Exostoses

We performed a prospective genotype-phenotype study using molecular screening and clinical assessment to compare the severity of disease and the risk of sarcoma in 172 individuals (78 families) with hereditary multiple exostoses. We calculated the severity of disease including stature, number of exostoses, number of surgical procedures that were necessary, deformity and functional parameters and used molecular techniques to identify the genetic mutations in affected individuals. Each arm of the genotype-phenotype study was blind to the outcome of the other. Mutations EXT1 and EXT2 were almost equally common, and were identified in 83% of individuals. Non-parametric statistical tests were used.

There was a wide variation in the severity of disease. Children under ten years of age had fewer exostoses, consistent with the known age-related penetrance of this condition. The severity of the disease did not differ significantly with gender and was very variable within any given family. The sites of mutation affected the severity of disease with patients with EXT1 mutations having a significantly worse condition than those with EXT2 mutations in three of five parameters of severity (stature, deformity and functional parameters). A single sarcoma developed in an EXT2 mutation carrier, compared with seven in EXT1 mutation carriers. There was no evidence that sarcomas arose more commonly in families in whom the disease was more severe.

The sarcoma risk in EXT1 carriers is similar to the risk of breast cancer in an older population subjected to breast-screening, suggesting that a role for regular screening in patients with hereditary multiple exostoses is justifiable.

Hereditary multiple exostoses (HME), also known as multiple hereditary exostoses, multiple osteochondromatosis or diaphyseal aclasis, is an autosomal dominant inherited trait with an incidence of at least one in 50 000, which makes it one of the most common inherited musculoskeletal conditions.1 Multiple cartilage capped exostoses develop during childhood and adolescence in the metaphyseal region of long bones, resulting in short stature and deformity, often requiring corrective surgery.1,2 Most lesions stagnate and ossify when skeletal growth is complete, but occasionally one might grow more aggressively or reactivate as a chondrosarcoma.3"5 The severity of dis ease in HME varies considerably. Some individuals achieve normal height with few exostoses, whereas others are badly affected. Radiological studies confirm this variation, yet the factors that determine severity of the disease are not known.1,2

It is possible that some, or all, of the variation is due to genetic differences. In conditions such as achondroplasia, a single gene is affected in the same way in all patients. In patients with HME, two different genes, each on different chromosomes, are associated with the condition. It has also been shown that a variety of mutations occur in these genes in HME. It has been suggested that genetic variation could account for the difference in severity of the disease.6"11 Knudson12 demonstrated a mathematical relationship between observed numbers of tumours and cellular mutation rates in familial neoplastic traits. The risk of malignant change might also be related to osteochondroma load and severity of the disease. Therefore, we designed a prospective study to assess the risk of sarcoma formation and the variation in severity with different genetic mutations, to determine whether a particular mutation had a morbid effect.

Patients and Methods

Our prospective study comprised 172 individuals (78 families) with HME. The researchers collecting the genotype data were 'blind' to the results of the phenotype data, and vice versa. Index patients with solitary osteochondromas were excluded.

The phenotype can be characterised in several ways. Most relevant to disease activity is the number of exostoses, especially as genetic epidemiology predicts this to be a major determinant of the risk of sarcoma. ' Assessing the number of osteochondromas by clinical examination alone has previously been shown to correlate with radiological data,2 suggesting that in population studies, clinical examination provides a reliable estimate of the severity of disease and avoids the theoretical risk of malignant change associated with ionising radiation.14,15

Population-based studies have defined some deformities which are inherent in HME. These include short ulnae, dislocated proximal radio-ulnar joint, forearm bowing, genu vara/valga and short stature generally. ' Handicap, however, is reflected in functional parameters, as well as anatomical ones. It is recognised that some osteochondromas have greater disabling effects than others (such as those close to the distal ulnar epiphysis) and an assessment of the disease phenotype has to take this into account." Disability (functional effects) include reduced range of movement of elbow, forearm and knee. The severity of these deforming and functional features of HME are graded in Table I. Standing height was measured and compared with age and gendermatched data from Buckler-Tanner.16 The number of exostoses, the number of previous surgical operations and the magnitude of the deforming and functional effect of osteochondroma were compared in separate genotypephenotype analyses.

Patients were selected by referral from orthopaedic surgeons, geneticists and occasionally they presented themselves, between 1996 and 2000. All had the clinical manifestations of HME.

