Clinical Genetics 1976: 9: 245-249
Inheritance of hypoplastic left heart syndrome (HLHS): Further observations* L. G. BROWNELLAND M. H. K. SHOKEIR** Department of Paediatrics, University of Manitoba and Children’s Centre, Winnipeg, Manitoba, Canada
From a survey of newborn infants using truncate ascertainment of families through affected offspring, 64 cases of hypoplastic left heart syndrome (HLHS) were found. Pedigree information was obtained from 42 families. The incidence of HLHS, based on 64 affected infants, was 36.5/100,000 births. The incidence of HLHS with aortic and/or mitral atresia was 25.1/100,000 births. The recurrence among later siblings was about 2 %. AutosomaI recessive hypothesis of transmission of this disorder was not supported. Using Edwards’ formula, the observed recurrence among later siblings was compatible with polygenic inheritance. Received 23 May, accepted for publication 20 June 1975
Hypoplastic left heart syndrome (HLHS), described first by Lev (1952), encompasses a broad spectrum of abnormalities of the left side of the heart ranging from hypoplasia of the aortic arch with an underdeveloped left ventricle, to more severe forms with atresia of the aorta, atretic aortic and mitral valves, hypoplastic left atrium and vestigial left ventricle (Lumb & Dawkins 1960, Noonan & Nadas 1958, Sinha et al. 1968). The pathology, pathophysiology, and natural history of this disorder have been outlined previously (Shokeir 1971). Until recently, little has been published concerning familial clustering of cases of HLHS and the possible genetic contribution to their etiology. Brekke (1953) reported a
* Supported by MRC of Canada. ** Queen Elizabeth I1 Scientist.
family in which two siblings suffered aortic atresia and hypoplasia of the aortic orifice. Ehlers & Engle (1966) described a sibship of three, with two affected infants. In 1966 Rao et al. published a report of two successive siblings with hypoplastic ascending aorta, atretic aortic valve, and small left ventricle and atrium. Kojima (1969) reported on a family in which two of four siblings were affected. Familial occurrence of HLHS was also described by Saied & Folger (1972). Kaufman et al. (1972) described a sibship of three girls; the first- and third-born had HLHS with vertebral and renal anomalies, while the second child had only vertebral and renal anomalies. In 1974, Bjornstad & Michalsen discussed the occurrence of coexistent mitral and aortic valve
246
BROWNELL AND SHOKEIR
atresia with intact ventricular septum in siblings. In a previous study of the possible genetic etiology of this syndrome, Shokeir (1971) suggested autosomal recessive transmission as the operative mode of inheritance. Subsequently, Holmes et al. (1972, 1974) invoked the concept of polygenic inheritance to explain the pattern of transmission. We, therefore, embarked on a comprehensive study of the disorder in a localized multiethnic urban community - Winnipeg. This setting was in marked contrast to that of the first study (Shokeir 1971) which focused on a rural population with considerable ethnocentric distribution and appreciable endogamy. Present lnvestigatlon
A survey of infants affected with HLHS, born in the city of Winnipeg between January 1957 and June 1974, was conducted through a chart and post mortem review in Winnipeg hospitals. This permitted ascertainment of cases diagnosed clinically (with confirmation by cardiac catheterization) andlor at necropsy. To allow for the variation in terminology used by cardiologists and pathologists in describing given entities within the syndrome, the boundaries of our search were extended initially to include most congenital heart diseases. Ultimately, however, only the 64 infants with hypoplastic left ventricle and with left-sided obstruction severe enough to result in death, were included. The presence and severity of aortic and mitral atresia varied, as did the presence of endocardial fibroelastosis. Four cases with multiple extracardiac lesions or chromosomal aberrations were excluded. Family pedigrees were obtained from 42 of the 64 families ascertained through affected offspring. The remaining families could not be contacted. No parental con-
sanguinity could be detected. The sex ratio was 1.37:l (M:F), conforming to the previously observed excess of affected males (Monie & DePape 1950, Noonan & Nadas 1958), which however, was not significantly different from a 1:l ratio (xZ(1) = 1.56 0.15 < P < 0.25). There was no evidence of an increased number of spontaneous abortions, nor was there a significant association with maternal drug ingestion, infection, Xirradiation, or history of diabetes. The data thus generated were used to estimate the incidence of the disorder in the population and the rate of recurrence among later siblings. Furthermore, hypotheses of modes of genetic transmission were examined. The incidence of HLHS, based on 64 affected infants, was 36.5/100,000 births (Table 1). The incidence of HLHS with aortic and/or mitral atresia was 25.11 100,000 births, which is comparable to the value 261100,000 births obtained by Shokeir (1974), who used analogous criteria for his survery. The incidence of aortic andlor mitral atresia associated either with HLHS or with other conditions (transposition of the great vessels, single ventricle, or chromosomal aberrations) was 29.0/100,000 births, which is comparable to Mitchell and
Table 1 A comparison of previously determined inci-
dence estimates with those determined in t h e present study Estimate from Estimate from present previous study study 1. HLHS'
36.5/100,000
-
2. HLHS with aortic +/or mitral atresia
25.11100.000
26.0/100,000
3. Aortic +/or mitral atresia (with or without HLHS)
29.0/100,000
26.7/100,000
HLHS = hypoplastic left heart syndrome.
