Regulation of red blood cell production by erythropoietin: normal mouse marrow in vitro.

The response of immediate erythroid precursors to timed exposures of erythropoietin (Ep) were investigated by examining the formation of murine erythroid colonies in vitro. At various times after the initiation of normal marrow cultures with Ep, enough anti-erythropoietin serum was added to neutralize the actions of the hormone for the rest of the 48 h culture period in plasma clots. Results show that a significant number of erythroid precursors develop to a mature colony after very short Ep exposures (as little as 18 min), but they account for only about 13% of the total colonies generated when Ep is active for 48 h. Only if Ep is active for more than 6 h do additional colonies form, in direct proportion to the length of the Ep exposure. Maximum erythroid colony production requires prolonged exposure to Ep, and the longer Ep is active, the larger the percentage of erythroid colonies that reach 17 cells or more. The mechanism(s) behind these two effects appear intimately connected. These results suggest the existence of two populations of Ep-responsive cells in the immediate erythroid precursor compartment of mice: (1) A CFUE pool, which is on the verge of commitment into the recognizable erythron, and requires very short Ep exposures for this entry; and (2) a "pre-CFUE" stage which feeds into the CFUE pool in the presence of Ep, and which produces colonies only after longer Ep exposures. It is proposed that the pre-CFUE response to Ep involves 1-2 self-replications before entering the CFUE population.

Molecular and cellular radiobiology of heavy ions.

Quantitative studies at the BEVALAC have demonstrated some of the physical and radiobiological factors that promise to make accelerated heavy ions important for the therapy of cancer. The measured physical dose-biological effect relationships allow the safe and effective delivery of therapeutic schedules of heavy ions. Among the charged particle beams available, carbon, neon and helium ions in the "extended Bragg peak mode" have optimal physical and biological effectiveness for delivery of therapy to deep seated tumors. The depth-dose profiles of these beams protect intervening and adjacent tissues as well as tissues beyond the range of the particles. For the treatment of hypoxic tumors, silicon and argon beams are being considered because they significantly depress the radiobiological oxygen effect in the region of the extended Bragg ionization peak. The depth-effectiveness of the argon beam is somewhat limited, however, because of primary particle fragmentation. Silicon beams have a depth-dose profile which is intermediate between that of neon and argon, and are candidates to become the particle of choice for maximizing high LET particle effects. Heavy accelerated ions depress enzymatic repair mechanisms, decrease variations of radiosensitivity during the cell division cycle, cause greater than expected delays in cell division, and decrease the protective effects of neighboring cells in organized systems. Near the Bragg peak, enhancement of heavy particle effects are observed in split dose schedules. Late and carcinogenic effects are being studied. With the newly developed Repair-Misrepair theory we can quantitatively model most observations.

Treatment of cancer with heavy charged particles.

A clinical radiotherapeutic trial using heavy charged particles in the treatment of human cancers has accrued over 400 patients since 1975, 378 of whom were treated with particles and 28 with low LET photons as control patients. Heavy charged particle radiotherapy offers the potential advantages of improved dose localization and/or enhanced biologic effect, depending on particle selected for treatment. Target sites have included selected head and neck tumors, ocular melanomata, malignant gliomata of the brain, carcinoma of the esophagus, carcinoma of the stomach, carcinoma of the pancreas, selected juxtaspinal tumors and other locally advanced, unresectable tumors. A Phase III prospective clinical trial has been started in carcinoma of the pancreas using helium ions. Phase I-II studies are underway with heavier particles such as carbon, neon and argon ions in order to prepare for prospective Phase III trials. Silicon ions are also under consideration for clinical trial. These studies are supported by the United States Department of Energy and National Institutes of Health.

The historical background for large-animal studies with neutrons of various energies.

Studies of the biological effects of fast neutrons were started as soon as adequate sources became available, and the first of these was the cyclotron at Berkeley. Limitations on methods for physical dosimetry render most of the earliest studies nearly useless, and the first landmark studies with modern methodology were done at Berkeley in the 1950s. This report is an attempt to review and summarize all the work that took place from 1950 to 1965. The RBE for fast neutrons in the dog is generally in the range of 1.0 or less, depending upon reference radiation, the biological end point chosen, and the exposure geometry. Also the RBE seems to be relatively independent of the fast-neutron spectrum, at least for the range of values that has been studied. Mean after-survival times in irradiated dogs and goats indicate that the bone marrow syndrome usually predominates in large animals irradiated with fast neutrons. The gastrointestinal threshold for the dog is around 1200 cGy, and for neutrons the biological effectiveness for the GI syndrome still does not bring the GI threshold down to the point where it predominates over the bone marrow syndrome. As expected from biological first principles, for dose rates in excess of a few tens of centigrays per minute, dose rate seems to have no effect on the outcome with either neutron, gamma or X irradiations when the end point is lethality.

