PREGNANCY COMPLICATIONS IN TYPE IV EHLERS-DANLOS SYNDROME

50

termination in

Occasional

Survey

pregnancies

early gestation, at the very least, all such be considered at high risk of complications.

must

PATIENTS AND METHODS

PREGNANCY COMPLICATIONS IN TYPE IV EHLERS-DANLOS SYNDROME NOREEN L. RUDD KAREN A. HOLBROOK

CARL NIMROD PETER H. BYERS

Department of Pediatrics, Division of Medical Genetics, Alberta Children’s Hospital, Calgary, Alberta, Canada, and the Departments of Pathology, Biological Structure, and Medicine, University of Washington, Seattle, Washington, U.S.A.

Type IV Ehlers-Danlos syndrome (EDS IV), unlike other forms of EDS, appears to be associated with a high incidence of pregnancy complications. In a group of fourteen families 20 women with EDS IV were identified. The diagnosis was confirmed in at least 1 member of each family by in-vitro measurement of type III collagen production by dermal fibroblasts, and all affected subjects produced lower levels of the protein than controls. Of the 20 women identified, 10 had been pregnant, and 5 had died from pregnancy-related complications. The overall risk of death in each pregnancy in this group was 25%. The complications of pregnancy included rupture of bowel, aorta, vena cava, or uterus, vaginal laceration; and post-partum uterine haemorrhage. The severity and frequency of the complications in this type of EDS warrant careful counselling before pregnancy and care of all pregnant patients in a highrisk facility. Summary

During the past 5 years we (P.H.B., K.A.H.) have served as a diagnostic centre for inherited connective-tissue disorders. We have identified fourteen families (34 individuals) with type IV EDS. We have seen all available patients from eleven families, and extensive medical records and necropsy data were provided on the remaining patients. Tissue samples for fibroblast culture and for examination by electron microscopy were obtained from 24 patients, and at least 1 family member in each known family was subjected to biopsy and studied. 29 of the patients belonged to nine families in which inheritance was autosomal dominant. The remaining 5 were the only affected members of their families. All patients studied had abnormalities in production of type III procollagen-alterations in synthesis, in stability of assembled molecules, or in its secretion. There were 20 female and 14 male patients; 9 and 8, respectively, were still living at the time the families were identified. 1 female patient has died since then. No patient presented because of

problems due to pregnancy. At the time of this report 10 of the 20 women had had 20 2 pregnancies had been interrupted by

pregnancies (see table).

COMPLICATIONS OF PREGNANCY IN EDS TYPE IV

INTRODUCTION

THE Ehlers-Danlos syndrome (EDS: hyperelastosis cutis) is a clinically, genetically, and biochemically heterogeneous group of inherited connective-tissue disorders: 1-8 clinical and biochemical studies have been used to segregate nine or more distinct types.Although extraordinary skin extensibility, initially described by Van Meekeren,9 has been the clinical hallmark of EDS, physicians have become increasingly aware of the generalised nature of the connective-tissue abnormalities. Patients with type IV EDS have translucent or transparent thin skin, through which the venous pattern is readily seen; only slight joint hyperextensibility, which is often limited to the hands; fragile skin which feels normal; and few difficulties with scar formation. They may die from arterial and bowel rupture.1,6-8,1O,11 This variety of EDS has been studied biochemically, and the phenotype can result from several disorders in production of type III collagen. These defects include absent or decreased synthesislz-14 and abnormal secretion.",15,16 Although type III collagen is ubiquitously distributed in connective tissues, the gastrointestinal tract, vessel walls, and the uterus are particularly rich in it. Even though the complications of spontaneous rupture of bowel and major vessels have been well recognised in EDSIV, there has been little appreciation of the high frequency of pregnancy complications with lethal outcome for both the affected mother and her child. We now review the pregnancy experiences of 2 patients in detail, those of 8 others made known to us during review of family histories of patients whose cells and tissues we examined to establish the diagnosis of EDS IV, and of a small number of patients previously reported by other workers. Complications of pregnancy in women with EDS IV are severe enough to consider

