Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
BACKGROUND INFORMATION AND SCIENTIFIC PRINCIPLES 9 1 Background Information and Scientific Principles PHYSICS AND DOSIMETRY OF IONIZING RADIATION All living matter is composed of atoms joined into molecules by electron bonds. Ionizing radiation is energetic enough to displace atomic electrons and thus break the bonds that hold a molecule together. As described below, this produces a number of chemical changes that, in the case of living cells, can lead to cell death or other harmful effects. Ionizing radiations fall into two broad groups: 1) particulate radiations, such as high energy electrons, neutrons, and protons which ionize matter by direct atomic collisions, and 2) electromagnetic radiations or photons such as x rays and gamma rays which ionize matter by other types of atomic interactions, as described below. Absorption and Scattering of Photons Photons ionize atoms through three important energy transfer processes: the photoelectric process, Compton scattering, and pair production. For photons with low energies (<0.05 megaelectron volt [Me V]) the photoelectric process dominates in tissue. The photoelectric process occurs when an incoming photon interacts with a tightly bound electron from one of the inner shells of the atom, and causes the electron to be ejected with sufficient energy to escape the atom. Characteristic x rays and Auger electrons follow from this process, but the biological effects are due mainly to excitations and ionizations in molecules of tissue caused by the ejected electron. The probability of the photoelectric process occurring is strongly
BACKGROUND INFORMATION AND SCIENTIFIC PRINCIPLES 10 dependent on the average atomic number of the tissue with an equally strong inverse dependence on the photon energy. At higher photon energies (0.1-10 Me V), Compton scattering is the most probable process that takes place in irradiated tissue. It occurs when the photon energy greatly exceeds the electron binding energy, so that an orbital electron appears to the photon as a free electron. The photon scatters off the electron, giving up part of its energy to the electron, which proceeds to ionize and excite tissue molecules. The scattered photon with reduced energy continues to interact with other electrons and repeats the above process many times until the photon either escapes the absorbing material or its energy is sufficiently degraded for the photoelectric process to occur. Within the energy range of 0.1-10 Me V, the Compton process has a modest dependence on energy and is almost independent of atomic number. Above a threshold energy of 1.02 Me V, the pair-production process is possible. Here a photon converts its energy in the presence of an atomic nucleus to a positron-electron pair, which, in turn, proceeds to interact with tissue atoms and molecules, leading to eventual biological effects. When the positron slows down it is almost always annihilated with an electron, producing two 0.511 Me V photons. The probability of pair-production in tissue increases slowly with photon energy but does not outweigh that of the Compton process until the photon energy reaches 20 Me V. The process depends upon the average atomic number of the tissue. Photon Spectral Distributions As seen from the description presented above, the absorption and scattering of photons depend critically on photon energy. The initial photon energy depends on the source of the radiation. Gamma rays resulting from radioactive decay consist of monoenergetic photons with energies that do not exceed several Me V in energy. Because of scattering and absorption within the radioactive source itself and in the encapsulating material, the photons that are emitted do have a spectrum of energies but it is fairly narrow. Relatively broad energy distributions are the rule for x-ray photons produced from electrical devices. X rays are effectively produced by the rapid deceleration of charged particles (usually electrons) by a material of high atomic number. This results in a continuous distribution of energies with a maximum at an energy about one third that of the most energetic electron. As photons interact with matter, their spectral distribution is further altered in a complex manner as the photons transfer energy to the absorbing medium by the processes described above.
BACKGROUND INFORMATION AND SCIENTIFIC PRINCIPLES 11 Electron Spectral Distributions and LET When monoenergetic photons interact with a tissue medium, the electrons that are set in motion, particularly from the Compton process, proceed to interact with the atoms and molecules of the medium, losing energy through collisions and excitations, and are scattered in the process. The result is a complex shower of electrons, the energy distribution of which is continuously degraded as the electrons give up their energy to the medium at a rate defined by the electron stopping power of the medium. As the electron proceeds through tissue, it creates a track of excited and ionized molecules that, for energetic electrons, are relatively far apart. For example, the dimension of this spacing is such that there is a finite probability that the energetic electron can pass through a DNA molecule, with about 3 nm separating the two strands, without releasing any of its energy and therefore without causing damage. The spatial energy distribution, stated in terms of the amount of energy deposited per unit length of particle track, is defined as the linear energy transfer (LET) of the radiation. X rays and gamma rays set in motion electrons with a relatively low spatial rate of energy loss and thus are considered low LET radiations. The photon and electron energy degradation processes described above result in a broad distribution of LET values occurring in irradiated tissue. A typical value of LET for the electrons set in motion by cobalt-60 gamma rays (average energy 1.25 Me V) would be about 0.25 keV/Âµm. This can be contrasted with a densely ionizing 2 Me V alpha particle which produces about 1000 times more ionization per unit distance, 250 keV/Âµm. Such particles are characterized as high LET radiation. Knowledge of LET is important when considering the relative biological effectiveness (RBE) of a given radiation; LET is commonly used as a measure of radiation quality, as discussed below. Microdosimetry Various limitations in the concept of LET and absorbed dose in subcellular tissue volumes led to the introduction of microdosimetry. Microdosimetry takes account of the fact that energy deposition by ionizing radiations is a stochastic (random) process. Identical particles of the same energy interacting in a small volume of material deposit differing amounts of energy due to chance alone. The specific energy, z, is defined as the ratio Îµ/m where Îµ is the energy imparted by a single ionizing particle in a volume element of mass m. The mean value of z for a large number of particles is equal to the absorbed dose. The microdosimetric analogue to LET is the quantity lineal energy, defined as Îµ/d, where d is the mean chord length in the volume occupied by mass m. Distributions of absorbed dose in terms of lineal energy can be measured by proportional counters
BACKGROUND INFORMATION AND SCIENTIFIC PRINCIPLES 12 filled with tissue-equivalent gas at pressure levels appropriate for simulating spheres of tissue with diameters on the order of 1 Âµm. The principles of microdosimetry are extensively discussed in the BEIR IV report (NRC88) and ICRU report 36 (ICRU83). Energy TransferâKerma and Absorbed Dose The transfer of energy from photons to tissue takes place in two stages: (1) the interaction of the photon with an atom, causing an electron to be set in motion, and then (2) the subsequent absorption by the medium of kinetic energy from the high energy electron through excitation and ionization. The first stage can be identified with the quantity called kerma, K, which stands for kinetic energy released in the material. K = dEtr/dm, where dEtr is the kinetic energy transferred from photons to electrons in a volume element of mass dm. The second stage, energy absorption, is more important for understanding radiobiological effects. The absorbed dose, the energy absorbed per unit mass, differs from kerma in that the dose may be smaller due to lack of charged particle equilibrium, bremsstrahlung escaping from the medium, etc. Another difference is that the kerma refers to energy transfer at a point, whereas the energy is absorbed over a distance equal to the electron range. Of the two quantities, absorbed dose is the easier one to approach experimentally and can be determined by a number of well-defined techniques, including gas ionization methods, calorimetry, and thermoluminescent techniques. On the other hand, kerma is often more easily calculated. Radiation Chemical Effects Following Energy Absorption After the electron produced by a photon interaction passes through tissue, exciting and ionizing atoms and molecules, a number of important chemical events that precede the biological effects take place. Most of the energy absorption takes place in water, since cells are made up of more then 70% water. When an ionizing particle passes through a water molecule, it may ionize it to yield an ionized water molecule, H2O+, and an electron by the reaction: The electron can be trapped, polarizing water molecules to produce the so- called hydrated electron, eaq. On the other hand, the ionized water molecule, H2O +, reacts at the first collision with another water molecule to produce an hydroxyl radical, OHâ¢ according to the reaction:
BACKGROUND INFORMATION AND SCIENTIFIC PRINCIPLES 13 H2O+ + H2O â OHâ¢ + H3O+. The free radical OHâ¢ has an unpaired electron and is therefore highly reactive as it seeks to pair its electron to reach stability. At the high initial concentrations, certain back reactions occur producing hydrogen molecules, hydrogen peroxide and water. The initial species produced in water radiolysis can then be written as: Instead of being ionized, the water molecule may simply be excited according to the reaction: where H2O* is the excited molecule. But H2O* soon breaks up into the Hâ¢ radical and the OHâ¢ radical according to: H2O* â Hâ¢ + OHâ¢. As a result of the above processes, three important reactive species are produced: the aqueous electron, OHâ¢, and Hâ¢, with initial relative yields of about 45%, 45%, and 10%, respectively. These reactive species attack molecules in the cell leading to the production of biological damage. The OHâ¢ radical is believed to be the most effective of the three species in causing damage. Because it is an oxidizing agent, it can abstract a hydrogen atom from the deoxyribose moiety of DNA, for example, yielding a highly reactive site on DNA in the form of a DNA radical. Since this process arises from the irradiation of a water molecule rather than the DNA itself, the process is known as the indirect effect. Electrons set in motion by photons can, of course, directly excite or ionize cell macromolecules by direct interaction with the critical molecule. This is called the direct effect. Both mechanisms can produce cellular damage. There is strong evidence that the DNA is the most critical site for lethal damage, but other sites such as the nuclear membrane or the DNA- membrane complex may also be important. Ward (Wa88) has derived an approximation of the damage yields expected in various moieties of DNA within an irradiated cell, in which consideration is given to the direct deposition of energy in DNA and other molecules. Table 1-1 shows the amount of energy deposited per Gray in each moiety of DNA within a cell that is assumed to contain 6 pg of DNA.
BACKGROUND INFORMATION AND SCIENTIFIC PRINCIPLES 14 TABLE 1-1 Amount of Energy Deposited in DNA per Cell per Gray Constituent Mass per Cell (pg) eV Deposited Number of 60-eV Events Deoxyribose 2.3 14,000 235 Bases 2.4 14,700 245 Phosphate 1.2 7,300 120 Bound water 3.1 19,000 315 Inner hydration 4.2 25,000 415 SOURCE: J. F. Ward, C. L. Limoli, P. Calabro-Jones, and J. W. Evans (Wa88). Calculated from this is the number of events since 60 eV is the average amount of energy deposited per event. The yields of DNA damage necessary to kill 63% of mammalian cells (63% of cells killed means that, on average, each cell has sustained one lethal event) can be assessed for various lethal agents (Wa88), as shown in Table 1-2. The high efficiency with which ionizing radiation (and bleomycin) kill cells is not simply due to individual OH radical-induced lesions, as witnessed by the large-scale production of single-strand breaks with hydrogen peroxide. Ward et al. (WA87) suggest that the efficiency of cell killing by ionizing radiation at relatively low levels of DNA damage is due to the production of damage in more than one moiety in a localized region, i.e., lesions resulting from multiply damaged sites in a single location or locally multiply damaged sites (LMDS). Recent studies (Wi85, Gr85, Ei81), as analyzed by Ward (Wa88), support the importance of indirect effects of ionizing radiation in producing damage to intracellular DNA. This is of particular significance in view of the suggestion that most intracellular DNA damage is caused by direct ionization and that radicals produced in water cannot access the macromolecule. It appears from the above analysis (Wa88) that the volume of water in the DNA-histone complex (nucleosome) is at least equal to the DNA volume and that radiation- produced OH radicals in the water volume have ready access to the DNA molecule. Some of the current assessments of DNA damage caused by ionizing radiation in mammalian cells (Wa88) are as follows: (1) direct and indirect effects are both important; (2) the quantity of damage produced by ionizing radiation is orders of magnitude lower than for most other agents for equal cell- killing efficiency; (3) individual damage moieties are not biologically significant since they can be repaired readily by using the undamaged DNA strand as a template; (4) LMDS are more likely the lethal lesion in cellular
BACKGROUND INFORMATION AND SCIENTIFIC PRINCIPLES 15 DNA; these result from a high local energy deposition in the DNA (in such a volume, multiple radicals cause multiple lesions locally); (5) the individual lesions making up an LMDS can be widely separated on the opposite strands of the DNA; if they are separated too much, they could be repaired as individual lesions. TABLE 1-2 Yields of DNA Damage Necessary to Kill 63% of the Cells Exposed Agent DNA Lesion Number of Lesions per Cell per D37a Ionizing radiation ssB 1,000 dsB 40 Total LMDSb 440 DPCc 150 Bleomycin A2 ssB 150 dsB 30 UV light T<>T dimer 400,000 ssB 100 Hydrogen peroxide 0Â° ssB <2,600,000 37Â°C ? Benzo[a]pyrene 4,5-oxide Adduct 100,000 Aflatoxin Adduct 10,000 1-Nitropyrene Adduct 400,000 Methylnitrosourea 7-Methylguanine 800,000d O6-Methylguanine 130,000d 3-Methyladenine 30,000d 2-(N-Acetoxy-N-acetyl) amino- Adduct 700,000 fluorene Other similar aromatic amides produce about the same number of adducts per lethal event a D37 = dose of agent required to reduce survival of cells to 37% of the number exposed. b Calculated, LMDS = locally multiply damaged sites. c DPC = DNA-protein cross-links. d D calculated from individual exposures; no survival curves available. 37 SOURCE: J. F. Ward, C. L. Limoli, P. Calabro-Jones, and J. W. Evans (Wa88). Physics and Dosimetry of High-LET Radiation (Neutrons) Interactions of Neutrons with Tissue Elements When neutrons impinge on a tissue medium, they will either penetrate it without interacting with its constituent atoms or they will interact with its atoms in one or more of the following ways: (1) elastically, (2) inelastically, (3) nonelastically, (4) by capture reactions, or (5) through spallation processes.
