Recognition of thalidomide defects

8
76 Med Genet 1992; 29: 716-723 SYNDROME OF THE MONTH Recognition of thalidomide defects R W Smithells, C GH Newman Thalidomide: a brief history Thalidomide (alpha-phthalimido-glutarimide) was developed by the German firm Chemie Grunenthal as an anticonvulsant drug. Early trials showed it to be unsuitable for this pur- pose but indicated that it had sedative proper- ties. Furthermore, it had one remarkable prop- erty: overdoses simply caused prolonged sleep, not death. The drug was first marketed in Germany in 1957 under the name Contergan, and in the UK in April 1958 as Distaval. Later, compound preparations which combined tha- lidomide with other drugs were marketed for a wide variety of indications: Asmaval for asthma, Tensival for hypertension, Valgraine for migraine, and so forth. The promotion of these products laid great stress on the safety of thalidomide, based on the remarkable property described above. German paediatricians and geneticists began to see children with gross limb malfor- mations of a most unusual pattern. When two cases were shown at a paediatric meeting in Kassel by Kosenow and Pfeiffer in October 1960, few people present had ever seen similar limb defects. Wiedemann in 1961 described 13 affected infants who had been referred to him over a period of 10 months, and noted that this amounted to an epidemic. He drew attention to a number of associated malformations in these children, including congenital heart dis- ease, microphthalmos and coloboma, intestinal atresia, renal malformations, abnormal pinnae, and facial naevus. In November 1961, Lenz suggested that these deformities resulted from the mothers having taken thalidomide. By a remarkable coincidence, the same suggestion was made at much the same time by McBride in Australia. Confirmation of this suggestion came rapidly from all parts of the British Isles, Kenya, Japan, Sweden, Belgium, Switzerland, Lebanon, Israel, Peru, Canada, Brazil, the Netherlands, and the USA. The drug had been released only for clinical trials in the USA because of concerns following reports from Europe of irreversible peripheral neur- itis as a side effect of thalidomide. Conse- quently there were very few cases. By con- trast, it had been on sale over the counter in Germany, and there were consequently more affected children there than anywhere else. In the UK the drug was available on prescription only, but it was used very widely for, among other problems, common symptoms of early pregnancy. Thirty years later, subjects are still coming forward (albeit, in small numbers) with claims that they have birth defects which have (or may have) been caused by thalidomide taken by their mothers during early pregnancy, and that they should therefore be accepted as bene- ficiaries of whatever forms of financial assist- ance may be available. In the UK, this is the Thalidomide Trust. Thalidomide caused a wide variety of birth defects, not one of which was unique to that drug. Nevertheless, the nature and pattern of the defects are, in most cases, characteristic enough to be recognisable to an experienced eye. Indeed, many of the present beneficiaries of the Trust have been accepted on the basis of clinical judgements, without direct evidence of thalidomide exposure. As the number of doc- tors with wide experience of this problem is small (possibly only three in the UK) and will become smaller with the passage of time, and as new claims continue to arise, it seems timely to record, in as much detail as possible, the observations which underlie these clinical judgements. As well as describing the defects and pat- terns associated with thalidomide, it will be appropriate to discuss 'differential diagnosis', that is, recognisable defects and syndromes which, to a greater or lesser degree, resemble thalidomide defects. Some of these are un- likely to confuse an experienced eye: others can present considerable difficulty. In considering a new claim, attention will obviously be paid to the claimant's date of birth. In the UK, thalidomide was first mar- keted in April 1958, initially in very small quantities, so it is not to be expected that it would damage anybody born before January 1959 (unless supplies had been obtained from West Germany). Sales in the UK stopped in November 1961. Had consumption stopped at the same time, no 'thalidomide babies' should have been born after August 1962. However, many people still had tablets containing thali- domide in their homes, and either did not hear promptly of its dangers, or did not realise that their tablets contained thalidomide. In fact, children with defects accepted as attributable to thalidomide (though not necessarily with documentary evidence of prescription) con- tinued to be born up to about May 1963, and, very exceptionally, beyond this date. In trying to establish a yardstick or bench- mark for bona fide thalidomide defects, it is necessary to start with cases in which there is 5 North Grange Mews, North Grange Road, Leeds LS6 2EW. R W Smithells Leon Gillis Unit, Queen Mary's Hospital, Roehampton. C G H Newman Correspondence to Professor Smithells. 716 on January 14, 2022 by guest. Protected by copyright. http://jmg.bmj.com/ J Med Genet: first published as 10.1136/jmg.29.10.716 on 1 October 1992. Downloaded from

