The Genetics of Paget’s Disease of the Bone

Robin J. Leach, Frederick R. Singer and G. David Roodman

Departments of Cellular and Structural Biology (R.J.L.), Pediatrics (R.J.L.), and Medicine (G.D.R.), The University of Texas Health Science Center at San Antonio, San Antonio, Texas 78229-3900; Department of Medicine (G.D.R.), The Veterans Administration Hospital, San Antonio, Texas 78284; and John Wayne Cancer Institute at St. John’s Hospital and Health Center (F.R.S.), Santa Monica, California 90404

Address correspondence and requests for reprints to: G. D. Roodman, M.D., Ph.D., Department of Medicine/Hematology (7880), University of Texas Health Science Center at San Antonio, 7703 Floyd Curl Drive, San Antonio, Texas 78229-3900. E-mail: roodman{at}uthscsa.edu


    Introduction
 Top
 Introduction
 Histocompatibility leukocyte...
 Family studies
 HLA linkage studies
 Chromosome 18 linkage studies
 Role of receptor activator...
 Role of RANK in...
 PDB and osteosarcoma
 PDB and paramyxovirus
 Conclusion
 References
 
In 1877, Sir James Paget first described the localized skeletal disease that is now known as Paget’s disease of the bone (PDB) (1). The disease is primarily a disorder of the osteoclast with marked increase in bone resorption, followed by abundant new bone formation. In this report, we present a brief overview of the genetics of PDB; a complete review was recently published by Singer and Leach (2).

PDB has a familial tendency (3, 4, 5), suggesting that there is a genetic predisposition. In a study of the frequency and characteristics of the familial aggregation of PDB in Spain, Morales-Piga et al. (6) found that 40% of their index cases had at least one first-degree relative affected with PDB. In the pedigrees they reported, PDB seemed to be transmitted through either parent, suggesting an autosomal dominant mode of inheritance.

Siris (7) conducted an epidemiological study of PDB in the United States, using questionnaires completed by 864 patients with physician-diagnosed PDB, and compared these results to 500 controls of similar age. A history of PDB was noted in a first-degree relative in 12% of the patients, compared with only 2% of controls. The risk of first-degree relatives of a pagetic patient developing PDB was seven times greater than for an individual without an affected relative. The cumulative risk for developing PDB up to age 90 for a first-degree relative of a patient was 9% compared with 2% in individuals with unaffected relatives. This is good evidence that a gene, or genes, plays a role in the acquisition of the disease.


    Histocompatibility leukocyte antigen (HLA) association studies
 Top
 Introduction
 Histocompatibility leukocyte...
 Family studies
 HLA linkage studies
 Chromosome 18 linkage studies
 Role of receptor activator...
 Role of RANK in...
 PDB and osteosarcoma
 PDB and paramyxovirus
 Conclusion
 References
 
The role of the major HLA loci on chromosome 6 in many human diseases makes it a logical starting point for studying the genetics of PDB. Several studies since 1975 have examined the possibility that there is an association between HLA and PDB (8, 9, 10, 11, 12). Although HLA typing of class I antigens revealed no significant associations, significant associations were observed between class II antigens and PDB in two studies (11, 12). An increased incidence of HLA-DQW1 and HLA-DR2 was found in a preliminary report of 53 patients in Los Angeles (11). A second study of 25 Ashkenazi Jews in Israel revealed an increased incidence of HLA-DR2 in this population (12).


    Family studies
 Top
 Introduction
 Histocompatibility leukocyte...
 Family studies
 HLA linkage studies
 Chromosome 18 linkage studies
 Role of receptor activator...
 Role of RANK in...
 PDB and osteosarcoma
 PDB and paramyxovirus
 Conclusion
 References
 
Although multiple families have been reported with PDB, the average number of affected individuals in these families is three. Thus, these families are not sufficiently large for performing classical linkage analysis. However, there have been a few larger kindreds that have been used for linkage analysis. As reviewed below, analysis has identified two loci, one on chromosome 6 and one on chromosome 18, as susceptibility loci for PDB. Both loci seem to play a role in PDB, and there is evidence from other studies that there are other loci yet to be identified.


