The neoplastic transformation potential of mammography X rays and atomic bomb spectrum radiation

Considerable controversy currently exists regarding the biological effectiveness of 29 kVp X rays which are used for mammography screening. This issue must be resolved to enable proper evaluation of radiation risks from breast screening. Here a definitive assessment of the biological effectiveness of 29 kVp X rays compared to the quality of radiation to which the atomic bomb survivors were exposed is presented for the first time. The standard radiation sources used were (a) an atomic bomb simulation spectrum and (b) 2.2 MeV electrons from a strontium-90/yttrium-90 (90Sr/90Y) radioactive source. The biological end point used was neoplastic transformation in vitro in CGL1 (HeLa x human fibroblast hybrid) cells. No significant difference was observed for the biological effectiveness of the two high-energy sources for neoplastic transformation. A limiting relative biological effectiveness (RBE(M)) of 4.42 +/- 2.02 was observed for neoplastic transformation by 29 kVp X rays compared to these two sources. This compares with values of 4.67 +/- 3.93 calculated from previously published data and 3.58 +/- 1.77 when the reference radiation was 200 and 220 kVp X rays. This suggests that the risks associated with mammography screening may be approximately five times higher than previously assumed and that the risk-benefit relationship of mammography exposures may need to be re-examined.     

Electron spectra and the RBE of X rays

For an assessment of the possible difference in effectiveness between mammography X rays and conventional X rays, the energy and LET spectra of the released electrons are examined. At photon energies below 20 keV and above 100 keV, the energy of the electrons increases with increasing photon energy, which implies that higher-energy photons produce less densely ionizing radiation and are therefore somewhat less effective per unit dose. However, in the intermediate energy range from 20 keV to 100 keV-the range that is relevant to medical diagnostics-the change from the photoelectric effect to the Compton effect causes a transient decrease of electron energies. The ionization density is therefore similar for 200 kVp X rays and 30 kVp mammography X rays, and the distributions of dose in LET suggest an RBE of 30 kVp mammography X rays compared to 200 kVp X rays of up to 1.3. This is in line with an earlier assessment by Brenner and Amols in terms of microdosimetric data, but it is strongly at variance with a recent claim that X rays for mammography are about four times more effective at small doses than conventional X rays and that they cause a correspondingly greater risk for breast cancer. Since LET need not be the only relevant factor, general response functions are examined here that specify-at low dose-the effect per electron of initial energy E and account, for example, for a particular role of the electron range. It is shown that, with any response per electron track that is a nondecreasing function of its starting energy, the low-dose RBE of the mammography X rays relative to the 200 kVp X rays must be substantially less than 2. The Auger electron that accompanies most photoelectrons, but only a minority of the Compton electrons, may increase the effectiveness of the mammography X rays somewhat, but it cannot explain the reported high values of the RBE.     

Is there reliable experimental evidence for a low-dose RBE of about 4 for mammography X rays relative to 200 kV X rays?

recently reported, on the basis of observations of neoplastic transformation in human hybrid CGL1 cells, a low-dose relative biological effectiveness (RBE(M)) of 4.3 for mammography X rays (29 kV) relative to 200 kV X rays. With reference to data in the literature, they inferred a factor of about 8 relative to 60Co gamma rays and concluded that this result is relevant to risk estimation. However, the conclusions do not appear to be valid. The data from the transformation study exhibit uncertainties in the statistical analysis that preclude any generalization of the inferred RBE(M). The data selected or inferred from the literature are likewise insufficient to support the stated RBEs. Our own uniform data set for the yields of dicentrics was obtained for widely varying photon energies with blood samples from the same donor, and it avoids interindividual variations in sensitivity as well as the differences in methodology that are associated with interlaboratory comparisons. Our data provide RBE(M) values for 29 kV X rays of 1.64 +/- 0.27 relative to 220 kV X rays and 4.75 +/- 1.67 and 6.12 +/- 2.51 relative to 60Co gamma rays.     

