ELSEVIER

Reproductive Toxicology, Vol. 11, No. 4, pp. 583-587, 1997 0 1997 Elsevier Science Inc.

Printed in the USA. All rights reserved 0890.6238/97 $17.00 + .OO

PI1 SO890-6238(97)00006-3

GROWTH AS A MANIFESTATION OF TERATOGENESIS: LESSONS FROM HUMAN FETAL PATHOLOGY

MASON BARR, JR.

Teratology Unit, Departments of Pediatrics, Pathology, and Obstetrics, University of Michigan, Ann Arbor, Michigan

INTRODUCTION

Prior to my arrival from Michigan to spend a year with Bob Brent, he sent me one of his latest reprints on the subject of intrauterine growth retardation (1). While my original aim in going to Philadelphia was to try out the ruby laser as a surgical tool in the embryo, Bob got me working on uterine vascular ligation in the rat and its effects on fetal growth. It became evident that certain basic information was needed before we could interpret the results of the ligations. So, together with Ron Jensh, we started a series of work defining the roles of some natural variables on fetal growth (2-4).

When my year in Philadelphia was up, I returned to Michigan and set up my own laboratory to continue these studies in rats. All went well until a pathology resident appeared in my office one day with some questions about

an autopsy done on an infant with simelia. He could not answer any of my questions about the finer details of morphology. In short order I switched from rat to human fetuses. But my time with Bob was not wasted; he taught me the value of very careful measurement, data collec- tion, and statistical analysis. More significantly, he got me hooked on the idea that growth aberration is an important manifestation of teratogenesis. As I started my new career as a fetal pathologist, I discovered that the available data concerning normal fetal measurements were incomplete and often inaccurate, particularly in regard to variation about the mean. (William Werten- baker is said to have observed “Norms are human ob- sessions; nature rarely bothers to construct one.“) There was only one thing to be done-collect the information myself, using the obsessive-compulsive techniques that Bob Brent taught me when I worked with rat pups.

This manuscript was contributed as part of the April 1996 sym- posium held by Jefferson Medical College of Thomas Jefferson Uni- versity, Philadelphia, PA, honoring Dr. Robert L. Brent, M.D., Ph.D., D.Sc. (bon), for his administrative, educational, and research contribu- tions.

Address correspondence to Mason Barr, Jr., M.D., Pediatric Genetics-Teratology, D1109 MPB 0718, University of Michigan, Ann Arbor, MI 48 109.

It took some time and patience to amass enough data to have a go at defining “normal.” Cooperatively,

with Tom Shepard (5) and Will Blackbum (6), a decent sample size of “normal” (and by that I mean morpho- logically normal and otherwise typical) fetuses was

collected. From the start, a variety of measurements had been carefully recorded for each fetus examined, normal or abnormal. This gave us several ways to validate that

fetal growth had in fact been normal or typical, and it allowed the exclusion of those fetuses with aberrant growth patterns. After considerable experimentation with

curve-fitting techniques, I was able to derive efficient descriptions for both the somatic and visceral measure- ments (6).

In this sort of work the primary interest has usually been to correlate measurement to gestational age. How- ever, dating by the last menstrual period is widely

considered to be a shaky proposition (7) and my data confirm that impression (8) (Figure 1). With the more

frequent use of early ultrasonography to validate dating, more accurate age estimates have become more com-

mon. Nonetheless, a good deal of the data we have on ages are obviously inaccurate. Accordingly, we ap- proached the definition of normal using fetal weight, crown-rump length, and brain weight as the primary standards. As more accurate age data become available, we are beginning to be able to make the correlation with gestational age and are finding that normal growth is more predictable (i.e., smaller variance about the mean)

than had been appreciated in studies that accepted “LMPs” as valid information. I believe our data are sound for fetal weights up to about 2000 g; above that they must be regarded as best estimates since the sample size drops off drastically.

Once we had an idea of what normal was, it was possible to go back to the larger collection of fetuses and look for growth aberrations among abnormal fetuses and those sharing some particular gestational experiences. A number of these analyses have been done and some have been reported; I will touch only on some of those that I find particularly intriguing.

583

584 Reproductive Toxicology Volume 1 I, Number 4, 1997

1 2 3 4 5 6 7 9 9 ‘10’11’12’13’14’15’16’17’18’1 9’20’21‘22’23’24’25’26’27’29’29’30’31’

Day stated as first day of last menstrual period (LMP)

Fig. I. Frequency distribution of the days listed as the first day of the last menstrual period, demonstrating a nonrandom distribution.