BACKGROUND INFORMATION AND SCIENTIFIC PRINCIPLES 16 Elastic scattering is the most important interaction in tissue irradiated with neutrons at energies below 20 MeV. This would include the energy range for fission neutrons (<10 Me V), neutrons produced with 16 Me V deuterons bombarding a beryllium target (<20 Me V), and neutrons produced with 150 keV deuterons on tritium (<20 Me V). The neutron, an uncharged particle, interacts primarily by collisions with nuclei in the absorbing medium. If the total kinetic energy of the neutron and the nucleus remains unchanged by the collision, the collision is termed elastic. During an elastic collision, the maximum energy is transferred from the neutron to the nucleus if the two masses are equal. In soft tissue, the most important neutron interaction is with hydrogen. There are three reasons for this: (1) Nearly two-thirds of the nuclei in tissue are protons, (2) the energy transfer with protons is maximal (about one- half), and (3) the interaction probability (cross-section) for hydrogen is larger than that for any other element. The result is that about 90% of the energy absorbed in tissue from neutrons with energy of less than 20 Me V comes from protons that are recoiling from elastic collisions. The remaining energy is absorbed by other recoiling tissue nuclei in the following decreasing order of importance: oxygen, carbon, and nitrogen. Inelastic scattering refers to reactions in which the neutron interacts with the nucleus but is promptly reemitted with reduced energy and usually with a changed direction. The scattering nucleus, which is left in an excited state, then emits a nuclear deexcitation gamma ray. For neutrons with kinetic energies of greater than 10 Me V, inelastic scattering contributes to energy loss in tissue; about 30% of the energy deposited in tissue by 14-Me V neutrons, for example, comes from inelastic interactions. The important inelastic interactions of neutrons in soft tissue are not with hydrogen but with carbon, nitrogen, and oxygen. Nonelastic scattering defines reactions in which the neutron-nucleus interaction results in the emission of particles other than a single neutron such as alpha particles and protons [e.g., 16O(n,Î±)13C, 14 N(n,p)14C]. The cross- sections for nonelastic scattering in tissue become significant at energies greater than 5 Me V and increase as the neutron energy approaches 15 Me V. These reactions are usually accompanied by deexcitation gamma rays, but their importance is due to the high LET of the charged particles emitted, especially alpha particles. At neutron energies greater than 20 Me V, even though nonelastic cross-sections do not increase appreciably, nonelastic processes become increasingly important contributors to the total dose because of the increased average energy of the charged particles resulting from the interaction. The capture of low-energy neutrons in the thermal and near-thermal regions provides a significant contribution to tissue dose. The reactions of importance are 14N(n,p)14C and 1H(n, Î³)2H. The former reaction produces
BACKGROUND INFORMATION AND SCIENTIFIC PRINCIPLES 17 locally absorbed energy of 0.62 Me V from the proton and the recoil nucleus. The latter reaction yields a 2.2-Me V gamma ray that, in general, deposits energy at a distance from the capture site and that has a reasonable probability of escaping altogether from a mass as large as a rodent. For thermal neutrons the 14N(n,p)14C reaction is the major contributor of absorbed energy in tissue samples with a dimension of less than 1 cm because of the short range (<10 Âµm) of the 0.58-Me V proton. However, for larger masses of tissue (e.g., the human body), the 2.2-Me V gamma rays from the 1H(n, Î³)2H reaction are a significant dose contributor. In the spallation process the neutron-nucleus interaction results in the fragmentation of the nucleus with the emission of several particles and nuclear fragments. The latter are heavily ionizing, so the local energy deposition can be high. Several neutrons and deexcitation gamma rays also can be emitted, yielding energy carriers that escape local energy deposition. The spallation process does not become significant until neutron energies are much greater than 20 Me V. In summary, elastic and nonelastic scattering and the capture process are by far the most important reactions in tissue for neutrons in the fission energy range. Inelastic and nonelastic scattering begin at about 2.5 and 5 Me V, respectively, and become important at an energy of about 10 Me V. As the neutron energy goes higher, nonelastic scattering and spallation reactions increase in importance, and elastic scattering becomes of less importance for energies greater than 20 Me V. POPULATION EXPOSURE TO IONIZING RADIATION IN THE UNITED STATES A new assessment of the average exposure of the U.S. population to ionizing radiation has recently been made by the National Council on Radiation Protection and Measurements (NCRP87b). Six main radiation sources were considered: natural radiation and radiation from the following five man-made sources: occupational activities (radiation workers), nuclear fuel production (power), consumer products, miscellaneous environmental sources, and medical uses. For each source category, the collective effective dose equivalent was obtained from the product of the average per capita effective dose equivalent received from that source and the estimated number of people so exposed. The average effective dose equivalent for a member of the U.S. population was then calculated by dividing the collective effective dose equivalent value by the number of the U.S. population (230 million in 1980). As discussed below, the dose equivalent is defined as the product of the absorbed dose, D, and the quality factor Q, which accounts for
BACKGROUND INFORMATION AND SCIENTIFIC PRINCIPLES 18 differences in the relative biological effectiveness of different types of radiation. The effective dose equivalent relates the dose-equivalent to risk. For the case of partial body irradiation, the effective dose equivalent is the risk- weighted sum of the dose equivalents to the individually irradiated tissues. TABLE 1-3 Average Annual Effective Dose Equivalent of Ionizing Radiations to a Member of the U.S. Population Dose Equivalenta Effective Dose Equivalent Source mSv mrem mSv % Natural Radonb 24 2,400 2.0 55 Cosmic 0.27 27 0.27 8.0 Terrestrial 0.28 28 0.28 8.0 Internal 0.39 39 0.39 11 Total natural â â 3.0 82 Artificial Medical x-ray diagnosis 0.39 39 0.39 11 Nuclear medicine 0.14 14 0.14 4.0 Consumer products 0.10 10 0.10 3.0 Other Occupational 0.009 0.9 <0.01 <0.3 Nuclear fuel cycle <0.01 <1.0 <0.01 <0.03 Fallout <0.01 <1.0 <0.01 <0.03 Miscellaneousc <0.01 <1.0 <0.01 <0.03 Total artificial â â 0.63 18 Total natural and artificial â â 3.6 100 a To soft tissues. b Dose equivalent to bronchi from radon daughter products. The assumed weighting factor for the effective dose equivalent relative to whole-body exposure is 0.08. c Department of Energy facilities, smelters, transportation, etc. SOURCE: National Council on Radiation Protection and Measurements (NCRP87b). As seen in Table 1-3 and Figure 1-1, three of the six radiation sources, namely radiation from occupational activities, nuclear power production (the fuel cycle), and miscellaneous environmental sources (including nuclear weapons testing fallout), contribute negligibly to the average effective dose equivalent, i.e., less than 0.01 millisievert (mSv)/year (1 mrem/year). A total average annual effective dose equivalent of 3.6 mSv (360 mrem)/ year to members of the U.S. population is contributed by the other three sources: naturally occurring radiation, medical uses of radiation, and radiation from consumer products. By far the largest contribution (82%) is made by natural sources, two-thirds of which is caused by radon and its
BACKGROUND INFORMATION AND SCIENTIFIC PRINCIPLES 19 decay products. Approximately equal contributions to the other one-third come from cosmic radiation, terrestrial radiation, and internally deposited radionuclides. The importance of environmental radon as the largest source of human exposure has only recently been recognized. FIGURE 1-1 Sources of radiation exposure to the U.S. population (NCRP87b). The remaining 18% of the average annual effective dose equivalent consists of radiation from medical procedures (x-ray diagnosis, 11% and nuclear medicine, 4%) and from consumer products (3%). The contribution by medical procedures is smaller than previously estimated. For consumer products, the chief contributor is, again, radon in domestic water supplies, although building materials, mining, and agricultural products as well as coal burning also contribute. Smokers are additionally exposed to the natural radionuclide polonium-210 in tobacco, resulting in the irradiation of a small region of the bronchial epithelium to a relatively high dose (up to 0.2 Sv per year) that may cause an increased risk of lung cancer (NCRP84). Uncertainties exist in the data shown in Table 1-3. Uncertainties for exposures from some consumer products are greater than those for exposures from cosmic and terrestrial radiation sources. The estimates for the most important exposure, that of lung tissue to radon and its decay products, have many associated uncertainties. Current knowledge
BACKGROUND INFORMATION AND SCIENTIFIC PRINCIPLES 20 of the average radon concentration, the distribution of radon indoors in the United States, and alpha-particle dosimetry in lung tissue is limited. In addition, knowledge of the actual effective dose equivalent is poorly quantified. Further uncertainties are caused by difficulties in combining data for exposure from different sources that actually are from different years, mainly from 1980 to 1983. RADIOBIOLOGICAL CONCEPTS Experiments on radiation-induced cell killing have given rise to a number of radiobiological principles and concepts. Many of these principles and concepts are inferred to apply to mutagenesis and carcinogenesis, as well as to cell killing, although this is often not known for certain since it is not possible to perform comparable experiments with all of these endpoints. Some of the major concepts are discussed below. The first concept is that the principal target for radiation-induced cell killing is DNA. Although it is not the exclusive target, it is generally the most consequential. While the evidence for this conclusion is circumstantial, it is also convincing (Le56). As noted above, the consequences of the absorption of radiant energy arise from excitations and ionizations along the tracks of the charged particles that are set in motion when radiant energy is absorbed. Biological damage may be a consequence of a direct interaction between the charged particles and the DNA molecule, or the biological effects may be mediated by the production of free radicals (Mi78). In the latter case, which is the indirect action of radiation, the absorption of the radiation may occur in, for example, a water molecule, and the consequent free radical produced may diffuse to the DNA, where it gives up its energy to produce a biological lesion. In the case of sparsely ionizing radiations, such as x rays and gamma rays, about two-thirds of the biological effects are produced by this indirect action, and this component of the radiation damage is amenable to modification by a variety of physical and chemical factors. As the quality of the radiation changes from low to high LET, the balance shifts from the indirect action to the direct action. The second major concept concerns the shape of the dose-response relationship. With cell lethality, R, as the endpoint, the dose-response relationship for low-LET radiations often approximates a linear-quadratic function of the dose, D. R = Î±D + Î²D2. The relative importance of the linear and quadratic terms varies widely for different cells and tissues. The ratio Î±/Î², which is the dose at which the linear and quadratic contributions to the biological effect are equal,
BACKGROUND INFORMATION AND SCIENTIFIC PRINCIPLES 21 may vary from about 1 Gray (Gy) to more than 10 Gy. As the LET of the radiation is increased, the ratio Î±/Î² also increases for a given cell or tissue, and for very high LET radiations, survival (1-R) approximates an exponential function of dose at doses of interest. For carcinogenesis in laboratory animals, dose-response relationships with a wide variety of shapes have been reported. At higher doses there is the complication of a balance between increased cell transformation and increased cell killing. The linear-quadratic formulation had its origins in the 1930s, when it was used to fit data for radiation induced chromosome aberrations (Sa40). Many chromosome aberrations appear to be the consequence of the interaction between breaks in two separate chromatids. This applies to aberrations, such as dicentrics, that lead to cell lethality, as well as to aberrations such as translocations that, in some cases, lead to cancer through the activation of an oncogene. Thus, the interpretation of the linear-quadratic formulation is that the characteristic shape of the dose-response curve reflects a predominance of single-track events, which are proportional to the dose at low doses and low dose rates, and of two-track events which are proportional to the square of the dose and result in the upward bending of the cancer induction curve at high doses received at high dose rates. This biophysical model has been challenged in recent years, largely on the basis of data with soft x rays, which are highly effective biologically even though the length of the secondary tracks they produce is too short to enable a single track to break two independent chromosomes (Th86). Hence, although the data have been interpreted in terms of the more conventional linear- quadratic formulation (Br88), an alternative model has been proposed in which all biological damage is presumed to result from single track effects, with the additional factor of a repair process that saturates at higher doses. Biological experiments that allow an unequivocal choice to be made between the models have not yet been performed. The third concept is that the biological consequence of a given dose of radiation varies with the quality of the radiation. With cell killing as the endpoint, the relative biological effectiveness (RBE) of many types of radiation has been studied in detail (Ba63). Although the RBE varies with the LET of the radiation, it also varies with the dose, dose rate, type of cell or tissue used to score the biological effect, and the endpoint in question (Br73, Ba68). The pattern of variation of the RBE with LET appears to be similar for mutagenesis as for cell killing, but it has not been established to be the same for carcinogenesis as an endpoint. The quality factor (Q) rather than RBE is widely used in radiation protection. The International Commission on Radiological Protection (ICRP) has suggested, however, that the quality factor should be based on a microdosimetric quantity such as lineal energy (ICRU86).
BACKGROUND INFORMATION AND SCIENTIFIC PRINCIPLES 22 For cell lethality as an endpoint, cell sensitivity to radiation varies as a function of its stage in the cell cycle. This is the fourth major radiobiological concept. In general, cells are most sensitive in the G2 phase or in mitosis, and they are most resistant during the phase of DNA synthesis (Si66, Ter63). In the case of mutagenesis, it appears, in some instances at least, that the most sensitive phase of the cycle is G1. There is little or no information concerning the variation of cellular sensitivity with the phase of the cell cycle for oncogenic transformation in vitro. The fifth concept is that the effect of a given dose may be influenced greatly by the dose rate. The influence of the dose-rate effect has been widely studied and is well established for cell lethality as an endpoint (Ha64, Ha72). In general, the effectiveness of a given dose tends to decrease with decreasing dose rate. In the case of low-LET radiations, the reduced effectiveness of a dose delivered at low dose rates is a consequence of the interaction of a number of factors, most notably the repair of sublethal damage, the redistribution of the cells within the mitotic cycle, and the compensatory cellular proliferation during a protracted exposure. In the case of high-LET radiations, the dose-rate effect is much reduced, at least those components of it that are a consequence of repair and redistribution. These general considerations appear to be equally valid for mutagenesis and carcinogenesis, although there is some evidence that for high-LET radiations, protracting an exposure may lead to an increase in the induction of cancer and mutations (Ha79, 80, He88, Hi84, Vo81, Ul84, and Fr77) in some situations. There is the important practical problem of allowing for dose-rate effects in the analysis of site-specific cancer risks (see Chapter 4). The cumulative knowledge of dose rate factors in experimental radiobiology was summarized in NCRP Report 64 (NCRP80) and has been discussed in several reports of the United Nations Scientific Committee on the Effects of Atomic Radiation (UNSCEAR) concerned with risk estimates for the carcinogenic effects of radiation (UN77, UN86). These reports noted that any value from 2 to 10 for the extent to which a given dose of low-LET radiation may be assumed to decrease in effectiveness at low dose rates, could be rationalized on the basis of experiments with laboratory animals, but suggested a factor of 2.5 for use in risk assessment for human leukemia at low doses and dose rates. They further suggested that this risk be multiplied by 5 to get the risk for all cancers. There are scant human data that allow an estimate of the dose-rate effectiveness factor (DREF). If the apparently nonlinear dose-incidence curve for leukemia in atomic bomb survivors (see Chapter 5) is assumed to reflect a linear quadratic relationship between the incidence and the dose, the contribution of the quadratic dose term can be expected to be reduced at low doses and low dose rates. According to this interpretation, fitting linear and
BACKGROUND INFORMATION AND SCIENTIFIC PRINCIPLES 23 linear-quadratic models to these data, the ratio of the linear coefficients for the two fits yields an estimate of the DREF. This Committee's analysis in Chapter 5 yields a DREF of 2. This compares with the estimate of 2.25 made by the BEIR III Committee, based on essentially the same data set but with the obsolete T65D dose estimates (see Annex 4B). TABLE 1-4 Summary of Dose-Rate Effectiveness Factors for Low-LET Radiation Source of Data Observed Full Limited for Single Best Range of Values Narrow Range Estimate of Values Human leukemia â â 2.1 (present report) BEIR III â â 2.0 to 2.5 Laboratory animal studies Specific locus 3â10 3â7 5 mutation Reciprocal transloc. 5â10 5â7 5 Life shortening 3â10 3â5 4 Tumorigenesis 2â10 2â5 4 The much more extensive animal data include four basic sets from which DREF values can be derived. These include (1) the induction of specific locus mutations, (2) the induction of reciprocal chromosome translocations, (3) life shortening induced by whole-body external irradiation, and (4) tumor induction in small mammals. All of these studies are relevant for the selection of DREF values for estimating human risks for neoplastic disease. Table 1-4 provides a summary of the experimental findings for these categories of radiation injury. The observed full range of values in Table 1-4 closely reflects findings from many individual studies. The upper limit of 10 for all four endpoints is a repeatedly observed value; there are some higher values, but these are not recurring findings. The lower limit depends on exact experimental conditions regarding instantaneous dose rate, protraction period, fractionation pattern, and, for tumorigenesis, the specific type of tumor involved. The narrow range recognizes that the upper limit may include some experimental conditions that are not entirely relevant. For example, the highest values come from studies of the effects of continuous daily irradiation until death, which may be an unlikely circumstance for humans except as a result of natural background radiation. The single best estimate values are appropriate for all low-dose-rate, low-LET radiation exposures delivered intermittently, or even continuously, over periods of months to years. The sixth radiobiological concept is that a variety of chemicals can modify the cell killing effects of radiation. Oxygen and other agents that mimic oxygen by being electron affinic tend to sensitize cells to the effects
BACKGROUND INFORMATION AND SCIENTIFIC PRINCIPLES 24 of a given dose of radiation, while radical scavengers, such as sulfhydryl compounds, tend to protect cells (Mo36, Pal84, Pat49, and Yu80). In general, the redox status of the cell affects its response to radiation. There is little available evidence suggesting that the same considerations apply to mutagenesis and carcinogenesis. The seventh radiobiological concept is that modifiers exist which have little influence on cell killing but may greatly modify the multistep process of carcinogenesis and its in vitro counterpart, oncogenic cell transformation (Ha87). These modifiers include: (1) hormones (Gu80); (2) tumor promoters, that is, agents that do not affect initiation but that dramatically affect the later stages of carcinogenesis in vivo or transformation in vitro (Ke80); (3) protease inhibitors, such as antipain (Bo79, Ke81). These factors, which have little influence on cell lethality, can exert a profound effect on the response to radiation when carcinogenesis, transformation or both, are the endpoints being studied. Indeed, such biological factors can dwarf in magnitude the effect of such physical factors as radiation quality and dose rate. Promoters, for example, can alter the shape of the dose- response relationship and can modify the absolute frequency of transformation produced by a given dose of radiation. This is discussed in more detail in Chapter 3. Differences in Relative Biological Effectiveness (RBE) among Radiations Absorbed dose (which is most often referred to simply as dose) is a physical quantity that, all other things being equal, correlates well with biological effect. However, when the quality of radiation changes, absorbed dose alone no longer specifies biological effect. In other words, a given absorbed dose of x rays, does not necessarily result in the same biological effect as the identical dose of neutrons or alpha particles. To characterize this difference, the concept of RBE was introduced; that is, the RBE of radiation 1 relative to that of radiation 2 is the inverse ratio of the doses of each, (D2/D1) required to produce the same biological effect. When the dose-response relationships for the two types of radiation differ in shape, RBE is necessarily dependent on the level of the effect that is considered and should be specified as such. In the 1963 ''Report of the RBE Committee to the International Commission on Radiological Protection and the International Commission on Radiological Units and Measurements" (ICRP63), the comparison of low-LET or standard radiation was designated as x rays, gamma rays, electrons, or positrons of any specific ionization; and an RBE of unity was assigned to any radiation with an average LET in water of 3.5 keV/Âµm or less. RBE values relative to this standard were then tabulated for a variety of LET values and biological endpoints as a basis for deriving the risk per
BACKGROUND INFORMATION AND SCIENTIFIC PRINCIPLES 25 unit dose of any high-LET radiation relative to the risk per unit dose of the standard low-LET radiation at low doses and dose rates. In the ICRP-ICRU, 1963 report, it was pointed out that knowledge of the RBE of different types of radiation is used in two ways in radiological protection: first, to provide a basis for setting occupational dose limits for high- LET radiation in relation to accepted limits for low-LET radiation (and to allow the reverse procedure for certain bone-seeking isotopes) and, second, to provide a basis for summing the doses of radiations of different qualities to which a person may have been exposed. This latter use of RBE generally has only limited validity, however, since the prediction of biological effects on the basis of doses of different radiations weighted by their RBEs is a correct procedure only if (1) radiations act independently (a condition rarely met), and (2) their dose-response curves are linear. An example illustrating this point is the fact that the biological effect of neutron fields contaminated by various amounts of photons cannot be predicted from knowing the neutron RBE only, except, perhaps, at very low doses. The ICRP-ICRU Report clearly differentiated the radiobiological concept of RBE from that of the quality factor (now designated Q). Conceptually, Q has a meaning similar to that of RBE; however, it was recognized that Q may not necessarily be identical to RBE. Q is defined as the ratio of occupational exposure dose limits, while RBE values are determined experimentally from radiobiological data. Thus, the concept of Q cannot be considered independently of the general philosophy that is to be applied to the derivation of dose limits for different radiations in the context of radiation protection. In dealing with the limited data on RBE then available, particularly on the more relevant endpoints of mutagenesis and carcinogenesis, it was assumed that the dose-response curve for high-LET radiation generally tended to be linear, at least at low doses. For the low-LET standard radiation, discussion oriented largely around the linear quadratic dose-response curve, with an initial linear component dominating at low doses and dose rates. The linear component of the low-LET radiation curve, interpreted as resulting from a single-track mechanism, was thought to be due almost entirely to the high-LET radiation regions at the end of particle tracks. The slope of this linear component of the total dose-response curve was expected to be largely independent of dose rate and dose fractionation. Dose-rate effects were expected only at higher doses, where the dose squared or multitrack mechanisms were associated with the nonlinear component of the overall dose-effect curve. A similar formulation has been used repeatedly in the literature, including a report by the National Council on Radiological Protection and Measurements, NCRP Report 64 (NCRP80).