The effect of 29 kV X rays on the dose response of chromosome aberrations in human lymphocytes

The induction of chromosome aberrations in human lymphocytes irradiated in vitro with X rays generated at a tube voltage of 29 kV was examined to assess the maximum low-dose RBE (RBE(M)) relative to higher-energy X rays or 60Co gamma rays. Since blood was taken from the same male donor whose blood had been used for previous irradiation experiments using widely varying photon energies, the greatest possible accuracy was available for such an estimation of the RBE(M), avoiding the interindividual variations in sensitivity or differences in methodology usually associated with interlaboratory comparisons. The magnitude of the linear coefficient alpha of the linear-quadratic dose-effect relationship obtained for the production of dicentric chromosomes by 29 kV X rays (alpha = 0.0655 +/- 0.0097 Gy(-1)) confirms earlier observations of a strong increase in alpha with decreasing photon energy. Relating this value to previously published values of alpha for the dose-effect curves for dicentrics obtained in our own laboratory, RBE(M) values of 1.6 +/- 0.3 in comparison with weakly filtered 220 kV X rays, 3.0 +/- 0.7 compared to heavily filtered 220 kV X rays, and 6.1 +/- 2.5 compared to 60Co gamma rays have been obtained. These data emphasize that the choice of the reference radiation is of fundamental importance for the RBE(M) obtained. A special survey of the RBE(M) values obtained by different investigators in the narrow quality range from about 30 to 350 kV X rays indicates that the present RBE is in fairly good agreement with previously published findings for the induction of chromosome aberrations or micronuclei in human lymphocytes but differs from recently published findings for neoplastic transformation in a human hybrid cell line.     

Dependence of the yield of DNA double-strand breaks in Chinese hamster V79-4 cells on the photon energy of ultrasoft X rays

Induction of DNA DSBs by low-LET radiations reflects clustered damage produced predominantly by low-energy, secondary electron "track ends". Cell inactivation and induction of DSBs and their rejoining, assayed using pulsed-field gel electrophoresis, were determined in Chinese hamster V79-4 cells irradiated as a monolayer with characteristic carbon K-shell (CK) (0.28 keV), aluminum K-shell (AlK) (1.49 keV), and titanium K-shell (TiK) (4.55 keV) ultrasoft X rays under aerobic and anaerobic conditions. Relative to (60)Co gamma rays, the relative biological effectiveness (RBE) for cell inactivation at 10% survival and for induction of DSBs increases as the photon energy of the ultrasoft X rays decreases. The RBE values for cell inactivation and for induction of DSBs by CK ultrasoft X rays are 2.8 +/- 0.3 and 2.7 +/- 0.3, respectively, and by TiK ultrasoft X rays are 1.5 +/- 0.1 and 1.4 +/- 0.1, respectively. Oxygen enhancement ratios (OERs) of approximately 2 for cell inactivation and induction of DSBs by ultrasoft X rays are independent of the photon energy. The time scale for rejoining of DNA DSBs is similar for both ultrasoft X rays and 60Co gamma rays. From the size distribution of small DNA fragments down to 0.48 kbp, we concluded that DSBs are induced randomly by CK and AlK ultrasoft X rays. Therefore, ultrasoft X rays are more efficient per unit dose than gamma radiation at inducing DNA DSBs, the yield of which increases with decreasing photon energy.     

RBE of 10 kV X rays determined for the human mammary epithelial cell line MCF-12A

The dependence of relative biological effectiveness (RBE) on photon energy is a topic of extensive discussions. The increasing amount of in vitro data in the low-energy region indicates this to be a complex dependence that is influenced by the end point and cell line studied. In the present investigation, the RBE of 10 kV X rays (W anode) was determined relative to 200 kV X rays (W anode, 0.5 mm copper filter) for cell survival in the dose range 1-10 Gy and for induction of micronuclei in the range 0.5-3.6 Gy for MCF-12A human mammary epithelial cells. The RBE for cell survival was found to increase with decreasing dose, being 1.21+/-0.03 at 10% survival. Considerably higher values were obtained for micronucleus induction, where the RBE(M) obtained from the ratio of the linear coefficients of the dose-effect curves was 2.6+/-0.4 for the fraction of binucleated cells with micronuclei and 4.1+/-1.0 for the number of micronuclei per binucleated cell. These values, together with our previous data, support a monotonic increase in RBE with decreasing photon energy down to the mean energy of 7.3 keV used in the present study.     

