This article is a US Government work, and, as such, is in the public domain in the United States of America.
Osteosarcoma incidence rates in the United States peak in adolescence and in the elderly. The international patterns of osteosarcoma incidence in children have been described, whereas those for young, middle age or elderly adults have not. Using the Cancer Incidence in Five Continents, International Agency for Cancer Research database we compared incidence rates for children and adolescents (age 0–24 years), the middle age group (25–59 years) and elderly (≥60 years) persons by world regions and individual countries. Overall, worldwide osteosarcoma incidence rates were quite similar in the younger age groups. The greatest variation in incidence rates was observed in the elderly. Published 2009 UICC
Bone tumors make up about 3–5% of childhood cancers and less than 1% of cancers in adults.1–4 Of these, osteosarcoma is the most commonly diagnosed primary malignant bone tumor.1, 5–7 Individuals with localized osteosarcoma have an average 5-year survival of about 80% but those with metastatic disease have much worse outcomes.8 Its incidence is bimodally distributed by age with peaks in adolescence and in the elderly.1, 8–12 Osteosarcoma incidence in childhood and adolescence seems to be relatively consistent throughout the world;5, 13–17 however, international comparisons of incidence for other age groups have not been described.
We recently showed that osteosarcoma in children and adolescents, middle ages, and elderly persons have unique epidemiologic features in the United States using data from the Surveillance, Epidemiology and End Results program.8, 18 Osteosarcoma incidence in the youngest cases (age 0–24 years) was greatest in Asian/ Pacific Islanders, whereas it was greatest in Blacks and Whites in the middle age group (age 25–59 years) and elderly (age 60+ years) patients, respectively.8 This study also found that the first osteosarcoma incidence peak occurred at younger ages in women than in men, and from 1973 to 2004 osteosarcoma incidence increased in the youngest age group, variable in the middle age group and decreased in the elderly. The anatomic site distributions were more diverse for the older age groups compared with the 0–24 group, and survival rates varied by age group, anatomic site and disease stage.8 The disparate features found among these 3 age groups suggest that they should be studied as separate groups.
Here, we expanded the characterization of osteosarcoma incidence to include all age groups: children and adolescents (age 0–24 years), middle ages (25–59 years) and elderly (60+ years) based on data from the Cancer Incidence in Five Continents, International Agency for Cancer Research database, to better understand the international epidemiology of osteosarcoma. Advancing understanding of osteosarcoma epidemiology has the potential to identify new etiologic insights in this complex disease.
Material and methods
The world incidence rates and case counts were estimated for osteosarcoma from the Cancer Incidence in Five Continents, International Agency for Cancer Research CancerBase No. 7, CI5-annual detailed data set.19 This database includes information from regional cancer registries in Colombia, Ecuador, Canada, United States, India, Israel, Japan, Kuwait, Philippines, Singapore, Thailand, France, Italy, The Netherlands, Poland, Spain, Switzerland, England and Australia; and national registries in Costa Rica, Denmark, Estonia, Iceland, Slovakia and Scotland. We included data from both nationally covered countries (unbiased) and regional cancer registries, which may not be representative of the whole country, to maximize estimates of international coverage. The geographic coverage by country and registry are described in more detail in the referenced database.19 Incidence rates were estimated for each regional registry (for the countries with more than 1 registry), country, world region and for all of the countries combined (the world rate). For countries with multiple regional registries and for the world estimates, data were pooled from all of the registries for each country, or all of the available countries for the world rate, and the incidence is calculated using the pooled data. All osteosarcoma incidence rates are per 1,000,000 and are age standardized to a world population to account for differences in the age distribution of patients between countries.
All osteosarcoma cases were classified according to the International Classification of Disease for Oncology (ICD-O-2, codes 9180-9200).20 Data are presented by age group (0–24, 25–59, and 60–85+ years) and by sex. These age groups were chosen to focus on the 2 incidence peaks in children and adolescents and the elderly, and the plateau in the middle age group. Geographic region and time period are described in the table and figure. Rates that are based on less than 10 cases are considered unstable and should be interpreted with caution (these estimates are indicated by italics).
