TERATOLOGY 44599-604 (1991)
Limb Reduction Defects in Humans Associated With Prenatal lsotretinoin Exposure RENATA RTXXO. EDWARD J. LAMMER. ENRICO PARANO. ~~
~
~
~
~~~~
LORENZO PAVONE, AND J. CRAIG ARGYLE Clinzca Pediatrica, Uniuersitu di Catania, Catania, Italy 95125 (R.R.,EP., L.P.); California Birth Defects Monitoring Program, Emeryuille, California 94608 (E.J.L.): and Department of Pathology, Wilford Hall Medical Center, Lackland AFB, Texas 78236 (J.C:A.)
ABSTRACT Retinoic acid has long been used to induce limb reductions defects in experimental animal studies. No limb malformations, however, have been reported among malformed retinoic acid-exposed human fetuses from case reports o r epidemiologic studies. We report a child and a fetus with limb reduction malformations following maternal use of isotretinoin (13-cis-retinoic acid) during the first trimester of pregnancy. The child had a unilaterally absent clavicle and nearly absent scapula, with a short humerus and short, synostotic forearm bones. He also had ventriculomegaly and developmental delay, minor dysmorphic facial features, and a short sternum with a sterno-umbilical raphe. The fetus had a unilaterally absent thumb with normal proximal bony structures. Other findings included hydrocephalus, craniofacial anomalies, thymic agenesis, supracristal ventricular septa1 defect, single umbilical artery, anal and vaginal atresia, and urethral agenesis with dysplastic, multicystic kidneys. Although the limb malformations were quite dissimilar, a number of anomalies that are frequently found among isotretinoin-exposed fetuseslinfants were present in both cases. This increases the probability that retinoic acid caused these limb defects, but a causal association cannot be conclusively drawn on the basis of these two retrospective case reports. Case 1 Retinoic acid is known to be teratogenic in humans (Lammer et al., '85a) and a numC.A. was born to a 17-year-old mother ber of experimental animal species (Shene- and 24-year-old father after a full-term unfelt, '70; Webster et al., '86; Fantel et complicated pregnancy. The family history al., '76). In humans, the use of isotretinoin was unremarkable and the parents were not (13-cis-retinoic acid) for the treatment of consanguineous. He was their first child. severe nodulocystic acne has led to many During the first month after conception the inadvertent embryonic and fetal exposures. mother took 50 mglday of isotretinoin for These exposures have induced malforma- acne. She took isotretinoin for 10 days, but tions principally affecting craniofacial, cen- the gestational day of the final dose of tral nervous system, cardiac, and thymic isotretinoin is uncertain. Sometime during development. Although retinoic acid-in- the first two months of gestation, she also duced limb defects have been studied exper- had a fever for three days that was not imentally as a model system for limb dys- treated. The mother reported that she felt morphogenesis, no human fetus or infant fetal movements regularly. Labor was sponwith primary limb reduction defects follow- taneous and the vaginal delivery was uning exposure to isotretinoin has been re- complicated. Birthweight was 3,500 gm ported. We report a child and a fetus with (50th percentile), length was 50.5 cm (50th prenatal exposure to isotretinoin who were percentile), and head circumference (OFC) delivered with limb reduction defects and who had other abnormalities that have been reported previously in the retinoic acid emReceived November 5 , 1990; accepted August 14, 1991 bryopathy . 0 1991 WILEY-LISS, INC.
600
R. RIZZO ET AL.
Fig. 1. Case 1. Note short, internally rotated humerus attached to thorax, and sterno-umbilical raphe.
Fig. 2. Case 1. Absence of right clavicle. Small anomalous bone (arrow) posterior to right ribs is presumably a markedly underdeveloped scapula.