The maximum low-dose RBE of 17.4 and 40 keV monochromatic X rays for the induction of dicentric chromosomes in human peripheral lymphocytes

Schmid et al. recently reported on the maximum low-dose RBE for mammography X rays (29 kV) for the induction of dicentrics in human lymphocytes. To obtain additional information on the RBE for this radiation quality, experiments with monochromatized synchrotron radiation were performed. Monochromatic 17.4 keV X rays were chosen for comparison with the diagnostic mammography X-ray spectrum to evaluate the spectral influence, while monochromatic 40 keV X rays represent a higher-energy reference radiation, within the experiment. The induction of dicentric chromosomes in human lymphocytes from one blood donor irradiated in vitro with 17.4 keV and 40 keV monochromatic X rays resulted in alpha coefficients of (3.44 +/- 0.87) x 10(-2) Gy(-1) and (2.37 +/- 0.93) x 10(-2) Gy(-1), respectively. These biological effects are only about half of the alpha coefficients reported earlier for exposure of blood from the same donor with the broad energy spectra of 29 kV X rays (mean energy of 17.4 keV) and 60 kV X rays (mean energy of 48 keV). A similar behavior is evident in terms of RBEM. Relative to weakly filtered 220 kV X rays, the RBEM for 17.4 and 40 keV monochromatic X rays is 0.86 +/- 0.23 and 0.59 +/- 0.24, respectively, which is in contrast to the RBEM of 1.64 +/- 0.27 for 29 kV X rays and 1.10 +/- 0.19 for 60 kV X rays. It is evident that the monochromatic radiations are less effective in inducing dicentric chromosomes than broad-spectrum X rays with the corresponding mean energy value. Therefore, it can be assumed that, for these X-ray qualities with broad energy spectra, a large fraction of the effects should be attributed predominantly to photons with energies well below the mean energy.     

Comparative microdosimetry of photoelectrons and Compton electrons: an analysis in terms of generalized proximity functions

A current discussion on mammography screening is focused on claims of high relative biological effectiveness (RBE) of mammography X rays compared to conventional 200 kV X rays. An earlier assessment in terms of the electron spectra of these radiations has led to the conclusion that the RBE is bound to be less than 2, regardless of specific model assumptions and the microdosimetric properties of electrons. The present study extends this result in terms of the microdosimetric proximity function, t(x), for electrons, which is essentially the spatial auto-correlation function of energy within particle tracks. If pairs of DNA lesions, e.g. chromosome breaks or deletions, bring about the observed damage, the value t(x) determines for a specified radiation the relative frequency of pairs of lesions a distance x apart. The effectiveness of the radiation is thus proportional to an average of the values of t(x) over the distances, x, for which lesions can combine. The analysis suggests that 15 keV electrons can have a low-dose relative biological effectiveness (RBE(M)) of 1.6 relative to 40 keV electrons if the interaction distances do not exceed about 1 micro m. An extension of the concept, the reduced proximity function, t(delta)(x), permits the inclusion of models with an energy threshold, such as delta = 100 eV, 500 eV or 2 keV, for the formation of each of the DNA lesions. This makes it possible to assess the potential impact of the Auger electrons which accompany most photoelectrons, but only a minority of the Compton electrons. It is found that the Auger electrons could make photoelectrons substantially more effective than Compton electrons at energies below 10 keV but not at energies above 15 keV. The conclusions obtained for the RBE of 15 keV electrons relative to 40 keV electrons will be roughly representative of the RBE of mammography X rays relative to conventional 200 kV X rays.     

Radiobiology of ultrasoft X rays. I. Cultured hamster cells (V79)

Ultrasoft X rays (approximately less than keV) provide a useful probe for the study of the physical parameters associated with the induction of biological lesions because the spatial scale of their energy depositions is of nanometer dimensions, comparable to that of critical structures within the cell. We report on cell-killing experiments using cultured hamster cells (V79) exposed to carbon K (0.28 keV), aluminum K (1.5 keV), copper K (8.0 keV), and 250 kVp X rays, under oxic and hypoxic conditions, and as a function of cell-cycle phase. Our principal results are: RBE increases with decreasing X-ray energy; OER decreases with decreasing X-ray energy; and cell-cycle response is similar for all X-ray energies. Our RBE results confirm earlier observations using ultrasoft X rays on mammalian cells. The shapes of fitted curves through the data for each energy are statistically indistinguishable from one another, implying that the enhanced effectiveness is purely dose modifying. The results reported herein generally support the view that single-track effects of radiation are predominantly due to very local energy depositions on the nanometer scale, which are principally responsible for observed radiobiological effects.     

