genet smith-lemli-opitz syndrome: variable clinical …smith-lemli-opitz syndrome table 1...

58Med Genet 1998;35:558-565

Smith-Lemli-Opitz syndrome: a variable clinicaland biochemical phenotype

A K Ryan, K Bartlett, P Clayton, S Eaton, L Mills, D Donnai, RM Winter, J Burn

AbstractWe have reviewed all known UK cases ofSmith-Lemli-Opitz syndrome. Among 49cases with proven 7-dehydrocholesterolreductase deficiency, half had been termi-nated or had died in infancy. The mini-mum incidence is 1 in 60 000. The frequentoccurrence of hypospadias may accountfor 71% of recognised cases being male.Important common features whichemerged include short thumbs, severephotosensitivity, aggressive behaviour, andatrioventricular septal defect. The typicalfacial appearance becomes less obviouswith age and 20% of cases did not have 2/3toe syndactyly. Biochemical measure-ments of serum 7-dehydrocholesterol didnot correlate with clinical severity.(7Med Genet 1998;35:558-565)

Keywords: Smith-Lemli-Opitz syndrome; mental retar-dation; cholesterol; 7-dehydrocholesterol

Department ofHumanGenetics, University ofNewcastle upon Tyne,19 Claremont Place,Newcastle upon TyneNE2 4AA, UKA K RyanJ Burn

The James SpenceInstitute of ChildHealth, Royal VictoriaInfirmary, Newcastleupon Tyne, UKL BartlettS Eaton

Institute of ChildHealth, London, UKP ClaytonS EatonK MillsRM Winter

Regional GeneticService, St Mary'sHospital, Manchester,UKD Donnai

Correspondence to:Dr Ryan.

Received 22 October 1997Revised version accepted forpublication 5 January 1998

Smith-Lemli-Opitz (SLO) syndrome is an

autosomal recessive mental retardation/multiple congenital anomaly syndrome firstdescribed by Smith et al' in 1964. Theincidence is estimated to be 1 in 20 000 to 1 in50 000 with a carrier frequency of 0.014.2 3

Before 1993, diagnosis was based on character-istic clinical features, including mental retarda-tion, failure to thrive, dysmorphic facies, cleftpalate, congenital heart disease, hypospadias,and 2/3 toe syndactyly. There is a wide clinicalspectrum, ranging from mild developmentaldelay and syndactyly to lethal multiorgananomalies.2 4

In 1993 and 1994, Irons et ar and Tint et altshowed that very low levels of cholesterol andmarkedly raised levels of the cholesterolprecursor 7-dehydrocholesterol (7DHC) were

present in subjects with both the severe and themilder forms of SLO. In 1995, Honda et af

showed that SLO resulted from a primarydefect in the enzyme 7-dehydrocholesterolreductase, which normally converts 7DHC tocholesterol. Hypocholesterolaemia and raised7DHC are present in blood and other tissues.Cholesterol is vital for the normal structure ofcell and mitochondrial membranes, develop-ment of the central nervous system, and myeli-nation, and for normal synthesis and metabo-lism of steroids, bile acids, and vitamin D. It ispostulated that the excess of 7DHC or

deficiency of cholesterol or both results inabnormal plasma membrane and myelin for-mation. The protein product of the vertebrateearly embryonic patterning gene, Sonic Hedge-hog, undergoes autoproteolysis and covalent

linkage with cholesterol, forming a cholesterolmodified active product. Holoprosencephalyand other malformations present in SLO maybe caused by incomplete or abnormal modifi-cation of the Sonic Hedgehog protein.8 Al-though a deficiency of the 7DHC reductaseenzyme is suspected as the primary abnormal-ity, little is currently known about the functionand structure of the enzyme or its regulation.Work is presently under way to fully character-ise and purify the enzyme. Positional cloningtechniques are currently under way to identifyputative genes in the 7q32.1 region, followingthe discovery of a de novo balanced transloca-tion in a patient reported by Wallace et ar andother patients with chromosome rearrange-ments in the 7q region." To date, themetabolic defect has not been associated withany gene in the 7q32 region.

It has been suggested that dietary interven-tion at an early age may improve theneurological and physical outcome in thesechildren.4 12 Trials of high dose cholesterol andbile acid supplementation have been intro-duced to a small number of patients andpreliminary reports suggest that the childrenare better behaved and there are some reportsof acceleration of neurological progress. Con-clusions on the value of therapy are dependentupon a clear delineation of the natural historyof the disease.With the advent of a biochemical marker for

diagnosis, the "well defined clinical spectrum",born of "minimal diagnostic criteria",13 hasbecome amenable to independent evaluation.We have reviewed all known SLO patients inthe UK in order to establish the clinicalspectrum.

Subjects and methodsPatients were ascertained through geneticistsand paediatricians in the UK and through theUK SLO support group. Ethical approval forthe study was obtained from the Newcastle andNorth Tyneside Health Authority Joint EthicsCommittee. A total of 86 cases were initiallyidentified as having "SLO". Thirty-seven caseswere excluded from the study for a variety ofreasons. One patient with a biochemically con-firmed diagnosis refused to take part in thestudy. Six patients had had a clinical diagnosisbut were found not to have SLO on biochemi-cal testing. Two adults with a clinical diagnosiswere not tested biochemically because ofparental wishes. There were 28 cases who haddied on whom we were unable to obtain tissuesamples for biochemical confirmation; of these,11 had features in keeping with the diagnosis ofSLO.