247
INHERITANCE OF HLHS
co-workers’ value of 26.7/100,000 births for the same category (Mitchell et al. 1973). The hypothesis of autosomal recessive transmission was tested by an a posteriori method of estimating the proportion of affected individuals, or segregation ratio. First, assuming single ascertainment, the segregation ratio Po was estimated by the simple sib method (Weinberg 1927). Secondly, assuming truncate ascertainment of the original 62 families and, thus, multiple ascertainment of the 42 families interviewed, the segregration ratio PI was estimated by Elandt-Johnson’s method (1971). In theory (Haldane 1938, Steinberg 1959), if this disorder is transmitted as a autosomal recessive trait we should find: Po < 0.250 < Pi However, we found that POwas 0.019 and PI was 0.030. Therefore, because even the upper limit segregation ratio, PI, was significantly lower than the ratio 0.250 expected with autosomal recessive inheritance, this hypothesis cannot be supported ( ~ 2Total (7) = 25.318, P < .001; x 2 combined (1) = 23.183, P < .001; x 2 difference (6) = 2.135, 0.900 < P < 0.950) (Elandt-Johnson 1970). This hypothesis was tested further by the a priori method of Apert (McKusick 1969), (Table 2). This allowed calculation of the
Table 2
Family data with expected number affected based on Apert’s a priori method Sibship size
No. of sibships
No. of sibs
Observed affected
7 6 5 12 8 2 1 1 42
7 12 15 48 40 12 7
7
Expected affected ~~~
1 2 3 4 5 6 7 8 Totals
X~ 3.76, 0.05
a 149
< P < 0.10.
7 5 12 9 2 1 1 44
7.000 6.858 6.485 17.556 13.112 3.650 2.020 2.222 56.903
expected number affected, assuming recessive inheritance and taking into account the number of sibships and their respective sizes. This expected value, 58.903, was not significantly greater than the observed number affected, 44 ( x 2 (1) = 3.76, 0.05 < P < 0.10). Accordingly, autosomal recessive inheritance could not be rejected on the basis of the a priori method of segregation analysis. Using Edwards’ formula (1960) the expected recurrence among later siblings, 1.91 %, was not significantly different from the 1.87 % recurrence observed among later siblings. Discussion
At the outset it had been our intention to delineate discrete clinicogenetic entities within HLHS, each individually recognizable with a possibly distinct mode of transmission and risk of recurrence. However, because of the size of our sample and the low recurrence within the sample, this proved impossible. Based on the data for the syndrome as a whole, the results of the a posteriori test and lack of discernible parental consanguinity leave little support for the autosomal recessive hypothesis of transmission of this condition. However, the a priori method of Apert failed to reject the hypothesis of autosomal recessive transmission (0.05 P < 0.10). This perhaps indicates the shortcoming of Apert’s method in discriminating the various modes of inheritance. The low observed recurrence among later siblings as predicted by the polygenic inheritance model lends support to this proposed mode of determination. We can conclude, therefore, that autosomal recessive inheritance alone cannot account for our observations, while polygenic inheritance alone may. Nevertheless, the polygenic model is very
Inheritance of hypoplastic left heart syndrome (HLHS): Further observations* L. G. BROWNELLAND M. H. K. SHOKEIR** Department of Paediatrics, University of Manitoba and Children’s Centre, Winnipeg, Manitoba, Canada
From a survey of newborn infants using truncate ascertainment of families through affected offspring, 64 cases of hypoplastic left heart syndrome (HLHS) were found. Pedigree information was obtained from 42 families. The incidence of HLHS, based on 64 affected infants, was 36.5/100,000 births. The incidence of HLHS with aortic and/or mitral atresia was 25.1/100,000 births. The recurrence among later siblings was about 2 %. AutosomaI recessive hypothesis of transmission of this disorder was not supported. Using Edwards’ formula, the observed recurrence among later siblings was compatible with polygenic inheritance. Received 23 May, accepted for publication 20 June 1975
