2 ~) 4.8 (4.3) 6 ( - 2 bx) 1.3 (0.9)

DS86 Sample

Control 2 0.8 2 ( - 1 b) 0.6 (0.3) 0.01-0.09 2 4.5 1 ( - 1 c) 1.8 (0.0) 0.10-0.49 2 ( - 1 ~) 3.4 (1.8) 0 0.0 0.50-0.99 4 ( - 1 ~) 25.0 (20.0) 0 0.0

1.00_1.99 8 } 750 3 } 375 2.00+ 1 0 Total 19 ( -2" ) 4.9 (4.4) 6 ( - 2 ux) 1.4 (0.09)

aDown's syndrome. bRetarded sibling. CJapanese B encephalitis at age 4. *Taken from (50) and (51).

age) have been grouped to reflect these known phases in normal development. Four categories measured from the presumed mo- ment of fertilization have been used: 0-7 , 8-15, 16-25, and 26 or more weeks. In the first period, the precursors of the neurons and neuroglia have emerged and are mitotically active. In the second, a rapid increase in the number of neurons occurs; they migrate to their ultimate developmental sites and lose their capacity to divide, becoming perennial cells. In the third, differentiation in situ accelerates, synaptogenesis that began about the eighth week increases, and the definitive cytoarchitecture of the brain un- folds. The fourth period is one of continued architectural and cellular differentiation and synaptogenesis of the cerebrum; with at the same time, accelerated growth and development of the cerebellum. Severe mental retardation. Thirty of the 1,544 individuals included in the sample on whom DS86 doses can be computed (doses are not available for 55 survivors in this sample) terminated in a child with severe mental retardation and 18 of these, or 60%, had disproportionately small heads, that is, a head with a circumference more than two standard deviations below the mean observed among the 1599 births in the entire sample (5, 36-38, 67).

When the prenatally exposed survivors are distributed over the four age groupings previously described, and the frequency of mentally retarded individuals is examined in the light of their doses and the age at which they were irradiated, the following emerges (see Table 7).

First, the highest risk of severe mental retardation occurred when exposure was at 8 to 15 weeks after fertilization (50). There is no apparent increased risk prior to the eighth week nor after the 25th. The number of women who were pregnant and whose infants did not survive, however, is unknown, and thus it is conceivable that embryos exposed in the first eight weeks following fertilization with brain damage were aborted.

Second, within this critical period, damage expressed as the frequency of subsequent severe mental retardation appears to be

TABLE 8

LINEAR-RESPONSE RELATIONSHIP AND THRESHOLD OF SEVERE MENTAL RETARDATION*

Gestational Age (weeks)

T65DR Dose DS86 Dose

b S b b S b (Gy) (Gy) T (Gy) (Gy)

All

8-15

16-25

All

8-15

16-25

Grouped Data

0.300 b 0.069 0.18 0.498 b 0.100 (L=0, U=0.51)

0.462 b 0.091 0 0.613 b 0.101 (L=0, U=0.20)

0.206 sug~ 0.113 0.23 0.497 a 0.23 l (L=0, U =0.68)

Individual Data

0.330 b 0.070 0.23 a 0.419 b 0.062 (L =0.08, U=0.33)

0.369 b 0.057 0 0.996 b 0.146 (L=0, U =0.24)

0.464 sug~ 0.237 0.54 a 0.532 a 0.234 (L = 0.04, U=0.71)

0.47 a (L=0.18, U=0.61)

0.20 (L=0, U=0.55)

0.64 ~ (L=0.21, U-->0.64)

0.43 a (L=0.23, U=0.56)

0.40 (L=0, U =0.56)

0.70 ~ (L =0.21, U=0.99)

*Taken from (50) and (51). aSignificant difference, p<0.05. bSignificant difference, p--<0.01. Su~gSignificant difference, p--<O. 10.


TABLE 9

LINEAR-RESPONSE RELATIONSHIP AND THRESHOLD OF SEVERE MENTAL RETARDATION WITHOUT PROBABLE NONRADIATION RETARDED CASES*

Gestational Age (weeks)

T65DR Dose DS86 Dose

b S b b S b (Gy) (Gy) T (Gy) (Gy)

All

8-15

16-25

All

8-15

16-25

Grouped Data

0.312 b 0.073 0.21 a 0.521 b 0.106 (L=0.11, U=0.58)

0.566 b 0.108 0.18 0.743 b 0.121 (L=0, U=0.59)

0.213 s"gg 0.113 0.23 0.497 ~ 0.231 (L=0, U=0.67)

Individual Data

0.335 b 0.072 0.25 a 0.445 ~' 0.061 (L=0.11, U=0.53)

0.409 b 0.057 0.17 1.095 b 0.158 (L=0, U=0.31)

0.463 a 0.231 0.53 a 0.539 a 0.233 (L = 0.04, U=0.71)

0.51 a (L =0.21, U=0.63)

0.39 a (L =0.12, U =0.60)

0.64 a (L=0.210 U-->0.64)

0.50 ~ (L=0.30, U=0.61)

0.46 a (L=0.23, U=0.62)

0.70 ~ (L=0.21, U=0.99)

*Taken from (50) and (51). aSignificant difference, p--<0.05. bSignificant difference, p--<0.01. suggSignificant difference, p--0.10.

linearly related to the dose received by the fetal tissues, and there is little evidence of a threshold. About forty-five percent of fetuses exposed to 1.0 Gy in this period will be mentally retarded; this is a risk more than fifty times greater than that in the unexposed comparison group. There is still disagreement, however, concerning the effects of exposures at low levels (for example, 0.01 Gy or less) on intelligence. Third, a period of lesser vulnerability appears to exist in the interval 16-25 weeks after fertilization. However, here a threshold does seem to exist; no increase in cases is seen at doses of less than 0.50 Gy.

