---

II. TABLE OF CONTENTS

---

Each and every cell in the human body is constantly engaged in a life and death struggle to survive "in spite of itself." Normal physiological processes needed for cell survival generate toxic oxidative products that are damaging, even mutagenic, and potentially carcinogenic. Yet cells and people survive because of the cell's remarkable capacity to repair the majority, if not all, of this oxidative damage. We don't know, however, the relationship between this normal oxidative damage and the high frequency of cancers that exist in all human populations. Is cancer a price we pay for the very biological processes that keep us alive?

We are also constantly exposed to low levels of natural background radiation from cosmic radiation and from naturally occurring radioactive materials in air, soils, water, and even living things. Research has taught us that while even low levels of radiation induce biological damage, the damage is similar to the oxidative damage induced by normal cellular processes. Thus a critical, yet unanswered, question in radiobiology is whether the biological damage induced by low doses and low dose rates of radiation is repaired by the same cellular processes and with the same efficiency as normal oxidative damage that is the way of life for every living cell.

This Program Plan will outline a research strategy to determine if low dose and low dose-rate radiation presents a health risk to people that is the same as or greater than the health risk resulting from the oxidative by-products of normal physiological processes. This information is essential for future decisions made to protect people from adverse health risks from exposure to radiation.

Extensive research on the health effects of radiation using standard epidemiological and toxicological approaches has been used for decades to characterize responses of populations and individuals to high radiation doses, and to set exposure standards to protect both the public and the workforce. These standards were set by using modeling approaches to extrapolate from the cancers observed following exposure to high doses of radiation to predicted but unmeasureable changes in cancer frequency at low radiation doses. The use of models was necessary because of our inability to detect changes in cancer incidence following low doses of radiation. Historically, the predominant approach has been the Linear-no-Threshold model which assumes that each unit of radiation, no matter how small, can cause cancer. As a result, radiation-induced cancers are predicted from low doses of radiation for which it has not been possible to directly demonstrate cancer induction.

Most of the projected radiation exposures associated with human activity over the next 100 years will be to low dose and low dose-rate radiation from medical tests, waste clean-up, and materials associated with nuclear weapons and nuclear power production that remain in the environment. The major type of non-medical radiation exposures will be low Linear Energy Transfer (LET) ionizing radiation (primarily X-and gamma-radiation) from fission products. Thus, the DOE Low Dose Radiation Research Program will initially concentrate on studies of low-LET exposures delivered at low total doses and low dose-rates. As the program develops, investigation of low dose high-LET radiation will be included as it becomes appropriate.

The overriding goal of this program is to ensure that human health is adequately and appropriately protected. It currently costs billions of dollars to protect workers and the public from exposure to man-made radiation, often at exposure levels lower than the natural background levels of radiation. If it could be demonstrated that there is no increased risk associated with these exposures, significant savings could be realized in risk management programs.

To address the effects of very low levels of exposure to ionizing radiation, the research program will build on advances in modern molecular biology and instrumentation, not available during the previous 50 years of radiation biology research. It will concentrate on understanding the relationships that exist between normal endogenous processes that deal with oxidative damage and processes responsible for the detection and repair of low levels of radiation-induced damage. Research will focus on understanding cellular processes responsible for recognizing and repairing normal oxidative damage and radiation-induced damage. If the damage and repair induced by low dose radiation is qualitatively the same as for normal oxidative damage, it is possible that there are levels of damage that the body can handle. In contrast, if the damage from ionizing radiation is qualitatively different from normal oxidative damage, then its repair, and the hazard associated with it, may be unique. To understand the relationship between normal oxidative damage and radiation-induced damage, studies will be conducted at very low doses and dose-rates and the perturbation of the normal physiological processes will be characterized at all levels of biological organization -from genes to cells to tissues to organisms. Research needs are identified in five interrelated areas:

This research program plan is intended to serve as an outline of research needs and challenges for research scientists, DOE program staff, and regulators. It is not intended to identify the specific experimental strategies needed to solve the challenge of understanding the health risk from exposure to low doses of radiation. Specific research strategies should be developed and proposed by individual scientists using this research plan as a guide and with input from DOE program staff and regulators.

Research conducted in this program will be used to make better estimates of health risks from exposures to low levels of radiation, information that is critical to adequately and appropriately protect people and to make the most effective use of our national resources.

Summary of Funding Needs

Return to the top

---

IV. INTRODUCTION

Estimates of cancer risks following exposure to ionizing radiation are based on epidemiological studies of exposed human populations, principally the Japanese atomic bomb survivors. While analyses of these populations provide relatively reliable estimates of risks for high dose and high dose rate exposures, it is the effects of low doses and low dose rates that present the greatest health concerns for radiation workers and the general population today. The risks of cancer and mutations produced by very low doses remain a critical unresolved issue because they cannot be directly measured in exposed populations. Conceptually, we are forced to estimate risks for low-doses and for doses received as chronic protracted exposures or low dose fractionated exposures by applying various dose response models to available high dose data.

Currently, overall estimates of low dose risks are based on empirical linear fits of existing human data from relatively high dose exposures that have been adjusted for low-dose and low dose-rate exposures. This approach has generally been adopted by those responsible for assessing radiation risks. 4, 9 However, some have argued that this approach is inappropriate, greatly overestimating cancer risks. Among those who believe that current protection standards overestimate risks, many argue that a threshold for radiation-induced cancer exists. This is a critical, but as yet unresolved, issue because of the potential societal and economic impact of decisions upon which these estimates of risk are based. Epidemiological data by themselves are not capable of resolving the critical questions at hand; moreover, conventional radiation biology experimental approaches have gone as far as they can toward addressing low dose issues.

