American J o u r n a l of Medical Genetics 36:333-335 (1990)
Frequency of Congenital Heart Defects in Patients With Hemophilia Kerry B. Jedele, Virginia V. Michels, Hymie Gordon, and Gerald S. Gilchrist Departments of Medical Genetics (K.B.J., V.V.M., H.G.) and Pediatrics and the Comprehensive Hemophilia Center (G.S.G), Mayo Clinic and Mayo Foundation, Rochester, Minnesota Most structural congenital heart defects (CHD) are thought to be multifactorially determined, but the precise causal factors usually are unknown. One may postulate that vascular events, such as hemorrhage in the developing embryo, could influence morphogenesis of the heart. One method of studying this hypothesis is to determine the frequency of CHD in persons with heritable bleeding diatheses and their families. We reviewed retrospectively medical and family histories of 120 hemophilia A and 14 hemophilia B patients seen in our Genetics Department. The family histories included 1,126 maternal relatives of hemophiliac patients. We also reviewed the family histories of 138 patients with X-linked disorders who did not have bleeding diatheses or syndromes associated with CHD; these histories included 960 maternal relatives. There was one confirmed case of a CHD in 134 hemophilia patients, giving a frequency of 0.75% compared to 0.8%in the general population at birth. There was no apparent difference in the frequency of CHD in hemophilia A and B patients compared to the general population or in the relatives of hemophilia patients as compared to control individuals.
KEY WORDS Bleeding diatheses, birth defects (heart), coagulopathy INTRODUCTION Congenital heart defects (CHD) occur in 8/1,000 births LMitchell et al., 19711. CHD can be caused by chromosome abnormalities, single gene defects, and teratogens such as drugs, viruses, and certain maternal diseases; but the specific cause of 90% of CHD is unknown [Nora and Nora, 19781.Most CHD are thought to be multifactorially determined, i.e., due to the interacReceived for publication August 4, 1989; revision received November 27, 1989. Address reprint requests to Virginia V. Michels, M.D., Department of Medical Genetics, Mayo Clinic, Rochester, MN 55905.
0 1990 Wiley-Liss, Inc.
tion of multiple genetic and equally unknown environmental agents. A case-control study has been reported in which there was a n excess of heritable coagulopathies and anemias in newborn infants with CHD and in their first-degree relatives [Ferencz et al., 19841.It is possible that coagulopathies might affect cardiac morphogenesis during embryogenesis by alterations in hemodynamics, abnormal formation of the vascular bed, hemorrhage, or some other mechanism. This altered cardiac morphogenesis could result in a n excess of CHD among patients with coagulopathies or among children born to women who have or are carriers for coagulopathies. To examine more specifically the relationship between 2 types of congenital coagulopathies (hemophilia A and B) and CHD, we performed a retrospective study of hemophilia A and B patients and their maternal relatives.
MATERIALS AND METHODS The medical and family histories of all 120 hemophilia A and 14 hemophilia B patients seen in the Department of Medical Genetics of the Mayo Clinic were reviewed. The family histories of these patients included 1,126 maternal first- and second-degree relatives. The medical and family histories of 138 non-hemophiliac patients and their 960 maternal first- and second-degree relatives were reviewed to form the control population. This control group consisted of patients seen in the Department of Medical Genetics for other conditions not associated with abnormal bleeding or CHD. Ofthese 138 control patients, 74 had X-linked diseases including ocular albinism (5), Becker muscular dystrophy (111, Duchenne muscular dystrophy (91, Fabry disease (171, Alport syndrome (7), adrenoleukodystrophy (12), X-linked immunodeficiency ( 5 ) , and Kallmann syndrome (8). Because of insufficient number of control individuals with definite X-linked disease not associated with coagulopathies or CHD, we added 64 control patients with hereditary motor and sensory neuropathy (Charcot-Marie-Tooth disease) which can be inherited as a n autosomal dominant or X-linked condition. These control individuals were chosen since the recording of their family histories would have been the same as for other X-linked disorders, with particular attention being paid to maternal relatives. The family histories were obtained from the patients or their parents in a uniform fashion by a staff genealo-
Jedele et al.