BACKGROUND INFORMATION AND SCIENTIFIC PRINCIPLES 26 With higher-LET radiations, the initial linear term generally extends to higher doses than those seen with low-LET radiations. Frequently, it is as difficult to demonstrate a quadratic term with high-LET radiations as it is to demonstrate the initial linear term with low-LET radiations. On the basis of the linear-quadratic model, the RBE derived from data obtained at high dose rates would be expected to be highly dependent on dose, with a sharp increase in RBE as the dose decreases (Figure 1-2). With decreasing dose rate, the slope of the high-LET curve would be expected to change only minimally. With low-LET radiation, however, at very low doses or with higher doses at low dose rates (or with a very high degree of fractionation), the curve would ultimately be expected to become linear with a slope equal to that of the linear component of the linear-quadratic dose-response curve. Thus, with the limiting conditions of very low dose, any dose at very low dose rates, or both, the limiting RBE should be equal to the slope of the high-LET dose- response relationship, divided by the slope of the linear term of the linear- quadratic dose-response relationship. This ratio was designated in ICRP-ICRU 63 as RBEm, which is the maximum RBE which is obtained at minimal doses. Thus, emphasis was put on RBE values that were obtained at very low doses, very low dose rates, or both, which were considered to be most relevant to radiation protection. It was made clear by ICRP-ICRU, 1963 that essentially all of the increase in RBE at low doses is caused by a decrease in the slope of the low-LET curve as the dose decreases. This is a basic problem with the current definition of RBE in which low-LET radiation is the "standard" relative to which RBE is evaluated. Currently, the biological effectiveness of all photon and electron radiations are assumed to be the same, although there is experimental evidence that medium energy (200-250 kVp) x rays are twice as effective as Cobalt-60 gamma rays for low doses on the order of 1 rad, at least for some endpoints such as oncogenic transformation and chromosome aberrations (Bo83, Un76, Sc74). Microdosimetric measurements lead to similar conclusions (E172). Factors Affecting RBE Radiation Quality (LET) The current use of LET as a measure of radiation quality is based essentially on (1) its simplicity (easy to calculate, easy to understand), and (2) the recognition that there exists an association between the spatial patterns of energy deposition and biological effectiveness. As such, LET is a reasonable qualitative index for ranking radiations on an ordinal scale of biological effect. For quantitative predictions, however, LET has severe limitations (ICRU83, ICRU86).
BACKGROUND INFORMATION AND SCIENTIFIC PRINCIPLES 27 FIGURE 1-2 Dependence of RBE on dose and dose rate for situations in which a linear- quadratic dose-effect relationship applies. The four curves correspond (from top to bottom) to increasing values of the dose rate. The RBE shown here is representative of such endpoints as chromosomal damage or cell killing. To provide a more adequate description of energy deposition and, implicitly, radiation quality, a number of microdosimetric-based concepts have been developed in the past 20 years. These range from lineal energy (the stochastic counterpart of LET) to distributions of distances between elementary deposits of energy (proximity functions) and radial dose distributions. These quantities are often used in making more successful predictions of RBE as a function of both radiation type and dose. In practical applications the fact remains, however, that they are used only by a restricted group of specialists, so that LET continues to dominate common perceptions of radiation quality (see Glossary). Variation of RBE with LET For charged particles of defined LET in the track segment mode, RBE has been determined as a function of LET, by using monolayers of mammalian cells and scoring cell lethality, mutation, and oncogenic transformation as biological endpoints. In all cases, RBE increases with LET, reaching a maximum at about 100 keV/Âµm, and subsequently falling
BACKGROUND INFORMATION AND SCIENTIFIC PRINCIPLES 28 for higher-LET values. In general, a given LET predicts the same biological effect for a given dose if it is produced by particles with different masses and charges, such as protons, deuterons, or helium ions (Figure 1-3). However, the concept of LET breaks down, and in the case of very heavy particles having an atomic number close to that of uranium, anomalous results have been reported, together with a complex relationship between RBE and LET (Kr82). There is some evidence that, in the same cell system, higher RBE values are found for mutation than for cell lethality, even at the same radiation dose. FIGURE 1-3 Radiobiological effectiveness, RBE, as a function of linear energy transfer, LET, in cells of human origin, with cell lethality or mutation at the HGPRT locus as endpoints (Co77, He88). Variation of RBE with Dose Rate and Fractionation For low-LET radiations, the consensus is that decreasing the dose rate or dividing a given dose into a number of fractions spread over a period of time reduces the biological effectiveness. In most cases, for high-LET radiations such as neutrons, the effect of a given dose is relatively unchanged when the dose rate is lowered or when fractionation is used. In a few important instances, including neoplastic transformation in vitro, carcinogenesis in experimental animals, and mutagenesis, dose protraction by use of a low dose rate or by fractionation actually enhances the biological effectiveness of a given dose (Figures 1-4, 1-5). The overall conclusion is that the RBE of high-LET radiations compared with that of low-LET radiations may be larger for a low dose rate than for a single acute exposure at a high dose rate.
BACKGROUND INFORMATION AND SCIENTIFIC PRINCIPLES 29 FIGURE 1-4 Hypothetical dose-effect curves for high-LET radiation (upper two curves) and low-LET radiation. It is assumed that lowering the dose rate, (dashed line) results in enhancement of the effect for the high-LET field and a decrease in the yield for the low-LET radiation. This situation has been observed in certain transformation experiments. Variation of RBE with the Biological System or Endpoint Used Even for a given dose or dose per fraction, the RBE of a given type of radiation can vary greatly according to the cell or tissue exposed and according to the endpoint scored. At higher doses and with cell lethality as an endpoint, there is a strong tendency for RBE values to be higher for cells and tissues in which the x-ray dose-response relationship has a large initial shoulder and for RBE values to be lower for cells and tissues for which the cell survival curve more closely approximates a simple exponential function of dose. For lower doses and dose rates and with mutation, neoplastic transformation, or carcinogenesis in vivo as an endpoint, a wide range of RBEm values has been reported. Values have ranged from less than 10 to greater than 100. The Need for the Concept of RBE It would be desirable to have human dose-response information, and therefore risk estimates, for somatic and genetic effects for all types of
BACKGROUND INFORMATION AND SCIENTIFIC PRINCIPLES 30 radiations, including x rays, neutrons, and alpha particles. Human risk estimates for low-LET radiations are available for many effects from various populations, including the Japanese atomic-bomb survivors; however, the recent revision of the dosimetry from Hiroshima and Nagasaki essentially negates previous RBE estimates for neutrons obtained from the Japanese data (see Annex 4-2). For neutrons, therefore, human risk estimates must result from a two-step process, namely, low-LET effects data from human studies and RBE estimates from animal experiments. FIGURE 1-5 RBE versus dose for the curves of Figure 1-4. Dashed line, low dose rate; solid line, high dose rate. The body of radiobiological data available indicate that, in principle, RBE increases with decreasing dose, with limiting higher values generally reached at low doses or at low dose rates. This relationship results from the fact that the dose-response for low-LET radiation is often a linear-quadratic function of dose, whereas for neutrons it approximates a linear function of dose. In general, the biological effects of x rays or gamma rays decrease with fractionation or reduction in the dose rate, whereas with neutrons the effectiveness per rad remains the same or even increases as the dose rate is reduced or the time over which the dose is delivered is protracted. For this reason, the RBE is usually quite different for a protracted exposure from that for a single acute exposure.
BACKGROUND INFORMATION AND SCIENTIFIC PRINCIPLES 31 The limiting value of the RBE at low doses or low dose rates varies with the tissue or cell irradiated (Br73, Fi69, Fi71). This has been documented extensively with cell lethality as an endpoint; but there appears to be at least as much variation between systems when carcinogenesis, mutation, or transformation in vitro is the endpoint. The limiting value of the RBE also varies by a factor of about 2, depending on whether x rays or gamma rays are used as the low-LET radiation (Bo83). This is consistent with the difference in microdosimetric spectra that are characteristic of 250-keV x rays, as compared with those which are characteristic of high-energy gamma-rays (E172). There is some evidence, at least in C3H10T1/2 cells, that the RBE of neutrons relative to x rays may depend on the level of tumor promoting agent present, since TPA has a larger influence on the incidence of oncogenic transformation induced by x rays than neutrons (Ha82). RBE was a relatively simple concept when it was first introduced, during an era in which radiobiological experimentation was characterized by measurements of the dose which was lethal to 50% of the laboratory animals (LD50) (Bo78). It has now become a complex quantity as a result of the sophistication of the biological systems that are available. While the RBE is complicated by its dependence on dose and dose rate, there is no prospect, at present, that this useful concept can be dropped. A vast body of additional human data will be needed before the concept of RBE can be replaced. However, selection of an appropriate RBE in a specific situation is often difficult. An intensive review of RBE values from experimental systems, including in vitro studies and studies of carcinogenesis in laboratory animals, leads to the conclusion that, for fission spectrum neutrons, RBE values range from about 2 to greater than 100 (ICRU86). In the analysis of a-bomb survivor data in Chapter 4 of this report, the committee elected to assume a value of 20 for the RBE of bomb neutrons relative to gamma rays for radiocarcinogenesis. This is consistent with the value of Q recommended by national and international groups concerned with radiation protection (NCRP87a, ICRP85). It is also consistent with many experimentally determined RBE values obtained for a variety of tumors in experimental animals, although it was recognized that lower, as well as higher, values have been reported for some neoplasms. EFFECTS OF RADIATION ON GENES AND CHROMOSOMES The Genome The human genome is composed of DNA that is contained principally in the chromosomes and, to a much lesser extent, in the mitochondria. The chromosomes, of which there are 23 pairs, contain about 6 Ã 109
BACKGROUND INFORMATION AND SCIENTIFIC PRINCIPLES 32 pairs of DNA bases (3 Ã 109 per haploid set of chromosomes) and each chromosome includes a single supercoiled molecule of DNA associated with chromosomal proteins. The organization of this material can be visualized microscopically only to a limited degree. With contemporary cytogenetic techniques, fixed chromosomal metaphase spreads reveal 500 or so bands, although refined techniques can reveal about 2,000 bands per haploid set of chromosomes. The total number of genes is unknown but has been estimated to be in the range of 50,000 to 100,000 per haploid set of chromosomes. This genetic material comprises approximately 3,000-4,000 units of recombination (centimorgans). Thus, a visible chromosomal band at a resolution of 500 bands per haploid set, may include 6 Ã 103 kilobase pairs (kb) of DNA, 100-200 genes, and 6-8 centimorgans of recombining genome. The range in gene size is extreme, with some of the order of magnitude of 10 kb, the retinoblastoma gene about 200 kb, and the muscular dystrophy gene almost 2000 kb of DNA. The parts of genes translated into proteins constitute a minority of total DNA, with many proteins being coded for one kb or so of DNA. Some of the untranslated DNA is important in the regulation of gene expression, while much DNA seems to be extragenic and of unknown function. Not only does the genome recombine in each generation but it can also undergo mutation, a term applied here to denote all changes in chromosomes, their genes, and their DNA. Thus, alterations in chromosome number and structure are included, as are changes that are not visible microscopically. These latter, submicroscopic changes include deletions, rearrangements, breaks in the sugar-phosphate backbone, errors in DNA replication, and base alterations. Most mutations occur during cellular replication. Mutation occurs in both germ cells and somatic cells, although it is much less apparent in somatic cells unless the mutation occurs during tissue proliferation, as happens with some congenital defects and with cancer. On the other hand, many mutations in the germ line are lethal during embryonic development. Thus, the same mutation might be more common in somatic cells than in germ cells because of the lack of tissue-specific selection against it. Chromosomal Abnormalities Three classes of chromosomal abnormalities are known to occur in both germ cells and somatic cells. The best known changes in the germ line are those that affect chromosomal number. Thus, Down syndrome is the result of a mutation in which a parental (usually maternal) germ cell acquires two copies of chromosome 21 as a result of chromosomal nondisjunction during gametogenesis. Fertilization by a normal sperm then yields a zygote with 47 chromosomes. Such trisomy is common at
BACKGROUND INFORMATION AND SCIENTIFIC PRINCIPLES 33 conception, although trisomy (and monosomy) for most chromosomes is invariably lethal to the embryo. The cause of the increase in trisomy with advancing maternal age has focused on differences between male and female gametogenesis. In the female, oogonial mitoses occur during fetal life, and maturation of eggs proceeds to the dictyotene stage, where it is arrested until the time of ovulation. Eggs in a 40-year-old woman have been at this stage for twice as long as in a 20-year-old woman. In contrast, male gametogenesis continues without interruption from puberty to death. Changes in chromosome number can also occur in somatic cells, although the frequency is difficult to estimate because of selection against monosomic and trisomic cells. However, in cancer cells such changes are common. A second class of chromosomal abnormality is the chromosomal break. When a chromosome break occurs in the cell cycle before DNA replication (G1 or early S phase), it will be observed at the following mitosis as a chromosome break (both chromatids are broken). If the break occurs later in the S phase or in the G2 phase, it will be observed as a chromatid break. For each such break that is observed, there may be many others that rejoin and are not observed. Single breaks, both chromosomal and chromatid, are readily induced by ionizing radiation, and their number increases linearly with dose. A third class of visible chromosomal abnormality is the structural rearrangement, which embraces unstable forms, such as rings and dicentrics, and stable forms, including interstitial deletions, inversions, and translocations. These result from the inappropriate joining of two breaks at different sites. The number of these aberrations is generally proportional to the square of the x-ray dose, since two events are necessary. However, there is also a linear component, because a single densely ionizing tail of a particle track can produce both events, so that a linear-quadratic equation more properly describes the dose- response relationship (see Figure 1-6). At low doses only the linear term dominates. Neutrons, on the other hand, because they are more densely ionizing particles, often produce two breaks as the result of a single event, so the dose- response relationship is more nearly linear. At low doses, neutrons are much more biologically effective; i.e., the RBE of neutrons relative to that of x rays is significantly greater than unity. The frequency of two-break aberrations in human lymphocytes irradiated in culture approximates 0.1 aberration per cell per Sv in the low-to-intermediate dose range (L181). The frequency of such aberrations is increased correspondingly in radiation workers, as well as in accidentally or therapeutically irradiated persons, in whom it may serve as a biological dosimeter (L181; IAEA86). Since chromosome aberrations are preponderantly deleterious to the cells in which they occur, the affected cells tend to be gradually eliminated with time.