was 34.5 cm (50th percentile). Several anomalies were present a t birth, but the neonatal period was uneventful. He was clavicle, and a short humerus that directly able to hold his head erect at 5 months and attached at the thorax (Fig. 2). The right sit alone at 7 months. At 7 months, he was radius and ulna were also short, broad, and admitted at the Pediatric Clinic of Catania completely synostotic (Fig. 3). Digital subtraction arteriography demonstrated that for a febrile convulsion. At 7 months, his weight was 8,400 gm the right subclavian artery terminated just (25th percentile), length was 81 cm (75th distal to the origin of the internal mampercentile), and OFC was 48.5 cm (90th per- mary artery, which was enlarged. A large centile). His head was dolichocephalic with right superior (supreme) intercostal artery a prominent forehead. His ears were large entered the right upper limb and apparently and round. The nose was short with a provided its arterial supply. The seizures have not recurred. At age 2.5 rounded tip and the philtrum was broad. There was mild micrognathia. The neck was years, his development is delayed; he has long. There was marked thoracic asymme- just begun walking and speaks only a few try with hypoplasia of the right upper chest, syllables. shoulder girdle, and limb. The right clavicle Case 2 and scapula were not palpable. The short, An 18-year-old GI Po woman conceived internally rotated right humerus was attached to the thorax a t the level of the 4th- while taking 60 mglday of isotretinoin for 5th ribs (Fig. 1). Both radius and ulna were cystic acne. She used it for 62 days after short. The fingers of the right hand were conception. The pregnancy was otherwise shorter than the left and had camptodac- unremarkable until hydrocephalus and a tyly. The sternum was short. A visible, cystic kidney were prenatally diagnosed by raised fibrous cord extended for 4 cm be- ultrasonography. The pregnancy was termitween the umbilicus and the xiphoid (Fig. nated a t 22 weeks. The placenta was small, 1).The right testis was undescended. 92 gm (normal 120-230 gm), and there was The routine laboratory findings were a single umbilical artery. There were no normal: a urine metabolic screen, amino ac- other gross or microscopic abnormalities of ids, lactate, pyruvate and CPK. An ophthal- the placenta. mological examination, electrocardiogram, Fetal growth was normal, weight was 425 electroencephalogram, and intravenous py- gm (50th percentile), crown-rump length elogram were also normal. A high-resolu- was 19.0 cm (50th percentile at this weight), tion karyotype showed normal 46,XY chro- and length was 26.0 cm. The head was mosomes. Brain MRI and CT scans revealed large, and the external ears were small, asymmetrically enlarged ventricles. Skele- malformed, and low-set (Fig. 4). The mantal x-rays showed nearly complete absence dible was small. The brain could not be exof the right scapula, agenesis of the right amined microscopically. The eyes were nor-
RETINOIC ACID-INDUCED LIMB MALFORMATIONS
601
Fig. 5. Case 2. Absent left thumb.
Fig. 3. Case 1. Complete synostosis of short right radius and ulna. Camptodactyly of fingers.
Fig. 4. Case 2. Enlarged fetal head due to hydrocephalus and malformed, lowset external ear.
ma1 grossly and microscopically. Examination of the thoracic cavity revealed no thymus and microscopic sections of the superior mediastinum did not reveal any thymic tissue. A small, cystic, mediastinal structure lined by squamous, respiratorytype epithelium was found. The lungs were small for birthweight and weighed 6.3 gm (5th percentile; Shepard et al., '88). An aberrant right upper lobe bronchus originated above the carina. The right lung consisted of a lower lobe and a fused upper and middle
lobe. The left lung had no fissures. The heart weighed 2.9 gm, the foramen ovale was patent, and a ventricular septa1 defect (2 mm) was found just inferior to the pulmonic valve. The external genitalia were ambiguous and no external vaginal opening or anus were present. Within the abdominal cavity, the distal colon ended as a dilated pouch. A normal uterus and left fallopian tube and ovary were identified, but no adnexal structures were found on the right. The left kidney was large (2.1 cm) and multicystic. The right kidney (1.2 cm) was medially displaced and multicystic. Ureters and a bladder were found, but no urethra. Microscopically, the dysplastic kidney contained multiple small parenchymal cysts consistent with renal dysplasia secondary to obstruction. The left adrenal gland was smaller than the right but had a normal microscopic appearance. The left thumb and thenar eminence were absent (Fig. 5). Whole body radiographs revealed no other ossseous anomalies. A normal left radius and ulna were present. The temporal bones were markedly dysplastic. DISCUSSION
This fetus and infant represent the first reports of primary limb reduction defects following isotretinoin during early pregnancy. McBride ('85) previously reported a child with terminal transverse limb reductions of all four extremities whose mother had taken isotretinoin from days 26 to 40 of gestation. The weight of evidence, however, indicated that these limb reductions occurred secondary to amniotic band amputations rather than as primary malformations (Lammer et al., '85b). Limb reduction defects have been reported following exposure