Unilateral renal agenesis

To determine whether compensatory growth of the kidney occurs during fetal life we studied 20 human specimens with unilateral renal agenesis as an isolated

defect. We found that the total renal mass in unilateral renal agenesis was 83% of that in weight-matched controls (9). By histologic study, a uniform increase in all nephron elements was found. There are several unanswered questions: Why does a fetus with adequate placental clearance of metabolic wastes have compensa-

tory renal growth? Could there be an induced negative feedback system involving a renotropic factor? If so, what is the initial stimulus? Initially, we thought that the stimulus might be oligohydramnios (lo), but the suspi- cion proved unfounded when a larger number of fetuses with prolonged amniorrhea was studied; their renal

weights were not increased (Figure 2). So the phenom- enon of compensatory renal growth seems to apply only to unilateral renal agenesis, and our original questions remain to be answered.

Fetal hypotension

Based on only a few cases with appropriate data, fetuses with ACE inhibitor fetopathy appear to have shorter than normal arms (11). Working from the suspi- cion that the physiologic basis of ACE inhibitor fetopa- thy is hypotension, it was hypothesized that relative growth lag of the limbs is a marker of an adaptive shift

of circulation in the compromised fetus, diminished to the periphery to preserve central perfusion. To test this notion, arm lengths of other fetuses suspected of suffer- ing from hypotension were compared to normals. Based on a still small sample (and no direct evidence of their blood pressure) shortening of the limbs does seem to be associated with fetal hypotension (unpublished obser- vations). At the very least, this suggests that further research to assess fetal blood pressure and its correlation with limb growth might be fruitful.

Interestingly, because we suspected both that the donor twin in the twin-twin transfusion syndrome must

be chronically hypotensive and that the renal tubular dysgenesis found in ACE inhibitor fetopathy is second-

ary to hypotension-associated decreased renal blood flow (1 I), we looked at the kidneys of donor twins and found that they show evidence of proximal convoluted tubule dysgenesis (12). On the other hand, fetuses with Turner syndrome (vide infra) may also be hypotensive and they certainly have short limbs (Figure 4), but they do not have renal tubular dysgenesis.

Twin-twin tmnsfusion

In studies of monozygotic vs. dizygotic twins we found that monozygotic twins with diamniotic, mono- chorionic placentation had a high degree of brain-sparing growth restriction in the smaller twin and cardiac hyper- plasia in the larger twin (13,14; Figure 3). We think that these findings are most likely caused by the hemody- namic inequalities that exist in the twin-twin transfusion syndrome. Contrary to the assumption that hemodynamic imbalance is not likely to be significant until there is a 20 to 30% body weight difference between the twins, we found evidence of cardiac hyperplasia in the recipient twin even when the body weight difference was still insignificant. This is relevant not only to the well being of the fetus in utero; it also predicts the possibility of a difficult adaptation to extrauterine life by the larger, recipient twin when the interfetal communication is

severed at birth.

Turner syndrome The vast majority of fetuses with Turner syndrome

die by midgestation. The typical appearance of this

n Ammorrhea 4-13 days (n 30)

Amnlorfhea 14-27 days (n lr3)

D Amniorhea 28+ days (“18)

Fig. 2. Deviation of the mean observed Z-Score from normal for various somatic and visceral measurements by duration of amniorrhea (body weight; crown-heel length; crown-rump length; occipito-frontal head circumference; arm, leg and foot lengths; visceral weights.

Growth as a manifestation of teratogenesis 0 M. BARR 585

Larger Twin Smaller Twin I L 8, I

6 . .

6 ; ._ : .: t. .

5 4 . .

. . . ,... : .I

z . : .

z 2 i

8 0~

‘. .z : .i : C: .

; .: . 0

$ 0 Ikr ” ,.,

: ., :*, : :‘. :‘I. : : 1 *I / : ! : : :-

-2 :: Y

: z ,;,o

L. 1.93 i .,~

:;.

-4 f -10 0 10 20 30 40 50 60 70

Percent difference I” body weights

*Larger Twn ~&nailer Twin

Fig. 3. Somatic and visceral growth profile of monozygotic twins at midgestation. The standard used is body weight. The bars represent the mean observed Z-Score ? 1 SD. The bottom graph shows the observed Z-Scores of hearts plotted against percentage body weight differences between the twins of a set.