Radiation-induced division delay in 9L spheroid versus monolayer cells.

The technique of percentage labeled mitoses was used to compare radiation-induced division delay in 9L rat gliosarcoma cells growing as spheroids or as exponential monolayers. The length of delay induced by each of five X-ray doses was determined as the difference between control and irradiated cultures in the time required to reach the half-height of the first peak of labeled mitoses. Spheroid cells were delayed significantly longer than monolayer cells; the slopes of the dose responses were 32 and 13 min/Gy, respectively. Cells in small spheroids (150 micron diameter) were delayed to the same extent as cells in large spheroids (800 micron diameter). Like the contact effect previously observed as enhanced radiation survival of cells grown as spheroids, the increased radiation-induced delay may be a consequence of the growth of cells in three-dimensional contact.

Differences in the X-ray sensitivity of cells in different regions of the sandwich, a diffusion-limited system for cell growth.

The sandwich system was recently developed as a tumor analog; like spheroids, sandwiches are diffusion-limited multicellular systems which exhibit a necrotic center and a viable cell border. Using sandwiches of the 9L cell line, we compared the X-ray sensitivity of cells in the inner half of the viable border, adjacent to the necrotic center, with that of cells in the outer half of the viable border, adjacent ot the medium. No cells were hypoxic at the time of irradiation. The cells in the inner half of the viable border exhibited an increased radioresistance over cells in the outer half. The effect was dose multiplying with a multiplying factor of 1.5. Besides the sandwich studies, the X-ray sensitivity of 9L plateau monolayer cultures (induced by starvation) was compared to exponentially growing monolayer cultures. The plateau cultures exhibited an increased radioresistance over the exponentially growing cultures. The effect was also dose multiplying with a multiplying factor of 1.5.

The RBE-LET relationship for rodent intestinal crypt cell survival, testes weight loss, and multicellular spheroid cell survival after heavy-ion irradiation.

This report presents data for survival of mouse intestinal crypt cells, mouse testes weight loss as an indicator of survival of spermatogonial stem cells, and survival of rat 9L spheroid cells after irradiation in the plateau region of unmodified particle beams ranging in mass from 4He to 139La. The LET values range from 1.6 to 953 keV/microns. These studies examine the RBE-LET relationship for two normal tissues and for an in vitro tissue model, multicellular spheroids. When the RBE values are plotted as a function of LET, the resulting curve is characterized by a region in which RBE increases with LET, a peak RBE at an LET value of 100 keV/microns, and a region of decreasing RBE at LETs greater than 100 keV/microns. Inactivation cross sections (sigma) for these three biological systems have been calculated from the exponential terminal slope of the dose-response relationship for each ion. For this determination the dose is expressed as particle fluence and the parameter sigma indicates effect per particle. A plot of sigma versus LET shows that the curve for testes weight loss is shifted to the left, indicating greater radiosensitivity at lower LETs than for crypt cell and spheroid cell survival. The curves for cross section versus LET for all three model systems show similar characteristics with a relatively linear portion below 100 keV/microns and a region of lessened slope in the LET range above 100 keV/microns for testes and spheroids. The data indicate that the effectiveness per particle increases as a function of LET and, to a limited extent, Z, at LET values greater than 100 keV/microns. Previously published results for spread Bragg peaks are also summarized, and they suggest that RBE is dependent on both the LET and the Z of the particle.

Differential repair of potentially lethal damage in exponentially growing and quiescent 9L cells.

