627J Med Genet 1995;32:627-632

Congenital anomalies and genetic disorders infamilies of children with central nervous systemtumours

Susan M Jones, Peter C Phillips, Patricia T Molloy, Beverly J Lange,Michael N Needle, Jaclyn A Biegel

Division of HumanGenetics andMolecular Biology,5th Floor,Wood Building,The Children'sHospital ofPhiladelphia andthe University ofPennsylvania Schoolof Medicine,34th Street & CivicCentre Boulevard,Philadelphia,PA 19104, USAS M JonesJ A Biegel

Division of Neurology,The Children'sHospital ofPhiladelphia andThe University ofPennsylvania Schoolof Medicine,Philadelphia,PA 19104, USAP C PhillipsP T MolloyM N Needle

Division of Oncology,The Children'sHospital ofPhiladelphiaand The University ofPennsylvania Schoolof Medicine,Philadelphia,PA 19104, USAP C PhillipsP T MolloyB J LangeM N Needle

Correspondence to:Dr Biegel.Received 16 December 1994Revised version accepted forpublication 27 March 1995

AbstractMedical genetic histories of 165 childrenwith primary central nervous system(CNS) tumours and 4599 relatives oftheseprobands were examined to identify birthdefects or genetic disorders that may beassociated with the aetiology of CNS tu-mours. Twelve primary malignancies werefound in 329 (4%) of the parents of prob-ands. Two of 99 half sibs but no full sibshad malignancies. Twenty-four percent offamilies had histories warranting con-

sultation for an inherited disorder or birthdefect. Single instances of malformationsor genetic disorders were reported in 36families and several disorders were re-

ported in more than one family, includingfamilial hypercholesterolaemia (4), olivo-pontocerebellar atrophy (2), and familialabdominal aortic aneurysm (2). Althoughrecurring abnormalities were not iden-tified in probands, it is possible that one

or more of the birth defects or geneticdisorders observed in probands or rel-atives may be associated with CNStumourigenesis.

(J Med Genet 1995;32:627-632)

Recent advances in molecular genetics andcytogenetics have increased our understandingof the genetic basis of cancer.'` However, al-though brain tumours are the most commonsolid tumours in children, the genetic ae-

tiologies of childhood primary central nervoussystem (CNS) tumours are largely unknown.Some genetic disorders, particularly the

phakomatoses, are associated with an increasedincidence of CNS tumours. Furthermore, mo-lecular genetic abnormalities identified in thesesyndromes are likely to have important im-plications for our understanding of tumouri-genesis. For example, identification andcloning of the neurofibromatosis 1 (NFl)gene' has permitted detection of germlinemutations in NF1 patients7 and somatic mut-ations of the NFl gene in sporadic cases ofNFl associated tumours.8 Studies of the VonHippel-Lindau syndrome (VHLS) gene9-"2 andRET have shown similar findings."3

In contrast to the well characterised geneticpredisposition to tumours observed in NF1and VHLS, the vast majority ofchildhood CNStumours are sporadic and develop in childrenwithout identified risk factors.'4 However, as

genes for specific disorders are identified, on-cology programs will face new issues con-cerning patient and family risk assessment. Fewoncology programs are organised to providesuch service, since genetic risk assessment andcounselling have traditionally been managed byclinical genetics services. Similarly, few clinicalgenetics services are prepared to provide gen-etic counselling to large volumes of patientswith different types of malignancies. The iden-tification of genes associated with inheritedpredisposition to malignancy may result in in-creased physician referral and patient demandfor genetic consultation regarding cancer.

Because of the paucity of knowledge re-garding the aetiologies of primary paediatricbrain tumours and the relative absence of ap-propriate models for clinical genetic diagnosticand counselling services in childhood braintumour patients, the Division of Human Ge-netics and Molecular Biology and the Divisionof Oncology at The Children's Hospital ofPhiladelphia jointly undertook a survey of fam-ily histories of children with primary CNS tu-mours. The primary objective of this study wasto identify constellations of single occurrencesof congenital anomalies or genetic disorderswhich could be associated with CNS tumouri-genesis. Such associations may help identifyareas of the genome in which to search forcandidate genes for CNS malignancies. Anadditional objective was to examine families'use of genetic services when referral for suchservice was made in the setting of the Neuro-oncology service. This paper describes the or-ganisation and structure of this project andsummarises the findings in 4599 relatives in165 families.