termination: 1 (case 10) because of concern about inheritance of EDS IV and possible complications of pregnancy ; the second (case 1) because of the ruptured colon at 8 weeks’ gestation. 12 of the 16 live-born infants had EDS IV. 2 infants died during complications of delivery; 1 was normal (case 1) and we did not determine whether the other was affected (case 2). The remaining infants were normal. The high proportion ofinfants born with EDS IV reflects the mode of ascertainment of these families. The 5 women in this series who died did so in the peripartum period (table). The detailed obstetric histories of those who died with documented uterine rupture are given below.

voluntary

Case 1 The patient

was a

33-year-old woman. She was born at 7 months’

gestation weighing 1130 g and had a unilateral clubfoot and hypoplastic leg. From infancy she had bruised easily. After 8 weeks’ gestation in a previous pregnancy at age 31 she had suffereda spontaneous rupture of her sigmoid colon. A therapeutic abortion was done at the time of operation. She was thin, 152 cm tall, with a triangular thin face and prominent eyes. She had mildly atrophic scars on her elbows, knees, and left eyelid from trauma and on her abdomen from surgery. She had thin skin through which the venous vasculature was visible. The skin over her knees and elbows was slightly hyperextensible. Joint hypermobility was limited to her digits, although one thumb was fixed in a subluxed position. Her hands were thin and had a

51

prematurely aged appearance. Her left foot was hypoplastic and the calf was tender. Several ecchymoses were present. She presented to the obstetrics service after 20 weeks of her second pregnancy with a painful haemorrhage into her left calf following minor ankle trauma. (A similar episode had occurred 3 years previously). At the time of examination tests of platelet count and function and of clotting factors were normal. Ultrasound studies confirmed gestational age and showed that the patient’s great vessels were of normal size; there valve

was no

evidence of mitral-

prolapse. An electrocardiogram had nonspecific ST-T wave

changes. The patient’s father had had episodes of bleeding which were hard to control and died of a ruptured viscus in his 30s; her mother died at

’

age 30 of unknown causes. No medical records were available. Her sister is apparently normal. The patient was readmitted at 28 weeks’ gestation in premature labour. Isoxsuprine infusion (24 mg/h) was started and labour ceased. 12 h later she began to complain of epigastric pain and nausea and her vomitus was blood tinged.’No source of bleeding was identified, although gastritis was suspected. No arterial abnormalities were detected by repeat ultrasound examination at this time. 32 h after admission, the obstetrician decided to let labour proceed. 4 h after isoxsuprine infusion was stopped the patient’s blood pressure was found to be 140/100 mm Hg. Shortly thereafter she suddenly collapsed and no pulse or blood pressure could be detected. Resuscitative measures failed. . At necropsy the tissues were unusually easy to dissect. The thoracic cavity contained approximately 1000 ml fresh blood and the abdominal cavity another 250 ml fresh blood. There was a large aortic dissection with additional accumulation of blood in the mediastinum. The thoracic aorta was torn transversely in two places (fig. 1). The first was a 1’ 5 cm tear 3 cm distal to the origin of the left

Fig. 2-Complete myometrial

tear

of left

posterolateral aspect

of

uterus of case 1.