All patients were informed of the programme and agreed to take part. Ethical approval was obtained for sample collection and analysis. A clinical history was obtained and a family pedigree. The clinical data was collected by clinical examination carried out by a single examiner (MF) who was blind to all genotypic data.

Genomic DNA was extracted either from blood, using the Nucleon DNA extraction Kit (Scotlab, Lanarkshire, UK), or from buccal swabs by a previously described protocol.10 The lab investigator was blind to phenotypic data. The initial mutational analysis of EXTl and EXT2 was carried out by conformation-sensitive gel electrophoresis and fluorescent single-strand conformation polymorphism analysis, but the greater part of the study was carried out using denaturing high-performance liquid chromatography (DHPLC).10,11 EXTl and EXT2 exons and their flanking intronic regions were amplified by polymerase chain reaction (PCR) from genomic DNA isolated from proband samples by the use of specific primer pairs. For DHPLC analysis the initial primer pairs used were as previously ment Analysis System (Transgenomic, Cheshire, UK).11 The full sequence of the PCR product for all fragments were entered into the Transgenomic WAVEMaker 3.4 program (Transgenomic). This enabled the temperature and gradient conditions for each fragment to be selected. The crude PCR products, which had been denatured, followed by gradual re-annealing, were injected into a DNASep column (Transgenomic). The column mobile phase consisted of a linear acetonitrile gradient in a 0.1 M triethylamine acetate buffer (TEAA). The calculated gradient, at a flow rate of 0.9 ml/ min, was run for all the amplicons at the relevant column temperature for each fragment.

Where a variant Chromatograph was obtained, the fragment was amplified by PCR using the same primers and the DNA sequence determined. The PCR product was purified through a spin column, using a QIAquick PCR purification kit (Qiagen, Crowley, UK), then sequenced in both directions using BigDye Terminator Cycle Sequencing Ready Reaction kit (PE-Applied Biosystems, Warrington, UK) and run on an ABI 377 DNA Sequencer (PE-Applied Biosystems). DNA from both affected and unaffected family members was sequenced.

Statistical analysis. Individuals from the same family cannot be viewed as truly independent of each other. Therefore, except for analyses of age and gender, all analyses were undertaken on the mean value of the numbers of exostoses, deformity and functional grading, in each family. No matter how many family members were assessed, only one datum point was generated for each family. Statistical analysis of all data was undertaken using non-parametric tests (chi-squared, Wilcoxon rank-sum test, and Kendall's test of rank correlation).

Results

Between 1996 and 2000, we assessed the genotype and phenotype of 172 individuals from 78 families with HME. Genotyping revealed 71 individuals (34 families) with EXTl mutation, 72 (27 families) with EXT2 and 29 (17 families) with no identifiable mutation. The majority (50 of 61 families with known mutations) were loss-of-function mutations.

The number of palpable exostoses varied considerably from site to site. Their anatomical frequency in our series is recorded in Figure 1. Age-distribution is shown in Figure 2. The number of exostoses in individuals ranged from zero to 172, but did not demonstrate a significant correlation with age (rank correlation coefficient = 0.027, p = 0.296). A trend towards fewer exostoses, however, was seen in children up to the age of ten years (Fig. 2) who were almost equally represented in both EXTl (n = 18) and EXT2 (n = 16) mutation groups.

There were 82 men and boys and 90 women and girls. There was a trend towards more severe disease in the men and boys than in the women and girls, but this failed to reach statistical significance in the several parameters of severity which were assessed (median number of exostoses 27 and 22, respectively; z = 1.840, ρ = 0.066). Boys and girls were almost equally represented in the group aged less than ten years. Gender was not significantly different between families with EXTl and EXT2 mutations.

Definite loss-of-function mutations were found in 50 families and missense mutations in 11. In 17 families no mutation could be identified. When the numbers of exostoses were compared in families with loss-of-function and missense mutations, no significant difference was noted (median number of exostoses 29 and 24, respectively; z = 0.993, p = 0.321). Of 14 new unpublished mutations in our data-set, 13 were loss-of-function and one a missense mutation.

