Among investigations providing the basis for radiation protection standards worldwide, the atomic-bomb survivor study is the most long-standing and extensive ever undertaken.

by Warren K Sinclair, president emeritus, US National Council on Radiation Protection and Measurements, and RERF visiting director

This article is abridged from a talk given at ceremonies held in June 1995 in Hiroshima and Nagasaki to mark RERF’s 20th anniversary. Attending the event were present and past employees as well as representatives of citizen’s groups, the local and national governments, RERF’s funding agencies, and the US National Academy of Sciences. This article originally appeared in RERF Update 8(1):6-8, 1996.

---

We are here this afternoon to celebrate the 20th year since the Radiation Effects Research Foundation (RERF) was founded, thus beginning the unique Japanese-American collaboration that succeeded 27 years of atomic-bomb survivor studies conducted by the Atomic Bomb Casualty Commission (ABCC). These studies have led to what we now recognize, worldwide, as the most informative findings on the delayed effects of ionizing radiation on man ever obtained.

‘Discovery’ of radiation

This year is 1995, which marks a century of man’s awareness of ionizing radiation as a factor in his life. Röntgen discovered x rays in 1895 and put them to manifold use at once, including in medicine. In 1896, Becquerel discovered the radioactivity of uranium, which was followed by the discovery of other radioactive substances, such as radium by Marie Curie. Obviously, radioactivity and ionizing radiation have been present naturally in the universe since the beginning of time, but man finally became aware of them just 100 years ago. The discovery of radioactivity opened the field of nuclear physics and quickly led Rutherford and others to an understanding of the structure and composition of the atom and, indeed, of nuclei themselves.

Today, we recognize that radiation is an inevitable part of our lives and that natural background radiation is ubiquitous on Earth. It consists of radon exposure, external terrestrial exposure, internal radionuclides, and cosmic radiation, which result in a dose to the individual of about 3 mSv per year (NCRP Report 93, Bethesda, Maryland, USA, 1987). Assuming linearity at low doses, the risk of cancer associated with natural background may be of the order of 1%, ie, about 1/20th of the cancer risk due to all other causes. In the US, individuals receive on average about another 0.6 mSv per year from manmade sources, mainly medical.

 

Putting radiation to work

Man also has put ionizing radiation to a broad range of uses, including industrial, agricultural, medical, and nuclear power, thus providing the opportunity for further exposure to manmade sources (International Atomic Energy Agency, Highlights of Activities, IAEA, Vienna, 1993). With these uses–some involving large amounts of radiation and radioactivity–comes the inevitability of accidents. However, in spite of widespread use, comparatively few fatal accidents have occurred in radiation work. By 1995, a total of 389 accidents had taken place–some 3,000 persons exposed significantly and 112 fatalities (not counting possible later cancer deaths) (Radiation Emergency Assistance Center/Training Site Accident Registries, Oak Ridge Institute for Science and Education, Oak Ridge, Tennessee, USA, 1995).

These activities occupationally expose radiation workers (about 4 million worldwide) to an average of about 2 mSv per year, which is about equal to a doubling of natural background radiation except for radon. This effectively doubles the radiogenic component of their risk.

The good news about radiation protection is that in the US, while the number of workers has grown from about 500,000 persons to 2,000,000 between 1960 and 1990, the average dose of those exposed has decreased by more than a factor of 2 because of the “as low as reasonably achievable” philosophy and other radiation protection pressures. In the US nuclear industry, the decrease has been more rapid and greater–more than a factor of 3 (from over 6 mSv/y on average in 1980 to under 2 mSv/y on average in 1990).

Except for accidents involving high doses, radiation protection today is not concerned with direct deterministic effects–erythema, cataract, sterility–because the low doses involved in most occupational settings are below the thresholds for these effects. However, stochastic effects–cancer and genetic effects–may be caused at low frequency after low-dose exposures. The major radiation protection question is: “What is the frequency (or risk) of cancer after a specific low dose of ionizing radiation?”

Source of most radiation risk information: Atomic-bomb survivors

All of our information on these risks comes from exposed human populations, of which by far the most important are the survivors of the atomic bombings at Hiroshima and Nagasaki, although some medically exposed populations and some occupationally exposed populations provide important complementary information.

Cancer induction after ionizing radiation is, of course, a delayed effect. Leukemia has a minimum latency of two years and a peak incidence at about six to eight years. Solid tumors have a minimum latency of about five to ten years, and thereafter incidence rises as the natural cancer rate increases with age at least up to 40 years after exposure.

Internationally, the responsibility of informing the world community on risk estimates for induced cancer is undertaken by the United Nations Scientific Committee on the Effects of Atomic Radiation (UNSCEAR) and, in the US, by the Advisory Committee on the Biological Effects of Ionizing Radiation (BEIR) of the US National Academy of Sciences (NAS). These bodies consider the relevant worldwide data from exposed populations and estimate the risk. Internationally, the International Commission on Radiological Protection (ICRP) and, in the US, the National Council on Radiation Protection and Measurements (NCRP)–informal professional bodies–use these risk estimates as the basis for recommendations on limits of exposure for workers and the public. Governments usually frame their radiation safety legislation on the ICRP and NCRP recommendations.

I have been personally involved with UNSCEAR from the 1977 report onwards and with the ICRP since 1977 (especially during the drafting of the ICRP 1990 recommendations), and I also have been president of the NCRP through most of that period. Consequently, I have considered these radiation risks from both international and national viewpoints.

