Daşu, A. and Denekamp, J. Inducible Repair and Intrinsic Radiosensitivity: A Complex but Predictable Relationship? Two groups have proposed a simple linear relationship between inducible radioresistance in a variety of mammalian cells and their intrinsic radiosensitivity at 2 Gy (Lambin et al. , Int.J. Radiat. Biol. 69, 279–290, 1996; Alsbeih and Raaphorst, unpublished results, 1997). The inducible repair response (IRR) is quantified as a ratio, α S /α R , i.e. the slope in the hypersensitive low-dose region, α S , relative to the α R term of the classical linear-quadratic formula. These proposals imply that the intrinsic radiosensitivity at clinically relevant doses is directly linked to the cell’s ability to mount an adaptive response as a result of exposure to very low doses of radiation. We have re-examined this correlation and found that the more extensive data set now available in the literature does not support the contention of a simple linear relationship. The two parameters are correlated, but by a much more complex relationship. A more logical fit is obtained with a log-linear equation. A series of log-linear curves are needed to describe the correlation between IRR and SF2, because of the spectrum of α/β ratios among the cell lines and hence the confounding effect of the β term at a dose of 2 Gy. The degree of repair competence before irradiation starts could also be a major factor in the apparent magnitude of the amount of repair induced. There appears to be a systematic difference in the data sets from different series of cell lines that have been obtained using flow cytometry techniques in the laboratory in Vancouver and using dynamic microscope imaging at the Gray Laboratory. We suggest that the use of a brief exposure to a laser beam in flow cytometry before the cells are irradiated might itself partially induce a stress response and change the DNA repair capacity of the cells. The clinical consequences of the relationship for predicting the benefits of altered fractionation schedules are discussed.

Buthionine sulfoximine (BSO) has been used to deplete glutathione (GSH) in V79-379A cells in vitro, and the effect on the efficiency of oxygen and misonidazole (MISO) as radiosensitizers has been determined. Treatment with 50 or 500 μM BSO caused a rapid decline in GSH content to less than 5% of control values after 10 hr of exposure (<tex-math>$t_{1/2}=1.6\ {\rm hr}$</tex-math>). Removal of BSO resulted in a rapid regeneration of GSH after 50 μM BSO, but little regeneration was observed over the subsequent 10-hr period after 500 μM. Treatment with either of these two concentrations of BSO for up to 14 hr did not affect cell growth or viability. Cells irradiated in monolayer on glass had an oxygen enhancement ratio (OER) of 3.1. After 10-14 hr pretreatment with 50 μM BSO, washed cells were radiosensitized by GSH depletion at all oxygen tensions tested. The OER was reduced to 2.6, due to greater radiosensitization of hypoxic cells than aerated ones by GSH depletion. GSH depletion had the effect of shifting the enhancement ratio vs <tex-math>$p{\rm O}_{2}$</tex-math> curve to lower oxygen tensions, making oxygen appear more efficient by a factor of approximately 2, based on the <tex-math>$p{\rm O}_{2}$</tex-math> required to give an OER of 2.0. In similar experiments performed with MISO, an enhancement ratio of 2.0 could be achieved with 0.2 mM MISO in anoxic BSO-pretreated cells, compared to 2.7 mM MISO in non-BSO-treated cells. Thus MISO appeared to be more efficient in GSH-depleted cells by a factor of 13.5. These apparent increases in radiosensitizer efficiency in GSH-depleted cells could be explained on the basis of radiosensitization of hypoxic cells by GSH depletion alone (ER = 1.29-1.41). The effect of GSH depletion was approximately equal at all sensitizer concentrations tested, except at high oxygen tensions, where the effect was insignificantly small. These results are consistent with hypoxic cell radiosensitization by GSH depletion and by MISO or oxygen acting by separate mechanisms.

