11 June 2007
Dear Sir
Re: Risk factors for osteosarcoma in young people in Cornwall: A case control study; Wright and Pheby, JEHR Vol 5 Iss 2 (2006)
In their paper published late last year, Wright and Pheby (1) presented incidence rates for osteosarcoma in young people that cannot be reproduced from their references. There are problems with the definition of the disease described, the geographic areas considered and the resident population of these areas.
The authors presented numbers and rates of osteosarcoma cases in the UK, the South West and their study area in the West of Cornwall (see Table 1). We have attempted to reproduce this table from their reference sources. Our results are shown in Table 2 (more recent data has been added where this is available). We were unable to find some of the incidence data allegedly available from the sources stated in the paper and in these cases we have substituted data from the United Kingdom Association of Cancer Registries (UKACR). The two tables are different in many respects.
The authors used The Office of National Statistics (ONS) as a source of incidence statistics. There are three points to note about this:
1) The ONS does not publish data on osteosarcoma. Osteosarcoma is a particular bone cancer defined by histological type. The ONS Cancer Registrations Series MB1 used by Wright and Pheby is a set of tables published annually detailing the cancers registered in England. It does not include detail of histological type and therefore cannot be used as a source of data on osteosarcoma. Data for the more generic bone cancer (defined as ICD-10 codes C40 and C41) is available. Our table presents data for this group of diseases, which includes osteosarcoma together with other conditions.
2) The ONS publishes cancer statistics for England (2), not the UK.
3) Table 1 in each MB1 release details of cancer registrations by sex, anatomical site (ICD-10 Code) and age. It is possible to use this data to look at incidence rates by each of these variables for England as a whole but not broken down by region. Table 4 in MB1 provides a regional break down of registrations but not by age. MB1 cannot therefore be used as a source of age-specific data for bone cancer in the South West of England. We have used the MB1 tables to source the number of bone cancers in people under the age of 25 in England.
Our data describing the number of bone tumours in people under 25 in the Government Office Region of the South West has been taken from the UKACR Cancer Information System (accessible through Cancer Registries and within the NHS by arrangement). This source was also used to produce data for the Kerrier District Council in the West of Cornwall. We chose this as the smallest area, which includes Wright and Pheby’s study area, for which population data is readily available. The population of Kerrier is roughly double the stated population of the study area.
Even assuming that Wright and Pheby presented cases of bone cancer rather than osteosarcoma, the numbers of cases in the two tables compare poorly, particularly, for the South West. Wright and Pheby present 210 cases for the 6 years 1995–2000, while the UKACR data presented in our table, which includes all bone cancer cases in the region, is 90. Wright and Pheby’s count therefore leads to an overestimate of the rate of bone cancer in the South West.
ONS publishes population estimates for a range of geographies. Wright and Pheby’s population data bears little resemblance to available ONS data [4]. The ONS population data for England exceeds the authors’ data for the whole of the UK. The ONS population for the South West is more than double that in the authors’ table.
Finally, we would expect the Wright and Pheby study period of 1995–2005 to be described as 11 years (i.e. January 1995 to December 2005), not the 9 stated in their table.
Osteosarcoma is a rare but often devastating disease amongst young people. There has been understandable concern about the higher than average incidence of bone cancer among young people in Cornwall in recent years and this disease is under constant review by the South West Public Health Observatory. There have been no registered cases of osteosarcoma among young people in the West of Cornwall since 2003.
Yours faithfully
Dr Julia Verne - Deputy Regional Director of Public Health & Director of SWPHO
Mr Andy Pring – Senior Information Analyst, South West Public Health Observatory
Dr Sean McPhail – Head of Cancer Analysis, South West Public Health Observatory
Response from the authors.
Dear Sir,
Verne et al have indicated that data pertaining to the incidence of osteosarcoma is not readily accessible. We agree with this, which is why we had to access more than one data source in order to present comparative data on it.
However, the presentation of these data was not the main purpose of our paper, which was to investigate possible risk factors for osteosarcoma in young people by studying a number of cases, with matched controls, that occurred in and around Helston in recent years (1). Dr. Verne and colleagues concede that “There has been understandable concern about the higher than average incidence of bone cancer among young people in Cornwall in recent years.” To state, though, that “... this disease is under constant review by the South West Public Health Observatory” is misleading. The SWPHO’s ‘constant review’, their own reports indicate, is entirely through monitoring the incidence of osteosarcoma via the South West Regional Cancer Registry. The cancer registry does not hold exposure data, so SWPHO has not conducted any analytical epidemiological research into the causes of osteosarcoma. Their previously published statement that “No single environmental risk factor for the cases in Helston could be identified” (2) is untenable, as they have not investigated any environmental risk factors.
The assertion of error in our presentation of national and regional population data is a complete red herring. The population data quoted by Verne et al are mid-year estimates for 1997, i.e. six years after the datum point of the 1991 census. It is well known that such estimates are very prone to error, and that these errors are cumulative, increasing year by year after the census. Verne et al state that “The ONS population for the South West is more than double that in the authors’ table”. There is a very simple explanation for this, which clearly Verne et al have overlooked. When they talk about ‘the South West’, they indicate that they mean the South West Government Office Region. When we mentioned the South West, we were referring to the former South Western health region, i.e. the counties of Gloucestershire, Somerset, Devon and Cornwall, and the former county of Avon, which was the region covered by the former South-Western Regional Cancer Registry when one of us (DP) was its Director, and which constitutes roughly half of the region for which the Government Office South West has responsibility.
The statement that “There have been no registered cases of osteosarcoma among young people in the West of Cornwall since 2003” is another red herring. We identified a very strong association between domestic radon levels and the occurrence of osteosarcoma. This merits further research. Osteosarcoma is a rare cancer. Clearly, most young people exposed to high levels of radon do not develop osteosarcoma. We do not know why some people exposed to ionising irradiation, from whatever cause, develop cancer, while many others do not. Such catastrophes appear to be random events, and it is well known that random events are not evenly distributed.
While our observation of the association between radon and osteosarcoma was, we believe, the first such observation in this age group, it was biologically plausible, and not surprising as ionising radiation has been shown to be a risk factor in other age groups. What is surprising is that, given the progress that has been made in elucidating the role of a potentially remediable risk factor for osteosarcoma, the SWPHO appears so keen to play down the role of radon, and not to cooperate with research that could save lives. We hope that they will rethink their attitude, as we would welcome a productive and fruitful research collaboration on this issue.
Yours faithfully,
Derek Pheby, Mel Wright
- 1) Wright M, Pheby D. Risk Factors for Osteosarcoma in Young People in Cornwall: A Case Control Study. Journal of Environmental Health Research 2006; 5(2):61-69.
- 2) Verne J, Harling C, Miles D. Investigation into an alleged cluster of osteosarcoma. Bristol, South West Cancer Intelligence Service, 2003.