Holt-Oram Syndrome: a Clinical Genetic Study J Med Genet: First Published As 10.1136/Jmg.33.4.300 on 1 April 1996

300_0fMed Genet 1996;33:300-307

Holt-Oram syndrome: a clinical genetic study J Med Genet: first published as 10.1136/jmg.33.4.300 on 1 April 1996. Downloaded from R A Newbury-Ecob, R Leanage, J A Raebum, I D Young

Abstract to clarify the spectrum of abnormalities and to A clinical and genetic study of the Holt- delineate the HOS phenotype led us to review Oram syndrome (HOS) has been carried the clinical features in our patients, and dis- out in the United Kingdom involving 55 tinguish the clinical features most helpful for cases designated Holt-Oram syndrome, counselling purposes. together with their parents and sibs. Data This study was carried out in conjunction from the clinical assessment of both fa- with a genetic linkage study which has shown milial and isolated cases were used to de- genetic heterogeneity in the Holt-Oram syn- fine the HOS phenotype and to outline drome, with one gene (HOS1) being localised the spectrum of abnormalities, especially to chromosome 12 in five out ofseven families.7 factors affecting severity. Skeletal defects No phenotypic differences could be detected affected the upper limbs exclusively and between the linked and unlinked families. were bilateral and asymmetrical. They ranged from minor signs such as clinodactyly, limited supination, and sloping Patients and methods shoulders to severe reduction deformities The study was carried out between March 1991 of the upper arm (4.5%). The radial ray and September 1993. Cases were ascertained was predominantly affected and the left by contacting clinical geneticists and paediatric side was more severely affected than the cardiologists and through the support group right. All affected cases showed evidence of REACH, for children with upper limb de- upper limb involvement. Cardiac defects ficiency. Patients and first degree relatives were were seen in 95% of familial cases and visited at home by RNE. A detailed family included both atrial septal defect (ASD, history was taken as well as all relevant medical 34%) and ventricular septal defect (VSD, and surgical details. Patients were examined 25%); 39% had only ECG changes. Cardiac carefully for any skeletal abnormalities or dys- involvement ranged from asymptomatic morphic features. Measurements were taken of conduction disturbances to multiple struc- the upper arm, lower arm, palm, thumb, and tural defects requiring surgery in infancy. middle finger. Where possible portable elec- Sudden death could be caused by heart trocardiography was performed. Information block. Inheritance was autosomal dom- was sought from x rays, electrocardiograms http://jmg.bmj.com/ inant with 100% penetrance and no evid- (ECG), and cardiac ultrasound examinations ence ofreduced fitness. Increasing severity previously undertaken. occurred in succeeding generations con- In order to assess the severity of ab- sistent with anticipation. normalities, clinical examination data were re- (J'Med Genet 1996;33:300-307) corded using a scoring system based on that devised by Gall et al and modified by Glad-

Key words: Holt-Oram syndrome; heterogeneity; ma- stone and Sybert9 (table 1). Measurements on September 26, 2021 by guest. Protected copyright. ternal effect; anticipation. were compared with age matched standards.'0 Statistical analysis was carried out using Mann- Whitney/Kruskall Wallis and Spearman Rank Several heart-hand syndromes have been iden- Correlation. tified of which the Holt-Oram (HOS) is the best known. The first formal report by Holt and was based on a four generation Oram in 1960' ASCERTAINMENT family in which the main skeletal lesion was a Sixty six affected subjects from 28 families were triphalangeal thumb; the heart defect was a Twelve families had three There referred to the study. Centre for Medical secundum atrial septal defect (ASD). or more affected members. There were 33 Genetics, City was also an unusual cardiac arrhythmia and their from Hospital, Hucknall vessels. Of males and 33 females; ages ranged Road, Nottingham evidence of hypoplastic peripheral NG5 lPB, UK nine subjects, four were confirmed as affected R A Newbury-Ecob were affected his- and five (not examined) by Table 1 Severity score system J A Raeburn ASD was in four but only con- I D Young tory. suspected one case. (ECG) Skeletal abnormalities firmed in Electrocardiography 0 No abnormality on physical or radiological examination Department of showed a variety of abnormalities including 1 Minor abnormalities including reduced thenar eminence, of the thumb Paediatric Cardiology, a long PR interval, sinus bradycardia, nodal clinodactyly, or hypoplasia Glenfield Hospital, 2 Triphalangeal or aplastic thumbs. Radial/ulnar hypoplasia Leicester, UK escape, and atrial fibrillation. Three family 3 Arms and forearms present but bone(s) missing R Leanage members had a triphalangeal thumb. The 4 Phocomelia was absent in one subject and another Cardiac abnormalities Correspondence to: thumb One case had a minor 0 Asymptomatic with no physical findings Dr Newbury-Ecob. had radial aplasia. only 1 Asymptomatic murmur or conduction defect Received 16 October 1995. anomaly of the clavicle. Subsequent reports 2 Structural heart abnormality not requiring surgery Revised version accepted have included various combinations of upper 3 Structural heart abnormality requiring surgery for publication 4 Lethal malformation 30 November 1995 limb abnormality and heart defect.`6 The need Holt-Oram syndrome: a clinical genetic study 301