    HLA linkage studies
 Top
 Introduction
 Histocompatibility leukocyte...
 Family studies
 HLA linkage studies
 Chromosome 18 linkage studies
 Role of receptor activator...
 Role of RANK in...
 PDB and osteosarcoma
 PDB and paramyxovirus
 Conclusion
 References
 
The first linkage studies used the HLA loci because of their highly polymorphic nature. In a study conducted in New York City (13), three families with 29 informative children (all over the age of 45) were used. These kindreds were typed at the HLA-A, -B, and -C loci. Using haplotype data, the investigators obtained a maximum LOD (the logarithm of the odds ratio of a particular locus being linked or not linked to the disease locus) score of 2.44 with 11% recombination. Because a LOD score of 3 is considered linkage, these data were only deemed "suggestive." In a second study performed in New Zealand (14), two additional families were identified that seemed to segregate the disease in an autosomal dominant fashion. Each family had four affected members. These families were genotyped for the HLA loci. Although significant linkage was not obtained with this study, the combined linkage analysis from both studies resulted in a maximum LOD score of 3.69 with 10% recombination. The combined data were considered sufficient for establishing a predisposition locus for PDB on chromosome 6 near the HLA loci. Two other family studies have been published that explored the linkage of the HLA loci with PDB (15, 16). Both studies reported no linkage between the HLA loci and a PDB predisposition gene. However, neither family was large enough to exclude linkage, and the relevance of these observations is unclear.


    Chromosome 18 linkage studies
 Top
 Introduction
 Histocompatibility leukocyte...
 Family studies
 HLA linkage studies
 Chromosome 18 linkage studies
 Role of receptor activator...
 Role of RANK in...
 PDB and osteosarcoma
 PDB and paramyxovirus
 Conclusion
 References
 
Familial expansile osteolysis (FEO) is a rare bone dysplasia, which is transmitted as an autosomal dominant trait (17) in a large kindred in Northern Ireland and a kindred in the United States. FEO shares some features with PDB. The bone lesions appear similar to early pagetic osteolytic lesions, although they occur at a much earlier age; also, these lesions never become sclerotic. Interestingly, osteoclasts from these patients contain paramyxoviral-like nuclear inclusions, which have also been reported in PDB (18). These results supported the hypothesis that FEO may be an allelic variant of PDB.

Hughes et al. (19) have used genetic linkage analysis to localize FEO to chromosome 18q. The disease shows a tight linkage with several polymorphic markers on chromosome 18q, with a maximum LOD score of 11.53 and no recombination. To further test the hypothesis that FEO is an allelic variant of PDB, linkage analysis has been performed between chromosome 18 markers and PDB disease kindreds.

The first published linkage study with chromosome 18 markers was performed by Cody et al. (20), using one large pedigree. The pedigree is presented in Fig. 1Go. In this study, 16 DNA samples were evaluated with a total of 12 markers from 18q. The maximum LOD score of 3.44 was obtained with genetic marker D18S42 with no recombination. This marker lies in the same region of chromosome 18q that is linked to FEO. Thus, this study supports the hypothesis that FEO and PDB result from mutations in the same locus or tightly linked loci.



View larger version (27K):
[in this window]
[in a new window]
 
Figure 1. Pedigree of kindred with PDB that is linked to the loci on chromosome 18q. Black symbols represent affected individuals. An asterisk indicates those individuals for whom the phenotype is unknown. The disease haplotype is indicated by a rectangle. [Adapted from Cody et al. (20 ).]

 
The following year, Haslam et al. (21) reported a study of eight families with familial PDB. Using seven polymorphic markers, they obtained a maximum LOD score of 2.97 with marker D18S42 with 5% recombination. Statistically, it seemed that the families were genetically heterogeneous. Only five of the eight families had positive LOD scores in this region, whereas the remainder seemed to be linked to another locus. There is also a preliminary report of a French pedigree that is clearly linked to 18q (22). This pedigree was analyzed with 12 chromosome 18 markers and gave a maximum LOD score of 3.10 with marker D18S68 at 1% recombination.

There have been several recent reports of families with PDB that do not seem to be linked to chromosome 18 (21, 23, 24). The most recent family was sufficiently large to exclude linkage across the entire interval (23). (A LOD score of -2 is needed to exclude linkage.) Unfortunately, the majority of the nonchromosome 18-linked families have not been evaluated with markers from chromosome 6. Until positive LOD scores are obtained with these "unlinked" families, it is difficult to estimate the number of predisposition loci for PDB.