558

on May 22, 2020 by guest. P

rotected by copyright.http://jm

g.bmj.com

/J M

ed Genet: first published as 10.1136/jm

g.35.7.558 on 1 July 1998. Dow

nloaded from

http://jmg.bmj.com/

Smith-Lemli-Opitz syndrome

Table 1 The main clinicalfeatures in 49 SLO subjects in this series and comparison withprevlous studies

Percentages ofpreviousClinicalfeatures No of affected* Percentage studies

GeneralMales 35 71 64t, 79:Mental retardation 23/25 92 lOOt, 100§Blonde hair 17/26 65Photosensitivity 13/24 54Abnormal sleep pattern 16/23 70

FaceMicrocephaly 32/40 80 63:Ptosis 26/44 59 321:Cataracts 6 12 18t, 22§, 53tAnteverted nares 34 69 841:Low set ears 23 47Cleft palate (hard or soft) 18 37 52t, 68:Tongue abnormalities 6 12 63tMicrognathia 33 67 100:t

Skeletal abnormalitiesPostaxial polydactyly 24 49 43t, 52§, 95:tToe syndactyly 43/48 90 89:t, 99t, 100§Short/proximally placed thumb 24 49 63:

Other abnormalitiesHeart defect 18 37 36§, 38t , 84tGenital anomaly 32/35 91 100:Renal anomaly 14 29 13t, 40§, 47tLung anomaly 12 24 33§, 56:Gastrointestinal abnormality 15 31 22§, 25t, 68:Brain structural anomaly 6/28 21 16§,26:

*Unless otherwise stated, the number of subjects with each abnormality is from a total of 49 cases.tCunniff et al,4 80 patients from the Kennedy Kreiger Institute, Baltimore.1:Curry et al,'" 19 severely affected patients from the USA.§Lin et al,2' 59 patients from the Kennedy Kreiger Institute, Baltimore.

Forty-nine cases from 43 sibships wereincluded in the study. Most patients had aninitial clinical diagnosis which was confirmedbiochemically following availability of the7DHC assay. Where the patient had died, ret-rospective biochemical confirmation of clinicaldiagnosis was obtained from tissue samples.However, if no tissue samples were available,parents were tested for raised 7DHC, indicat-ing carrier status (three cases).4 Prenatal diag-nosis was performed on amniotic fluid in onefetus. Plasma samples were collected and keptfrozen at -20°C until analysis. Serumcholesterol and 7DHC measurements wereperformed using gas chromatography-massspectrometry (GC-MS) of plasma sterolsfollowing solvent extraction.6 It is importantthat serum cholesterol measurements are madeby stable isotope dilution by GC-MS, as stand-ard cholesterol methods report serumcholesterol as the sum of cholesterol + 7DHC+ 8DHC. We have banked DNA and obtained

Figure 1 Frequency of severity scores for 49 SLO subjects.Malformations of the brain, oral cavity, heart, kidney,gastrointestinal tract, liver, lung, bowel, genitals, eyes, andskeleton were scored as O=normal, 1 =minor malformation,and 2=major malformation. These were added to obtain ascore of between 0 and 20.

lymphocyte cell lines from 20 patients and 14sets of parents.Twenty-four patients were alive, 20 were

dead (including one stillbirth), and five fetuseswere terminated. All patients' medical andgenetic notes were reviewed and all livingpatients were examined personally by one ofthe authors (AR). Seven of the cases in thisseries have been previously reported; four bySeller et all 16 (two unrelated fetuses and twosibs who died early in the neonatal period) andthree by McGaughran et al7 (a family of twosibs and prenatal diagnosis in a third termi-nated fetus).

All cases were assigned a severity score,initially devised by Bialer et al'5 to quantifydegree of severity and recently modified byKelley to remove the bias for early death (Kel-ley, personal communication). Malformationsof the brain, oral cavity, heart, kidney, liver,lung, bowel, genitals, eyes, and skeleton werescored as O=normal, 1 =minor malformation,and 2=major malformation. These were addedto obtain a score of between 0 and 20.

Parents of the 24 living patients were givenbehavioural and sleep questionnaires to com-plete. The behavioural questionnaire askeddetailed questions on developmental assess-ment (vision and hearing, manual dexterity,mobility, feeding, continence) and behaviour(feeding, dressing, communication, education,sleep, social behaviour, activities, movements,interests, self injury, aggression, pain aware-ness, emotional state). The sleep questionnaireasked for details on present and past sleepinghabits and behaviours and treatment of sleepdisturbances.

ResultsIn some instances, data were not available orwere not applicable, for example, the presenceof ptosis cannot be determined in fetuses andfeeding data were not available in those infantswho died in the early neonatal period. We havegiven numbers of patients with each abnormal-ity for each section, and where the denomina-tor is not the total 49 patients, this is specified.The main clinical features are summarised intable 1.

INCIDENCEFifteen subjects (10 male, five female) wereknown to have been born or terminated in theUK in 1995 and 1996. The male birth rate inthe UK is approximately 300 000 births peryear. Therefore, the minimum incidence is1/60 000 males per year and the minimumbirth prevalence is 1/67 000 males per year.Assuming females and males are affected inequal numbers (but that more males are diag-nosed because of the presence of genitalabnormalities), this is equivalent to a minimumcarrier frequency of 1 in 122 (0.8%). The par-ents of one child were distantly related.

