24Med Genet 1995;32:264-268

Alagille syndrome: family studies

F V Elmslie, A J Vivian, H Gardiner, C Hall, A P Mowat, R M Winter

AbstractAlagille syndrome (AGS) is one of themajor forms of chronic liver disease inchildhood with severe morbidity and amortality of 10 to 20%. It is characterisedby cholestasis of variable severity withpaucity of interlobular bile ducts and an-omalies ofthe cardiovascular system, skel-eton, eyes, and face. Previous studiessuggest a wide variation in the expressionof the disease and a high incidence ofnewmutations. To determine more accuratelythe rate of new mutations and to developcriteria for detecting the disorder in par-ents we systematically investigated par-ents in 14 families with an affected child.Clinical examination was supplementedby liver function tests, echocardiography,radiographic examination ofthe spine andforearm, ophthalmological assessment,and chromosome analysis. Six parents hadtypical anomalies in two or more systemspointing to the presence of autosomaldominant inheritance. Systematic screen-ing of parents for the features defined inthis study should improve the accuracy ofgenetic counselling.

( Med Genet 1995;32:264-268)

The Hospital for SickChildren, GreatOrmond Street,London WClN 3JH,UKF V ElmslieH Gardiner*C HallR M Winter

Moorfields EyeHospital, City Road,London EClV 2PD,UKA J Vivian

Department of ChildHealth, Variety ClubChildren's Hospital,King's CollegeHospital,Denmark Hill,London SE5 8RX, UKA P Mowat* Present address:The Glenfield Hospital,Groby Road,LeicesterLE3 9QP, UK.

Correspondence to: DrElmslie, Department ofPaediatrics, UCL MedicalSchool, The Rayne Institute,5 University Street, LondonWC1E 6JJ, UK.Received 12 September1994Revised version accepted forpublication 9 December1994

Alagille syndrome is a common cause ofcholestatic liver disease in childhood, with anestimated incidence of 1 in 70 000 live births.'The association between intrahepatic chole-stasis, a characteristic face, and a cardiacmurmur was described in 1969 by Alagille etal2 as a new and distinct form of cholestasisin infancy. Further reports by Watson andMiller' and Alagille et al' provided more

evidence for a new syndrome. Liver diseaseoccurs in association with paucity of inter-lobular bile ducts (intrahepatic biliary hypo-plasia), detectable on liver biopsy. It isaccompanied by cardiovascular abnormalities,in particular peripheral pulmonary stenosis,skeletal anomalies, and ophthalmological de-fects. The characteristic skeletal abnormalityis butterfly vertebrae caused by a persistentsagittal cleft through the vertebral body; thesemay fuse with time and therefore may notbe present in older patients with AGS. Otherskeletal abnormalities such as shortening ofthe bones of the forearm and hands andnarrowed interpedicular distance of the ver-

tebrae may be present.56 The eye abnormalityusually seen is posterior embryotoxon, an

abnormal prominence of Schwalbe's line.7Posterior embryotoxon is known to occur in8 to 15% of the normal population.8

In a review of 80 cases in 1987, Alagille

et al9 suggested that there were five cardinalfeatures of the syndrome: paucity of in-trahepatic bile ducts, cardiovascular ab-normalities, vertebral arch defects, posteriorembryotoxon, and a characteristic face. In thesame year, Mueller'" suggested that the diag-nosis could be made in the presence ofany threeof six features: intrahepatic biliary hypoplasia,peripheral pulmonary stenosis, posterior em-bryotoxon, butterfly vertebrae, a characteristicface, and a first degree relative with AGS. In1986, Byrne et all' found a deletion of theshort arm of chromosome 20 in a baby withintrauterine growth retardation, jejunal stenosisand dysmorphic facial features associated withperipheral pulmonary stenosis, vertebral ab-normalities, and cholestasis secondary topaucity of interlobular bile ducts. Reviewingprevious reports of monosomy 20p they foundthat all had some features ofAlagille syndrome.A further 13 cases ofdeletions of20p associatedwith AGS have been described'2 including onecase in which the deletion had been transmittedfrom an affected mother to her daughter. Thispoints to the existence of a locus or loci onchromosome 20p that are responsible for pro-ducing AGS.AGS is now established as being inherited