    Role of receptor activator of nuclear factor {kappa}B (RANK) in FEO
 Top
 Introduction
 Histocompatibility leukocyte...
 Family studies
 HLA linkage studies
 Chromosome 18 linkage studies
 Role of receptor activator...
 Role of RANK in...
 PDB and osteosarcoma
 PDB and paramyxovirus
 Conclusion
 References
 
RANK ligand (RANKL) is a newly described member of the tumor necrosis factor (TNF) family that has been identified as a critical osteoclastogenic factor (25, 26, 27). RANKL is expressed on marrow stromal cells and osteoblasts and seems to mediate the effects of most osteoclastogenic factors. In vivo studies have demonstrated that mice lacking RANKL develop severe osteopetrosis (28). Factors such as 1,25-(OH)2D3, interleukin 1, interleukin 11, and prostaglandin E2 seem to induce osteoclast formation indirectly by up-regulating RANKL expression on marrow stromal cells (29). These data suggest that RANKL may be the common mediator for the effects of most osteotropic factors on osteoclast formation.

The receptor for RANKL, RANK, is a member of the TNF receptor family and interacts with TNF receptor-associated factor 2 and translocates it to the nucleus to induce nuclear factor {kappa}B signaling. RANK is expressed on osteoclast precursors and osteoclasts, and overexpression of RANK can induce nuclear factor kB signaling in the absence of RANKL (30). Furthermore, deletion of RANK by homologous recombination results in osteopetrosis (31). These data confirm the critical role of RANK/RANKL in osteoclastogenesis.

Earlier this year, Hughes et al. (32) identified the FEO gene on chromosome 18. Using mutational analysis, they identified gene mutations in three different FEO families. Interestingly, all three families carried the identical mutation. The mutation was an 18-bp in-frame insertion in the signal peptide sequence of the RANK gene (also known as TNFRSF11A). The 18-bp insertion was the result of a tandem duplication of bases 84 through 101 in exon 1. The mutation resulted in stabilization of the RANK protein, which in turn increased RANK signaling (32). As noted above, this increased RANK signaling could result in increased osteoclast formation, although this has not been proven.


    Role of RANK in PDB
 Top
 Introduction
 Histocompatibility leukocyte...
 Family studies
 HLA linkage studies
 Chromosome 18 linkage studies
 Role of receptor activator...
 Role of RANK in...
 PDB and osteosarcoma
 PDB and paramyxovirus
 Conclusion
 References
 
With the identification of the mutated gene in FEO on chromosome 18, it is now possible to test the hypothesis that FEO and PDB are the result of mutations in the same gene. Hughes et al. (32) screened 90 sporadic PDB patients and found none with this 18-bp insertion mutation. They also performed mutation screening in members of four PDB families who had evidence of linkage to 18q. All four of these families had been used in the linkage study by Haslam et al. (21). In one of these families, a slightly larger duplication involving bases 75 through 101 was observed in exon 1 (32). The PDB disease family with the mutation in RANK was of Japanese ancestry. This mutation segregated with the disease in this family. No other mutations were observed in the other three PDB families.

In the PDB family with the RANK mutation, most of the affected individuals presented in their teens and early twenties with bone pain or deformity. In addition, all the patients were described as having dental problems and several had hearing impairment. This clinical description begs the question whether this family has PDB vs. a mild version of FEO. It is uncommon in PDB to have tooth loss, but this seems to be almost universal for FEO (33). In addition, hearing loss in PDB usually occurs later in life and is associated with the thickening of the skull, whereas hearing loss associated with FEO occurs early (33). Unfortunately, clinical information was not provided in the publication by Hughes et al. (32); thus, it is difficult to conclude that RANK mutations are associated with familial PDB, although it does not seem to play a major role in sporadic PDB.

To date, there have been two large families that demonstrated significant linkage to chromosome 18 (20, 22). It will be important to thoroughly evaluate the RANK gene in these families before making a conclusion concerning the role of RANK in familial PDB. The fact that the other families first described by Haslam et al. (21) did not show mutations in the RANK gene may be explained two different ways. First, the families may actually be linked to another locus. Clearly, there is more than one PDB predisposition gene, and these families are relatively small. Thus, linkage using these families can only identify a large region that carries the predisposition gene. It is possible that these families are truly linked to chromosome 18 and that the responsible gene is near the RANK locus, or that they are linked to another chromosome. Alternatively, the "mutation" in the RANK gene could be a promoter "mutation" that slightly increases the gene expression. Such a "mutation" may be only a polymorphism, and it may be difficult to demonstrate functional significance. Based on the data for the insertion mutation and the role of RANK in osteoclast differentiation, one would expect that any alterations in the RANK gene causing PDB would demonstrate increased RANK signaling.


