Measurements of LET distribution and dose equivalent onboard the Space Shuttle IML-2 (STS-65) and S/MM#4 (STS-79).

Space radiation dosimetry measurements have been made onboard the Space Shuttle STS-65 in the Second International Microgravity Laboratory (IML-2: 28.5 degrees x 300 km: 14.68 days) and the STS-79 in the 4th Shuttle MIR mission (S/MM#4: 51.6 degrees x 300-400km: 10.2 days). In these measurements, three kinds of detectors were used; one is a newly developed active detector telescope called "Real-time Radiation Monitoring Device (RRMD-I for IML-2 and RRMD-II with improved triggering system for S/MM#4)" utilizing silicon semi-conductor detectors and the other detectors are conventional passive detectors of thermoluminescence dosimeters (TLDs) and CR-39 plastic track detectors. The main contribution to dose equivalent for particles with LET > 5.0 keV/micrometer (IML-2) and LET > 3.5 keV/micrometer (S/MM#4) is seen to be due to galactic cosmic rays (GCRs) and the contribution of the South Atlantic Anomaly (SAA) is less than 5% (IML-2: 28.5 degrees x 300 km) and 15% (S/MM#4: 51.6 degrees x 400 km) in the above RRMD LET detection conditions. For the whole LET range (> 0.2 kev/micrometer) obtained by TLDs and CR-39 in these two typical orbits (a small inclination x low altitude and a large inclination x high altitude), absorbed dose rates range from 94 to 114 microGy/day, dose equivalent rates from 186 to 207 microSv/day and average quality factors from 1.82 to 2.00 depending on the locations and directions of detectors inside the Spacelab at the highly protected IML-2 orbit (28.5 degrees x 300 km), and also, absorbed dose rates range from 290 to 367 microGy/day, dose equivalent rates from 582 to 651 microSv/day and average quality factors from 1.78 to 2.01 depending on the dosimeter packages around the RRMD-II "Detector Unit" at the S/MM#4 orbit (5l.6 degrees x 400km). In general, it is seen that absorbed doses depend on the orbit altitude (SAA trapped particles contribution dominant) and dose equivalents on the orbit inclination (GCR contribution dominant). The LET distributions obtained by two different types of active and passive detectors, RRMDs and CR-39, are in good agreement for LET of 15 - 200 kev/micrometer and difference of these distributions in the regions of LET < 15 kev/micrometer and LET > 200 kev/micrometer can be explained by considering characteristics of CR-39 etched track formation especially for the low LET tracks and chemical etching conditions.     

Shuttle radiation dose measurements in the International Space Station orbits.

The International Space Station (ISS) is now a reality with the start of a permanent human presence on board. Radiation presents a serious risk to the health and safety of the astronauts, and there is a clear requirement for estimating their exposures prior to and after flights. Predictions of the dose rate at times other than solar minimum or solar maximum have not been possible, because there has been no method to calculate the trapped-particle spectrum at intermediate times. Over the last few years, a tissue-equivalent proportional counter (TEPC) has been flown at a fixed mid-deck location on board the Space Shuttle in 51.65 degrees inclination flights. These flights have provided data that cover the expected changes in the dose rates due to changes in altitude and changes in solar activity from the solar minimum to the solar maximum of the current 23rd solar cycle. Based on these data, a simple function of the solar deceleration potential has been derived that can be used to predict the galactic cosmic radiation (GCR) dose rates to within +/-10%. For altitudes to be covered by the ISS, the dose rate due to the trapped particles is found to be a power-law function, rho(-2/3), of the atmospheric density, rho. This relationship can be used to predict trapped dose rates inside these spacecraft to +/-10% throughout the solar cycle. Thus, given the shielding distribution for a location inside the Space Shuttle or inside an ISS module, this approach can be used to predict the combined GCR + trapped dose rate to better than +/-15% for quiet solar conditions.     

Space radiation enhancement linked to geomagnetic disturbances.