NATURAL HISTORYThe 49 cases in this study were from 43sibships. There were 35 males and 14 females(71% males). There were 113 known concep-tions in these 43 sibships: 90 births (39

559



g.bmj.com

/J M


g.35.7.558 on 1 July 1998. Dow

nloaded from

http://jmg.bmj.com/

Ryan, Bartlett, Clayton, et al

.~~~~~~~~~ ~............

Figure 2 (A) An 18 month old male with blonde hair downward slanting palpebralfissures, bilateral ptosis, and large right ear. (B) A 16 month oldfemale with blonde hair,left sided ptosis, a small nose with anteverted nares, and low set ears. (All photographsreproduced with permission.)

probands), 16 miscarriages, and seven termi-nated fetuses. Subsequent to the probands'births, there were nine affected SLO cases from28 conceptions (28%); six were liveborn andthree were terminated. Three of these subse-quently affected cases (one liveborn, twoterminated fetuses) were not included in thestudy because of insufficient clinical detail.For 44 pregnancies, a maternal obstetric his-

tory was available. Of these, 16 (36%) were

complicated by pregnancy induced hyperten-sion, oligohydramnios, or abnormal scans.Polyhydramnios was present in only one preg-

nancy. Five pregnancies were complicated byoligohydramnios; all had either hypoplastickidneys or bilateral renal agenesis. The

Figure 3 This boy has the typical "SLO face". He hasmicrocephaly, a narrow bitemporal diameter, downwardslanting palpebralfissures, ptosis, a short nose withanteverted nares, micrognathia, and low set ears.

antenatal ultrasound showed intrauterinegrowth retardation in seven patients andcongenital abnormalities in seven patients (onepregnancy had both features). Based on infor-mation from 30 mothers, half reported reducedfetal movements. Nuchal oedema was presentin two infants at birth and was present atnecropsy in another fetus at 22 weeksgestation.'5 Of the 41 patients for whom gesta-tion was known, seven (17%) were bornprematurely, before 37 weeks' gestation. In 39patients, delivery data were available; nine wereborn by caesarean section (23%) and 11patients (28%) were a breech presentation.

Five fetuses were terminated between 17 and23 weeks' gestation, owing to congenitalabnormalities detected on antenatal ultra-sound (4) or because of confirmation of SLOby amniocentesis following a previously af-fected sib (1). One infant was stillborn follow-ing intrauterine fetal death at 30 weeks' gesta-tion. There were 19 postnatal deaths; seven(37%) occurred within the first day of life, 12(63%) within the first month, and 18 (95%)within the first year. The oldest death was at 14years. Death was the result of congenital heartdisease or multiple congenital abnormalitiesincluding lung and renal hypoplasia in themajority of cases. Twenty-four patients werealive, with ages ranging between 4 months and38 years at the time of examination; the major-ity were aged less than 12 years. There were fiveadults, aged 22 to 38 years.The majority of patients were diagnosed as

having SLO at necropsy or within the firstmonth of life. All subjects were scored to ratetheir severity of congenital anomalies (range2-15, mean=6.51, SD 2.84) (fig 1). Themajority of patients had one or more majorstructural congenital abnormality in additionto facial dysmorphism and developmentaldelay. There were 11 patients without majororgan malformations but with one or moreminor malformations. For example, one malepatient had bilateral 2/3 toe syndactyly, righttalipes equinovalgus, mild dysmorphic fea-tures, mild learning difficulties, feeding diffi-culties, and ptosis (fig 2A). Another male infanthad congenital dislocation of the hip, bifiduvula, right undescended testis, and facial dys-morphism. One female had mild developmen-tal delay, bilateral 2/3 toe syndactyly, shortthumbs, mild dysmorphic features, and failureto thrive, but did not require nasogastric tubefeeding (fig 2B).Assessment of severity of mental retardation

in those patients living at least six months (25patients) was based upon developmental mile-stones, medical records, school performancereports, formal developmental assessment, andphysical examination. Two were in the normalrange, although they were aged only 8 monthsand 10 months at the time of assessment.Twenty-three of 25 (92%) had mental retarda-tion; four (16%) had mild, seven (28%) hadmoderate, and 12 (48%) had severe mentalretardation.There were 11 sibs from five families. Three

sets of sibs are alive and two sibships of two andthree sibs (two sibs and one terminated fetus)

560



g.bmj.com

/J M


g.35.7.558 on 1 July 1998. Dow

nloaded from

http://jmg.bmj.com/


Figure 4 The same male photographed at 3 months, 4years, 8 years, atnd 26 years. The typicalface of childhoodbecomes less obvious with time.

all died within the first month of life. The phe-notype was broadly concordant within sib-ships; their severity scores were (3,5), (9,9),(5,7), (8,6,8), and (6,3).The majority of patients had severe feeding

difficulties in infancy with poor suck, lack ofinterest in feeds, and vomiting. All but sixpatients required nasogastric tube feeding forbetween 4 months and 8 years. The vomitingand feeding difficulties tended to improve as

the children grew older, though gastrostomy/fundoplication was required in eight patients.Medical treatment for gastro-oesophageal re-

flux was generally ineffective.