gist and reviewed by one of the authors. These family histories were recorded before the authors were aware of the reported association of CHD with hemophilia. Since our clinic is a referral center for cardiac surgery, the geographic origin of patients with CHD was determined to evaluate for possible selection bias. The frequency of CHD in the hemophilia patients was compared to that of the general population by using a one-sample binomial test (two-sided), and the frequency of CHD in maternal relatives was compared to that in relatives of our control individuals. RESULTS Of 134 patients with hemophilia A or B, only one was diagnosed with CHD (patient 1). A second patient with confirmed CHD but uncertain coagulation status was found (patient 2). He was the brother of a confirmed hemophilia A patient. In a family with hemophilia B, the son of a non-hemophiliac maternal uncle died after cardiac surgery. In the control population, only one maternal uncle with CHD was recorded. There was only one confirmed CHD case in 134 hemophiliac patients, giving a frequency of 0.75% (95%confidence intervals = 0.00%-4.1%). If patient 2 were included because of a clinical history of bleeding, the frequency of CHD would be 2 CHD cases in 134 hemophiliacs, giving a frequency of 1.50%, which remains statistically insignificant. The incidence of CHD in total births in the general population is 0.8% [Mitchell et al., 19711. In the maternal first- and second-degree relatives of hemophilia patients, 2 of 1,126 (0.18%) had CHD, which was not significantly different from the one of 960 (0.10%) of relatives of control patients. As expected, these frequencies are lower than is found in the general population, because the information is based only on family history.
CLINICAL REPORTS Patient 1 This patient was a 5,010 g male infant noted to be cyanotic a t birth. Cardiac catheterization showed transposition of the great arteries (TGA) and atrial septal defect (ASD). Oozing occurred both from the femoral vein cut-down site and the umbilical stump, and laboratory evaluation showed hemophilia A (factor VIII less than 1%). At age 16 months he died of gram-negative septic shock after heart surgery. The mother’s 2 maternal-half-brothers were known to have hemophilia A. There was no family history of CHD. Patient 2 The older brother of a confirmed hemophilia A patient was apparently healthy a t birth and underwent circumcision a t one day of age. At age 2 days, focal hemorrhage from the circumcision began, which continued until the bleeding site was cauterized at age 3 days. No other abnormal bleeding was noted. On the fourth day of life, a heart murmur was found and the infant had a sudden cardiac arrest. Autopsy showed atrial and visceral situs solitus, aortic valvular atresia, hypoplastic left heart with endocardia1 fibroelastosis, hypoplastic mitral
valve, hypoplastic aorta, patent foramen ovale and patent ductus arteriosus. No coagulation studies were obtained. There was no family history of CHD.