    PDB and osteosarcoma
 Top
 Introduction
 Histocompatibility leukocyte...
 Family studies
 HLA linkage studies
 Chromosome 18 linkage studies
 Role of receptor activator...
 Role of RANK in...
 PDB and osteosarcoma
 PDB and paramyxovirus
 Conclusion
 References
 
In the majority of patients, PDB is asymptomatic (34). The 5–10% of patients with symptoms have bone pain with a wide range of complications, including increased fractures, deafness, and neurological findings (35). The most devastating complication of PDB is malignant transformation of the bone. Although these transformations are rare, occurring in less than 1% of patients, they contribute significantly to the morbidity and mortality of the disease. There are various forms of malignant transformation associated with PDB, but the most frequent is osteosarcoma. Osteosarcoma, like most cancers, is believed to result from a series of genetic alterations that transform the osteoblast to a malignant state. There is interesting evidence to imply that there is a molecular connection between osteosarcoma and PDB (36).

Loss of heterozygosity (LOH) mapping is a method used to identify tumor suppressor loci in cancer (37). Yamaguchi et al. (38) have used LOH mapping to examine osteosarcomas. They found high-frequency LOH on 3q, 13q, 17p, and 18q, suggesting that these chromosome arms contain tumor suppressor genes important in the development of osteosarcoma. Nellisery et al. (36) have further refined the location of this tumor suppressor gene between genetic markers D18S60 and D18S42 (see Fig. 2Go). This is the same region that codes the FEO and the familial PDB genes. This has led us to the hypothesis that there is a link between the predisposition gene for PDB and the tumor suppressor gene for osteosarcoma on chromosome 18q. Interestingly, seven of seven osteosarcomas from PDB patients showed LOH in this region (36). However, it is unclear, at present, whether the PDB predisposition gene and the osteosarcoma tumor suppressor gene are one and the same gene or two tightly linked genes on chromosome 18q.



View larger version (17K):
[in this window]
[in a new window]
 
Figure 2. Map of the minimal region of LOH on chromosome 18q in osteosarcoma tumors. The polymorphic loci and their estimated physical distances are shown across the top. Note that D18S51, D18S55, and D18S68 were not ordered in the contig. Below the map is the information gained from the LOH studies of each tumor. •, No LOH; {circ}, LOH. Uninformative markers have a dash. The arrows depict the region and the direction of LOH for each tumor. The minimal region of LOH in the sporadic osteosarcomas is denoted by the large shaded rectangle. [Adapted from Nellissery et al. (36 ).]

 
Molecular analysis by Hughes et al. (32) localized the RANK gene between markers D18S64 and D18S51, which was the same interval known to contain the FEO locus. Further mapping in our laboratory has demonstrated that the RANK gene lies between markers D18S60 and D18S42 (R. J. Leach, unpublished data), which is the critical region for the putative osteosarcoma tumor suppressor gene (see Fig. 2Go). These data suggest that the role of RANK in osteosarcoma needs to be thoroughly analyzed.


    PDB and paramyxovirus
 Top
 Introduction
 Histocompatibility leukocyte...
 Family studies
 HLA linkage studies
 Chromosome 18 linkage studies
 Role of receptor activator...
 Role of RANK in...
 PDB and osteosarcoma
 PDB and paramyxovirus
 Conclusion
 References
 
In addition to a genetic predisposition for PDB, many studies have suggested a potential viral etiology for PDB as well (39), because pagetic osteoclasts contain paramyxoviral-like nuclear inclusions. Although the presence of a virus in pagetic osteoclasts is still controversial, the majority of studies have supported either measles virus or canine distemper virus as the paramyxovirus present in pagetic osteoclasts and their precursors. It is possible that the combination of a genetic predisposition to PDB and chronic infection of osteoclast precursors by paramyxovirus may be required to express the pagetic phenotype in cells of the osteoclast lineage. However, until a virus is isolated or a paramyxoviral gene can be shown to induce PDB in an animal model, the role of the virus in the pathogenesis of PDB is still unclear.


    Conclusion
 Top
 Introduction
 Histocompatibility leukocyte...
 Family studies
 HLA linkage studies
 Chromosome 18 linkage studies
 Role of receptor activator...
 Role of RANK in...
 PDB and osteosarcoma
 PDB and paramyxovirus
 Conclusion
 References
 
PDB disease clearly has a hereditary component. Family studies demonstrate that there is more than one predisposition gene for PDB. The mapping of osteosarcoma tumor suppressor and a PDB predisposition locus on 18q makes this region of the genome of particular interest, and a region of future study. Until more extensive linkage analyses are performed on larger kindreds with classic PDB, the other susceptibility loci for PDB will remain elusive. The identification of PDB predisposition genes will increase our understanding of the pathogenesis of this disorder and could lead to alternate treatment strategies. In addition, the identification of these genes will be useful for understanding bone biology and the molecular basis of the alterations that give rise to PDB.