Space radiation dosimetry measurements have been made on board the Space Shuttle. A newly developed active detector called "Real-time Radiation Monitoring Device (RRMD)" was used (Doke et al., 1995; Hayashi et al., 1995). The RRMD results indicate that low Linear Energy Transfer (LET) particles steadily penetrate around the South Atlantic Anomaly (SAA) without clear enhancement of dose equivalent and some daily periodic enhancements of dose equivalent due to high LET particles are seen at the lower geomagnetic cutoff regions (Doke et al., 1996). We also have been analyzing the space weather during the experiment, and found that the anomalous high-energy particle enhancement was linked to geomagnetic disturbance due to the high speed solar wind from a coronal hole. Additional analysis and other experiments are necessary for clarification of these phenomena. If a penetration of high-energy particles into the low altitude occurs by common geomagnetic disturbances, the prediction of geomagnetic activity becomes more important in the next Space Station's era.     

A comparison of depth dependence of dose and linear energy transfer spectra in aluminum and polyethylene

A set of four tissue-equivalent proportional counters (TEPCs), with their detector heads at the centers of 0 (bare), 3, 7 and 9-inch-diameter aluminum spheres, were flown on Shuttle flight STS-89. Five such detectors at the centers of polyethylene spheres were flown 1 year earlier on STS-81. The results of dose-depth dependence for the two materials convincingly show the merits of using material rich in hydrogen to decrease the radiation exposure to the crew. A comparison of the calculated galactic cosmic radiation (GCR) absorbed dose and dose-equivalent rates using the radiation transport code HZETRN with nuclear fragmentation model NUCFRG2 and the measured GCR absorbed dose rates and dose-equivalent rates shows that they agree within root mean square (rms) error of 12.5 and 8.2%, respectively. However, there are significant depth-dependent differences in the linear energy transfer (LET) spectra. A comparison for trapped protons using the proton transport code BRYNTRN and the AP-8 MIN trapped-proton model shows a systematic bias, with the model underpredicting dose and dose-equivalent rates. These results show the need for improvements in the radiation transport and/or fragmentation models.     

Effective dose equivalent on the ninth Shuttle--Mir mission (STS-91)

Organ and tissue doses and effective dose equivalent were measured using a life-size human phantom on the ninth Shuttle-Mir Mission (STS-91, June 1998), a 9.8-day spaceflight at low-Earth orbit (about 400 km in altitude and 51.65 degrees in inclination). The doses were measured at 59 positions using a combination of thermoluminescent dosimeters of Mg(2)SiO(4):Tb (TDMS) and plastic nuclear track detectors (PNTD). In correcting the change in efficiency of the TDMS, it was assumed that reduction of efficiency is attributed predominantly to HZE particles with energy greater than 100 MeV nucleon(-1). A conservative calibration curve was chosen for determining LET from the PNTD track-formation sensitivities. The organ and tissue absorbed doses during the mission ranged from 1.7 to 2.7 mGy and varied by a factor of 1.6. The dose equivalent ranged from 3.4 to 5.2 mSv and varied by a factor of 1.5 on the basis of the dependence of Q on LET in the 1990 recommendations of the ICRP. The effective quality factor (Q(e)) varied from 1.7 to 2.4. The dose equivalents for several radiation-sensitive organs, such as the stomach, lung, gonad and breast, were not significantly different from the skin dose equivalent (H(skin)). The effective dose equivalent was evaluated as 4.1 mSv, which was about 90% of the H(skin).     

Detection of space radiation-induced double strand breaks as a track in cell nucleus

To identify DNA damage induced by space radiations such as the high linear energy transfer (LET) particles, phospho-H2AX (γH2AX) foci formation was analyzed in human cells frozen in an International Space Station freezer for 133 days. After recovering the frozen sample to the earth, the cells were cultured for 30 min, and then fixed. Here we show a track of γH2AX positive foci in them by immuno-cytochemical methods. It is suggested that space radiations, especially high LET particles, induced DSBs as a track. From the formation of the tracks in nuclei, exposure dose rate was calculated to be 0.7 mSv per day as relatively high-energy space radiations of Fe-ions (500 MeV/u, 200 keV/μm). From the physical dosimetry with CR-39 plastic nuclear track detectors and thermo-luminescent dosimeters, dose rate was 0.5 mSv per day. These values the exposed dose rate were similar between biological and physical dosimetries.     