Families with an EXTl mutation had significantly worse disease than those with an EXT2 mutation across several parameters of severity (Fig. 3); 19 of 34 families (56%) with EXTl consisted of individuals whose mean height was below the 25th centile, compared with six of 27 individuals (22%) in families with EXT2 (Χ^sup 2^ = 7.050, p = 0.008). The median family height in families with the EXTl mutation was on the 18th centile, and that for families with the EXT2 mutation was on the 50th centile. Individual deformity scores ranged from zero to 17 from a total of 18. Significantly greater deformity scores were recorded in families with the EXTl mutation than those with EXT2 (median score 5.0 and 1.8, respectively; z = 3.532, p = 0.0004). Individual functional scores ranged from zero to 16 from a total of 18. Significantly worse functional scores were recorded in families with the EXTl mutation than those with EXT2 (median score 6.0 and 3.1, respectively; z = 2.512, p = 0.012).

The number of exostoses per individual varied between zero and 172. The number of exostoses were similar in families with EXTl and EXT2 mutations (median 27 and 29, mean 37 and 29, respectively; z = 1.430, p = 0.153).

A total of ten chondrosarcomas occurred in nine families; seven in subjects within the study and three in relatives who had already died (Table H). All sarcoma patients inherited loss-of-function mutations in their defective EXT gene; seven sarcomas occured in EXTl mutation carriers, and only one in an EXT2 mutation carrier. No statistical difference was identified when the number of exostoses was compared in families with or without a family history of sarcoma (median of exostoses 37 and 28, respectively; z = 0.962, p = 0.336).

Discussion

Both epidemiological and clinical features of this condition allow its classification as a familial (mostly benign) neoplastic trait.2'13 This assertion is strengthened by molecular genetic evidence which strongly suggests a clonal aetiology for the osteochondroma and that a 'two-hit' model for HME pathogenesis is appropriate. Mutations of the germ line in the EXT tumour suppressor gene family have been identified in HME.6-10,12-17-22 The risk of sarcoma in HME is estimated, in population-based series, to be smaller than initially thought. Recent estimates would suggest a lifetime risk of about 2% to 4%.3,23-25 In published series, the EXT1 mutation has been identified in about 45% of HME patients and about 40% in the EXT2. The remaining 15% have no identified mutation of the germ line.6-11 In our group, the ratio of families with EXT1 and EXT2 mutations is similar to those in other published series, both self-selected and hospital-based.8,10,11 The anatomical distribution is also similar to published radiological skeletal surveys, although the numbers identified close to deep-seated joints (proximal femur, pelvis) are lower.1,5,25,26

Our results confirm an age-related phenotypic penetrance. This is widely acknowledged in HME and consistent with published data, which report an age-dependent disease penetrance and a median age of first diagnosis of HME5 of three years.1 A trend towards worse disease was observed in men and boys. These results are consistent with other studies and suggest only a limited role for the disease-modifying effects of gender.1,5

Although loss-of-function mutations result in no useful gene-product, missense mutations (substitution of one amino acid for another) might allow partial gene-function if the mutation did not occur at a critical site. Our data reveal that the majority of EXT mutations are associated with loss-of-function, which would be expected in a tumour-suppressor gene, requiring knock-out for disease-expression. Furthermore, most missense mutations (10 of 11) have been previously described,27 suggesting that they occur in crucial sites causing an effective 'knock-out' of the gene. Consequently, similar disease severities might be expected.

The findings of an earlier smaller genotype-phenotype study in HME concur with our results.28 Our group of patients is considerably larger, enabling the variables of age and gender to be examined. EXT1 mutation is found more frequently than EXT2 mutation in solitary exostoses, but it is not yet known why this gene has a more dominant role in severity of disease and sarcoma-risk in HME.22 It is likely that a large non-EXT gene component to severity of disease exists since family members frequently exhibit a wide range of HME scores (Fig. 4).

In regard to the association between the EXT1 mutation and chondrosarcoma, three questions arise; is it worth screening for pre-malignant disease, and if so, how and in whom? Given a lifetime risk of approximately 3%, most of which occurs in the third to fifth decades of life, it would seem reasonable to assess the annual risk of sarcoma in this age group to be 0.1%.3,22-24 Possibly this risk might be doubled in EXT1 mutation carriers. By comparison, the incidence of breast cancer is 0.2% per annum between the ages of 50 and 60 years;29 an age group in which mammographic screening is recommended in many countries. By this comparison, a screening programme in HME (in particular EXT1 mutation carriers) might seem to be appropriate.

