Human biology and epidemiology of childhood bone cancers: a review.
A. Polednak
1985-02-01
Abstract:Clinical case series and descriptive epidemiologic studies have suggested a relationship between childhood bone cancers and accelerated skeletal growth, but definitive studies are lacking. Anthropometric studies of osteosarcoma and E wing's sarcoma cases and normal controls, along with their families, are needed to assess the role of growth and maturation. Some descriptive epidemiologic features of these two histologic types of childhood bone cancer are reviewed, along with evidence for the role of hormones (i.e., growth hormone, somatomedins, androgens). Theoretical mechanisms involving hormones are outlined including their possible role as promoters or as factors interacting with environmental agents (such as ionizing radiation). Clinical-epidemiologic studies of families prone to osteosarcomas and to constellations of cancers (and/or skeletal diseases and malformations) may aid in distinguishing genetic and environmental factors and defining specific mechanisms (e.g., hormonal, cytogenetic) of susceptibility. Our understanding of the etiologies of osteosarcoma and E wing's sarcoma in children may depend on the integration of knowledge from various areas, including: the differentiation and growth of bone cells and their precursors; the hormonal control of human skeletal growth and maturation; epidemiologic studies of exposures to environmental agents; clinical-epidemiologic studies of families prone to certain cancers; and developments regarding the role of oncogenes and cytogenetic abnormalities in specific cancers. Although primary bone cancer is a rare disease, with average annual ageadjusted incidence rates per 100,000 of 1.0 for males and 0.6 for females in 1973-77 (Young et al. 1981), a substantial proportion of cancers in children and adolescents occurs at this site. In the population-based New York State Cancer Registry, for example, primary bone cancers comprised about 12.5% of all cancer cases diagnosed in 1976-79 among Upstate (i.e., New York State exclusive of New York City) residents aged 0-34 years. Clinical case series and epidemiologic studies, reviewed below, have suggested a possible relationship between childhood bone cancers and accelerated skeletal growth, but definitive analytic studies and investigations of interactions between skeletal growth and other factors are lacking. There are ever-increasing opportunities for integrating epixBureau of Chronic Disease Prevention, Division of Community Health and Epidemiology, New York State Department of Health, Albany, New York 12237. Human Biology, February 1985, Vol. 57, No. 1, pp. 1-26 © Wayne State University Press, 1985 This content downloaded from 157.55.39.127 on Sun, 26 Jun 2016 07:43:20 UTC All use subject to http://about.jstor.org/terms 2 Anthony P. Polednak demiologic data with information from several areas including the endocrinology and genetics of human growth, the biology of bone cells (i.e., their physiology, differentiation and growth) and ideas on carcinogenic mechanisms. This review examines the state of present knowledge on the etiology of childhood bone cancer, demonstrates the need for integrating information from these several areas and suggests directions for future research. Because of many epidemiologic differences between the two major histologic types of childhood bone cancers (i.e., osteosarcoma and E wing's sarcoma) (Miller, 1981), these cancers will be discussed separately wherever possible. Some Descriptive Epidemiologic Findings The hypothesis that risk of childhood osteosarcoma is related to rates of skeletal growth is based on the associations between osteosarcoma rates (by age, sex and anatomical site) and known patterns of growth in stature. Price (1955, 1958) recognized that the earlier occurrence of osteosarcomas in girls (vs. boys) corresponded to their earlier adolescent growth spurt, that the greater incidence in young males was related to their greater rates of growth and that the predilection of osteosarcomas for the lower end of the femur reflected maximal growth rates in this region. A rather similar hypothesis was advanced earlier by Johnson (1953), which was based on the association between bone cancer risk and antecedent bone diseases involving "excessive cell activity" as well as on growth rates. These observations regarding childhood osteosarcoma rates and known age and sex patterns in growth have been confirmed by others (Glass and Fraumeni 1970; Hems, 1970; Larsson and Lorentzon, 1974). Incidence data derived from the population-based New York State Cancer Registry for residents of New York State for 1974-80 were examined for age and sex differences in osteosarcoma and Ewings sarcoma. Included were 181 cases of osteosarcoma and 143 cases of Ewing's sarcoma diagnosed under 25 years of age (Table 1); population data reported from the 1980 Census were used for denominators. Average annual incidence rates by sex and age (Table 1) are also plotted in Figure 1. For osteosarcoma (Table 1 and Figure 1) incidence rates are slightly higher for females than for males in some age groups before 15 years, as reported in other studies. In contrast to females, male osteosarcoma rates show a distinct peak at age 17 years. These sex differences have been attributed to differences in growth patterns. That is, the adolescent height spurt begins earlier in girls than boys (10.5 vs. 12.5 years), reaches This content downloaded from 157.55.39.127 on Sun, 26 Jun 2016 07:43:20 UTC All use subject to http://about.jstor.org/terms