The largely US-funded ABCC and, more recently, the RERF have assessed the epidemiological data on excess cancer deaths in their program approximately every four years beginning in 1961. These data have formed the basis of input to UNSCEAR and BEIR periodically. It warrants noting that a risk estimate is actually a risk coefficient, ie, the number of excess cancers per unit population divided by the dose causing the excess. Thus, the dose also is important and also has been evaluated carefully at intervals for the RERF program, stimulated by both NCRP and NAS committees. The latest revision, known as Dosimetry System 1986, was approved for use at RERF by both US and Japanese national dosimetry committees.
The extent of the cancer risk information available to UNSCEAR has increased over time (Table 1), derived mainly from the ABCC-RERF Life Span Study (LSS) but supported by some other studies. Information has increased through the years so that a single value for leukemia risk in 1958 has evolved into individual risks for about ten organs and a “remainder” in 1994. The estimated lifetime risk (high dose rate) for all cancers was about 10%-12%/Sv in the 1988 and 1994 reports. I believe that will not differ greatly when LSS Report 12, which incorporates data up through 1990, is completed and published in 1996 [Pierce et al, Radiation Research 146:1-27, 1996]. BEIR committee evaluations follow a pattern similar to those of UNSCEAR and generally have found similar risks.

 

Why is the RERF Life Span Study more important than any other radiation-related study?

A summary of the number of cancer deaths altogether as of 1985 indicates 339 excess cancer deaths among about 6,000 cancer deaths. This is not a large number statistically, especially when it is broken down into individual cancer sites.

Nevertheless, the LSS, because of the number of persons involved and the range of doses received (up to high doses), has more power than any other study to do the following:

• to produce risk estimates for total cancer, mortality, and incidence;
• to produce risk estimates for 10-20 individual organs;
• to demonstrate the shape of the dose response;
• to find the lowest doses for which there are statistically significant risks (LSS Report 11: 0.2 Sv; LSS Report 12: 0.05 Sv);
• to examine the effect of variables such as age and sex;
• to follow the fate of the youngest cohorts: 0-9 years old and 10-19 years old at the time of the bombings;
• to demonstrate latency and whether the risk of solid tumors decreases with time; and
• to demonstrate cancer risks in sensitive groups such as fetuses.

We are now beginning to see in the early and imprecise results of occupational studies (from the US, UK, and Russia) risk data that are similar to those from the LSS as interpreted by the ICRP (UNSCEAR, 1994). This is important confirmation. 

It is not only for total risk estimation that the LSS is so powerful. The 1994 UNSCEAR report also considered risks by individual site from all sources as completely as possible. For some sites, such as the breast, the number of sources is extensive. The average risk derived from all studies is about the same as the LSS value, the standard against which all other studies are measured.

The RERF LSS not only has all this power with respect to cancer induction but it also yields information on noncancer effects as well, including acute effects, mental retardation among those exposed in utero, delayed noncancer effects, and genetic effects. Note that here I am focusing on the human effects studies arising from the LSS program. I am not addressing the entire RERF research program which includes the important radiobiological and other studies that have always been a part of the program.

 

Applications of the cancer mortality risk data

The RERF LSS risk data are invariably the standard for risk estimates and form the basis for the following:

• standards for radiation workers;
• standards for the public;
• probability of causation assessment in a wide variety of circumstances;
• assessing the impact of accidents;
• assessing the risk to the public of environmental exposures;
• assessing the risk to soldiers and others exposed to tests, etc; and
• assessing effects on special groups, eg, fetuses (mental retardation, etc).

Probably the most important single application is the underlying basis of low-dose radiation protection standards. The change in risk estimates based on RERF LSS data collected up through 1988 caused the ICRP (and NCRP) to lower limits for workers in 1990 for the first time in over 30 years from an average of 50 mSv/y to an average of 20 mSv/y. Limits for the public were similarly lowered.

 

The Future

For the data up through 1985, only 39% of the LSS population had died and been evaluated. For LSS Report 12, which covers the data up through 1990, the figure is about 44%. The change in percentages expected in the future can be calculated easily from the characteristics of the Japanese population (in 1990).

The most important groups, about which we so far know little, are the 0- to 9-year-old group and the 10- to 19-year-old group. The population demographics with respect to these groups are shown in Table 2. In 1990, only 6% and 14% of these two groups had died and could be evaluated. By the year 2010, 44% of the 10- to 19-year-old group are projected to have been evaluated, as compared with only 20% of the youngest (0- to 9-year-old) group. In my view, the epidemiological and statistical study must continue at least until that time and in modified form perhaps even longer.

 

Long-term projects require enormous patience on the part of everyone concerned: the study subjects, the scientists, the managers, and the funding sources, ie, the Japanese Ministry of Health and Welfare and the US Department of Energy. The latter especially has had severe funding pressures mainly, but not solely, because of fluctuations in the yen-dollar exchange rate. In spite of these difficulties, the RERF program must continue strongly for the sake of future world knowledge.

It is fitting, in this 20th year, to quote James Liverman, who, in 1975 as assistant administrator of DOE’s predecessor (the Energy Research and Development Agency), was instrumental on the US side in bringing about this unique binational research foundation. “I regard the binational RERF as a great move forward in Japanese and American science. . . . I am very proud of it,” he said in June 1995.

In closing, let me say that the RERF staff has not only made a great success of this ongoing binational project but it has contributed enormously and uniquely to the worldwide knowledge of radiation-induced cancer and to many other delayed effects as well. Without you and the splendid cooperation of the atomic-bomb survivors, the world scientific community would still be floundering on this important subject. However, we still have much to learn, especially about the youngest groups of survivors.

Thank you for what you have done so far. I wish you Godspeed in what you will do for the world scientific community tomorrow.