|
By GIFF JOHNSON A US study on cancer in the Marshall
Islands released this week estimates that about 530 cancers
attributable to radiation exposure could occur in the 13,940
people alive in the RMI in 1954 — and that more than 55
percent of these cancers are yet to develop or be diagnosed.
The report, prepared by the US government’s National
Cancer Institute (NCI) at the request of the US Senate
Committee on Energy and Natural Resources, also said that
the baseline number of cancers expected to naturally occur
in this population of 13,940 is about 5,600 — about half of
which have yet to develop because the population was young
in the 1950s, with a high percentage under 10 years of age.
The study was requested by the Senate committee in
preparation for hearings and review of the RMI’s changed
circumstances petition. It is dated September 2004 but did
not get into circulation in the RMI until earlier this month
after RMI officials became aware of it because it was
referenced in a US Congressional Research Service assessment
of the changed circumstances petition issued in mid-March
this year.
Is the new NCI study helpful to the RMI or the US
government in their opposing positions for and against
additional compensation for nuclear test damages?
Some US officials are taking the view that since the
Nuclear Claims Tribunal has awarded claims for more than
2,000 conditions, this shows that there has been “enough”
compensation because the NCI study indicates only 530
cancers could be related to fallout exposure (almost half of
the 2,000 Tribunal awards have been for non-cancerous health
conditions).
Officials in the RMI, however, see it as vindication of the
Tribunal’s administrative procedure for compensating certain
cancers without regard to proof of exposure because cancers
cannot be specifically identified as naturally occurring or
radiation caused.
“We estimate that the nuclear testing program in the
Marshall Islands will cause about 500 additional cancer
cases among Marshallese exposed during the years 1946-1958,
about a nine percent increase over the number of cancers
expected in the absence of exposure to regional fallout,”
the NCI study said.
“More than 85 percent of those radiation-related cases
would likely occur among those exposed in 1954 on the atolls
of Rongelap, Ailingnae, Ailuk, Mejit, Likiep, Wotho, Wotje
and posssibly Ujelang,” the study said.
Curiously, the study authors do not mention Utrik, one of
the so-called ‘four nuclear test-affected atolls’.
The study notes the generally poor quality of cancer data
in the RMI and other Pacific Islands, and notes that there
are some similarities between Micronesians and Polynesians
in Hawaii. The authors of the study based their estimated
baseline and excess risks cancers on “all-race rates” in the
US, adjusted to reflect age-standardized rates for ethnic
Hawaiians.
The authors of the study say that at lower dose levels
the relation between the size of the radiation dose and
radiation-related cancer risk (the higher the dose, the
higher the risk) which is well established. But this “cannot
be assumed to hold at the extremely high radiation doses to
the thyroid gland and the colon estimated for the
populations of Rongelap and Ailingnae,” the study says. The
study, therefore, estimated the cancer risk for these two
atolls at the upper limits — accounting for nearly 30
percent of the estimated 530 cancers caused by radiation.
In a letter to Senator Pete Domenici, who chairs the
Senate Energy Committee, Dr. Andrew C. von Eschenbach,
director of NCI, cautioned that the “estimated number of
cancers to be expected is highly uncertain because 1) dose
estimates are uncertain; 2) baseline cancer rates are
approximate; and 3) organ-specific doses estimated for some
atolls are so high that related radiation risks are not
easily derived based on findings from studies of other
opulations such as the Hiroshima and Nagasaki A-bomb
survivors.”
He added, however, that the doses were estimated to avoid
under-estimating the number of radiation-related cancers
that might be expected to occur. |