Transcript of Recognition of thalidomide defects

Page 1: Recognition of thalidomide defects

76 Med Genet 1992; 29: 716-723

SYNDROME OF THE MONTH

Recognition of thalidomide defects

R W Smithells, C G H Newman

Thalidomide: a brief historyThalidomide (alpha-phthalimido-glutarimide)was developed by the German firm ChemieGrunenthal as an anticonvulsant drug. Earlytrials showed it to be unsuitable for this pur-pose but indicated that it had sedative proper-ties. Furthermore, it had one remarkable prop-erty: overdoses simply caused prolonged sleep,not death. The drug was first marketed inGermany in 1957 under the name Contergan,and in the UK in April 1958 as Distaval. Later,compound preparations which combined tha-lidomide with other drugs were marketed for awide variety of indications: Asmaval forasthma, Tensival for hypertension, Valgrainefor migraine, and so forth. The promotion ofthese products laid great stress on the safety ofthalidomide, based on the remarkable propertydescribed above.German paediatricians and geneticists

began to see children with gross limb malfor-mations of a most unusual pattern. When twocases were shown at a paediatric meeting inKassel by Kosenow and Pfeiffer in October1960, few people present had ever seen similarlimb defects. Wiedemann in 1961 described 13affected infants who had been referred to himover a period of 10 months, and noted that thisamounted to an epidemic. He drew attentionto a number of associated malformations inthese children, including congenital heart dis-ease, microphthalmos and coloboma, intestinalatresia, renal malformations, abnormal pinnae,and facial naevus.

In November 1961, Lenz suggested thatthese deformities resulted from the mothershaving taken thalidomide. By a remarkablecoincidence, the same suggestion was made atmuch the same time by McBride in Australia.Confirmation of this suggestion came rapidlyfrom all parts of the British Isles, Kenya,Japan, Sweden, Belgium, Switzerland,Lebanon, Israel, Peru, Canada, Brazil, theNetherlands, and the USA. The drug hadbeen released only for clinical trials in theUSA because of concerns following reportsfrom Europe of irreversible peripheral neur-itis as a side effect of thalidomide. Conse-quently there were very few cases. By con-trast, it had been on sale over the counter inGermany, and there were consequently moreaffected children there than anywhere else. Inthe UK the drug was available on prescriptiononly, but it was used very widely for, amongother problems, common symptoms of earlypregnancy.

Thirty years later, subjects are still comingforward (albeit, in small numbers) with claimsthat they have birth defects which have (ormay have) been caused by thalidomide takenby their mothers during early pregnancy, andthat they should therefore be accepted as bene-ficiaries of whatever forms of financial assist-ance may be available. In the UK, this is theThalidomide Trust.Thalidomide caused a wide variety of birth

defects, not one of which was unique to thatdrug. Nevertheless, the nature and pattern ofthe defects are, in most cases, characteristicenough to be recognisable to an experiencedeye. Indeed, many of the present beneficiariesof the Trust have been accepted on the basis ofclinical judgements, without direct evidence ofthalidomide exposure. As the number of doc-tors with wide experience of this problem issmall (possibly only three in the UK) and willbecome smaller with the passage of time, andas new claims continue to arise, it seems timelyto record, in as much detail as possible, theobservations which underlie these clinicaljudgements.As well as describing the defects and pat-

terns associated with thalidomide, it will beappropriate to discuss 'differential diagnosis',that is, recognisable defects and syndromeswhich, to a greater or lesser degree, resemblethalidomide defects. Some of these are un-likely to confuse an experienced eye: otherscan present considerable difficulty.

In considering a new claim, attention willobviously be paid to the claimant's date ofbirth. In the UK, thalidomide was first mar-keted in April 1958, initially in very smallquantities, so it is not to be expected that itwould damage anybody born before January1959 (unless supplies had been obtained fromWest Germany). Sales in the UK stopped inNovember 1961. Had consumption stopped atthe same time, no 'thalidomide babies' shouldhave been born after August 1962. However,many people still had tablets containing thali-domide in their homes, and either did not hearpromptly of its dangers, or did not realise thattheir tablets contained thalidomide. In fact,children with defects accepted as attributableto thalidomide (though not necessarily withdocumentary evidence of prescription) con-tinued to be born up to about May 1963, and,very exceptionally, beyond this date.

In trying to establish a yardstick or bench-mark for bona fide thalidomide defects, it isnecessary to start with cases in which there is

5 North Grange Mews,North Grange Road,Leeds LS6 2EW.R W Smithells

Leon Gillis Unit,Queen Mary'sHospital,Roehampton.C G H Newman

Correspondence toProfessor Smithells.

716

on January 14, 2022 by guest. Protected by copyright.

http://jmg.bm

j.com/

J Med G

enet: first published as 10.1136/jmg.29.10.716 on 1 O

ctober 1992. Dow

nloaded from

Page 2: Recognition of thalidomide defects

Recognition of thalidomide defects

very good evidence of thalidomide intake inearly pregnancy. Unfortunately, the investig-ator has never seen the pregnant mother swal-low a tablet. Absolute certainty is thereforeunattainable, and we must settle for a scale ofprobability. The evidence may include thefollowing.(1) A dated prescription for thalidomide, the

date falling within or before (but not toolong before) the period ofembryonic sensit-ivity to the drug (34 to 50 days after thebeginning of the last menstrual period).