Through recent advances in cell and molecular biology and concomitant advances in chemical and biological technology, scientists have now created an extraordinary opportunity to definitively resolve this critical low dose issue. Research to decode the genome, to understand structure-function relationships for genes and proteins, and to apply molecular biology to medical problems has resulted in the development of new scientific resources and technologies. These can be modified and applied to basic problems in radiation biology. In association with the development of instrumentation, there has been an explosion of knowledge in the fields of molecular and cellular biology. For example, it is now possible to identify the genetic basis of many diseases, to clone and amplify individual genes, to grow a wide range of critical cell types associated with cancer, and to develop transgenic animal models. All these techniques help us understand and modify the expression and action of many genes. With new molecular techniques and the proper application of instrumentation, it will be possible to increase understanding of normal processes that repair oxidative and radiation-induced damage at the molecular, cellular and tissue levels, to evaluate molecular processes that modify the expression of these changes during cancer development, and to determine the role of low levels of radiation in these processes.

Over the last several years it has become clear that oxidative free radicals produced by normal cellular metabolism are involved in the production of endogenous DNA damage. The types of damages produced by these free radicals overlap with the majority of molecular damage produced by ionizing radiation. Cellular DNA repair mechanisms, that are highly conserved across species, evolved to remove these endogenous oxidative DNA damages and thus preserve genomic integrity. It is precisely because free radical-induced DNA damages are efficiently repaired that cells have low rates of spontaneous mutation. This raises two critical questions. Does low level ionizing radiation induce damage that can be efficiently repaired by the same or similar repair systems as endogenous damage? If so, does this result in a threshold for adverse effects induced by low doses of radiation?

There is ample evidence that DNA repair competence can influence radiation effects, including radiation-induced cancer. There is also accumulating evidence that even low doses of radiation can elicit numerous molecular responses that have the potential to influence the consequences of those exposures. Thus, a suggestive, but unproven, case can be made supporting the view that a threshold may exist at low doses of radiation. With the continuing development of sophisticated molecular biological approaches, together with new and evolving chemical and biophysical techniques, it is now possible to readdress the low-dose issue, including the likelihood of a threshold.

Coupled with advances in biological research, new technologies will have to be advanced, including new approaches to measure cellular damage following very low dose exposures and to determine molecular responses to that damage at the level of single genes or for small changes in gene expression. Much of this technology development will be facilitated by interactions with other ongoing programs such as the human genome and structural biology programs.

Recent epidemiological and genetic studies suggest there may be a large number of genetic polymorphisms in the human population. The potential of these polymorphisms to change the risk for cancer as a result of interactions with environmental factors, including low doses of radiation, is yet to be established and is a major thrust of this program. If the frequencies of polymorphisms that impact susceptibility to radiation-induced cancer are relatively high, they could significantly impact risk estimates at low doses for the population in general. It is now possible to identify, map, and clone the genes involved in radiation damage response functions, define the polymorphic frequencies of these genes in the population and determine their importance for susceptibility. This will provide the opportunity to directly determine their impact on cancer risk estimates after exposure to radiation. This effort will also be facilitated by interactions with the human genome program.

The focus of research in the Low Dose Radiation Research Program should be on doses of low linear energy transfer (LET) radiation that are at or below current workplace exposure limits. In general, research in this program should focus on total radiation doses that are less than or equal to 10 rads. Some experiments will likely involve selected exposures to higher doses of radiation for comparisons with previous experiments or for determining the validity of extrapolation methods previously used to estimate the effects of low doses of radiation from observations made at high doses.

Return to the top

---

V. PROGRAM OUTLINE

A. Low Dose Radiation vs. Endogenous Oxidative Damage -The Same or Different?

1) Key Question: How is the DNA damage produced by low dose ionizing radiation qualitatively and/ or quantitatively different from normal oxidative damage?

2) Description

Over the last several years it has become clear that oxidative free radicals produced by normal cellular metabolism are involved in the production of endogenous DNA damage. The types of damages produced by these free radicals overlap with the majority of molecular damage produced by low dose, low LET ionizing radiation. The majority of damage produced by low LET ionizing radiation is due to the radiolysis of water in the vicinity of the DNA molecule, leading to free radical-induced DNA damages, much of which is similar to that produced by endogenous free radicals. These free radicals damage the DNA sugars and bases producing single strand DNA breaks, base loss, and a large number of modified DNA bases. A much smaller number of double strand DNA breaks are produced by direct ionization of DNA or, possibly, by the processing of multiple single lesions produced in close proximity. Protein-DNA cross-links are also formed, but in very low amounts.

In spite of the fact that the frequency of double strand breaks is much lower than that of other types of damage, double strand breaks may be the major determinant that distinguishes normal oxidative damage from low dose radiation induced damage to DNA. 19, 20 In mammalian cells, the double strand break is considered to be the primary lesion involved in cellular lethality, mutagenesis, and, perhaps more significantly in terms of cancer risk, the lesion that is more difficult for cells to accurately repair. Clustered DNA damage that, at least at high radiation doses, appears to be unique to ionizing radiation may be particularly difficult to repair. Free radical-induced lesions present on a single strand of DNA have not generally been implicated in cell death and carcinogenesis because they are readily repaired by the cell's base excision repair. Although the impact of unrepaired DNA damage to vital genes cannot be ignored, it is likely that subsequent misprocessing leading to misrepaired DNA damage is largely responsible for a number of biological effects including cancer.