The alteration of potentially lethal damage repair by postirradiation treatment with hypertonic saline (0.5 M PBS) was investigated in exponentially growing and quiescent 9L cells in vitro. A single dose of X rays (8.5 Gy) immediately followed by a 30-min treatment with hypertonic PBS at 37 degrees C reduced the survival of exponentially growing 9L cells by a factor of 13-18 compared to survival of irradiated immediately and delayed-plated cells, while the survival of quiescent cells was reduced by only a factor of 5-8. Survival curves confirmed the relative resistance of the quiescent 9L cells versus exponentially growing 9L cells to X rays plus hypertonic treatment. Both the slope and the shoulder of the survival curve were reduced to a greater extent in exponentially growing cells than in the quiescent cells by hypertonic treatment. The response of quiescent cells cannot be explained by either the duration of hypertonic treatment or the redistribution of the cells into G1 phase. We show that quiescent 9L cells can recover from hypertonically induced potentially lethal damage when incubated under conditions which have been found to delay progression through the cell cycle, and postulate that an altered chromatin structure or an enhanced repair capacity of quiescent 9L cells may be responsible for their resistance.

Quiescence in 9L cells and correlation with radiosensitivity and PLD repair.

The onset of quiescence, changes in X-ray sensitivity, and changes in capacity for potentially lethal damage (PLD) repair of unfed plateau-phase 9L44 cell cultures have been systematically investigated. The quiescent plateau phase in 9L cells was the result of nutrient deprivation and was not a cell contact effect. Eighty-five to 90% of the plateau-phase cells had a G1 DNA content and a growth fraction less than or equal to 0.15. The cell kinetic shifts in the population were temporally correlated with a developing radioresistance, which was characterized by a larger shoulder in the survival curve of the quiescent cells (Dq = 5.71 Gy) versus exponentially growing cells (Dq = 4.48 Gy). When the quiescent plateau-phase cells were refed, an increase in radiosensitivity resulted which approached that of exponentially growing 9L cells. Delayed plating experiments after irradiation of exponentially growing cells, quiescent plateau-phase cells, and synchronized early to mid-G1-phase cells indicated that while significant PLD repair was evident in all three populations, the quiescent 9L cells had a higher PLD repair capacity. Although data for immediate plating indicated that 9L cells may enter quiescence in the relatively radioresistant mid-G1 phase, the enhanced PLD repair capacity of quiescent cells cannot be explained by redistribution into G1 phase. When the unfed quiescent plateau-phase 9L cells were stimulated to reenter the cell cycle by replating into fresh medium, the first G1 was extended by 6 h compared with the G1 of exponentially growing or refed plateau-phase 9L cells.(ABSTRACT TRUNCATED AT 250 WORDS)

Tumorigenic potential of high-Z, high-LET charged-particle radiations.

The potential for radiogenic neoplasia from charged-particle irradiation has been estimated using the Harderian gland of the mouse as a test system. Particles ranging in Z from Z = 1 (proton) to Z = 41 (niobium), in energy from 228 to 670A MeV, and in LET from 0.4 to 464 keV/microns were produced at the Lawrence Berkeley Laboratory BEVALAC. Expression of the tumorigenic potential of the initiated cells was enhanced by hormones from isogeneic grafts of pituitaries. The goal of the studies was to estimate the initial slope of the relationship between increased tumor prevalence at 16 months after irradiation and the dose received. Initial slopes were measured with good precision for 60Co gamma rays and the Bragg plateau beams of 228A MeV 4He ions, 600A MeV 56Fe ions, and 350A MeV 56Fe ions. The ratio of the initial slope for these ions to that of 60Co gamma rays give an estimate of the maximum RBE for radiogenic neoplasia. These values were 2.3 for the 4He ions, 40 for 600A MeV 56Fe, and 20 for 350A MeV 56Fe. In the studies reported here the prevalence of tumors as the result of pituitary isografts was not enhanced after irradiation with 56Fe ions. It remains to be seen how effective pituitary isografts are for enhancement of radiogenic neoplasia from other ions at different LET values. A risk analysis was undertaken using particle fluence rather than dose as the independent variable. This analysis provides a value for a "cross section" expressed in microns 2. This parameter expresses as the increase in proportion of mice with one or more Harderian gland tumors per unit increase in particle fluence. The plot of the cross section (risk coefficient) as a function of LET is monotonic, with no clear evidence of a maximum value of the risk coefficient for even the highest LET particle used.

Recovery from potentially lethal damage and recruitment time of noncycling clonogenic cells in 9L confluent monolayers and spheroids.