(2) A doctor's statement, preferably a swornaffidavit, that he supplied such a prescrip-tion at such a time, but kept no record of it.

(3) A mother's statement (preferably sworn)that she took thalidomide at the relevanttime, with an indication of its source(which is not necessarily a prescription).

(4) A mother's ability to identify the tablet shetook, when shown a selection of tablets.Fortunately, all tablets containing thalido-mide were readily identifiable and could berecognised or described by reasonably ob-servant people.

At this point it is appropriate to mention twoimportant paradoxes. First, although a fewmothers may have claimed untruthfully tohave taken thalidomide (acting in what theybelieved to be their child's best interests), avastly greater number of the mothers of accep-ted Trust beneficiaries denied any knowledgeof any drug consumption during pregnancy.There is a negative drug history in about 50%of accepted cases, and there is unlikely to beany such evidence in the future.The second paradox is that, bearing in mind

that 2 to 3% of all babies born have significantbirth defects, and that thalidomide consump-tion was widespread in 1960 to 1961, somemothers who undoubtedly took the drug whenpregnant (though probably outside the sensi-tive period) gave birth to babies with defectsquite unrelated to thalidomide. It is also pos-sible for a baby exposed to thalidomide duringthe sensitive period to be born with a variety ofdefects, of which some, but not all, are druginduced. Authorities differ about the possi-bility that a fetus exposed to thalidomide dur-ing the sensitive period might be born withoutbirth defects. If it happens, it is rare.The last general point to mention is the risk

of perpetuating an error. If a clinician acceptsas resulting from thalidomide a defect whichhas not, in fact, been described in cases withstrong documentary support, he is more readyto accept that defect the next time he meets it.A defect that has been accepted two or threetimes becomes built into his repertoire of 'tha-lidomide defects'. This underlines the neces-sity to start from those cases with good evid-ence of drug exposure.The following comments are based on 148

personally examined cases with good evidenceof thalidomide exposure in early pregnancy,with documentary support in 35 cases.

The pattern of thalidomide defectsThalidomide is associated in the public mind

with limb defects, and these certainly accountfor the majority of cases. However, almost anyorgan of the body could be affected. Thesecond major group of defects involves theears, the eyes, and the nerve supplies to theface, the eye muscles, and the lacrimal (tear)glands. Internal defects commonly affected theheart, the kidneys and urinary tract, the ali-mentary tract, and the genital tract, and nonewas unique to thalidomide. The early morta-lity rate among 'thalidomide babies' was about40%, largely as a result of serious internalmalformations. Consequently, internal defectsare much less common among survivors thanthey were among the whole group at birth.Most of the serious internal defects caused

problems at or soon after birth which eitherrequired treatment or led to death. Some de-fects of the kidneys and female genital tractwhich can only be shown by special tests didnot become apparent until many years afterbirth. It is possible that there are still unde-tected internal problems in people aged 30years or more, but as time passes it becomesincreasingly unlikely that any such hiddendefects will cause significant problems. Theywill therefore play little part in the diagnosis ofthalidomide damage in the future.A small but important group of thalidomide

related problems includes conditions whichare not present at birth but develop later.Abnormalities of the spine were recognisedearly, and of the knees rather later. Otherbones/joints may also be affected. It is to beexpected that the thalidomide damaged peoplewill be prone to the same ills as beset the rest ofthe population. A causal connection with thali-domide would be suggested if a particulardisease was more common among the thalido-mide population than among the general popu-lation from which they came, or if the diseasepresented at an unusual age or in an unusualway.

Some general features of thalidomidedamageAt birth, many thalidomide babies exhibited acentral facial naevus of the 'stork mark' varietyin the centre of the forehead (which is commonamong all babies) but spreading down over thenose and upper lip, and sometimes with a smallelement on the lower lip just below the ver-milion border. These birthmarks disappearedover one to two years and will not be seen infuture claimants, but there may be historical orphotographic evidence. Some children showedfacial features best understood by reference tophotographs.

Short stature, over and above the shortnessattributable to short leg bones, is seen moreoften than would be expected by chance, and isaccepted as attributable to thalidomide.

SYMMETRYAs the two sides of the embryo develop moreor less in parallel, and it is difficult to envisagea drug which reaches it via the blood streambeing distributed only to one side, one wouldexpect drug induced malformations of bilateral

717

on January 14, 2022 by guest. Protected by copyright.

http://jmg.bm

j.com/

J Med G

enet: first published as 10.1136/jmg.29.10.716 on 1 O

ctober 1992. Dow

nloaded from

Page 3: Recognition of thalidomide defects

Smithells, Newman

structures to be more or less symmetrical. Thisis, broadly, what is observed in practice, notonly with thalidomide, but with other terato-genic drugs and with defects of genetic origin.However, the extent of the symmetry variesaccording to the nature of the defect, both inthe closeness of the match between left andright, and in the proportion of cases which areappreciably asymmetrical. The significance ofsymmetry will be discussed in relation to eachdefect group.