Autosomal trisomies

“Phenotype” typically refers to structural features, although for many syndromes growth and functional

features are key components. Many factors influence prenatal growth, including heritable factors, environmen- tal factors, and factors intrinsic to the conceptus. Growth aberrations may be somatic or visceral, generalized or localized, transient or permanent, accelerated or restrict- ed, and of early or late onset. They may be correlated, singly or in combination, with structural abnormalities, functional abnormalities, death, or premature delivery. Somatic and visceral growth profiles of midgestation human fetuses with trisomy 21, 18, or 13 demonstrate

that each disorder has a characteristic pattern of growth aberration ( 17).

Some of the growth aberrations are of such a nature that they are useful for the in utero ultrasonographic screening of fetuses for autosomal trisomy. This is cer- tainly the case for trisomy 21, the most prevalent of the trisomies at midgestation (Figure 5). The shortness of

both the arms and legs has been noted by many investi- gators, but short limbs by themselves are not sufficiently specific indicators of trisomy 21 (18-23). To reduce the number of false positive screening tests and consequent

Turner Syndrome tjo Hydrops or Hygrorqas (n 4)

condition is a fetus with massive hydrops and huge

loculated hygromas around the neck. Because there is at

least moderate sparing of brain growth, brain weight was

selected as the best normative standard for other mor-

phometrics. Among the Turner fetuses were four with no

hydrops/hygroma; these had normal heart weights (Fig-

ure 4). The mean heart weight of 21 hydropic/hygroma-

tous cases was markedly reduced at the 0.45 centile

(mean Z-Score -2.61, 68% C.I. -3.53 to -1.68). This

heart weight reduction was independent of the presence of aortic coarctation or bicuspid aortic valve. For 20

fetuses with hydrops and hygromas, but normal karyo-

types, the mean heart weight was at the 78.5 centile

(mean Z-Score 0.79, 68% C.I. - 1.42 to 2.99). We have

postulated that the cardiac hypoplasia may be a primary defect in Turner syndrome and that its occurrence (and

its severity) is a contributory factor in the expression of the hydropic- hygromatous phenotype and the principal

element leading to intrauterine death in midgestation

(15,16).

Hydropa 6. Hygwmas (n 2?l

“=&YE ,z 0 0 Cl z s ?

Normal Kavo!ype Hydrops 6 Hygromas (n 20)

Fig. 4. Somatic and visceral growth profiles of fetuses with Turner syndrome (with and without hydrops and hygromas) and a comparison group of fetuses with hydrops and hygromas but normal karyotypes. The standard used is brain weight. The bars represent the mean observed Z-Score ? 1 SD.

586 Reproductive Toxicology

Trisomy 21 (n. 68) 4.0

30

Trisomy 18 (n ?4) F

Fig. 5. Somatic and visceral growth profiles of fetuses with trisomy 21 and trisomy 18. The standard used is body weight. The bars represent the mean observed Z-Score ? 1 SD.

amniocenteses, while maintaining or increasing the de- tection rate for trisomy 21, we have tested various combinations of limb measurements that we believe may increase both the specificity and sensitivity of ultrasono- graphic screening (24-27).

Trisomy 18 is often cited as an archetypal example of “symmetrical” intrauterine growth restriction. Our

data indicate that in trisomy 18, at midgestation, the growth aberration is anything but symmetrical (28; Fig- ure 5). The lung and adrenal weights are often reduced to a degree that suggests the neonate with trisomy 18 may experience physiologic problems with lung and adrenal function (17). This may, in turn, contribute to the early neonatal lethality of this syndrome.

Supranormal spleen and kidney weights and sub- normal lung weights are found in trisomy 13 (17; Figure 6). I wondered if the reduced lung weight might be a consequence of reduced fetal breathing activity associ- ated with holoprosencephaly. However, fetuses with isolated alobar holoprosencephaly do not show evidence of lung undergrowth. Other than having comparable degrees of reduced head circumference and brain weight, euploid fetuses with holoprosencephaly show no other growth pattern similarities to trisomy 13 fetuses. The reason for and significance of the splenomegaly and renomegaly in trisomy 13 remain to be discovered.