602
R. RIZZO ET AL.
to another synthetic retinoid, etretinate (Happle et al., '84). These malformations included two cases with a short thumb, absent finger, syndactyly, and a forearm defect. In another report, a malformed fetus was delivered of a woman who stopped etretinate therapy 4 months before conception but had detectable serum levels of etretinate and etretin when the pregnancy was terminated a t 10 weeks (Grote et al., '85). That fetus had a hypoplastic femur, absent tibia and fibula, and a single toe. Because limb reduction defects have not been found previously among case reports or systematic epidemiologic studies of isotretinoin exposure, is i t likely that the limb reduction defects of our report were caused by isotretinoin? For both cases, a number of anomalies were present that involve the structures most commonly affected in the retinoic acid embryopathy (Lammer et al., '85a; Lammer, '87). Case 1had ventriculomegaly, developmental delay, and minor craniofacial anomalies (rounded ears, short nose, micrognathia); case 2 had hydrocephalus, malformed ears, and DiGeorge sequence. It is uncommon, however, to find the additional anomalies reported for case 2: anal atresia, distal vaginal atresia, and urinary tract obstruction. The supraumbilical raphe with sternal defect found for case 1 has not been reported in retinoic acid embryopathy. This unusual malformation has been reported in association with large facial hemangiomas, but rarely with other malformations (Hersh, '85; Leiber, '82; Gotlieb e t al., '82). Our case lacked any hemangiomas, however, and does not represent this recently recognized syndrome of unknown etiology. The shoulder malformation of patient 1 is exceptionally rare and does not occur a s part of any Mendelian disorder that affects limb development (Temtamy and McKusick, '78). Absence of the scapula has been reported only when the distal limb is completely absent (Wood and Marchinski, '90). In the thalidomide embryopathy, the glenoid fossa was occasionally malformed in association with more distal limb malformations, but no scapular anomaly like that of patient 1 was reported. We did not identify any cytogenetic abnormalities or environmental exposures, other than isotretinoin, that might have caused the malformations of our two cases. In addition, the timing of the isotretinoin exposure for case 2 extended well into the period of limb develop-
ment, while the interval of exposure for case 1 was uncertain. How do these two types of limb reduction defects compare with the patterns of limb anomalies induced by retinoic acid in experimental animals? In the experimental setting, retinoic acid may cause limb duplications or reductions (Cusic and Dagg, '85; Sulik and Dehart, '88; Kochhar, '67; Tickle et al., '82) and the types of malformation depend upon the timing, dose, and duration of exposure. Kochhar ('77) and others have induced proximal upper limb anomalies by exposing mice to all-trans-retinoic acid on day 11 and later (Fraser and Travill, '78). Although the scapula is sometimes reduced to a small triangular bone in these experimental studies, nearly complete absence of the scapula and clavicle a s in our case 1has not been described. By treating a t earlier times (day 9.5) with 13-cis-retinoicacid, and with higher doses, more distal types of limb reductions were also induced (Sulik and Dehart, '88). These limb reductions included ectrodactyly of the preaxial andlor postaxial side. Preaxial and postaxial polydactyly also occurred. Absence of the first preaxial digit, as occurred in our case 2, was the most frequent distal reduction defect in some experimental studies (Kochhar, '77). In a pigtail monkey study of all-trans-retinoic acid (10 mglkg), Fantel et al. ('76) reported finding postaxial ectrodactyly as the most common limb anomaly, with several cases of syndactyly and one of bilateral short ulnas. In another primate study in which 41 pregnant cynomolgus monkeys were given isotretinoin in doses from 2-25 mglkglday, no limb malformations were observed in the offspring (Hummler et al., '90). Thus, retinoic acid seems capable of inducing a variety of limb reductions andlor duplications in vertebrates; the absent preaxial digit found in our case 2 has been commonly reported, while the degree of the scapular and clavicular reduction defects found in our case 1 have not, Several hypotheses have implicated different cell populations and processes in the pathogenesis of these experimentally induced limb reduction defects. Kochhar ('77) found no evidence that the cells of the apical ectodermal ridge (AER) were affected when all-trans-retinoic acid was given on day 11 or later. He implicated the underlying mesenchymal cell population as the susceptible cell population. In contrast, Sulik and De-
RETINOIC ACID-INDUCED LIMB MALFORMATIONS
Hart 088) reported excessive cell death of AER cells following an earlier isotretinoin exposure on day 9.5 in mice. They concluded that excessive cell death in the AER was the underlying abnormal developmental phenomenon, but that this may occur in concert with adverse effects on mesenchyma1 cells. Recent in situ hybridization studies using probes to mRNA transcripts of the retinoic acid receptor gamma showed that it was uniformly distributed throughout the limb bud mesenchyme from days 9.5 to 11.5, but not in the AER (Ruberte et al., '90). By day 12.5, the transcripts were more restricted to the precartilaginous mesenchymal condensations of the scapula, humerus, and proximal forelimb bones. In the related situation of the regenerating amphibian limb, it appears that retinoic acid acts on the mesodermal cells of the blastema, not the AER, to alter pattern formation of the regenerated limb (Maden, '84). Tickle and co-workers ('89) have also shown that, although retinoic acid can induce reorganization of the chick limb bud AER, the effects are mediated by the mesenchymal cells, not the AER cells. Thus, assuming that many of teratogenic efforts of retinoic are receptor mediated through its currently recognized receptors, it seems likely that most of the effects on limb development will result from primary influences on its mesenchyme. In summary, we report a fetus and infant with different limb reduction defects who were exposed to isotretinoin during early development. Case 2 had additional malformations characteristic of the retinoic acid embryopathy and the limb malformation, absent thumb, has been frequently reported in many experimental studies. Isotretinoin probably caused this limb reduction defect. For case 1,however, a causal connection is more uncertain. The absent scapula and clavicle of case 1 is an exceptionally rare human malformation. While it has not been reported in experimental studies, less severe malformations of the scapula have been seen. While case 1had additional malformations found in retinoic acid embyropathy, they were less characteristic than those of case 2, and the supraumbilical raphe-sternal defect has not been reported before. While our case reports suggest that isotretinoin causes limb malformations, they do not provide conclusive evidence of human teratogenicity.