3) Decision Making Value

We marvel at the differences in "metabolism" that exist between people not stopping to think that comparable differences in normal oxidative damage may exist between us. We live at high elevations like Denver or Salt Lake City (90 mrad/ year), at sea level (23 mrad/ year), and everywhere in between without realizing that there are 4-fold differences in natural cosmic background radiation that are simply dependent on elevation. "For significant sub-populations, the range of annual cosmic-ray dose equivalent exceeds an order of magnitude, i. e., from 150 to 5000 :Sv (15 to 500 mrem)." 10 In addition, the lung dose from radon in homes that contributes most of the natural radiation dose equivalent, varies between regions of the United States by more than an order of magnitude. 11 Research is needed to understand and quantify real, not calculated, differences or similarities in DNA damage induced by normal oxidative processes versus low doses or low dose rates of ionizing radiation in efforts to efficiently and effectively protect people from unnecessary and avoidable health risks. The problem facing scientists and policy makers today is that all the information for radiation-induced DNA damage is from information obtained at high doses; doses at which cells are traversed by multiple ionization tracks. There are simply no data at the low doses normally considered relevant to public health issues where a cell may only be traversed by a single electron track over a long period of time, e. g., one year. It is not difficult to imagine that the spectrum of damage at such low doses may be substantially different from that observed at high doses. Because the background of spontaneous damage from normal oxidative processes is fairly high, the question arises as to whether low levels of ionizing radiation actually make a significant addition to the background level of damage. 19 Thus, it is fundamental to the entire low dose issue to determine whether the amount and kinds of DNA damage produced at low doses of radiation are different from those normally produced within cells.

If the DNA damage produced by low doses of ionizing radiation is qualitatively similar to the damage produced by normal physiological processes then we can, as outlined below, determine if our normal damage defense mechanisms protect us from this additional damage. This could lead to a conclusion that the linear-no-threshold model is inappropriate for estimating health risks from low dose radiation. On the other hand, if low dose ionizing radiation produces unique types of damage not produced by normal oxidative processes and not removed by our damage defense mechanisms, then the linear-no-threshold model may be shown to be the most appropriate tools for estimating risk.

4) Recommendations and Costs.

Research is needed to understand and quantify real, not calculated, differences or similarities in DNA damage induced by normal oxidative processes versus low doses or low dose rates of ionizing radiation. This information is the foundation for the many aspects of the Low Dose Radiation Research Program. Although always needed, it was not previously attainable because critical resources and technologies were not available. Today, technologies and resources such as those developed as part of the human genome program, e. g., coupled capillary electrophoresis and mass spectrometry systems and DNA sequence information, have the potential to detect and characterize small differences in damage induced by normal oxidative processes and low doses of radiation. ⁷

Research is needed in two closely related and interdependent areas: technology development and basic research, including molecular dosimetry.

A significant investment in technology development will be required to expand current capabilities for delivering, identifying, and quantifying small amounts of oxidative or radiation-induced damage. Radically new technologies are likely not needed but current technologies will need to be modified. Methodologies having high sensitivity as well as high signal-to-noise ratios will be critical in this effort. A focused technology development effort consisting of two cycles of three-year grants should yield broadly useful and available methods for delivering and measuring small amounts and differences of oxidative damage in cells. An annual investment of approximately $4.2 million will be required for each of the first three years with approximately $1.6 million required for each of the next three years.

This program is ultimately intended to quantify risk to humans from exposures to low doses and dose rates of radiation. Before it can accomplish this goal it must define what is meant by low dose and dose rate. According to target theory, the "lowest" possible radiation dose is a single interaction between radiation-induced electrons and a target biomolecule. This interaction is postulated to set up a chain of events that is responsible for the biological effects. Of course the biological effects may require multiple electron interactions to trigger the events responsible for radiation-induced disease. An additional complication is that these initial interactions are, in turn, modulated by a variety of cell-, tissue-, or organism-level processes such as cell-cell communication, repair, and misrepair.

To understand human health effects induced by low doses of radiation it is necessary to study radiation-induced biological effects using a variety of recently developed techniques and technologies. A key aspect of this effort will be the definition of radiation dose at the molecular level. At the molecular level, the concept of dose, which is energy deposited per unit of mass, may not even be applicable. It may be necessary to define other measures of interaction between the energy deposited and the target molecule such as the number of electrons interacting with cells or tissues, their spatial and temporal energy distribution, the superposition of these events on the target molecule, and the number of targets that need to be traversed to initiate a biological response. This new type of approach may be needed to quantify "radiation dose" at the molecular level. To avoid confusion and to make the new concepts of dose acceptable to the operational health physicist, it is important that new molecular and cellular dose concepts be easily understood and related to conventional dose at the macro-scale, where energy per unit of mass can be easily used. This is essential so that the concepts of energy interaction used at the molecular level can be related to a conventional definition of dose, e. g., the Sievert, and used to estimate human health risks.

Similarly, a significant research effort will be required to characterize and quantify normal oxidative damage in cells and the incremental increases induced by low doses of ionizing radiation. We know that clusters of ionization produced by high doses of radiation result in complex DNA damage that is difficult to repair. We don't know if very low doses of low LET radiation, that is currently the focus of this research plan, have the same effect. Partnerships should be encouraged between laboratories involved in characterization and quantification of radiation and oxidative damage and groups with expertise in or developing new technology to facilitate progress in both areas simultaneously. An annual research investment of approximately $2.6 million will be required for each of the first three years increasing to approximately $4.2 million for the next three years as new technologies are developed and become more widely available. A critical goal of the research component of this program is to quantify levels of damage induced by normal oxidative processes and the incremental increases due to low dose radiation. Qualitative descriptions of differences and/ or similarities between the types of damage induced under both conditions are useful in the design and interpretation of experiments in other parts of the low dose radiation research program. However, to be most useful in risk models and for regulators these differences or similarities must be quantified.