in an autosomal dominant fashion,"-'6 but withextreme variability ofexpression and a high rateof new mutation. Several families, includingWatson and Miller's original families,3 havebeen described in which Alagille syndrome hasbeen transmitted from one generation to thenext with variation in the phenotype. In thecases so far published, a mild phenotype in theparent has led to a more severe phenotype inthe offspring, leading to the suggestion thatanticipation may occur in this disorder. In ad-dition, Shulman et all" suggested that in-heritance from the mother resulted in a moresevere phenotype leading this author to suggestsimilarities with the inheritance of myotonicdystrophy. However, there is little informationto date about the proportion of affected chil-dren who inherit the disorder from a parent.Nor is there published information on the sys-tematic evaluation of parents with regard tothe abnormalities present in AGS. Knowledgeof the minimal expression of the disease wouldenable more accurate counselling ofthe familiesof children with AGS.The aim of the study, therefore, was to

develop criteria to aid diagnosis in an affectedparent. In addition we wished to determinein what proportion of affected children therewas evidence of autosomal dominant in-heritance of the disease and whether maternalor paternal transmission of the disease hadan effect on the severity of the phenotype ofthe offspring.

264

265Alagille syndrome: family studies

O! Liver diseaseO~Cardiovascular abnormalityFL Posterior embryotoxonEl Skeletal malformation0 Unaffected

Family 1 Family 8

El 12 3

MI : 31111:3 11:4 13 1:5 M

Family 6

1:1 1:2 I:3 I4

111:1 111:2 111:3 H1:4 1111:5 1HA:

Family 7

:Li 1:2 1:3 1;4

ILl 11:211:3 11:4 11:5 11:6 l:7 11S 11:9

M1:I 111:2 M:3 111:IS

M1 I:2 J 1:4

MlI DL2 M3

Family 12

E I 1;2 1:3 1:4

HA: f:2 xJ:3 tJ:4 11.5 11:6 11:7

M:1111:2 111:3 1U:4 :5

Family 14

:1 3 1:4

ILIII 211.3 11:6

Figure 1 Pedigrees offamilies in which there was evidence of autosomal dominant inheritance.

Patients and methodsThe patients studied were ascertained throughthe Children's Liver Unit of King's CollegeHospital and The Hospital for Sick Children,Great Ormond Street, London. Index cases

were included in the study if they had a definitediagnosis of Alagille syndrome based on thecriteria of Mueller.'0 All 14 index cases, 24parents, and four sibs were examined at TheHospital for Sick Children. In four familiesonly one parent was examined, the remainderbeing unavailable for study. All index cases andtheir parents had the following investigationsperformed: liver function tests (LFTs) (serumbilirubin, alkaline phosphatase, glutamyl trans-peptidase, aspartate transaminase), serum

cholesterol, blood chromosomes, radiographsof the spine, forearm, and hand, echo-cardiography, anterior and posterior segmentophthalmic examination, and ocular elec-trophysiology. The bones of the hand and in-terpedicular distances of the spine were

measured. The results were analysed using an

unpaired t test to search for minor differencesbetween the two parent groups, affected andunaffected.

ResultsWe found definite evidence for autosomal dom-inant inheritance of the syndrome in six ofthe 14 families studied. The pedigrees of thefamilies displaying dominant inheritance areshown in fig 1. Subject II-7 in family 7 had aninteresting family history. Two sisters died ininfancy and another in childhood followingrepeated hospital admissiors. All were knownto have had heart murmurs but no other detailswere available. It is possible that they were alsoaffected.

In three cases the father was the affectedparent and in three the mother. In only onecase had the affected parent previously sus-pected that he was affected. Only three sibs(family 1 III-3, family 7 III-5, family 12 III-4)were examined and none of them underwent

Elmslie, Vivian, Gardiner, Hall, Mowat, Winter

blood tests or radiography. However, on clinicalexamination, there was no evidence of the dis-ease in any of the three.