603
ACKNOWLEDGMENTS
The authors acknowledge the able assistance of Jacqueline Ellis in preparing the manuscript. Thanks to Dr. Joe Rutledge for bringing case 2 to our attention. LITERATURE CITED Cusic A.M., and C.P. Dagg (1985) Spontaneous and retinoic acid-induced postaxial polydactyly in mice. Teratology, 31:49-59. Fantel A.G., T.H. Shepard, L.L. Newell-Morris, and B.C. Moffett (1976) Teratogenic effects of retinoic acid in pigtail monkeys (Macaca nemestrina). 1. General features. Teratology, 15.65-72. Fraser B.A., and A.A. Travill(1978)The effect of retinoic acid on chondrogenesis in the fetal hamster tibia in vivo. J . Embryol. Exp. Morph. 48:23-35. Gotlieb A., A. Hanukoglu, D. Fried, and M. Rosen (1982) Micrognathia associated with asternia and telangiectatic skin lesion. Syndrome Ident. 8:lO-12. Grote W., D. Harms, U. Janig, H. Kietzmann, U. Ravens, and I. Schwarze (1985) Malformation of fetus conceived 4 months after termination of maternal etretinate treatment. Lancet, i:1276. Happle R., H. Traupe, Y., Bounameaux, and T. Fisch (1984) Teratogene wirkung von etretinat beim menschen. Dtsch. Med. Wschr. 109,1476-1480. Hersh J.H. (1985) Sternal malformationhascular dysplasia association. Am. J. Med. Genet., 21:177-186. Hummler H., R. Korte, A.G. Hendrickx (1990) Induction of malformations in the cynomolgus monkey with 13-cis retinoic-acid. Teratology, 42:263-272. Kochhar D.M. (1967) Teratogenic activity of retinoic acid. Acta Pathol. Microbiol. Scand., 70:398-404. Kochhar D.M. (1977) Cellular basis of congenital limb deformity induced in mice by vitamin A. Birth Defects X111f1):111-154. Lammer E.J. (1987) Patterns of malformation among fetuses and infants exposed to retinoic acid (isotretinoin). In: J.A. McLachlan, R.M. Pratt, and C.L. Markert eds. Banbury Report 26: Developmental Toxicology: Mechanisms and Risk. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY, pp. 243-255. Lammer E.J., D.T. Chen, R.M. Hoar, N.D. Agnish, P.J. Benke, J.T. Braun, C.J. Curry, P.M. Fernhoff, A.W. Grix, I.T. Lott, J.M. Richard, S.C. Sun (1985a). Retinoic acid embryopathy. N. Engl J Med 313:837-841. Lammer E.J., D.B. Flannery, and M. Barr (198513) Does isotretinoin cause limb reduction defects? Lancet, ii: 328. Leiber B. (1982) Angeborene supraumbilikale mittelbauchrhaphe (SMBR) und kavernose gesichtshemangiomatose: Ein neues syndrom? Monatsschr. Kinderheilkunde, 130:84-90. Maden M. (1984) Does vitamin A act on pattern formation via the epidermis or the mesoderm? J . Exp. Zool., 230:387. McBride W.G. (1985) Limb reduction deformities in child exposed to isotretinoin in utero on gestation days 26-40 only. Lancet i:1276. Ruberte E., P. Dolle, A. Kurst, A. Zelent, G. MorrissKay, and P. Chambon (1990) Specific spatial and temporal distribution of retinoic acid receptor gamma transcripts during mouse embryogenesis. Development 108:213-222. Shepard T.H., M. Shi, G.W. Fellingham, M. Fujinaga, J.M. FitzSimmons, A.G. Fantel, and M. Barr (1988)
604
R. RIZZO ET AL.
Organ weight standards for human fetuses. Pediatr. Pathol. 8.513-524. Shenefelt R.E. (1970) Morphogenesis of malformations in hamsters caused by retinoic acid: relation to dose and stage a t treatment. Teratology, 5t103-118. Sulik K.K., and D.B. Dehart (1988) Retinoic acid-induced limb malformations resulting from apical ectodermal ridge cell death. Teratology, 37r527-537. Temtamy S., and V. McKusick (1978) The Genetics of Hand Malformations. Alan R. Liss, Inc. New York. Thaller C., and G. Eichele (1987) Identification and spatial distibution of retinoids in the developing chick limb bud. Nature, 327t625-628. Tickle C., B. Alberts, L. Wolpert and J. Lee (1982)Local application of retinoic acid to the limb bud mimics
the action of the polarizing region. Nature, 296.564566. Tickle C., A. Crawley, and J. Farrar (1989) Retinoic acid applications to chick wing buds leads to a dosedependent reorganization of the apical ectodermal ridge that is mediated by the mesenchyme. Development, I06:691-705. Webster W.S., M.C. Johnston, E.J. Lammer and K.K. Sulik (1986) Isotretinoin embryopathy and the cranial neural crest: an in vivo and in vitro study. J. Craniofac. Genet. Dev. Biol. 6:211-222. Wood V.E. and L. Marchinski (1990) Congenital anomalies of the shoulder. In: The Shoulder. C.A. Rockwood, and F.A. Matsen (eds.) W.B. Saunders, pp. 98148.
