Ms Christine Peat
National Assembly for Wales
New Crown Building
Cathays Park
CARDIFF
CF1 3NO
Dear Ms Peat
ADVICE FROM COMARE REGARDING THE INCIDENCE OF
CANCER IN THE CHEPSTOW AREA
I enclose a paper, discussed and agreed at the COMARE meeting
of 20 June 2002 which answers the specific questions asked by
the Welsh Assembly Government.
Yours sincerely
Professor BRYN BRIDGES
COMARE Chairman
----------------------------------------------------------------------------------------------------------------------
CANCER INCIDENCE IN THE CHEPSTOW AREA
At the request of the Welsh Assembly Government, COMARE have considered
the report by Dr Steward (Director of the Welsh Cancer Intelligence
and Surveillance Unit, WCISU) into cancer incidence in the Chepstow
area. This was prompted by a report from Green Audit suggesting
a leukaemia cluster within 10 km of the Oldbury nuclear power
plant. Dr Steward was kind enough to come to COMARE on 21 March
to present the results in person. Dr Steward's report found that
the number of cases of childhood myeloid leukaemia identified
by Green Audit was the same as those identified from the Welsh
Cancer Registry and WCISU. However neither COMARE nor WCISU agree
with the conclusions drawn from these data by Green Audit since
there was no a priori hypothesis that radiation should cause an
increase in myeloid leukaemia uniquely in the age groups reported
by Green Audit. The Welsh Assembly Government asked COMARE some
specific questions. These questions were considered by COMARE.
Does the study demonstrate or suggest a raised incidence
myeloid leukaemia in the 0-4 age group in the Chepstow area?
The incidence rate was higher but due to the small number of cases
(3 in 25 years and none in the latest 8 years, 1991-1998) this
may be due to chance. COMARE has agreed to consider myeloid leukaemia
in the 0-4 age group in the COMARE report on geographical variations.
Would radiation-related neoplasms be expected to be more
apparent in the 0-4 year age group rather than the 5-14 age group?
COMARE considers that radiation related childhood cancer would
not be expected to be concentrated primarily in the age range
0-4y. Therefore, to maximise statistical power to detect a real
effect, the whole age range 0-14y should be considered.
The majority of leukaemia in infants (under one year old) involve
rearrangements of the MLL gene. This abnormality has never been
associated with exposure to ionising radiation. Dr Steward's analysis
did not report the particular subtypes of leukaemia in the cases.
COMARE would be interested to comment on these data if they were
available.
Does the pattern of disease in the study point to man-made
radiation, eg, from nuclear discharges?
Exposure to radiation, at a dose sufficient to produce measurable
heath effects, would be expected to lead to an increase in all
leukaemias, with preservation of the background ratio between
lymphatic and myeloid leukaemia. This would be followed some years
later by a rise in the number of solid tumours. Acute lymphatic
leukaemia would be the dominant type, because that is the most
common type of leukaemia in 0-4 year olds. Dr Steward's study
did not show increased incidence of childhood leukaemia as a whole,
total childhood cancer, or adult breast or prostate cancer within
10 km of Oldbury nuclear power plant. The reported pattern of
cancer incidence does not suggest that man-made radiation is the
cause of either the childhood or the adult cases investigated.
Are the methods used in the study appropriate (eg, the
use of one-sided p values and the 90% confidence interval)?
In general, the methods are appropriate though it would be more
usual to use a 95% confidence interval and a two-sided test, which
is more conservative.
Top