DNA double-strand break misrejoining after exposure of primary human fibroblasts to CK characteristic X rays, 29 kVp X rays and 60Co gamma rays

The efficiency of ionizing photon radiation for inducing mutations, chromosome aberrations, neoplastic cell transformation, and cell killing depends on the photon energy. We investigated the induction and rejoining of DNA double-strand breaks (DSBs) as possible contributors for the varying efficiencies of different photon energies. A specialized pulsed-field gel electrophoresis assay based on Southern hybridization of single Mbp genomic restriction fragments was employed to assess DSB induction and rejoining by quantifying the restriction fragment band. Unrejoined and misrejoined DSBs were determined in dose fractionation protocols using doses per fraction of 2.2 and 4.4 Gy for CK characteristic X rays, 4 and 8 Gy for 29 kVp X rays, and 5, 10 and 20 Gy for 60Co gamma rays. DSB induction by CK characteristic X rays was about twofold higher than for 60Co gamma rays, whereas 29 kVp X rays showed only marginally elevated levels of induced DSBs compared with 60Co gamma rays (a factor of 1.15). Compared with these modest variations in DSB induction, the variations in the levels of unrejoined and misrejoined DSBs were more significant. Our results suggest that differences in the fidelity of DSB rejoining together with the different efficiencies for induction of DSBs can explain the varying biological effectiveness of different photon energies.     

Radiobiology of ultrasoft X rays. III. Normal human fibroblasts and the significance of terminal track structure in cell inactivation

Ultrasoft characteristic X rays from carbon (0.28 keV) are severely attenuated as they pass through biological material, causing a nonuniform distribution of dose to cell nuclei. Complications of studying ultrasoft X rays can be minimized in this context by using cells with very thin cytoplasm and nuclei (e.g., less than the attenuation length of the X rays), and which exhibit a more nearly exponential dose response to cell killing, such as normal human fibroblasts compared with V79 cells. Using this cell system, we report the relative biological effectiveness (RBE) of A1-K and C-K X rays to be near unity. Previous studies of cell inactivation by characteristic carbon X rays gave RBEs of 3 to 4, supporting the idea that localized energy depositions from secondary electrons and primary track ends represent the principal mode of biological action for other low-LET radiations. In part, the reported high RBEs result from the use of mean dose to describe energy deposited within the cell nuclei by these poorly penetrating radiations. Implicit in the use of mean dose is that cellular damage varies linearly with dose within a critical target(s), an assumption that is of questionable validity for cells that exhibit pronounced curvilinear dose responses. The simplest interpretation of the present findings is that most energy depositions caused by track-end effects are not necessarily more damaging than the sparsely ionizing component.     

Induction of incomplete and complex chromosome aberrations by 30 kVp X rays

X rays of 26-30 kVp are routinely used for mammography screening. For radioprotection purposes, a quality factor (Q) of 1 is assumed for all photon energies, but it is thought that the relative biological effectiveness (RBE) increases as the photon energy decreases. The analysis of radiation-induced chromosome aberrations is one of the most widely used methods to study the interaction between radiation and DNA. Here we present a FISH study on metaphases from peripheral blood samples irradiated with three different X-ray energies (30, 80 and 120 kVp). The study comprises two FISH approaches: one using pantelomeric and pancentromeric probes to evaluate the induction of incomplete chromosome aberrations and the other using mFISH to evaluate the induction of complex chromosome aberrations. The results indicate that exposure to 30 kVp X rays resulted in a modest increase in the induction of incomplete elements and complex aberrations compared to 80 and 120 kVp X rays.     

RBE of X rays of different energies: a cytogenetic evaluation by FISH

Mammography using 26-30 kVp X rays is routinely used in breast cancer screening. Discussion about the radiation-related risk associated with this methodology is ongoing. For radioprotection purposes, a quality factor of 1 has been assigned for all photon energies. However, the relative biological effectiveness (RBE) could increase as the photon energy decreases. Analyzing different biological parameters, for 30 kVp X rays, RBE values from 1 to 8 have been estimated. In the present study, a cytogenetic FISH evaluation of the RBE of 30, 80 and 120 kVp X rays has been done. Blood samples were irradiated with 10 doses from 0.05 to 3 Gy for each energy studied. The yields of translocations and dicentrics were determined by fluorescence in situ hybridization (FISH) using whole chromosome probes for chromosomes 1, 4 and 11 together with a pancentromeric probe. The alpha coefficients of the dose-effect curves for dicentrics, minimum number of breaks needed to produce exchange-type aberrations, and apparently simple translocations were used to estimate the RBE. Using the curves obtained for 120 kVp as a reference, the RBE values for dicentrics were 1.08+/-0.43 and 1.73+/-0.59 for 80 and 30 kVp X rays, respectively; for minimum number of breaks these values were 1.38+/-0.39 and 1.42+/-0.41, and for apparently simple translocations they were 1.26+/-0.40 and 1.51+/-0.47, respectively. Moreover, the induction of complex aberrations by these energies was compared. The percentage of complex aberrations relative to total aberrations showed a significant tendency to increase as X-ray energy decreased: 7.8+/-1.19, 9.8+/-1.6 and 14.1+/-1.9 for 120, 80 and 30 kVp, respectively (P<0.02).     