BACKGROUND INFORMATION AND SCIENTIFIC PRINCIPLES 34 FIGURE 1-6 Frequency of dicentric chromosome aberrations in human lymphocytes irradiated in vitro in relation to dose, dose rate, and quality of radiation (L181). Although chromosome aberrations can be induced by relatively low doses of radiation, only a small percentage of them is attributable to natural background radiation. The majority result from other causes, including certain viruses, chemicals, and drugs. The health implications, if any, of an increase in the frequency of such aberrations in circulating lymphocytes is uncertain. All of these classes of chromosomal abnormalities (non-diploid number, breaks, and structural rearrangements) occur as either germ line (constitutional) mutants or somatic mutants. The Down, Turner, and Klinefelter syndromes are all examples of abnormalities in chromosome number. Many examples of disease-specific constitutional deletions and rearrangements are known. There are no examples of constitutional breaks in all cells examined, but there are about 18 known heritable fragile sites, in which breakage at a specific site can be elicited under certain in vitro
BACKGROUND INFORMATION AND SCIENTIFIC PRINCIPLES 35 conditions, such as folate deficiency (He84). In addition, there are three recessively inherited conditions in which chromosomal breakage and rearrangement occur, namely, ataxia telangiectasia (AT), Fanconi's anemia (FA), and Bloom's syndrome (BS) (He87, Sc74). All three predispose a person to cancer. Patients with AT are unusually sensitive to ionizing radiation, as are their cells in vitro. Cells from patients with BS show a high rate of quadriradial figures, which are caused by homologous chromosomal exchanges, and a high rate of sister chromatid exchanges. A fourth recessive disorder, xeroderma pigmentosum (XP), is not associated with spontaneous chromosomal breakage, but it does predispose a person to chromosomal aberrations induced by ultraviolet light. XP predisposes a person to ultraviolet radiation-induced skin cancers. Somatic chromosome abnormalities can be found at a low rate in the general population, but they are found almost universally in cancer cells. Abnormalities of both number and form are typical. Cancer cells generate abnormalities at an increased rate, but some of them are so specific that they are regarded as being important in the origin of cancer (Ro84). Thus, about 90% of patients with chronic myelocytic leukemia have an aberration known as the Philadelphia chromosome in their leukemia cells. The Philadelphia chromosome is a reciprocal translocation between chromosomes 9 and 22. Every person with Burkitt lymphoma shows a translocation between chromosome 8 and chromosomes 14, 2, or 22; again, this is confined to the tumor cells. Several other tumor-specific translocations are known. Monosomy for chromosome 22 is common in people with meningiomas. Deletions of various chromosomes are found to be associated at a high frequency with certain cancers; e.g., deletion of the short arm of chromosome 3 (3p-) in persons with small-cell carcinoma of the lung and renal carcinoma; 1p-in persons with neuroblastoma; 11p-in persons with Wilms' tumor; and deletion of the long arm of chromosome 13 (13q-) in persons with retinoblastoma and osteosarcoma. There are also two other kinds of aberrations: homogeneous staining regions and double minute chromosomes; these are found in certain cancers, especially neuroblastoma and small-cell carcinoma of the lungs, and do not occur constitutionally. The most compelling evidence that a specific aberration may be causal for cancer can be seen in retinoblastoma and Wilms' tumor; that is, persons are predisposed to these tumors if they inherit the same type of constitutional deletion (at chromosome band 13q14 or 11p13, respectively) as is found confined to the tumor cells in other cases. This finding suggests that both the hereditary and the nonhereditary forms of these tumors are initiated by an abnormality at the same chromosomal site, with the abnormality being a visible chromosomal deletion in some cases and a submicroscopic mutation in others (Kn85).
BACKGROUND INFORMATION AND SCIENTIFIC PRINCIPLES 36 DNA Abnormalities Abnormalities in chromosome number are not necessarily associated with structural changes in DNA, but chromosomal breaks and aberrations involve such changes, as do the many mutations that are not visible microscopically. The mechanisms by which mutations are caused have of course, been of considerable interest. Some in vitro studies with DNA provide an example of the changes that can occur even at 37Â°C. For example, one of the most frequently noted changes is deamination of cytosine, in which cytosine is converted to uracil (Li74). Uracil then pairs with adenine instead of guanine, so the coding sequence is changed following replication. Deamination of adenine, although less frequent, also leads to mutation, because the product, hypoxanthine, pairs with cytosine instead of thymine (Li72). Another important change concerns the methylation of guanine, which may be caused by the presence of the active methyl donor S-adenosylmethionine (Ry82). This change alters both the geometry and the base pairing of guanine. Two products of thymine, the cyclobutane pyrimidine dimer and 6,4-pyrimidine-pyrimidone, which are produced by ultraviolet irradiation, distort the DNA helix (Mi85). Mutations would occur at much higher rates than are actually observed, if it were not for the existence of repair mechanisms. In the case of the thymine photoproducts noted above and in the case of bulky adducts of DNA with certain chemicals, repair proceeds via sequential steps, the first being a cutting of the abnormal strand of DNA on each side of the site of the abnormal nucleotide by an endonuclease. This leads to deletion of a DNA segment that includes the dimer or adduct. The gap, which may be enlarged by an exonuclease, is then filled by a polymerase-catalyzed DNA strand that is complementary to the intact strand of DNA. The final reaction is closure at the growing end by a ligase. This is the classical excision repair pathway first described for bacteria. It is a relatively slow process, but it is very accurate. Recognition of the DNA repair pathway came in humans with the discovery that the disease xeroderma pigmentosum involves a defect in excision repair (C168). This was the first known example of a DNA repair defect in humans. It is thought to account for the propensity of individuals with this disease to develop cancer of the skin, because ultraviolet light induces thymine photoproducts in the exposed skin cells. If not excised, these thymine photoproducts in turn impair faithful DNA replication, causing induced mutations and chromosome aberrations at an increased rate, as has been observed in vitro. Presumably, these mutations may occur in one or more ''cancer genes" that are involved in carcinogenesis in skin cells and melanocytes. Many spontaneous and induced mutations do not affect the gross configuration of DNA. Such mutations include those resulting from the
BACKGROUND INFORMATION AND SCIENTIFIC PRINCIPLES 37 removal, destruction, or mutation of bases; destruction of deoxyribose residues; and breakage of DNA chains. Such damage, which is common with exposure to ionizing radiation, is also corrected by excision repair, but an array of specific enzymes different from those employed in the classical mechanism is used (Li82). These mechanisms are much faster, but less accurate, so residual mutation is more likely. DNA chain breaks are, of course, associated with all chromatid and chromosome breaks. The dose-response curves for single-strand and double-strand breaks may both be linear with x rays, apparently because the former are caused by single ionizations and the latter are caused by the dense tails of ionization tracks. Most chain breaks are repaired following modification of the break termini, filling the defect with polymerase activity and ligation. It may be that the same ligase can function in both slow and fast repair processes. It has recently been reported that ligase deficiency is a feature of Bloom's syndrome (Ch87, Wi87). This would explain the propensity for chromosome breakage and aberration found in patients with that syndrome. It would also explain the increased mutation rate that has been reported in vitro (Vi83) and recently in vivo (La89). An important kind of damage to DNA, and one frequently produced by ionizing radiation, is removal of a base, with the formation of an apurinic or an apyrimidinic (AP) site (Li82). This damage can be repaired by an AP endonuclease that excises the remaining deoxyribose phosphate. There are reports that some cases of xeroderma pigmentosum and ataxia telangiectasia may have reduced AP endonuclease activity. After the creation of AP sites, the AP site itself can be mutagenic if the sites are not removed by AP endonuclease. During the next round of cell division, DNA polymerase may copy past the AP site by inserting a purine, usually adenine, without regard to what is present opposite the site in the other strand. This kind of repair is obviously prone to error. Alterations in DNA caused by deamination of cytosine or adenine and by disruption of purine or pyrimidine rings can also be repaired. The mismatched or degraded base is removed by one of several specific glycosylases, enzymes that are relatively abundant and rapidly acting, leaving an AP site, which is then handled by AP endonuclease as noted above (Li82). While genetic defects in these enzymes are not known in humans, bacterial mutants lacking uracil glycosylase show considerably altered mutation rates. One other alteration in DNA is processed in a unique way. As noted earlier, methylation of guanine (of an oxygen atom at position 6) may occur under physiological conditions, but it is also produced by certain alkylating agents. An unusual enzyme has been discovered that removes this methyl group and transfers it to a cysteine residue of the enzyme itself, restoring the DNA to its normal configuration, but inactivating the enzyme in the process (Ha83). This methyl transferase is literally a suicide enzyme; in
BACKGROUND INFORMATION AND SCIENTIFIC PRINCIPLES 38 fact, it is not strictly an enzyme because it is not regenerated. No inherited defect has been reported for this enzyme in humans, but it has been found that in some cancer cells the function of this enzyme is defective. It may be that such cancer cells undergo further mutations relevant to tumor progression more readily. If so, they should be more susceptible to killing by alkylating agents. Conclusions Although the human genome is highly stable in both germ-line and somatic cells, errors do occur in its transmission from one generation of individuals or cells to the next. These errors occur at a spontaneous rate that can be increased by environmental agents, including radiation. These errors can be so macroscopic that they are detectable cytogenetically, as in the case of abnormalities of number or structure of chromosomes. Other errors cannot be detected cytogenetically, but can be detected as changes in the nucleotide sequence of a gene. Many such errors (mutations) are repaired. The importance of the existence of repair mechanisms is underscored by the predisposition to cancer that is associated with some rare hereditary disorders in which one of these repair mechanisms is defective. INTERNALLY DEPOSITED RADIONUCLIDES: SPECIAL CONSIDERATIONS Exposure to ionizing radiation occurs from radionuclides deposited within the body as well as from sources outside the body. Differences in the characteristics of these two types of exposure must be considered when interpreting studies of irradiated populations and estimating the possible health effects of different patterns of irradiation. With an internally deposited radionuclide, the radionuclide enters the body at the time of exposure but the doses it delivers to various organs and tissues of the body continue to accumulate until the radionuclide is removed by physical or biological processes. Thus, the radiation is delivered to various organs gradually, at changing dose rates, over what may be an extended range of ages. An internally deposited radionuclide also frequently produces nonuniform irradiation to the organs and tissues in which or near which it is incorporated, depending on its radioactive emissions and metabolic characteristics. In this respect, the spatial and temporal patterns of the doses delivered by internally deposited radionuclides differ from those typically delivered by external irradiation (Figure 1-7). These and other differences in both dosimetry and biological response have a direct impact on the characteristics of the resulting dose-response relationships. Accordingly, any quantification of human health risks from
BACKGROUND INFORMATION AND SCIENTIFIC PRINCIPLES 39 FIGURE 1-7 Temporal patterns of dose distribution.