Limb Reduction Defects in Humans Associated With Prenatal lsotretinoin Exposure RENATA RTXXO. EDWARD J. LAMMER. ENRICO PARANO. ~~
~
~
~
~~~~
LORENZO PAVONE, AND J. CRAIG ARGYLE Clinzca Pediatrica, Uniuersitu di Catania, Catania, Italy 95125 (R.R.,EP., L.P.); California Birth Defects Monitoring Program, Emeryuille, California 94608 (E.J.L.): and Department of Pathology, Wilford Hall Medical Center, Lackland AFB, Texas 78236 (J.C:A.)
ABSTRACT Retinoic acid has long been used to induce limb reductions defects in experimental animal studies. No limb malformations, however, have been reported among malformed retinoic acid-exposed human fetuses from case reports o r epidemiologic studies. We report a child and a fetus with limb reduction malformations following maternal use of isotretinoin (13-cis-retinoic acid) during the first trimester of pregnancy. The child had a unilaterally absent clavicle and nearly absent scapula, with a short humerus and short, synostotic forearm bones. He also had ventriculomegaly and developmental delay, minor dysmorphic facial features, and a short sternum with a sterno-umbilical raphe. The fetus had a unilaterally absent thumb with normal proximal bony structures. Other findings included hydrocephalus, craniofacial anomalies, thymic agenesis, supracristal ventricular septa1 defect, single umbilical artery, anal and vaginal atresia, and urethral agenesis with dysplastic, multicystic kidneys. Although the limb malformations were quite dissimilar, a number of anomalies that are frequently found among isotretinoin-exposed fetuseslinfants were present in both cases. This increases the probability that retinoic acid caused these limb defects, but a causal association cannot be conclusively drawn on the basis of these two retrospective case reports. Case 1 Retinoic acid is known to be teratogenic in humans (Lammer et al., '85a) and a numC.A. was born to a 17-year-old mother ber of experimental animal species (Shene- and 24-year-old father after a full-term unfelt, '70; Webster et al., '86; Fantel et complicated pregnancy. The family history al., '76). In humans, the use of isotretinoin was unremarkable and the parents were not (13-cis-retinoic acid) for the treatment of consanguineous. He was their first child. severe nodulocystic acne has led to many During the first month after conception the inadvertent embryonic and fetal exposures. mother took 50 mglday of isotretinoin for These exposures have induced malforma- acne. She took isotretinoin for 10 days, but tions principally affecting craniofacial, cen- the gestational day of the final dose of tral nervous system, cardiac, and thymic isotretinoin is uncertain. Sometime during development. Although retinoic acid-in- the first two months of gestation, she also duced limb defects have been studied exper- had a fever for three days that was not imentally as a model system for limb dys- treated. The mother reported that she felt morphogenesis, no human fetus or infant fetal movements regularly. Labor was sponwith primary limb reduction defects follow- taneous and the vaginal delivery was uning exposure to isotretinoin has been re- complicated. Birthweight was 3,500 gm ported. We report a child and a fetus with (50th percentile), length was 50.5 cm (50th prenatal exposure to isotretinoin who were percentile), and head circumference (OFC) delivered with limb reduction defects and who had other abnormalities that have been reported previously in the retinoic acid emReceived November 5 , 1990; accepted August 14, 1991 bryopathy . 0 1991 WILEY-LISS, INC.
600
R. RIZZO ET AL.
Fig. 1. Case 1. Note short, internally rotated humerus attached to thorax, and sterno-umbilical raphe.
Fig. 2. Case 1. Absence of right clavicle. Small anomalous bone (arrow) posterior to right ribs is presumably a markedly underdeveloped scapula.