Table 8 gives the intercepts and slopes obtained when a linear model, without threshold, is fitted to the data in Table 7 with and without the inclusion of the 0-0.01 Gy group (the 'controls ') , and when the 'controls ' are pooled over all prenatal ages. Within the most vulnerable age group (irradiated at the 8th through the 15th week following fertilization), the rate of increase in incidence of severe mental retardation with dose is about 0.44 G y - 1 with an estimated standard error of about 0.09 G y - 1 (see Table 9).

Three of the severely mentally retarded children, all in Hi- roshima, are known to have, or have had, Down's syndrome, a fourth, also in Hiroshima, Japanese encephalitis in infancy, and a fifth, in Hiroshima had a retarded sibling. It is conceivable that, in these instances, the mental retardation was merely a part of the former syndrome or secondary to the infection, but in either event not radiation-related. Virtually the same regression coefficients were obtained when these five children were excluded from the analysis, and thus the main conclusions are not dependent upon the inclusion or exclusion of these individuals. Small head size. As previously stated, the small head sizes were two or more standard deviations below the mean head size of all of the individuals in the study sample. About 10% of these individuals with small head sizes were also mentally retarded. Among the mentally retarded, as earlier stated 18 out of 30 (60%) had small head sizes (73,74). It should be noted that head circumference was

not standardized against body size, and since mental retardation is often seen in individuals whose head circumferences are disproportionately small for their body sizes, the value just cited may be spuriously low. Intelligence tests. Intelligence has been described as the ability to manage oneself and one 's affairs prudently; to combine the elements of experience; to reason, compare, comprehend, use numerical concepts and combine objects into meaningful wholes; to have the faculty to organize subject-matter experience into new patterns; or to have the aggregate capacity to act purposefully, think rationally and deal effectively with one 's environment. Given such differences in definition, it is natural that the bases of measurement should vary. Intelligence tests differ one from another in the importance given to verbal ability, psychomotor reactions, social comprehension, and so on. Thus, the score attained by an individual will depend to some degree upon the type of test used; however, generally, individuals scoring high on one type of test tend to obtain high scores on other tests. Most intelligence tests are so structured that the distribution of test results follows an approximately normal curve, with some 95% percent of the population falling within two standard deviations of the mean. Individuals whose scores lie, consistently, two standard deviations or more below the mean would commonly be described as retarded. In the Japanese experience, the highest IQ achieved by any of the severely mentally retarded children on the Koga test was 64.

Schull, Otake and Yoshimaru (65) describe an analysis of Koga intelligence test scores (33,70) obtained in 1955 on survivors exposed prenatally. These results are summarized in Table 10, and can be briefly stated as follows: 1) there is no evidence of a radiation-related effect on intelligence among those individuals exposed within 0 -7 weeks after fertilization or in the 26th or subsequent weeks; 2) for individuals exposed at 8-15 weeks after fertilization, and to a lesser extent those exposed at 16-25 weeks,


TABLE 10

LINEAR REGRESSION COEFFICIENTS OF IQ SCORE ON T65DR AND DS86 DOSES*

Regression Coefficient

Gestational Standard Error Ages (weeks) b S b

T65DR SAMPLE

0-7 269 (0) - . 057 ( - . 057 ) 0.061 (0.061) 8-15 350 ( - 4 ) - . 2 1 2 b ( - . 1 6 8 b) 0.039 (0.042)

16-25 480 ( - 3) - . 207 b ( - . 133 a) 0.042 (0.053) 26+ 574 ( - 1) - .033 ( - .035) 0.055 (0.054) Total 1673 ( - 8 ) - .155 b ( - .111 b) 0.023 (0.025)

Heterogeneity ×2(3)= 11.07, p=0.01 [Heterogeneity X2(3) = 4.75, p = 0.01]

DS86 SAMPLE

0-7 179 (0) - . 027 ( - . 027 ) 8-15 256 ( - 4 ) - . 290 b ( - . 2 5 0 b)

16-25 352 ( - 3 ) - . 2 0 4 b ( .098 s~gg) 26+ 415 ( - 1) - . 0 4 2 ( - .044) Total 1202 ( - 8) - . 158 b ( - . 102 b)

0.053 (0.053) 0.042 (0.051) 0.044 (0.057) 0.050 (0.050) 0.024 (0.026)

Heterogeneity X2(3) = 22.30, p<0.01 [Heterogeneity X2(3) = 11.82, p<0.01]

The coefficients are expressed as change in IQ points per 0.01 Gy. *Taken from (50) and (51). "Significant difference, p-<0.05. bSignificant difference, p<0.01. suggSignificant difference, p-<0.10.

the mean tests scores, but not the variation in scores about the mean, are significantly heterogeneous among exposure categories; 3) the distribution of test scores suggests a progressive shift downwards in individual scores with increasing exposure; and 4) within the group most sensitive to the occurrence of clinically recognizable severe mental retardation, individuals exposed 8 through 15 weeks after fertilization, the diminution in intelligence score under the linear model is 21-33 points at 1 gray, based on the new dosimetry and the specific set o f observations used. School performance. As a part o f the assessment o f the effects on the developing embryonic and fetal brain of exposure to ionizing radiation, the school performances of prenatally exposed survivors in Hiroshima and a suitable comparison group have been studied (51). At the t ime this information was collected these children were 10 to 11 years old, and most had recently completed the fourth grade. The records themselves include information on school attendance, performance in various subjects, their behavior, and physical status.