PHYSICAL FINDINGSGrowthGrowth data were available for all 24 livingpatients. Average birth weight was 2700 g at 38weeks' gestation. Failure to thrive was seen in21/24 (88%), who were below the 3rd centilefor height/weight parameters. The majoritywere born with heights and weights within thenormal range but gradually fell across and then

Table 2 Range of oral abnormizalities in 49 SLO subjects

7No of affected

TongueHypoplastic 4Adherent to floor of the mouth IBifid 1Longitudinal grooves 3

TeethCrowded 3Large or widely spaced 4

Palate abnormalitiesCleft of the hard palate 7Cleft of the soft palate 11Cleft uvula 5High arched palate 14

below the centile charts by 6 months of age.Adult height ranged between 150 and 160 cmfor four males and was 152 cm in the only adultfemale (all <3rd centile).

FacialfeaturesFig 3 shows the typical facial appearance in latechildhood while fig 4 shows the evolution offacial appearance with age. Microcephaly withbitemporal narrowing was present in 32/40(80%). In the eight patients whose headcircumference was >=3rd centile, the majoritywere less than or equal to the 10th centile andonly one patient had a head circumference>50th centile.

Ptosis was present in 26/44 patients (59%)and in three it was unilateral. The palpebralfissures were either upward or downwardslanting in 13/49 patients (27%). Six of 49(12%) patients had cataracts (two unilateral,four bilateral) and 10 patients had hyperte-lorism. Vision was generally reported asnormal, although eight had strabismus. Therewere no reports of coloboma or retinalabnormalities.The nose was typically small with a broad,

flat nasal bridge and the nares were antevertedin 34 patients (69%), although this featurebecame less obvious with age. The ears, usuallyof normal size and shape, were low set in 23patients (47%) and were posteriorly rotated in16 patients (33%). The cheeks often had a fullappearance. Micrognathia was present in 33(67%) patients and became less obvious withage.Mouth shape was usually normal, but a few

patients had large or small mouths. Sevenpatients had dental abnormalities consisting ofeither large central front teeth or crowded teethas a result of a small mouth or palate (table 2).Six patients had hypoplastic or groovedtongues or both (bifid tongue in one patient).Thirty-seven (76%) had a palatal abnormality;18 (37%) had either a cleft of the soft or hardpalate and the remainder had either a cleftuvula or high arched palate. In addition to this,18 (37%) patients had broad alveolar margins,which, like the palate, was typically rugose inappearance (fig 5).

Cardiac defectsEighteen patients (37%) had congenital heartdisease. Atrioventricular septal defect waspresent in six patients (12%) and was associ-ated with coarctation in two cases. Two othercases had isolated coarctation and two had leftventricular hypoplasia. Three infants hadpatent ductus arteriosus (not associated withprematurity) and there was one of each of thefollowing: atrial septal defect, right ventricularhypoplasia, aortic stenosis, complex disease,and unknown type.

Skeletal abnornmalitiesSkeletal abnormalities were a frequent finding,including shortened limbs in 10 patients and abroad range of mild skeletal abnormalities,listed in table 3. Postaxial polydactyly waspresent in 26 patients (53%) and was seen twiceas frequently in the upper limbs compared to

'WI :-

#VP-,

561



g.bmj.com

/J M


g.35.7.558 on 1 July 1998. Dow

nloaded from

http://jmg.bmj.com/


Figure 5 Cleft of the soft palate, showing the typicalrugosity of the alveolar ridges and palate.

the lower limbs. Soft tissue syndactyly was seenin 43/48 patients (90%). The "classic" 2/3 toesyndactyly was seen in 39/48 (81%) patients(three unilateral, 36 bilateral) and, whilepresent in isolation in the majority, it was alsoseen in combination with various patterns of toesyndactyly. In only two patients was mild 3/4hand syndactyly present.

Genital anomaliesOf the 35 males in our series, 32 (91%) hadgenital abnormalities. Eight had phenotypicsex reversal (23%) and 11 (31 %) had ambigu-ous genitalia. Of the 13 remaining males, fourhad hypospadias and cryptorchidism, sevenhad hypospadias, and two had cryptorchidism.All females had apparently normal genitalia.Two adult males had delayed puberty; one ofthese males had had prophylactic bilateralorchidopexy as a child.

Skin and appendagesA total of 15/24 (63%) patients had markedphotosensitivity; one child has been shown tohave an ultraviolet A light photosensitivity (DrA Anstey, personal communication). All arereported to have marked dermal erythematousreactions on exposure to sunlight. Eczema wasreported in four patients. Of the 26 patientswhose hair colour was recorded, 17 (65%) hadblonde hair, eight had brown hair, and one hadred hair. The fair colouring was often incontrast to hair colouring in first degreerelatives.