Recent European measurements inside Biorack

The dosimetric package used inside Biorack on board STS76, STS81 and STS84 comprises passive detector stacks built from plastic nuclear track detectors (PNTDs), thermoluminescence detectors (TLDs) and one or two active DOSTEL (DOSimetric TELescope) units using planar silicon detectors. Five passive detector stacks were exposed at different places inside the BIORACK incubators and in different stowage positions. DOSTEL units were exposed inside the 22 degrees C incubator in all flights. Mission integrated dose measurements, particle fluence rates and neutron doses are obtained from the passive detector stacks. These results are complemented by time resolved particle counts and dose rates and linear energy transfer (LET) spectra separately for the contribution of the trapped particles and the galactic cosmic rays (GCR) as a result of the DOSTEL measurements. In addition, it was possible to investigate the anisotropy of the radiation field inside Biorack by the use of a second DOSTEL unit on STS84. Since all exposures are during a solar minimum period, the total radiation exposure is of a similar extent for all flights, although position differences in dose rate up to a factor of two are observed. Particle fluence rates show lower variations. Mission averaged mean quality factors (Q) determined from the LET spectra are 2.0+/-0.1; the deduced dose equivalent rates range from 631 to 716 microSv/day.     

Inactivation of aerobic and hypoxic cells from three different cell lines by accelerated (3)He-, (12)C- and (20)Ne-ion beams

The LET-RBE spectra for cell killing for cultured mammalian cells exposed to accelerated heavy ions were investigated to design a spread-out Bragg peak beam for cancer therapy at HIMAC, National Institute of Radiological Sciences, Chiba, prior to clinical trials. Cells that originated from a human salivary gland tumor (HSG cells) as well as V79 and T1 cells were exposed to (3)He-, (12)C- and (20)Ne-ion beams with an LET ranging from approximately 20-600 keV/micrometer under both aerobic and hypoxic conditions. Cell survival curves were fitted by equations from the linear-quadratic model and the target model to obtain survival parameters. RBE, OER, alpha and D(0) were analyzed as a function of LET. The RBE increased with LET, reaching a maximum at around 200 keV/micrometer, then decreased with a further increase in LET. Clear splits of the LET-RBE or -OER spectra were found among ion species and/or cell lines. At a given LET, the RBE value for (3)He ions was higher than that for the other ions. The position of the maximum RBE shifts to higher LET values for heavier ions. The OER value was 3 for X rays but started to decrease at an LET of around 50 keV/micrometer, passed below 2 at around 100 keV/micrometer, and then reached a minimum above 300 keV/micrometer, but the values remained greater than 1. The OER was significantly lower for (3)He ions than the others.     

Comparisons of LET distributions measured in low-earth orbit using tissue-equivalent proportional counters and the position-sensitive silicon-detector telescope (RRMD-III)

Determinations of the LET distribution, phi(L), of charged particles within a spacecraft in low-Earth orbit have been made. One method used a cylindrical tissue-equivalent proportional counter (TEPC), with the assumption that for each measured event, lineal energy, y, is equal to LET and thus phi(L) = phi(y). The other was based on the direct measurement of LETs for individual particles using a charged-particle telescope consisting of position-sensitive silicon detectors called RRMD-III. There were differences of up to a factor of 10 between estimates of phi(L) using the two methods on the same mission. This caused estimates of quality factor to vary by a factor of two between the two methods.     

The response of a spherical tissue-equivalent proportional counter to iron particles from 200-1000 MeV/nucleon

The radiation environment on board the space shuttle and the International Space Station includes high-Z and high-energy (HZE) particles that are part of the galactic cosmic radiation (GCR) spectrum. Iron-56 particles are considered to be one of the most biologically important parts of the GCR spectrum. Tissue-equivalent proportional counters (TEPCs) are used as active dosimeters on manned space flights. These TEPCs are further used to determine the average quality factor for each space mission. A TEPC simulating a 1-microm-diameter sphere of tissue was exposed as part of a particle spectrometer to (56)Fe particles at energies from 200-1000 MeV/nucleon. The response of TEPCs in terms of mean lineal energy, y(F), and dose mean lineal energy, y(D), as well as the energy deposited at different impact parameters through the detector was determined for six different incident energies of (56)Fe particles in this energy range. Calculations determined that charged-particle equilibrium was achieved for each of the six experiments. Energy depositions at different impact parameters were calculated using a radial dose distribution model, and the results were compared to experimental data.     