Osteosarcoma incidence rates among individuals ≤24 years were generally consistent worldwide with peaks around puberty. Most rates ranged from 3 to 5 per million in men (average 4.3) and 2 to 4 per million in women (average 3.4) (Table I). Osteosarcoma was more common in men than women in most countries. The overall world male-to-female ratio of osteosarcoma in ages 0–24 years was 1.43:1. Incidence peaked in men at age 15–19 years and in women at 10–14 years in all regions studied (Fig. 1). Incidence rates in women were as high or higher than those observed in men less than 15 years (age 0–14 years) almost everywhere, but then increased later in puberty for men.
Table I. Incidence of Osteosarcoma by Age Group and Sex
Country (yr), regional registry
ASR, world age-standardized rate per 1,000,000; n, number of cases.
Italics indicate potentially unstable rates based on less than 10 cases.
Countries with national registries.
Includes data from registries in Los Angeles and San Francisco.
Canada has national and regional registries, both data are shown. The world rates include data from all listed countries and registries.
The majority of osteosarcoma incidence rates among persons aged 25–59 years were in the range of 1.5–2.5 per million in men for each country (average 1.9) and 0.5–2 in women (average 1.36) (Table I). There was a plateau of incidence observed in all regions in this age group, although more variation was observed in Latin America (Fig. 1). Osteosarcoma was more common in men than in women in most countries, with a male-to-female ratio of 1.28:1.
There was a second peak of osteosarcoma incidence in the elderly observed in the majority of countries and regions, although it was comparatively lower in Europe (excluding the United Kingdom; Fig. 1). Osteosarcoma incidence rates were more varied among the elderly; however, the majority were in the range of 2.5–5 per million in men (average 4.0) and 1.5–4 in women (average 3.1) (Table I). Men aged 75+ years had strikingly higher incidence rates in Australia (15–18) and Canada (10–11), and for male patients aged 80+ years in the United Kingdom (11.6) when compared with other countries (rates in the range of 1–7 for men aged 75+ years) (Fig. 1). Osteosarcoma was more common in men aged 60+ than women of the same age group in most countries, although rates were higher in women in Western Europe (Table I). The worldwide male-to-female ratio was 1.01:1 for this age group.
This is the first report of international osteosarcoma incidence in subjects of all ages to our knowledge. The recent study of incidence and survival patterns in the United States found that osteosarcoma incidence, time trends, survival, pathologic subtype and anatomic site were varying among children and adolescents, middle ages and elderly persons.8 This suggested that there may be important osteosarcoma incidence differences in these age groups throughout the world.
In individuals ≤24 years, the incidence rates were minimally variable between countries. Previously, a higher incidence of childhood osteosarcoma was reported in Southern Europe,15 particularly in Italy.5 We also found a slightly higher incidence in Italy (males: 5.3, females: 4.5) for ages 0–24 years, particularly for men in Florence and Torino. However, the highest rates were in Latin America (males: 7.0–7.6, females: 3.5–4.9). Rates were particularly high in all age groups in the Philippines and Ecuador. A high incidence of osteosarcoma has been observed in an earlier report in 2 African countries, Sudan and Uganda, compared with the relative frequency in populations of European origin.14 This is consistent with the finding that young and middle aged individuals of African American descent in the United States had higher rates of osteosarcoma than whites.8 It is important to note that other differences between ethnic groups may be observed, but data are less complete in those populations.
The earlier worldwide incidence peak observed in adolescent girls compared with boys links osteosarcoma to bone growth and an earlier occurrence of the adolescent growth spurt in girls, supporting the idea that bone growth and/or hormonal changes during puberty may have a role in osteosarcoma pathogenesis.21–23 The observation that the vast majority of adolescent tumors develop in the long bone epiphyses of the lower limbs1, 8, 24, 25 lends further support to this hypothesis.