Table 2 Details of cases referred to study March Table 3 Frequency and type of common upper limb and 1991-September 1993 thorax abnormalities in 55 patients with HOS Index cases 66 Abnormality No affected (%) Relatives found to be affected during study 6 Total affected cases included in study 72 Familial Isolated J Med Genet: first published as 10.1136/jmg.33.4.300 on 1 April 1996. Downloaded from Personally examined 57 n=44 n=11 Dead 5 Unavailable or insufficient details 8 Hand Total unaffected relatives 59 Total affected and unaffected 131 Thumb Absent 19 (43) 8 (73) Hypoplasia 17 (39) 4 (36) Triphalangeal 8 (18) 2 (18) Syndactyly 5 (11) 2 (18) 4 months to 70 years. Thirteen cases had Long 6 (14) 0 no Normal 7 (16) 0 family history of HOS and are referred to in Hypoplastic thenar eminence 31 this study as isolated. Fifteen patients were (70) 11 (100) not Fingers available for examination; these included five Clinodactyly 28 (64) 5 (45) affected subjects who had died whose details Brachydactyly 19 (43) 9 (82) Hypoplasia 7 (16) were obtained from family history (5) and from Absent 3 (7) 3 (27) case notes (2), PM details (2), and photographs Syndactyly 4 (9) 2 (18) (4). Twelve examined patients had not had Normal 10 (23) 0 ECG and echocardiography. Six relatives, Lower arm thought previously to be unaffected, were found Radius Hypoplasia 18 (41) 6 (55) to be affected during the study examination. Aplasia 10 (23) 3 (27) In total, 57 patients were examined. Details of Ulna the cases are given in table 2. Cases were Hypoplasia 18 (49) 2 (18) included only if the heart and radial ray defects Aplasia 0 4 (36) previously described in HOS were present in at Limited supination 27 (61) 5 (45) least one affected family member. Two patients Linmited extension elbow 22 (50) 7 (64) from one family who had upper limb ab- Normal 15 (34) 1 (9) normalities but no evidence of cardiac ab- Upper arm normality within the family were excluded from Humerus the study. Hypoplasia 20 (45) 8 (73) Abnormal head 22 (50) 7 (64) Aplasia 0 1 (9) Normal 22 (50) 1 (9) Clinical features Shoulder girdle UPPER LIMB ABNORMALITIES (TABLE 3) Clavicles Familial cases Hypoplasia 33 (78) 7 (64) Prominent acromioclavicular joint 11 (25) 5 (45) In all cases upper limb abnormalities were Normal 10 (23) 2 (18) present and were clinically detectable. The ra- Thorax dial ray was predominantly affected, with ulnar http://jmg.bmj.com/ involvement only when the radius was also Pectus excavatum 18 (41) 1 (9) involved and with a lesser degree of severity. Hypoplasia pectoralis major 29 (66) 5 (45) Transverse reduction defects were not seen.

Minor abnormalities, including hypoplasia of the thenar eminence, limited supination of the on September 26, 2021 by guest. Protected copyright. forearm, and narrow sloping shoulders, occurred without other evidence of skeletal involvement. Severe reduction abnormalities were found in two familial cases (4 5%). Fig 1 shows typical hand abnormalities in HOS.