Materials and methodsA genetic counsellor met families of childrenwho were followed by the Neuro-oncology Ser-vice for primary brain or spinal cord tumours,were not adopted, and did not have NF1. NF1patients were excluded because the NFl genehas been identified, the mode of inheritanceand recurrence risk of this disorder are known,and families of NF1 patients receive geneticcounselling through The Children's Hospitalof Philadelphia Neurofibromatosis Clinic. Abrief prenatal history of the proband was ob-tained. Standard medical information was re-corded about the patient as well as his/herparents, grandparents, great grandparents, sibs,aunts, uncles, and first cousins. If the history

on Septem

ber 27, 2020 by guest. Protected by copyright.

http://jmg.bm

j.com/

J Med G

enet: first published as 10.1136/jmg.32.8.627 on 1 A

ugust 1995. Dow

nloaded from

62ones, Phillips, Molloy, Lange, Needle, Biegel

Table 1 Identification of increased familial risk of malignancy, birth defects, and genetic disease

Did anyone in the family have a tumour that began in the brain or spinal cord? If so, how old was the person when this was diagnosed?Did anyone in the family have cancer? If so, at what site(s) did this begin? How old was the person when this was diagnosed?Did anyone in the family have a benign, or non-cancerous, tumour? If so, where was this tumour? How old was the person when this was diagnosed?Did anyone in the family have spina bifida or anencephaly, which is failure of the spine or brain to close and form properly during early pregnancy?Did anyone in the family have a birth defect or defects? If so, what was the birth defect or defects? What was the cause of this, if known?Were any of your relatives mentally retarded or "slow"? If so, was the cause of mental retardation investigated? If so, what was the cause, if determined?Did anyone in the family have seizures or epilepsy?Did any of the women or partners of any of the men in the family have miscarriages or stillbirths? If so, how many? At what time during the pregnancy

did the loss occur? Was the cause of pregnancy loss determined? (For stillborns: Did the fetus have any birth defects?)Did anyone in the family die in infancy? If so, what was the cause of death?Did anyone in the family die in childhood or as a teenager? If so, what was the cause of death?Did anyone in the family die as a relatively young adult, that is, in the 20s, 30s, or 40s? If so, what was the cause of death?Did anyone in the family have a growth disorder, that is, was anyone larger or smaller than expected for family norms? Did anyone have a medicaldisorder that caused them to be tall or short, other than having or being treated for a brain tumour?Did anyone in the family have irregularities of the skin such as moles or birthmarks? If so, describe these. Where were they? How many were there?

indicated an area where information about ad-ditional relatives was desired, the pedigree wasexpanded accordingly. However, these ad-ditional relatives were not included in datatabulation. In addition to obtaining routinemedical information, the genetic counsellorasked a series of questions (table 1) to identifydisorders suggestive of an increased familialrisk of malignancy, birth defects, or geneticdisease. Interviews with families ofnew patientstook place within a month of the diagnosisand with families of established patients duringvisits for continuing treatment or long termfollow up care. Most contacts occurred in theoutpatient Neuro-oncology Clinic; some famil-ies were seen when their child was admitted tothe hospital.The genetic counsellor reviewed the family

history with the attending physician if the pedi-gree showed an indication for further studies,including a clinical genetics consultation. Anintellectually impaired or growth delayed childwhose disorder was deemed to be secondaryto treatment for malignancy was not considereda candidate for referral. Upon agreement fromthe physician, either the counsellor or physicianexplained the indication for genetic con-

sultation to the family. If the parents wishedto pursue this, the counsellor facilitated thereferral. Ifthe family elected genetic service, thereported diagnosis was confirmed by medicalrecords review. Such verification was notsought if the family declined service. Brief gen-etic counselling for indications of increased riskfor an autosomal recessive disorder in futurechildren because of ethnic background, for ex-

ample, sickle cell anaemia for black couples,was provided by the genetic counsellor at thetime of history taking; such counsellings were

not included in data tabulation.

Table 2 Number and percentages of tumours by histologyHistology Number Percentage of total

(rounded to nearestwhole number)

Primitive neuroectodermal tumour 61 36High grade glioma 18 11

(anaplastic astrocytoma, glioblastoma multiforme)Low grade glioma 39 24

(fibrillary or pilocytic astrocytoma)Brainstem glioma (diffuse) 3 2Brainstem glioma (focal) 7 4Ependymoma 21 13Other 16 10Total 165 100

ResultsDuring the 21 months from September 1992 toJune 1994, 167 families were asked to provide a

family history; 165 families participated. Formost families the history was contributed bythe mother alone (62%); in some families thiswas provided by both parents (25%) or thefather alone (10%); in a few families this was

contributed by the patient or maternal grand-mother (3%). A broad spectrum of primaryCNS tumours was represented among thesepatients (table 2). The distribution of tumourhistology in this study is comparable to that inpublished reports from other centres,'5-17 withthe exception of our higher incidence of prim-itive neuroectodermal tumours (PNETs). Thismay reflect referral pattem, as The Children'sHospital of Philadelphia Neuro-oncology Pro-gram is a PNET research centre. The 165families reported a total of 4599 relatives. Forprobands the average age at diagnosis was 8years 10 months; the youngest patient was aged3 months and the oldest patient, with a PNET,was aged 38 years at diagnosis.Of329 parents about whom information was

