J Med Genet 1996;33:193-196

Arachnoid cyst and chronic subduralhaematoma in a child with osteogenesisimperfecta type III resulting from thesubstitution of glycine 1006 by alanine in thepro cL2(I) chain of type I procollagen

W G Cole, T P Lam

AbstractThe features of a child with osteogenesisimperfecta type III (01 III) resulting fromthe heterozygous substitution of glycine1006 by alanine in the pro a2(I) chain oftype I procollagen were studied. He wasborn at term with the clinical features ofsevere OI, including deep grey-bluesclerae. He had severe osteopenia and alllong bones were smaller than normal withcortical thinning, metaphyseal expansion,poor metaphyseal modelling, and multiplefractures. However, the vertebrae, pelvis,and shoulder girdle were of normal shapeand there were few rib fractures. Histo-logical examination of the calvarium andtibial shaft showed woven bone withoutlamellar bone or Haversian systems. Theshafts of the long bones were widenedowing to repeated fractures. Progressiveenlargement ofthe calvarium occurred be-tween 3 and 4-5 months of age owing tobilateral chronic subdural haematomataand a large arachnoid cyst in the Sylvianfissure. The cyst was probably de-velopmental in origin while the subduralcollections were probably the result ofperinatal skull trauma. The cyst and thesubdural collections resolved followingdrainage but ventricular dilatation withnormal cerebrospinal fluid pressure fol-lowed. The proband is the first reportedcase of 01 with a glycine substitution byalanine in the pro a2(I) chain of type Iprocollagen.(JMed Genet 1996;33:193-196)

Centre for the Studyof HeritableConnective TissueDisorders,Research Institute,The Hospital for SickChildren,555 University Avenue,Toronto, Ontario,Canada M5G 1X8W G ColeT P Lam

Correspondence to:Professor Cole.

Received 5 July 1995Revised version accepted forpublication 13 November1995

Key words: arachnoid cyst; subdural haematoma; os-teogenesis imperfecta type III.

Osteogenesis imperfecta type III (OI III) isusually caused by dominant negative mutationsof the COLlAl or COL1A2 genes that encodethe oal (I) and at2(I) chains of type I collagen,respectively. 13 We previously reported the bio-chemical findings in the first case ofOI resultingfrom a Gly substitution by Ala in the at2(I)chain of type I collagen.4 The proband was

heterozygous for a transversion of G-3287 toC in ot2(I) cDNA, which converted the GGCcodon for Gly 1006 to GCC for Ala. The pointmutation was located in exon 49 of COL1A2.

..

Figure 1 Clinical appearance at 6 months of age.

In this paper we describe the clinical, radio-graphical, and pathological features of this childwith OI III.

Case reportCLINICAL HISTORYThe male proband was the first child of anunrelated 33 year old father and 28 year oldmother. Both parents were healthy and neitherhad any clinical or biochemical features of OI.The proband was born by caesarian sectionwith a birth weight of 3045 g which was on the15th centile for the period of gestation. He hadthe typical clinical features of 01 III (fig 1).'2His sclerae were deep grey-blue. Crepitus, asa result of fresh fractures, was present in mostof the long bones of the limbs.

RADIOGRAPHICAL FEATURESThe skeleton was poorly ossified. At birth allthe long bones were smaller than normal withcortical thinning, metaphyseal expansion, and

193

on Decem

ber 6, 2021 by guest. Protected by copyright.

http://jmg.bm

j.com/

J Med G

enet: first published as 10.1136/jmg.33.3.193 on 1 M

arch 1996. Dow

nloaded from

Cole, Lanm

Figure 4 Lateral radiograph of the skull at 1 month ofage. The base of the skull, orbits, and face are betterossified than the calvarium.

Figure 2 Anteroposterior radiograph of the chest at Imonth of age. The chest is narrow proximally. The ribs arenarrow with expanded metaphyses. The vertebrae andshoulder girdle are well formed.

multiple fresh and healing fractures (figs 2 and3). The calvarium was poorly mineralised butthe occiput, base of the skull, and orbits werebetter ossified (fig 4). There were numerousWormian bones in the occipital sutures. Themidface was hypoplastic.

PROGRESSHis respiratory function was satisfactory at birthas there were few rib fractures. The mobilefresh fractures of the femora, tibiae, and oneforearm healed rapidly, but multiple new frac-tures occurred with normal handling of thebaby.At 1 month of age, his head circumference

was 39 cm (75th centile). Between 3 and 4 5months of age, his mother noted that his cal-varium was progressively enlarging. The headcircumference of 48 cm was beyond the 98thcentile of 45 cm and he had the clinical featuresof raised intracranial pressure.A CT scan of the head showed bilateral

subdural haematomata, a large arachnoid cyst

Figure 3 Radiograph of the pelvis at 1 month of age. The pelvic shape is normzal. Thefemoral metaphyses are expanded. There are fractures of the femoral shafts andmetaphyses.