Thumb The thumb was the most commonly affected structure (84%). In contrast to the largest previous study,2 no patients had involvement of the thumb only. Even mild thumb hypoplasia was accompanied by hypoplasia of the thenar eminence (70%), limited supination ofthe forearm (61 %), or narrowing of the shoulders. Triphalangeal thumb was only found in eight patients (18%); 19 (43%) had absence of one or both thumbs. Other frequent thumb abnormalities included a hypoplastic or rudi- mentary thumb (17, 39%) or a long thumb (D) owing to elongation of the proximal phalanx (6, 14%). Five patients (11%) had syndactyly Figure 1 Typical hand abnornalities in HOS. Hands offour affected people showing (A) bilateral opposable triphalangeal thumbs, (B) right triphalangeal thumb, left absent of the first and second digits and seven patients thumb, (C) bilateral hypoplastic thumbs, and (D) normal thumbs with fifth finger (16%) had no detectable abnormality of the clinodactyly and brachydactyly of the second to fifth fingers. thumb. 302 Newbury-Ecob, Leanage, Raeburn, Young

Fingers Reduced shoulder movement, particularly ab- Hand measurements showed reduced finger duction, was found. Hypoplasia ofthe overlying length in 19 cases (43%), including four in shoulder girdle musculature, especially the pec- whom no thumb abnormality was present. toralis and the was common. major deltoids, J Med Genet: first published as 10.1136/jmg.33.4.300 on 1 April 1996. Downloaded from Hypoplasia of individual fingers (16%) most commonly involved the index and middle fin- Asymmetry. Abnormalities of the upper limb gers. Three patients had absence of one or were always bilateral and often asymmetrical more fingers. This occurred only where the (69 5%). Where this was the case the left side thumb was also absent. Clinodactyly of the was affected more than the right in 90O6% of fifth finger was a common finding (64%). cases and the right more than the left in 9 4%. Fig 1B shows typical hands with absence of the left thumb and a right sided triphalangeal Lower arm thumb. In fig 2 a severely affected child has In the lower arm, radial hypoplasia was seen phocomelia ofthe left arm but less severe radial in 18 cases (41 %) and was invariably ac- aplasia on the right. companied by a lesser degree of ulnar hypoplasia. No patients had involvement of the ulna Ten Isolated cases only. (23%) had complete radial aplasia. The of The commonest functional abnormality was spectrum abnormalities differed in the small group of isolated cases (table 3). In par- limited supination of the forearm seen in 27 four cases (61 %). This was not always associated ticular, patients had predominantly ulnar with radiological evidence of radioulnar aplasia with involvement of the ulnar ray of the syn- hand. One child had a ostosis but was sometimes associated with lim- unilateral transverse ited extension of the elbow. reduction anomaly ofthe second to fifth fingers with hypoplasia of the thumb and radial and ulnar hypoplasia, an anomaly not seen in any Upper arm familial cases. Five patients had exclusively Upper arm involvement unilateral defects. In general, abnormalities in leading to rhizomelic isolated cases were shortening was detected in almost half of more severe. For example, all 73% of the isolated cases had cases (45%), including four cases with no ob- aplasia of the thumb as to 43% of familial a vious radial hypoplasia. Phocomelia was seen opposed cases, in two familial cases (4 5%) (fig 2). greater proportion had radial and ulnar involvement, and two (18%) had bilateral phocomelia. In total, seven of the isolated cases were atypical for Holt-Oram syndrome as com- Shoulder with the Proximal reduction defects were not pared abnormalities seen in the familial present cases. Full details of these are in without distal defects, with the exception of given table 4. shoulder abnormalities which were especially http://jmg.bmj.com/ common (77%), giving a characteristic ap- pearance of narrow, sloping shoulders, this being because of the combination of short hypoplastic clavicles, hypoplasia of the head of the humerus, and decreased musculature. Al- ternatively, prominence of the lateral third of the clavicle and acromioclavicular joint gave a bony prominence to the shoulder (fig 3). on September 26, 2021 by guest. Protected copyright.

Figure 3 Typical shoulder abnormalities in HOS. _~~~~~J"S i'k. Narrow sloping shoulders, loss of usual contour owing to hypoplasia of the head of the humerus, prominent lateral Figure 2 Photograph and x ray of severely affected child showing left radial aplasia, third of clavicle, hypoplasia ofpectoralis major, and mild ulnar aplasia, and hypoplastic humerus, and right radial hypoplasia and absent first ray. pectus excavatum. Holt-Oram syndrome: a clinical genetic study 303

Table 4 Details of seven atypical cases cardiac defect; of these 11 required surgery in Case No Cardiac abnormality Predominant skeletal abnormality childhood and five were operated on in adult life. Ten had an asymptomatic structural lesion 1 Isolated Atrial septal defect Unilateral ulnar aplasia (VSD or MVP). All three patients with com- Ventricular septal defect J Med Genet: first published as 10.1136/jmg.33.4.300 on 1 April 1996. Downloaded from Pulmonary stenosis plete heart block became symptomatic and re- 2 Isolated Ventricular septal defect Unilateral ulnar aplasia quired insertion of a pacemaker. Aortic stenosis Detailed family histories showed that one 3 Isolated Dextrocardia Unilateral radial aplasia affected relative had died postoperatively fol- 4 Isolated Atrial septal defect Unilateral digital Ventricular septal defect reduction defect lowing closure of an ASD. Three patients from Aortic stenosis one family had died suddenly of undiagnosed 5 Isolated Atrial septal defect Unilateral radial aplasia heart abnormality. Necropsy on one, a child 6 Isolated Tetralogy of Fallot Bilateral radial aplasia aged 6 months, showed a large undiagnosed 7 Isolated Tetralogy of Fallot Bilateral radial aplasia ASD. No postmortem information was available on the other two cases who died suddenly.