available, 12 (4%) were reported to have a

malignancy. Six mothers each reported one

tumour. The mother of a PNET patient was

reported to have died at 27 years of com-

plications of a brain tumour; however, thiscould not be confirmed, as the family deniedpermission for records review. Three mothershad breast cancer, one had ovarian cancer, andone had thyroid cancer. Six fathers reportedthe following tumours: optic pathway glioma;a benign pharyngeal tumour diagnosed at 29years and subsequently a testicular tumourdiagnosed at 30 years; bladder cancer; non-

melanoma skin cancer (2); and an unknownprimary malignancy.No probands had children. No family re-

ported malignancy in any of the 261 full sibs ofprobands. One family reported that the paternalhalf sister of a female proband with a sub-ependymoma had died of complications of a

neuroblastoma; another family reported thatthe paternal half brother of a male probandwith a pineal germinoma died of complicationsof testicular cancer. No family showed the clas-sic constellation of tumours associated withthe Li-Fraumeni syndrome, characterised bysarcomas, breast cancer, brain tumours, andleukaemia.1' 19

One or more indications for genetic con-

sultation were identified in 39 of the 165 famil-

(1)(2)(3)(4)(5)(6)(7)(8)

(9)(10)(1 1)(12)(13)

628

on Septem

ber 27, 2020 by guest. Protected by copyright.

http://jmg.bm

j.com/

J Med G

enet: first published as 10.1136/jmg.32.8.627 on 1 A

ugust 1995. Dow

nloaded from

Congenital anomalies and genetic disorders in families of children with central nervous system tumours

Table 3 Indications for genetic consultation in paediatric central nervous system tumour patients and relatives (relatives identified by relationship topatient)

Indication Evaluation

(A) Patients with additional medical findings(1) Neurofibromatosis 1 and pulmonary stenosis: Referred for genetic re-evalua

r/o neurofibromatosis-Noonan syndrome(2) Basal cell naevus syndrome, r/o Referred for genetic consultat:(3) Fibrodysplasia ossificans progressiva (FOP) Previous genetic consultation(4) Charcot-Marie-Tooth disease (autosomal Previous genetic consultation

dominant, type otherwise unspecified)(5) a Ehlers-Danlos syndrome type III Previous genetic consultation

b Developmental delay (identified after Referred for genetic consultat:initial genetics consultation

(6) Duchenne muscular dystrophy (DMD) Previous genetic consultation(7) Fragile X syndrome Previous genetic consultation(8) Multiple minor malformations, undiagnosed Referred for genetic re-evalua

(9) Hair patteming anomaly (frontal upsweep) Referred for genetic consultati(same patient as case 22)

(10) Hair patteming anomaly (lateral upsweep) Referred for genetic consultati(11) Hair patteming anomaly (vertex scalp curl) Referred for genetic consultati(12) Duplicated cystic left kidney Referred for genetic consultat:(13) Third branchial cleft sinus tracts Referred for genetic consultati(14) Preauricular pit Referred for genetic consultati(15) Pyloric stenosis Referred for genetic consultat

Pyloric stenosis, fh of(patemal uncle affected)

(B) Patients with relatives with birth defects or genetic disorders 6h =family history of)(16) Wiskott-Aldrich (WA) syndrome, fh of Previous genetic consultation

(matemal first cousin once removed affected;matemal uncle possibly affected)

(17) Von Hippel-Lindau syndrome (VHLS), fh of Previous genetic consultation(father affected)

(18) a Familial hypercholesterolaemia, fh of Neuro-oncologist provided ge(mother affected) consultationb Pseudocholinesterase deficiency, fh of Referred for genetic consultat:(mother affected)

(19) Familial hypercholesterolaemia, ffi of Neuro-oncologist provided ge(matemal grandfather and other relatives affected) consultation

(20) Familial hypercholesterolaemia, fh of Neuro-oncologist provided ge(mother and other relatives affected) consultation

(21) Familial hypercholesterolaemia, fh of Neuro-oncologist provided ge(maternal grandmother affected) consultation

(22) Olivopontocerebellar atrophy, fh of Referred for genetic consultat:(patemal grandmother affected)(same patient as case 9)

(23) Olivopontocerebellar atrophy, fh of Referred for genetic consultat(matemal great aunt and others affected)

(24) Abdominal aortic aneurysm, fh of Referred for genetic consultat(patemal great aunts, great uncles, and first cousin affected)

(25) Abdominal aortic aneurysm, fh of Referred for genetic consultat(mother and matemal grandmother affected)

(26) Aortic aneurysm, fh of Referred for genetic consultat(patemal grandfather affected)