Neoplastic cell transformation by heavy ions

We have studied the induction of morphological transformation by heavy ions. Golden hamster embryo cells were irradiated with 95 MeV 14N ions (530 keV/microns), 22 MeV 4He ions (36 keV/microns), and 22 MeV 4He ions with a 100-microns Al absorber (77 keV/microns) which were generated by a cyclotron at the Institute of Physical and Chemical Research in Japan. Colonies were considered to contain neoplastically transformed cells when the cells were densely stacked and made a crisscross pattern. It was shown that the induction of transformation was much more effective with 14N and 4He ions than with gamma or X rays. The relative biological effectiveness (RBE) relative to 60Co gamma rays was 3.3 for 14N ions, 2.4 for 4He ions, and 3.3 for 4He ions with a 100-microns Al absorber. The relationship between RBE and linear energy transfer was qualitatively similar for both cell death and transformation.     

Radiobiology of ultrasoft X rays. IV. Flat and round-shaped hamster cells (CHO-10B, HS-23)

The results reported earlier in this series indicated that the relative biological effectiveness (RBE) of ultrasoft X rays decreases with decreasing cell thickness, approaching unity for the thinnest cells used, plateau-phase human skin fibroblasts (HSF). The possible dependence of RBE on the configuration of the cell nucleus is investigated further in this paper using two CHO cell lines that attach well and have similar intrinsic radiosensitivities to 60Co gamma rays. One of the lines forms monolayers similar to V79 cells, while the other remains more spherical during growth. We find an increasing RBE with decreasing X-ray energy for both of these cell lines, consistent with our results using V79 cells. Also consistent with our results obtained with 10T1/2 and HSF cells, we find an increasing RBE with increasing cell thickness. The possible dependence of RBE on radiosensitivity and the use of the concept of mean dose for ultrasoft X rays is discussed.     

The RBE of neutrons for induced mitotic gene conversion in "error-prone repair" defective yeast

Mitotic gene conversion was induced in the diploid yeast strain D7.rad6 which lacks "error-prone repair" and thus does not mutate. Neutrons (14.5 MeV), 60Co gamma rays, and 150 kVp X rays delivered under oxic or anoxic conditions were compared for their ability to induce gene conversion. Doses were chosen to minimize cell killing. A lack of induced mutation in this strain at the ilv1-92 allele was confirmed. Gene conversion of the trp5-27/trp5-12 alleles was induced with a linear dose response, and the yield of convertants per gray was significantly enhanced over yields reported previously for a wild-type stain. The relative biological effectiveness (RBE) of neutrons relative to low-LET radiations was found to be about 2.2 for either oxic or anoxic radiation in contrast to wild-type where the oxic RBE was 1.7 and the anoxic RBE 2.7. Absence of the rad6 function was therefore associated with an altered RBE for the conversional end point. The oxygen enhancement ratio (OER) for gene conversion was found to be about 1.7 for all radiations in contrast to the wild type where the OER for neutrons was 1.7, but for low-LET radiations it was 2.7. As repair of ionizing damage in the rad6 strain did not lead to mutation, owing to the loss of "error-prone repair," the changes in yield, RBE, and OER were consistent with the hypothesis that some of the lesions processed by wild type to generate mutations could, in the rad6 strain, lead instead to gene conversion.     

Radiobiology of ultrasoft X rays. II. Cultured C3H mouse cells (10T1/2)

In the first paper of this series (Radiat. Res. 110, 396-412 (1987], using V79 cells, we reported that the relative biological effectiveness (RBE) of ultrasoft X rays was found to increase with decreasing energy, and the oxygen enhancement ratio (OER) was found to decrease with decreasing energy. In this report, we present RBE and OER results for 10T1/2 cells that are known to grow uniformly flat and are considerably thinner than V79 cells. Thus the variation in dose across the cell nucleus is considerably reduced. The OER results agree well with our earlier V79 results. However, the RBE values for 10T1/2 cells compared to V79 cells are systematically less for all soft X rays and especially for 0.28 keV carbon-K (1.3 compared to 3.4 for V79 cells). Some plausible explanations are presented to reconcile the apparent discrepancy between V79 and 10T1/2 results.     

