Part 1. The Latency Period
The breast-cancers induced in a population, by the radiation received in a single year, are spread out over many years. Some of the breast-cancers induced in a population by ionizing radiation become clinically evident very soon after exposure, but most of them become clinically evident 10, 20, 30, 40, 50 (and probably more) years later.
Women developing breast-cancer today in their 20s, 30s, and 40s, for example, have to consider the possibility that their breasts received radiation during infancy or childhood. In such cases, the radiation exposure would be a prime suspect for causation, especially if they have an inherited predisposition to cancer. Chapter 3 deals with recent evidence for radiation induction of early-onset breast-cancer.
Many more women, developing breast-cancer in their 50s, 60s, 70s, and 80s, may have "paid the price" for radiation possibly received decades earlier --- as early as the 1910s, 1920s, 1930s, 1940s and 1950s. Of course, it will be impossible for many women of such ages (if they are still alive) to find out if they are the individuals who received breast-irradiation so long ago. This book is likely to astonish them with the number of ways in which it could have happened to them when they were too young to remember it --- during their infancy and childhood.
The time-delay, between the radiation exposures and the diagnosis of clinical cancer, is called the "incubation period" or the "latency period."
We know for certain, from the study of the Atomic-Bomb Survivors in Japan, that the latency period varies enormously among individuals. Is this variation explained by differences in age, at the time of irradiation? Not fully. Among survivors who received their radiation exposure in August 1945, the latency period sometimes varies by decades even among survivors who were approximately the same age in 1945. The A-Bomb Study tells us beyond any doubt that each exposure to extra radiation has its total cancer-consequences spread out over many years. (Such evidence is shown in Gofman, Table 17-B, for example.)
Is there a limit to the latency period? Yes. It can not possibly be greater than the remaining lifespan of the irradiated group's most enduring member. Thus, the range of the latency period can be greater for a group irradiated during infancy and childhood, than for a group irradiated at age 60.
Part 2. Latency and Today's Breast-Cancer Problem
The "spread-out" nature of radiation-induced cancer has a very important implication for today's breast-cancer problem. If we want to learn what fraction of recent, current, and future breast-cancer is due to medical (and other) irradiation --- and is therefore preventable --- our evaluation has to begin as early as 1920.
For the sake of illustration, let us consider the 1940s. If a large female population of mixed ages received breast irradiation in 1940, 1941, 1942, 1943, 1944, etc., the breast-cancer rates observed in 1950, in 1960, in 1970, in 1980, in 1990, and very probably in 2000 and 2010 also, will include some breast-cancer cases produced by irradiation in each of those years of the 1940s.
The maximum duration of the cancer latency-period is not yet known. The existing human evidence is reviewed and discussed in Chapter 17 of our 1990 book. The results from the two longest, largest human studies are explored in detail. They are the study of the A-Bomb Survivors and the study of adult patients irradiated for relief of the pain from ankylosing spondylitis (a serious disease of the spine).
The A-Bomb Study is particularly important, because it includes females irradiated at all ages, including children below age 10 --- who have the greatest risk from irradiation. The follow-up evidence through 1985 on cancer deaths shows that the radiation effect is certainly not "over" (finished, gone) in 1985 for people irradiated by the bombs 40 years earlier.
No study yet exists which includes the full lifespan of younger participants. For example, A-Bomb Survivors who were under age 10 in 1945 are under age 60 in 1995. Therefore, no one knows whether radiation's carcinogenic effect will endure for the entire lifespan of the youngest irradiated group, or not. Many experts today assume that it will. We, too, use that assumption in this book. Only future follow-ups of the A-Bomb Survivors will reveal this part of the "story" for certain. The A-Bomb Study is a biomedical resource of unique value.
Specifically Breast Cancer: What We Know about Latency Now
In 1994, the report entitled "Incidence of Female Breast Cancer among Atomic Bomb Survivors, 1950-1985" was published. We will refer to it in later chapters also. Here, our focus is on what it says about a key question of latency:
Can breast irradiation during infancy and childhood really cause radiation-induced breast-cancer in adulthood?
Some physicians seem incredulous. Nonetheless, the answer is "Yes." During the 1970s, the idea was barely accepted --- even as a possibility --- by segments of the radiation community. In 1981, when the evidence was still thin, we predicted that the answer would turn out to be "Yes" (Gofman, 1981, Chapter 7). The subsequent evidence for "Yes" is so strong that it elicits an emphatic statement (below) from an ordinarily cautious source: Analysts at the Radiation Effects Research Foundation, known as RERF. Funded by the governments of the United States and Japan, RERF controls all the databases of the A-bomb survivors. The 1994 report on breast-cancer incidence (just cited above) was done by RERF analysts: Masayoshi Tokunaga, Charles Land, and four more. Near the paper's end, citing their own data and the data of Hildreth 1989, these analysts state (Tokunaga 1994, p.220):
"At this point, there can be little doubt that radiation exposure of breast tissue during early childhood and infancy can contribute to the risk of breast cancer during adult life," (When irradiation causes an elevated cancer-risk for an individual, it causes an elevated cancer-rate for a group.)