Hypoplastic left heart syndrome (HLHS), described first by Lev (1952), encompasses a broad spectrum of abnormalities of the left side of the heart ranging from hypoplasia of the aortic arch with an underdeveloped left ventricle, to more severe forms with atresia of the aorta, atretic aortic and mitral valves, hypoplastic left atrium and vestigial left ventricle (Lumb & Dawkins 1960, Noonan & Nadas 1958, Sinha et al. 1968). The pathology, pathophysiology, and natural history of this disorder have been outlined previously (Shokeir 1971). Until recently, little has been published concerning familial clustering of cases of HLHS and the possible genetic contribution to their etiology. Brekke (1953) reported a
* Supported by MRC of Canada. ** Queen Elizabeth I1 Scientist.
family in which two siblings suffered aortic atresia and hypoplasia of the aortic orifice. Ehlers & Engle (1966) described a sibship of three, with two affected infants. In 1966 Rao et al. published a report of two successive siblings with hypoplastic ascending aorta, atretic aortic valve, and small left ventricle and atrium. Kojima (1969) reported on a family in which two of four siblings were affected. Familial occurrence of HLHS was also described by Saied & Folger (1972). Kaufman et al. (1972) described a sibship of three girls; the first- and third-born had HLHS with vertebral and renal anomalies, while the second child had only vertebral and renal anomalies. In 1974, Bjornstad & Michalsen discussed the occurrence of coexistent mitral and aortic valve
246
BROWNELL AND SHOKEIR
atresia with intact ventricular septum in siblings. In a previous study of the possible genetic etiology of this syndrome, Shokeir (1971) suggested autosomal recessive transmission as the operative mode of inheritance. Subsequently, Holmes et al. (1972, 1974) invoked the concept of polygenic inheritance to explain the pattern of transmission. We, therefore, embarked on a comprehensive study of the disorder in a localized multiethnic urban community - Winnipeg. This setting was in marked contrast to that of the first study (Shokeir 1971) which focused on a rural population with considerable ethnocentric distribution and appreciable endogamy. Present lnvestigatlon
A survey of infants affected with HLHS, born in the city of Winnipeg between January 1957 and June 1974, was conducted through a chart and post mortem review in Winnipeg hospitals. This permitted ascertainment of cases diagnosed clinically (with confirmation by cardiac catheterization) andlor at necropsy. To allow for the variation in terminology used by cardiologists and pathologists in describing given entities within the syndrome, the boundaries of our search were extended initially to include most congenital heart diseases. Ultimately, however, only the 64 infants with hypoplastic left ventricle and with left-sided obstruction severe enough to result in death, were included. The presence and severity of aortic and mitral atresia varied, as did the presence of endocardial fibroelastosis. Four cases with multiple extracardiac lesions or chromosomal aberrations were excluded. Family pedigrees were obtained from 42 of the 64 families ascertained through affected offspring. The remaining families could not be contacted. No parental con-
sanguinity could be detected. The sex ratio was 1.37:l (M:F), conforming to the previously observed excess of affected males (Monie & DePape 1950, Noonan & Nadas 1958), which however, was not significantly different from a 1:l ratio (xZ(1) = 1.56 0.15 < P < 0.25). There was no evidence of an increased number of spontaneous abortions, nor was there a significant association with maternal drug ingestion, infection, Xirradiation, or history of diabetes. The data thus generated were used to estimate the incidence of the disorder in the population and the rate of recurrence among later siblings. Furthermore, hypotheses of modes of genetic transmission were examined. The incidence of HLHS, based on 64 affected infants, was 36.5/100,000 births (Table 1). The incidence of HLHS with aortic and/or mitral atresia was 25.11 100,000 births, which is comparable to the value 261100,000 births obtained by Shokeir (1974), who used analogous criteria for his survery. The incidence of aortic andlor mitral atresia associated either with HLHS or with other conditions (transposition of the great vessels, single ventricle, or chromosomal aberrations) was 29.0/100,000 births, which is comparable to Mitchell and
Table 1 A comparison of previously determined inci-
dence estimates with those determined in t h e present study Estimate from Estimate from present previous study study 1. HLHS'
36.5/100,000
-
2. HLHS with aortic +/or mitral atresia
25.11100.000
26.0/100,000
3. Aortic +/or mitral atresia (with or without HLHS)
29.0/100,000
26.7/100,000
HLHS = hypoplastic left heart syndrome.