While growth aberrations may most commonly be inherent to the genotype, some of the observed patterns suggest the need for investigation of the interaction between physiology and growth of the trisomic fetus. Some disturbances in fetal physiology may not be specific to any one syndrome, but might have general application irrespective of the syndrome. For example, if our hypothesis about the relation of short limbs with fetal hypotension is generally true, finding growth restriction

Volume 11, Number 4, 1997

of the limbs could be a tip-off to the physiologic de- rangement and allow a focused approach to fetal therapy.

The finding of an autosomal trisomy is often con- sidered a sufficient explanation of fetal or neonatal death. This is a rather incurious approach. As pointed out by Rushton (29), an abnormal karyotype does not really explain why that particular fetus died. We should be better informed if we could understand whether or not

the presence of small adrenals, small hearts, small lungs, large spleens, or large kidneys has anything to do with

the clinical course. If some 20% of trisomy 21 fetuses die between midgestation and term (30), our question should be why and how do they die? I suspect that there is a physiologic cause of death in the case of many Turner syndrome fetuses. The capacity of the small heart to provide sufficient circulation seems eventually to be exceeded, resulting in right-sided heart failure.

Warkany (31) coined the term “terathanasia” in recognition of the lethality of many malformations and anomalies. He suggested that, if nature generally aborts abnormal fetuses, perhaps the survival of some repre- sents a failure of a natural screening process and the reasons for this failure are worthy of study. The finding of aberrant growth patterns suggests additional ways to

address Warkany’s challenge.

T!isq.ey 1s Holoprosencephaly (n 8)

I . I I -I- I- I

I . - - ,

No!malKaryotYPs Holoprosencephaly (” 3)

Fig. 6. Somatic and visceral growth profiles of fetuses with trisomy 13 (with and without holoprosencephaly) and a com- parison group of fetuses with holoprosencephaly but normal karyotypes. The standard used is body weight. The bars repre- sent the mean observed Z-Score -t 1 SD.

Growth as a manifestation of teratogenesis 0 M. BARR 587

CONCLUSION

In my laboratory detailed analysis of fetal somatic and visceral growth is a standard feature of the autopsy. The concepts explored here are used not only in inter- pretation of the autopsy findings. They are used prospec- tively in the evaluation of fetuses in utero in the attempt

to achieve improved reproductive outcome. An adven- ture, already under way, is the use of such technology as ultrasonic imaging and Doppler flow studies to see if the alterations in fetal physiology that are suggested by the findings from the autopsy room can be reliably detected in utero. I believe that coordinated studies among the perinatologist, neonatologist, ultrasonographer, and fetal pathologist offer exciting possibilities for improving

pregnancy outcome.

1.

2.

3.

4.

5.

6.

7.

8.

9.

10.

11.

12.

13.

14.

REFERENCES

Brent RL, Jensh RP. Intrauterine growth retardation. In Woollam DHM, ed. Advances in teratology, vol. 2. London: Logos Press; 1967:139-227. Barr M, Jensh RP, Brent RL. Fetal weight and intrauterine position in rats. Teratology. 1969;2:231-46. Barr M, Jensh RP, Brent RL. Prenatal growth in the albino rat: effect of number, intrauterine position and resorptions. Am J Anat. 1970;128:413-28. Jensh RP, Brent RL, Barr M. The litter effect as a variable in teratologic studies of the albino rat. Am J Anat. 1970; 128: 185-92. Shepard TH, Shi M, Fellingham GW, Fujinaga M, Fitzsimmons JM, Fantel AG, Barr M. Organ weight standards for human fetuses. Pediatr Pathol. 1988;8:513-24. Barr M, Blackburn WR, Cooley NR. Human fetal somatic and visceral morphometrics. Teratology. 1994;49:487-96. Alexander GR, Allen MC. Conceptualization, measurement, and use of gestational age. I. Clinical and public health practice. J Perinatol. 1996;16:53-9. Barr M. Use of the last menstrual period (LMP) in timing human pregnancies. Teratology. 1981;23:134. Hartshorne N, Shepard T, Barr M. Compensatory renal growth in human fetuses with unilateral renal agenesis. Teratology. 1991;44: 7-10. Barr M. Amniotic volume and fetal renal weight. Teratology. 1987;35:67A-68A. Pryde PG. Sedman AB, Nugent CE, Barr M. Angiotensin-convert- ing enzyme inhibitor fetopathy. J Am Sot Nephrol. 1993;3:1575- 82. Martin RA, Jones KL, Mendoza A, Barr M, Benirschke K. The effect of ACE inhibition on the fetal kidney: decreased renal blood flow. Teratology. 1992;46:3 17-21. Pridjian G, Nugent CE, Barr M. Twin gestation: influence of placentation on fetal growth. Am J Obstet Gynecol. 1991;165: 1394-401. Barr M. Twin growth profiles and identification of twin-twin

15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

25.