was 34.5 cm (50th percentile). Several anomalies were present a t birth, but the neonatal period was uneventful. He was clavicle, and a short humerus that directly able to hold his head erect at 5 months and attached at the thorax (Fig. 2). The right sit alone at 7 months. At 7 months, he was radius and ulna were also short, broad, and admitted at the Pediatric Clinic of Catania completely synostotic (Fig. 3). Digital subtraction arteriography demonstrated that for a febrile convulsion. At 7 months, his weight was 8,400 gm the right subclavian artery terminated just (25th percentile), length was 81 cm (75th distal to the origin of the internal mampercentile), and OFC was 48.5 cm (90th per- mary artery, which was enlarged. A large centile). His head was dolichocephalic with right superior (supreme) intercostal artery a prominent forehead. His ears were large entered the right upper limb and apparently and round. The nose was short with a provided its arterial supply. The seizures have not recurred. At age 2.5 rounded tip and the philtrum was broad. There was mild micrognathia. The neck was years, his development is delayed; he has long. There was marked thoracic asymme- just begun walking and speaks only a few try with hypoplasia of the right upper chest, syllables. shoulder girdle, and limb. The right clavicle Case 2 and scapula were not palpable. The short, An 18-year-old GI Po woman conceived internally rotated right humerus was attached to the thorax a t the level of the 4th- while taking 60 mglday of isotretinoin for 5th ribs (Fig. 1). Both radius and ulna were cystic acne. She used it for 62 days after short. The fingers of the right hand were conception. The pregnancy was otherwise shorter than the left and had camptodac- unremarkable until hydrocephalus and a tyly. The sternum was short. A visible, cystic kidney were prenatally diagnosed by raised fibrous cord extended for 4 cm be- ultrasonography. The pregnancy was termitween the umbilicus and the xiphoid (Fig. nated a t 22 weeks. The placenta was small, 1).The right testis was undescended. 92 gm (normal 120-230 gm), and there was The routine laboratory findings were a single umbilical artery. There were no normal: a urine metabolic screen, amino ac- other gross or microscopic abnormalities of ids, lactate, pyruvate and CPK. An ophthal- the placenta. mological examination, electrocardiogram, Fetal growth was normal, weight was 425 electroencephalogram, and intravenous py- gm (50th percentile), crown-rump length elogram were also normal. A high-resolu- was 19.0 cm (50th percentile at this weight), tion karyotype showed normal 46,XY chro- and length was 26.0 cm. The head was mosomes. Brain MRI and CT scans revealed large, and the external ears were small, asymmetrically enlarged ventricles. Skele- malformed, and low-set (Fig. 4). The mantal x-rays showed nearly complete absence dible was small. The brain could not be exof the right scapula, agenesis of the right amined microscopically. The eyes were nor-
RETINOIC ACID-INDUCED LIMB MALFORMATIONS
601
Fig. 5. Case 2. Absent left thumb.
Fig. 3. Case 1. Complete synostosis of short right radius and ulna. Camptodactyly of fingers.
Fig. 4. Case 2. Enlarged fetal head due to hydrocephalus and malformed, lowset external ear.
ma1 grossly and microscopically. Examination of the thoracic cavity revealed no thymus and microscopic sections of the superior mediastinum did not reveal any thymic tissue. A small, cystic, mediastinal structure lined by squamous, respiratorytype epithelium was found. The lungs were small for birthweight and weighed 6.3 gm (5th percentile; Shepard et al., '88). An aberrant right upper lobe bronchus originated above the carina. The right lung consisted of a lower lobe and a fused upper and middle
lobe. The left lung had no fissures. The heart weighed 2.9 gm, the foramen ovale was patent, and a ventricular septa1 defect (2 mm) was found just inferior to the pulmonic valve. The external genitalia were ambiguous and no external vaginal opening or anus were present. Within the abdominal cavity, the distal colon ended as a dilated pouch. A normal uterus and left fallopian tube and ovary were identified, but no adnexal structures were found on the right. The left kidney was large (2.1 cm) and multicystic. The right kidney (1.2 cm) was medially displaced and multicystic. Ureters and a bladder were found, but no urethra. Microscopically, the dysplastic kidney contained multiple small parenchymal cysts consistent with renal dysplasia secondary to obstruction. The left adrenal gland was smaller than the right but had a normal microscopic appearance. The left thumb and thenar eminence were absent (Fig. 5). Whole body radiographs revealed no other ossseous anomalies. A normal left radius and ulna were present. The temporal bones were markedly dysplastic. DISCUSSION
This fetus and infant represent the first reports of primary limb reduction defects following isotretinoin during early pregnancy. McBride ('85) previously reported a child with terminal transverse limb reductions of all four extremities whose mother had taken isotretinoin from days 26 to 40 of gestation. The weight of evidence, however, indicated that these limb reductions occurred secondary to amniotic band amputations rather than as primary malformations (Lammer et al., '85b). Limb reduction defects have been reported following exposure