Table 3 Skeletal abnormalities in 49 SLO subjects

No ofaffected Percenitage

GeneralDislocated/subluxed hips 8/44 18Symmetrical limb shortening 6 12Upper limb shortening 4 8

Hand abnormalitiesShort/proximally placed thumbs 24 49Postaxial polydactyly 24 49Brachydactyly of 1 or more fingers 10 20Ulna deviation of fingers 7 14Clinodactyly 4 8Overriding index finger 3 6

Foot abnormalitiesSoft tissue syndactyly 43/48 90Postaxial polydactyly 14 29Talipes calcaneovalgus 13 27Short/broad toes 8 16Overriding toes 3 6Varus deformity 2 4

Miscellaneous featuresFourteen (29%) patients had structural renalabnormalities, consisting of bilateral renalagenesis/hypoplastic kidneys (6), single kidney(3), cystic kidneys (2), lack of fetal lobulations(1), and renal tract dilatation (2). In twopatients with renal hypoplasia, the ureters andbladder were also hypoplastic. Functional gas-trointestinal problems were frequently seen.Constipation was a major problem in 11patients, while three had chronic diarrhoea.One patient with intractable diarrhoea hadmarked improvement in symptoms oncecholesterol therapy was implemented. Struc-tural anomalies were present in 14 (29%)patients, and several had more than oneaffected organ. Defects observed were malrota-tion (4), pyloric stenosis (6), and Hirschs-prung's disease (4). Liver abnormalities wereinfrequent, with cholestatic liver disease beingpresent in three patients and only one patienthaving an atretic gall bladder. Ten patientswere reported as having frequent lower respira-tory tract infections; these were often the resultof aspiration of feeds and tended to improve asvomiting decreased or following gastrostomyinsertion. Twelve patients had hypoplastic/incomplete lobulation of the lungs. Six patientshad excess skin folds/nuchal oedema over thenape of the neck and skin webbing was noted inthree patients (groin, neck and axilla, andelbow).

Brain structureTwenty-eight patients had either brain imagingor postmortem examinations. Apart frommicrocephaly, malformations were present insix patients (21 %). These consisted of myelinmaturation delay (2), lissencephaly and agen-esis of the corpus callosum (1), hypoplasia ofthe corpus callosum (1), choroidal cyst (1), anda hypoplastic cerebellum (1). Of note, holo-prosencephaly was not reported in any patient.Seizures were seen transiently in only twoinfants.

BEHAVIOURAL PHENOTYPEA pattern of behavioural characteristicsemerged from questionnaires completed by theparents of 23 living children and adults aged 6months and older. A striking feature was theabnormal sleeping pattern, seen in 16 (70%)patients. As infants, the majority were sleepyand floppy. In the first few years of life, childrenfrequently slept as little as two to three hours anight, without catch up sleep during the day.The majority were difficult to settle and wouldwake during the night and in the early hours ofthe morning. Sedatives were ineffective in allbut one case. Most children gradually requiredmore sleep as they aged, sleeping for seven toeight hours a night by the age of 5-6 years,although one boy was 9 years old before hissleeping pattern normalised. Common behav-ioural characteristics included self injuriousbehaviour in eight (35%) and aggressivebehaviour in 12 (52%) (fig 6). Ritualisticbehaviours were seen in 12 (52%); these com-monly included playing the same audio orvideo cassette repeatedly, obsessions about

562



g.bmj.com

/J M


g.35.7.558 on 1 July 1998. Dow

nloaded from

http://jmg.bmj.com/


Figure 6 This 25 year oldfemale has blonde hair, micnose with anteverted nares, and micrognathia. The abrcself injurious behaviour.

placement of objeexample, spinningMany of the childand loving to taffectionate withparents reported texcessively sensitiN

LABORATORY DATA

Forty-seven patieformed. All were i

abnormalities at 7determined by Y1logical tissue.

In three cases, twas confirmed aftto have raised 7Dparental heterozyg46 cases, all had ration of one case,8DHC (the 8-dewithout any dete(had a severe, butpatients, both preiand 7DHC valuesserum cholesterolmmol/l (mean= 1.

ment 7DHC rang(mean=397, SD 2ues against each oi

tested by first ordenificant correlatiou

in

a)

oin

C/)

a)

C,)

14-

12-

10

8

6L-

2

8*_

0 200 4

Figure 7 Very weak coand severity score in 19

_... variables: 7DHC v cholesterol (r2=0.29),7DHC v severity (r2=0. 16), and cholesterol vseverity (r2=0.33). Fig 1 shows the frequencydistribution of clinical severity while fig 7shows the very weak correlation of severityscores with 7DHC levels.

DIETARY INTERVENTIONFifteen patients were on cholesterol supple-

-=--aJ u mentation. While there has been no dramaticimprovement in developmental ability of any ofthese patients, anecdotal reports of improve-ment in alertness, interest in the environment,and decreased aggression have been reported.Physical changes included improvement ineczema and photosensitivity and resolution ofchronic diarrhoea. One child was thought to be

-- e# more irritable on treatment and the cholesterolwas ceased temporarily. This has subsequentlybeen restarted without adverse effect. Apart

-rocephaly, bilateral ptosis,a small from this one report, no other complications ofision on her left cheek was caused by the diet were seen.