Table 5 Cardiac abnornalities in 55 cases of HOS Defect No of affected (%) Isolated cases By definition all isolated cases had a structural Total Total 44 familial 11 isolated cardiac lesion. Five patients had an ASD, two with ASD alone, two with ASD and a VSD, Structural Secundum atrial septal defect 15 (34) 5 (45) and one with ASD and pulmonary stenosis (PS). Ventricular septal defect 11 (25) 4 (36) Two further patients had a VSD, one of whom Mitral valve prolapse 3 (7) Tetralogy of Fallot 2 (18) also had aortic stenosis (AS), and two cases had Aortic stenosis 2 (18) tetralogy of Fallot. There was one case with a Primum atrial septal defect 1 (2) Patent ductus arteriosus 1 (9) sinus venosus ASD, one with patent ductus ar- Sinus venosus atrial septal defect 1 (9) teriosus (PDA), and one case of dextrocardia Dextrocardia 1 (9) Pulmonary stenosis 1 (9) with otherwise normal cardiac anatomy. ECG abnormality only 17 (39) 0 (0) Normal heart 4 (9) 0 (0) ASSOCIATED ABNORMALITIES Eighteen (41%) of the familial patients had a pectus deformity ofwhom 15 had an underlying CARDIAC ABNORMALITIES (TABLE 5) Familial cases cardiac defect. A number of non-cardiac, non- Full cardiac details were available for 55 skeletal abnormalities were also observed, most patients and 95% had evidence of cardiac in- particularly in the eyes. Sixteen patients wore had normal clin- glasses for a refractive error, five had a squint, volvement. Only two patients and two had Duane anomaly. Skin findings ical examination, ECG, and echocardiography. http://jmg.bmj.com/ Of the familial cases, 25 (57%) had a structural included multiple pigmented naevi (1), large heart lesion, 22 (50%) had a septal defect, 15 cafe au lait pigmentation (1), haemangioma (34%) an ASD, and 11 (25%) VSD, including (1), and accessory nipple (1). four patients with both. One patient had a complete atrioventricular septal defect and complete heart block. Three cases had mitral Genetic aspects fam- MODE OF INHERITANCE valve prolapse (MVP): two from the same on September 26, 2021 by guest. Protected copyright. ily had MVP and first degree heart block and The pedigrees were consistent with autosomal one patient had MVP, an ASD, and complete dominant inheritance with many examples of heart block (CHB). In total, three patients had male to male transmission and no statistically complete heart block, two with septal defects significant difference between the numbers of and one without. Seventeen (39%) had no affected males and females. Twenty two affec- structural defect but had an abnormal ECG ted parents had a total of 36 affected and 36 with evidence of conduction defect (table 6). unaffected offspring. The commonest ECG abnormality was first No parental age effect was observed. This degree heart block (PR interval >0 2 sec) which was assessed by comparing the mean parental was seen in 71%. Right and left bundle branch age of sporadic cases with the mean general block, left axis deviation, and bradycardia were population parental ages. The mean paternal also seen. The majority (85%) had an asympto- and maternal ages in the study were 29-6 years matic cardiac lesion. Sixteen (38%) patients (range 26-35 years) and 27-6 years (range required cardiac surgery for correction of their 25-35 years) respectively, as compared with the general population figures of 30-19 and 27 15 years respectively. Table 6 ECG findings in 17 patients with no structural Penetrance was complete. Careful ex- lesion amination of all 17 obligate heterozygotes ECG Affected (%) showed clinically detectable abnormalities. Biological fitness was calculated from the Long PR interval 12 (71) Bradycardia 6 (35) numbers of offspring born to affected and un- Left axis deviation 7 (41) affected subjects. Twelve affected parents had Right axis deviation 8 (47) Left bundle branch block 2 (12) 13 unaffected sibs. The affected parents had P mitrale 1 (6) 24 offspring, indicating that there was no re- Complete heart block 1 (6) duction in fitness. 304 Newbury-Ecob, Leanage, Raeburn, Young