602
R. RIZZO ET AL.
to another synthetic retinoid, etretinate (Happle et al., '84). These malformations included two cases with a short thumb, absent finger, syndactyly, and a forearm defect. In another report, a malformed fetus was delivered of a woman who stopped etretinate therapy 4 months before conception but had detectable serum levels of etretinate and etretin when the pregnancy was terminated a t 10 weeks (Grote et al., '85). That fetus had a hypoplastic femur, absent tibia and fibula, and a single toe. Because limb reduction defects have not been found previously among case reports or systematic epidemiologic studies of isotretinoin exposure, is i t likely that the limb reduction defects of our report were caused by isotretinoin? For both cases, a number of anomalies were present that involve the structures most commonly affected in the retinoic acid embryopathy (Lammer et al., '85a; Lammer, '87). Case 1had ventriculomegaly, developmental delay, and minor craniofacial anomalies (rounded ears, short nose, micrognathia); case 2 had hydrocephalus, malformed ears, and DiGeorge sequence. It is uncommon, however, to find the additional anomalies reported for case 2: anal atresia, distal vaginal atresia, and urinary tract obstruction. The supraumbilical raphe with sternal defect found for case 1 has not been reported in retinoic acid embryopathy. This unusual malformation has been reported in association with large facial hemangiomas, but rarely with other malformations (Hersh, '85; Leiber, '82; Gotlieb e t al., '82). Our case lacked any hemangiomas, however, and does not represent this recently recognized syndrome of unknown etiology. The shoulder malformation of patient 1 is exceptionally rare and does not occur a s part of any Mendelian disorder that affects limb development (Temtamy and McKusick, '78). Absence of the scapula has been reported only when the distal limb is completely absent (Wood and Marchinski, '90). In the thalidomide embryopathy, the glenoid fossa was occasionally malformed in association with more distal limb malformations, but no scapular anomaly like that of patient 1 was reported. We did not identify any cytogenetic abnormalities or environmental exposures, other than isotretinoin, that might have caused the malformations of our two cases. In addition, the timing of the isotretinoin exposure for case 2 extended well into the period of limb develop-
ment, while the interval of exposure for case 1 was uncertain. How do these two types of limb reduction defects compare with the patterns of limb anomalies induced by retinoic acid in experimental animals? In the experimental setting, retinoic acid may cause limb duplications or reductions (Cusic and Dagg, '85; Sulik and Dehart, '88; Kochhar, '67; Tickle et al., '82) and the types of malformation depend upon the timing, dose, and duration of exposure. Kochhar ('77) and others have induced proximal upper limb anomalies by exposing mice to all-trans-retinoic acid on day 11 and later (Fraser and Travill, '78). Although the scapula is sometimes reduced to a small triangular bone in these experimental studies, nearly complete absence of the scapula and clavicle a s in our case 1has not been described. By treating a t earlier times (day 9.5) with 13-cis-retinoicacid, and with higher doses, more distal types of limb reductions were also induced (Sulik and Dehart, '88). These limb reductions included ectrodactyly of the preaxial andlor postaxial side. Preaxial and postaxial polydactyly also occurred. Absence of the first preaxial digit, as occurred in our case 2, was the most frequent distal reduction defect in some experimental studies (Kochhar, '77). In a pigtail monkey study of all-trans-retinoic acid (10 mglkg), Fantel et al. ('76) reported finding postaxial ectrodactyly as the most common limb anomaly, with several cases of syndactyly and one of bilateral short ulnas. In another primate study in which 41 pregnant cynomolgus monkeys were given isotretinoin in doses from 2-25 mglkglday, no limb malformations were observed in the offspring (Hummler et al., '90). Thus, retinoic acid seems capable of inducing a variety of limb reductions andlor duplications in vertebrates; the absent preaxial digit found in our case 2 has been commonly reported, while the degree of the scapular and clavicular reduction defects found in our case 1 have not, Several hypotheses have implicated different cell populations and processes in the pathogenesis of these experimentally induced limb reduction defects. Kochhar ('77) found no evidence that the cells of the apical ectodermal ridge (AER) were affected when all-trans-retinoic acid was given on day 11 or later. He implicated the underlying mesenchymal cell population as the susceptible cell population. In contrast, Sulik and De-
RETINOIC ACID-INDUCED LIMB MALFORMATIONS
Hart 088) reported excessive cell death of AER cells following an earlier isotretinoin exposure on day 9.5 in mice. They concluded that excessive cell death in the AER was the underlying abnormal developmental phenomenon, but that this may occur in concert with adverse effects on mesenchyma1 cells. Recent in situ hybridization studies using probes to mRNA transcripts of the retinoic acid receptor gamma showed that it was uniformly distributed throughout the limb bud mesenchyme from days 9.5 to 11.5, but not in the AER (Ruberte et al., '90). By day 12.5, the transcripts were more restricted to the precartilaginous mesenchymal condensations of the scapula, humerus, and proximal forelimb bones. In the related situation of the regenerating amphibian limb, it appears that retinoic acid acts on the mesodermal cells of the blastema, not the AER, to alter pattern formation of the regenerated limb (Maden, '84). Tickle and co-workers ('89) have also shown that, although retinoic acid can induce reorganization of the chick limb bud AER, the effects are mediated by the mesenchymal cells, not the AER cells. Thus, assuming that many of teratogenic efforts of retinoic are receptor mediated through its currently recognized receptors, it seems likely that most of the effects on limb development will result from primary influences on its mesenchyme. In summary, we report a fetus and infant with different limb reduction defects who were exposed to isotretinoin during early development. Case 2 had additional malformations characteristic of the retinoic acid embryopathy and the limb malformation, absent thumb, has been frequently reported in many experimental studies. Isotretinoin probably caused this limb reduction defect. For case 1,however, a causal connection is more uncertain. The absent scapula and clavicle of case 1 is an exceptionally rare human malformation. While it has not been reported in experimental studies, less severe malformations of the scapula have been seen. While case 1had additional malformations found in retinoic acid embyropathy, they were less characteristic than those of case 2, and the supraumbilical raphe-sternal defect has not been reported before. While our case reports suggest that isotretinoin causes limb malformations, they do not provide conclusive evidence of human teratogenicity.
