Epidemiology Research Unit
Centre Hospitalier de I'Universite de Montreal- Hotel-Dieu, and
Department of Nutrition, Faculty of Medicine, University of Montreal
American Academy of Neurological and Orthopaedic Surgeons

Introduction

Characteristics of bone cancer occurrence:
Bone cancer is not a common cancer compared to many other types of cancer. The principal malignant tumors of bone are: a) osteosarcomas that occur mostly in the leg bones of children and young adults; this form is more frequent among girls under 15 and boys over 15; its incidence is higher among nonwhites than whites(1); b) chondro-sarcomas that usually afflict people over 40 years of age; this is a slow-growing tumor that often starts in the pelvic bones; and c) Ewing's sarcoma, a cancer that impacts mainly children and teenagers; this form infiltrates large bones such as those of the thigh, upper arm, shin or pelvis; two times as many males are affected as females; a fast-growing tumor, its incidence is almost 9-fold higher among whites than blacks.

According to the U.S. Surveillance, Epidemiology and End Results Program(2), osteosarcomas contribute 36% of all types of bone cancer, followed by chondrosarcomas and Ewing's sarcomas with around 30% and 16% respectively. The incidence of osteosarcoma appears to be more frequent in two periods of life, during adolescence and old age.

Regarding the geographical distribution of incidence (Table 1, Figures 1 and 2) and according to the Unit of Descriptive Epidemiology of the International Agency for Research on Cancer (IARC), only a few countries in Africa have reliable statistics on bone cancer. Among them, Mali has the highest standardized rate among males (1.4 per 100,000), while Algeria exhibits the highest rate among females (1.2 per 100,000), with a male/female ratio ranging from 0.75 to 1.55(3).

In the Americas, Chinese males in Hawaii have the highest incidence rate of bone cancer (6.4 per 100,000). Actually, this is the highest rate in the world. Among females, Paraguay has the highest incidence rate in the region (1.6 per 100,000). The highest male-female ration (9.0) in the world is found among Japanese Americans in Los Angeles, California. In the United States, Filipino males and Japanese females have the lowest incidence rates for bone cancer. Canada has a moderate frequency rate of bone cancer within the America. In Canada, Quebec seems to have the highest, and New Brunswick, the lower rate of bone cancer patients.

In Europe, Poland (Wilesia) has the highest incidence rate of this disease in both males and females, with 2.4 and 1.5 per 10,000 respectively, while Italy shows the lowest rates for males and females. Both France and Sweden have the highest male/female ratio (3.5).

In Australia, the incidence rate of this disease ranges from 0.3 per 100,000 in Tasmanian males to 2.0 in Maori males. In females, except for Maoris (1.2 per 100,000), most regions have an age standardized rate below 1.0 per 100,000. In general, this continents has the highest male/female ratio (around 1.6), indicating a higher incidence rate of bone cancer among males.

In Asia, the Philippines (Manila) has the highest rate of males while Thailand (Khon Kaen) has the highest frequency in females. It is interesting that Singapore Indians have the lowest male/female ratio of bone cancer in the world (0.5), while Israeli on-Jews show the second highest worldwide male/female ratio (4.0).

In brief, among males, Chinese in Hawaii have the highest incidence rate of bone cancer, while Khon Kaen of Thailand has the highest rate among women in the world(3).

With regard to the mortality rate from bone cancer, we compared existing statistics on different continents for two time periods (1984-86 and 1994-96). In general, the mortality rate of the world's population decreased in males from 1.01 to 0.91. In females, the rate in 1984-86 was 0.78, which fell to 0.63 in 1994-96 (Table 2).

The mortality rates from bone cancer rose significantly among both males (from 0.47 to 0.80) and females (from 0.41 to 1.04) in Africa, indicating a 7% increase among males and an increment of more than 15% among females for a period of 10 years. Although the number of cases is small, it still indicates a significant elevation of mortality from bone cancer. In the Americas, a 0.3% increase in male and a 0.4% rise in female mortality from bone cancer have been reported. Interestingly, this disease shows a higher frequency among Asian males compared to the rate 10 years ago (Table 2); the rate has remained unchanged among females. The two continents of Europe and Oceania have seen a sharp and significant decrease of the mortality rate. For example, in Europe, the rate fell by around 2.4% in males and by 1.8% in females. In Oceania, these declines were 2.4% in males and 2.9% in females.

The rate reduction could be mostly due to better and advanced treatment of bone cancer as well as disease prevention through public education and awareness in avoiding known risk factors. In general, it seems that the mortality rates from bone cancer are decreasing.


                   Etiological Factors for Bone Cancer

Bone cancer comprises around 5% of all cancers among children in developed countries, particularly in North America and Europe(4). There is little information about risk factors for bone cancer in adults, while some data are available on the etiology of this disease among children. Therefore, in this section, we will review the existing evidence of risk factors for bone cancer in children.