Current practice is in favour of clinical examination once every year or two. There are no data on the number of malignancies that develop between screening visits, nor on the efficacy of such a programme. In our series only 13.5% of palpable exostoses were found in the axial skeleton (Fig. 1), while in a large recent report of chondrosarcomas in HME, 80% of tumours occurred in the axial skeleton.30 Worries, therefore, exist as to whether deep-seated axial exostoses or chondrosarcomas can be identified by clinical examination alone. Radiological surveys deliver a significant radiation dose, and have theoretical risks in a condition now identified as a familial neoplastic trait. Magnetic resonance imaging of the axial skeleton is a promising line of investigation in screening and positron-emission tomography has also been used to study exostosis activity. Both imaging modalities may have a role in screening, however, no published data on their efficacy exist.

Given that families with a history of sarcoma do not have a significantly greater median HME score than those with no such history, it seems that severity of disease alone cannot be utilised as a marker of the risk of malignant change. EXT gene testing may have a role, but at least one sarcoma has been reported in an EXT2 mutation carrier,21 in addition to the patient reported here. Thus, all known patients with HME may benefit from inclusion in a screening programme.

The Arthritis Research Campaign generously supported this work with a project grant (P0540).

No benefits in any form have been received or will be received from a commercial party related directly or indirectly to the subject of this article.

References

1. Schmale GA, Conrad EU 3rd, Raskind R. The natural history of hereditary multiple exostoses. J Bone Joint Surg [Am] 1994:76-A:986-92.

2. Porter DE, Emerton ME, Villaneuva-Lopez F, Simpson AHRW. A clinical and radiographie analysis of osteochondromas and growth disturbance in hereditary multiple exostoses. J Pediatr Orthop 2000:20:246-50.

3. Jaffe HL Hereditary multiple exostoses. Arch Pathol Lab Med 1943:36:335-57.

4. Solomon L. Diametric growth of the epiphyseal plate. J Bone Joint Surg [Br] 1966:48-B:170-7.

5. Solomon L. Hereditary multiple exostosis. J Bone Joint Surg [Br] 1963:45-B:292-304.

6. Cook W, Raskind W, Blanton SH, et al. Genetic heterogeneity in families with hereditary multiple exostoses. Am J Hum Genet 1993;53:71-9.

7. Ahn J, Ludecke H-J, Lindow S, et al. Cloning of the putative tumour suppressor gene for hereditary multiple exostoses (EXT1). Nat Genet 1995;11:137-43.

8. Stickens D. Clines C, Burbee D, et al. The EXT2 multiple exostoses gene defines a family of putative tumour suppressor genes. Nat Genet 1996;14:25-33.

9. Wuyts W, Van Hul W, Wauters J, et al. Positional cloning of a gene involved on hereditary multiple exostoses. Hum Mol Genet 1996;5:1547-57.

10. Phillippe C, Porter DE, Emerton ME, et al. Mutation screening of EXT1 and EXT2 genes m patients with hereditary multiple exostoses. Am J Hum Genet 1997;61: 520-8.

11. Dobson-Stone C, Cox R, Lonie L, et al. Comparison of fluorescent single-strand conformation polymorphism analysis and denaturing high-performance liquid chromatography for detection of EXT1 and EXT2 mutations in hereditary multiple exostoses. Eur J Hum Genet 2000;8:24-32.

12. Knudson Jr AG. Mutation and cancer: statistical study of retinoblastoma. Proc Natl Acad Sci USA 1971;68:661-5.

13. Porter DE, Simpson AHRW. The neoplastic pathogenesis of solitary and multiple osteochondromas J Palhol 1999;188:119-25.

14. Chou RH, Wong GB, Kramer JH, et al. Toxicities of total-body irradiation for pediatric bone-marrow transplantation. Int J Radiat Oncol Biol Phys 1996;34:843-51.

15. Langenskiold A, Edgren W. Imitation of chondrodysplasia by localised roentgen ray injury - an experimental study of bone growth. Acta Chir Scand 1950;99:353.

16. Child Growth Foundation. Tanner-Whitehouse revised growth and development record, British longitudinal standards, 1995. London: Child Growth Foundation, 1996.