Return to the top

B. Understanding biological responses to radiation and endogenous damage.

1) Key Question: Do molecular, cellular, or tissue responses to radiation modify the processing of damage from radiation and endogenous sources, mainly oxidative metabolism, and/ or determine if damaged cells are eliminated, inhibited, or expressed as cancer?

2) Description

Knowing the types of damage produced by low dose ionizing radiation and the differences and/ or similarities of that damage to normal oxidative damage are key first steps in understanding potential health risks from low dose radiation. Only by understanding these difference and/ or similarities can we determine if and how low doses of ionizing radiation affect cells, tissues, and people. However, it is the biological effects of this radiation-induced damage, not the damage itself, that determines the health risks to people. Thus, several questions need to be answered before we can accurately evaluate the health risks from exposure to low doses of ionizing radiation: Do the same things happen in cells, tissues, and people exposed to high and low doses of ionizing radiation? Do they happen the same way? In the end, the goal of the research described in this section of the Low Dose Radiation Research Plan is to determine if health risk is directly proportional to radiation dose regardless of the dose. Understanding the mechanisms of and the dose-effect relationships for the biological effects of low doses of ionizing radiation will provide the scientific basis in support of or against the existence of a threshold for adverse effects induced by low doses of ionizing radiation.

Cellular pathways for recognizing damage, for signaling information on damage throughout the cell and to other cells, and for responding to damage are key elements in damage repair and processing. While there has been a significant amount of research defining radiation- induced genes and radiation-induced stress responses in mammalian cells, the relative contribution of a particular inductive response to the cellular consequences, e. g., survival, apoptosis, cancer, has been examined in detail for only a few genes such as p53 or PKC. At low doses of radiation no relationships between radiation-induced responses and other oxidative stresses have yet been defined. Most radiation-induced gene changes reported to date are transient events, occurring at a specific time following exposure and then decreasing some time thereafter. The kinetics of these responses appear to vary with radiation dose, radiation quality, and cell type but systematic studies on specific radiation-induced responses have not been carried out. It must be determined which genes and proteins are specifically induced in response to low doses of ionizing radiation, how these relate to other oxidative stresses, and, importantly, how the induced proteins affect endpoints relevant to radiation-induced cancer.

In mammalian cells, the principal DNA repair pathways involved in the repair of ionizing radiation induced DNA damage are base excision repair and non-homologous end-rejoining. Base excision repair, which evolved to protect cells against endogenous damage, removes radiation-induced single DNA lesions, base damages, single strand breaks, and sites of base loss. Together these types of damage have been estimated to account for about 70% of radiation-induced DNA damage. 19 This simple DNA repair pathway is well understood and is highly homologous between bacteria and humans with many of the proteins involved exhibiting up to 40% identity. This pathway is relatively error free in most instances. Interestingly, a confounder specific to ionizing radiation is that multiple single lesions in DNA formed in close proximity to one another are recognized by the enzymes of the base excision repair pathway but their processing can result in a double strand break.

In contrast to the types of DNA damage described above, double strand breaks in mammalian cells are generally repaired by non-homologous end-rejoining. This type of repair does not require that the ends of the two recombining molecules have any sequence homology, i. e., ends of broken DNA molecules that don't belong together or that have pieces missing can actually be joined by this process. Although less well characterized than excision repair, this pathway is extremely important with respect to radiation effects. This is because radiation-induced double strand breaks, while lower in frequency than most other types of radiation-induced damage, are the major threat to the integrity of the genome because of the problems associated with their repair. Isolated mammalian cells and mice defective in components of this pathway are hypersensitive to the cytotoxic effects of ionizing radiation. Recent studies of cancer prone human populations have served to underscore the potential importance of this pathway.

Because of the nature of the damage, the non-homologous end-rejoining pathway may be more error prone. Thus, processing of DNA double strand breaks leads to mutations, chromosomal aberrations, and, perhaps, genomic instability (see below). 5 These consequences can also reveal important information relevant to the low dose question. For example, newer chromosome painting techniques have revealed that an unexpectedly large proportion of radiation-induced chromosome aberrations is due to exchanges requiring multiple breaks and involving multiple chromosomes. 14 Less sensitive techniques had previously indicated that such rearrangements appeared to be simple exchange events between chromosomes. These newer results present a clear challenge to current theories including key aspects that underpin the linear-no- threshold dose response.

The biological effects of radiation can be affected by responses at many levels -from molecules to tissues. Evidence that molecular, cell and tissue responses can influence radiation and endogenous oxidative effects is challenging current radiobiological theory underpinning the linear-no-threshold model. For example, over the last decade, a number of studies have demonstrated an apparent adaptive response in cells irradiated with small doses of ionizing radiation. 21 These cells exhibit an increased resistance to the induction of radiation effects from subsequent higher doses of ionizing radiation. Although the initial endpoint in these studies was chromosome aberrations, adaptive responses to mutation, cytotoxicity, and cancer induction have been observed in cultured cells and in mice. It is likely that radiation-induced adaptation involves changes in DNA repair, signal transduction and/ or cell cycle kinetics. Most evidence indicates the adaptive response is related to oxidative stress and is associated with excision repair, although restriction enzymes that produce double strand breaks have also been shown to induce the adaptive response to ionizing radiation. Clearly, the adaptive response has the potential to impact adverse health risks and estimates of risk from low doses of radiation by altering DNA damage from both radiation and endogenous sources.

