. It is commonly stated that “any radiation dose, no matter how small, can cause cancer”. The basis for that statement is the linear-no threshold theory (LNT) of radiation carcinogenesis. According to LNT, if 1 Gy (100 rads) of exposure gives a cancer risk R, the risk from 0.01 Gy (1 rad) of exposure is R/100, the risk from 0.00001 Gy (1 millirad) is R/100,000, and so on. Thus the cancer risk is not zero regardless of how small the exposure.
However, over the past several years, a strong sentiment has developed in the community of radiation health scientists to regard risk estimates in the low-dose region based on LNT as being grossly exaggerated or completely negligible. For example, the 6000 member Health Physics Society, the principal organization for radiation protection scientists, issued a position paper (HPS 1996) stating “Below 10 rad ….risks of health effects are either too small to be observed or are non-existent”. A similar position statement was issued by American Nuclear Society. When the Health Physics Society Newsletter asked for submission of comments on validity of LNT, there were about 20 negative comments submitted and only a single comment supportive of LNT. In a worldwide poll conducted by the principal on-line discussion group of radiation protection professionals (RADSAFE), the vote was 118 to 12 against LNT A 2001 Report by the French Academy of Medicine concluded that LNT is “without any scientific validity, and an elaborate joint study by the French Academy of Medicine and the French Academy of Sciences (Aurengo et al 2005) strongly condemned the use of LNT. While U.S. official agencies have been slower to accept this position, the U.S, National Council on Radiation Protection and Measurements (NCRP) stated in NCRP Publication No. 121 (NCRP 1995) “Few experimental studies and essentially no human data can be said to prove or even provide direct support for the [LNT] concept”, and in NCRP Publication No.136 (NCRP-2001) stated “It is important to note that the rates of cancer in most populations exposed to low level radiation have not been found to be detectably increased, and in most cases the rates appear to be decreased”. A group of scientists opposing use of LNT (Radiation Science and Health) submitted several hundred papers supporting their position to National Research Council.
Beyond failure of LNT, there is substantial evidence that low level radiation may be protective against cancer; a view known as “hormesis”. There is an International Hormesis Society which sponsors an annual International Scientific Conference and publishes a peer reviewed scientific journal and a regular newsletter
The purpose of this paper is to review the basis for LNT and to present some of the mostly recent information that has caused this strong shift in sentiment. Other recent reviews have been published with somewhat different approaches to similar objectives (Feinendegen 2005, 2005a, Tubiana 2005).
PROBLEMS WITH THE BASIS FOR LINEAR-NO THRESHOLD THEORY
The original basis for linear-no threshold theory (LNT), as that theory emerged in the mid-twentieth century, was theoretical and very simple. A single particle of radiation hitting a single DNA molecule in a single cell nucleus of the human body can initiate a cancer. The probability of such a cancer initiation is therefore proportional to the number of such hits, which is proportional to the number of particles of radiation, which is proportional to the dose. Thus the risk is proportional to the dose – this is linear-no threshold theory.
An important problem with this simple argument is that factors other than initiating events affect the cancer risk. Human bodies have biological defense mechanisms which prevent the vast majority of initiating events from developing into a fatal cancer (Pollycove 2001). A list of some of the most important examples including how they are affected by low level radiation follows (Feinendegen 2005):
--- Our bodies produce repair enzymes which repair DNA damage with high efficiency, and low level radiation stimulates production of these repair enzymes.
---Apoptosis, a process by which damaged cells “commit suicide” to avoid extending the effects of the damage, is stimulated by low level radiation. A similar effect is achieved by premature differentiation and maturation to senescence.
---The immune system is important for preventing mutations from developing into a cancer; there is abundant evidence that low level radiation stimulates the immune system, but high radiation levels depress it.
---The overwhelmingly most important cause of DNA damage is corrosive chemicals (reactive oxygen species -ROS); there are processes for scavenging these out of cells, and low level radiation stimulates these scavenging processes (Kondo 1993). Elevated ROS levels have been shown to initiate a broad array of biochemical reactions that are stress responses, leading to the conclusion that “the best protection against stress is stress itself” (Finkel 2000)..