BACKGROUND INFORMATION AND SCIENTIFIC PRINCIPLES 40 exposure to ionizing radiation must consider, first, the determination of risk factors for exposure situations in which adequate data on dose and response are available, and second, the relative importance of various dosimetric or response factors that can alter the resulting risk estimates. This applies to both external and internal irradiation conditions. In this section, the general characteristics of exposure and dose-response relationships are discussed for internally deposited radionuclides as they apply to estimation of site-specific radiation-induced cancer risks in exposed human populations (see Chapter 4). The health effects of radon progeny and other internally deposited alpha- emitting radionuclides were examined in depth in the BEIR IV report (NRC88). Radionuclide Dose-Modifying Factors The intake of radionuclides can occur by inhalation, ingestion, injection, and absorption through the skin and mucous membranes or through cuts and abrasions (ICRP79). The relative importance of these different routes of intake depends on the particular exposure situation considered, for example, environmental or occupational exposure, accidental exposure, or medical administration of radionuclides. Each of these exposure routes has its own characteristic pattern of initial deposition on or in various parts of the body such as the lungs, the gastrointestinal tract, or skin. As long as a radionuclide is present at one of these sites of intake, the surrounding tissue will be irradiated, and the extent of this irradiation will be determined by dosimetric factors (see the section on physics and dosimetry earlier in this chapter and see below). A portion of the radionuclide present at these sites of intake may dissolve and be absorbed into the blood. Once uptake to body fluids has occurred, the radionuclide will be deposited in other organs and tissues, depending on its physical and chemical properties. Chemical, physical, and biological processes can also influence the effective retention time for a given radionuclide, thereby influencing the period of time during which the irradiation of the surrounding tissues occurs. The description and quantification of the deposition, retention, and excretion of internally deposited radionuclides are generally well understood. The most extensive reviews of metabolic and dosimetric data for the different radionuclides currently available are those given by the International Commission on Radiological Protection (ICRP) Publication 30 (ICRP79). Additional information on the dosimetric approaches incorporated in the current ICRP system is available in reports by Johnson (Jo85) and the National Council on Radiation Protection and Measurements (NCRP85). The methodology and values given by ICRP were assembled for radiological protection planning purposes. Thus, the values chosen for the various
BACKGROUND INFORMATION AND SCIENTIFIC PRINCIPLES 41 parameters are conservative; that is, they can lead to overestimates of risk factors. These values may not be appropriate for estimation of risk when the organ and tissue doses received by exposed individuals are considered. Some of the relevant data have been derived from human studies, in particular studies on the deposition of inhaled particles and gases (e.g., radon progeny in uranium miners), whole-body retention of radionuclides with emissions that are detectable outside the body (e.g., radiocesium from worldwide fallout), and excretion (feces and urine) samples collected primarily in occupational exposure situations (e.g., transuranic radionuclides). Concentrations of radionuclides in some tissues have also been measured at autopsy. The remainder of the data have been and continue to be obtained from studies of various species of laboratory animals conducted under controlled laboratory conditions. The study of laboratory animals makes it possible to examine radionuclide biokinetics and metabolism, for which human data are sparse or nonexistent, and to determine the effects of various modifying factors on the resulting dosimetry (NRC88). Each laboratory animal species has its own anatomic and physiological characteristics that need to be considered when the resulting dosimetric parameters are extrapolated to human exposure situations. For instance, the mechanical clearance of insoluble particles from the pulmonary region is strongly species-dependent; mice and rats clear these particles by mucociliary activity much more rapidly than do guinea pigs, dogs, or humans (Sn83, Sn84). Knowledge of these differences is necessary for appropriate dose calculations in studying dose-response relationships in different species as surrogates for humans. Similarly, the hepatic turnover of actinide and lanthanide radionuclides in mice and rats is considerably faster than that in dogs and nonhuman primates (ICRP86, Bo74). Other factors that need to be considered when determining the dose received by critical cells include an identification of the target cells of concern and how the patterns of cellular irradiation are influenced by nonuniform radionuclide deposition or clearance, age, and health status (Sm84, Fi83). Radionuclide Response-Modifying Factors There are only a few groups of human subjects with radionuclide burdens of sufficient magnitude to produce long-term biological effects. Major groups in this category include patients treated with 226Ra, 224Ra, Thorotrast (232Th and progeny), or 131I; uranium miners exposed to 222Rn and its progeny; and uranium workers exposed to 238U, 235 U, and 234U. All of these study populations, except those exposed to 131I, involve people exposed to high-LET (alpha) radiations and were discussed in detail in the BEIR IV report (NRC88).
BACKGROUND INFORMATION AND SCIENTIFIC PRINCIPLES 42 With the exception of the special case of exposure of the human thyroid to 131I,discussed in Chapter 5, long-term biological effects of internally deposited low-LET-emitting radionuclides have not been observed in human populations. Estimations of the potential health risks of such radionuclides must be sought by other means. To provide such data, a large number of life-span studies on the effects of radionuclides have been conducted in laboratory animals. Major studies currently in progress include those examining the effects of inhaled 239PuO in baboons (Me88), inhaled 238PuO or 239PuO in dogs (Pa86, Mc86), 2 2 2 inhaled 239Pu(NO3)4 in dogs (Pa86), inhaled fission products (90Sr, 144Ce, 137Cs, 91Y, and 90Y) in relatively soluble or insoluble forms in dogs (Mc86), injected 226Ra, in dogs (Go86), and intravenously injected 239Pu, 226Ra, 228Th, 228Ra, and 90Sr in dogs (Wr86). Comparative life-span studies involving large numbers of rats exposed to low doses of high-or low-LET radiation include studies of inhaled 239PuO2 (Sa88), inhaled 144CeO2 (Lu87a) and thoracic or whole-body x- irradiation (Lu87b). These studies should provide a critical link between observations on laboratory animals and existing human data. It is expected that such studies, many of which are currently nearing completion (Th86), will contribute to our understanding of the relative importance of possible risk modifiers such as dose, dose rate, nonuniformity of dose distribution, species, age, health status, and exposure to other carcinogenic agents in combination with radiation. USE OF ANIMAL STUDIES Observations on the biological effects of ionizing radiation began to be made soon after the discovery of x rays in 1895. Already in 1896, there were reports of dermatitis and alopecia in those experimenting with x-ray generators (Fu54). By 1902-1903, the first reports had appeared describing skin carcinomas on the hands of radiologists, and less than a decade later, sarcomas had been induced in rats by repeated exposure to irradiation (Fu54). In 1906, from studies of radiation effects on the testes of goats, J. Bergonie and L. Tribondeau formulated their well-known generalization that: X-rays are more effective on cells which have a greater reproductive activity; the effectiveness is greater on those cells which have a longer dividing future ahead, on those cells the morphology and function of which is least fixed (Translation by G.H. Fletcher, Be06). Two decades later, H. J. Muller reported the mutagenic effects of radiation on the germ cells of Drosophila melanogaster (Mu28). During the 1920s congenital abnormalities were recognized in children whose mothers had been irradiated while they were pregnant (Go29), and in the following decades radiation teratogenesis was widely investigated in mice and rats
BACKGROUND INFORMATION AND SCIENTIFIC PRINCIPLES 43 (Ru54). Research in these areas and on the systemic cellular and molecular effects of ionizing radiations have continued in a variety of animal species, in parallel with continued observation on radiation effects in humans. In many respects the human data and the animal data are complementary. There are several important areas in which the human data are inadequate for risk estimation and must be interpreted in the light of concepts developed from experiments with animals. In particular, information from experimental animals is useful for human risk assessment in the following areas: 1. prediction of the effects of high-LET external radiations, including neutrons; 2. prediction of the effects of low or varying dose rates and of various patterns of fractionation of exposure to low-LET radiations, high- LET radiations, or both; 3. clarification of the mechanisms of radiogenic damage including mutagenesis, carcinogenesis, and developmental effects; this is crucial to the development of appropriate interpretations and mathematical models of radiation effects in humans; and 4. prediction on the uptake, distribution, retention, dose distribution, and biological effects of internally deposited radionuclides for which there are inadequate data in humans. The validity of quantitative extrapolation from animals to humans is of great concern. Such a procedure may be defined better as the "transposition" of concepts and parameters derived from animal studies to humans in order to compensate for inadequate or unavailable information. Opinions vary about appropriate methods for extrapolating data and concepts between species, but there are times when it is essential. It is unlikely that humans are so physiologically unique among mammalian species as to invalidate selective use of animal data. Consideration has been given recently in two areas to direct extrapolation of dose-incidence ratios for carcinogenesis from experimental animals to humans. The first of these includes the use of ratios of the relative effectiveness of two internally deposited radionuclides in animals in order to estimate the relative risks of the nuclides for man when human data are available on only one of the nuclides. This was examined in the BEIR IV report (NRC88). The second is the direct application of the relative risk (per Gy) of cancer in animals to prediction of the relative risk of cancer in irradiated humans (St88). Much of the information on radiation-induced and spontaneous mutation rates in humans is based on chromosomal aberrations and specific locus mutations in somatic cells, the latter primarily in culture systems. Estimations of human genetic risk are thus made in the light of dose-response
BACKGROUND INFORMATION AND SCIENTIFIC PRINCIPLES 44 relationships and mechanistic considerations derived from experimental studies on inherited genetic effects, primarily in mice, and interpreted in the light of the large body of data from other biological systems. Radiation-induced mutation is a process which is completed within a relatively short interval after exposure. Interspecies extrapolation of experience with dose-mutation effects can therefore be done with somewhat greater confidence than can comparable extrapolations of effects on multistage prolonged processes such as carcinogenesis and life shortening. There are extensive experimental data concerning radiation effects on embryogenesis with specific reference to the development of gross abnormalities of the central nervous system and disruption of neuroblast proliferation, migration, differentiation, and establishment of neural pathways. Measurements of effects of exposure during embryogenesis on neurological function, including learning capacity and cognition, are less common and more difficult to perform in experimental animals. Although interpretation and application of the experimental data to human risk estimation requires careful comparison of equivalent developmental stages, the data are valuable in complementing sparse human information. EPIDEMIOLOGICAL STUDIES: SPECIAL CONSIDERATIONS Epidemiologic studies are a critical tool in assessing radiation risks, since they alone provide data directly applicable to humans. However, epidemiologic studies of individuals exposed to radiation have methodologic limitations which should be kept in mind when assessing the results of such studies. This section briefly summarizes these concerns. Further discussion of these issues can be found in standard textbooks on epidemiologic methods (Ma70, Ro86). Most epidemiologic studies of low-LET radiation have focused on cancer as the outcome. This discussion of epidemiologic methods and their limitations also focuses on cancer, although most of the considerations also apply to studies of other outcomes. High-Dose Studies The use of high-dose studies to quantify risk estimates involves a two- stage process. First, risk parameters that apply to the particular high-dose group under observation must be estimated from the empirical data. Second, mathematical models must then be used to extrapolate from the experience of the specific high-dose population to that of the low-dose population of interest, taking into account differences both in exposure factors such as dose and dose rate and host factors such as age, sex and race. Both steps are, of course, subject to error, and the assumptions and limitations involved in the second step will be discussed in detail later in
BACKGROUND INFORMATION AND SCIENTIFIC PRINCIPLES 45 this chapter. The problems and limitations involved in the first step are discussed here. Studies reported to date have essentially been of the retrospective cohort type. Populations receiving high doses of low-LET radiation are rare, and exposure to such doses is unlikely to occur in the future, apart from the therapeutic irradiation of patients. Such studies are subject to both sampling variability and bias. Sampling variability should generally be adequately expressed by the confidence intervals around the parameters estimated by the particular mathematical model, but bias represents a greater problem. Biases in epidemiology are generally classified as resulting from selection, information, or confounding. Selection bias can be defined as arising from any design problem that tends to make the study subjects unrepresentative of their source population. Such a bias can prevent generalization of the results. For example, if the survivors of the atomic bombings at Hiroshima and Nagasaki were healthier than the general population, their susceptibilities to radiation carcinogenesis could be different from those of the general population. In addition, selection may lead to internally biased results when the follow-up is selective. This occurs when those individuals selected for follow-up are different for differing categories of exposure and when that difference is associated with a differing underlying cancer risk. For example, if only 50% of the atomic-bomb survivors had been followed, and there were more smokers in the high-dose group that were followed than in the low-dose group, there would be an excess of lung cancer in the high-dose group that was not caused by radiation. Such a selection bias is likely to occur only when there is substantial loss to follow-up. It is unlikely that this plays a role in the major high-dose epidemiologic studies on which risk estimates are currently based, since follow-up has been essentially complete for these studies. Information bias, which refers to any process which distorts the true information on either exposure or disease status, is likely to be of more importance than selection bias. Misclassification of exposure is likely to be a major potential source of error in making risk estimates. Nondifferential misclassification with respect to exposure level leads to an underestimation of risk and tends to reduce any upward curvature in the dose-response relationship. This occurs, for example, when the distribution of errors in dose estimates is the same in the diseased and the nondiseased, as will generally be the case for most cohort studies. Other biases may be more subtle. Misclassification of disease status is particularly important when such status is determined from death certificates which are often unreliable for a number of cancer types. These errors are more likely to be differential, i.e., dependent upon a subject's exposure status, and could bias a dose-response curve away from the null.