Even cells that are not irradiated can be affected by the irradiation of a neighboring cell. Recently, several laboratories have demonstrated changes in gene expression, increases in sister chromatid exchanges, and the induction of chromosomal instability in cells not directly irradiated but rather in proximity to irradiated cells. ⁸ Biological changes in cells not traversed by radiation have been called "bystander" effects. The mechanisms involved to induce bystander effects are under investigation and will help understand the mode of action of radiation. To date, bystander effects have only been associated with high LET radiation. It is important for this program to determine if these effects can be induced by exposure to low LET radiation delivered at low total doses or dose-rates. Demonstration of a by-stander effect for low doses of low LET radiation could, potentially, suggest an increased risk from low doses of radiation above the risks already predicted by linear no-threshold models.

The induction of genomic instability is postulated to be the underlying event that leads to the cascade of genetic changes that results in the genetic diversity observed in most solid cancers. It has now been clearly demonstrated that radiation can induce changes in cells that result in an increase in mutations and chromosome aberrations and a decrease in the cloning efficiency of the progeny of irradiated cells many population doublings after irradiation. Genomic instability has been demonstrated in both in vitro systems ⁵ and in vivo using mice. ¹³ What appears to be unique about radiation-induced genomic instability is its high frequency suggesting that it is not produced as the result of a change in a single gene or even a group of genes. Since the target for induction of genomic instability is located in the cell nucleus the high frequency suggests the target size is likely to encompass a large fraction of the genome. ⁶

Tissues have also been shown to play a deciding role in the ultimate fate of cancer or precancerous cells. For example, the extracellular matrix (ECM), the mass of fibrous and globular proteins that surrounds cells, performs a critical role in dictating a tissue's organization and function. 1 Communication networks have been demonstrated between the nucleus, cells, and their microenvironment. Surprisingly, ECM can actually trip switches deep within the nucleus and spur the genes themselves into action. ECM has been shown to play a critical role in the reversion of breast cancer cells to normal cell function in culture and in dramatically reducing tumors in mice. The notion that cancer is the result of not just genetic change, but an interweaving of mutation and changes in developmental regulation and tissue structure will have a profound impact on how we view cancer induction, diagnosis, and prognosis. It will also impact the way that we estimate cancer risk, especially from low dose exposures from which only small number of precancerous, and potentially inhibitable or reversible, changes might be expected.

Finally, the role of the entire organism in determining the ultimate biological responses and health risks from low doses of radiation need to be considered. Age-related changes in all of the responses described above could change the nature and biological impact of radiation-induced damage, resulting in age-dependent changes in health risk from low doses of radiation.

3) Decision Making Value

There is both suggestive and direct evidence that biological changes and responses induced by high doses of radiation may not always be the same at low doses of radiation. Some of these changes and responses will likely have no effect on the ultimate health risk from low dose radiation but others could be critical determinants of health risks from low dose radiation exposure. These various changes and responses need to be sorted out so that they are most useful to those charged with estimating health risk from low dose radiation exposures. Understanding the mechanisms of and the dose-effect relationships for the biological effects of low doses of ionizing radiation will provide the scientific basis in support of or against the existence of a threshold for adverse effects induced by low doses of ionizing radiation. As previously noted, a problem facing scientists and policy makers today is that all the information for radiation-induced DNA damage and the responses to that damage is from information obtained at high doses. There are simply no data at the low doses normally considered relevant to public health issues. Thus, it is fundamental to the entire low dose issue to determine if things happen the same way in cells, tissues, and people exposed to high and low doses of ionizing radiation and if the same things even happen?

If the biological changes and responses induced by low doses of ionizing radiation are similar qualitatively and quantitatively to those induced by high doses of radiation, then the linear-no-threshold model may be most appropriate for estimating health risks from low dose radiation. On the other hand, there may be real differences in some biological changes and responses induced at low radiation doses relative to those induced at high doses. If such differences are demonstrated, then it will need to be determined if the linear-no-threshold model overestimates (or even underestimates) cancer risk from low dose radiation.

4) Recommendations and Costs

Research is needed to understand and quantify real, not extrapolated or assumed, differences or similarities in biological changes and responses observed following exposures to low doses or low dose rates of ionizing radiation. This research covers the breadth of radiation and cancer biology from the initial recognition and processing of radiation damage by a cell to the potential development of cancer. Not all research, no matter how important to our understanding of the mechanisms of cellular responses to low dose radiation or of cancer development, will necessarily be useful for estimating health risks from low dose radiation or in choosing low dose radiation risk models. However, understanding and quantifying key aspects of the biological changes and responses induced by low dose radiation is likely to have dramatic impacts on our ability to efficiently and effectively protect people from unnecessary and avoidable health risks.

Research will benefit from the rapidly increasing availability of DNA sequence data from humans and other model organisms including mouse, yeast, fruit fly, etc. Recently developed technologies for characterizing and quantifying gene expression should be exploited. In some cases, further improvements in these technologies will be needed, such as increases in the sensitivity for detecting and quantifying gene expression. Cytogenetic techniques that couple traditional cytogenetic approaches with advances in molecular biology and automation will likely be useful in efforts to determine how accurately low dose radiation damage is repaired. Advances in the use and development of model organisms and of advanced systems for studying "normal" cells in culture should also be exploited to study the more complex interactions of cells and tissues in determining the biological effects of low dose radiation.

Research is needed that addresses the following key questions:

Do cells recognize and respond to low doses of ionizing radiation the same way that they do to high doses of radiation? As previously discussed, much of the damage induced by radiation and normal oxidative processes are similar. Research should concentrate on damage that is unique to low doses of radiation and on differences or similarities between biological responses following high versus low doses of radiation. It must be determined which genes and proteins are specifically induced in response to low doses of ionizing radiation, how these relate to other oxidative stresses, and, importantly, how the induced genes and proteins affect endpoints relevant to radiation-induced cancer. It must also be determined if the ability and efficacy of cells to recognize and repair radiation damage is affected by the radiation dose.