cts, or repeated actions, for Discussionself or objects repeatedly. Based on our figures from male patients born

ren were described as happy and diagnosed with SLO in 1995 and 1996 inthe extent of being over the UK, we suggest a minimum incidence ofstrangers. In four cases, 1/60 000 and a carrier frequency of at least

hat their children's feet were 1/122. Continuing difficulty in diagnosis{e to tactile stimulus. means this is undoubtedly an underestimate. A

single example of parental consanguinity is inkeeping with the hypothesis that SLO is a rela-

nts had a karyotype per- tively common disorder. For counselling pur-reported as normal, with no poses, a carrier frequency of 1-2% is appropri-q32. One patient had sexing ate: These figures are in keeping with the)robe identification on histo- estimates of prevalence of 1/20 000-1/50 000

in the northern European population and thehe clinical diagnosis of SLO American population respectively, and a carrierzer their parents were shown frequency of 1-2%.2 31 There appears to be aHC levels, compatible with wide variation in carrier frequency in differenttote levels. For the remaining ethnic groups, with a higher frequency inised 7DHC, with the excep- northern Europeans and a low frequency inwho had markedly raised Asian and Afican races.' Whether this will

'hydro isomer of 7DHC), prove to be because of some selective advan-ctable 7DHC. This patient tage or other factors is unknown. The truetypical phenotype. For 19 birth incidence could be determined if the

treatment serum cholesterol entire newborn population for a given locationwere available. Pretreatment was screened. This is now theoretically possi-ranged from 0.2 to 3.1 ble; Zimmerman et af0 have recently published

71, SD 0.84) and pretreat- a method of screening dried filter paper bloodted from 143 to 991 imol/l specimens for raised 7DHC and low75). Correlation of these val- cholesterol.ther and severity scores were With the advent of a biochemical marker, therregression analysis. No sig- already wide clinical spectrum has expandedn was found between these even further. None of the features in SLO arepathognomonic or obligatory for diagnosis.

While the majority of cases are diagnosed in thenewborn period because of multiple congenitalabnormalities and facial dysmorphism, therewere 11 patients in this series without majorstructural abnormalities, but with mild dys-

* * morphic features and developmental delay.Patients representing the mild end of the spec-trum require a high level of suspicion for diag-nosis. Three patients in our series were

I I I I diagnosed after the clinician sent serum for00 600 800 1000 1200 analysis to "exclude SLO"; they were only7DHC mmol/I mildly dysmorphic, but had at least two minor)rrelation between 7DHC values anomalies consistent with the diagnosis ofpatients. SLO. This suggests that more patients with

563



g.bmj.com

/J M


g.35.7.558 on 1 July 1998. Dow

nloaded from

http://jmg.bmj.com/


developmental delay, dysmorphic features, andcompatible minor congenital abnormalitiesshould be considered for biochemical testing ofthe cholesterol pathway, even in the absence ofmore severe physical abnormalities. Recently, asecond defect in the conversion of lanosterol tocholesterol, desmosterolosis, was described.2This and other aberrations of the cholesterolbiosynthetic pathway can be detected usingGC-MS.The "SLO face" is easily recognised in the

majority of patients, although may be subtle.We feel that the striking features are micro-cephaly with narrow bitemporal diameter, pto-sis, a short nose with anteverted nares, palatalabnormality (ridged, high arched, cleft uvula,or cleft palate), and broad alveolar ridges. Aspatients age, the face becomes less typical, andso should not be relied on for diagnosis.

Chorionic villus biopsy and amniocentesissterol analysis have proven to be reliable forprenatal diagnosis and is useful in thosefamilies at high risk of having an affectedchild.22 24 Features seen on scan include intra-uterine growth retardation, renal abnormali-ties, congenital heart disease, polydactyly,nuchal oedema, decreased fetal movements,and oligohydramnios.24 26 Recently, prenataldiagnosis of SLO based on suggestive abnor-malities seen on antenatal scan has beenreported.2 Sterol analysis could be performedat the same time as chorionic villus biopsy oramniocentesis in those pregnancies affected bycompatible congenital abnormalities. Similarly,low maternal serum oestriol values have beenassociated with SLO affected pregnancies.2 22The non-specific nature of the structuralanomalies in SLO detectable by anomaly scan-ning raises the question of routinemeasurement of 7DHC in amniotic fluidcollected for chromosome analysis on the basisof an abnormal scan. Given that the number ofcases detected could be of the order of 1/1000diagnostic tests, routine adoption is notjustified without formal evaluation.

Eighteen (37%) patients had congenitalheart disease; a disproportionate increase inatrioventricular septal defect was seen, afinding noted in previous reports.'6 2S 2'6 We didnot, however, find a significant increase in atrialseptal defect or anomalous pulmonary venousdrainage. None of our patients, nor anyreported SLO patients, have had conotruncalanomalies. The atrioventricular septal defectexcess may reflect alterations in cell morphol-ogy and loss of cellular attachments, resultingin defective migration of cells during develop-ment of the endocardial cushions.Marked photosensitivity, with a striking ery-

thematous reaction on brief exposure tosunlight, was seen in 63% of patients. Sunburnis a major problem and children are generallykept indoors as burning may occur within a fewminutes of exposure to sunlight or through thecar window. The photosensitivity has beenpostulated to be the result of the accumulationof7DHC in fibroblasts or abnormal cell mem-brane structure caused by deficientcholesterol. '

Limb abnormalities were commonly seen,though it is of interest that only 81% had the"hallmark" 2/3 toe syndactyly. No patient wasreported to have holoprosencephaly, a malfor-mation which has been reported in SLO andhas been implicated to be, in part, the result ofthe effect of the Sonic Hedgehog gene.' Giventhe severity of the neurological impairment inSLO, the rarity of epilepsy is worthy of note.Cunniff et al described two patients from theirseries of 80 who died of cholestatic liverdisease. The addition of a further three patientsin this series with cholestatic liver diseasemakes it likely that this is an uncommon, butserious manifestation of the SLO phenotype.