DISCUSSION Reports have been published of patients with hemophilia A and B with transposition of great arteries (TGA) [Roskos et al., 19831, hemophilia B with tricuspid atresia [Lawson et al., 19751,hemophilia A with double outlet right ventricle with a small ventricular septal defect [Leggett et al., 19841, hemophilia A with ostium primum defect and cleft mitral valve [Lusher et al., 19741, and congenital dysfibrinogenemia with tetralogy of Fallot [Branson et al., 19771. Patients with primary coagulopathies [Brockman et al., 1972; Meagher et al., 1981; Ma et al., 19791 (2 hemophilia A patients, one von Willebrand disease patient) who required aortic valve replacement with or without mitral valve replacement have also been reported; these patients presented a t 49, 49, and 62 years of age. One patient had a bicuspid aortic valve, but the cause of the valvular disease in the other patients was not specified and may have been a n acquired condition. In 1984 a population-based study of CHD and various hematologic disorders in infants and their first-degree relatives was claimed to show a n excess of coagulopathies and anemias but no hernoglobinopathies among the relatives of infants with CHD [Ferencz et al., 19841. However, of the 7 in 1,020 cases identified in that study with CHD and a blood disorder, 2 had “fetal anemias,” one had thrombocytopenia and spherocytosis due to trisomy 18, one had pancytopenia, and one had anemia and spherocytosis. Of the 2 patients with heritable coagulopathies, one had Williams syndrome and the mother had von Willebrand disease, and the other, a female, had a father with hemophilia [Ferencz e t al., 19841. Thus, only 2 in 1,020 cases with CHD may have had a heritable coagulopathy. Experimental animal data have led to speculations about a casual relationship between hemodynamic alterations in utero and abnormal cardiovascular development [Jaffee, 19781. Trypan blue has been demonstrated to cause hypervolemia and consequent abnormal formation of heart tubes suggestive of partial TGA and left heart hypoplasia in animals [Jaffee, 19781.Grabowski and Schroeder [19681 showed that hypoxia in chick embryos produced a n “edema syndrome” with severe hypervolemia, hematoma formation, and bradycardia. Increased resistance to vascular outflow in the developing heart in experimental models had produced ventricular septal defects [Jaffee, 19781. Therefore, one may speculate that abnormal coagulation in cardiac primordia may result in spontaneous bleeding that alters hemodynamic patterns and results in structural heart malformations. Abnormal coagulation with bleeding and hematoma formation may also result in vascular disruption to the developing heart and subsequent malformation, analogous to the demonstration of first and second branchial arch derivative abnormalities following localized hematoma formation in the mouse model of Goldenhar syndrome [Poswillo, 19733. In addition. endothelial cells have been demonstrated to moduce factor VIII and von Willebrand factor (vWF) [ker-
Congenital Heart Defects in Hemophilia Patients
encz et al., 19841. Perhaps the lack of endothelial production of factor VIII in hemophilia A patients affects the hemodynamics of the developing vascular bed in the embryonic heart. These mechanisms might lead one to expect a n increase of CHD in patients with underlying coagulopathies such as hemophilia A and B. Therefore, we studied this population to determine if these expectations were fulfilled by the finding of increased CHD in these patients. It is important to differentiate between primary coagulopathies in patients with CHD and coagulopathies secondary to CHD. There is extensive documentation that abnormal hemostasis of varying causes can be associated with both cyanotic and non-cyanotic CHD [ColonOtero et al., 19871. Frequent findings include thrombocytopenia [Ekert et al., 19701, prolonged bleeding time [Ekert et al., 19701, loss of vWF multimers [Gill et al., 19861, shortened fibrinogen half-life [Thies et al., 19821, decreased factor V levels LEkert et al., 19701, and deficiency of vitamin K-dependant clotting factors synthesized in the liver [Henriksson et al., 19791.The degree of hypoxia and hyperviscosity (polycythemia) appears to correlate with these hemostatic defects [Ekert et al., 1970; Thies et al., 1982; Suarez et al., 1984; Gross et al., 19681. Correction of the abnormal hemodynamic state may normalize acquired vWF abnormalities [Henriksson et al., 19791. By confining our study to hemophilias A and B, we avoided the confounding effects of these secondary coagulation defects and evaluated only primary coagulopathies that were present in utero. Our findings show no apparent difference in the frequency of CHD in our hemophilia A and B patients compared to the general population. However, the sample size of this study is too small to exclude up to a fivefold increase in the risk of CHD along hemophilia patients. We did not detect a n increase in the frequency of CHD in the families of hemophiliacs. The low numbers of CHD patients in the maternal first- and second-degree relatives of both the hemophilia and control populations are presumably owing to inaccurate information about the family histories; we have no reason to believe that bias existed in the eliciting of family histories of hemophiliac vs. control patients. Because CHD is a relatively common form of birth defect, it is not suprising that a n occasional child with hemophilia A or B might also have a CHD by chance.