Do cells repair DNA damage induced by low doses of ionizing radiation the same way that they do damage induced by high doses of radiation? Understanding the repair or misrepair of radiation induced damage is dependent on understanding the nature of damage induced by low and high doses of radiation outlined in Section A above, Low Dose Radiation vs. Endogenous Damage -The Same or Different? The repair of radiation-induced DNA damage is of fundamental importance to all aspects of a cell and/ or an organism's responses to radiation exposure. The fidelity of the repair and damage processing systems will significantly affect the dose response curve for cancer induction, particularly at low doses. Ineffective repair or misrepair of radiation damage and subsequent processing of this unrepaired or misrepaired damage can significantly impact genomic integrity resulting in radiation-induced mutations, chromosomal aberrations, chromosomal stability, and cancer. Quite simply, if radiation-induced damage is faithfully repaired and processed, a threshold is expected. On the other hand, if repair and subsequent processing can lead to errors at low doses as it does at high doses, an expectation of a threshold is not warranted.

Additional understanding of the molecular mechanisms involved and in the closely linked damage signaling pathways will provide information relevant to the faithful repair of specific lesions, the molecular responses of cells to specific lesions and the consequences of cellular processing of radiation-induced damage compared to that of endogenous damage. Many of these consequences can be assessed using rapidly developing molecular cytogenetic technology such as combinatorial fluorescence in situ hybridization (FISH). Because cytogenetic effects represent the synthesis of damage induction, repair and processing, these new technologies provide the opportunity to directly test certain key predictions of models of radiation effects at low doses. Substantially more information is also needed on 1) the underlying repair processes; 2) the role of DNA sequence and chromatin structure in determining radiation response and target size for biological endpoints relevant to cancer; and 3) how and if the processing of damage induced by low doses of radiation leads to alterations in gene expression, changes in cell-cell or cell-matrix communication, mutations, chromosomal aberrations, and genomic instability.

Do low doses of radiation "protect" cells against subsequent low doses of ionizing radiation? If low doses of radiation regularly and predictably induce a protective response in cells to subsequent low doses of radiation this could have a substantial impact on estimates of adverse health risk from low dose radiation. Do radiation-induced adaptive responses protect cells against DNA damage caused by other carcinogenic agents, as well as radiation? How long do these protective effects persist? The generality and the extent of this apparent adaptive response in cells irradiated with small doses of ionizing radiation needs to be quantified.

Are the potentially damaging effects of low dose radiation amplified by interactions between cells? It is important for this program to determine if these so-called by-stander effects can be induced by exposure to low LET radiation delivered at low total doses or dose-rates. If such an effect is demonstrated and quantifiable, it could, potentially, increase estimates of risk from low dose radiation. This by-stander effect, in essence, "amplifies" the biological effects of a low dose exposure by effectively increasing the number of cells that experience adverse effects to a number greater than the number of cells directly exposed to radiation.

Is genetic instability, a key step in the development of cancer, induced or initiated by low doses of radiation? Current evidence suggests that radiation damage can lead to genetic instability in the progeny of irradiated cells. Does DNA repair and processing of radiation damage lead to instability? Is susceptibility to instability under genetic control? Information is needed on the relationship between radiation damage, DNA repair, and genetic instability and on the underlying mechanisms of how the processing of damage might lead to instability in the progeny of irradiated cells several generations later. Further, while there has been considerable speculation about the role of such instability in radiation-induced cancer, its role in this process remains to be determined.

Is the development of cancer induced by low (versus high) doses of radiation affected by normal tissues that surround the potential cancer cells? The ability of an irradiated cell to escape normal tissue regulatory processes or of a tissue to inhibit the further progression of precancerous cells may be differentially affected by high versus low doses of radiation. Exposure-and dose-response studies should be conducted to determine if the basic mechanisms of radiation action change as a function of total radiation dose and dose rate. High doses of ionizing radiation induce matrix and tissue disorganization, cell killing, changes in cell proliferation kinetics, induction of a multitude of genes and growth factors, and extensive chromosome and genetic damage. Many of these changes may be essential steps in radiation-induced cancer. It is important to determine if low doses of ionizing radiation can induce these biological changes. It will also be important to determine if cancer can be induced by doses that are too low to produce such changes.

The research described in this section is long range, basic research that will require regular monitoring to ensure that it stays focused on questions and results that will be useful in estimating health risks from low doses of radiation. In general, research results that are quantifiable will be most useful. Anticipated progress will be incremental, depending on results from previous experiments and research efforts. Thus, it is anticipated that research in this component of the program will continue for the duration of the program. A series of three to four cycles of two to three year grants is anticipated, with the focus of each subsequent cycle dependent on the results of the previous cycle.

Return to the top

C. Thresholds for low dose radiation -fact or fiction?

1) Key Question: Are there radiation doses or energies below which there is no significant biological change or below which the damage induced is effectively dealt with by normal cellular processes?

2) Description

The goal of the research described in this section of the Low Dose Radiation Research Plan is to determine if there are radiation doses or energies below which there is no significant biological change or below which damage can be effectively dealt with by normal cellular processes. If there are, then there should be no regulatory concern for exposures below these thresholds since there will be no increase in risk. The previous two sections of the Low Dose Radiation Research Plan outlined a research strategy to determine if:

This information will be used by scientists to develop computational techniques, e. g., algorithms and advanced mathematical approaches, that can be used to determine health risks from low doses of ionizing radiation. The new information derived from cellular and molecular studies together with available data from epidemiologic and animal studies will be incorporated into these models.