Measurement of the initial levels of DNA damage in human lymphocytes induced by 29 kV X rays (mammography X rays) relative to 220 kV X rays and gamma rays

Experiments using the alkaline comet assay, which measures all single-strand breaks regardless of their origin, were performed to evaluate the biological effectiveness of photons with different energies in causing these breaks. The aim was to measure human lymphocytes directly for DNA damage and subsequent repair kinetics induced by mammography 29 kV X rays relative to 220 kV X rays, 137Cs gamma rays and 60Co gamma rays. The level of DNA damage, predominantly due to single-strand breaks, was computed as the Olive tail moment or percentage DNA in the tail for different air kerma doses (0.5, 0.75, 1, 1.5, 2 and 3 Gy). Fifty cells were analyzed per slide with a semiautomatic imaging system. Data from five independent experiments were transformed to natural logarithms and fitted using a multiple linear regression analysis. Irradiations with the different photon energies were performed simultaneously for each experiment to minimize interexperimental variation. Blood from only one male and one female was used. The interexperimental variation and the influence of donor gender were negligible. In addition, repair kinetics and residual DNA damage after exposure to a dose of 3 Gy were evaluated in three independent experiments for different repair times (10, 20, 30 and 60 min). Data for the fraction of remaining damage were fitted to the simple function F(d) = A/(t + A), where F(d) is the fraction of remaining damage, t is the time allowed for repair, and A (the only fit parameter) is the repair half-time. It was found that the comet assay data did not indicate any difference in the initial radiation damage produced by 29 kV X rays relative to the reference radiation types, 220 kV X rays and the gamma rays of 137Cs and 60Co, either for the total dose range or in the low-dose range. These results are, with some restrictions, consistent with physical examinations and predictions concerning, for example, the assessment of the possible difference in effectiveness in causing strand breaks between mammography X rays and conventional (150-250 kV) X rays, indicating that differences in biological effects must arise through downstream processing of the damage.     

Occurrence of mammary tumors in rats after exposure to tritium beta rays and 200-kVp X rays

The RBE for tritium was estimated in reference to 200-kVp X rays, using acceleration of breast tumor appearance in the female Sprague-Dawley rat as the end-point. Chronic X-ray doses of 0.3-2.0 Gy were delivered over 10 days. Intraperitoneal injections of tritiated water ranging in concentrations from 45 to 370 MBq/100 g body wt were administered, followed by four additional injections at 2-day intervals and half of the initial concentrations. Seventy-five percent of the total tritium dose was delivered to the mammary gland within the first 10 days and 95% within the first 20 days after the start of the tritium exposure. RBE estimations were based on various criteria including the tumor incidence per Gy at 450 days postirradiation and the time required to induce tumors in 50% of the animals at risk. The results suggest that tritium beta rays are about 1.1-1.3 times more effective than chronic 200-kVp X rays for acceleration of the appearance of rat mammary tumors. However, the uncertainties involved in these calculations are such that the effects of tritium beta rays could not be reliably distinguished from those of chronic 200-kVp X rays. Measured differences in RBE values were slightly larger for the comparison between acute and chronic X rays than for the comparison between chronic tritium beta rays and chronic X rays.     

The effectiveness of monoenergetic neutrons at 565 keV in producing dicentric chromosomes in human lymphocytes at low doses

The induction of dicentric chromosomes in human lymphocytes from one individual irradiated in vitro with monoenergetic neutrons at 565 keV was examined to provide additional data for an improved evaluation of neutrons with respect to radiation risk in radioprotection. The resulting linear dose-response relationship obtained (0.813 +/- 0.052 dicentrics per cell per gray) over the dose range of 0.0213-0.167 Gy is consistent with published results obtained for irradiation with neutrons from different sources and with different spectra at energies lower than 1000 keV. Comparing this value to previously published "average" dose-response curves obtained by different laboratories for (60)Co gamma rays and orthovoltage X rays resulted in maximum RBEs (RBE(m)) of about 37 +/- 8 and 16 +/- 4, respectively. However, when our neutron data were matched to low-LET dose responses that were constructed several years earlier for lymphocytes from the same individual, higher values of RBE(m) resulted: 76.0 +/- 29.5 for (60)Co gamma rays and 54.2 +/- 18.4 for (137)Cs gamma rays; differentially filtered 220 kV X rays produced values of RBE(m) between 20.3 +/- 2.0 or 37.0 +/- 7. 1. The results highlight the dependence of RBE(m) on the choice of low-LET reference radiation and raise the possibility that differential individual response to low-LET radiations may need to be examined more fully in this context.