REFERENCES Branson HE, Schmer G, Dillard DH (1977): Fibrinogen Seattle: A qualitatively abnormal fibrinogen in a patient with tetralogy of Fallot. Am J Clin Pathol 67:236-240. Brockman SK, April1 SN, Rabiner FS (1972):Aortic valve replacement in hemophilia: Report of a case. JAMA 222:660-661. Colon-Otero G, Gilchrist GS, Holcomb GR, Ilstrup DM, Bowie EJW (1987):Preoperative evaluation of hemostasis in patients with congenital heart disease. Mayo Clin Roc 62:379-388. Ekert H, Gilchrist GS, Stanton R, Hammond D (1970):Hemostasis in cyanotic congenital heart disease. J Pediatr 76221-230, Ferencz C, Rubin JD, McCarter R J , Wilson PD, Boughman JA, Brenner JI, Neil1 CA, Perry LW, Hepner SI, Downing J W (1984): Hematologic disorders and congenital cardiovascular malformations: Converging lines of research. J Med 15:337-354. Gill JC, Wilson AD, Endres-Brooks J , Montgomery RR (1986):Loss of the largest von Willebrand factor multimers from the plasma of patients with congenital cardiac defects. Blood 67:758-761. Grabowski CT, Schroeder RE (1968):A time-lapse photographic study of chick embryos exposed to teratogenic doses of hypoxia. J Embryo1 Exp Morphol 19:347-362. Gross S, Keefer V, Liebman J (19681: The platelets in cyanotic congenital heart disease. Pediatrics 42:651-658. Henriksson P, Varendh G, Lundstrom N-R (1979):Haemostatic defects in cyanotic congenital heart disease. Br Heart J 41:23-27. Jaffee OC (1978): Hemodynamics and cardiogenesis: The effects of physiologic factors on cardiac development. In Rosenquist GC, Bergsma D (eds):“Morphogenesis and Malformation of the Cardiovascular System.” New York: Alan R. Liss, Inc., for the National Foundation-March of Dimes. BD:OAS XIV (7):393-404. Lawson R, Rullman D, Brodeur M, Starr A (1975): Tricuspid atresia with Christmas disease (hemophilia B). J Thorac Cardiovasc Surg 69:585-588. Leggett PL, Doyle D, Smith WB, Culpepper W 111, Cooper S, Ochsner J L (1984): Elective cardiac operation in a patient with severe hemophilia and acquired factor VIII antibodies. J Thorac Cardiovasc Surg 87:556-560. Lusher JM, Ravindranath Y, Arciniegas E, Green E (1974):Open heart surgery in a hemophiliac patient. Am J Dis Child 127:708-711. Ma DDF, Chang VP, Concannon AJ, Lau J (1979):Aortic valve replacement in a patient with von Willebrands disease. Aust NZ J Surg 49:247-250. Meagher PD, Rickard KA, Richards JG, Baird DK (1981):Aortic and mitral valve replacement in a patient with severe haemophilia A. Aust N Z J Med 11:76-79. Mitchell SC, Korones SB, Berendes HW (1971): Congenital heart disease in 56,109 births. Circulation 43:323-322. Nora JJ, Nora AH (1978):The evolution of specific genetic and enviromental counselling in congenital heart disease. Circulation 57:205-213. Poswillo D (1973): The pathogenesis of the first and second branchial arch syndrome. Oral Surg 35:302-328. Roskos RR, Gilchrist GS, Kazmier FJ, Feldt RH, Danielson GK (1983): Management of hemophilia A and B during surgical correction of transposition of the great arteries. Mayo Clin Proe 52:182-186. Suarez CR, Menendez CE, Griffin AJ, Ow EP, Walenga JM, Fareed J (1984): Cyanotic congenital heart disease in children: Hemostatic disorders and relevance of molecular markers of hemostasis. Semin Thromb Hemost 10:285-289. Thies W-R, Gobel U, Liersch R (1982):Fibrinogen half-life in children with cyanotic congenital heart disease. Eur J Pediatr 139:43-47.