BACKGROUND INFORMATION AND SCIENTIFIC PRINCIPLES 46 Finally, confoundingâi.e., distortion of risk estimates due to the association of both exposure and disease with some other covariate, such as smokingâis unlikely to be of substantial importance in affecting risk estimation based on comparison of groups of individuals with varying degrees of exposure, but it could be of importance when an unexposed control group is also used in the estimation procedure. For example, the characteristics of the ''not in the city" group in the Japanese atomic-bomb survivor study may be somewhat different from the group exposed to the radiation, and if these characteristics are associated with differing cancer risks, such confounding would have an effect on the risk estimates. This may be a particular problem with studies of patients irradiated for medical conditions if risk estimation is carried out with an unexposed comparison group, such as the general population: the condition for which irradiation is used could well be associated with an altered cancer risk. The three types of bias discussed above could all play roles in affecting the internal validity of risk estimates (i.e., the validity of the results for the particular population being studied). However, even in the absence of such biases, there remains a fundamental problem in extrapolating the risks from one population to another, for example, from the Japanese to North Americans. The method of such extrapolation depends on the mathematical model chosen; and, although empirical evidence may be available from studies carried out in both countries, there often is considerable uncertainty about the validity of the procedure that is used. The quantitative risk estimates developed in Chapter 4 of this report are based primarily on extrapolation from studies of populations exposed to high doses of radiation over relatively short periods of time. The rationale for this approach is that only these studies provide sufficiently precise estimates of risk at any dose. Risk estimates for low doses and protracted exposure could therefore be in error because of (1) an inappropriate mathematical model, or (2) biases in the high-dose epidemiologic studies used to estimate the parameters of the chosen model, as discussed above. The committee has attempted to mitigate the first problem by using sufficiently general model classes that include most of the widely accepted alternatives and by providing estimates of the range of uncertainty in the estimates. In general, the estimates of risks derived in this way for doses of less than 0.1 Gy are too small to be detectable by direct observation in epidemiologic studies. However, it is important to monitor the experience of populations exposed to such low levels of radiation, in order to assess whether the present estimates are in error by some substantial factor. Low-Dose Studies A number of low-dose studies have reported risks that are substantially
BACKGROUND INFORMATION AND SCIENTIFIC PRINCIPLES 47 in excess of those estimated in the present report. These include risks to populations exposed to high background levels of radiation, diagnostic x rays and fallout from nuclear weapons testing or nuclear accidents, and to individuals with occupationally derived exposures. Some of these studies are discussed in more detail subsequently. Although such studies do not provide sufficient statistical precision to contribute to the risk estimation procedure per se, they do raise legitimate questions about the validity of the currently accepted estimates. The discrepancies between estimates based on high-dose studies and observations made in some low-dose studies could, as indicated above, arise from problems of extrapolation. An alternative explanation could be inappropriate design, analysis, or interpretation of results of some low-dose studies. This section discusses the particular methodologic problems which can arise in such studies, and the section on low-dose studies in Chapter 7 summarizes a number of these studies and assesses their results, taking into account the methodological limitations discussed here. The problem of random error caused by sampling variability is relatively more important for low-dose than for high-dose studies. (Sampling variation means the range of results to be expected by exact replication of the study, if this were possible; its major determinant is sample size and its distribution across exposure and disease categories.) To understand why this is so, suppose that two studies were conducted, one in a population exposed to 1 Gy and one in a population exposed to 0.01 Gy, in which similar sample sizes and designs were used, and suppose that the resulting standard errors on the log relative risk were the same. Thus, suppose the relative risk in the high-dose population was 11 with 95% confidence intervals of 5.5 and 22 and the relative risk in the low- dose population was 1.1 with confidence intervals of 0.55 and 2.2. The point estimates on the relative risk coefficient from the two studies would be identical at 10/Gy, but the confidence intervals on the high-dose estimate are 4.5 and 21 and on the low dose estimate are â4.5 and 12.0. This comparison emphasizes the importance of considering sampling variability in assessing the results of low-dose studies. In fact, the problem of sampling variation is even more serious than this simple example would indicate. The standard error of the relative risk in a simple 2 Ã 2 table of exposure by disease status is determined primarily by the size of the smallest cell in the table, which is usually the number of exposed cases. In most studies of low-dose effects, this cell may be quite small, so the resulting standard error is larger than that for high-dose studies, even if the overall sample sizes were the same. In general, systematic biases are also relatively more important for the objectives of low-dose studies than they are for those of high-dose studies. Because of the existence of more and larger populations exposed to low doses, low-dose studies are often ecological (correlational) or case-control
BACKGROUND INFORMATION AND SCIENTIFIC PRINCIPLES 48 studies rather than cohort studies. The ecological and case-control studies are particularly prone to bias in their design. Selection bias is a major potential problem in case-control studies: the major concern is over the appropriateness of the control group. This is a particular problem for those studies in a medical setting. Information bias leading to misclassification of either exposure or disease status, if random, leads to underestimated risk, and several low-dose studies could well involve substantial systematic misclassification, for example, misclassification because of recall bias by cases in case-control studies. Similarly, tumors which can be induced radiogenically could be overestimated in radiation-exposed individuals. Confounding may be more important for low-dose than for high-dose studies. An observed relative risk of 2 is much more likely to be produced solely by confounding than a relative risk of 10 (Br80). The possibility of confounding can only be judged on a study-by-study basis, but some generalizations are possible. Ecological correlation studies, such as the studies of areas with high levels of background radiation, are probably the most susceptible to confounding. Residents of areas with high levels of background radiation are likely to differ in many ways from those in areas with low levels of background radiation. This could affect cancer rates, but data on the relevant characteristics are unlikely to be available for analysis. As an example, exposure to radiation from terrestrial sources may vary with housing structure, which, in turn, may reflect a socioeconomic status that correlates with such factors as smoking and alcohol use. This possibility alone generally makes such studies uninterpretable, and when the ecological fallacy discussed below is also considered, these two problems alone are enough to make such studies essentially meaningless. Case-control studies, on the other hand, generally offer the greatest opportunity to control for confounding by matching or obtaining information on definable covariates for use in analysis. However, the extent to which this has been done varies from study to study. It is necessary, of course, to collect data on such confounders, and, if the confounders are not recognized in advance, the appropriate data may not be available. Finally, three other potential biases of low-dose studies should be mentioned (Be88). The first is the ecological fallacy, that is, that in correlational studies, any excess risk occurring in a population with increased exposure may be occurring in individuals other than the individuals who are actually receiving the excess exposure. Second, is the possibility of selective reporting. Epidemiologists are more likely to report and journal editors are more likely to accept positive findings than null findings. Thus, information in the literature on populations exposed to low doses of radiation may be slanted in favor of those studies that show higher risks than the conventional estimates, since those that show estimates consistent with the
BACKGROUND INFORMATION AND SCIENTIFIC PRINCIPLES 49 accepted values would not be seen as significant. The magnitude of this potential effect is unquantifiable, but it almost certainly exists and plays a role in the plethora of low-dose studies with a reported positive risk. Third, there is the problem of multiple comparisons. This arises if a number of tests of significance are made with respect to elevated risks for a number of cancer sites. Such a process invalidates the conventional value quoted for the test of significance and leads to more significant results than nominally would be expected by chance. For example, in following a cohort of occupationally exposed individuals, if comparisons are made for 10 cancer sites with a p value of 0.05, which nominally would be expected 5% of the time by chance for a single comparison, significant excesses would arise 40% of the time by chance for at least one of those outcomes. Interpretation of such results must be guided by prior hypotheses, and by consistency of results among studies, a major criterion for causality. RISK ASSESSMENT METHODOLOGY Need for Models in Risk Assessment One of the major aims of this report, as of previous BEIR reports, is to provide estimates of the risks of cancer resulting from various patterns of exposure to ionizing radiation. In principle, such estimates could be derived by identifying a group of individuals with similar exposures and similar backgrounds and following them to compare the proportion of the group who eventually developed cancer with the proportion who developed cancer in a comparable unexposed group or in the general population. For situations in which it is not possible to measure the risks directly, statistical models must be used to derive estimates. Large sample sizes are needed in any such comparisons, to minimize random variation; the rarer the disease and the smaller the effect of exposure, the larger the sample needs to be. For example, the BEIR III report estimated that a single exposure to 0.1 Gy (10 rads) of low-LET radiation might cause, at most, about 6,000 excess cases of cancer (other than leukemia and bone cancer) per million persons, as opposed to a natural incidence of about 250,000. To identify this number as a statistically significant excess, a cohort of about 60,000 people with the same exposure would have to be followed for a lifetime, or an even larger number of people would have to be studied if follow-up were for a shorter period of time. Under ideal conditions, a case-control study to identify the same excess would have to consist of at least 120,000 cases and 120,000 controls. It is unlikely that such large groups with similar exposures could be identified, let alone feasibly studied. Furthermore, even if the random variation could be overcome by the large sample sizes needed, estimates of such small
BACKGROUND INFORMATION AND SCIENTIFIC PRINCIPLES 50 excess risks (2%) could easily be biased by confounding, misclassification, or selection effects. Epidemiologists generally agree that excess risks of less than 50% are difficult to interpret causally (Br80). In practice, therefore, it is necessary to obtain risk estimates by extrapolation from smaller and less homogeneous groups who have been exposed to larger doses by using statistical dose-response models. The second problem is that there are many other factors that are known to contribute to cancer risks or to modify the effects of radiation on cancer risks, and these factors need to be taken into account. While it is theoretically possible to control for such factors by cross-classifying the data into subgroups that are homogeneous with respect to all relevant factors, it is again unlikely that sufficiently large subgroups will be available to allow for stable estimates, particularly if the number of factors is large. For investigating lung cancer, for example, it might be necessary to control for sex, age, time since exposure, and smoking habit; if four levels were used for grouping each factor other than sex, a total of 128 subgroups would be needed, each of which would need to be the minimum size if risk estimates specific to each group were to be observed directly. Since this is not generally feasible, it is necessary to rely on multivariate statistical models to identify the consistent patterns across the variables simultaneously and to predict the risks for subgroups in which the sample sizes are inadequate. The third problem is that direct estimates of lifetime risk can only be obtained after an exposed population has been followed for a lifetime. Few populations have been followed so long, and even the atomic-bomb survivors, one of the populations followed for the longest period, has been followed only for just over 40 years. As the risks for many cancers in this population are still elevated, it is an open question whether the excess risk will continue for the remainder of the population's life and, if so, at what rate. It is not appropriate to wait until follow-up is complete, however, since interim estimates of risk are needed now for public health purposes. Again, to provide such estimates, one must fall back on statistical models that adequately describe the data available so far and the range of uncertainty around them. Epidemiologic data have increasingly been called on to help resolve claims for compensation by exposed individuals. Because a radiation-induced cancer is clinically indistinguishable from cancers caused by other factors, such claims must be settled on the "balance of probabilities," in other words, by determining what was the most likely cause, given the individual's history of exposure to radiation, and taking into account confounding and modifying factors. The calculation of these probabilities of causation depends on the availability of suitable multivariate exposure-response models. A recent National Institutes of Health working group
BACKGROUND INFORMATION AND SCIENTIFIC PRINCIPLES 51 (NIH85) has provided tables of such probabilities; these were based on data that were available at the time. Approaches to Model Construction and Fitting Exposure-Time-Response Models The last 20 years have seen a rapid increase in the use of multivariate models in the analysis of epidemiologic studies. The incidence of cancer and other diseases that are characterized as binary endpoints (present or absent) has usually been analyzed in terms of either the logistic model for the probability of disease P(z) as a function of exposure and other variables, where z = (z,â¦,zp), or the proportional hazards model for the instantaneous rate of disease, Î»(t \ z), at age t, where Î± and Î² are unknown regression coefficients that must be estimated, and Î»Âº(t) is the baseline rate in unexposed subjects (z = 0). These functions have a number of desirable mathematical properties that make them convenient to use under a wide range of circumstances, but they are not based on any particular biological theory. Thus, while they are useful for describing patterns and testing associations in which there is relatively little prior knowledge or biological theory, more reliable predictions can be made by using models that exploit such prior knowledge. The Committee has chosen, instead, to base its reanalyses of original epidemiologic data and risk assessments on the radiobiological principles and theories of the carcinogenesis process that are described elsewhere in this report. From this discussion, several considerations have emerged that need to be considered in designing statistical models. Dose-Response Relations Radiobiological theory indicates that at low doses, the risk of a biological lesion being formed should depend linearly on dose if a single event is required or on the square of dose if two events are required. It is commonly held that high-LET radiation can cause lesions by the traversal of a single particle, but that for low-LET radiation, either one or two photons might be required. At higher doses, radiation can cause cell sterilization or cell death, which competes with the process of malignant transformation. The probability of avoiding sterilization and death follows the usual laws of survival, which indicate that it should have
BACKGROUND INFORMATION AND SCIENTIFIC PRINCIPLES 52 a negative exponential dependence on either dose or the square of dose (again, depending on whether one or two events are needed). When these principles are combined, one obtains the general dose-response model used in the BEIR III report and extensively throughout the radiation literature: where D is the radiation dose, and F(D) is the incidence rate of cancer, a quantity that will be defined more rigorously below. Dependence on Time Cancer rates vary over several orders of magnitude as a function of age, and the excess risk caused by radiation exposure also varies as a complex function of age and time since exposure. Numerous mathematical theories of carcinogenesis have been devised to predict the dependence of incidence rates on exposure, age, and other time-related factors, but so far none has won universal acceptance and there have been few attempts to fit these models to epidemiologic data. Although the committee felt that stronger inferences about lifetime risk might be possible by exploiting these biomathematical models, it was unable to arrive at a consensus as to the particular models to use. Thus, there remains a need for simpler methods of summarizing the basic patterns of excess risk over time that do not depend on unproven hypotheses. Because leukemia and bone cancer appear to differ in temporal distribution from other cancers, these have generally been treated separately. Leukemia and Bone Cancer Following an instantaneous exposure to radiation, the rates of leukemia and bone cancer appear to follow a wave like pattern, rising within 5 years after exposure and then returning to near baseline rates within 30 years. For populations that have been followed for at least that long, no problems of projection arise. One simply models the risk of leukemia over the study period as a function of dose, F(D), and treats that as a lifetime excess risk estimate. The only complication is that the parameter estimates in F(D) may depend on sex s and age at exposure t. For populations with incomplete follow-up, the BEIR III Committee (NRC80) modelled the mortality rate, Î» (s,t,D), and applied that estimate as a constant to the period from 2 to 27 years after exposure. All Other Cancers In contrast to the rates for leukemia and bone cancer, the rates for most other cancers appear to have remained in excess for as long as most exposed populations have been followed. Whether they will continue to remain elevated for the rest of the population's life remains an important unanswered question, but most risk assessments have been based on the assumption that they will, although not necessarily