(27) Intracranial aneurysms, fh of Referred for genetic consultat(matemal grandmother and great grandmother affected)

(28) Mitral valve prolapse syndrome, fh of, r/o Referred for genetic consultat:(mother and two sisters affected)

(29) Huntington's disease, fh of Referred for genetic consultat:(patemal grandfather affected)

(30) Adult onset polycystic kidney disease, fh of, r/o Referred for genetic consultati(patemal grandmother, great aunt, andgreat grandmother possibly affected)

(31) Ichthyosis, autosomal dominant, type Previous genetic consultationotherwise unspecified, fh of(mother and other matemal relatives affected)

(32) a Marfan syndrome, fh of, r/o Referred for genetic consultati(matemal aunt and uncle possibly affected)b Sickle cell trait, fh of Previous genetic consultation(mother is sickle haemoglobin heterozygote)

(33) a 3 thalassaemia trait, fh of Previous genetic consultation(father affected)Sickle cell trait, fh of(mother affected)b Down's syndrome, fh of Previous genetic consultation(patemal uncle affected)

(34) Retinitis pigmentosa, fh of Referred for genetic consultati(father and patemal uncle affected)

(35) Trisomy 13, fh of Previous genetic consultation(stillbom sister affected)

(36) Down's syndrome, matemal fh of x 3 Referred for genetic consultati(matemal cousins affected)

(37) Microcephaly and developmental delay, fh of Referred for genetic consultati(brother affected)

(38) Developmental delay, fh of Referred for genetic consultati(matemal uncle affected)

(39) Spina bifida, fh of Referred for genetic consultati(two sibs affected)

(40) Cleft lip and short fifth fingers, fh of Referred for genetic consultati(mother affected)

Comments

ition

tion

tion

tion

tion

tiontiontiontiontiontion

-netic

tion-netic

-netic

netic

tion

tion

tion

tion

tion

tion

tion

tion

tion

ion

ion

ion

ion

ion

ion

ion

Neurofibromatosis-Noonan syndrome

DeclinedPatient has spontaneous mutation for FOPFather and brother affected

Mother affected; matemal grandmother and othermaternal relatives possibly affected

Mother DMD heterozygoteMother and other matemal relatives fragile X positiveMacrocephaly, autosomal dominant: mother affectedBitemporal narrowingDeclined

DeclinedDeclinedBranchio-oto-renal syndrome, possibleIsolated malformationDeclinedDeclined

Mother possible WA heterozygote

Patient at 50% risk to have VHLS allele:has brain tumour not characteristic of VHLS

Declined

Patient's serum cholesterol within normal limits(WNL)Patient's serum cholesterol WNL

Patient's serum cholesterol WNL

Declined

Declined

Familial abdominal aortic aneurysm, autosomaldominantDeclined

No evidence of Marfan syndrome or other inbornerror of connective tissueDeclined

Mitral valve prolapse, autosomal dominant

Declined

Declined

Patient not affected

Declined

Mother's future children at 25% risk to havehaemoglobinopathy if future partner ishaemoglobinopathy heterozygoteParents' future children at 25% risk to have sickle cellf thalassaemia

Parents' future children not at increased risk tohave Down's syndromeDeclined

Parents' future children at no greater than 1% riskto have trisomy 13Declined

Declined

Declined

Declined

Declined

ies (24%) for a total of 44 indications. Table 3 Case 1 was initially thought to have NFl.summarises these. Previous genetic evaluations Although NF1 patients were not generally in-for 14 of these indications had been performed. cluded in this study, the presence ofpulmonaryIn two of these cases (1 and 8) we determined stenosis in this child prompted referral to rulea need to re-evaluate the original diagnoses. out the neurofibromatosis-Noonan syndrome.

629

on Septem

ber 27, 2020 by guest. Protected by copyright.

http://jmg.bm

j.com/

J Med G

enet: first published as 10.1136/jmg.32.8.627 on 1 A

ugust 1995. Dow

nloaded from

63ones, Phillips, Molloy, Lange, Needle, Biegel

Case 8 was a child with multiple minor mal-formations whose diagnosis remained unknownafter initial genetic assessment. Families of twoother cases (5 and 32) of the previously coun-selled group reported a newly identified secondindication for genetic evaluation. Case 5 waspreviously diagnosed with Ehlers-Danlos syn-drome type III (ED III), inherited from hisintellectually normal mother. Developmentaldelay in this child was identified after the initialgenetic assessment, and at his Neuro-oncologyClinic visit referral was made to the clinicalgenetics service for re-evaluation. Case 32 wasa child whose mother had had genetic coun-selling for sickle cell trait. She reported relativeswith medical findings strongly suggestive ofMarfan syndrome. These four families werereferred for clinical genetic services, along withthe remaining 26 families not having previousgenetic consultation. Thus, 30 ofthe 165 famil-ies were referred for genetic services. No re-ferrals were made for advanced maternal age,as no mothers over the age of 35 years who werepregnant or planning pregnancy were seen.

