Wing Study and Multiple Myeloma

[Andy Hull <hull@mail.sep.bnl.gov>]

By coincidence, I happened to have a copy of the 1981 GAO Report, "Problems
In Assessing the Cancer Risks of Low-Level Ionizing Radiation Exposure" on
my desk "to be read" stack, having been reminded of it by Carol Marcus in
an item she had put out on Radsafe some months ago in connection with the
Rocketdyne study.
Having now scanned it for information pertinent to the recent Wing Study,
I'm passing on some of what I found.  

First, leukemia and myeloma are uncontrolled divisions of white blood cells.
In the former, the body is flooded with stem cell progeny. In the latter,
the cells are not found in the blood, but remain in the bone marrow, where
they cause destructive lesions. In the December 26 issue of NCHS's Monthly
Vital Statistics Report, gives the  Mortality Rate for Multiple Myeloma
(ICD 203) for the twelve months ending with April 1997 as 3.9/100,000. The
American Cancer Society's Cancer Facts and Figures-1997, provides 20 year
trends and a breakdown by gender.  The rates per 100,000 for 1971-73 were
Male 2.8 and Female 1.9 and for 1991-93 Male 3.8 and Female  2.6. Second,
as has already seen stated, it seems doubtful that the study includes
individuals to provide meaningful dose-effect information. For example, the
GAO Report states "It has been estimated that obtaining reliable estimates
of the excess breast cancer risk of women exposed to one rad of x-rays
would require a 10-year    study of 50 million exposed women and 50 million
controls. Even if the breast dose were 10 rads, the required sample size
would be 1 million  women."  Even taking into account the lower background
rate of multiple myeloma, thus making an excess of it more apparent than
for breast cancer, the required number of exposed workers for a valid study
seems large compared to those that were included in this study.  

Third, multiple myeloma is relatively rare. Thus, in a small population,
the chance occurrences (or non-occurrence) of one or two cases could
produce a large chance meaningless variation, a point made by GAO in its
consideration of the . 
1997 study of Hanford workers by Mancuso, Stewart and Kneale (Chapters 15
and 16 ," GAO Analysis of the Hanford Data".). It concluded "The number of
multiple myeloma and pancreatic cancers is at present too small to
distinguish fluctuations from effects definitively".  Perhaps the Wing
Report offers more assurance, but this remains to be seen.

Andrew P. Hull
S&EP-BNL,
Upton,NY 11973
Ph.  516-344-4210
Fax  516-344-3105
e mail:   hull@mail.sep.bnl.gov