In the first four years of e lementary schooling the Japanese student is exposed to training in some seven different subjects ranging from language through science to physical education. Their achievements or performance in these subjects can be summarized as follows: Damage to the 8-15 week fetal brain appears to be linearly related to the absorbed dose, as judged by the relationship of average school performance score to dose (see Table 11). Damage to the fetus exposed at 16-25 weeks after fertilization is similar to that seen in the 8 -15-week group. This trend is stronger, however , in the earliest years of schooling. In the groups exposed within 0 -7 weeks fol lowing fertilization, or 26 or more weeks after fertilization, there is no evidence of a radiation- related effect on scholastic performance. As will be noted, these

TABLE 11

THE LINEAR RELATIONSHIP OF SCHOOL PERFORMANCE SCORE TO DS86 DOSE*

Regression Estimates Gestational Number of Ages (weeks) Cases b S b

First Grade

0-7 106 0.0023 0.0032 8-15 225 - 0.0115 b 0.0022

16-25 267 - 0.0097 b 0.0024 26+ 323 0.0023 0.0036 All 921 - 0.0070 b 0.0014

Heterogeneity X 2 = 20.48, p<0.01

Second Grade

0-7 107 0.0036 0.0034 8-15 224 -0.0127 b 0.0022

1 6-25 268 - 0.0096 b 0.0024 26+ 324 0.0001 0.0036 All 923 - 0.0076 b 0.0014

Heterogeneity ×2 = 21.34, p<0.01

Third Grade

0-7 107 0.0012 0.0038 8-15 221 -0.0117 b 0.0025

16-25 265 -0.0101 b 0.0025 26+ 319 -0.0006 00037 All 912 -0.0074 b 0.0015

Heterogeneity X 2= 12.62, p<0.01

Fourth Grade

0-7 5& 0.0172 a 0.0084 8-15 204 - 0.0095 a 0.0042

16-25 260 -0.0109 b 0.0026 26+ 321 -0.0035 0.0032 All 841 - 0.00898 0.0018

Heterogeneity ×2= 4.42, p=0 .22

*Taken from (50) and (51). aSignificant difference, p -0 .05 . bSignificant difference, p--<0.01. suggSignificant difference, p-<0.10.

results parallel those previously found in prenatally exposed survivors with respect to achievement in standard intelligence tests in childhood. Convulsions. Seizures are a frequent sequela of impaired brain development , and therefore, could be expected to affect more children with radiation-related brain damage than children without. Dunn and her colleagues (8) have described the incidence, and type, o f seizures among survivors prenatally exposed to the atomic bombing o f Hiroshima and Nagasaki, and their association with specific stages of development at the time of irradiation. Histories of seizures were obtained at biennial routine clinical examinations starting at the age of two years. These clinical records were used to classify seizures as febrile or unprovoked (without precipitating cause).

Seizures were not recorded among individuals exposed 0 -7 weeks after fertilization at doses higher than 0.10 Gy. After irradiation at 8-15 weeks after fertilization, the incidence of


seizures was highest among individuals with doses exceeding 0.10 Gy and was linearly related to the level of fetal exposure. This obtains for all seizures without regard to the presence of fever or precipitating causes, and for unprovoked seizures. When the 22 cases of severe mental retardation were excluded, the increase in seizures was only suggestively significant and then only for unprovoked seizures. There was no increase in recorded seizures due to exposure at later stages of development.

The risk ratios for unprovoked seizures, following exposure within the 8th through the 15th week after fertilization, are 4.4 (90% confidence interval: 0.5-40.9) after 0.10-0.49 Gy and 24.9 (4.1-191.6) after 0.50 or more Gy when the mentally retarded are included, and 4.4 (0.5--40.9) and 14.5 (0.4-199.6), respectively, when they are excluded. Uncertainties. Many uncertainties are associated with these estimates of risk. They include the limited nature of the data, especially on mental retardation and convulsions, the appropriate- ness of the comparison group, errors in the estimation of the tissue absorbed doses and the prenatal ages at exposure, and other confounding factors in the postbomb period, including nutrition and disease, which could play a role. At the present time, no satisfactory assessment of their contribution can be made at this late date (65). Given the present uncertainties, since most of these extraneous sources of variation would have a greater impact at high than low doses, and thus produce a concave upwards dose-response function, the prudent course would be to assume that the dose-response relationship is not materially altered other than additively by these potential confounders. This would have the effect of overestimating the risk at low doses where greatest regulatory concern exists. Increasing the exposure safety factor would result in the same effect without the implication that, at the present time, we can state with certainty that effects occur at low dose levels.