BACKGROUND INFORMATION AND SCIENTIFIC PRINCIPLES 53 at the same level. The BEIR III committee (NRC80) and much of the radioepidemiologic literature has relied on two simple models for projecting risks of these cancers: absolute risk and relative risk models. Letting Î»[T,D(t)] represent the incidence rate of cancer at age T resulting from an instantaneous exposure to dose D at age t, and letting Î»Âº(t) represent the baseline rate in unexposed persons, the two models can be represented as follows: where F(D) is given by Equation (1-3) with Î±0 constrained to 0 for the absolute risk model and 1 for the relative risk model. The BEIR III Committee adopted two minor modifications to these models: first, the excess risk was taken to be 0 for the first 10 years following exposure; second, the coefficients of F(D) were allowed to depend on sex and age at exposure. These modifications were extended in the BEIR IV Committee's (NRC88) reanalyses of the data on radon and lung cancer by adopting a general relative risk model of the form: where Î±1 is the average slope of a linear dose-response relationship and f (T), g(t), and h(T â t) represent modifying effects of age at risk, age at exposure, and time since exposure to be estimated, respectively. A general model of this type is also used in this report, except that the dose term Î±1D is replaced by (Î±2D + Î±3D2). Incorporation of Other Risk Factors In addition to the time-related factors discussed above, there are numerous risk factors that have been identified as having a direct effect on cancer rates; some of these may also modify the effects of radiation exposure on cancer rates; Unfortunately, there are relatively few studies that have assessed these other risk factors in combination with radiation. For lung cancer, the most important risk factor is smoking. The BEIR IV Committee (NRC88) has reviewed the studies reporting on the joint effects of smoking and radiation exposures and concluded that there was evidence of a synergistic (greater than additive) effect, but that there was also some evidence that the effect was less than multiplicative. They did not, however, consider the three-way interaction of age, smoking, and radiation. For low-LET radiation, the only data available on this point came from the Japanese atomic-bomb survivors and appeared
BACKGROUND INFORMATION AND SCIENTIFIC PRINCIPLES 54 to be too sparse to merit further modeling by incorporating age. Good human data on the interaction between radiation and other exposures do not appear to exist. The present Committee has therefore decided not to pursue analysis of interaction effects further at this time. Approaches to Model Fitting The approach that is taken to fitting risk models to epidemiologic data depends on the form in which the data are available. Some of the more complex models require access to the raw data on individual subjects and their entire history of exposures. However, most models can be fitted with very little loss of information by placing the subjects into subgroups with similar values of the relevant characteristics, particularly dose and age at exposure, and then tabulating their person-years at risk and the numbers of cases of each type of cancer as a function of age and time since exposure. The study data can then be summarized by two arrays, one of person-years, Y ijkl, for dose group i, age at exposure group j, attained age group k, and time since exposure interval l, and one of numbers of cancers, N ijklm, in each subgroup ijkl from each type of cancer m. Admittedly, the numbers of cases in most of the cells will be small, but this does not pose a problem for the method of analysis to be used. Next, one assumes that the numbers of cases in each cell follows a Poisson distribution, with the expected value given by the product of the rate predicted by the model and the person-years for that cell. The data can then be fitted by the technique of maximum likelihood. The likelihood is the probability of the observed data given a particular choice of model parameters, which, in this circumstance, is obtained from the product of the Poisson probabilities for each cell of the cross-tabulation. A Newton-Raphson search is used to find the parameter values which maximize this likelihood. Confidence limits and significance tests can be derived from large sample theory (Co74). The committee used a computer program known as AMFIT for fitting a general class of regression models for the Poisson data. Further details of the fitting program can be found in Annex 4C to Chapter 4. In any model fitting analysis, it is important to know how well the model describes the data. There are several aspects to this question. First, one would like an overall assessment of whether the model fits; such an assessment is known as a goodness-of-fit test. A poor fit might be an indication either that the chosen model is incorrect or that there is some problem with the data; a good fit does not prove that the model is correctâit simply means that there is insufficient evidence that the model is wrong. Next, assuming that the model fits, one would like to know the range of parameter values that is also consistent with the data; this range is known as a confidence interval and is important in evaluating the uncertainty in
BACKGROUND INFORMATION AND SCIENTIFIC PRINCIPLES 55 the fitted model. Next, one would like to be assured that the model is not unduly influenced by a few observations at the expense of the bulk of the data or by the inclusion of variables that are too highly correlated to be separated. Techniques to identify these types of problems are known as diagnostics and were used by the Committee throughout these analyses, as discussed in Annex 4F. Special Problems Pooling Data From Multiple Studies For many cancer sites, information was available from more than one epidemiologic study, raising the issue of how these data should be combined for risk assessment purposes. Because the studies generally differed in the nature of the exposures, the populations, and numerous methodological details, it was considered inappropriate to simply combine all of the raw data into a single data set. Instead, each of the studies for which original data were available to the committee were analyzed separately to obtain an estimate of the relevant parameters and their uncertainties. Formal tests of homogeneity were carried out to assess whether any differences in results could reasonably be ascribed to chance. If the results appeared to be consistent, an overall estimate could be obtained by a matrix weighted average and an estimate of the uncertainty of the pooled estimate could easily be derived. On the other hand, if the results appeared to be discrepant, the committee had to make a subjective judgment as to the quality and relevance of each of the studies. Use of Animal Data The committee felt strongly that its risk assessments should be based on human data to the extent that they were available and that animal data should be used only to address questions for which human data were unavailable or inadequate. Questions in the latter category included the RBE of neutrons and gamma rays and the effect of dose rate. Treatment of the RBE One of the problems for which the human data are inadequate is that of estimating the RBE for neutrons. The BEIR III Committee (NRC80) attempted to estimate the RBE for leukemia from the data from Japanese atomic-bomb survivors and then applied their estimate to the data on solid tumors. Aside from the inappropriateness of treating this point estimate as if it were known with certainty, the approach is no longer valid because reassessment of the atomic- bomb dosimetry has largely eliminated the differences in responses between Hiroshima and Nagasaki on which the previous estimate of the RBE was based. It therefore became necessary for the present Committee to rely on animal data for this purpose. For all analyses of the Radiation Effects Research Foundation
BACKGROUND INFORMATION AND SCIENTIFIC PRINCIPLES 56 (RERF) data, a value of 20 for the RBE for neutrons was assumed as a fixed constant. The justification for this choice is given in Chapter 4. Projection of Lifetime Risk Estimates Once the epidemiologic and animal data are summarized in the form of an exposure-time-response model, the final stage of risk assessment involves the calculation of lifetime risk for patterns of exposure of particular interest. This is done with standard life table (mortality table) techniques (Bu81). Consider the case of lifetime exposure at a constant annual rate. A life table analysis would proceed as follows. Starting with a hypothetical population of 1 million newborn infants, the first column in the life table gives the number of infants that are expected to survive to each age. The second column gives the cancer rate predicted by the exposure-time-response model, and the third column gives the number of cases of cancers that would result; this is given by the product of the first two columns. The fourth column gives the number of deaths from other causes, based on current mortality rates, which are not assumed to depend on radiation. The number of survivors at the beginning of the next age interval is therefore the number at the start of the interval minus the number of radiogenic and nonradiogenic deaths, and the process continues until the entire cohort is dead (although, in practice, the calculations are usually terminated at age 100). The total number of excess cases of cancer is estimated by subtracting the number of deaths obtained from a similar life table for persons with no radiation exposure. For protracted exposures, these calculations assume that each increment of exposure contributed independently to the cancer rates. Thus, the risk at age T is given by the background rate plus the sum over the entire exposure history of the excess rate attributable to each exposure increment; that is, if D(t) represents the history of radiation doses at each age t and [T,D(t)] represents the postulated dependence of cancer rates on age and each increment of exposure then the risk from the entire history of exposure is given by: This implies that the rate is a function of cumulative exposure (possibly weighted by a function of age at exposure or time since exposure). There is evidence, however, that the contributions of extended exposures are not simply additive: for low-LET radiation, protracted exposures appear to be less hazardous than instantaneous exposures of the same total dose, possibly because sublesions caused by the first event can be repaired before additional events occur; for high-LET radiation, the effect may simply be additive, or protracted exposures may even be more hazardous, possibly
BACKGROUND INFORMATION AND SCIENTIFIC PRINCIPLES 57 because subsequent radiation exposure can promote already initiated cells. The committee acknowledges this problem but, as explained earlier in this chapter, it does not believe that sufficient information is available to deal with this question in a definitive manner. The committee therefore chose to retain the assumption of independence for the calculations but to present the results in such a way that the reader can make adjustments for protracted exposure when warranted. Uncertainty of the Risk Estimates Unlike the BEIR III Committee (NRC80) which presented a range of lifetime risks based on relative and absolute risk models for several choices of dose-response functions, the present committee has chosen to assess the uncertainty of the projected lifetime risks by using a Monte Carlo simulation approach. The committee's preferred exposure-time-response model for a particular site of cancer or group of sites was characterized by a vector of parameter estimates and a covariance matrix which describes the uncertainty in each parameter. By repeated sampling from the set of possible parameter values, with sampling probabilities determined by their covariance matrices, 1,000 sets of possible parameters were obtained. Each combination was then applied to the life table calculation described above to obtain a set of predicted lifetime risks. The resulting distribution, presented in Chapter 4, gives a measure of the statistical uncertainty in the committee's risk estimates under the preferred model. Other sources of uncertainty, external to the preferred model and its statistical uncertainty, are discussed in Annex 4F. A number of other models fit the data nearly as well. The Monte Carlo simulation could, in principle, have been extended to include sampling over alternative models. However the committee invoked a number of nonstatistical criteria, e.g., biological plausibility, to chose between alternative models, and felt that using a simple goodness-of-fit criteria as weights in the Monte Carlo simulation would not adequately reflect this process. Life table results are presented in Annex 4D for a number of alternative models. It is of interest that the range of life table risks estimated under these alternative models is less than the uncertainty estimated by the Monte Carlo simulation. REFERENCES Ad87 Adams, L. M., S. P. Ethier, and R. L. Ullrich. 1987. Enhanced in vitro proliferation and in vivo tumorigenic potential of mammary epithelium from BALB/c mice exposed in vivo to gamma-radiation and/or 7,12-dimethylbenz(a)anthracene. Cancer Res. 47:4425-4431.
BACKGROUND INFORMATION AND SCIENTIFIC PRINCIPLES 58 Ba63 Barendson, G. W., H. M. D. Walter, J. F. Fowler, and D. K. Bewley. 1963. Effects of different ionizing radiations on human cells in tissue culture. III. Experiments with cyclotron- accelerated alpha particle and deuterons. Radiat. Res. 18:106. Ba68 Barendson, G. W. 1968. Responses of cultured cells, tumors, and normal tissues to radiation of different linear energy transfer. Curr. Top. Radiat. Res. Q. 4:293-356. Be88 Begg, C. B., and J. A. Berlin. 1988. Publication bias: a problem in interpreting medical data (with discussion). J. Roy. Statist. Soc. Ser A. 151:419-463. Be06 Bergonie, J., and L. Tribondeau. 1906. De quelques resultats de la radiotherapie et essai de fixation d'une technique rationelle. C. R. Sceances Acad. Sci. 143:983. (English translation by G. H. Fletcher, Radiat. Res. 11:587, 1959.) Bo74 Boecker, B. B., and R. G. Cuddihy. 1974. Toxicity of 144Ce Inhaled as 144CeCl3 by the beagle: Metabolism and dosimetry. Radiat. Res. 60:133-154. Bo78 Bond, V. P., C. B. Meinhold, and H. H. Rossi. 1978. Low dose RBE and Q for X ray compared to gamma-ray radiations. Health Phys. 34(5):433-438. Bo79 Borek, C., R. Miller, C. Pain, and W. Trom. 1979. Conditions for inhibiting and enhancing effects of the protease inhibitor antipain on X-ray induced neoplastic transformation in hamster and mouse cells. Proc. Natl. Acad. Sci. USA 76:1800-1803. Bo83 Borek, C, E. J. Hall, and M. Zaider. 1983. X rays may be twice as potent as X rays for malignant transformation at low dose. Nature 301:156-158. Br88 Brenner, D.J. 1988. Concerning the nature of the initial damage required for the production of radiation induced exchange aberrations. Int. J. Radiat. Biol. 52: 805-809. Br80 Breslow, N. E., and N. E. Day. 1980. Statistical Methods in Cancer Research, vol. 1: The Analyses of Case Control Studies. Publication No. 32. Lyon, France: IAR Scientific Publications. Br73 Broerse, J. J., and G. W. Barendsen. 1973. Relative biological effectiveness of fast neutrons for effects on normal tissue. Curr. Top. Radiat. Res. Q. 8:305-350. Bu81 Bunger B. M., J. R. Cook, and M. K. Barrick. 1981. Life Table Methodology for Evaluating Radiation Risk: An Application Based on Occupational Exposures . Health Phys. 40:439-455. Ch87 Chan, J. Y. H, F. F. Becker, J. German, and J. H. Ray. 1987. Altered DNA ligase I activity in Bloom's syndrome cells. Nature 325:357-359. C168 Cleaver, J. E. 1968. Defective repair replication of DNA in xeroderma pigmentosum. Nature 218:652-656. Co74 Cox, D. R. and D. V. Hinkley. 1974. Theoretical Statistics. London: Chapman and Hall. Co77 Cox, R., J. Thacker, and D. T. Goodhead. 1977. Inactivation and mutation of cultured mammalian cells by aluminum characteristics, ultrasoft X rays, and radiation of different LET. Int. J. Radiat. Biol. 31:561-576. Ei81 Eisenberg, H., and G. Felsenfeld. 1981. Hydrodynamic studies of the interaction between nucleosome core particles and core histones. J. Mol. Biol. 150:537-555. E172 Ellett, W. H. and L. A. Braby. 1972. The Microdosimetry of 250 kVp and 65 kVp X Rays, 60Co Gamma Rays, and Tritium Beta Particles. Radiat. Res. 51:229-243. Fi69 Field, S. B. 1969. The relative biological effectiveness of fast neutrons for mammalian tissues. Radiology 93:915-920. Fi71 Field, S. B. and S. Hornsey. 1971. RBE values for cyclotron neutrons for effects
BACKGROUND INFORMATION AND SCIENTIFIC PRINCIPLES 59 on normal tissues and tumors as a function of dose and dose fractionation. Eur. J. Cancer 7:151-169. Fi83 Fisher, D. R. 1983. Current concepts in lung dosimetry. Report PNL-SA-11049. Springfield, Va.: National Technical Information Service. Fr77 Fry, R. J. M. 1977. Radiation carcinogenesis. Int. J. Radiat. Oncol. Biol. Phys. 3:219-226. Fu54 Furth, J., and E. Lorenz. 1954. Carcinogenesis by ionizing radiations. Pp. 1145-1201 in Radiation Biology, vol. I, part II, A. Hollaender, ed. New York: McGraw-Hill. Go86 Goldman, M., L. S. Rosenblatt, and S. A. Book. 1986. Lifetime radiation effects research in animals: An overview of the status and philosophy of studies at University of California- Davis Laboratory for Energy Related Health Research. Pp. 53-65 in Life-Span Radiation Effects Studies in Animals: What Can They Tell Us?, R. C Thompson and J. A. Mahaffey, eds. U.S. Department of Energy Report CONF-830951. Springfield, Va.: National Technical Information Service. Go29 Goldstein, L., and D.P. Murphy. 1929. Etiology of the ill-health in children born after maternal pelvic irradiation. II. Defective children born after post-conception pelvic irradiation. Am. J. Roentgenol. 22:322-331. Gr53 Gray, L. H., A. D.Conger, M. Ebert, S. Hornsey, and O. C. A. Scott. 1953. The concentration of oxygen dissolved in tissues at the time of irradiation as a factor in radiotherapy. Br. J. Radiol 26:638-648. Gr85 Greulich, K. O., J. Ausio, and H. Eisenberg. 1985. Neucleosome core particle structure and structural changes in solution. J. Mol. Biol. 186:167-173. Gu80 Guernsey, D. L., A. Ong, and C. Borek. 1980. Thyroid hormone modulation of x ray induced in vitro neoplastic transformation. Nature 288:591-592. Ha64 Hall, E. J., J. S. Bedford. 1964. Dose rate: Its effect on the survival of HeLa cells irradiated with gamma rays . Radiat. Res. 22:305-315. Ha72 Hall, E. J. 1972. Radiation dose-rate: A factor of importance in radiobiology and radiotherapy. Br. J. Radiol. 45:81-97. Ha87 Hall, E. J., and T. K. Hei. 1987. Oncogenic transformation by radiation and chemicals. Pp. 507-512 in Proceedings of the 8th International Congress of Radiation Research, E. M. Fielden, J. F. Fowler, J. H. Hendry, and D. Scott, eds. Taylor and Francis. Ha79 Han, A., and M. M. Elkind. 1979. Transformation of mouse C3H/101/2 cells by single and fractionated doses of X rays and fission-spectrum neutrons. Cancer Res. 39:123-130. Ha80 Han, A., C. K. Hill, and M. M. Elkind. 1980 Neoplastic transformation of IOT 1/2 cells by 60Co gamma-rays: Evidence of repair of damage at reduced dose rate. Int. J. Radiat. Biol. 37:585-589. Ha82 Han, A. and M. M. Elkind. 1982. Enhanced transformation of mouse 10T 1/2 cells by 12-O- tetradecanoyl phorbol-13- acetate following exposure to x-rays or to fission-spectrum neutrons. Cancer Res. 42:477-483. Ha83 Harris, A. L., P. Karran, and T. Lindahl. 1983. O6-Methylguanine-DNA methyltransferase of human lymphoid cells: Structural and kinetic properties and absence in repair-deficient cells. Cancer Res. 43:3247-3252. He84 Hecht, F., and G. R. Sutherland. 1984. Fragile sites and cancer breakpoints. Cancer Genet. Cytogenet. 12:179-181. He87 Hecht, F., and B. K. Hecht. 1987. Chromosome changes connect immuno-deficiency and cancer in ataxia telangiectasia. Am. J. Pediatr. Hematol. Oncol. 9:185-188.