Arrow points

to

fetal

fingers.

longer be heard. After stabilisation with fluids

an emergency section was begun. The abdomen was full of blood and amniotic fluid, and the fetus was free in the abdominal cavity having ruptured through the posterior wall of the uterus. All the tissues were extremely friable. The surgical report indicated. that "Constant oozing and bleeding was a serious problem, and it was difficult to identify tissues and place sutures... To our dismay, even the vaginal tissues were extremely friable, and placing an Oschner clamp on the vaginal wall, as ordinarily is done, cut through like butter ..." Despite attempted surgical repair and 7 units of blood, haemostasis could not be achieved and the patient died. caesarean

Tissue Culture and Electron Microscopy Dermal fibroblasts were grown from explants of skin of case 1 and her fetus at the time of death and from the affected son of case 2. Tissues from the same 3 subjects were fixed and examined by light and electron microscopy by procedures established previously. 16 The production of type III procollagen the cells in culture was used to confirm the diagnosis of EDS IV. 5 Cells in early passage were plated at a density of 2.5 5 x 105/35 mm culture dish and allowed to attach and spread overnight. They were then preincubated with medium which contained Dulbecco-Vogt modified Eagle medium (Gibco), 10% newborn calf serum, and 50 tg/ml ascorbic acid. After 4 h preincubation, the cells were

by

Fig. 1-Thoracic which

a

aorta

of

case

1

demonstrating

two tears between

dissection occurred.

incubated with 100 Ci/ml 2,3(3H]-proline (New England Nuclear, 25 Ci/mmol) for 24 h in the growth medium that contained 50 g/ml ascorbic acid but no serum. At the end of the labelling period the radiolabelled medium and cell-layer proteins were harvested separately in the presence of protease inhibitors and then examined by electrophoresis in 5% sodium dodecylsulphate polyacrylamide

gels. 18 subclavian artery; it involved the intima and media. There dissection between this and a second tear 1 cm above the diaphragm, which was virtually circumferential and involved all three layers of the vessel. The uterus had an irregular 4 cm tear in the left posterolateral region (fig. 2) which had extended completely through the myometrium. There was partial separation of the fundally implanted placenta. The fetus appeared normal. was a

Case 2 The patient was a 33-year-old woman in whom the diagnosis of EDS had been considered because of easy bruising. She had readily visible veins over her chest and varicosities of the veins of both legs. Two previous pregnancies had been uncomplicated. Her first child, through whom we identified this family when he was 26 years old, had EDS IV. The third pregnancy progressed normally and she entered labour without difficulty. During the third stage of labour contractions stopped abruptly, blood pressure fell suddenly, and uterine bleeding began. At that point the fetal heart tone could no

RESULTS AND DISCUSSION

The diagnosis of EDS IV was confirmed in our patients or their affected family members by measurement of production of type III procollagen by dermal fibroblasts in culture. Cells from case 1 produced a near-normal amount of procollagen, but almost all the secreted collagenous protein was type I procollagen and only a very small amount was type III procollagen (fig. 3). Skin from the patient was less than half the normal thickness and the collagen bundles were very small throughout the dermis (fig. 4). Both findings are characteristic of EDS IV. Cells from the fetus produced a normal amount of collagen and normal proportions of types I and III procollagen. Cells from the affected son of case 2 produced a normal amount of collagen but there was an abnormality in secretion of type III procollagen so that some protein accumulated intracellularly (not shown).

52

Fig. 3-Radioautofluorogram of JH-proline-labelled proteins secreted by control cells (a) and cells from case 1 (b). Virtually no proal(III) was seen in medium from the patient’s cells.

Although

previously appreciated, pregnancy among a very high rate of maternal our none of which were families, Among complication. identified because of pregnancy-related problems, 20 pregnancies resulting in 16 live-born children occurred in 10 affected women. No pregnancies have occurred in the remaining 10, most of whom are below child-bearing age. There were 2 voluntary terminations; 1 child died owing to maternal death (case 1); and another died during labour because of maternal uterine rupture (case 2). 12 children were women