Figure 1 Symmetrical reduction deformities of all limbswith loss of preaxial upper limb digits. Note hypoplasticscrotum.

UPPER LIMBSThe most characteristic defects are reductiondeformities, that is, the loss of part or thewhole of one or more bones. The bones of theupper limb are affected in a remarkably regularorder, starting with the thumb, followed bythe radius, the humerus, the ulna,, and finallythe fingers on the ulnar side of the hand(middle., ring, and little fingers). Conse-quently., the defect which falls just short ofcomplete absence of the arm (upper limb ame-lia) consists of one or more digits attacheddirectly to the shoulder. If the defect is a littleless severe still, there is a lump of bone inter-posed between the shoulder and the digits., theorigin of this lump being impossible to deter-mine on anatomical grounds. If the reductiondeformity is less still, it is possible to identifythe individual long bones (humerus, ulna).Not surprisingly, when substantial parts

of bones are missing, the muscles normally

attached to them are hypoplastic, but theextent of this muscle hypoplasia does notalways correspond exactly to the loss of bone.For example, the muscles of the shoulder andupper arm may be markedly hypoplastic evenwhen the humerus is of normal length. Thedegree of underdevelopment of the thenareminence does not always reflect the size of thethumb bones.When the long bones of the arm are affected,

their relationships to one another may be dis-turbed. The radius and ulna may be partly orwholly fused, limiting rotation of the forearm.There may be humeroulnar fusion, preventingmovement at the elbow, in which case thehumerus and ulna are usually short.

Thumbs and thenar musclesCareful examination and x rays of the thumbsare often extremely informative. The humanembryo is thought to be sensitive to thalido-mide from approximately 34 to 50 days afterthe start of the mother's last menstrual period,although individual structures may only besensitive for part of that time. The thumbsappear to be the first part of the skeleton to beaffected, and the last. Early involvement maybe associated with a range of facial defects (seebelow) or more extensive involvement of theupper limbs: late involvement may be associ-ated with anorectal stenosis.Complete absence of the thumbs is far more

common than thumb deformity. Next mostcommon is hypoplasia of the thumb and thenarmuscles: these small, thin thumbs are com-monly fused in part to the adjacent indexfinger.Thumbs affected by thalidomide quite often

contain three phalanges (which may only beshown on x ray), and may resemble, in size andposition, a fifth finger rather than a thumb, aproblem which often necessitated an operationto create an opposable digit. If the thumb isaffected early by a short exposure to thalido-mide, the radius may escape intact. Equally, ifthe thumb is affected late, the radius may beunaffected. However, on the basis of cases withsound evidence of thalidomide exposure, itseems likely that the thumb is never com-pletely normal if the radius is abnormal. If theradial defect is very slight, so may the thumbdefect be, but the combination of a severelyabnormal or absent radius with a normalthumb militates strongly against thalidomideas a cause.

RadiusReduction deformities of the radius tend tostart at the distal end and extend towards theelbow. In addition to being short, a thalido-mide damaged radius tends to be bowed, andmay be thicker than normal. It is occasionallyfused to some degree with the ulna. If theradius is shorter than the ulna, which is com-mon, the wrist and hand cannot be in normalalignment with the forearm but are rotatedtowards the radial side (radial club hand).

718

on January 14, 2022 by guest. Protected by copyright.

http://jmg.bm

j.com/

J Med G

enet: first published as 10.1136/jmg.29.10.716 on 1 O

ctober 1992. Dow

nloaded from

Page 4: Recognition of thalidomide defects

Recognition of thalidomide defects

HumerusThe humerus also tends to be affected from thedistal to the proximal end. However, theshoulder joint is often weak and liable torecurrent dislocation, even with a normallength humerus, because of hypoplasia of themuscles surrounding the shoulder. A weakshoulder is therefore common even though thehumeral head and glenoid cavity are bothpresent. The lower end of the humerus issometimes fused to the upper end of the ulna.

UlnaThe characteristic shape of the proximal end ofthe ulna is retained in less severe defects andmakes the ulna identifiable. In extreme cases,it loses its shape, and the residual knob of bonewhich is often seen on x rays can only beassumed to be of ulnar origin because of thesequence of events already described.

FingersThe thumb develops in association with theradius; the middle, ring, and little fingers withthe ulna; and the index finger usually, butmore variably, with the radius. Consequently,if the radius is affected, the thumb, and oftenthe index finger, will be affected, but the otherthree fingers are usually present if there is anypart of the ulna present. They may be thin,flexed, and weak. If the ulna is absent, theremay be three (exceptionally four), two, one, orno fingers at the shoulder.