Assessment of radiation exposure level for hydrobionts in some special industrial ponds at the "Mayak" PA

Evaluation of the radionuclide content in the ecosystem components (water, sediments, aquatic organisms) of industrial reservoirs-storages of liquid radioactive waste of the "Mayak" PA (reservoirs R-4, R-10, R-11, R-17, R-9) and the estimation of the absorbed dose rate in aquatic organisms of these reservoirs using the software package ERICA Assessment Tool 1.0 May 2009 have been performed. Gradient of the absorbed dose rate for the detected taxonomic groups of hydrobionts in the series of the studied reservoirs R-11 --> R-10 --> R-4 --> R-17 --> R-9 was almost equal to one order of magnitude. The estimated absorbed dose rate for phytoplankton ranged from 5.4 x 10(0) mGy/day (R-11) to 4.0 x 10(4) mGy/day (R-9), for zooplankton--from 6.4 x 10(-1) mGy/day (R-11) to 3.8 x 10(3) mGy/day (R-9), for zoobenthos (chironomids)--from 5.6 x 10(0) mGy/day (R-11) to 1.1 x 10(3) mGy/day (R-17), for fish (roach)--from 8.0 x 10(-1) mGy/day (R-11) to 1.9 x 10(1) mGy/day (R-4).     

Problems of component discrimination in space radiation dosimetry

Resolving the LET spectrum of environmental radiation in space for assessing dose equivalents creates special problems due to superposition effects. Three components of the radiation field in space, trapped protons, tissue disintegration stars, and neutrons, contribute the bulk of the total dose equivalent. While lack of discrimination of neutron recoil and trapped primary protons does not interfere with correct determination of the combined dose equivalent as such, the simultaneous bursts of several low-energy protons and alpha particles from tissue disintegration stars completely defy LET-resolution with conventional instrumentation. So far, the tissue star dose has been determined only semiquantitatively from nuclear emulsion data. The neutron spectrum in space shows a markedly higher relative fluence in the region beyond 5 MeV than the fission neutron spectrum. Therefore, its LET spectrum centers less heavily on LET values near the proton Bragg Peak. This would call for assigning a QF value of less than 10 to the neutron dose in space. Still more serious shortcomings exist with regard to LET interpretation of heavy primaries.     

Comparisons of LET distributions for protons with energies between 50 and 200 MeV determined using a spherical tissue-equivalent proportional counter (TEPC) and a position-sensitive silicon spectrometer (RRMD-III)

Experiments have been performed to measure the response of a spherical tissue-equivalent proportional counter (TEPC) and a silicon-based LET spectrometer (RRMD-III) to protons with energies ranging from 50-200 MeV. This represents a large portion of the energy distribution for trapped protons encountered by astronauts in low-Earth orbit. The beam energies were obtained using plastic polycarbonate degraders with a monoenergetic beam that was extracted from a proton synchrotron. The LET spectrometer provided excellent agreement with the expected LET distribution emerging from the energy degraders. The TEPC cannot measure the LET distribution directly. However, the frequency mean value of lineal energy, y(-)(f), provided a good approximation to LET. This is in contrast to previous results for high-energy heavy ions where y(-)(f) underestimated LET, whereas the dose-averaged lineal energy, y(-)(D), provided a good approximation to LET.     

ESR spin trapping of hydroxyl radicals in aqueous solution irradiated with high-LET carbon-ion beams

The aim of this study was to quantify the hydroxyl radicals (*OH) produced when aqueous solutions are decomposed by high-linear energy transfer (LET) 290 MeV/nucleon carbon-ion beams using an electron spin resonance (ESR) spectrometer. Aerated cell culture medium containing 200 mM 5,5-dimethyl-1-pyrroline-N-oxide (DMPO) was irradiated with doses of 0 to 20 Gy with an LET of 20 to 90 keV/ micro m. We were able to obtain ESR spectra 10 min after irradiation, and the formation of *OH and hydrogen atoms was confirmed by radiolysis of deuterium oxide and ethanol containing DMPO. Our results showed that the yield of *OH by carbon-ion radiolysis increased in proportion to the absorbed dose over the range of 0 to 20 Gy. Furthermore, we discovered that the yield of *OH decreased linearity as LET increased logarithmically from 20 to 90 keV/ micro m. The generation of *OH by carbon-ion radiolysis at LETs of 20, 40, 60, 80 and 90 keV/ micro m was 64, 58, 52, 49 and 50%, respectively, of that for low-LET X radiolysis. These unique findings provide a further understanding of the indirect effect of high-LET radiation.     