247
INHERITANCE OF HLHS
co-workers’ value of 26.7/100,000 births for the same category (Mitchell et al. 1973). The hypothesis of autosomal recessive transmission was tested by an a posteriori method of estimating the proportion of affected individuals, or segregation ratio. First, assuming single ascertainment, the segregation ratio Po was estimated by the simple sib method (Weinberg 1927). Secondly, assuming truncate ascertainment of the original 62 families and, thus, multiple ascertainment of the 42 families interviewed, the segregration ratio PI was estimated by Elandt-Johnson’s method (1971). In theory (Haldane 1938, Steinberg 1959), if this disorder is transmitted as a autosomal recessive trait we should find: Po < 0.250 < Pi However, we found that POwas 0.019 and PI was 0.030. Therefore, because even the upper limit segregation ratio, PI, was significantly lower than the ratio 0.250 expected with autosomal recessive inheritance, this hypothesis cannot be supported ( ~ 2Total (7) = 25.318, P < .001; x 2 combined (1) = 23.183, P < .001; x 2 difference (6) = 2.135, 0.900 < P < 0.950) (Elandt-Johnson 1970). This hypothesis was tested further by the a priori method of Apert (McKusick 1969), (Table 2). This allowed calculation of the
Table 2
Family data with expected number affected based on Apert’s a priori method Sibship size
No. of sibships
No. of sibs
Observed affected
7 6 5 12 8 2 1 1 42
7 12 15 48 40 12 7
7
Expected affected ~~~
1 2 3 4 5 6 7 8 Totals
X~ 3.76, 0.05
a 149
< P < 0.10.
7 5 12 9 2 1 1 44
7.000 6.858 6.485 17.556 13.112 3.650 2.020 2.222 56.903
expected number affected, assuming recessive inheritance and taking into account the number of sibships and their respective sizes. This expected value, 58.903, was not significantly greater than the observed number affected, 44 ( x 2 (1) = 3.76, 0.05 < P < 0.10). Accordingly, autosomal recessive inheritance could not be rejected on the basis of the a priori method of segregation analysis. Using Edwards’ formula (1960) the expected recurrence among later siblings, 1.91 %, was not significantly different from the 1.87 % recurrence observed among later siblings. Discussion
At the outset it had been our intention to delineate discrete clinicogenetic entities within HLHS, each individually recognizable with a possibly distinct mode of transmission and risk of recurrence. However, because of the size of our sample and the low recurrence within the sample, this proved impossible. Based on the data for the syndrome as a whole, the results of the a posteriori test and lack of discernible parental consanguinity leave little support for the autosomal recessive hypothesis of transmission of this condition. However, the a priori method of Apert failed to reject the hypothesis of autosomal recessive transmission (0.05 P < 0.10). This perhaps indicates the shortcoming of Apert’s method in discriminating the various modes of inheritance. The low observed recurrence among later siblings as predicted by the polygenic inheritance model lends support to this proposed mode of determination. We can conclude, therefore, that autosomal recessive inheritance alone cannot account for our observations, while polygenic inheritance alone may. Nevertheless, the polygenic model is very