Three issues do warrant further concern here; these include the shape of the dose-response function, the existence of a threshold in the dose-response, and the effects of dose fractionation. Within the period of maximum vulnerability, virtually without exception, the data previously presented can be satisfactorily approximated by more than one dose-response function, generally a linear or a linear-quadratic model. Given that neuronal death, mismanaged migration, and faulty synaptogenesis could all play a role in the occurrence of mental retardation or cortical dysfunction more generally, and that each could have its own different dose- response relationship, there is little or no prior basis for presuming that one or the other of these models better describes the fundamental biological events involved. The " t rue" model, therefore, remains a matter of conjecture, and it seems unlikely that epidemiological studies alone will never be able to determine what the " t rue" one may be, although such studies are valuable in the determination of whether a relationship does or does not exist. Perforce the model used in the estimation of risk must rest on a series of considerations, not all of which are biological. Most importantly, the risk estimated should be a prudent one, minimiz- ing risk wherever such exists. This argues for the use of a linear dose-response since presumably at lower doses, where the evidence of an effect is weakest, risk is apt to be overestimated, although use of such a relationship rather than increasing the safety factor is still a subject of debate.

Although a linear or a linear-quadratic dose-response relationship describes the observed frequency of severe mental retardation in the 8th through the 15th week adequately, there could be a threshold with DS86 dosimetry. As Otake, Yoshimaru and Schull (50) have shown, the estimation of the value of this presumed threshold is not straightforward. When all of the cases of mental retardation are included in the analysis, the lower bound of the estimate of the threshold includes zero, that is to say, a threshold

cannot be shown to exist by statistical means. However, if the two cases of Down's Syndrome in the 8-15-week period are excluded, the 95% lower bound of the threshold appears to range from 0.12 to 0.23 Gy. It should be noted, however, that the imposition on the data of a linear model with a threshold gives rise to a rate of increase with dose that predicts virtually every fetus exposed to 1.0 Gy or more will be retarded. This is at variance with the actual observations, but this would not necessarily be true if a curvilinear model with a threshold were fitted. The DS86 dosimetry suggests a threshold in the 16-25-week period of 0.21--0.70 Gy. There is still disagreement, however, whether a threshold does truly exist.

Little is known about the effects of chronic or fractionated exposures to ionizing radiation on the developing human embryo and fetus. Given the complexity of brain development and the differing durations of specific developmental phenomena, it is reasonable, however, to assume that dose fractionation will have some effect. The hippocampus, for example, and the cerebellum continue to have limited neuronal multiplication, and migration does occur in both organs. Changes continue in the hippocampus and cerebellum into the first and second years of life. Continuing events such as these may show dose-rate effects differing from those associated with the multiplication of the cells in the periventricular areas of the cerebrum, or the migration of neurons to the cerebral cortex.

Most of the information available on the effects of dose rate involves the experimental exposure of rodents, and must be interpreted with due regard to the differences between species in developmental timing and rates relative to birth. Brizzee and Brannon [(6), see also (24)] have examined cell recovery in the fetal brain of rats. The incidence and severity of tissue alterations generally varied directly with dose, and were clearly greater in single dose than in split dose groups with the same total exposure.

In summary, reevaluation of the Japanese atomic bomb survi- vor data has provided a new perspective on the periods of sensitivity of the developing brain to radiation-related damage, and the possible nature of the dose-response relationship. These findings have been described in some detail previously; briefly, and as they specifically concern risk estimation, they are as follows:

The period of maximum vulnerability to radiation appears to be the time from approximately the beginning of the eighth through the fifteenth week after fertilization, that is, within the interval when the greatest proliferation of neurons and their migration to the cerebral cortex occur. A period of lesser vulnerability occurs in the succeeding period from the sixteenth through the twenty-fifth week after fertilization. The latter accounts for about a fourth of the apparently radiation-related cases of severe mental retardation. The least vulnerable period is the initial seven weeks after fertilization during which no radiation-related cases of severe mental retardation have been seen.

The data on intelligence tests and school performance suggest the same two gestational periods of vulnerability to radiation, the first period showing the greatest sensitivity. More importantly, these data suggest a continuum of effects on the developing brain of exposure to ionizing radiation; indeed, the downwards shift seen in the distribution of IQ scores with increasing exposure predicts reasonably well the increase in severe mental retardation actually observed. This suggests, in turn, that the impact of exposure to ionizing radiation will be related to where in the normal continuum of cortical function an individual would have resided if unexposed.

Exposure in utero--other human data. Numerous studies of the possible role of ionizing radiation in the origin of central nervous system abnormalities (13,40) have been published, but few, aside from the Japanese experience, provide a reliable basis for risk estimation. Generally, there is little information on the exposures,


or on the ages after fertilization at the time of exposure. However, Granroth (16), in Finland, has examined the association of diagnostic x-ray examinations with the occurrence of defects of the central nervous system. The data, drawn from the Finnish Registry of Congenital Malformations, reveal a significant increase in central nervous system abnormalities, primarily anencephaly, hydrocephaly and microcephaly, among newborn infants exposed in utero, when contrasted with time-area-matched control subjects. No estimate is given of the fetal absorbed dose. Moreover, as the author notes, the majority of these infants were exposed because of the clinical suspicion of either maternal pelvic or fetal anomaly and, therefore, the exposures were unlikely to have occurred at a time when abnormalities, such as anencephaly, are induced (41). Accordingly, it seems unlikely that the results reflect a teratogenic effect of radiation.