Scapula and clavicleThe scapula and clavicle are not subject todeformity as the arm bones are, but where thearm is small or absent and the muscles corres-pondingly defective, the scapula and clavicleare of little or no use and may therefore besmaller than normal. The scapula in particularmay be grossly underdeveloped, and when thehumerus is severely deficient, the glenoidcavity is underdeveloped or absent.Not infrequently x rays show bones of cur-

ious shapes which do not closely resemble anyof the three long bones of the arm. These mayrepresent aggregates derived from more thanone primitive bone (radius+ ulna, humerus+ulna).

Symmetry of upper limb defectsBilateral symmetry is more marked with de-fects of the upper limbs than it is with lowerlimb or non-limb defects. Nevertheless, thereis almost invariably a small difference betweenthe two sides, often confined to the digits.Commonly, for example, there are three digitson one side and four on the other, or there maybe the same number on both sides, but the sizeof one, or the degree of fusion between adjac-ent digits, differs. There may be completeamelia on one side and a single digit on theother. It is unusual to find the number of digitson the two sides differing by more than one(bearing in mind that some very rudimentary

thumbs either necrosed and dropped off spon-taneously or were surgically removed soonafter birth).When the long bones of the arms are

involved, there is usually a small difference inthe extent of the reduction deformity. Thegreater the difference between the two sides,the less likely are the defects to be of drug orgenetic origin, but it is not possible to drawrigid lines. Complete amelia on one side with anormal upper limb on the other is, at the least,highly unlikely to be attributable to any drug.

LOWER LIMBSThe majority of people with thalidomide de-fects of the upper limbs have normal lowerlimbs. A minority have defects of all limbs.Defects of the lower limbs with normal upperlimbs are uncommon.

.P

,'Ir ."

I.1.-

A

Figure 2 Severe reduction deformities of legs withhypoplastic labia. Normal length arms with 'fingerised'thumbs.

The pattern of lower limb defects is morevariable than in the upper limbs, and thedegree of bilateral symmetry is less marked;symmetry is most frequently seen with themost severe defects. The long bones are thefirst to be affected, although talipes (club foot)and congenital dislocation of the hip may bothoccur without reduction deformities. Com-plete lower limb amelia is rare, especially insurvivors. In the most severe cases commonlyseen, the feet (usually abnormal) arise from thehip areas, often with one or more unidentifi-able small bony lumps shown on x rays

719

on January 14, 2022 by guest. Protected by copyright.

http://jmg.bm

j.com/

J Med G

enet: first published as 10.1136/jmg.29.10.716 on 1 O

ctober 1992. Dow

nloaded from

Page 5: Recognition of thalidomide defects

Smithells, Newman

between the pelvis and the feet. Lesser degreesof reduction deformity show much more vari-able patterns, as described below.

FemurThe femur is quite often the only long bone tobe affected, although the tibia or tibia + fibulamay be affected as well, or with a normalfemur. The upper end is the first to go, whichinevitably prevents the formation of a normalhip joint. The lower end is usually the last partto be preserved. In addition to shortening, thefemur may be bowed or angulated. Early frac-ture of its upper end, which may not ossify forsome years after birth, may result in separatedor angulated portions appearing on x ray aftermid-childhood.

TibiaThe tibia may be the only long bone to beaffected, but the femur is often affected as well,and if the tibia is significantly shorter thannormal, the fibula has no option but to be shortor bent. In contrast to the femur, the tibiatends to be affected first at its lower end,compromising the integrity of the ankle joint.Like the femur, it may be bowed as well asshort, in which case the fibula is almost boundto be bowed as well.

FibulaWhen a single long bone remains visible on xray (with or without additional bony lumps) ittends to be slender and straight, with endslittle thicker than the shaft, that is, it resemblesa fibula much more closely than it does a femuror tibia. It therefore appears to be analogous tothe ulna in the upper limb in being the lastlong bone to disappear.

Feet and toesTalipes (club foot) can occur without any limbreduction deformity. It is, of course, a verycommon birth defect, but it occurs more com-monly in people exposed to thalidomide inutero than would be expected by chance.Talipes is virtually constant if the tibia orfibula or both are affected, or if all three longbones are affected. It may also be seen whenthe femur is affected but the tibia and fibula arenormal. When there is virtually no structurebetween the feet and the hips, the feet areinevitably in an abnormal posture.

In contrast to fingers, where the general ruleis 'five or fewer', the rule for toes is 'five ormore'. Supernumerary and bifid toes areusually on the side of the big toes. There maybe as many as eight (possibly more) toes oneach foot, and the number on the two feetrarely differs by more than one. Supernumer-ary toes were often removed surgically soonafter birth.

Hips, knees, and anklesHips may be dislocated or unstable at birth,

and may develop Perthes' disease later. Kneesmay be unstable at birth or develop arthritislater or both. There may be recurrent disloca-tion of the patellae. Ankles may be deformedor unstable.

SPINECongenital absence of part of the sacrum isthought rarely to be a manifestation of thalido-mide damage. Later changes in the spine (lossof joint space, anterior fusion of vertebralbodies) affect principally the low thoracic andlumbar spine.