In four of these six families there was a

history of miscarriage, all occurring at over

10 weeks' gestation. Two mothers had two

miscarriages, and a further two had one each,out of a total of 22 pregnancies. In the non-

familial group of eight families there were two

miscarriages out of a total of 22 pregnancies.Although there superficially appears to be an

excess of miscarriages in the familial group,

this did not reach statistical significance usingx2 with Fisher's exact test.The clinical findings and results of in-

vestigations in the affected parents are pres-

ented in table 1. All affected parents hadposterior embryotoxon and at least one othermajor syndromic feature. Five had ab-normalities of the spine and eye. In three,midline notches on the vertebral end plateswere present representing fused butterfly ver-

tebrae. Four also had a short ulna. Two hadanomalous optic discs and a pigmentary ret-inopathy. Electrophysiology of the eye was nor-

mal in all cases, including the parent withpigmentary retinopathy. Three had pulmonarymurmurs on clinical examination but in thetwo who underwent echocardiography no ab-normality was detected. In only two was thereany abnormality of the liver function tests, inone a mildly raised bilirubin, and in the othera mild rise of alkaline phosphatase. The motherin family 14 and the father in family 6 hada history of jaundice in infancy which was

unexplained and recovered spontaneously. Inall parents blood chromosomes were normal.Using an unpaired t test, no significant differ-ence was found between affected and un-

affected parents in any of the followingparameters: lengths of the bones in the hand,lumbar interpedicular distance, aspartate trans-aminase, albumin, bilirubin, and triglycerides.However, the alkaline phosphatase levels were

significantly higher in affected parents with a

p value of <0 05.The severity of the disease in the affected

parent varied and did not correlate with theseverity in the child. Neither did the sex of the

parent influence the severity of disease in thechild. For example, in family 1, the affectedmother (II-3) gave no history of jaundice inchildhood and had no biochemical evidence ofliver disease. She did have skeletal ab-normalities and posterior embryotoxon (table1). She had a daughter (III-2) with severe liverdisease, requiring liver transplantation. III-2also required surgery for an anomalous leftcoronary artery at the age of 1 year and hadperipheral pulmonary stenosis. In contrast,both the mother of family 12 (II-2) and herdaughter (III3) had abnormalities in all foursystems.

DiscussionEvidence for autosomal dominant inheritancewas present in six of the 14 families studied,representing 43% of those studied. The ma-

jority of children therefore represented possiblenew mutations. We found no evidence for an-

ticipation as was suggested by Shulman et al."7There was no significant difference in thephenotype of children who had inherited thedisorder from their mother as compared withthose who inherited from their father, sug-

gesting that comparisons with myotonic dys-trophy are not valid.Although previous studies have suggested

that the new mutation rate is high in thisdisorder, the number of children with the syn-

drome attributable to a new mutation has beenfar from clear. Few family studies have beenpublished and the numerous case reports ofthe syndrome may be biased towards reportingof those in which there is a family history. Ithas therefore been difficult to elucidate theproportion of affected children who representnew mutations, a figure that would aid geneticcounselling. Our data suggest that at least 50%of those with AGS represent new mutations,deduced from the fact that investigation oftheirparents was normal. These results concur withsegregation analysis performed by Dhorne-Pol-let et al."6AGS has frequently been described as a dis-

ease displaying variable penetrance. Strictly,penetrance means the frequency with which

Table 1 Characteristics of affected parents

Family

1 6 7 8 12 14

Parent affected Mother Father Father Father Mother MotherJaundice in infancy No Yes No No No NoLFTs Normal Raised Normal Normal Bilirubin Not done

alkaline 23 tmol/lphosphatase (ref range(131 U/I) < 12 gmol/l)

Cholesterol Not done 8-1 5-1 Not done 4-8 Not done(mmol/l)Echocardiography Normal Normal but Normal but Normal Not done, Not done

pulmonary pulmonary pulmonarymurmur present murmur murmur

Radiography Midline notches Scoliosis, Tight scoliosis, Midline notches Midline notches Normal spine,D6-D8, short short ulna short ulna on vertebral on vertebral end short 4thulna end plates plates D7, D8, metacarpal