Part 3. Why Does the Latency Period Vary among Individuals?
We are often asked an intelligent double-question: "Why does the incubation time vary, and why does it take so long for cancer to become manifest clinically after exposure to a carcinogen such as ionizing radiation?"
The answer to this question has more than one part.
First, we must emphasize that there is no evidence for thinking that all cases take "so long." For example, if a thousand cancer cases are "committed" to occur sooner or later, due to genetic lesions induced by ionizing radiation received by a group of people in 1940, some of the cases probably occur by 1945. Some may even occur immediately.
Why do we say "probably" and "may"? Why can't we say for certain? The answer lies in something called "the small numbers problem." The cases committed in 1940 and occurring by 1945 look clinically just like all the other cancers occurring in the years 1940 through 1944. The presence of the cases committed in 1940, therefore, has to be detected by epidemiology. But if the accumulated number of radiation-induced cases (committed in 1940 and full-blown by 1945) is still very small in 1945, relative to the number of cases occurring anyway, their presence can not be proven statistically even if they are really there. This is the common --- and dreaded --- "small numbers problem" in epidemiology.
"Last Straws," "Early Straws," and Luck
One reason for thinking that some cases may occur immediately derives from the multi-step genetic model of cancer development (Chapter 1). Radiation-induced cases could develop very quickly in individuals who have already accumulated several carcinogenic lesions in a vulnerable cell. For such a cell, a radiation-induced lesion could be the "last straw which broke the camel's back." And the latency period could be extremely short.
But compared with "last straws" from radiation exposure, "early straws" must occur more frequently. Individuals with "early straws" from a radiation exposure will develop cancer only after a cell accumulates additional carcinogenic lesions, so their cancers are likely to show a longer latency period than "last straw" cancers. The multi-step genetic model goes a long way toward explaining the variation in the latency period.
It is important to emphasize that only some irradiated individuals with "early straws" ever develop a full set of cancer-lesions, and that not everyone who is irradiated by the same dose of radiation develops cancer later. Because luck plays a role, it is not possible to predict which individuals will develop the extra cancers when a group of people receives radiation. Nonetheless, we (and others) can predict the total number of individuals who will develop radiation-induced cancer from a known amount of radiation exposure, because we have real-world evidence patiently accumulated by numerous researchers over many decades.
How the "Host" of a Cancer Determines Its Latency Period
The length of the latency period almost certainly depends on the "host" of a radiation-induced cancer. As hosts, people vary in the number and strength of carcinogenic lesions which are inherited, and individuals vary in the number and strength of carcinogenic lesions which are induced by various non-inherited factors, both before and after a specific exposure to ionizing radiation. So it would be really surprising if latency did not vary from person to person, following a specific radiation exposure.
Ionizing radiation is known to induce a great variety of genetic lesions, most especially chromosomal abnormalities. Because chromosomes are the structures produced by the winding of the DNA around specific proteins, chromosomal lesions are genetic lesions. It is abundantly clear now that cancer cells generally show chromosome abnormalities of structure, or number, or both. Moreover, the variety of the abnormalities is also abundantly clear. It is very reasonable to expect that the speed of cancer development varies with the particular chromosomal damage which is present in a cell. That is one aspect of saying that the "strength" of carcinogenic lesions varies.
Part 4. The Radiation Effect: "Most Cases" vs. "Early-Onset Cases"
About 22 percent of all deaths per year in the USA are due to cancer. As everyone knows, cancer rates generally increase with advancing age. There is a special set of childhood cancers, but they are fortunately rare. As for leukemia in the USA, about 5 percent of it occurs before age 20, so it is not predominantly a childhood malignancy.
What we see for cancer in general (with some exceptions) is a rising rate of new "spontaneous" cases per 10,000 adults, as they advance in age. Some of the so-called spontaneous cases are really radiation-induced cases. Nonetheless, in epidemiologic studies of radiation-induced cancer, analysts are observing the consequence when some people have been exposed to extra radiation, while a comparable control-group has not been exposed to the extra dose.
The observed effect of extra radiation is to cause the spontaneous rates ("background" rates) of cancer to increase, by percentages which depend upon the amount of extra radiation exposure and on the age at which the extra exposure occurs. Most of the spontaneous cancer-cases appear during middle-age and beyond, and so do most of the cases induced by extra exposure to ionizing radiation.
Most cases, but certainly not all.
There is very important evidence from the A-Bomb Study of radiation-induced early-onset breast-cancer --- cases occurring before age 35 --- in women who were under age 20 at the time of the bombings. That is the subject of Chapter 3.
# # # # #