26.

27.

28.

29.

30.

31.

transfusion in midgestation. Proc Greenwood Genet Ctr. 1996;15: 151-2. Barr M, Oman-Ganes L. Subnormal heart weight and hydropic Turner syndrome in midgestation: a cause of death. Proc Green- wood Genet Ctr. 1989;8: 169. Barr M. Cardiac hypoplasia as a primary defect and cause of death in Turner syndrome fetuses. Proc Greenwood Genet Ctr. 1994;13: 61-2. Barr M. Growth profiles of human autosomal trisomies at midges- tation. Teratology. 1994;50:395-8. Cuckle H, Wald N, Quinn J, Royston P, Butler L. Ultrasound fetal femur length measurement in the screening for Down’s syndrome. Br J Obstet Gynaecol. 1989;96:1373-8. Fitzsimmons J, Droste S, Shepard TH, Pascoe-Mason J, Chinn A, Mack LA. Long-bone growth in fetuses with Down syndrome. Am J Obstet Gynecol. 1989;161:1174-7. Rodis JF, Vintizileos AM, Fleming AD, Ciarleglio L, Nardi A, Feeney L, Scorza WE, Campbell WA, Ingardia C. Comparison of humerus length with femur length in fetuses with Down syndrome. Am J Obstet Gynecol. 1991;165:1051-6. Benacerraf BR, Neuberg D, Frigoletto FD. Humeral shortening in second-trimester fetuses with Down syndrome. Obatet Gynecol. 1991;77:223-7. Rotmensch S, Luo JS, Liberati M, Belanger K, Mahoney MJ, Hobbins JC. Fetal humeral length to detect Down syndrome. Am J Obstet Gynecol. 1992;166:1330-4. Nyberg DA, Resta RG, Luthy DA, Hickok DE, Williams MA. Humerus and femur length shortening in the detection of Down syndrome. Am J Obstet Gynecol. 1993;168:534-8. Johnson MP, Barr M, Evans MI. Third trimester progression of aneuploid-specific aberrant growth patterns of trisomy 21. Am J Hum Genet. 1990;47:A278. Johnson MP, Barr M, Treadwell MC, Michaelson J, Isada NB, Pryde PG, Dombrowski MP, Cotton DB, Evans MI. Fetal leg and femur:foot length ratio: a marker for trisomy 21. Am J Obstet Gynecol. 1993;169:557-63. Johnson MP, Michaelson JE, Barr M, Treadwell MC, Isada NB, Dombrowski MP, Pryde PG, Addis J, Evans Ml. Sonographic screening for trisomy 21: fetal humerus:foot length ratio, a useful new marker. Fetal Diagn Therapy. 1994;9:130-8. Johnson MP, Michaelson JE, Barr M, Treadwell MC, Hume RF, Dombrowski MP, Johnson A, Bichalski J, Evans MI. Combining humerus and femur length for improved sonographic identification of pregnancies at increased risk for trisomy 21. Am J Obstet Gynecol. 1995;172:1229-35. Johnson MP, Barr M, Qureshi F, Drugan A, Evans MI. Symmet- rical intrauterine growth retardation is not symmetrical: organ- specific gravimetric deficits in midtrimester and neonatal trisomy 18. Fetal Therapy. 1989;4:110-9. Rushton DI. The nature and causes of spontaneous abortion with normal karyotypes. In H Kalter, ed. Issues and reviews in teratol- ogy, vol. 3. New York: Plenum Press; 1985:21-63. Hook EB. Genetic counseling and prenatal cytogenetic services: policy implications and detailed cost-benefit analyses of programs for the prevention of Down syndrome. In Porter IH, Hook EB, eds. Service and education in medical genetics. San Diego: Academic Press; 1979:29-54. Warkany J. Terathanasia. Teratology. 1978; 17: 187-92.