D10,short ulna

Ophthalmic Posterior Posterior Posterior Posterior Posterior Posteriorexamination embryotoxon embryotoxon, embryotoxon embryotoxon embryotoxon, embryotoxon

divergent pigmentarystrabismus, retinopathy,anomalous discs anomalous discs

266

Alagille syndrome: family studies

those carrying the gene express it. Completepenetrance implies that all who have the geneexpress it, variable penetrance or incompletepenetrance implies that there are some peoplewho carry the gene but do not express it. Theonly possible historical evidence for incompletepenetrance in Alagille syndrome is the familydescribed by Mueller et al'3 in which two ap-parently normal parents had two affected chil-dren. In no published case has the disease beenseen to skip a generation. We did not examineany of the grandparents but there was no evid-ence from the family histories of the unaffectedparents of non-penetrance of the gene. It ispossible that the expression of the disease wasvery mild in the family of Mueller et all3 andwent undetected, but an alternative explanationwould be that of gonadal mosaicism. Onemechanism known for non-penetrance is im-printing. Reviewing published family data forevidence of imprinting we found no evidencefor its presence in AGS. In our group threechildren inherited Alagille syndrome from theirfathers and three from their mothers; there wasno difference in phenotype according to theparent of the origin. Our findings do not sup-port imprinting as an important mechanism in

Figure 2 (Above) Proband and affected father from family 7. (Below) Proband fromfamily 8 with affected father and unaffected mother.

AGS. Only one sib in our group appeared tobe affected, out of eight, which is at odds withautosomal dominant inheritance with completepenetrance. However only three sibs were ex-amined, and none underwent investigation. Itis therefore possible that some of them mayhave been mildly affected and went undetected.

It has been postulated that AGS may rep-resent a contiguous gene syndrome.'8 None ofthe probands had dysmorphic features ad-ditional to those seen in AGS. Three of the 14probands had significant learning difficulties ormotor developmental delay. Of these, one hadataxic cerebral palsy in addition to AGS at-tributed to hypoxic-ischaemic encephalopathy,and another had severe liver disease requiringtransplantation at the age of 3 years. Post-transplantation he began to make rapid de-velopmental progress. We found no firm evid-ence to suggest that AGS is a contiguous genesyndrome.

In addition, we found no evidence for theexistence of anticipation in this disorder. Inthree families the affected child appeared tohave similar disease severity to the parents. Ifsubjects II-3, II-4, and II-5 in family 7 wereaffected, their early deaths imply that theirdisease was more severe than that of both II-7and III- 1. It is more likely that those that aremildly affected survive and reproduce and thosethat have more severe disease die or are unableto reproduce, resulting in the superficial ap-pearance of anticipation. In addition, there wasno evidence for a "maternal factor" resultingin increased severity of disease when the diseasewas transmitted from an affected mother to heroffspring.There is increasing evidence that AGS is not

as benign as was originally thought, and thereare reports oflong term complications (notably,hepatocellular carcinoma and late onset liverfailure) occurring in AGS.'9 It is not knownwhether those parents retrospectively as-certained run a risk of developing the com-plications associated with the syndrome,although there is no documented case of thesecomplications occurring in a patient with noclinical or biochemical evidence of liver disease.Until more is understood about the long termnatural history of this disease it will be difficultto be reassuring when counselling these famil-ies.

In order to provide accurate genetic coun-selling to the families of children with Alagillesyndrome it is important to be able to dis-tinguish between those that represent newmutations and those that have inherited AGSfrom an affected parent. The extreme variabilityof expression of the syndrome has made this adifficult task. Based on this small study a setof major and minor criteria was developed foraiding diagnosis in a family which presents withan affected child. Major criteria are establishedfor the diagnosis of children with Alagille syn-drome,910 but similar criteria do not apply toadults with the disorder. The major and minorcriteria are shown in table 2. All parents ofchildren with AGS will fulfil at least one of themajor criteria, that of having a first degreerelative with Alagille syndrome. In addition, all

267

Elmslie, Vivian, Gardiner, Hall, Mowat, Winter

Table 2 Major and minor criteria for use in diagnosis ofparents

Major criteria Minor criteria

History of prolonged jaundice in infancy Alkaline phosphatase > 103 U/irequiring investigation Short ulnaPulmonary murmur Pigmentary retinopathyPosterior embryotoxon Anomalous optic discsVertebral end plate notchesFirst degree relative with Alagillesyndrome (omit if proband)

Table 3 Number of major and minor criteria present inaffected parents

Parent Major Minor

1 3 16 4 37 3 18 3 012 4 314 3 0

our affected parents had at least two majorcriteria, and in those that had been completelyinvestigated, between one and three minor cri-teria (table 3). In our group of six affectedadults, this represented the minimal expressionof the syndrome.The variability of expression can make it

difficult to be categorical about whether a par-ent is affected or not. It is clear from our

evaluation of 14 families that a detailed history,examination, and investigation are required todistinguish the affected group from the un-

affected group, and the development of majorand minor criteria for diagnosis may help inthe future evaluation of families.