Table 7 Correlation between severity scores for cardiac 7' and skeletal abnormality A Skeletal score >_ 6

0 1 2 3 4 O a)I J Med Genet: first published as 10.1136/jmg.33.4.300 on 1 April 1996. Downloaded from Cardiac score 0 0 1 0 0 0 ) +5 1 0 7 7 3 0 i1 VCOc 2 0 2 5 3 0 co _X 3 0 2 3 9 2 s..cJ cn 4 4 0 0 0 0 0 E2n Spearman rank r=0-4894, p =0001. 0 0 Lt ) 3O

SEVERITY 2 When scored separately, skeletal and cardiac abnormalities showed a statistically significant 1 positive correlation (table 7). Although only 11 III two of the total number of affected patients Generation (4-5%) had severe phocomelia, 39% patients had a skeletal severity score of .3 indicating a serious reduction anomaly of the upper limb; 5.5F B 34% had abnormalities of the forearms and 5.0 _ hands only and 27% had only minor ab- > -_ normalities of the hands. For cardiac ab- ._a) 4)X normalities 36% scored 3, indicating a 4.5 H az + structural lesion requiring surgery or symp- C . 4.0 H tomatic arrhythmia, 23% scored 2 denoting o X, } 0E. 00 3.5 K 5.5 r -a5 8 L n 3.0 _ >- 5.0 H 2.5 _ g a) Sib 1 Sib 2 Parent Offspring CO CzO-~dv 'aen 4.5 Group az + K C0 Figure 6 Evidence for anticipation in HOS. Mean .- .s (A) E combined severity score for nine three generation families OcoX 4.0 H showing significant increase in severity in third generation. 0 0 Kruskall Wallis X2= 10 89, p= 0 0043. (B) Mean

combined severity score for sib pairs compared to http://jmg.bmj.com/ .8 LO CO parent-offspring pairs showing greater correlation between CD 3.5 H sibs than between parent and offspring.

3.0 Male Female an asymptomatic structural defect, and 39% Sex scored 1 owing to an ECG abnormality. There was no difference in severity between Figure 4 Mean combined severity score with 95% on September 26, 2021 by guest. Protected copyright. confidence interval for 22 females and 22 males showing males and females (fig 4). Mean scores for no difference in severity. (Mann-Whitney U= 221, p = offspring of affected mothers were greater than 0 52.) for offspring of affected fathers (fig 5). Analysis of severity of abnormalities in different generations within a family showed increasing 7.0 r severity in successive generations with a sig- 6.5 nificant increase in severity from generation I to III (fig 6A). There appeared to be a greater 6.0 correlation between severity in sib pairs than between parent and offspring (fig 6B). 0 fin 5.5 + C z .' 5.0 Discussion

co 0 C0 4.5 In this study we have reviewed the clinical and genetic features of HOS based on the largest Be. n ° 4.0 known series, consisting of 72 patients, 57 of LOl CO a) whom have been personally examined. This 3.5 has allowed us to define the clinical phenotype and delineate the spectrum of abnormalities 3.0 FIath- Mother Father seen in familial cases. In this paper we have Parent focused particularly on those clinical features which help to distinguish the diagnosis in isol- Figure 5 Maternal effect in HOS. Mean combined ated cases and those which enable affected severity score for 11 offspring of affectedfathers showing greater severity among offspring of affected mothers. parents to be alerted to the range of severity (Mann-Whitney U= 305, p= 0 0473.) likely to be seen in future affected children. Holt-Oram syndrome: a clinical genetic study 305

Table 8 Minimal diagnostic criteria for Holt-Oram diagnostic criteria (table 8). It has become syndrome evident that a number of isolated cases have Radial ray defect features outside this spectrum and possibly Septal defect (atrial or ventricular) or AV block in at least one represent other heart-hand associations. family member J Med Genet: first published as 10.1136/jmg.33.4.300 on 1 April 1996. Downloaded from