EARS, EYES, AND ASSOCIATED DEFECTSThe second most common group of defects,grouped because of their tendency to occur in avariety of combinations and permutations,involves developmental abnormalities of theear and eye, and abnormalities of innervationof the external ocular muscles, the facialmuscles, and the tear glands. Other thalido-mide defects in this anatomical region includecleft palate, bifid uvula (which may be thoughtof as a minimal degree of cleft palate), andchoanal atresia. The overall facial featureswhich characterise a number of affected sub-jects are better illustrated than described.

EarsEar defects tend to be bilateral and fairlysymmetrical. In the most extreme cases, thepinna is completely absent (anotia) and theexternal auditory meatus is a blind pit. Such anear is inevitably profoundly deaf. There maybe fleshy skin tags (accessory auricles) wherethe ear should be. Less severe is microtia, inwhich there is an attempt to form an ear: hereagain there may be accessory auricles. It isoften easy to decide that one pinna is smallerthan the other, but less easy to decide whetherthe difference is any greater than the normalasymmetry of ears, and whether a pair ofpinnae of the same size are smaller than theyshould be. If the pinnae are normal, the exter-nal auditory meati (usually both) may be nar-row or tortuous or both. These narrow meatireadily become obstructed by wax or, lesscommonly, cholesteatomata, leading to recur-rent deafness.

Facial palsyWeakness of the facial muscles (usually affect-ing the whole face, but occasionally only part)is much more often unilateral than bilateral,and is almost invariably associated with anotiaor microtia on the same side. It is also com-monly associated with the other defects ofinnervation described below.

EyesThe most common structural defects of theeyes are coloboma of the iris, with or withoutcoloboma of the retina, and underdevelopmentof the globe leading to anophthalmos or micro-

Figure 3 Radiograph ofdorsolumbar spine showinganterior fusion andfragmentation ofvertebrae, irregularity ofintervertebral spaces, andloss of lumbar lordosis.Spinal flexion was greatlyrestricted.

720

on January 14, 2022 by guest. Protected by copyright.

http://jmg.bm

j.com/

J Med G

enet: first published as 10.1136/jmg.29.10.716 on 1 O

ctober 1992. Dow

nloaded from

Page 6: Recognition of thalidomide defects

Recognition of thalidomide defects

phthalmos. Coloboma and microphthalmosare quite often associated. Both defects are

- predominantly bilateral.Dermoid cysts on the surface of the eye are

3less common and tend to occur in associationwith anotia or microtia (which may lead todiagnostic difficulties: see Differential dia-gnosis below). Abnormal eyes are, for obviousreasons, associated with poor visual acuity, butvision may also be poor in structurally normal

Figure 4 Extensive centralfacial naevus, severemicrophthalmia withcoloboma of the left andright eyes. This child alsohad severe reductiondeformities of all fourlimbs.

eyes.Defects of eye movements are nearly always

associated with ear defects, often with facialweakness, and tend to be bilateral. Most com-monly, abduction of the eye is restricted orentirely absent, sometimes wrongly describedas a VIth nerve palsy, the fault being in thebrain stem connections, not in the lower motorneurone. Next most commonly, both abduc-tion and adduction are affected. Rarely, eyemovements may be even more restricted, oraberrant eye movements may occur (for ex-ample, when looking to the right, the left eyedeviates upwards). Defects of eye movementmay also present diagnostic difficulties (seeDifferential diagnosis below).

In the tear-saliva syndrome (crocodiletears), tears are secreted rather than salivawhen food is eaten, and tears may not besecreted in association with crying. This re-sults from wrong nerve connections, probablyin the brain stem. It may be bilateral or unila-teral. Affected subjects usually also have de-fects of eye movements and abnormalities ofthe ears. Crocodile tears are not unique tothalidomide.

SquintNot surprisingly, squint is common in associ-ation with any of the eye abnormalities de-scribed above although not, as a rule, if there isonly a defect of eye movements. It also appearsto be more common in children affected bythalidomide than in the general population,although there is nothing specific about theirsquints.

Structural defects of the ears and eyes canscarcely be missed, and facial weakness isunlikely to be overlooked unless it is veryslight. In contrast, defects of eye movements,even when quite extensive, can easily bemissed if the range of movement is not fullytested. Similarly, crocodile tears cannot beobserved and are only recognised by asking theright question.

Cleft lip and palateThese occur among persons affected by thali-domide more often than in the general popula-tion. The deformities appear not to differ fromother facial clefts.

Summary of external defects associated

with thalidomide (figs 1 to 4)Upper limbsShoulder: hypoplasia of shoulder muscles, sca-

pula, clavicle.

Arm: total absence, prominent acromioclavi-cular joint.Upper arm: reduction deformity of humerus(upper end).Elbow: humeroulnar, radioulnar fusion.Forearm: reduction deformities of radi-us > ulna.Hand: deformities usually related to those offorearm (preaxial emphasis, for example,radial club hand).Fingers: absence, hypoplasia, fixed flexion,syndactyly (preaxial emphasis).Thumb: absence, hypoplasia, triphalangy,non-opposable.