In reviewing the biochemistry of 19 patients,there was limited correlation between pretreat-ment values of serum cholesterol, 7DHC, andseverity of congenital abnormalities. Whileother groups also found little correlationbetween 7DHC and clinical severity, Cunniff etal and Tint et aP" found an inverse correlationbetween clinical severity and cholesterol level.The poor correlation between 7DHC,cholesterol, and clinical severity taken togetherwith the distribution of severity scores in ourseries points to the presence of one or moreimportant modifier genes in the population.8DHC is an isomer of 7DHC; neither are

normally detected in controls, but are presentin SLO homozygotes. There was one patient inour series with raised 8DHC, without anydetectable 7DHC. This case had a severe, butotherwise typical phenotype. The absence of7DHC may be artefactual, as 7DHC may bemore prone to photodegradation or auto-oxidation compared to 8DHC and cholesterol.Alternatively, it is possible that this case mayhave had a different inborn error of cholesterolmetabolism.A high cholesterol diet can increase serum

cholesterol and reduce 7DHC in SLO patients.It has been suggested that it may improvedevelopmental outcome. Given the wide clini-cal spectrum, small numbers, and placeboeffect on development and perceptions ofbehaviour, it will be difficult to prove an effectwithout fully randomised trials. For thisreason, we have made no specific comment onthe effect of dietary intervention and mustawait results of continuing studies. While therehas been no dramatic improvement in develop-mental ability in any of the patients seen in ourseries, anecdotal reports of improvement inbehaviour and alertness were noted by someparents. Any improvement in behaviour such asaggressive outbursts and sleep patterns isworthwhile, making these children and adultsmore manageable for their care givers. Impor-tantly, very few adverse effects of the diet havebeen reported. Recent reports by Elias et al"and Nwokoro and Mullvihill32 have notedimprovements in growth parameters althoughimprovements in behaviour are still largelysubjective.

In summary, we have found a wide spectrumof abnormalities in patients with SLO. Sincethe advent of a biochemical marker, the clinicalfeatures are being refined and some patientshave been diagnosed with minimal features. We

564



g.bmj.com

/J M


g.35.7.558 on 1 July 1998. Dow

nloaded from

http://jmg.bmj.com/


believe clinicians should have a low thresholdfor biochemical testing in patients who are dys-morphic and have developmental delay.

The authors would like to thank Drs M Barrow, J Beesley, CBerry, D Brenton, L Brueton, T Cole, D Davidson, I Ellis, PFarndon, D Fitzpatrick, F Flinter, A Fryer, C Garrett, A Green,R Greenham, E Hatchwell, E Hoskyns, F Howard, H Hughes, JHurst, S Huson, J Keeling, J McGraughran, C McKeown, JMorgan, B Mueller, E Neurkla, T Perham, M Porteous, MSeller, F Stewart, J Tolmie, G Woods, E Wraith, and I Young forallowing access to patients and their medical records and HMcConicky and Dr J Collins who performed many of the devel-opmental and neurological assessments. The behavioural andsleep questionnaires were kindly supplied by Dr T Bernie.Thanks also to the patients with SLO and their families withoutwhom this study would not have been possible. This work wasfunded by Action Research and the Borwick Trust.

1 Smith DW, Lemli L, Opitz JM. A newly recognizedsyndrome of multiple congenital anomalies. _7 Pediatr1964;64:210-17.

2 Lowry RB, Yong S. Boderline normal intelligence in theSmith-Lemli-Opitz (RSH) syndrome. Am _7 Med Geniet1980;5: 137-43.

3 Kelley RI. A new face for an old syndrome. Am _7 Med Getnet1997;68:251-6.

4 Cunniff C, Kratz LE, Moser A, Natowicz MR, Kelley RI.Clinical and biochemical spectrum of patients withRSH/Smith-Lemli-Opitz sydrome and abnormalcholesterol metabolism. Am _7 Med Genet 1997,68:263-9.

5 Irons M, Elias ER, Salen G, Tint GS, Batta AK. Defectivecholesterol biosynthesis in Smith-Lemli-Opitz syndrome.Latncet 1993;341:1414.

6 Tint GS, Irons M, Elias ER, et al. Defective cholesterol bio-synthesis associated with the Smith-Lemli-Opitz syn-drome. N Engl_ Med 1994;330: 107-13.

7 Honda A, Tint GS, Salen G, Batta AK, Chen TS, Shefer S.Defective conversion of 7-dehydrocholesterol tocholesterol in cultured skin fibroblasts from Smith-Lemli-Opitz syndrome homozygotes. ] Lipid Res 1995;36:1595-601.

8 Kelley IK, Roessler E, Hennekam CM, et al. Holoprosen-cephaly in RSH/Smith-Lemli-Opitz syndrome: does abnor-mal cholesterol metabolism affect the function of sotnichedgehog? An _7 Med Genet 1996;66:478-84.