Sixteen of the indications for genetic coun-selling were for a history in the proband ofbirth defect(s) or additional medical findings(table 3A). Five patients (cases 3, 4, 5a, 6, 7)and their families had been evaluated by aclinical geneticist or a non-geneticist physicianfor diagnoses of, respectively, fibrodysplasiaossificans progressiva, autosomal dominantCharcot-Marie-Tooth disease (type otherwiseunspecified), ED III, Duchenne muscular dys-trophy, and fragile X syndrome. These familiesdid not require further clinical genetics followup for these diagnoses. Patients with the re-maining 11 indications were referred for initialor repeat genetic consultation; of these, fivefamilies (45%) elected evaluation. Case 1 wasfound upon re-evaluation to have the neuro-fibromatosis-Noonan syndrome. The family ofcase 2, a medulloblastoma patient, declinedreferral to rule out the basal cell naevus syn-drome in the patient. The referral was promp-ted by the mother's report that her daughterhad numerous naevi in addition to the me-dulloblastoma, and that the child's father hadnumerous naevi and paranoid schizophrenia.No other syndromes associated with tumouri-genesis were identified in probands.

Twenty-eight of the indications were for pos-sible increased risk for a birth defect or geneticdisorder in living or future family membersbecause of the presence of an affected rel-ative(s) (table 3B). Seven of these families hadbeen evaluated by a clinical geneticist or a non-geneticist physician and did not require furtherfollow up regarding these diagnoses. The re-maining families had not had genetic con-sultation and were referred for such service.The neuro-oncologists provided genetic con-sultation for the four families (cases 18-21)with histories consistent with familial hyper-cholesterolaemia (FH), one of the most com-mon autosomal dominant disorders in humans.These probands had normal serum cholesterollevels. Of the remaining 17 families who hadnot had a genetics consultation, three electeda genetics evaluation.

Most parents who declined a genetics as-sessment cited as the reason for declining thewish to avoid more medical evaluation andstress. Other reasons included: fear of beingidentified as having a genetic "responsibility"for the tumour or other disorder; fear of beingidentified as affected with or at risk for a seriousgenetic disorder; lack of interest in de-termination of affected or at risk status for agenetic disorder; inability to obtain cooperationfrom relatives for the evaluation; preoccupationwith other family problems.One family (case 22) reported that the

proband's paternal grandmother had olivo-pontocerebellar atrophy (OPCA). Anotherfamily (case 23) reported that the proband'smaternal grandmother's two sisters had thisdisorder, and that the father of this sibship,who died of lung cancer, had a "neurologicaldisorder". The grandmother herself was re-ported to have been diabetic at her death aged60 years of a "heart attack". These diagnoseswere not confirmed, as both families declinedgenetic consultation. In both probands thetumour was a posterior fossa PNET. Fourfamilies reported histories consistent withfamilial hypercholesterolaemia. Two families(cases 24 and 25) reported a history of familialabdominal aortic aneurysm (FAAA). One fam-ily (case 24) elected genetic consultation, andthe reported diagnosis was confirmed in fiverelatives in two generations.

DiscussionThe primary objective of this study was toidentify paediatric CNS tumour patients withmedical or family history findings potentiallyrelated to tumour aetiology. Birth defects orgenetic disorders present in more than oneproband or relative would suggest that cyto-genetic and molecular genetic analyses of thesepeople might provide information useful inthe localisation of genes associated with CNStumourigenesis. Such approaches successfullyled to the identification of the retinoblastomagene at 13ql420 and the Wilms' tumour geneWT-1 at 11p1321 and have implicated aBeckwith-Wiedemann syndrome locus on1lp15.22One strategy for identifying factors related to

tumourigenesis has been to look for structuralabnormalities in children with cancer. Meheset al23 reported a significantly increased pre-valence of minor anomalies in both childrenwith cancer and their sibs, although no specificassociation of a given pattern of anomalieswith a particular malignancy was identified.Schumacher et a124 found an increased in-cidence of rib anomalies in 1000 children withsome, but not all, types of cancer. Both groupssuggested that the malignancy and minor an-omalies might be different aspects of a singleabnormality in embryogenesis. However, nei-ther study evaluated family histories of pro-bands for genetic disorders or malformationsyndromes.Although we did not necessarily expect to

find a common malformation syndrome inprobands that was previously unrecognised, we

630

on Septem

ber 27, 2020 by guest. Protected by copyright.