Cells that have been grown as multicell tumor spheroids exhibit radioresistance compared to the same cells grown in monolayers. Comparison of potentially lethal damage (PLD) repair and its kinetics was made between 9L cells grown as spheroids and confluent monolayers. Survival curves of cells plated immediately after irradiation showed the typical radioresistance associated with spheroid culture compared to plateau-phase monolayers. The dose-modification factor for spheroid cell survival is 1.44. Postirradiation incubations in normal phosphate-buffered saline (PBS), conditioned media, or 0.5 M NaCl in PBS reduced the differences in radiosensitivity between the two culture conditions. Postirradiation treatment in PBS or conditioned medium promoted repair of potentially lethal damage, and 0.5 M NaCl prevented the removal of PLD and allowed the fixation of damage resulting in lower survival. Survival of spheroid and monolayer cells after hypertonic NaCl treatment was identical. NaCl treatment reduced Do more than it did the shoulder (Dq) of the survival curve. PLD repair kinetics measured after postirradiation incubation in PBS followed by hypertonic NaCl treatment was the same for spheroids and for plateau-phase monolayers. The kinetics of PLD repair indicates a biphasic phenomenon. There is an initial fast component with a repair half-time of 7.9 min and a slow component with a repair half-time of 56.6 min. Most of the damage (59%) is repaired slowly. Since the repair capacity and kinetics are the same for spheroids and monolayers, the radioresistance of spheroids cannot be explained on this basis. Evidence indicates that the time to return from a Go (noncycling G1 cells) state to a proliferative state (recruitment) for cells from confluent monolayers and from spheroids after dissociation by protease treatment may be the most important determinant of the degree of PLD repair that occurs. Growth curves and flow cytometry cell cycle analysis indicate that spheroid cells have a lag period for reentry into a proliferative state. Since plating efficiency remains high and unchanging during this period, one cannot account for the delay on the basis of the existence of a large fraction of Go cells which are not potentially clonogenic. The cell cycle progression begins in 6-8 h for monolayer cells and in 14-15 h for spheroids. It is hypothesized that the slower reentry of spheroid cells into a cycling phase allows more time for repair than for the rapidly proliferating monolayer cells.

The relative biological effectiveness of 670 MeV/A neon as a function of depth in water for a tissue model.

Linear energy transfer (LET infinity) spectra of identified charge fragments and primaries, produced by nuclear interactions of 670 MeV/A neon in water, were measured along the unmodulated Bragg curve of the neon beam. The relative biological effectiveness (RBE) values for spermatogonial cell killing, as reported on the basis of weight loss assay of mouse testes irradiated with beams of approximately constant single LET infinity, were summed over the particle LET infinity spectra to obtain an effective RBE for each charged-particle species, as a function of water absorber thickness. The resultant values of effective RBE were combined to obtain an effective RBE for the mixed radiation field. The RBE calculated in this way was compared with experimental RBEs obtained for spermatogonial cell killing in the mixed radiation field produced by neon ions traversing a thick water absorber. Discrepancies of 10-40% were observed between the calculated RBE and the RBE measured in the mixed radiation field. Part of this discrepancy can be attributed to undetected low-Z fragments, whose contribution is not included in the calculation, leading to an overestimated value for the calculated RBE. On the other hand, calculated values 10% greater than the measured RBE are explained as track structure effects due to the higher radial ionization density near neon tracks relative to the ionization density near the silicon tracks used to fit the RBE vs LET infinity data.

Radiation-induced renal damage: the effects of hyperfractionation.

The response of mouse kidneys to multifraction irradiation was assessed using three nondestructive functional end points. A series of schedules was investigated giving 1, 2, 4, 8, 16, 32, or 64 equal X-ray doses, using doses per fraction in the range of 0.9 to 16 Gy. The overall treatment time was kept constant at 3 weeks. Kidney function was assessed from 19 to 48 weeks after irradiation by measuring changes in isotope clearance, urine output, and hematocrit. The degree of anemia (assessed from the hematocrit measurements) is a newly developed assay which is an early indicator of the extent of renal damage after irradiation. All three assays yielded steep dose-effect curves from which the repair capacity of kidney could be estimated by comparing the isoeffective doses in different schedules. There was a marked influence of fractionation, with increasing dose being required to achieve the same level of damage for increasing fraction number, even between 32 and 64 fractions. The data are well fitted by a linear quadratic dose-response equation, and analysis of the data in this way yields low values (approximately 3.0 Gy) for the ratio alpha/beta. This would suggest that hyperfractionation , using extremely small X-ray doses per fraction, would spare kidneys relative to tumors and acutely responding tissues.

Response of colony-forming units-spleen to heavy charged particles.