Lower limbsHip: congenital dislocation.Thigh: reduction deformity of femur (upperend).Knee: patellar dislocation.Lower leg: reduction deformity of tibia> fi-bula.Foot: deformities usually related to those of leg(for example, club foot).Toes: polydactyly, bifid toes (preaxial em-phasis).

CraniofacialCharacteristic facies in some cases.Central facial naevus, fading over one to twoyears.Eyes: anophthalmia, microphthalmia, col-oboma of iris/retina, conjunctival dermoidcyst.Ears: anotia, microtia, accessory auricles; atre-sia, stenosis, tortuosity of external auditorymeatus.Neurology: facial palsy, restricted eye move-ments, tear-saliva syndrome.

StatureOften short because of poor growth/osteo-chondritis of spine/progressive kyphosis.

External genitaliaHypoplasia of scrotum/labia with severe lowerlimb deficiency.

FUNCTIONAL PROBLEMSA number of thalidomide damaged personsexhibit a variety of neurodevelopmental prob-lems: mental handicap, dyslexia, autism, orepilepsy. These problems appear indis-tinguishable from the same conditions inpeople not affected by thalidomide, but theyhave occurred more often than would beexpected by chance and have thereforegenerally been accepted as attributable to thedrug when associated with more characteristicfeatures.

Summary of internal defects associatedwith thalidomideHeart: patent ductus arteriosus, VSD, ASD,and pulmonary stenosis in survivors. Com-plex, especially conotruncal, lesions were seenamong early deaths.

721

on January 14, 2022 by guest. Protected by copyright.

http://jmg.bm

j.com/

J Med G

enet: first published as 10.1136/jmg.29.10.716 on 1 O

ctober 1992. Dow

nloaded from

Page 7: Recognition of thalidomide defects

Smithells, Newman

Urinary tract: absent, horseshoe, ectopic,hypoplastic, rotated kidney; hydronephrosis,megaureter, ectopic ureter, vesicoureteric re-flux, inert bladder.Genital tract: undescended, small, or absenttestis, hypospadias, cyst of hydatid of Mor-gagni; vaginal atresia, interruption of the Fal-lopian tube, bicornuate uterus.Alimentary tract: duodenal atresia, pyloric ste-nosis, inguinal hernia, imperforate anus withfistula, anorectal stenosis, anteriorly displacedanus. (Congenital absence of appendix and gallbladder have been noted at necropsy.)Orofacial: cleft palate, high arched palate, bifiduvula, palatal palsy, cleft lip, choanal atresia,small mandible, conjunctival dermoids; ab-sent, overcrowded, or maioccluded teeth.Skeletal: sacral agenesis, hemivertebrae, ribanomalies.Neurodevelopmental problems: mental handi-cap, epilepsy, dyslexia, receptive dysphasia,behaviour disorder (including autistic andhyperkinetic), involuntary movements. Someof these defects have been recorded only onceor twice, and the association with thalidomidemay be coincidental, but most of the defectslisted have been seen more frequently thanwould be expected by chance.A number of acquired diseases (for example,

coeliac disease, diabetes, multiple sclerosis)have been seen, but no more frequently thanexpected by chance. A causal relationship isunlikely.

Differential diagnosisLIMB DEFECTSSome of the conditions which have causedproblems in the past will not do so in the futurebecause they are associated with perinatal or

early childhood death. These include shortlimbed dwarfism, which should not presentdiagnostic difficulty (for example, achondro-genesis, thanatophoric dwarfism, severe osteo-genesis imperfecta), and pseudothalidomidesyndrome (Roberts syndrome, SC syndrome),

ttll.A,.

A~~

Figure Postaxial reduction deformities of hand and

foot (different people), not caused by thalidomide.

an autosomal recessive disorder which in-cludes limb reduction deformities.

Life expectancy is more variable in the TAR(thrombocytopenia-absent radius) syndrome,an autosomal recessive disorder in whichthrombocytopenia tends to improve and maynot be evident after the neonatal period, and inwhich absent radii are associated with normalthumbs, and in the Cornelia de Lange syn-drome, in which the limb defects are bizarreand asymmetrical, and other features oftensuggest the diagnosis at birth.

Radial aplasia is a feature of Fanconi's pan-myelopathy, but the blood changes indicatethe diagnosis. The family history may indicateautosomal dominant radial aplasia (thoughnot, of course, new mutations). Radial andexternal ear defects may be associated withdeafness, eye, cardiac, and dental defects in thelacrimo-auriculo-dento-digital (LADD) syn-drome. The maternal history will help toidentify diabetic embryopathy (which does notclosely resemble thalidomide embryopathy).Amniotic band lesions most often affect a

single limb, are rarely symmetrical, and re-semble 'congenital amputations'. Ring constric-tions may be present on one or more limbs.