Received June 19, 2000.

Revised September 13, 2000.


    References
 Top
 Introduction
 Histocompatibility leukocyte...
 Family studies
 HLA linkage studies
 Chromosome 18 linkage studies
 Role of receptor activator...
 Role of RANK in...
 PDB and osteosarcoma
 PDB and paramyxovirus
 Conclusion
 References
 

  1. Paget J. 1877 On a form of chronic inflammation of bones (osteitis deformans). Med Chir Trans. 60:37–64.
  2. Singer FR, Leach RJ. 2000 Genetics of Paget’s disease of bone. In: Econs MJ, ed. The genetics of osteoporosis and metabolic bone disease. Totowa, NJ: Humana Press; 309–318.
  3. Jacobs TP, Michelsen J, Polay JS, D’Adamo AC, Canfield RE. 1979 Giant cell tumors in Paget disease of bone: familial and geographic clustering. Cancer. 44:742–747.[Medline]
  4. Sofaer JA, Holloway SM, Emery A. 1983 A family study of Paget disease of bone. J Epidemiol Community Health. 37:226–231.[Abstract]
  5. Mirra JM. 1986 CPC case: Paget disease and giant cell tumor. N Engl J Med. 314:105–113.[Medline]
  6. Morales-Piga AA, Rey-Rey JS, Corres-Gonzales J, Garcia-Sagredo JM, Lopez-Abente G. 1995 Frequency and characteristics of familial aggregation in Paget disease of bone. J Bone Miner Res. 10:663–670.[Medline]
  7. Siris ES. 1994 Epidemiological aspects of Paget disease: family history and relationship to other medical conditions. Semin Arthritis Rheum. 23:222–225.[Medline]
  8. Roux H, Mercier P, Maestracci D, Eisinger J, Recordier AM. 1975 HL-A et maladie de Paget. Rev Rhum Mal Osteoartic. 42:661–662.[Medline]
  9. Cullen P, Russel GG, Walton RJ, Whiteley J. 1976 Frequencies of HLA-A and HLA-B histocompatibility antigens in Paget’s disease of bone. Tissue Antigens. 7:55–56.[Medline]
  10. Gordon MT, Cartwright EJ, Mercer S, Anderson DC, Sharpe PT. 1994 HLA polymorphisms in Paget’s disease of bone. Semin Arthritis Rheum. 23:229.[Medline]
  11. Singer RF, Mills BG, Park MS, Takemura S, Terasaki PI. 1985 Increased HLA-DQWI antigen pattern in Paget’s disease of bone. Clin Res. 33:574A.
  12. Foldes J, Shamir S, Brautbor C, Schermann L, Menczel J. 1991 HLA-D antigens and Paget’s disease of bone. Clin Orthop. 266:301–303.[Medline]
  13. Fotino M, Haymovits A, Falk CT. 1977 Evidence for linkage between HLA and Paget’s disease. Transplant Proc. 9:1867–1868.[Medline]
  14. Tilyard MW, Gardner RJM, Milligan L, Cleary TA, Stewart RDH. 1982 A probable linkage between familial Paget’s disease and HLA loci. Aust NZ J Med. 12:498–500.[Medline]
  15. Moore SB, Hoffman DL. 1988 Absence of HLA linkage in a family with osteitis deformans (Paget’s disease of bone). Tissue Antigens. 31:69–70.[Medline]
  16. Kim GS, Kim SH, Cho JK, et al. 1997 Paget bone disease involving young adults in 3 generations of a Korean family. Medicine. 76:157–169.[CrossRef][Medline]
  17. Wallace RG, Barr RJ, Osterberg PH, Mollan RA. 1989 Familial expansile osteolysis. Clin Orthop. 248:265–277.[Medline]
  18. Dickson GR, Shirodira PV, Kanis JA, Beneton MN, Carr KE, Mollan RAB. 1991 Familial expansile osteolysis: a morphological, histomorphometric and serological study. Bone. 12:331–338.[Medline]
  19. Hughes AE, Shearman AM, Weber JL, et al. 1994 Genetic linkage of familial expansile osteolysis to chromosome 18q. Hum Mol Genet. 3:359–361.[Abstract]
  20. Cody JD, Singer FR, Roodman GD, et al. 1997 Genetic linkage of Paget disease of the bone to chromosome 18q. Am J Hum Genet. 61:1117–1112.[CrossRef][Medline]
