This study of primary bone cancer in 0 - 39 years olds in northern England is the most recent analysis that has examined incidence and survival for this age group. Cases were limited to those aged less than forty years because this is the upper age-limit for entry into most clinical trials [16, 23]. Furthermore, this age range coincides with the peak age-incidence distribution for the majority of malignant primary bone tumours. It is likely that a substantial proportion of the 0 - 39 year old patients have been treated according to trial protocols (although we did not have access to data on trial entry). The basic case incidence and survival data were accurately recorded by the registry. However, data on stage and treatment were not consistently and reliably recorded. Therefore, some of the differences in survival may be due to differences in case mix with respect to stage or treatment regime.
Although there is some evidence for an improvement in survival from Ewing sarcoma during the period 1981 - 2000, cases aged 15 - 39 years had worse outcome than those aged 0 - 14 years. In contrast, for osteosarcoma, there was no significant change in survival throughout the study period and no evidence that cases aged 15 - 39 years had worse outcome than childhood cases (aged 0 - 14 years).
The incidence rates reported in the present study were similar to those reported in other countries with mainly white populations. In addition, the predominance of osteosarcoma and Ewing sarcoma found in this study are also typical patterns reported in ethnically similar populations (5% were minority ethnic population) [24, 25].
Incidence of osteosarcoma, chondrosarcoma and Ewing sarcoma was higher among males than females. This is consistent with previous literature, which has reported that although males and females tend to show similar incidences in childhood [11, 15, 24], when an extended age range is included, all three diagnostic groups are more common in males .
Osteosarcoma and Ewing sarcoma both had incidence peaks in the 15 - 29 age group. This is consistent with previous studies, which have widely reported incidence of these two diagnostic groups to rise after the onset of puberty, when young people are undergoing a growth spurt and bones experience rapid growth [8, 11, 24].
This study found an increase in the incidence of osteosarcoma, but no temporal changes in the incidence of Ewing sarcoma and chondrosarcoma during the period 1981 - 2002. In both the USA and Europe, overall bone cancer incidence across all age groups has been reported to have remained steady over the last 30 years [25, 26].
We have previously analysed the survival of children resident in northern England and the West Midlands during the same time period . For children we found that whilst there was an improvement in survival from Ewing sarcoma, there was no improvement for osteosarcoma. The present study confirms that the lack of improvement in survival for osteosarcoma also applies at ages 15 - 39 years. However, it has indicated a worse outcome for Ewing sarcoma cases in age group 15 - 39 years. Another recent study from the USA analysed data from the Surveillance, Epidemiology and End Results Program and has found that there has been no statistically significant improvement in survival from osteosarcoma (at all ages) from 1984 to 2004 .
The most recent national data from Great Britain reported five-year survival rates of 53% and 51% for cases aged 0 - 39 years during 1990 - 1994 . Our data have shown that survival for northern England was better for osteosarcoma (58%) but worse for Ewing sarcoma (43%). Data for patients of all ages, from the United States National Cancer Database, also reported worse five-year survival (51.2%) for osteosarcoma and better five-year survival (50.2%) for Ewing sarcoma . Better survival of osteosarcoma compared with the US study is likely to be due to the exclusion of patients aged more than 40 years from the present study. Another study of osteosarcoma from Finland reported five-year survival of 65% during 1981 - 1990 . However, survival rates in northern England were similar to 15 - 24 year olds from a multi-centre (twenty countries) European study, which found five-year survival rates of 58% for osteosarcoma and 42% for Ewing sarcoma .
A number of studies have reported worse survival from osteosarcoma in adults aged more than forty years [27, 28, 31, 32]. It is very well recognized that those aged more than 40 years have poorer prognosis than younger patients . In our study of 0 - 39 year olds, we only found worse outcome for cases of Ewing sarcoma aged 15 - 39 years, but not for osteosarcoma. However, some other studies have specifically studied adolescent and young adult cases and have indicated worse outcome in this age group for both osteosarcoma and Ewing sarcoma compared with children [22, 34, 35].
The worse survival in cases of Ewing sarcoma aged 15 - 39 years is consistent with previous studies and may be due to a number of different factors including treatment, delays in diagnosis, metastatic disease, site and the stage of the tumour [36–39]. We were not able to investigate these factors within the scope of the present study. However, these issues will be addressed in future studies that focus on the Teenage and Young Adult (TYA) age group. A special National Cancer Research Institute TYA Clinical Studies Development Group has been formed and will oversee analyses of a new national TYA cancer dataset.