The linear-no-threshold model of radiation induced cancer states that cancer risk increases as a linear function of dose. From such a model it follows that even the smallest dose of radiation is theoretically capable of producing at least some cancers. It therefore becomes important to establish the validity of this model at very low doses. At issue is whether there are thresholds below which no excess cancer or genetic damage is induced. This is the topic of the previous two sections of this program plan and is a difficult issue to approach experimentally because of the inability to actually measure cancers produced by very low doses.

Several types of thresholds have been suggested. There are statistical or practical dose thresholds below which no increase in cancer can be detected because of the severe statistical limitations imposed by the high background rate of cancer and the low frequency of radiation induced cancer. There are potential energy thresholds related to the physical characteristics of the radiation itself, especially for low LET radiation, where the amount of energy deposited in a biological system is not adequate to cause biological damage. 2 Finally, biological thresholds have been postulated to exist that depend on biological processes, such as those outlined in Section B above, Understanding Biological Responses to Radiation and Endogenous Damage, acting on radiation induced damage or responses. The goal here is to determine if biological or energy thresholds exist following very low doses of ionizing radiation and to incorporate that information into new computational algorithms or advanced mathematical approaches that can be used to determine health risks from low doses of ionizing radiation.

3) Decision Making Value

In the absence of clear or useful scientific data, standards for exposure to low doses of radiation in the workplace or the environment are currently based on default conservative assumptions. Extensive observational and epidemiological data is available on the health effects of high levels of radiation exposure. Extrapolation of these data to low doses form the basis for current radiation standards. The research described in this Low Dose Radiation Research Plan will form the basis for a new scientific data set that will underpin future standards for and estimates of risk from exposure to low doses of ionizing radiation.

4) Recommendation and Cost

The principal focus of research in this component of the Low Dose Radiation Research Plan is to develop methods to synthesize or model new, quantitative, information (including molecular information) on low dose radiation induced damage and biological responses to that damage into a low dose radiation risk model. The goal of this research program is to develop scientifically defensible tools and approaches for determining risk that are widely used, accepted, and understood. Research should include, but not be limited to development of computational techniques, e. g., algorithms and advanced mathematical approaches, for use in determining risk, that model new quantitative information from cellular and molecular studies together with available data from epidemiologic and animal studies.

It is important to remember that if thresholds exist, there may be a series of thresholds depending on the biological system and endpoint being studied and the radiation doses being used. For example, there could be different thresholds for biological effects at the molecular, cellular, organ, and whole animal or human level. It is also possible that when dose is redefined at the cellular and molecular level, new paradigms of energy interaction may develop and it may be necessary to abandon the use of at this level. In order to provide a logical framework for relating dose and biological effect at each level of organization from the molecular to the whole organism, the advanced computational models based on quantitative experimental data developed in this program will be critical and useful.

A secondary, but essential component of this component of the Low Dose Radiation Research Plan, will be the design and conduct of additional biological experiments to address specific questions or predictions made by these new computational approaches. These biological experiments, though likely complementary to research conducted as part of Section B above, Understanding Biological Responses to Radiation and Endogenous Damage, will be designed and conducted in collaboration with modelers.

It is anticipated that three to four cycles of two to three year grants will be funded as part of these studies. Focused biological studies will be funded in parallel with computational studies beginning after the first cycle of computational grants. Anticipated progress will be incremental, depending on results from previous experiments and research efforts. Thus, it is anticipated that research in this component of the program will continue for the duration of the program.

D. Genetic factors that affect individual susceptibility to low dose radiation

1) Key Question: Do genetic differences exist making some individuals more sensitive to radiation induced damage? Are these individuals or related subpopulations at increased risk for radiation induced cancer?

2) Description

During the last decade there has been a progressive increase in understanding of the genetic contribution to complex diseases including cancer. Molecular studies examining the genetic component of susceptibility to cancer have identified a number of genes that confer susceptibility, and the number of such genes continues to increase. It is likely that there are also individual differences in susceptibility to radiation-induced cancer.

Recent developments have suggested a link between cellular responses to ionizing radiation and cancer susceptibility. Dose response kinetics for the induction of certain types of chromosome damage also correlate with cancer susceptibility although this correlation is only phenomenological. There is clear evidence in mice for genetic control of susceptibility to radiation-induced genomic instability that may extend to cancer susceptibility as well. Is there a similar susceptibility in humans? Further, gene products involved in the recognition and repair of DNA damage have been shown to be physically associated in cells whereas those same gene products are apparently disrupted in individuals with heritable diseases associated with genomic instability and cancer. ^{3, 12} Functional associations linking cell cycle, apoptosis and double strand break repair have also been defined, offering additional gene pathways that may be involved in cancer susceptibility. ²²

Genes associated with several different cancer prone diseases have also been shown to be associated with some form of alteration in DNA repair. Cells deficient in the ATM gene (the recently isolated gene associated with the disease Ataxia Telangiectasia) have defective damage response mechanisms, are sensitive to ionizing radiation, and have increased levels of spontaneous and radiation-induced chromosome aberrations. A protein complex associated with non-homologous chromosomal end-rejoining is defective in patients with Nijmegen breakage syndrome. ^{3, 18} Individuals with either Ataxia Telangiectasia or Nijmegen breakage syndrome are cancer prone, radiation sensitive, and demonstrate increased levels of chromosomal instability. Interestingly, the BRCA1 and 2 genes, found to be defective in many patients predisposed to breast and ovarian cancer, also appear to be involved in DNA double strand break repair pathways.