603
ACKNOWLEDGMENTS
The authors acknowledge the able assistance of Jacqueline Ellis in preparing the manuscript. Thanks to Dr. Joe Rutledge for bringing case 2 to our attention. LITERATURE CITED Cusic A.M., and C.P. Dagg (1985) Spontaneous and retinoic acid-induced postaxial polydactyly in mice. Teratology, 31:49-59. Fantel A.G., T.H. Shepard, L.L. Newell-Morris, and B.C. Moffett (1976) Teratogenic effects of retinoic acid in pigtail monkeys (Macaca nemestrina). 1. General features. Teratology, 15.65-72. Fraser B.A., and A.A. Travill(1978)The effect of retinoic acid on chondrogenesis in the fetal hamster tibia in vivo. J . Embryol. Exp. Morph. 48:23-35. Gotlieb A., A. Hanukoglu, D. Fried, and M. Rosen (1982) Micrognathia associated with asternia and telangiectatic skin lesion. Syndrome Ident. 8:lO-12. Grote W., D. Harms, U. Janig, H. Kietzmann, U. Ravens, and I. Schwarze (1985) Malformation of fetus conceived 4 months after termination of maternal etretinate treatment. Lancet, i:1276. Happle R., H. Traupe, Y., Bounameaux, and T. Fisch (1984) Teratogene wirkung von etretinat beim menschen. Dtsch. Med. Wschr. 109,1476-1480. Hersh J.H. (1985) Sternal malformationhascular dysplasia association. Am. J. Med. Genet., 21:177-186. Hummler H., R. Korte, A.G. Hendrickx (1990) Induction of malformations in the cynomolgus monkey with 13-cis retinoic-acid. Teratology, 42:263-272. Kochhar D.M. (1967) Teratogenic activity of retinoic acid. Acta Pathol. Microbiol. Scand., 70:398-404. Kochhar D.M. (1977) Cellular basis of congenital limb deformity induced in mice by vitamin A. Birth Defects X111f1):111-154. Lammer E.J. (1987) Patterns of malformation among fetuses and infants exposed to retinoic acid (isotretinoin). In: J.A. McLachlan, R.M. Pratt, and C.L. Markert eds. Banbury Report 26: Developmental Toxicology: Mechanisms and Risk. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY, pp. 243-255. Lammer E.J., D.T. Chen, R.M. Hoar, N.D. Agnish, P.J. Benke, J.T. Braun, C.J. Curry, P.M. Fernhoff, A.W. Grix, I.T. Lott, J.M. Richard, S.C. Sun (1985a). Retinoic acid embryopathy. N. Engl J Med 313:837-841. Lammer E.J., D.B. Flannery, and M. Barr (198513) Does isotretinoin cause limb reduction defects? Lancet, ii: 328. Leiber B. (1982) Angeborene supraumbilikale mittelbauchrhaphe (SMBR) und kavernose gesichtshemangiomatose: Ein neues syndrom? Monatsschr. Kinderheilkunde, 130:84-90. Maden M. (1984) Does vitamin A act on pattern formation via the epidermis or the mesoderm? J . Exp. Zool., 230:387. McBride W.G. (1985) Limb reduction deformities in child exposed to isotretinoin in utero on gestation days 26-40 only. Lancet i:1276. Ruberte E., P. Dolle, A. Kurst, A. Zelent, G. MorrissKay, and P. Chambon (1990) Specific spatial and temporal distribution of retinoic acid receptor gamma transcripts during mouse embryogenesis. Development 108:213-222. Shepard T.H., M. Shi, G.W. Fellingham, M. Fujinaga, J.M. FitzSimmons, A.G. Fantel, and M. Barr (1988)
604
R. RIZZO ET AL.
Organ weight standards for human fetuses. Pediatr. Pathol. 8.513-524. Shenefelt R.E. (1970) Morphogenesis of malformations in hamsters caused by retinoic acid: relation to dose and stage a t treatment. Teratology, 5t103-118. Sulik K.K., and D.B. Dehart (1988) Retinoic acid-induced limb malformations resulting from apical ectodermal ridge cell death. Teratology, 37r527-537. Temtamy S., and V. McKusick (1978) The Genetics of Hand Malformations. Alan R. Liss, Inc. New York. Thaller C., and G. Eichele (1987) Identification and spatial distibution of retinoids in the developing chick limb bud. Nature, 327t625-628. Tickle C., B. Alberts, L. Wolpert and J. Lee (1982)Local application of retinoic acid to the limb bud mimics
the action of the polarizing region. Nature, 296.564566. Tickle C., A. Crawley, and J. Farrar (1989) Retinoic acid applications to chick wing buds leads to a dosedependent reorganization of the apical ectodermal ridge that is mediated by the mesenchyme. Development, I06:691-705. Webster W.S., M.C. Johnston, E.J. Lammer and K.K. Sulik (1986) Isotretinoin embryopathy and the cranial neural crest: an in vivo and in vitro study. J. Craniofac. Genet. Dev. Biol. 6:211-222. Wood V.E. and L. Marchinski (1990) Congenital anomalies of the shoulder. In: The Shoulder. C.A. Rockwood, and F.A. Matsen (eds.) W.B. Saunders, pp. 98148.