Figure 5 CT scan of the head at 4-5 months of age. Thechronic subdural haematomata (SDH) and the right sidedarachnoid cyst (AC) are labelled. The ventricles areenlarged. Contrast enhanced the membrane between thesubdural collections and the subarachnoid space.

arising from the right Sylvian fissure, and vent-ricular dilatation (fig 5). A right temporal crani-otomy was undertaken to treat these disorders.The temporal bone was extremely thin. Thedura mater was expanded and bluish. Thesubdural collections contained xanthochromicfluid and were, therefore, classified as chronicsubdural haematoma. The lining membrane ofthe right subdural haematoma was adherent tothe membrane surrounding the arachnoid cyst.There was no obvious communication betweenthem. The arachnoid cyst contained clear fluidwithout any evidence of haemorrhage. Thesurrounding neural anatomy was normal. Thecyst wall was laid open. A catheter drained thecyst, subdural space, and basal cistern into theperitoneum. The subdural haematoma and thearachnoid cyst resolved but the ventricular sizesincreased without increased cerebrospinal fluidpressure (fig 6). The shunt was revised toa ventriculoperitoneal shunt. There were noneurological signs at any time.When last reviewed at 3 years of age he was

having few fractures. The long bones of thelimbs were abnormally wide and their corticeswere abnormally thin. His head circumference

194

on Decem

ber 6, 2021 by guest. Protected by copyright.

http://jmg.bm

j.com/

J Med G

enet: first published as 10.1136/jmg.33.3.193 on 1 M

arch 1996. Dow

nloaded from

Arachnoid cyst and chronic subdural haematoma in a child with osteogenesis imperfecta type III

Figure 6 CT scan of the head at 13 months of age. Thesubdural haematomata and the arachnoid cyst haveresolved. The ventricles are dilated.

had stabilised. He had delayed gross motordevelopment but no other neurological ab-normalities. He had gained head control andsome trunk control but showed no signs ofwishing to stand. He had ligamental laxity andmuscle hypotonia. He also had severe dent-inogenesis imperfecta.

PATHOLOGICAL FEATURESThe calvarial bone removed during the tem-poral craniotomy was very thin and the innerand outer tables were fused. Histological ex-amination showed woven bone without anytrabecular bone. The cortex of the tibial shaftalso consisted of woven bone with numerous

plump osteoblasts and little intervening bonematrix. There was no trabecular bone and noHaversian systems.The membrane surrounding the subdural

haematomata contained dense collagenousconnective tissue with focal areas ofgranulationtissue. The membrane surrounding the arach-noid cyst was lined with typical spindle shapedarachnoid cells.

DiscussionThe proband was born at term as is usual forbabies with OI III.' Radiographs of neonateswith OI III may show thin or thick long bonesalthough none of them resembles the broadcrumpled long bones ofOI II.`2 In the proband,most of the shafts of the long bones wereabnormally wide owing to poor metaphysealmodelling and to thickening of the bone fromexternal fracture callus. Thickening ofthe tibiaeand femora was more marked at 2-5 years thanat birth. It is uncertain whether the shafts ofthe proband's long bones will progressivelynarrow during childhood as reported in othercases with 01 II1.Cephalhaematoma, subdural haematoma,

and intracranial haemorrhage are recognisedcomplications of trauma to the skull duringthe latter part of the pregnancy and duringdelivery.5 These anomalies are usually evidentat birth or in the neonatal period. The prob-

and's chronic subdural collections are unusualin OI but were probably the result of shearinginjuries to the skull in the perinatal period.6Arachnoid cysts have not previously been

reported in OI. They are usually developmentalin origin and can be familial.78 Subarachnoidcysts and subdural haematomata may also co-exist.9 We propose that the proband's cyst wasdevelopmental rather than traumatic in originas it did not contain xanthochromic fluid. Ba-bies with developmental arachnoid cysts oftenpresent at a later age than the proband.7 It islikely that his early presentation was becauseof the added chronic subdural collections. Al-though the subdural collections and the cystresolved with treatment, ventricular dilatationand cerebral atrophy remained.There is only one other report of the bone