http://jmg.bm

j.com/

J Med G

enet: first published as 10.1136/jmg.32.8.627 on 1 A

ugust 1995. Dow

nloaded from

Congenital anomalies and genetic disorders in families of children with central nervous system tumours

hypothesised that by surveying a paediatricCNS tumour population we might identify arare syndrome or single malformation po-tentially associated with CNS tumourigenesis.Fifteen of 165 patients in our study populationhad birth defects(s) or genetic disorders. Theseappeared to be random and no consistent pat-terns of associated disorders emerged. None-theless, a disorder seen only once in the studypopulation may be aetiologically associatedwith the tumour. For example, familial ad-enomatous polyposis (FAP) occurs with anincidence of 1/8000 live births.25 With ap-proximately 258 000 000 people in the UnitedStates (US),26 there are an estimated 32 250US FAP patients. The gross localisation of thegene, APC (adenomatous polyposis coli), forFAP and the allelic Gardner syndrome (GS)began with cytogenetic identification of an in-terstitial deletion of chromosome 5q in a men-

tally retarded/multiply malformed patient withFAP and a negative family history for FAP.27This permitted linkage studies and deletionmapping of tumours which eventually led tothe cloning of APC.2829 Nine other patientswith FAP or GS, mental retardation, major or

minor malformations, and an interstitial de-letion of chromosome 5q have been re-

ported.114 Ten such patients per 32 250 USFAP patients give a ratio of one such syn-dromic patient for every 3225 US FAP subjects.Thus, even in a large series of FAP patients, itwould be most unlikely that even two patientswould have a malformation syndrome relatedto an interstitial deletion of 5q. This exampleshows that the infrequent occurrence of an

associated disorder in a series ofcancer patientsdoes not exclude a relationship between thedisorder and the malignancy, not does it pre-clude the use of that disorder in the iden-tification ofgenes aetiologically associated withthe malignancy.

Several disorders (FH, FAAA, OPCA) were

reported in more than one family. FH is an

autosomal dominant disorder. In some kind-reds FAAA is inherited as an autosomal dom-inant disorder with age dependent penetrance(McKusick No 100070).3 The OPCAs are a

heterogeneous group ofdisorders characterisedby progressive ataxia, parkinsonism, and patho-logical changes in the olivary nuclei, pons, andcerebellum. Many cases are sporadic and mayhave an acquired rather than genetic aetiology.OPCA has been described in a number offamilies as an autosomal dominant disorder;inheritance may be autosomal recessive in a

minority of families.36 37 It is not known whetherthe probands with family histories of OPCA orfamilial abdominal aortic aneurysm are het-erozygotes for these disorders. Although themultiple occurrences ofFH, OPCA, and FAAAin our study population are noteworthy, we

cannot be certain that the repetitions representa constellation or that these disorders conferupon gene carriers an increased risk for CNStumours. Nonetheless, it is possible that theobserved associations are because of a geneticlesion common to both the heritable disorderand the tumour. Of particular interest is thatthe cerebellum is both a site affected in OPCA

and the site of origin of each of these twoprobands' PNETs. Furthermore, OPCA is avery rare disorder, and its occurrence in two of165 families ascertained through a probandwith a CNS tumour is unusual if one assumesthat there is no relationship between OPCAgene(s) and genes aetiologically associated withposterior fossa PNETs.We have shown that a substantial proportion

(24%) of paediatric CNS tumour patients havemedical or family histories that warrant geneticevaluation. This finding is not unexpected,given that the identification of additional in-dications for genetic consultation when a familyhistory is taken has been previously reported.38In our population most indications related torisk assessment and counselling for disorderswith no known association with CNS tumours.Of 11 probands forwhom genetic consultationswere indicated because of the presence of as-sociated findings, five underwent evaluation.In one case (case 1) establishment of the diag-nosis of the neurofibromatosis-Noonan syn-drome contributed to the understanding of theaetiology of the tumour, as patients with thisdisorder have an increased risk of developingCNS tumours. This patient and a possible caseof basal cell naevus syndrome were the onlyprobands identified with syndromes associatedwith CNS tumourigenesis.

Families had been referred for genetic ser-vices for only 14 of44 (32%) indications beforecontact with the counsellor in the Neuro-oncology service. The reasons for this under-referral are not known, but may relate to lackofphysician access to a thorough family history,the physician's and family's focus on urgentmedical issues regarding the tumour, or lack ofphysician recognition of indications for geneticconsultation. The absence of referral was not-able for several reasons. First, some of thesedisorders constituted major health problems,because they either shortened life (for example,familial hypercholesterolaemia), caused sig-nificant morbidity (for example, Huntington'sdisease), or conferred handicap (for example,retinitis pigmentosa). Second, for some dis-orders accurate recurrence risk assessment(for example, for Down's syndrome) or pre-symptomatic diagnosis (for example, for ab-dominal aortic aneurysm) through laboratoryanalysis or examination was available. Third,for some disorders medical intervention to re-duce morbidity in affected subjects could beoffered (for example, monitoring of cardio-vascular status and use of antihypertensiveagents to slow aortic dilatation in Marfan syn-drome patients). Families elected genetic as-sessment for seven of 21 (33%) indications ofpossible increased risk for a disorder in livingor future family members because of the pres-ence of an affected relative(s). However, fourof these families were provided genetic con-sultation in a convenient manner by their child'sphysician during Neuro-oncology clinic. Ifthese families had refused consultation ar-ranged through another service, the ratewould have been only 14%. In contrast, 45%of couples whose child warranted genetic con-sultation because of associated findings elected