  21. Haslam SI, Van Hul W, Morales-Piga A, et al. 1998 Paget’s disease of bone: evidence for a susceptibility locus on chromosome 18q and for genetic heterogeneity. J Bone Miner Res. 13:911–917.[Medline]
  22. Lucotte G. 1999 Genetic linkage of Paget’s disease to chromosome 18q in a French pedigree (Abstract). Bone. 24:31S.
  23. Nance MA, Nuttall FQ, Econs MJ, et al. 2000 Heterogeneity of Paget disease of the bone. Am J Med Genet. 92:303–307.[CrossRef][Medline]
  24. Hocking L, Slee F, Haslam SI, et al. 2000 Familial Paget’s disease of bone: Pattern of inheritance and frequency of linkage to chromosome 18q. Bone. 26:577–580.[CrossRef][Medline]
  25. Lacey DL, Timms E, Tan HL, et al. 1998 Osteoprotegerin ligand is a cytokine that regulates osteoclast differentiation and activation. Cell. 93:165–176.[Medline]
  26. Yasuda H, Shima N, Nakagawa N, et al. 1998 Osteoclast differentiation factor is a ligand for osteoprotegerin/osteoclastogenesis-inhibitory factor and is identical to TRANCE/RANKL. Proc Natl Acad Sci USA. 95:3597–3602.[Abstract/Free Full Text]
  27. Hofbauer LC, Heufelder AE. 1998 Osteoprotegerin and its cognate ligand: a new paradigm of osteoclastogenesis. Eur J Endocrinol. 139:152–154.[Medline]
  28. Kong YY, Yoshida H, Sarosi I, et al. 1999 OPGL is a key regulator of osteoclastogenesis, lymphocyte development and lymph-node organogenesis. Nature. 397:315–323.[CrossRef][Medline]
  29. Yasuda H, Shima N, Nakagawa N, et al. 1998 Identity of osteoclastogenesis inhibitory factor (OCIF) and osteoprotegerin (OPG): a mechanism by which OPG/OCIF inhibits osteoclastogenesis in vitro. Endocrinology. 139:1329–1337.[Abstract/Free Full Text]
  30. Anderson DM, Maraskovsky E, Billingsley WL, et al. 1997 A homologue of the TNF receptor and its ligand enhance T-cell growth and dendritic-cell function. Nature. 390:175–179.[CrossRef][Medline]
  31. Dougall WC, Glaccum M, Charrier K, et al. 1999 RANK is essential for osteoclast and lymph node development. Genes Dev. 13:2412–2424.[Abstract/Free Full Text]
  32. Hughes AE, Ralston ST, Marken J, et al. 2000 Mutations in TNFRSF11A affecting the signal peptide of RANK, cause familial expansile osteolysis. Nat Genet. 24:45–49.[CrossRef][Medline]
  33. Hughes AE, Barr RJ. 1996 Familial expansile osteolysis: a genetic model of Paget’s disease. In: Sharpe PT, ed. The molecular biology of Paget’s disease. Austin, TX: R.G. Landes Company; 179–200.
  34. Dalinka MK, Aronchick JM, Haddack Jr JG. 1983 Paget’s disease. Orthop Clin North Am. 14:3–19.[Medline]
  35. Cawley MI. 1983 Complication of Paget’s disease of bone. Gerontology. 29:276–287.[Medline]
  36. Nellissery MJ, Padalecki SS, Brkanac Z, et al. 1998 Evidence for a novel osteosarcoma tumor-suppressor gene in the chromosome 18 region genetically linked with Paget disease of bone. Am J Hum Genet. 63:817–824.[CrossRef][Medline]
  37. Hansen MF, Cavenee WK. 1988 Tumor suppressors: recessive mutations that lead to cancer. Cell. 53:173–174.[Medline]
  38. Yamaguchi T, Toguchida J, Yamamuro T, et al. 1992 Allelotype analysis in osteosarcomas: frequent allele loss on 3q, 13q, 17p, and 18q. Cancer Res. 52:2419–2423.[Abstract]
  39. Singer FR, Roodman GD. 1996 Paget’s disease. In: Bilezikian J, Raisz L, Rodan G, eds. Principles of bone biology. San Diego: Academic Press; 969–977.