---Radiation can alter cell cycle timing. This can extend the time before the next cell division (mitosis). Damage repair is most effective before the next mitosis, so changing this available time can be important (Elkind M, personal communication).Altered cell timing can also affect DNA repair processes in many ways by changing chemical processes (Boothman 1996)
---Various other effects of low level radiation on cell survival have been observed and are referred to as “low dose hypersensitivity”, “increased radiation radioresistance”, and “death inducing effects” (Bonner 2004)
It is now recognized that development of cancer is a much more complex process than was originally envisioned. The role of “bystander effects”, signaling between neighboring cells relevant to their radiation experiences, is now recognized to be an important, albeit poorly understood, factor In fact it seems that tissue response, and even whole organ response, rather than just cellular response, must be considered (Aurengo et al 2005) .
There is also apparently obvious evidence for failure of the original simple model. For example, the number of initiating events is roughly proportional to the mass of the animal – more DNA targets mean more hits. Thus the simple theory predicts that the cancer risk should be approximately proportional to the mass of the animal. But the cancer risk in a given radiation field is similar for a 30 g mouse and a 70,000 g human. As another example, our very definition of dose, based on the energy absorbed per unit mass of tissue, which is proportional to the number of radiation hits per unit target mass, would be misleading if only the total number of hits (which is proportional to the number of initiating events) were relevant regardless of the target mass.
A detailed theoretical approach to evaluating the validity of LNT is based on the commonly accepted idea that double strand breaks (DSB) in DNA molecules are the principal initiating event in causing cancer. But DSB are also caused by endogenous corrosive chemicals, reactive oxygen species (ROS).In fact the DNA damage caused by radiation is mostly due to the production of ROS by the ionizing effects of the radiation on omnipresent water. It is estimated that endogenous ROS causes about 0.1 DSB per cell per day, whereas 100 mSv (10 rem) of radiation, which is close to the upper limit of what is normally called low level radiation, causes about 4 DSB per cell (Feinendegen 2005). Assuming that the number of cancers is proportional to the number of DSB, a 100 mSv dose of radiation would increase the lifetime (28,000 days x 0.1 DSB/day) risk of cancer by only about (4 / 2800=) 0.14%, whereas LNT predicts an increase of 1%. From this it is concluded that the underlying assumption of LNT that cancer initiating events are the controlling factor in determining the dose-response relationship for radiation is a serious over-simplification
DIRECT EXPERIMENTAL CHALLENGES TO THE BASIS FOR LNT A direct demonstration of the failure of the basis for LNT derives from microarray studies determining what genes are up regulated and down regulated by radiation.. It is found that generally different sets of genes are affected by low level radiation than by a high level dose. For example, in one study of mouse brain (Yin et al 2003), 191 genes were affected by a dose of 0.1 Sv but not by a dose of 2.0 Sv, 213 genes were affected by 2.0 Sv but not by 0.1 Sv, while 299 genes were affected by both doses. The 0.1 Sv dose induced expression of genes involved in protective and repair functions while down-modulating genes involved in unrelated processes.
A similar study with even lower doses on human fibroblast cells (Golder-Novoselsky et al 2002) found that a dose of 0.02 Sv caused more than 100 genes to change their expression, and these were generally different than the genes affected by 0.5 Sv. The former group was heavily weighted by stress response genes
Several other microarray studies have shown that high radiation doses which serve as the calibration for application of LNT, are not equivalent to an accumulation of low radiation doses (Tubiana 2005).
Sophisticated experimental techniques have been developed for observing the effects of a single alpha particle hitting a single cell. It was found (Miller 1999) that the probability for transformation to malignancy from N particle hits on a cell is much greater than N times the probability for transformation to malignancy from a single hit. This is a direct violation of linear-no threshold theory, indicating that estimated effects based on extrapolating the risk from high exposure, represented by N hits, greatly exaggerates the risk from low level exposure as represented by a single hit.
A very clear demonstration of a threshold response, in contrast to LNT, was found in tumor induction by irradiation throughout life of mouse skin (Tanooka 2001). For irradiation rates of 1.5 Gy/week, 2.2 Gy/week, and 3 Gy/week, the percentage of mice that developed tumors was 0%, 35%, and 100% respectively..