not

with EDS IV has

affected with EDS IV which reflects the manner in which the patients were ascertained. 5 of the 10 women who had been pregnant died in the peripartum period (table). 1 died from uterine rupture during labour at term, 1 from uterine rupture and arterial rupture during premature labour, 2 from vascular rupture in the immediate post-partum period, and the 5th probably from uterine rupture during labour. The surviving 5 women accounted for the remaining 9 pregnancies. Other pregnancy complications in this group included perineal tears (1) and sustained post-partum bleeding (2). The age; parity, or underlying biochemical mechanism of the disease did not appear to predict survival in these patients. Major complications of pregnancy in EDS IV can occur in the antenatal period, during labour and delivery, and post 10,20 As illustrated by our case 1, the antenatal period partum. can be accompanied by rupture of bowel (probably unrelated to the pregnancy itself) and major vessels. The role of increased vascular volume and of hypertension in arterial rupture are not clear. Although bruising is common and venous varicosities are frequent, rupture of low-pressure vessels is uncommon; however, 1 patient who apparently had EDS IV died with a ruptured pulmonary artery during the 7th month of her 2nd pregnancy.20 The periods of greatest risk to women with EDS IV are labour and delivery and post partum. Extensive perineal tears, avulsion of the bladder, and prolapse of the bladder and uterus have been described among affected women. 21,22 Uterine haemorrhage in the immediate post-partum period seems to be quite frequent, and it is occasionally uncontrollable without hysterectomy. 1 patient in the series reported by Barabas’° died from aortic rupture in the immediate post-partum period after her fourth pregnancy. The precise incidence of pregnancy complications in women with EDS IV is as yet unknown and, because patients

Fig. 4-Sections of partial-thickness skin from and full-thickness skin from case 1 (b). Case 1 has

a

a

normal control (a)

substantial decrease in thickness of the dermis.

High

magnification segments of the reticular dermis show large collagen bundles bordered by elastic fibres in the control (c) contrasted with fine collagen bundles, thin elastic fibres, and large areas of interstitial space in case 1 (d). a

and b reduced by a third from

x

75; c and d reduced by a third from x 250.

with complications are more likely to be reported, difficult to determine from medical, genetic, and obstetric literature. Since none of our patients was selected because of problems in pregnancy, this survey of our families with EDS IV represents an estimate of the magnitude of pregnancy complications. The maternal mortality rate among women in our families is 25% (5 deaths in 20 pregnancies). Thus, from a group of patients not selected for pregnancy complications, the frequency of pregnancy-related death is higher, for example, than in the Marfan syndrome, another inherited connective-tissue disorder in which pregnancy-related deaths are known to occur. 23 Pregnancy in patients with EDS IV presents formidable problems to their physicians. We suggest that in women in whom the diagnosis of EDS IV is suspected on clinical grounds, every effort should be made to confirm or exclude the diagnosis biochemically. For women not yet pregnant, a careful discussion of the complications should be undertaken. The mode of inheritance should be considered since, if wished, prenatal diagnosis of the disorder may be feasible. Given such an opportunity, the patient and her family can participate in an informed decision about reproduction. For the patient in early pregnancy, similar considerations pertain and discussion about prenatal diagnosis and elective termination are appropriate. Neither our patient (case 1) nor another previously described 10 (case 11 in table) had

53

complications after elective termination before 16 weeks’ gestation. One was performed by suction evacuation, the other by hysterotomy. If the pregnant EDS IV patient is first seen late in pregnancy, she must be cared for as a high-risk patient. Vaginal delivery and delivery by elective caesarean section have both been advocated for such patients. However, on the basis of the experience of patients in our group there does not appear to be a mode which has fewer complications. We suggest that all patients should be carefully monitored during labour and for several days post partum so that prompt intervention

can

be initiated in the

event

of vessel

or

uterine

rupture. We cannot emphasise too strongly that EDS IV is a distinct type ofEDS which differs in clinical and biochemical features from all other forms. Those complications which affect women with EDS IV are not to be expected in patients with other forms of EDS.z4 We thank Dr John Butt, Chief Medical Examiner, Province of Alberta, for his cooperation in tissue collection for diagnostic studies; Mary Hoff and Karen David for excellent technical assistance; and Marion Brown and Diane Dawson for preparing the manuscript. This study was supported in part by funds from the Alberta Children’s Foundation and grants from the U.S. Public Health Service (AM 21557, GM 07266), the March of Dimes Birth Defects Foundation (6-298), and the Poncin Scholarship Fund. P. H. B. is an Established Investigator of the American Heart Association.