631

on Septem

ber 27, 2020 by guest. Protected by copyright.

http://jmg.bm

j.com/

J Med G

enet: first published as 10.1136/jmg.32.8.627 on 1 A

ugust 1995. Dow

nloaded from

6Jones, Phillips, Molloy, Lange, Needle, Biegel

genetic evaluation for their child. This suggeststhat parents were likely to be uninterested inseeking genetic information with no knownrelevance to their child's tumour, but were

more likely to use such service when it mightcontribute to understanding of their child'shealth and, possibly, the cause of the tumour.The degree of malignancy of the tumours inthe families declining genetic services did notappear to be different from that of the familieselecting service, suggesting that this was not a

determinant of families' use of genetic services.This study, designed as a hypothesis gen-

erating project, provides a basis for subsequentstudies. Future larger series that confirm theobservations of the associated disorders notedin our patients would support the possibilitythat links exist between genes responsible forthese disorders and genes associated withtumourigenesis. A larger study also would havemore power to identify multiple occurrences

of those disorders seen only singly in our popu-lation. In addition, findings from the presentstudy suggest that paediatric oncology groupsplanning to incorporate genetic evaluation as

part of a multidisciplinary service will improveuse of genetic service by arranging this in theoncology clinic rather than the genetics clinic.Lastly, we have identified a group of probandsand relatives whose family histories suggest thatcytogenetic and molecular genetic analysis ofthese people may be useful in the identificationof genes associated with CNS tumourigenesis.

This work was supported by the Ethel Brown Foederer Fundfor Excellence (JAB, PCP), the National Institutes of HealthCA 46274 JAB), the National Institute of Neurological Dis-orders and Stroke NS 31102 (PCP), and the Ronald McDonaldChildren's Charities (PCP, JAB).

1 Eng C, Ponder BR. The role of gene mutations in the

genesis of familial cancers. FASEB 1993;7:910-9.2 Knudson AG. Antioncogenes and human cancer. Proc Natl

Acad Sci USA 1993;90:10914-21.3 Kelley MJ, Johnson BE. Genetic mechanisms of solid

tumorigenesis. Adv Intern Med 1994;39:93-122.4 Cawthon RM, Weiss R, Xu G, et al. A major segment

of the neurofibromatosis type 1 gene: cDNA sequence,genomic structure and point mutations. Cell 1990;62:193-201.

5 Viskochil D, Buchberg AM, Xu G, et al. Deletions and a

translocation interrupt a cloned gene at the neuro-

fibromatosis type 1 locus. Cell 1990;62: 187-92.6 Wallace MR, Marchuk DA, Andersen LB, et al. Type 1

neurofibromatosis gene: identification of a large transcriptdisrupted in three NFI patients. Science 1990;249:181-6.

7 Gutmann DH, Collins FS. The neurofibromatosis type 1

gene and its protein product, neurofibromin. Neuron 1993;10:335-43.

8 Li Y, Bollag G, Clark R, et al. Somatic mutations in the

neurofibromatosis 1 gene in human tumors. Cell 1992;69:275-81.

9 Latif F, Kalman T, Gnarra J, et al. Identification of the von

Hippel-Lindau disease tumor suppressor gene. Science

1993;269: 1317-20.10 Yao M, Latif F, Orcutt ML, et al. Von Hippel-Lindau

disease: identification of deletion mutations by pulsed-field gel electrophoresis. Hum Genet 1993;92:605-14.

11 Gnarra JR, Tory K, Weng Y, et al. Mutations of the VHLtumour suppressor gene in renal carcinoma. Nature Genet1994;7:85-90.

12 Shuin T, Kondo K, Torigoe S, et al. Frequent somaticmutations and loss of heterozygosity of the von Hippel-Lindau tumor suppressor gene in primary human renalcell carcinomas. Cancer Res 1994;54:2852-5.