Track structure in DNA irradiated with heavy ions

The spatial properties of trapped radicals produced in heavy-ion-irradiated solid DNA at 77 K have been probed using pulsed electron paramagnetic double resonance (PELDOR or DEER) techniques. Salmon testes DNA hydrated to 12 water molecules per nucleotide was irradiated with 40Ar ions of energy 100 MeV/nucleon and LET ranging from 300 to 400 keV/microm. Irradiated samples were maintained at cryogenic temperature at all times. PELDOR measurements were made using a refocused echo detection sequence that allows dipolar interaction between trapped radicals to be observed. The EPR spectrum is attributed to electron loss/gain DNA base radicals and neutral carbon-centered radicals that likely arise from sugar damage. We find a radical concentration of 13.5 x 10(18) cm(-3) in the tracks and a track radius of 6.79 nm. The cross section of these tracks is 144 nm2, yielding a lineal radical density of 2.6 radicals/nm. Based on the yields determined previously for particles having calculated LET values of 300-400 keV/microm and our measured lineal density, we obtain an LET of 270 keV/microm, which is in good agreement with the calculated range of values. These measurements of radical density and spatial extent provide the first direct experimental determination of track characteristics in irradiated DNA.     

Radioprotection of mouse colony forming units-spleen against heavy-charged particle damage by WR 2721

The effectiveness of S-2-(3-aminopropylamino)ethylphosphorothioic acid (WR 2721) to protect against the heavy-charged particle beams with dose-averaged LET infinity's ranging from 26 to 260 keV/micron was studied using the marrow colony forming units-spleen as a model system. WR 2721 (400 mg/kg) was injected ip 30 min before whole-body irradiation in the plateau ionization region of the Bragg curve. Significant protection was observed at 26, 51, and 135 keV/micron LET values where the data were collected with 20Ne, 28Si, and 40Ar ions, respectively. The largest component of protection was the slope change, where at LET values of 26 and 51 keV/micron the DMFs (slope) were 2.1 and 2.3, respectively, which are very close to the gamma-ray value of 2.4 (gamma LET approximately equal to 0.2 keV/micron). Protection, however, decreased with increase in LET from 51 to 135 keV/micron to the DMF value of 1.2 and no significant protection was observed against 56Fe ions at 260 keV/micron. Significant increases in extrapolation number occurred with gamma rays and neon particles. The results are discussed in terms of charged particle track structure, radiation chemistry, and potential clinical applications.     

Monte Carlo mixture model of lifetime cancer incidence risk from radiation exposure on shuttle and international space station

Estimating uncertainty in lifetime cancer risk for human exposure to space radiation is a unique challenge. Conventional risk assessment with low-linear-energy-transfer (LET)-based risk from Japanese atomic bomb survivor studies may be inappropriate for relativistic protons and nuclei in space due to track structure effects. This paper develops a Monte Carlo mixture model (MCMM) for transferring additive, National Institutes of Health multiplicative, and multiplicative excess cancer incidence risks based on Japanese atomic bomb survivor data to determine excess incidence risk for various US astronaut exposure profiles. The MCMM serves as an anchor point for future risk projection methods involving biophysical models of DNA damage from space radiation. Lifetime incidence risks of radiation-induced cancer for the MCMM based on low-LET Japanese data for nonleukemia (all cancers except leukemia) were 2.77 (90% confidence limit, 0.75-11.34) for males exposed to 1 Sv at age 45 and 2.20 (90% confidence limit, 0.59-10.12) for males exposed at age 55. For females, mixture model risks for nonleukemia exposed separately to 1 Sv at ages of 45 and 55 were 2.98 (90% confidence limit, 0.90-11.70) and 2.44 (90% confidence limit, 0.70-10.30), respectively. Risks for high-LET 200 MeV protons (LET=0.45 keV/micrometer), 1 MeV alpha-particles (LET=100 keV/micrometer), and 600 MeV iron particles (LET=180 keV/micrometer) were scored on a per particle basis by determining the particle fluence required for an average of one particle per cell nucleus of area 100 micrometer(2). Lifetime risk per proton was 2.68x10(-2)% (90% confidence limit, 0.79x10(-3)%-0. 514x10(-2)%). For alpha-particles, lifetime risk was 14.2% (90% confidence limit, 2.5%-31.2%). Conversely, lifetime risk per iron particle was 23.7% (90% confidence limit, 4.5%-53.0%). Uncertainty in the DDREF for high-LET particles may be less than that for low-LET radiation because typically there is very little dose-rate dependence. Probability density functions for high-LET radiation quality and dose-rate may be preferable to conventional risk assessment approaches. Nuclear reactions and track structure effects in tissue may not be properly estimated by existing data using in vitro models for estimating RBEs. The method used here is being extended to estimate uncertainty in spacecraft shielding effectiveness in various space radiation environments.     