9 Wallace M, Zori RT, Alley T, et al. Smith-Lemli-Opitz syn-drome in a female with a de novo, balanced translocationinvolving 7q32: probable disruption of an SLOS gene. AmJMed Genet 1994;50:368-74.

10 Curry CJR, Carey JC, Golabi M, et al. Multiple congenitalanomalies with male pseudohermaphroditism and frequentearly lethality. Am] Med Genet 1987,26:45-57.

11 Berry R, Wilson H, Robinson J, et al. Apparent Smith-Lemli-Opitz syndrome and Miller-Dieker syndrome in afamily with segregating translocation t(7 17)(q34;p13.1).AmJy Med Genet 1989;34:358-65.

12 Acosta PB. Theoretical and practical aspects of dietarytreatment of Smith-Lemli-Opitz syndrome. Int Pediatr1995;10:37-40.

13 Bergsma D, ed. Birth defects comlpendiunm. London: Macmil-lan Press, 1979:963.

14 Kelley RI. Diagnosis of Smith-Lemli-Opitz syndrome bygas chromatography/mass spectrometry of 7-dehydro-cholesterol in plasma, amniotic fluid and cultured skinfibroblasts. Clitn Chini Acta 1995;236:45-58.

15 Seller MJ, Flinter FA, Docherty Z, Fagg N, Newbould M.Phenotypic diversity in the Smith-Lemli-Opitz syndrome.Clinl Dysmorphol 1997;6:69-73.

16 Seller MJ, Russell J, Tint GS. Unusual case of Smith-Lemli-Opitz syndrome "type II". AmJ7Med Genet 1995;56:265-8.

17 McGaughran JM, Clayton PT, Mills KA, Moore L, DonnaiD. Prenatal diagnosis of Smith-Lemli-Opitz syndrome. AmJ Med Geniet 1995;56:269-71.

18 Bialer MG, Penchaszadeh VB, Kahn E, et al. Female exter-nal genitalia and Mullerian duct derivatives in a 46,XYinfant with the Smith-Lemli-Opitz syndrome. Ami . MedGetnet 1987;28:723-31.

19 Opitz JM. RSH/SLO (Smith-Lemli-Opitz) syndrome: his-torical, genetic and developmental considerations. Ami iMed Geniet 1994;50:344-6.

20 Zimmerman PA, Hercules DM, Naylor EW. Direct analysisof filter paper blood specimens for identification of Smith-Lemli-Opitz syndrome using time-of-flight secondary ionmass spectrometry. Am]l Med Geniet 1997;68:300-4.

21 Clayton PT, Mills K, Keeling J, Fitzpatrick D.Desmosterolosis: a new inborn error of cholesterol biosyn-thesis. Lancet 1996;348:404.

22 Abuelo DN, Tint GS, Kelley R, et al. Prenatal detection ofthe cholestrol biosynthetic defect in the Smith-Lemli-Opitzsyndrome by the analysis of amniotic fluid sterols. AmnMed Genet 1995;56:281-5.

23 Mills K, Mandel H, Montemagno R, et al. First trimesterprenatal diagnosis of Smith-Lemli-Opitz syndrome (7-dehydrocholesterol reductase deficiency). Pediatr Res 1996;39:816-19.

24 Rossiter JP, Hofman KJ, Kelley RI. Smith-Lemli-Opitzsyndrome: prenatal diagnosis by quantification ofcholesterol precursors in amniotic fluid. Am _7 Med Geniet1995;56:272-5.

25 Stewart FJ, Nevin NC, Dornan JC. Prenatal diagnosis ofSmith-Lemli-Opitz syndrome. Am _7 Med Genet 1995;56:286-7.

26 Johnson JA, Aughton DJ, Comstock CH, et al. Prenataldiagnosis of Smith-Lemli-Opitz syndrome, type II. Am]Med Genet 1994;49:240-3.

27 McKeever PT, Young ID. Smith-Lemli-Opitz syndrome II:a disorder of the fetal adrenals? _7 Med Genet 1990;27:465-6.

28 Lin AE, Ardinger HH, Ardinger RH Jr, Cunniff C, KelleyRI. Cardiovascular malformations in Smith-Lemli-Opitzsyndrome. Ami .7 Med Genet 1997;68:270-8.

29 Tint GS, Salen G, Batta AK, et al. Correlation of severityand outcome with plasma sterol levels in variants of theSmith-Lemli-Opitz syndrome. _7 Pediatr 1995;127:82-7.

30 Dehart DB, Lanoue L, Tint GS, Sulik KK. Pathogenesis ofmalformations in a rodent model for Smith-Lemli-Opitzsydrome. Ami . Med Genet 1997;68:328-37.

31 Elias ER, Irons MB, Hurley AD, Tint GS, Salen G. Clinicaleffects of cholesterol supplementation in six patients withthe Smith-Lemli-Opitz syndrome. Amn 7 Med Genet1997;68:305-10.

32 Nwokoro NA, Mulvihill JJ. Cholesterol and bile acidreplacement therapy in children and adults with Smith-Lemli-Opitz (SLO/RSH) syndrome. Am . Med Genet1997;68:315-21.

565



g.bmj.com

/J M


g.35.7.558 on 1 July 1998. Dow

nloaded from

http://jmg.bmj.com/

genet smith-lemli-opitz syndrome: variable clinical …smith-lemli-opitz syndrome table 1...

Documents