Correspondence should be addressed to P. H. B., Department of Pathology SM-30, University of Washington, Seattle WA 98195, U.S.A. REFERENCES

1. Beighton P. The Ehlers-Danlos syndrome. London: Heinemann, 1970. 2. McKusick VA. Heritable disorders of connective tissue, 4th ed. St Louis: Mosby, 1972. 3. Hollister DW. Heritable disorders of connective tissue: Ehlers-Danlos syndrome. Pediatr Clin N Am 1978; 25: 575-91. 4. Pinnell SR. Disorders of collagen. In: Stanbury JB, Wyngaarden JB, Frederickson DS, eds. The metabolic basis of inherited disease New York McGraw-Hill, 1978 1366-94. 5. Bernstein P, Byers PH. Disorders of collagen metabolism. In: Bondy PK, Rosenberg LE, eds. Metabolic control and disease, 8th ed. Philadelphia: WB Saunders, 1980: 1089-1153. 6. Hollister DW, Byers PH, Holbrook KA. Genetic disorders of collagen metabolism. Adv Hum Genet 1982; 12: 1-87. 7. Byers PH. Inherited disorders of collagen biosynthesis: Ehlers-Danlos syndrome, the Marfan syndrome and osteogenesis imperfecta. In: Spittel JA, ed. Clinical medicine.

Philadelphia: JB Lippincott (in press). 8. Byers PH, Barsh GS, Holbrook KA. Molecular mechanisms of connective tissue abnormalities in the Ehlers-Danlos syndrome. Collagen Rel Res 1981; 1: 475-89. 9. Van Meekeren JA. De dilatabilitate extraordinana cutis. In: Observations mediconchirugicae. Chap. 32, Amsterdam, 1682. Cited in McKusick VA. Heritable disorders of connective tissue, 4th ed. St Louis: Mosby, 1972. 10. Barabas AP. Vascular complications in Ehlers-Danlos syndrome. J Cardiovasc Surg 1972, 13: 160. Holbrook KA, McGillivray B, MacLeod PM, Lowry RB. Clinical and ultrastructural heterogeneity of type IV Ehlers-Danlos syndrome. Hum Genet 1979; 47: 141-50. 12. Pope FM, Martin GR, Lichtenstein JR, et al. Patients with Ehlers-Danlos syndrome type IV lack type III collagen. Proc Natl Acad Sci USA 1975; 72: 1314-16. 13. Pope FM, Martin GR, McKusick VA. Inheritance of Ehlers-Danlos type IV syndrome. J Med Genet 1977; 14: 200-04. 14. Aumailley M, Kneg T, Dessau W, Müller PK, Timpl R, Bricaud H. Biochemical and immunological studies of fibroblasts derived from a patient with Ehlers-Danlos syndrome type IV. Arch Dermatol Res 1980; 269: 169-77. 15.Byers PH, Holbrook KA, Barsh GS, Smith LT, Bornstein P. Altered secretion of type III procollagen in a form oftype IV Ehlers-Danlos syndrome: biochemical studies in cultivated fibroblasts. Lab Invest 1981; 44: 336-41. 16. Holbrook KA, Byers PH. Ultrastructural characteristics of the skin in a form of EhlersDanlos syndrome type IV: storage in the rough endoplasmic reticulum. Lab Invest 1981; 44: 342-50. 17. Bornstein P, Sage H Structurally distinct collagen types Annu Rev Biochem 1980; 49: 957-1014. 18. Laemmli UK Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 1970; 227: 680-85. 19. Pope FM, Jones PM, Wells RS, Lawrence D. EDS IV (acrogeria): new autosomal dominant and recessive types. J Roy Soc Med 1980, 73: 180-85. 20. Pearl W, Spicer M Ehler-Danlos syndrome. South Med J 1981; 74: 80-82. 21. Beighton P. Obstetric aspects of the Ehlers-Danlos syndrome Br J Obstet Gynaecol 1969; 76: 97-104. 22. Stoddard FJ, Myers RE. Connective tissue disorders in obstetrics and gynecology. Am J Obstet Gynecol 1968; 102: 240-52. 23 Pyeritz RE. Maternal and fetal complication of pregnancy in the Marfan syndrome. Am J Med 1981; 47: 785-90. 24. Taylor DJ, Wilcox I, Russell JK. Ehlers-Danlos syndrome during pregnancy: a case report and review of the literature. Obstet Gynecol Surv 1981; 36: 277 11.