Survival of colony-forming units-spleen (CFU-S) was measured after single doses of photons or heavy charged particles from the BEVALAC. The purposes were to define the radiosensitivity to heavy ions used medically and to evaluate relationships between relative biological effectiveness (RBE) and dose-averaged linear energy transfer (LET infinity). In in vitro irradiation experiments. CFU-S suspensions were exposed to 220 kVp X rays or to 20Ne (372 MeV/micron) or 40Ar (447 MeV/micron) particles in the plateau portion of the Bragg curve. In in vivo irradiation experiments, donor mice from which CFU-S were harvested were exposed to 12C (400 MeV/micron). 20Ne (400 or 670 MeV/micron), or 40Ar (570 MeV/micron) particles in Bragg peaks spread to 4 or 10 cm by spiral ridge filters. Based on RBE at 10 survival, the maximum RBE of 2.1 was observed for 40Ar particles characterized by an LET infinity of approximately 100 keV/micron. Lower RBEs were determined at lower or higher estimated values of LET infinity and ranged from 1.1 for low energy 40Ar particles to 1.5-1.6 for low energy 12C and 20Ne. The responses of CFU-S are compared with responses of other model systems to heavy charged particles and with the reported sensitivity of CFU-S to neutrons of various energies. The maximum RBE reported here, 2.1 for high energy 40Ar particles, is somewhat lower than values reported for fission-spectrum neutrons, and is appreciably lower than values for monoenergetic 0.43-1.8 MeV neutrons. Low energy 12C and 20Ne particles have RBEs in the range of values reported for 14.7 MeV neutrons.

An experimental compartmental flow model for assessing the hemodynamic response of intracranial arteriovenous malformations to stereotactic radiosurgery.

Stereotactic radiosurgery has proven to be an effective method of treating selected inaccessible or inoperable arteriovenous malformations (AVMs) of the brain. Radiation-induced obliteration of successfully-treated AVMs, however, occurs only after some latent period after treatment, depending on size, location, and dose. An experimental compartmental flow model is proposed to describe the hemodynamic alterations in the AVM as a result of the pathophysiological changes after radiosurgery, and to analyze temporal alterations in AVM blood flow rates and pressure gradients before complete obliteration. In representative small (low-flow, 150 ml/min) and large (high-flow, 440 ml/min) AVMs, it is found that increases in pressure gradients across certain vascular structures within the AVM occur during the normal course of radiation-induced flow decrease and AVM obliteration. The magnitude of these pressure alterations, however, may be within the normal physiological variations in cerebrovascular blood pressure. The effects of partial-volume irradiation of the AVM is examined by limiting radiosurgical treatment to varying portions of the flow compartments within the model. It is found that alterations in pressure gradients persist in unirradiated vascular shunts, even after complete obliteration of the treated AVM volume. These pressure alterations may increase the probability of hemorrhage from the untreated shunts of the AVM and cause redistribution of regional cerebral blood flow resulting in increased flow through these untreated shunts.

Imaging by injection of accelerated radioactive particle beams.

The process of imaging by detection of the annihilation gamma rays generated from positron emitters which have been injected into a patient by a particle accelerator has been studied in detail. The relationships between patient dose and injected activity have been calculated for C-11, N-13, C-15, F-17, and Ne-19 and measured for C-11 and Ne-19 with good agreement with the calculations. The requirements for imaging of the small amounts of activity that can be injected safely have been analyzed in terms of one specific application of the radioactive beam injection technique, that of Bragg peak localization in support of radiotherapy by heavy ions. The characteristics of an existing camera with sufficient sensitivity and spatial accuracy for that task are described. Results of the calculations of radioactive beam flux requirements are shown.

Radiation damage to glucose concentrating capacity and cell survival in kidney tubule cells: effects of fractionation.