His+ reversions caused in Salmonella typhimurium by different types of ionizing radiation

The yield of his+ reversions in the Ames Salmonella tester strain TA2638 has been determined for 60Co gamma rays, 140 kV X rays, 5.4 keV characteristic X rays, 2.2 MeV protons, 3.1 MeV alpha particles, and 18 MeV/U Fe ions. Inactivation studies were performed with the same radiations. For both mutation and inactivation, the maximum effectiveness per unit absorbed dose was obtained for the characteristic X rays, which have a dose averaged linear energy transfer (LET) of roughly 10 keV/micron. The ratio of the effectiveness of this radiation to gamma rays was 2 for inactivation and about 1.4 for the his+ reversion. For both end points the effectiveness decreases substantially at high LET, i.e., for the alpha particles and the Fe ions. The composition of the bottom and the top agar was the one recommended by Maron and Ames [Mutat. Res. 113, 173-215 (1983)] for application in chemical mutagenicity tests. The experiments with the less penetrating radiations differed from the usual protocol by utilization of a technique of plating the bacteria on the surface of the top agar. As in an earlier study [Roos et al., Radiat. Res. 104, 102-108 (1985)] greatly enhanced yields of mutations, relative to the spontaneous reversion rate, were obtained in these experiments by performing the irradiations 6 h after plating, which differs from the conventional procedure to irradiate the bacteria shortly after plating.     

Heavy ion effects on cellular DNA: strand break induction and repair in cultured diploid lens epithelial cells

The relative biological effectiveness (RBE) for the induction of DNA strand breaks and the efficiency of repair of these breaks in cultured diploid bovine lens epithelial cells was measured, using accelerated heavy ions in the linear energy transfer (LET)-range up to 16,200 keV/micron. At LET values above 800 keV/micron, the number of DNA strand breaks induced per particle increases both with the atomic number of the projectile and with its kinetic energy. About 90 per cent or more of the strand breaks induced by ions with an LET of less than 10,000 keV/micron are repaired within 24 h. Repair kinetics show a dependence on the particle fluence (irradiation dose). At higher particle fluences a higher proportion of non-rejoined breaks is found, even after prolonged periods of incubation. At any LET value, repair is much slower after heavy-ion exposure than after X-irradiation. This is especially true for low energetic particles with a very high local density of energy deposition within the particle track. At the highest LET value (16,200 keV/micron), no significant repair is observed.     

The effect of dose rate on radiation-induced neoplastic transformation in vitro by low doses of low-LET radiation

The dependence of the incidence of radiation-induced cancer on the dose rate of the radiation exposure is a question of considerable importance to the estimation of risk of cancer induction by low-dose-rate radiation. Currently a dose and dose-rate effectiveness factor (DDREF) is used to convert high-dose-rate risk estimates to low dose rates. In this study, the end point of neoplastic transformation in vitro has been used to explore this question. It has been shown previously that for low doses of low-LET radiation delivered at high dose rates, there is a suppression of neoplastic transformation frequency at doses less than around 100 mGy. In the present study, dose-response curves up to a total dose of 1000 mGy have been generated for photons from (125)I decay (approximately 30 keV) delivered at doses rates of 0.19, 0.47, 0.91 and 1.9 mGy/min. The results indicate that at dose rates of 1.9 and 0.91 mGy/min the slope of the induction curve is about 1.5 times less than that measured at high dose rate in previous studies with a similar quality of radiation (28 kVp mammographic energy X rays). In the dose region of 0 to 100 mGy, the data were equally well fitted by a threshold or linear no-threshold model. At dose rates of 0.19 and 0.47 mGy/min there was no induction of transformation even at doses up to 1000 mGy, and there was evidence for a possible suppressive effect. These results show that for this in vitro end point the DDREF is very dependent on dose rate and at very low doses and dose rates approaches infinity. The relative risks for the in vitro data compare well with those from epidemiological studies of breast cancer induction by low- and high-dose-rate radiation.