Overall, few genes or genetic conditions have been identified as potential susceptibility genes for cancer or radiation sensitivity. Currently, there is insufficient information to determine the total number of potential susceptibility genes, to estimate the frequency of polymorphisms in these genes in the population, and to assess the impacts on radiation-induced health risk that they pose. For example, patients who develop secondary cancers following radiation therapy could represent a subset of the population who are radiation sensitive, including to low doses of radiation. Molecular technologies provide powerful new ways to analyze the mammalian genome and address these issues. As this area of research matures, more complex issues of genetic interactions, including gene modifiers and gene-gene interactions and their impact on radiation-induced cancer will be able to be addressed.

3) Decision Making Value

Studies focusing on genetic susceptibility to radiation-induced cancer will improve understanding of low dose risks and will create opportunities for new basic knowledge of potential wide-ranging importance. The extent to which these studies impact current and future risk policy depends on the frequency of susceptibility genes in the general population and the ability of those genes to significantly influence low dose risks. If there are enough people who are unable to properly respond to and process radiation damage, then any model of radiation risk to the general population suggesting a threshold would appear to be untenable. Such information will also create opportunities to specifically identify susceptible individuals as well as provide insight into approaches to modify such susceptibility. Information derived from this research will need to be coupled with methods used to determine health risk. It is unlikely, given the anticipated frequencies of potential "susceptibility" genes, that estimates of population risk will be affected. However, information on genetic susceptibility to low doses of radiation could have dramatic impacts on estimates of individual risk.

Eventually we will have an understanding of all human genes and the possible role that some subset of these genes plays in determining individual susceptibility to radiation and to cancer. While attaining this level of understanding will be a major international achievement far beyond the scope of this research project, major challenges and uncertainties regarding the use of this information will remain. These challenges and uncertainties strike at the very heart of issues being actively discussed today and include issues of individual rights, genetic privacy, workplace discrimination, and health insurance discrimination to name a few.

If we had the capability today to identify all people with increased susceptibility to radiation induced damage what would we or could we do with that information? Would we keep them out of jobs or environments where they might receive even the smallest preventable radiation exposure? Would we tell them their risks and let them choose? Would we release this information to their employers? Their physicians? Their insurance companies? Their relatives? These are not and will not be easy decisions and are outside of the scope of this program.

Similarly, decisions will need to be made regarding the development of radiation exposure guidelines. The overriding goal of this research program is to provide information that can be used to ensure the adequate and appropriate protection of human health. What is adequate and appropriate? No risk at any cost? Acknowledged but acceptable minimal risk? Would we or should we protect all of society from radiation exposures that pose a health risk to the most sensitive among us? Again, difficult but unavoidable questions that will arise from research on the genetic susceptibility to low dose radiation that is an important part of this program.

4) Recommendations and Costs

The Low Dose Radiation Research Program should have three main goals in terms of genetic susceptibility to low dose radiation:

Research in these three areas will strongly complement ongoing initiatives at the National Institutes of Health.

The National Human Genome Research Institute (NHGRI) is funding research to identify common variants in the coding regions of the majority of human genes identified during the next five years. The goal is to develop a catalog of all common variants in all human genes. The NHGRI is also working to create a map of at least 100,000 single nucleotide polymorphisms, the most common polymorphisms in the human genome representing single base-pair differences between two copies of the same gene. These so-called SNPs will be a boon for mapping genes for complex diseases and traits such as cancer, cancer susceptibility, and susceptibility to low dose radiation.

The National Institute of Environmental Health Science (NIEHS) is funding research as part of the Environmental Genome Project to understand the impact and interaction of environmental exposures on human disease. The NIEHS project includes efforts to understand genetic susceptibility to environmental agents that will allow more precise identification of the environmental agents that cause disease and the true risks of exposures. Its principal focus is on chemicals. Thus, the focus on radiation in the Low Dose Radiation Research Program is highly complementary. Initially, the Environmental Genome Project will focus on categories of genes including: xenobiotic metabolism and detoxification genes; hormone metabolic genes; receptor genes; DNA repair genes; cell cycle genes; cell death control genes; genes mediating immune and inflammatory responses; genes mediating nutritional factors; genes involved in oxidative processes and, genes for signal transduction systems.

Efforts in the Low Dose Radiation Research Program should be coordinated with activities at the NHGRI and NIEHS in particular to prevent duplicative effort and to facilitate rapid progress. Coordination can include, but should not be limited to, joint planning, joint meetings of program staff and/ or funded investigators, joint solicitations, or co-funding of research grants.

Identification of potential susceptibility genes and polymorphisms in those genes is only the first (and perhaps the easiest) step in the program to characterize and understand genetic susceptibility. Determining the biological significance of these genetic polymorphisms with respect to cancer and radiation sensitivity is the ultimate goal and the more difficult task. The international human genome project, structural biology research, and the NHGRI and NIEHS efforts described above play important roles in determining which polymorphisms are most likely to influence gene function. Population genetics and computational biology approaches will be required to estimate the potential impact on estimates of population and individual risk. Genetic epidemiology approaches will also be needed to relate specific polymorphisms and combinations of polymorphisms with cancer risk. Inbred mouse strains and other model organisms with well-characterized differences in susceptibility to radiation-induced cancer are also important tools for identifying significant polymorphisms. Direct assessment of the biological significance of candidate "susceptibility genes" can also be undertaken using animal models such as knock-out and knock-in mice, mice with specific genes removed or added.

It is anticipated that three cycles of three-year grants will be funded as part of these studies. Research efforts will likely scale-up in the later years of the program as DNA sequence information and information on genetic polymorphisms becomes more broadly available from this and other program. Anticipated progress will be incremental, depending on results from previous experiments and research efforts. Thus, it is anticipated that research in this component of the program will continue for the duration of the program.