DIFFERENTIAL DIAGNOSIS LIMB ABNORMALITIES The differential diagnoses include the dom- These involved primarily the radial ray. Prox- inantly inherited heart-hand syndromes. Heart- imal abnormalities ofthe arm and forearm were hand syndrome II (Tabatznik) involves the always accompanied by distal abnormalities in combination of type D brachydactyly with the hand, though the thumb was not necessarily supraventricular tachycardia. 4 Patients may abnormal. The shoulders were often involved also have mild facial dysmorphism and mild even when involvement ofthe hands was minor. mental retardation. Ruiz De La Fuente and This feature seems to be specific to HOS and Prieto'5 described heart-hand syndrome III is not seen in cases ofisolated radial ray defects, with type C brachydactyly plus sick sinus syn- so the presence of a relatively subtle ab- drome. The long thumb brachydactyly syn- normality such as narrow sloping shoulders is drome described by Hollister and Hollister"6 a useful diagnostic pointer. Characteristically shows similarities to HOS. There is dominant there was asymmetry of the limb abnormalities inheritance ofa long thumb owing to elongation with a tendency for the left side to be more distal to the PIP joint with index finger bra- affected than the right. This contrasts with chydactyly, clinodactyly, narrow shoulders owing other radial ray defects which generally show to short clavicles, and pectus excavatum. right sided preponderance." Asymmetry is The cardiac abnormality is a conduction defect. otherwise unusual for single gene disorders There is also rhizomelic limb shortening. affecting the upper limb. Isolated cases of HOS need to be dis- tinguished from recessive conditions which include heart-hand abnormalities, for example, CARDIAC ABNORMALITIES thrombocytopenia-absent radius (TAR)." This The incidence of cardiac defects (95%) was differs from HOS in that radial aplasia is seen similar to that seen in previous studies.2 While with the thumb present. In HOS radial aplasia ASD was the structural lesion in the first family would invariably be accompanied by hypoplasia described,' subsequent reports have included or absence of the thumb. ASD may be seen VSD and other defects.29 Five families in this and also tetralogy of Fallot. A number of other study had no ASD. For example in one family abnormalities may be found in TAR, including a father with a conduction abnormality but no involvement of the lower limb, which are not septal defect had two sons both of whom had seen in HOS. Radial ray defects are well re- a same in Fanconi and VSD and the conduction disturbance. cognised syndrome'8 specific http://jmg.bmj.com/ In another family a father with a conduction investigation to exclude chromosomal breakage abnormality had a daughter with a VSD. All is required. Radial defects in the sporadic VAC- had radial ray abnormalities which were in- TERL association are more usually unilateral. 9 distinguishable from families with an ASD. Radial defects occur with Duane anomaly in Following localisation of a gene for HOS to the DR or Okihiro syndromes as well as being 12q, analysis in these two families was con- reported in HOS.20 Two patients (mother and sistent with linkage to this locus. This supports son) from this series had Duane anomaly and the notion that an ASD is not a requirement HOS. Both had cardiac and upper limb in- on September 26, 2021 by guest. Protected copyright. for the diagnosis. volvement. There was no history of deafness. All families had at least one member with a septal defect. First degree heart block was the commonest defect and may be seen with or VARIATION IN EXPRESSION without a septal defect. Two affected subjects Our observations have confirmed previous re- from one family showing linkage to 12q had ports of marked intra- and interfamilial vari- mitral valve prolapse in conjunction with first ation with cardiac involvement varying from degree heart block. Mitral valve prolapse has asymptomatic arrhythmias to large or multiple previously been described in HOS.29 Seven septal defects leading to death in infancy. Skel- patients had a combination ofleft axis deviation etal defects ranged from minor features such with first degree heart block. This ECG pattern as clinodactyly or hypoplasia of the thenar em- may be seen in families with dominantly in- inence to severe reduction defects of the upper herited non-syndromic ASD.'2 A conduction limb. In larger families there was evidence of defect may therefore be a forme fruste of HOS, a specific intrafamilial pattern as has been de- possibly reflecting previous spontaneous clos- scribed previously.2 22 Three families showed a ure of a septal defect. None of the rarer cardiac tendency to more severe cardiac involvement conditions seen in the isolated group and re- and mild limb abnormalities, whereas 12 ported previously in isolated cases, for example, showed more severe limb defects and less severe tetralogy of Fallot,'3 was seen in the familial cardiac involvement. group in whom the most complex defect was an atrioventricular septal defect. Data from this study have enabled us to INCIDENCE OF SEVERE ANOMALIES define the HOS phenotype based on the fea- Families seeking information regarding HOS tures seen in familial cases and to set minimal usually want to know both the spectrum of 306 Newbury-Ecob, Leanage, Raeburn, Young