Genetics:
The possible role of genetics in the etiology of bone cancer was suggest in a cohort of 850 children with bone cancer in England(5). In a meta-analysis of five studies, twin cases developed osteosarcoma as well as retinoblastoma(6), while no twins with Ewing's sarcoma or non-specified bone cancer showed a affected co-twin. In general, family aggregation of osteosarcoma is rare. Recently Longhi et al(7) reported four cases of osteosarcoma in two siblings and in a father and son. These patients had no other tumors in their family history and had negative p53 mutations.

It seems that the frequency of certain types of cancer, such as melanoma and brain tumors (OR=1.9) or stomach cancer (OR=2.0), is higher among the first-degree relatives of cases of Ewing' s sarcoma of bone and soft tissues(8).

In a multicentre study of osteosarcoma in the United States and Canada(9), 3% of subjects exhibited germ-line p53 mutations. In another case-control investigation in Spain(10), a significantly higher proportion of children with osteosarcoma had HLA-Al1 and HLA-B7 antigens. Such a high frequency ofHLA-A11 antigen was also found in Japan in 1990(11), while the possible role of HLA-B7 reported by Barona et al'o was not confirmed by any other authors.

Socio-demographics:
Holly et al.(12) observed that regular intake of mixed vitamin supplements during childhood decreases the risk of bone tumors (RR=0.4; 95%CI 0.1-1.4), but another group(13) reported no association between vitamin or mineral supplements and bone tumors. Hartley et al(14) examined the possible role of breast-feeding in bone cancer prevention. They concluded that there was no association. Several other studies(14-16) found no association between bone tumors and maternal or paternal smoking during pregnancy, while another case-control investigation(17) of Ewing's sarcoma showed that both paternal and maternal smoking during pregnancy increased the risk of this disease in children. The relative risk was 2.0 from maternal smoking and 3.7 from paternal exposure. It should be mentioned here that if both parents were smokers, the risk was more than 7-fold higher compared with non-smoker parents.

Holly et al.(12) observed no association between maternal smoking during pregnancy and the risk of osteosarcoma. No association was reported for maternal alcohol intake during pregnancy and bone cancer in general(14), Ewing's sarcoma(12) or osteosarcoma(13). Winn et al(17) discerned no correlation between maternal coffee consumption during pregnancy and the risk of Ewing's sarcoma.


Ionizing radiation:
In a follow-up study of 220 children with tuberculosis treated by radium-224, around 16% developed bone cancer, particularlyosteosarcoma(18). In a case-control study(19), no association was found between osteosarcoma and postnatal exposure to diagnostic ionizing radiation.

Maternal reproductive history:
There appears to be no association between maternal age, number of previous pregnancies as well as difficulty of becoming pregnant and bone cancer(17). Mothers who use medications for nausea and vomiting during pregnancy have a significantly higher risk of having children with Ewing's sarcoma.

Two case-control studies(l3,19) observed no association between maternal estrogen use in the year before or during pregnancy and the risk of osteosarcoma in children up to age 24 years.

Medical History:
Schumaker et al. (20) reported an excess of rib abnormalities among Ewing's sarcoma patients, but not in osteosarcoma cases. A case-control study in the United States found a high rate of hernias and heart conditions among Ewing's sarcoma subjects(l7). The high frequency of hernias occurred early in life. This finding was not confirmed by another case-control investigation in California(12) or by a comparison of England and Canada(21).

Other factors:
In a large cohort of male, world class Finnish athletes, with 30 to 60 years of follow-up, no cancer sites showed a significant excess, but there was a significantly higher risk of sarcoma of the bone and soft tissue(22). It was suggested that the history of injuries during their active sports period may have been the major cause of the high frequency of bone and soft tissue sarcoma.

                                               Conclusion

In general, bone cancer is not a common malignant disease, and perhaps for this reason its etiology is not known. Furthermore, very few studies have been conducted in this field. Among different types of bone cancer, osteosarcomas contribute the highest proportion (36% ), followed by chondrosarcomas (30%) and Ewing's sarcoma (16%).

The highest incidence rate of bone cancer in males is found among American Chinese in Hawaii. Among females, the population of Khon Kaen in Thailand exhibits the highestincidence rate of this cancer in the world. American Japanese have the highest sex ratio for bone cancer, indicating that the frequency of the disease is 9-fold higher among males than females.

There appears to be no consistent association between osteosarcomas and maternal estrogen use during pregnancy, which has been suspected. It is also evident that maternal occupation, birth weight and the presence of congenital anomalies do not playa role in the etiology of osteosarcomas.

Reports on the genetics of bone cancer are not conclusive, but the frequency of certain types of cancer appears to be higher among the first-degree relatives of bone cancer cases.

Parental smoking habits during pregnancy may increase the risk of bone cancer among children. Ionizing radiation heightens the risk of some types of bone cancer. The possible role of pesticides in the etiology of bone cancer has been suggested but needs further investigation.