Neumeister (42) has described the findings on 19 children exposed in utero to doses between 0.015 and 0.1 Gy. No instances of severe mental retardation are recorded, but developmental age at the time of exposure was not taken into consideration. Meyer and colleagues (35) failed to find evidence of an increased frequency of severe mental retardation among 1455 women who were exposed to small doses of radiation in utero as a result of diagnostic pelvic examinations of their mothers. It seems uncertain, however, whether their case-finding mechanism would have identified women who were mentally retarded, and, of course, the increased probability of premature death among such individuals would lead to underrepresentation of the retarded later in life. In addition, exposure must commonly have occurred late in pregnancy, after the most vulnerable period. Other studies, such as those of Oppenheim et al. (48) and Nokkentved (43), are similarly inappropriate for the estimation of radiation effects.

At present, there is no evidence of radiation-related cerebellar damage without concomitant damage to the cerebrum among the survivors exposed prenatally. It may be difficult to identify such damage for several reasons. First, Purkinje cells, the only efferent neurons in the cerebellum, are proliferating and migrating in the same developmental period as the neuronal cells that populate the cerebral cortex, and thus damage to precursors or differentiated Purkinje cells would occur at the same time and may be insepa- rable from damage to those cells that give rise to the cerebral cortex (75). Second, the granular neurons, the most numerous nerve cells in the cerebellum, retain their proliferative abilities after birth and could, in theory, repopulate areas of the developing cerebellum damaged by radiation. To the extent that this occurs, granular cell damage might be mitigated. Estimates of the risk of damage to the cerebellum following prenatal exposure, based on fixed or progressive neurologic deficit, are presently not possible.

COMPARISON OF HUMAN AND ANIMAL DATA

The goal in selecting a test for prediction is the demonstration that a selected group of agents with known effects in humans cause identifiable changes when the test is performed on animals. Conversely, agents which are known not to affect humans should be ineffective in the test in animals. This goal should not be equated to the "sensitivity" of the test. That is, the lowest dose which can be detected by a test has no theoretical importance in the determination of the accuracy with which results of a test in animals can be applied to prediction of effects in humans.

The known variations in behavioral effects of differing doses and time of exposure of the fetus to radiation offer a salutary warning about the complexity of selecting tests which are likely to be good predictors, even when the outcome in humans is estab- lished. For example, mental retardation is a known outcome of short-term gestational exposure to ionizing radiation in the human. Although various tests can detect altered rat behavior in response

to the doses which cause human mental retardation, it is not known which rat test is most predictive for mental retardation and not some other component of irradiation damage. It would be unwise to assume that reduction in body weight, for example, is the best general test for mental retardation, even if the correlation in animals exposed to ionizing radiation were excellent.

One way to evaluate the predictive value of a behavioral test is to estimate the accuracy with which the test predicts a structural abnormality or other nervous system changes caused by the agent in the same animal. This approach requires a quantifiable change in a behavior for comparison with a quantifiable effect on the nervous system. Simple correlations can be calculated which make use of the variability inherent in both measurements. Other predictive estimates can be obtained from multiple regression and covariance methods. The use of multiple regression in this context has recently been discussed (45). Once specific neurological damage can be equated to effects on specific behavioral tests in the same animals, the prediction of damage in one species from data in another species acquires a rational basis. From the available data, therefore, it can be concluded that: 1) following exposure of rat fetuses during the period of formation of the cortical plate, good correlations are found in postnatal thinning of the cerebral cortex, postnatal deficiencies in complex behaviors (including learning in complex problems) and body weight in the neonatal period; 2) for the same gestational period of exposure, correlation with simple motor behaviors is only fair; 3) marked morphological and functional alterations from exposure during gestational periods, other than during cortical plate formation, may be difficult to detect with commonly used behavioral tests; 4) when significant effects are reported for behavioral tests of all types, there is fair correlation of alteration of behavior with dose of radiation.

At this stage there is good evidence with which to select tests which detect some kinds of changes in rat behavior as a consequence of brain damage from fetal irradiation, but the interpreta- tion of their meaning and therefore, their predictive validity for humans, is very uncertain, particularly at low levels of exposure. Considerable additional work needs to be done, particularly in the low dosage area. Animal data do provide some insights into the mechanisms which may be operative at low doses, but at present, substantial low dose data are lacking.

RELEVANCE OF PROPOSED EPA DEVELOPMENTAL NEUROTOXICITY BATTERY TO THE EFFECTS OF ION/ZING RADIATION

For this discussion, it is assumed that some repairable damage to the fetal CNS is part of normal development and doses of radiation which cause effects in this range are not distinguishable from the risks encountered by any fetus. The assumption is that death and abnormal migration or differentiation of a few neurons is within the compass of a normal brain. Nevertheless, the data from animal experiments clearly document dose-response effects. Some measurements, particularly acute cell death from radiation exposure, have yet to demonstrate a threshold. Therefore, the limits of tolerance of the nervous system must be assumed to be low in the absence of precise data for specific functions. Damage to the animal brain can be seen shortly after irradiation as low as 0.25 Gy. Behavioral changes do appear at higher dosages; certainly at about 1.0 Gy. The animal and human studies thus far exhibit remarkable similarities.