The glucose concentrating capacity of cultured LLC-PK1 kidney epithelial cells has been measured after single and fractionated doses of X-rays. Steady-state glucose concentrating capacity (ratio of glucose concentration inside to outside cell) can be measured using radiolabelled analogues of glucose which are actively transported but not metabolized. These cells can be stimulated to increase their glucose concentrating capacity (up-regulation) by a reduction in the glucose concentration of the growth medium. However, after X-ray irradiation the cells have a reduced capacity to respond to up-regulation. This effect can be measured 7 days after irradiation and before radiation-induced cell killing affects the cell population. The previously reported radiosensitivity of this function to single doses of X-rays (in the range 1-16 Gy) was confirmed. Surprisingly, no significant sparing of this effect could be measured by fractionation of the X-ray dose into two or four fractions. However, the cells showed a significant fractionation effect if clonogenic survival was measured using the standard cell survival assay. These early effects have different fractionation response from the later phases of tissue damage, measured months to years after irradiation, which do show sparing due to fractionation and are thought to be mainly due to changes in cell survival. The lack of sparing by fractionation to the functional damage may suggest a different target from that which determines cell survival. These results support the hypothesis that radiation damages cellular functions, separately from cell replication.

Accelerated heavy ions and the lens. IX. Late effects of LET and dose on cellular parameters in the murine lens.

Lenses of mice irradiated with 250 MeV protons, 670 MeV/amu 20Ne, 600 MeV/amu 56Fe, 350 MeV/amu 56Fe, 600 MeV/amu 93Nb or 593 MeV/amu 139La ions were evaluated by analysing cytopathological indicators which have been implicated in the cataractogenic process. The LETs ranged from 0.39 to 953 keV/microns and the fluences from 1.31 x 10(3)/mm2 to 5.12 x 10(7)/mm2. The lenses were assessed 64 weeks post-irradiation in order to observe the late effects of LET and dose on the target cell population of the lens' epithelium. Our studies showed that growth-dependent pathological changes occurred at the cellular level as a function of dose and LET. For a given particle dose, as the LET rose, the number of abnormal mitotic figures, micronuclei frequency, and the disorganization of meridional rows increased to a maximum and then reached a plateau or decreased. For particles of the same LET, the severity of meridional rows disorganization and micronuclei frequency increased with increasing dose. The numbers of cells surviving at late times post-irradiation were comparable with those of controls. In addition, the cellular density was similarly unaffected. These observations are consistent with the current theory of the mechanism of radiation cataractogenesis which posits that genomic damage to the epithelial cells surviving the exposure is responsible for opacification.

A comparison of water equivalent thickness measurements: CT method vs. heavy ion beam technique.

The purpose of this study was to evaluate existing X-ray CT methods for determining water-equivalent path length from body surface to a tumour site. In the method, the CT numbers are used to obtain linear attenuation coefficients which provide a measure of the electron density. These numbers are averaged over the energy spectrum of the diagnostic X-ray beam and other parameters which have a dependence on energy. From range measurements with heavy charged particles, it is also possible to obtain an independent and direct measure of electron density along the beam path. In the results reported here, the beam path electron density or water-equivalent path length was measured with charged particle beams, using radiation sensitive diodes as target markers. To minimise error which would be introduced by motion of the target volume, a frozen dog cadaver was used. Comparison was made between the water-equivalent path length measured with high energy particle beams, and the water-equivalent path length estimated from an X-ray CT image of the same target volume by the methods presently used in charged particle therapy (Chen et al. 1979). There was good agreement between the values determined directly with neon or helium ion beams, but when these values were compared with estimated path lengths derived from X-ray CT data, it was observed that the CT range could be in error by as much as 11% for adverse conditions of marked inhomogeneity and the presence of high atomic number bone. Under the best conditions of moderate inhomogeneity and absence of bone, the derived CT range values agreed reasonably well with the direct measurements.

The RBE for renal damage after irradiation with 3 MeV neutrons.

Mouse kidneys were locally irradiated with single doses or up to 8 fractions of 240 kV X rays or 3 MeV neutrons. Damage was assessed from measurements of urine output, isotope clearance or haematocrit levels. All three assays gave steep dose-response curves by 4-5 months after irradiation. The repair capacity of the kidney was considerable after X-irradiation but was very small after irradiation with neutrons. Thus the RBE increased sharply with increasing fractionation. After large doses, an RBE of 2.3-2.5 was measured, rising to 4.5-5.1 after 8 fractions of 4 to 5 Gy X rays. Linear-quadratic analysis of these data has allowed RBE's to be calculated outside the measured dose range. The limiting RBE predicted at very low doses per fraction is 20 to 26, whereas at extremely high doses it would be as low as 1.2 to 1.4. This indicates that high RBE values may be seen in a slow turnover tissue after low doses per fraction (within the clinically relevant range) although this may not be evident after larger doses. Such high RBE's arise because of the shape of the underlying X-ray dose-response curve rather than the shape of the neutron curve.