the condition and the likelihood of serious Table 9 Severity of skeletal defect as a function of abnormality. In this study one in three cases cardiac defect had a significant reduction deformity of the Cardiac defect Mean severity score upper limb although only 1 in 22 had pho- Skeletal defect J Med Genet: first published as 10.1136/jmg.33.4.300 on 1 April 1996. Downloaded from comelia. The only previous estimate of the Atrial septal defect 2-66 incidence of phocomelia was 1 in 10.3 This Ventricular septal defect 2-28 difference may reflect underascertainment of No structural defect 1-89 milder cases in previous studies. The most severe heart cases in cardiac terms were three premature deaths and two children who required surgery in the first year of life because dominant disorder with a severe phenotype one of severe cardiac failure. Many septal defects would expect to see such an increase in severity either did not require treatment or, in the case since mildly affected cases are more likely to of ASD, were treated electively in childhood reproduce, and following the birth of a severely or adulthood. affected child family size is restricted. However, A statistically significant correlation was ob- fig 6B indicates that there is a greater cor- served between the severity of limb and heart relation between offspring than between parent abnormalities in a given person. No difference and offspring, which is supportive evidence for in severity between the sexes was observed. anticipation. The mean severity score for each This is in contrast to previous studies which generation increased significantly although not have shown greater severity in females.38 all individual meioses led to an increase in severity score. Maternal meioses were never accompanied by a decrease in severity. How- GENETIC ASPECTS ever several paternal meioses did lead to a The pedigrees were clearly indicative of auto- decrease in severity as measured by this somal dominant inheritance. A previous finding method. This does not in itself rule out an- of abnormal segregation with more affected ticipation, as similar observations have been than unaffected offspring born to affected par- made in myotonic dystrophy where the pro- ents' was not confirmed. It has been suggested posed underlying genetic mechanism is a "re- in previous studies that abnormal segregation treat of the triplet repeat".27 is the result of maternal fetal loss.2324 No such The evaluation ofanticipation and a parental effect was observed in this study. effect in a genetic disorder usually involves Penetrance was found to be complete. Al- documentation of age of onset. This approach though x rays sometimes showed additional cannot be applied to a congenital abnormality findings, all obligate heterozygotes showed clin- for which ranking or grading of severity for ical abnormalities on detailed examination. No statistical analysis is necessary. Complete ob- examples of non-penetrance or gonadal mo- jectivity in clinical assessment of severity is saicism were encountered. Previous reports of difficult to achieve. However, our observations the proportion of isolated cases attributable to are consistent with both anticipation and a http://jmg.bmj.com/ new mutations vary from 03 to 0O85. Of the maternal effect although we recognise that con- cases referred to in this study only 8% were firmation or otherwise awaits isolation and considered to be sporadic after examination of characterisation of the HOS gene(s). both parents and investigation with x ray, ECG, Meanwhile prospective parents can be alerted and echocardiography. This lower figure may to the fact that a child born with HOS to an reflect incomplete ascertainment because of affected parent has a 1 in 3 chance of having preferential referral of families for linkage ana- a severe reduction abnormality of the upper on September 26, 2021 by guest. Protected copyright. lysis in our study. In three cases referred with no limb with a 1 in 22 risk of phocomelia. If an significant family history, clinical abnormalities ASD is present there is a greater risk of a were found in one of the parents. serious limb abnormality than if a VSD or conduction defect occurs alone (table 9). Se- verity is likely to be greater if the transmitting ANTICIPATION parent is female. In our study there were no Several reports offamilies with HOS have men- cases of severe limb abnormality occurring in tioned an observed tendency to increasing isolation. Therefore the detection of a severe severity of abnormalities in successive reduction defect before or after birth indicates generations.462425 McKusick26 was the first to that there is a high chance that a structural report this in a family in which the grandparent cardiac lesion will also be present. had an ASD and triphalangeal thumbs, the The authors with to thank the clinicians who allowed us to parent had an ASD and absent thumbs, and the approach their patients including Dr M Baraitser, Professor J Burn, Dr N Dennis, Professor D Donnai, Dr P Farmdon, grandchild was born with a VSD and bilateral Professor P Harper, Dr H Hughes, Dr S Huson, Dr R Mueller, phocomelia. We examined this phenomenon in Dr V Murday, Dr 0 Quarrell, Dr J Sampson, Dr J Tolmie, and the late Dr David Siggers; Mr K Hosie for his help visiting this study. Although families with three or more families; Dr E Blank for his helpful comments regarding this affected generations showed increasing severity manuscript; Dr C A Parkin and Dr K Martin for the cytogenetic analyses; and Mrs R Goodwin for typing the manuscript. We of both skeletal and cardiac abnormalities as are most grateful to the families and patients with HOS for measured by severity scores (fig 6A) this is not their encouragement and support. This research was supported by the British Heart Foundation and the Trent Health Dir- in itself proof of anticipation. We would expect ectorate of Research and Development. a degree of ascertainment bias since previous generations are often detected retrospectively 1 Holt M, Oram S. Familial heart disease and skeletal ma- nifestations. Br Heart J 1960;22:236-42. and families are more likely to be identified if 2 Smith AT, Sack GH, Taylor GJ. Holt-Oram syndrome. J the index case is severe. Furthermore, in a Pediatr 1979;95:538-43. Holt-Oram syndrome: a clinical genetic study 307