Postnatal behavioral test batteries such as that proposed by the Environmental Protection Agency are an important step in refining our abilities to determine possible subtle long-term effects of prenatal irradiation on offspring. With the appropriate tests, behavioral changes may be seen at even lower dosage levels. It is understood that such behavioral test batteries are an addition to morphologic evaluations and that the use of specific behavioral


tests in a battery are dependent upon the types of questions being asked. Therefore, it is important not to restrict such a battery to a specific set of tests to the exclusion of others. In the Nagasaki and Hiroshima studies, for example, recording IQ's in 1955 of radiation exposed children undoubtedly seemed superfluous at that time, as was the acquisition of the childrens' school records; these data recently being shown to be important.

CONCLUSION

The following conclusions can be drawn from the available evidence for the human and animal responses to ionizing radiation: 1) dose-response data show marked parallels in all species,

including humans, when comparable developmental periods are compared; 2) at exposure levels of 1.0 Gy and above, serious consequences, particularly mental retardation, occur with high probability when the fetuses are exposed during the period of neuronal migration to the cortical plate of the developing telen- cephalon; 3) at exposure levels of 1.0 Gy and above, many behavioral tests in animals show significant alterations when the fetuses are exposed during cortical plate formation; 4) behavioral test batteries such as those proposed by the EPA will aid in determining long term radiation effects provided there is sufficient latitude concerning the inclusion of specific behavioral tests within the battery; 5) more studies are needed in the low dosage range if we are to fully understand those levels of ionizing radiation which are most likely to be of importance to the human condition.

REFERENCES

1. Altman, J.; Bayer, S. A. Prenatal development of the cerebellar system in the rat. J. Comp. Neurol. 179:23-28; 1978.

2. Bornhausen, M. Analysis of behavioral changes induced by prenatal irradiation. In: Kriegel, H.; Schmal, G. B.; Gerber, G. B.; Stieve, F. E., eds. Radiation risks to the developing nervous system. Stuttgart: Gustav Fischer; 1986:283-291.

3. Brent, R. L. Radiation teratogenesis. Teratology 21:281-298; 1980. 4. Brent, R. L. Effects of ionizing radiation on growth and development.

Contr. Epidem. Biostat. 1:147-183; 1979. 5. Blot, W. J.; Miller, R. W. Small head size following in utero

exposure to atomic radiation, Hiroshima and Nagasaki. Atomic Bomb Casualty Commission Technical Report 35-72; 1972.

6. Brizzee, K. R.; Brannon, R. B. Cell recovery in foetal brain after ionizing radiation. Int. J. Radiat. Biol. 21:375-388; 1972.

7. D'Amato, C. J.; Hicks, S. P. Development of the motor system: effects of radiation on developing corticospinal neurons and locomotor function. Exp. Neurol. 70:1-23; 1980.

8. Dunn, K.; Yoshimaru, H.; Otake, M.; Annegers, J. F.; Schull, W. J. Prenatal exposure to ionizing radiation and subsequent development of seizures. Radiation Effects Research Foundation Technical Report 5-88; 1988.

9. Fowler, H.; Hicks, S. P.; D'Amato, C. J. Effects of fetal irradiation on behavior in the albino rat. J. Comp. Physiol. Psychol. 55:309-314; 1962.

10. Furchtgott, E.; Echols, M. Activity and emotionality in pre- and neonatally X-irradiated rats. J. Comp. Physiol. Psychol. 51:541-545; 1958.

11. Furchtgott, E.; Echols, M.; Openshaw, J. W. Maze learning in pre- and neonatally X-irradiated rats. J. Comp. Physiol. Psychol. 51: 178-180; 1958.

12. Furchtgott, E.; Jones, J. R.; Tacker, R. S.; Deagle, J. Aversive conditioning in prenatally X-irradiated rats. Physiol. Behav. 5:571- 576; 1970.

13. Gazzara, R. A.; Altman, J. Early postnatal X-irradiation of the hippocampus and discrimination learning in adult rats. J. Comp. Physiol. Psychol. 95:484--495; 1981.

14. Goldstein, L.; Murphy, D. P. Etiology of the ill-health in children born after maternal pelvic irradiation. Part 2. Defective children born after post-conception pelvic irradiation. Am. J. Roentgenol. 22: 322-331; 1929.

15. Graham, T. N.; Marks, A.; Ershoff, B. H. Effects of prenatal X-irradiation on discrimination learning in the rat. Proc. Soc. Exp. Biol. Med. 100:78-81; 1959.

16. Granroth, G. Defects of the central nervous system in Finland. IV. Associations with diagnostic X-ray examinations. Am. J. Obstet. Gynecol. 133:191-194; 1979.

17. Hashizume, T.; Maruyama, T.; Nishizawa, K.; Nishimura, A. Dose estimation of human fetus exposed in utero to radiations from atomic bombs in Hiroshima and Nagasaki. J. Radiat. Res. 14:346-362; 1973.

18. Hicks, S. P.; Brown, B. L.; D'Amato, C. J. Regeneration and malformation in the nervous system, eye, and mesenchyme of the mammalian embryo after radiation injury. Am. J. Pathol. 33:459-482; 1957.

19. Hicks, S. P.; D'Amato, C. J.; Lowe, M. J. The development of the

mammalian nervous system. J. Comp. Neurol. 113:435-470; 1959. 20. Hicks, S. P.; D'Amato, C. J. Effects of ionizing radiation on

mammalian development. In: Woolum, D. H. M., ed. Advances in teratology, vol. 1. London: Losos Press; 1966:196-250.