Poland anomaly is unilateral, the hand de-fect being associated with agenesis of part ofthe pectoralis major muscle. There may behomolateral deficiency of the breast, nipple, orribs.

In the femur-fibula-ulna (FFU) syndrome,the named bones are principally affected, con-trasting with thalidomide which affects theradius and humerus before the ulna, and thetibia before the fibula. The defects may be veryasymmetrical.The major difficulty is presented by the

Holt-Oram syndrome, an autosomal dominantdisorder usually affecting the hands and fore-arms symmetrically, and associated in almostall cases with congenital heart disease, princi-pally atrial septal defect. The family historyoften helps, but new mutations occur.

EYES, EARS, ETCFive syndromes need to be considered here,one of which can cause considerable difficulty.Goldenhar syndrome (oculoauriculovertebraldysplasia), which merges with hemifacial mi-crosomia, is characterised by microtia, acces-sory auricles, epibulbar dermoids, and abnor-malities of the cervical spine.Wildervanck syndrome (seen predominantly

in girls) is characterised by malformed ears,deafness, and defects of the cervical spine.Thalidomide rarely affects the cervical spine.Mobius syndrome may manifest as facial/

ocular palsies.Duane syndrome is a disorder of ocular

movements characterised by (1) decreased ab-duction, (2) decreased adduction, (3) retrac-tion of the globe on adduction, (4) oblique riseor depression on adduction, (5) partial closureof the eyelids on adduction, (6) deficient con-vergence. It may be bilateral or unilateral. Anassociation with other defects, especially of thehands and ears, was described as long ago as

722

on January 14, 2022 by guest. Protected by copyright.

http://jmg.bm

j.com/

J Med G

enet: first published as 10.1136/jmg.29.10.716 on 1 O

ctober 1992. Dow

nloaded from

Page 8: Recognition of thalidomide defects

Recognition of thalidomide defects

1918. Some or all of these features, togetherwith the Marcus Gunn 'jaw winking' pheno-menon, occasionally accompany thalidomidefacial defects.The LADD syndrome has been considered

above.To confuse the picture still further, medical

publications contain examples of children whoappear to show a mixture of features of morethan one syndrome, for example Mobius syn-drome and Poland anomaly. Whether thesechildren are manifesting two separate syn-

dromes, an entirely different syndrome, or

some unusual 'intermediate' manifestation can

only be a matter for speculation and furtherresearch.

Selected bibliography (in chronological order)Knapp K, Lenz W. Die Formen der Thalidomid-Embryo-

pathie. Roentgen-Europ 1962;5:105-27.Leck IM, Millar ELM. Incidence of malformations since the

introduction of thalidomide. BMJ 1962;ii:16-20.Smithells RW. Thalidomide and malformations in Liverpool.

Lancet 1962;i: 1270-3.Taussig H. A study of the German outbreak of phocomelia.

JAMA 1962;180:1106-14.

von Weicker H, Bachmann KD, Pfeiffer RA, Gleiss J. Thalido-mid-Embryopathie. Dtsch Med Wochenschr 1962;87:1597-607.

Lenz W, Das Thalidomid Syndrom. Fortschr Med 1963;81:148-55.

Smithells RW, Leck I. The incidence of limb and ear defectssince the withdrawal of thalidomide. Lancet 1963;i:1095-7.

HMSO. Deformities caused by thalidomide. Reports on PublicHealth and Medical Subjects No 112. London: HMSO,1964.

Kreipe U. Missbildungen innere Organe bei Thalidomid-Embryopathie. Arch Kinderheilkd 1967;176:33-61.

Henkel L, Willert HE. Dysmelia: a classification and a patternof malformation in a group of congenital defects of thelimbs. J Bone joint Surgery (B) 1969;51:399-414.

Kajii T, Kida M, Takahashi K. The effect of thalidomideintake during 113 human pregnancies. Teratology 1973;8:163-6.

Smithells RW. Defects and disabilities of thalidomide children.BMJ 1973;i:269-72.

Yang T, Shen Cheng C, Wang C. A survey of thalidomideembryopathy in Taiwan. J Formosan Med Assoc 1977;76:546-62.

Quibell EP. The thalidomide embryopathy: an analysis fromthe UK. Practitioner 1981;225:721-6.

Smithells RW, Clark H. Thalidomide children: how are theynow? In: Ferguson A, ed. Advanced medicine 20. London:Pitman, 1984:184-90.

Newman CGH. Teratogen update: clinical aspects of thalido-mide embryopathy - a continuing preoccupation. Terato-logy 1985;32:133-44.

Newman CGH. The thalidomide syndrome: risks of exposureand spectrum of malformations. Clin Perinatol1986;13:555-73.

Stromland K, Miller M, Cook C. Ocular teratology. SurvOphthalmol 1991;35:429-46.

723

on January 14, 2022 by guest. Protected by copyright.

http://jmg.bm

j.com/

J Med G

enet: first published as 10.1136/jmg.29.10.716 on 1 O

ctober 1992. Dow

nloaded from