Byers PH,

Industrial Medicine GENETIC SUSCEPTIBILITY TO SCLERODERMA-LIKE SYNDROME INDUCED BY VINYL CHLORIDE K. I. WELSH C. M. BLACK R. M. BERNSTEIN A. E. WALKER L. J. CATOGGIO A. R. MCGREGOR J. K. LLOYD JONES West Middlesex University Hospital, Isleworth, Middlesex; Royal National Hospital for Rheumatic Diseases, Bath; Tissue Typing

Laboratory, Guy’s Hospital, London, Department of Dermatology, Chesterfield Royal Hospital, Chesterfield; Hammersmith Hospital, Du Cane Road, London; and Department of Rheumatology, Harlow Wood Orthopaedic Hospital, Nr Mansfield, Nottinghamshire

Vinyl chloride (VC) monomer can induce a scleroderma-like syndrome in a proportion of workers exposed to it during production of polyvinyl chloride. As part of a 5-year follow-up study HLA A, B, and DR antigens and anti-centromere and anti-scleroderma-70

Summary

antibodies were determined in 44 such workers. 21 of these had severe and 23 mild forms of vinyl-chloride disease. 50 patients with "classical" scleroderma and 148 healthy hospital workers acted as controls. 11 of the 21 patients classified as having severe VC disease were DR3 positive, and 8 of these had both B8 and DR3 antigens. None of the 23 patients with mild disease carried either antigen. The HLAantigen frequencies in VC disease mirrored those found in scleroderma (raised DR5 frequency and increased linkage disequilibrium between B8 and DR3). There were, however, significant differences in the frequency of autoantibodies in the two conditions. INTRODUCTION

VINYL chloride (VC) was first polymerised to polyvinyl chloride (PVC) in Germany in the 1930s, but it was many years before reports of liver changes’ and acro-osteolysis2 in workers involved in PVC production began to appear. These and many subsequent reports3 indicate that exposure to vinyl chloride can lead to a syndrome characterised by sclerotic changes in the skin, skin nodules, clubbing of the fingers,

osteolysis, Raynaud’s phenomenon, thrombocytopenia, portal fibrosis and impaired hepatic function, and pulmonary fibrosis. Excessive fatigue, muscle and joint pains, central nervous system symptoms, and impotence have also been described. Although only some workers who had been in contact with vinyl chloride developed symptoms, no studies on genetic markers in those affected have been reported. Vinyl-chloride disease is, however, clinically very similar to scleroderma, and the observations by ourselves and others that susceptibility to scleroderma is linked, albeit weakly, to HLA markers4-6 led us to study 44 of the 53 patients with symptoms of vinyl-chloride disease originally described by Ward et al. These patients are men who had developed either mild or severe forms of the disease after exposure to vinyl chloride. They were studied, as part ofa 5-year follow-up, for HLA-region allotypes, Gm allotypes, complement allotypes, and autoantibodies to centromere, Scl-70, and collagen subtypes. In addition we were able to determine the number of workers from the same factories who had been exposed to vinyl chloride but had no symptoms of vinyl-chloride disease.