21. Hicks, S. P.; D'Amato, C. J. Development of the motor system: hopping rats produced by prenatal irradiation. Exp. Neurol. 70:24-39; 1980.

22. Hicks, S. P.; D'Amato, C. J. Effects of radiation on development, especially of the nervous system. Am. J. Forensic Med. Pathol. 1:309-317; 1980.

23. Hoel, D. G. Radiation risk estimation models. Environ. Health Perspect. 75:105-107; 1987.

24. Jacobs, L. A.; Brizzee, K. R. Effects of total-body X-irradiation in single and fractionated doses on developing cerebral cortex in rat foetus. Nature 210:31-33; 1966.

25. Jensh, R. P.; Brent, R. L. Effects of 0.6 Gy prenatal X-irradiation on postnatal neurophysiologic development in the Wistar rat. Proc. Soc. Exp. Biol. Med. 181:611-619; 1986.

26. Jensh, R. P.; Brent, R. L. The effect of low-level prenatal X- irradiation on postnatal development in the Wistar rat. Proc. Exp. Biol. Med. 184:256-263; 1987.

27. Jensh, R. P.; Brent, R. L. The effects of prenatal X-irradiation on the appearance of reflexes and physiologic markers in the neonatal rat. Radiat. Res. 116:416-426; 1988.

28. Jensh, R. P.; Brent, R. L.; Vogel, W. H. Studies concerning the effects of low level prenatal X-irradiation on the postnatal growth and adult behavior in the Wistar rat. Int. J. Radiat. Biol. 50:1060-1081; 1986.

29. Jensh, R. P.; Brent, R. L.; Vogel, W. H. Studies on the effect of 0.4 Gy and 0.6 Gy prenatal X-irradiation on postnatal adult behavior in the Wistar rat. Teratology 35:53-61; 1987.

30. Kerr, G. D. Organ dose estimates for the Japanese atomic bomb survivors. Health Physics 37:487-508; 1979.

31. Kimler, B. F.; Norton, S. Behavioral changes and structural defects in rats irradiated in utero. Int. J. Radiat. Oncol. Biol. Physiol. 15: 1171-1177; 1988.

32. Kiyono, S.; Seo, M.; Shibagaki, M. Effects of environmental enrich- ment upon maze performance in rats with microcephaly induced by prenatal X-irradiation. Jpn. J. Physiol. 31:769-773; 1981.

33. Koga, Y. Two intelligence test methods viewed in relation to evaluated intelligence. Collection of Reports in Commemoration of Dr. Matsumoto. Stud. Psychol. Arts 923-988; 1937 (in Japanese).

34. Konermann, G. Brain development in mice after prenatal irradiation: modes of effect manifestation; dose-response-relationships and the RBE of neutrons. In: Kriegel, H.; Schmahl, W.; Gerber, G. B.; Stieve, F. E., eds. Radiation risks to the developing nervous system. Stuttgart: Gustav Fischer; 1986.

35. Meyer, M. B.; Tonascia, J. A.; Merz, T. Long-term of effects of prenatal X-ray on development and fertility of human females. In: Biological and environmental effects of low-level radiation. II. Vienna: IAEA; 1976:273-277.

36. Miller, R. W. Delayed effects occurring within the first decade after exposure of young individuals to the Hiroshima atomic bomb. Pediatrics 18:1-18; 1956.

260 S C H U L L , N O R T O N A N D JENSH

37. Miller, R. W.; Blot, W. J. Small head size after in utero exposure to atomic radiation. Lancet ii:784-787; 1972.

38. Miller, R. W.; Mulvihill, J. J. Small head size after atomic irradiation. Teratology 14:355-358; 1976.

39. Mullenix, P.; Norton, S.; Culver, B. Locomotor damage in rats after X-irradiation in utero. Exp. Neurol. 48:310-324; 1975.

40. Murphy, D. P. Maternal pelvic irradiation. In: Murphy, D. P., ed. Congenital malformations. 2nd ed. Philadelphia: Lippincott; 1947.

41. Muller, F.; O'Rahilly, R. Cerebral dysraphia (future anencephaly) in a human twin embryo at stage 13. Teratology 30:167-177; 1984.

42. Neumeister, K. Findings in children after radiation exposure in utero from X-ray examination of mothers. In: Late biological effects of ionizing radiation, vol. 1. Vienna: IAEA; 1978:119-134.

43. Nokkentved, K. Effects of diagnostic radiation upon the human fetus. Copenhagen: Munksgaard; 1968.

44. Norton, S. Behavioral changes in preweaning and adult rats exposed prenatally to low ionizing radiation. Toxicol. Appl. Pharmacol. 83:240-249; 1986.

45. Norton, S. Correlation of cerebral cortical morphology with behavior. Toxicol. Indust. Health 5:247-255; 1989.

46. Norton, S.; Kimler, B. F. Correlation of behavior with brain damage after in utero exposure to toxic agents. Neurotoxicol. Teratol. 9: 145-150; 1987.

47. Norton, S.; Mullenix, P.; Culver B. Comparison of the structure of hyperactive behavior in rats after brain damage from X-irradiation, carbon monoxide and pallidal lesions. Brain Res. 116:49~57; 1976.

48. Oppenheim, B. E.; Griem, M. L.; Meier, P. An investigation of effects of prenatal exposure to diagnostic X rays. In: Biological and environmental effects of low-level radiation, vol. II. Vienna: IAEA; 1976:249-260.