3 Holmes LB. Congenital heart disease and upper-extremity 15 Ruiz De La Fuente S, Prieto F. Heart-hand syndrome III. deformities. A report of two families. N Engl Med 1965; A new syndrome in three generations. Hum Genet 1980; 272:437-44. _J 55:43-7. 4 Gardner RJM, Buckfield PM, Veale AMO. A girl with severe 16 Hollister DW, Hollister WG. The "long thumb" bra- expression of the Holt-Oram syndrome. Clin Genet 1973; chydactyly syndrome. Am J Med Genet 1981;8:5-16. 4:417-21. 17 Hall JG. Thrombocytopenia and absent radius (TAR) syn- 5 Sanchez-Cascos A. Holt-Oram syndrome. Acta Paediatr drome. JMed Genet 1987;24:79-83. J Med Genet: first published as 10.1136/jmg.33.4.300 on 1 April 1996. Downloaded from Scand 1967;56:313-17. 18 Glanz A, Fraser FC. Spectrum of anomalies in Fanconi 6 Lewis KB, Bruce RA, Baum D, Motulsky AG. The upper anaemia. _7 Med Genet 1982;19:412-16. limb-cardiovascular syndrome. An autosomal dominant 19 Khoury MJ, Cordero JF, Greenberg F, James LM, Erickson genetic effect on embryogenesis. JAMA 1965;193: 1080-6. JD. A population study of the VACTERL association. 7 Terrett JA, Newbury-Ecob E, Cross GS, et al. Holt-Oram Pediatrics 1983;71:815-20. syndrome is a genetically heterogeneous disease with one 20 Chang CHJ. Holt-Oram syndrome. Radiology 1987;88:470- locus mapping to human chromosome 12q. Nature Genet 83. 1994;6:401-4. 21 Basson CT, Cowley GS, Solomon SD, et al. The clinical 8 Gall JC, Stem AM, Cohen MM, Adams MS, Davidson and genetic spectrum of the Holt-Oram syndrome (heart- RT. Holt-Oram syndrome: clinical and genetic study of hand syndrome). N Engl _J Med 1994;330:885-91. a large family. Am_Hum Genet 1966;18:187-200. 22 Kaufman RL, Rimoin DL, McAlister WH, Hartmann AF. 9 Gladstone I, Sybert VP. Holt-Oram syndrome: penetrance Variable expression of the Holt-Oram syndrome. Am. of the gene and lack of maternal effect. Clin Genet 1982; 21:98-103. Dis Child 1974;127:21-5. 10 Hall JG, Froster-Iskenius UG, Allanson JE. Handbook of 23 Brons JTJ, Van Geijn HP, Wiladimiroff JW, et al. Prenatal normal physical measurements. Oxford: Oxford University ultrasound diagnosis of Holt-Oram syndrome. Prenat Press, 1989. Diagn 1988;8:175-81. 11 Carroll RE, Louis DS. Anomalies associated with radial 24 Van Regemorter N, Haumont D, Kirkpatrick C, et al. dysplasia. _7 Pediatr 1974;84:409-11. Holt-Oram syndrome mistaken for thalidomide em- 12 Bosi G, Sensi A, Calzolavi E, Scortano M. Familial atrial bryopathy - embryological considerations. Eur _7 Pediatr septal defect with prolonged atrioventricular conduction. 1982;138:77-80. 25 Temtamy SA, McKusick VA. The genetics of hand mal- Am _J Med Genet 1992;43:641. 13 Brans YW, Lintermans JP. The upper limb cardiovascular formations. Birth Defects 1978;XIV(3):117-33. syndrome. Am _7 Dis Child 1972;124:779-83. 26 McKusick VA. Medical genetics. _7 Chronic Dis 1961;14: 14 Silengo MC, Biagioli M, Guala A, Lopez-Bell G, Lala R. 1-198. Heart-hand syndrome II. A report of Tabatznik syndrome 27 Brook JD. Retreat of the triplet repeat. Nature Genet 1993; with new findings. Clin Genet 1990;38:105-13. 3:279-87. http://jmg.bmj.com/ on September 26, 2021 by guest. Protected copyright.