                                               References

1. Parkin, D.M., Stiller, C.A., Draper, G.I., Bieber, C.A., Terracini, B. & Young, IL., eds (1988) International Incidence of Childhood Cancer (IARC Scientific Publication No.87), Lyon, IARC Press.
2.Ries, L.A.G., Wingo, P.A.; Miller, D.S., Howe, H.A., Weir, H.K., Rosenberg, H.M., Vernon, S.W., Cronin K., & Edwards, B.K. (2000) The annual report to the nation on the status of cancer, 1973-1997, with a special section on colonrectal cancer. Cancer 88: 2398-2424.
3.Ferlay, I., Bray, F., Pisani, P., & Parkin, D.M. (2001) Globocan 2000. Cancer incidence, Mortality and prevalence worldwide, Version 1.0, IARC Cancer Base No.5, Lyon, IARC Press.
4. Parkin, D.M., Stiller, C.A., Draper, G.I., & Bieber, C.A..(1993). The international incidence of childhood cancer. IntI Cancer 42:511-520.
5.Narod, S.A., Stiller, C., & Lenoir, G.M. (1991) An estimate of the heritable fraction of childhood cancer. Br J Cancer 63: 993-999.
6.Buckley, I.D., Buckley, C.M., Breslow, N.E., Draper, G.I., Robertson, P.K., & Mack, T.M. (1996) Concordance for childhood cancer in twins. Med Pediat Oncol 26: 223-229.
7. Longhi, A., Benassi, M.S., Molendini, L., Macchiagodena, M., Picci, P., & Bacci, G. (2001) Osteosarcoma in blood relatives. Oncol Rep 8:131-136.
8.Novakovic, B., Goldstein, A.M., Wexler, L.H., & Tucker, M.A. (1994) Increased risk of neuroectodermal tumors and stomach cancer in relatives of patients with Ewing's sarcoma family of tumors. J Natl Cancer Inst 86:1702-1706.
9. McIntyre, I.F., Smith-Sorensen, B., Friend, S.H., Kassel, I., Borresen, A.L., Yan, Y.A., Risso, C., Sato, I., Barbier, N., Miser, I. et al. (1994) Gerrnline mutations of the p53 tumor suppressor gene in children with osteosarcoma. J Clin Oncol12:925-30.
10. Barona, P., Sierrasesumaga, L., Antillan, F., & VillaElizaza, I. (1993) Study of HLA antigens in patients with osteosarcoma. Hum Heredit 43: 311-14.
11.Shimizu, T., Chigira, M., Nagase, M., Wanatabe, H., & Udagwa, E. (1990) HLA phenotypes in patients who have osteosarcoma. J Bone Joint Surg 72A: 68- 70.
12.Holly, E.A., Aston, D.A., Ahn, D.K., & Kristiansen, J.J. (1992) Ewing's bone sarcoma, paternal occupational exposure, and other factors. Am J Epidemiol 135: 122-129.
13. Gelberg, K.H., Fitzgerald, E.F., Hwang, S. & Dubrow, R. (1997) Growth and development and other risk factors for osteosarcoma in children and young adults. Int J Epidemiol 26: 272-278.
14.Hartley, A.L., Birch, I.M., McKinney, P.A., Teare, M.D., Blair, V., Carrette, I., Mann, I.R., Draper, G.I., Stiller, C.A., Johnston, H.E., Cartwright, R.A., & Waterhouse, I.A.H. (1988) The inter-regional epidemiological study of childhood cancer (IRESCC): case-control study of children with bone and soft tissue sarcomas. Br J Cancer 58: 838-842.
15. McKinney, P.A., & Stiller, C.A. (1986) Maternal smoking during pregnancy and the risk of childhood cancer (letter). Lancet Aug 302: 519-520.
16. Soharan, T., Lancashire, R., Prior, P., Peck, E., & Stewart, A. ( 1995) Childhood cancer and parental use of alcohol and tobacco. Am Epidemiol 5: 354-359.
17.Winn, D.M., Li, F.P., Robison, L.L., Mulvihill, 1.1., Daigle, A.E., & Fraumeni, I.F. Ir. (1992) A case-control study of the etiology of Ewing's sarcoma. Cancer Epidemiol Biom Prev 1: 525-532.
18. Spiess, H., & Mays, C.W. (1973) Protraction effect on bone-sarcoma inducton of 224 Ra in children and adults. In: Sanders, C.L., Busch, R.H., Ballou, I.E. & Mahlum, D.D., eds. Radionuclide Carcinogenic. Proceedings of the Twelfth Annual Hanford Biology Symposium at Richland, Washington, May 10-12, 1972 (Atomic Energy Commission Symposium Series 29), Oak Ridge, TN, USAEC Technical Information Centre, CONF720505, pp. 437-450.
19.Operskalski, E.A., Preston-Martin, S., Henderson, B.E., & Visscher, B.R. (1987) A case-control study of osteosarcoma in young persons. Am I Epidemiol 126:118-126.
20.Schumaker, R., Mai, A., & Gutjahr, P. (1992) Association of rib anomalies and malignancy in chilhood. Eur J Pediat 151: 432-434.
21.Narod, S.A., Hawkins, M.M., Robertson, C.M., & Stiller, C.A. (1997) Congenital anomalies and childhood cancer in Great Britain. Am. I. Human Genet 60: 474485.
22. Pukkala, E., Kaprio, I., Koskenvuo, M., Kujala, U., & Sarna, S. (2000) Cancer incidence among Finnish world class male athletes. Int J Sports Med 21:216-20.