17. Mertens F, Rydholm A, Kriecbergs A, et al. Loss of chromosone band 8q24 in sporadic osteocartilaginous exostoses. Genes Chromosomes Cancer 1994;9:8-12.

18. Wu Y-O, Heutink P, de Vries BBA, et al. Assignment of a second locus for multiple exostoses to the pericentric region of chromosome 11. Hum Mol Genet 1994;3:167-71.

19. Hecht JT, Hogue D, Strong LC, et al. Hereditary multiple exostoses and chrondrosarcoma: linkage to chromosome II and loss of heterozygosity for EXT-linked markers on chromosomes II and 8. Am J Hum Genet 1995;56:1125-31.

20. Raskind WH, Conrad EU, Chansky H, Matsushita M. Loss of heterozygosity in chondrosarcoma for markers linked to hereditary multiple exostoses loci on chromosomes 8 and 11. Am J Hum Genet 1995;56:1132-9.

21. Hecht JT, Hogue D, Wang Y, et al. Hereditary multiple exostoses (EXT): mutational studies of familial EXT1 cases and EXT-associated malignancies. Am J Hum Genet 1997;60:80-6.

22. Bovee J, Cleton-Jansen A, Wuyts W, et al. EXT mutation analysis and loss of hererozygosity insporadic and hereditary osteochondromas and secondary chondrosarcomas. Am J Hum Genet 1999;65:689-98.

23. Solomon L. Chondrosarcoma in hereditary multiple exostosis. A Afr Med J 1974;48: 671-6.

24. Wicklund CL, Pauli RM, Johnston D, Hecht JT. Natural history study of hereditary multiple exostoses Am J MedGenet 1995;55:43-6.

25. Voutsinas S, Wynne-Davies R. The infrequency of malignant disease in diaphyseal aclasis and neurofibromatosis. J Med Genet 1983;20:345-9.

26. Shapiro F, Simon S, Glimcher MJ. Hereditary multiple exostoses: anthropometrie, roentgenographic and clinical aspects. J Bone Joint Surg [Am] 1979:61-A:815-24.

27. Wuyts W, Van Hul W. Molecular basis of multiple exostoses: mutations in the EX1 and EX2 genes. Hum Mutat 2000;15:220-7.

28. Francannet C, Cohen-Tanugi A, LeMerrer M, et al. Genotype-phenotype correlation in hereditary multiple exostoses. J Med Genet 2001;38:430-4.

29. Iselius L, Slack J, Littler M, Morion NE. Genetic epidemiology of breast cancer in Britain. Ann Hum Genet 1991:55:151-9.

30. Ahmed AR, Tan T-S, Unni KK, et al. Secondary chondrosarcoma in osteochondroma: report of 107 cases. Clin Orthop 2003;11:193-206.

D. E. Porter, L. Lonie, M. Fraser, C. Dobson-Stone, J. R. Porter, A. P. Monaco, A. H. R. W. Simpson

From the University of Edinburgh, Edinburgh, Scotland and University of Oxford, Oxford, England

* D. E. Porter, FRCS, Tr and Orth, Senior Lecturer

* A. H. R. W. Simpson, FRCS, Professor

Edinburgh University Department of Orthopaedic Surgery, New Royal Infirmary of Edinburgh, Little France, Edinburgh EH16 4SU, UK.

* L. Lonie, MSc, Research Assistant

* C. Dobson-Stone, MBiochem, Research Assistant

* A. P. Monaco, PhD, Director

The Henry Wellcome Building for Genomic Medicine, Roosevelt Drive, Oxford OX3 7BN, UK.

* M. Fraser, BA, Research Nurse

* J. R. Porter, MB, BS, Research Assistant

Nuffield Department of Orthopaedic Surgery, University of Oxford, Nuffield Orthopaedic Centre, Windmill Road, Headington, Oxford 0X3 7LD, UK.

Correspondence should be sent to Mr D. E. Porter.

©2004 British Editorial Society of Bone and Joint Surgery

doi: 10.1302/0301-620X.86B7. 14815 $2.00

J Bone Joint Surg IBr] 2004;86-B:1041-6.

Received 25 July 2003; Accepted 1 October 2003

Copyright British Editorial Society of Bone & Joint Surgery Sep 2004
Provided by ProQuest Information and Learning Company. All rights Reserved