Mouse kidneys have been analyzed at sacrifice, 9 months after single-dose and fractionated irradiation, using wet and dry weight, a biochemical determination of hydroxyproline, and quantitation of dilated renal tubules in histological preparations. Dose-response curves have been constructed to determine the sensitivity and precision of the assays and to study the influence of dose fractionation on a variety of radiation responses of the kidney. There was a marked loss of kidney weight, measured either wet or dry, with maximum changes from control values by factors of 3 and 5, respectively. The wet:dry weight ratio increased with X-ray dose, indicating that relative fluid content was increased even 9 months after irradiation. This could be partly attributed to dilated renal tubules. Total collagen content per kidney, determined by a hydroxyproline assay, showed a less marked dose dependence, with a maximum increase of a factor of 1.4. However, hydroxyproline per dry weight increased by a factor of 7, and this ratio proved to be the most sensitive and precise measure of radiation damage. The "fibrosis" that is detected in histological sections appears to be more a relative than an absolute alteration in connective tissue. The loss of parenchymal cell mass, particularly in the proximal tubules, is the predominant factor; the increase in the absolute amount of collagen per kidney contributes to a lesser degree. The influence of radiation dose fractionation was analyzed using a linear-quadratic response model. The α/β ratios were between 0.9 and 2.9 Gy.

The response of mouse blood vessels to radiation has been studied by locally irradiating mesenteric vessels, in an exteriorized segment, using a strontium-90 β applicator. Four different assays of damage were made when the animals were sacrificed from 6 weeks to 18 months after irradiation. In all studies sham-irradiated animals were compared with those receiving 20, 30, and 45 Gy. Vein diameter was studied by direct observation, vascular volume and permeability by isotope techniques, and the cellularity of the arterioles from histological sections. Within 3 months changes were seen in all these parameters, with a gradual recovery toward normal by 12 months, and a second phase of damage appearing at 18 months. The time course of changes in vascular permeability appeared to match the loss of endothelial cells lining the arterioles, but there was a significant correlation only at 6 weeks. Fibrosis was not seen before 12 months, but at 18 months the vessel walls were fibrosed, sclerotic, and occasionally the lumen was occluded. At no time was any change in the dependent ileum seen. The blood supply to this tissue was maintained in spite of the severe radiation damage in the vessels.

The radiation response of the mouse urinary bladder was assessed by measuring the urination frequency after localized irradiation with 1.8-MeV electrons. The onset of radiation-induced incontinence is late, no changes were observed before 5 months after doses of up to 40 Gy (4000 rad). Beyond this time good dose-response curves were obtained above a threshold dose of 15 Gy. The time of onset of radiation damage was earlier after higher doses, but in all cases the maximum damage was observed by 12 months. There was some recovery of bladder function, especially in the lower dose groups, by 16 months. The increased frequency at 5-12 months was associated with epithelial denudation and a loss of the specialized polyploid surface cells. Fibrosis of the deeper layers was not apparent until after 12 months. The increased urination frequency responses are correlated with the delayed compensatory proliferation of the epithelium and with the resistance of the bladder to inflation under applied pressure, which also gave a dose response relationship at 18 months.

The radiosensitizing effect of misonidazole (Ro 07-0582) is presented for four types of mouse tumor. The interval between administration of the drug and irradiation has been found to have a large effect on the extent to which hypoxic tumor cells are sensitized. The optimal time for irradiation differed in the different tumors and ranged from 15 to 120 min. This delayed effectiveness probably relates to the time required for diffusion of the drug to the critical hypoxic regions and for its concentration in those regions to reach a maximum. Drug concentrations have been measured in whole blood and in the tumors using gas-liquid chromatography. The tumor concentrations ranged from 20 to 70% of the blood level in the four tumors; in man it was 40-107%. The in vivo radiosensitizing effect of the drug at these measured concentrations in tumors is similar to that achieved in vitro. The apparent loss in radiosensitizer effectiveness in mouse tumors in vivo relative to in vitro results is thus due to the lower concentration in tumors than in serum, because of the short metabolic half-life in mice.

Two electron-affinic nitroimidazoles have been investigated as potential radiosensitizers using tumor regrowth as a measure of response. Both drugs, metronidazole, (May and Baker), and Ro-07-0582, (Roche), were effective radiosensitizers of hypoxic tumor cells. These sensitizers have previously been shown to be specific in that they are effective only on hypoxic cells and give no sensitization of well-oxygenated cells. Flagyl (metronidazole) gave an enhancement ratio of about 1.6 at 0.75 mg/g body weight. Ro-07-0582 sensitized both naturally occurring hypoxic cells in air-breathing mice, and acutely hypoxic cells in clamped tumors. It had a small effect when administered after irradiation with 2000 rads, and caused very slight slowing of tumor growth in unirradiated mice. Ro-07-0582 is the most effective sensitizer so far tested, giving an enhancement ratio of up to 2.1 when 1 mg/g is administered intraperitoneally. Since no enhancement of well-oxygenated normal tissue has been observed, this could correspond to a therapeutic gain factor of 2.1.