BACKGROUND INFORMATION AND SCIENTIFIC PRINCIPLES 60 He88 Hei, T. K., D. J. Chen, D. J. Brenner, and E. J. Hall. 1988. Mutation induction by charged particles of defined linear energy transfer. Carcinogenesis 9:1233-1236. Hi84 Hill, C. K., A. Han, and M. M. Elkind. 1984. Fission-spectrum neutrons at a low dose rate enhance neoplastic transformation in the linear, low dose region (0-10 cGy). Int. J. Radiat. Biol. 46:11-15. IAEA86 International Atomic Energy Agency. 1986. Biological dosimetry: Chromosomal Aberration Analysis for Dose Assessment. Technical Reports Series No. 260. Vienna: International Atomic Energy Agency. ICRU83 International Commission on Radiation Units and Measurements. 1983. Microdosimetry. ICRU Report 36. Bethesda, Md.: International Commission on Radiation Units and Measurements. ICRU86 International Commission on Radiation Units and Measurements. 1986. The Quality Factor in Radiation Protection. ICRU Report 40. Report of Joint Task Group of the ICRP and the ICRU. ICRP63 International Commissions on Radiological Protection and on Radiological Units and Measurements. 1963. Report of the RBE Committee to the International Commissions on Radiological Protection and on Radiological Units and Measurements. Health Physics 9:357-386. ICRP79 International Commission on Radiological Protection. 1979-1988. Limits for intakes of radionuclides by workers. ICRP Publication 30. Vols. 2(3/4), 3, 4(3/4), 5, 6(2/3), 7, 8, and 19(4). Oxford: Pergamon. ICRP86 International Commission on Radiological Protection. 1986. The metabolism of plutonium and related elements. ICRP Publication 48. Oxford: Pergamon. Jo85 Johnson, J. R. 1985. Internal dosimetry for radiation protection, Chapter 6 in The Dosimetry of Ionizing Radiation, (K. R. Kase, B. E. Bjarngard, and F. H. Attix, eds. Orlando: Academic Press, Inc. Ke78b Kennedy, A. R., and J. B. Little. 1978. Protease inhibitors suppress radiation-induced malignant transformation in vitro. Nature 276:825-826. Ke81 Kennedy, A. R. and J. B. Little. 1981. Effects of protease inhibitors on radiation transformation in vitro. Cancer Res. 41:2103-2108. Ke78a Kennedy, A. R., S. Monjal, C. Heidelberger, and J. B. Little. 1978. Enhancement of X ray transformation by 12-O-tetradecanoyl phorbol 13 acetate in a cloned line C3H mouse embryo cells. Cancer Res. 38:439-443. Ke80 Kennedy, A. R., G. Murphy, and J. B. Little. 1980. Effect of time and duration of exposure to 12-O-tetradecanoyl-phorbol-13-acetate aon x-ray transformation of C3H 10T1/2 cells. Cancer Res. 40:1915-1920. Kn85 Knudson, A. G. 1985. Hereditary cancer, oncogenes, and antioncogenes. Cancer Res. 45:1437-1443. Kr82 Kraft, G., W. Kraft-Weyrather, H. Meister, H. G. Miltenburger, R. Roots, and H. Wulf. 1982. The influence of radiation quality on the biological effectiveness of heavy charged particles . Pp. 743-73 in Radiation Protection, J. Booz and H. G. Ebert eds. Commission of European Communities. La87 Laird, N. M. 1987. Thyroid cancer risk from exposure to ionizing radiation: A case study in the comparitive potency model. Risk Analysis 7:299-309. La89 Langlois, R. G., W. L. Bigbee, R. H. Jensen, and J. German. 1989. Evidence for increased to vivo mutation and somatic recombination in Bloom's syndrome. Proc. Natl. Acad. Sci. 86:670-674. Le56 Lea, D. E. 1956. DEA: Actions of radiations on living cells, 2nd ed. Cambridge, England: Cambridge University Press.
BACKGROUND INFORMATION AND SCIENTIFIC PRINCIPLES 61 Le42 Lea and Catcheside. 1942. The mechanism of the induction by radiation of chromosome aberrations in Tradescantia. J. Genet. 44:216-249. Li82 Lindahl, T. 1982. DNA repair enzymes. Annu. Rev. Biochem. 1:61-87. Li72 Lindahl, T. and B. Nyberg. 1972. Rate of depurination of native DNA. Biochemistry 11:3610-3618. Li74 Lindahl, T. and B. Nyberg. 1974. Heat-induced deamination of cytosine residues in deoxyribonucleic acid. Biochemistry 13:3405-3410. Ll81 Lloyd, D.C., and R. J. Purrott. 1981. Chromosome aberration analysis in radiological protection dosimetry. Radiat. Protect. Dosim. 1:19-28. Lu87a Lundgren, D. L., F. F.Hahn, W. C. Griffith, R. G. Cuddihy, P. J. Haley, and B. B. Boecker 1987. Effects of relatively low-level exposure of rats to inhaled 144CeO2. III. Pp. 308-12 in Inhalation Toxicology Research Institute Annual Report 1986-1987, J. D. Sun, and J. A. Mewhinney, eds. U.S. Department of Energy Report LMF-120. Springfield, Va.: National Technical Information Service. Lu87b Lundgren, D.L., F. F. Hahn, W. C. Griffith, R. G. Cuddihy, F. A. Seiler, and B. B. Boecker. 1987. Effects of relatively low-level thoracic or whole-body exposure of rats to X-rays. I. Pp. 313-317 in Inhalation Toxicology Research Institute Annual Report 1986-1987. J. D. Sun and J. A. Mewhinney, eds. U.S. Department of Energy Report LMF-120 . Springfield, Va.: National Technical Information Service. Ma70 MacMahon, B., and T. F. Pugh. 1970. Epidemiology Principles and Methods. Boston: Little, Brown and Company. Mc86 McClellan, R. O., B. B. Boecker, F. F. Hahn, and B. A. Muggenburg. 1986. Lovelace ITRI Studies on the toxicity of inhaled radionuclides in beagle dogs. Pp. 74-96 in Life-Span Radiation Effects Studies in Animals: What Can They Tell Us?, R. E. Thompson and J. A. Mahaffey, eds. U.S. Department of Energy Report CONF-830951. Springfield, Va.: National Technical Information Service. Me88 Metivier, H., R. Masse, G. Rateau, D. Nolibe, and J. Lafuma. 1988. In press. New data on the toxicity of 239PuO2 in baboons. Proceedings of the CEC/CEA/DOE-Sponsored Workshop on Biological Assessment of Occupational Exposure to Actinides. Versailles, France, May 30-June 2, 1988. Mi78 Michaels, H. B., and J. W. Hunt. 1978. A model for radiation damage in cells by direct effect and by indirect effect: A radiation chemistry approach. Radiat. Res. 74:23-24. Mi85 Mitchell, D. L., C. A. Haipek, and J. M. Clarkson. 1985. (6-4) Photoproducts are removed from the DNA of UV-irradiated mammalian cells more efficiently than cyclobutane pyrimidine dimers. Mutat. Res. 143:109-112. Mo36 Mottram, J. C. 1936. Factor of importance in radiosensitivity of tumors. Br. J. Radiol. 9:606-614. Mu28 Muller, H.J. 1928. The effects of X-radiation on genes and chromosomes. Science 67:82. NCRP80 National Council on Radiation Protection and Measurements. 1980. Influence of dose and its distribution in time on dose-response relationships for low LET radiations. Report No. 64. Washington, D.C.: National Council on Radiation Protection and Measurements. NCRP84 National Council for Radiation Protection and Measurements (NCRP). 1984. Evaluation of Occupational and Environmental Exposures to Radon and Radon Daughters in the United States. NCRP Report No. 78. Bethesda, Md.: National Council on Radiation Protection and Measurements.
BACKGROUND INFORMATION AND SCIENTIFIC PRINCIPLES 62 NCRP85 National Council for Radiation Protection and Measurements (NCRP). 1985. General Concepts for the Dosimetry of Internally Deposited Radionuclides. NCRP Report No. 84. Bethesda, Md.: National Council on Radiation Protection and Measurements. NCRP87a National Council on Radiation Protection and Measurements (NRCP). 1987. Recommendations on limits for exposure to ionizing radiation. Report No. 91. Bethesda, Md.: National Council on Radiation Protection and Measurements. NCRP87b National Council on Radiation Protection and Measurements (NCRP). 1987. Ionizing Radiation Exposures of the Population of the United States. Report No. 93. Washington, D.C.: National Council on Radiation Protection and Measurements. NIH85 National Institutes of Health. 1985. Report of the National Institutes of Health Ad Hoc Working Group to Develop Radioepidemiological Tables. NIH Publication 85-2748. Washington, D.C.: Superintendent of Documents, Government Printing Office. NRC80 National Research Council, Committee on the Biological Effects of Ionizing Radiations (BEIR III). 1980. The Effects on Populations of Exposure to Low Levels of Ionizing Radiation. Washington, D.C: National Academy Press. 524 pp. NRC88 National Research Council, Committee on the Biological Effects of Ionizing Radiations (BEIR IV). 1988. Health Risks of Radon and Other Internally Deposited Alpha-Emitters. Washington, D.C.: National Academy Press. 602 pp. Pal84 Palcic, B. and L. D. Skarsgard. 1984. Reduced oxygen enhancement ratio at low doses of ionizing radiation. Radiat. Res. 100:328-339. Pa86 Park, J.F., G. E. Dagle, H. A. Ragan, R. E. Weller, and D. L. Stevens. 1986. Current status of life-span studies with inhaled plutonium in beagles at Pacific Northwest Laboratory. Pp. 455-470 in Life-Span Radiation Effects Studies in Animals: What Can They Tell Us?, R. E. Thompson and J. A. Mahaffey, eds. U.S. Department of Energy Report CONF-830951. Springfield, Va.: National Technical Information Service. Pat49 Patt, H. M., E. B. Tyree, R. L. Straube, and D. E. Smith. 1949. Cysteine protection against x- irradiation. Science 110:213. Ro86 Rothman, K. J. 1986. Modern Epidemiology. Boston: Little, Brown and Co. Ro84 Rowley, J. D. 1984. Biological implications of consistent chromosome rearrangements in leukemia and lymphoma. Cancer Res. 44:3159-3168. Ru54 Russell, L. B. 1954. The effects of radiation on mammalian prenatal development. Pp. 861-918 in Radiation Biology, vol. I, part II. A. Hollaender, ed. New York: McGraw-Hill. Ry82 Rydberg, B., and T. Lindahl. 1982. Nonenzymatic methylation of DNA by the intracellular methyl group donor S-adenosyl-L-methionine is a potentially mutagenic reaction. EMBO J. 1:211-216. Sa88 Sanders, C. E., K. E. Lauhala, J. A. Mahaffey, and K. E. McDonald. 1988. Low-level 239PuO2 lifetime studies. Pp. 31-34 in Pacific Northwest Laboratory Annual Report for 1987 to the DOE Office of Energy Research. Part 1. Biomedical Sciences. U.S. Department of Energy Report PNL-6500 Part 1. Springfield, Va.: National Technical Information Service. Sa40 Sax, K. 1940. An analysis of X ray induced chromosome aberrations in Tradescantia. Genetics 25: 41-66. Sc74 Schmidt E., G. Rimpl, and M. Bauchinger. 1974. Dose-Response Relation
BACKGROUND INFORMATION AND SCIENTIFIC PRINCIPLES 63 of Chromosome Aberations in Human Lymphocytes After in Vitro Irradiation with 3-MeV Electrons. Radiat. Res. 57:228. Sc74c Schroeder, T. M., and J. German. 1974. Bloom's syndrome and Fanconi's anemia: Demonstration of two distinctive patterns of chromosome disruption and rearrangement. Humangenetik 25:299-306. Sh87 Shimizu, Y., H. Kato, W. J. Schull, D. L. Preston, S. Fujita, and D. A. Pierce. 1987. Lifespan report 11, part 1. Comparison of Risk Coefficients for Site Specific Cancer Mortality Based on the DS86 and T65DR Shielded Kerma and Organ Doses. RERF TR 12-87. Hiroshima: Radiation Effects Research Foundation. Si66 Sinclair, W. K., and R. A. Morton. 1966. X ray sensitivity during the cell generation cycle of cultured Chinese hamster cells. Radiat. Res. 29:450-474. Sm84 Smith, H., and G. Gerber, eds. 1984. Lung modelling for inhalation of radioactive materials, Proceedings of a meeting jointly organized by the Commission of the European Communities and the National Radiological Protection Board, Oxford, March 26-28, 1984. Report EUR9384EN. Sn83 Snipes, M. B., B. B. Boecker, and R. O. McClellan. 1983. Retention of monodisperse or polydisperse aluminosilicate particles inhaled by dogs, rats and mice. Toxicol. Appl. Pharmacol. 69:345-362. Sn84 Snipes, M. B., B. B. Boecker, and R. O. McClellan. 1984. Respiratory tract clearance of inhaled particles in laboratory animals. Pp 63-71 in Lung Modelling for Inhalation of Radioactive Materials, H. Smith and G. Gerber, eds. Report EUR 9384EN. St88 Storer, J. B., T. J. Mitchell and R. J. M. Fry. 1988. Extrapolation of the relative risk of radiogenic neoplasms across mouse strains and to man. Radiat. Res. 114:331-353. Ter63 Terasima, T. and L. J. Tolmach. 1963. Variations in several responses of HeLa cells to X- irradiation during the division cycle . Biophys. J. 3:11-33. Th86 Thacker, J., R. E. Wilkinson, and D. T. Goodhead. 1986. The induction of chromosome exchange aberrations by carbon ultrasoft X-rays in V79 hamster cells. Int. J. Radiat. Biol. 49:645-656. Th86d Thompson, R. C., and J. A. Mahaffey, eds. Life-Span Radiation Effects Studies in Animals: What Can They Tell Us? Report CONF-830951. Springfield, Va.: National Technical Information Service. Th81 Thomson, J. F., F. S. Williamson, D. Grahn, and E. J. Ainsworth. 1981. Radiat. Res. 86 (3):559-572, 572-588. Ul84 Ullrich, R. L. 1984. Tumor induction in BALB/c mice after fractionated or protracted exposures to fission-spectrum neutrons. Radiat. Res. 97:587-597. Un76 Underbrink, A., A. Kellerer, R. Mills, and A. Sparrow. 1976. Radiat. Environ. Biophys. 13:295. UN77 United Nations Scientific Committee on the Effects of Ionizing Radiation (UNSCEAR). 1977. Genetic effects of radiation. pp. 425-564. In Sources and Effects of Ionizing Radiation. Report A/32/40. Thirty Second Session, Supplement No. 40. New York: United Nations. UN86 United Nations Scientific Committee on the Effects of Ionizing Radiation (UNSCEAR). 1986. Genetic Effects of Radiation. Pp. 7-164 in Ionizing Radiation: Sources and Biological Effects. Report A/41/16. Forty First Session, Supplement No. 16. New York: United Nations. Vi83 Vijayalaxmi, H. J.Evans, J. H. Ray, and J. German. 1983. Bloom's syndrome: Evidence for an increased mutation frequency in vivo. Science 221:851-853.
BACKGROUND INFORMATION AND SCIENTIFIC PRINCIPLES 64 Vo81 Vogel, H. H., and H. W. Dickson. 1981. Abstracts of the 29th Annual Meeting of the Radiation Research Society, Minneapolis. Radiat. Res. 87(2):453. Wa87 Ward, J. F. 1988. DNA damage produced by ionizing radiation in mammalian cells: identities, mechanisms of formation and reparability. Prog. Nucleic Acids Res. Mol. Biol. 35:96-128. Wa88 Ward, J. F., C. L. Limoli, P. Calabro-Jones, and J. W. Evans. 1988. Radiation vs. chemical damage to DNA. Anticarcinogenesis and Radiation Protection, O. F. Nygaard, M. Simic, and P. Cerutti, eds. New York: Plenum. Wi85 Widom, J., and A. Klug. 1985. Structure of the 300 A chromotin filament: X-ray diffraction frm oriented samples. Cell 43:207-213. Wi87 Willis, A.E., and T. Lindahl 1987. DNA ligase I deficiency in Bloom's syndrome. Nature 325:355-357. Wr86 Wrenn, M. E., G. N. Taylor, W. Stevens, C. W. Mays, W. S. S. Jee, R. D. Lloyd, D. R. Atherton, F.W. Bwenger, S. C. Miller, J. M. Smith, L. R. Shabestan, L. A. Woodbury, and B. J. Stover. 1986. DOE life-span radiation effects studies in experimental animals at University of Utah Division of Radiobiology. Pp. 32-52 in Life-Span Radiation Effects Studies in Experimental Animals: What Can They Tell Us?, R. C. Thompson, and J. A. Mahaffey, eds. U.S. Department of Energy Report CONF-830951. Springfield, Va.: National Technical Information Service. Yu80 Yuhas, J. M., J. M. Spellman, and F. Cullo. 1980. The role of WR2721 in radiotherapy and/or chemotherapy. Pp. 303-308. Radiation Sensitizers, L. Brady, ed. New York: Masson.