histology in patients with OI III and definedtype I collagen mutations.'° Our findings ofsevere osteopenia, woven bone, plump os-teoblasts with minimal intervening matrix, andlack of lamellar bone and Haversian systemswere also reported in a case of OI III resultingfrom the substitution of Gly 427 by Arg inthe triple helical domain of the oa2(I) chain.'0However, these changes are not specific forOI III as they are also shared by babies withlethal perinatal 01 II resulting from the sub-stitutions of Gly by Arg at residues 391, 667,and 976, Gly by Val at residues 256, 973, and1006 of the ocl (I) chain, and Gly 700 by Aspin the oc2(I) chain."1-'3There are no other reported glycine sub-

stitutions at residue 1006 of the oa2(I) chain.However, a substitution of Gly 1006 by Val inthe ocl (I) chain produces lethal perinatalO0 II. 12 14 Only two Gly to Ala substitutions havebeen reported in the ocl(I) chain, at residues910 and 928, and both produced the OI IIphenotype. 1415

This work was supported by the Medical Research Council andthe Research Institute of The Hospital for Sick Children. Wethank R Humphreys for the neurosurgical findings.

1 Sillence DO, Senn A, Danks DM. Genetic heterogeneity inosteogenesis imperfecta. 7 Med Genet 1979;16:101-16.

2 Sillence DO, Barlow KK, Cole WG, Dietrich S, Garber AP,Rimoin DL. Osteogenesis imperfecta type III delineationof the phenotype with reference to genetic heterogeneity.AmJrHum Genet 1986;23:821-32.

3 Byers PH, Wallis GA, Willing MC. Osteogenesis imperfecta:translation of mutation to phenotype. J Med Genet 1991;28:433-42.

4 Lu J, Costa T, Cole WG. A novel G1006A substitution inthe oa2(I) chain of type I collagen produces osteogenesisimperfecta type III. Hum Mutat 1995;5: 175-8.

5 Bailey JA. Forms of dwarfism recognizable at birth. ClinOrthop 1971;76:150-9.

6 Tokoro K, Nakajima F, Yamataki A. Infantile chronic sub-dural hematoma with local protrusion of the skull in acase of osteogenesis imperfecta. Neurosurgery 1988;22:595-8.

7 Martinez-Lage JF, Poza M, Sola J, Puche A. Congenitalarachnoid cyst of the lateral ventricles in children. ChildsNerv Syst 1992;8:203-6.

8 Pomeranz S, Constantini S, Lubetzki-Kom I, Amir N.Familial intracranial arachnoid cysts. Childs Nerv Syst1991;7:100-2.

9 Eustace S, Toland J, Stack J. CT and MRI of arachnoid cystwith complicating intracystic and subdural haemorrhage. JComput Assist Tomogr 1992;16:995-7.

10 Sztrolovics R, Glorieux FH, Travers R, van der Rest M,Roughly PJ. Osteogenesis imperfecta: comparison of mo-lecular defects with bone histological changes. Bone 1994;15:321-8.

11 Cole WG, Chow CW, Rogers JG, Bateman JF. The clinicalfeatures of three babies with osteogenesis imperfecta re-sulting from the substitution of glycine by arginine in thepro at(I) chain of type I procollagen. J Med Genet 1990;27:228-35.

195

on Decem

ber 6, 2021 by guest. Protected by copyright.

http://jmg.bm

j.com/

J Med G

enet: first published as 10.1136/jmg.33.3.193 on 1 M

arch 1996. Dow

nloaded from

Cole, Lam

12 Cole WG, Patterson E, Bonadio J, Campbell PE, FortuneDW. The clinicopathological features of three babies withosteogenesis imperfecta resulting from the substitution ofglycine by valine in the prooal(I) chain of type I pro-collagen. 7 Med Genet 1992;29: 112-8.

13 Cohen-Solal L, Zylberberg L, Sangalli A, Gomez Lira M,Mottes M. Substitution of an aspartic acid for glycine 700in the x2(I) chain of type I collagen in a recurrent lethaltype II osteogenesis imperfecta dramatically affects the

mineralization of bone. ]f Biol Chem 1994;269:14751-8.14 Lamande SR, Dahl HH, Cole WG, Bateman JF. Char-

acterization of point mutations in the collagen COLlAland COL1A2 genes causing lethal perinatal osteogenesisimperfecta. Biol Chem 1989;264:15809-12.

15 Valli M, Sangalli A, Rossi A, et al. Osteogenesis imperfectaand type-I collagen mutations. A lethal variant caused bya Gly9 10 --Ala substitution in the alpha 1(I) chain. Eur]Biochem 1993;211:415-9.

196

on Decem

ber 6, 2021 by guest. Protected by copyright.

http://jmg.bm

j.com/

J Med G

enet: first published as 10.1136/jmg.33.3.193 on 1 M

arch 1996. Dow

nloaded from