An interesting difference between the age groups is the overall 1:1 male-to-female ratio of osteosarcoma cases in the elderly compared with the younger age groups male predominance (ages ≤24, 1.43:1; ages 25–59, 1.28:1). In the United States using Surveillance, Epidemiology and End Results incidence data, osteosarcoma in the elderly was found to occur slightly less commonly in men than in women (0.9:1).8 The rates were strikingly high for elderly male patients in the United Kingdom (particularly in the North Western and Oxford Regions), Australia and Canada, to our knowledge the first time this has been reported. The interpretation of data from some Asian and Latin American countries was limited because some of these rates are based on less than 10 cases and may be unstable. Rates were not available for each elderly age stratum in these 2 regions, so it was difficult to determine an incidence peak in the elderly; interestingly, numbers of cases reported and the overall rates were comparatively lower in these regions.
The greater geographic variation observed in elderly patients could be due to several reasons: (i) diagnosis and/or classification differences among countries; (ii) differences in completeness of cancer registration; (iii) differences in environmental exposures that could alter risk of primary osteosarcoma; (iv) exposures, such as irradiation, that could increase the chance of secondary osteosarcoma; and/or (v) different genetic components to osteosarcoma in the elderly, such as Paget's disease.26 In elderly patients, osteosarcoma is often considered a secondary neoplasm attributed to the sarcomatous transformation of Paget's disease of bone.27, 28 We were unable to separate osteosarcomas occurring with Paget's disease from those that did not with this database. There is marked geographic variation observed in the prevalence of Paget's disease, with a high prevalence noted in the United Kingdom, Australia and North America, and lower levels in Asia and the Middle East,29, 30 which is similar to the high and low regions of elderly osteosarcoma incidence.
Competing risks due to different environmental and/or genetic risk factors could also be involved in the geographic variation observed in the elderly patients. For example, osteosarcoma in this age group may occur in body sites that have been previously irradiated for a different cancer,27, 28, 31 a variable that we were unable to assess. Differences based on geographic location in access to radiotherapy and other cancer treatment modalities could also exist that would affect the rates of secondary osteosarcoma.
It is theoretically possible that vitamin D deficiency could contribute to risk of osteosarcoma since vitamin D is required for bone development and vitamin D deficiency has been associated with risk of developing colon, prostate, breast and several other cancers.32, 33 Vitamin D deficiency is recognized as pandemic34–38 and the elderly may be particularly affected.37, 39 One could also speculate that the increased rates in subjects of African descent are related to differences in vitamin D absorption. This hypothesis in osteosarcoma susceptibility has yet to be tested.
It is difficult to compare international survival rates for osteosarcoma because there are few estimates in the literature, few available data sources for older persons, and time frames have been reported differently. However, data from the Automated Childhood Cancer Information System (for ages 0–19 years from 1993–1997)40 and the Eurocare-3 Study: Survival of Cancer Patients in Europe (for ages 0–14 years from 1990–1994)41, 42 show that the 5-year survival rates for children and adolescents are similar in most European countries (55–75%), except for lower rates observed in Slovakia (19–20%), Estonia (26%) and Denmark (39%). Similar 5-year survival rates have also been observed in the United States (62%, ages 0–24; 68%, ages 0–14)8, 43 and Japan (51%, ages 0–14).44 The lower survival rates observed in Estonia, Slovakia and Denmark could be a consequence of limited access to early diagnosis and appropriate treatment.43 The 5-year osteosarcoma survival rates in the United States8 and Japan44 increased significantly from the 1970s to the 1980s (US: 51 to 63%; Japan: 18 to 51%), but have not improved further since the 1980s to the 1990s in the United States8 and Europe.42, 45 This lack of improvement during the last decade indicates an international need for more efficient management and new treatment strategies.
In summary, worldwide osteosarcoma incidence rates varied mostly in the elderly, with little geographic variation among the younger age groups. These data illustrate the importance of separating osteosarcoma incidence data by these age groups and inclusion of this information in more studies and databases.
The authors thank anonymous reviewers for their helpful comments.