13 Goodfellow PJ. Inherited cancers associated with the RETproto-oncogene. Curr Opin Genet Dev 1994;4:446-52.

14 Kuijten RR, Bunin GR. Risk factors for childhood braintumors. Cancer Epidemiol Biomarkers Prev 1993;2:277-88.

15 Young JL Jr, Ries LG, Silverberg E, Horm JW, MillerRW. Cancer incidence, survival and mortality for childrenyounger than age 15 years. Cancer 1986;58:598-602.

16 Bondy ML, Lustbader ED, Buffler PA, Schull WJ, HardyRJ, Strong LC. Genetic epidemiology of childhood braintumors. Genet Epidemiol 199 1;8:253-67.

17 Gold EB, Leviton A, Lopez R, et al. The role of familyhistory in risk of childhood brain tumors. Cancers 1994;73:1302-10.

18 Li FP, Fraumeni JF. Soft-tissue sarcomas, breast cancer andother neoplasms: a familial syndrome? Ann Intern Med1969;71:747-52.

19 Li FP, Fraumeni JF, Mulvihill JJ, et al. A cancer familysyndrome in twenty-four kindreds. Cancer Res 1988;48:5358-62.

20 Friend S, Bernards R, Rogelj S, et al. A human DNAfragment with properties of the gene that predisposesto retinoblastoma and osteosarcoma. Nature 1986;323:643-6.

21 Coppes MJ, Campbell CE, Williams BRG. The role ofWT-1 in Wilms tumorigenesis. FASEB 7 1993;7:886-95.

22 Ping AJ, Reeve AE, Law DJ, Young MR, Boenke M, Fein-berg AP. Genetic linkage of Beckwith-Wiedemann syn-drome to 11p15. AmJHum Genet 1989;44:720-3.

23 Mehes K, Signer E, Pluss HJ, Muller HJ, Stalder G. In-creased prevalence of minor anomalies in childhood malig-nancies. EurJ Pediatr 1985;144:243-9.

24 Schumacher R, Mai A, Gutjahr P. Association of rib an-omalies and malignancy in childhood. Eur7 Pediatr 1992;151:432-4.

25 Fraumeni JF. Genetic factors. In: Holland JF, Frei E, eds.Cancer medicine. 2nd ed. Philadelphia: Lea & Feibiger,1982:5-12.

26 United States Census Bureau. Current population report P-251104: population projections of the United States by age,sex, race and Hispanic origin for 2050. Washington: UnitedStates Government Printing Office, 1993.

27 Herrera L, Kakati S, Gibas L, Pietrzak E, Sandberg AA.Gardner syndrome in a man with an interstitial deletionof 5q. Am J Med Genet 1986;25:473-6.

28 Kinzler KW, Nilbert NC, Su L-K, et al. Identification ofFAP locus genes from chromosome 5q2 1. Science 1991;253:661-5.

29 Groden J, Thliveris A, Samowitz W, et al. Identification andcharacterization of the familial adenomatous polyposis coligene. Cell 199 1;66:589-600.

30 Hockey KA, Mulcahy MT, Montgomery P, Levitt S, De-letion ofchromosome 5q and familial adenomatous polyp-osis. J Med Genet 1989;26:61-8.

31 Kobayashi T, Narahara K, Yokoyama Y, et al. Gardnersyndrome in a boy with interstitial deletion of the longarm of chromosome 5. Am I Med Genet 1991;41:460-3.

32 Cross I, Delhanty J, Chapman P, et al. An interchromosomalinsertion causing 5q22 deletion and familial polyposis coliin two generations. _7 Med Genet 1992;29: 175-9.

33 Hodgson SV, Coonar AS, Hanson PJV, et al. Two cases

of 5q deletions in patients with familial adenomatouspolyposis: possible link with Caroli's disease. _7 Med Genet1993;30:369-75.

34 Lindgren V, Bryke CR, Ozcelik T, Yang-Feng TL, FranckeU. Phenotypic, cytogenetic and molecular studies of threepatients with constitutional deletions of chromosome 5 inthe region of the gene for familial adenomatous polyposis.Am J Hum Genet 1992;50:988-97.

35 McKusick VA. Mendelian inheritance in man. 10th ed. Bal-timore: Johns Hopkins University Press, 1992:3.

36 Gilman S, Bloedel JR, Lechtenberg R. Disorders of the ce-

rebellum. Philadelphia, FA Davis Co, 1981:244-50.37 Rosenberg RN, Grossman A. Hereditary ataxia. Neurol Clin

1989;7:25-36.38 Rubin SP, Malin J, Maidman J. Genetic counseling before

prenatal diagnosis for advanced maternal age: an im-portant medical safeguard. Obstet Gynecol 1983;62:155-9.

632

on Septem

ber 27, 2020 by guest. Protected by copyright.

http://jmg.bm

j.com/

J Med G

enet: first published as 10.1136/jmg.32.8.627 on 1 A

ugust 1995. Dow

nloaded from