49. O'Rahilly, R.; Gardner, E. The developmental anatomy and histology of the human central nervous system. In: Vinken, P. J.; Brnyn, G. W., eds. Handbook of clinical neurology, vol. 30. New York: North-Holland Publ. Co.; 1977:17-40.

50. Otake, M.; Yoshimaru, H.; Schull, W. J. Severe mental retardation among the prenatally exposed survivors of the atomic bombing of Hiroshima and Nagasaki: A comparison of the T65DR and DS86 dosimetry systems. Radiation Effects Research Foundation Technical Report 16-87.

51. Otake, M.; Schull, W. J.; Fujikoshi, Y.; Yoshimaru, H. Effect on school performance of prenatal exposure to ionizing radiation in Hiroshima: A comparison of the T65DR and DS86 dosimetry systems. Radiation Effects Research Foundation Technical Report 2-88.

52. Rakic, P. Mode of cell migration to the superficial layers of fetal monkey neocortex. J. Comp. Neurol. 145:61-83; 1972.

53. Rakic, P. Timing of major ontogenetic events in the visual cortex of the Rhesus monkey. In: Brain mechanisms in mental retardation. New York: Academic Press, Inc.; 1975:3-40.

54. Rakic, P. Prenatal genesis of connections subserving ocular dominance in the rhesus monkey. Nature 261:467-471; 1976.

55. Rakic, P. Prenatal development of the visual system in the rhesus monkey. Philos. Trans. R. Soc. Lond. [B] 278:260; 1977.

56. Rakic, P. Neuronal-glial interaction during brain development. Trends Neurosci. 4:184-187; 1981.

57. Rakic, P. Limits of neurogenesis in primates. Science 227:1054- 1055; 1985.

58. Rakic, P. Specification of cerebral cortical areas. Science 241: 170-241; 1988.

59. Rakic, P. Defects of neuronal migration and the pathogenesis or cortical malformations. Prog. Brain Res. 73:15-37; 1988.

60. Rakic, P.; Sidman, R. L. Histogenesis of cortical layer in human cerebellum, particularly the lamina dissecans. J. Comp. Neurol. 139:473-500; 1970.

61. Riley, E. P.; Vorhees, C. V., eds. Handbook of behavioral teratology. New York: Plenum Press; 1986:v-522.

62. Roesch, W. C., ed. Reassessment of atomic bomb radiation dosimetry in Hiroshima and Nagasaki: Final report. Hiroshima, Radiation Effects Research Foundation; 1987.

63. Schneider, B. F.; Norton, S. Postnatal behavioral changes in prenatally irradiated rats. Neurobehav. Toxicol. 1:193-197; 1979.

64. Schneider, B. F.; Norton, S. Development of neurons in the forebrain of rats following prenatal X-irradiation. Neurotoxicology 1:525-532; 1980.

65. Schull, W. J.; Otake, M.; Yoshimaru, H. Effect on intelligence test score of prenatal exposure to ionizing radiation in Hiroshima and Nagasaki: A comparison of the T65DR and DS86 dosimetry system. Radiation Effects Foundation Technical Report 3-88; 1988.

66. Seo, M. L.; Inouye, M.; Kiyono, S.; Shibagaki, M. Effects of environmentally differential rearing upon maze performance in prenatally irradiated microcephalic rats. Teratology 26:221-227; 1982.

67. Tabuchi, A.; Hirai, T.; Nakagawa, S.; Shimada, K.; Fugita, J. Clinical findings on in utero exposed microcephalic children. Atomic Bomb Casualty Commission Technical Report. 28-67; 1967.

68. Tamaki, Y.; Inouye, M. Brightness discrimination learning in a Skinner box in prenatally X-irradiated rats. Physiol. Behav. 16: 343-348; 1976.

69. Tamaki, Y.; Inouye, M. Avoidance of and anticipatory responses to shock in prenatally X-irradiated rats. Physiol. Behav. 22:701-705; 1979.

70. Tanebashi, M. Intelligence and intelligence tests. In: Koga, Y., ed. Outline of educational psychology. 20th ed. Tokyo: Kyodo Shuppan- sha; 1972:128-158 (in Japanese).

71. Wallace, R. B.; Graziadei, R.; Werboff, J. Behavioral correlates of focal hippocampal X-irradiation in rats. II. Behavior related to adaptive function in a natural setting. Exp. Brain Res. 43:207-212; 1981.

72. Werboff, J.; Havlena, J.; Sikov, M. R. Effects of prenatal X- irradiation on activity, emotionality and maze-learning ability in the rat. Radiat. Res. 16:441-452; 1962.

73. Wood, J. W.; Johnson, K. G.; Omori, Y. In utero exposure to the Hiroshima atomic bomb: Follow-up at twenty years. Atomic Bomb Casualty Commission Technical Report 9-65; 1965.

74. Wood, J. W.; Johnson, K. G.; Omori, Y.: Kawamoto, S.; Keehn, R. J. Mental retardation in children exposed in utero. Hiroshima- Nagasaki, Atomic Bomb Casualty Commission Technical Report 10-19; 1966.

75. Zecevic, N.; Rakic, P. Differentiation of Purkinje cells and their relationship to other components of developing cerebellar cortex in man. J. Comp. Neurol. 167:27-47; 1976.

Documents

Ionizing radiation and the developing brain