We wish to thank all the families who participated in the studyfor their patience and enthusiasm, Dr Alastair Baker for hishelp in family ascertainment, Ms Vanda Gooch for performingechocardiography, and Dr Guan Lim for his help and en-couragement.

1 Danks DM, Campbell PE, Jack I, Rogers J, Smith AL.Studies of the aetiology of neonatal hepatitis and biliaryatresia. Arch Dis Child 1977;52:360-7.

2 Alagille D, Habib EC, Thomassin N. L'atresie des voiesbiliaries intrahepatiques avec voies extrahepatiques permeablechez l'enfant. Paris: Editions Medicales Flammarion, 1969:301-18.

3 Watson GH, Miller V. Arteriohepatic dysplasia: familialpulmonary arterial stenosis with neonatal liver disease.Arch Dis Child 1973;48:459.

4 Alagille D, Odievre M, Gautier M, Dommergues P. Hepaticductular hypoplasia associated with characteristic facies,vertebral malformations, retarded physical, mental andsexual development, and cardiac murmur. J Pediatr 1975;86:63.

5 Singcharoen T, Partridge J, Jeans WD, Baddeley H. Ar-teriohepatic dysplasia. Br _J Radiol 1986;59:509-1 1.

6 Rosenfield NS, Kelly MJ, Jensen PS, Cotlier E, RosenfieldAT, Riely CA. Arteriohepatic dysplasia: radiologic featuresof a new syndrome. A3rR 1980;135:1217-23.

7 Riely CA, Cotlier E, Jensen P, Klatskin G. Arteriohepaticdysplasia: a benign syndrome of intrahepatic cholestasiswith multiple organ involvement. Ann Intern Med 1979;91:520-7.

8 Waring GO, Rodrigues MM, Laibson PR. Anterior chambercleavage syndrome. A stepladder classification. Surv Oph-thalmol 1975;20:3-37.

9 Alagille D, Estrada A, Hadchouel M, Gautier M, OdievreM, Dommergues JP. Syndromic paucity of interlobularbile ducts (Alagille syndrome or arteriohepatic dysplasia).Review of 80 cases. J Pediatr 1987;110: 195-9.

10 Mueller RF. The Alagille syndrome (arteriohepatic dys-plasia). _J Med Genet 1987;24:621-6.

11 Byrne JLB, Harrod MLE, Friedman JM, Howard-PeeblesPN. Del (20p) with manifestations of arteriohepatic dys-plasia. Am J. Med Genet 1986;24:673-8.

12 Anad F, Burn J, Matthews D, et al. Alagille syndrome anddeletion of 20p.J Med Genet 1990;27:729-37.

13 Mueller RF, Pagon RA, Pepin MG, et al. Arteriohepaticdysplasia: phenotypic features and family studies. ClinGenet 1984;25:323-31.

14 Greenwood RD, Rosenthal A, Crocker AC, Nadas AS.Syndrome of intrahepatic biliary dysgenesis and cardio-vascular malformations. Pediatrics 1976;58:243-7.

15 Henriksen NT, Langmark F, Sorland SJ, Fausa 0, LandaasS, Aagenaes 0. Hereditary cholestasis combined withperipheral pulmonary stenosis and other anomalies. ActaPaediatr Scand 1977;66:7-15.

16 Dhorne-Pollet S, Deleuze JF, Hadchouel M, Bonaiti-PellieC. Segregation analysis of Alagille syndrome. J Med Genet1994;31:453-7.

17 Shulman SA, Hyams JS, Gunta R, Greenstein RM, CassidySB. Arteriohepatic dysplasia (Alagille syndrome): extremevariability among family members. AmJMed Genet 1984;19:325-32.

18 Schnittger S, Hofer C, Heidemann P, Beernann F,Hansmann I. Molecular and cytogenetic analysis of aninterstitial 20p deletion associated with syndromic in-trahepatic ductular hypoplasia (Alagille syndrome). HumGenet 1989;83:239-44.

19 Schwarzenberg SJ, Grothe RM, Sharp HL, Snover DC,Freese D. Long-term complications of arteriohepatic dys-plasia. Am _t Med 1992;93:171-6.

268