Search

It is generally accepted that hypoxia is a common occurrence in many experimental and human tumours and that it is a major cause of local failure after radiotherapy. Many attempts have been made over the last years to eliminate this problem. One of the manoeuvres to improve tumour oxygenation is to manipulate the binding affinity of oxygen (O2) and haemoglobin (Hb). Previous studies have shown that some antilipidaemic drugs (clofibrate and its analogues) can reduce the Hb/O2 binding affinity and sensitize various animal tumours to radiation. The present study evaluated the ability of clofibrate to sensitize in situ a mouse carcinoma (CaNT) to radiation. Clofibrate at 1.5 mmol/kg increased the tumour radiosensitivity, when it was administered per os 2-6 h before a single radiation dose or 2 to 4 h before each of 10 fractions in 5 days. In both the single dose and fractionated regimens, the radiosensitizing effect was drug dose-dependent, but was only statistically significant at doses from 1.0 to 2.0 mmol/kg. These results suggest that clofibrate may be an effective radiosensitizer at radiation doses that are clinically relevant. Further experiments need to be carried out to evaluate clofibrate analogues for their radiosensitizing properties. Clofibrate tolerable doses in man have to be determined first in order to know if clofibrate and analogues could play a role in the clinical management of tumours where hypoxia may limit the outcome of radiotherapy.

It has been recognised for many years that anaemic cancer patients have a particularly poor prognosis (see recent reviews [4,8]). New data is regularly appearing in the literature, extending our knowledge to include many tumor sites. The evidence is now overwhelming that for most of these, local tumor control by radiotherapy is compromised in patients who are anaemic before and during radiotherapy. The role of the radiobiologist must be to offer an explanation for the clinical observations and to suggest a means of compensating for the problem in those patients whose treatment is prejudiced by anaemia. This statement assumes a cause and effect relationship between anaemia and tumor curability, supposedly through an impairment of oxygen transport to the tumor cells. We must then consider the consequences of a reduction in circulating haemoglobin levels in model tumor systems in animals, combine that information with our knowledge of physiological mechanisms and attempt to reconcile our conclusions with the clinical findings. We should be aware of course, that a failure to achieve this could be the result of inadequacies of the mouse model or because anaemia is simply not the cause of clinical radioresistance but rather a consequence, along with radiosensitivity of particular tumor characteristics. What do the experimental animal data reveal? We will consider two clinically important questions: 1) Are tumors in mice with lower than normal haemoglobin levels more resistant to radiation and what is the temporal relationship between duration of anaemia and sensitivity? 2) Does the restoration of haemoglobin levels before radiotherapy change sensitivity?

Enhancement of CCNU cytotoxicity by misonidazole (MISO) was studied in three tumours and two normal tissues in the mouse. The 3 experimental tumours (SCCVII/St, EMT6 and KHT) showed very different sensitivities to CCNU alone, but MISO enhanced the cell killing in ech case. The effect was not always dose-modifying, so that the CCNU dose range for the greatest enhancement was different in each of the tumours. In all 3 tumours, enhancement increased with dose of MISO. The effect on two normal tissues, marrow (CFU-S) and testis (spermatogonia), was also investigated. Enhancement of marrow toxicity could be demonstrated only at CCNU doses greater than 12.5 mg/kg, so that at lower CCNU doses there was a therapeutic gain equal to the tumour enhancement ratio. The spermatogonia effect, however, showed enhancement by MISO similar to that seen in the tumours at all CCNU doses up to 20 mg/kg.

The response of 14 mouse tumour sub-lines to the radiosensitizing action of a large single dose of misonidazole (MISO) has been assessed by regrowth delay. In 13 of these, significant enhancement of radiation effect occurred under ambient conditions, indicating sensitization of naturally hypoxic cells. The enhancement observed (SER') varied with the radiation dose, as would be predicted for a mixed oxic/hypoxic cell population. The maximum SER' in these 13 tumours did not depend on histology or regrowth rate. The 14th tumour, a slow-growing sarcoma, was not sensitized under ambient conditions, but showed marked sensitization when clamped to produce acutely hypoxic cells. This is consistent with no hypoxic cells occurring naturally in a sarcoma with a slow rate of growth. Faster-growing variants of this tumour showed radiosensitization under ambient conditions. The slow-growing carcinoma, RH, however, appears to contain hypoxic cells and did show sensitization. The cytotoxic action of MISO was compared with the radiosensitization by administering it after irradiation in 8 of the tumour lines. In 2 tumours no cytotoxicity was observed. In the rest cytotoxicity was significant, but much smaller than the sensitization observed when MISO was administered before irradiation. These regrowth-delay data have been used to calculate hypoxic fractions in 3 ways. Estimates of hypoxic fraction ranged from less than 0.1% in the slow sarcoma to greater than or equal to 30% in several tumours. There is considerable variation in the estimate, according to the technique used.

The ability of three nitroimidazoles [SR-2508, misonidazole (MISO), and benznidazole] with differing octanol-water partition coefficients to enhance the cytotoxicity of the nitrosourea 1-(2-chloroethyl)-3-cyclohexyl-1-nitrosourea (CCNU) was evaluated in two mouse tumors (the KHT sarcoma and the SCC VII/St carcinoma). These results were compared with the effect on two normal tissues (bone marrow CFU-S and testis spermatogonia). When given as a large single dose, benznidazole was more effective than MISO in enhancing the cytotoxicity of CCNU to both tumors. SR-2508 had no effect. The advantage of benznidazole over MISO was lost, however, because benznidazole gave more toxicity in the normal tissues than MISO. In experiments where the nitroimidazoles were administered by multiple small injections to maintain a blood plasma level between 50 and 100 micrograms/ml, benznidazole was also more effective than MISO in enhancing CCNU cytotoxicity in the tumors. In each case, enhancement was rather less than that obtained with large single injections. Again, however, benznidazole did not produce a consistently greater therapeutic gain than MISO because it also enhanced normal tissue toxicity while MISO did not. SR-2508 was ineffective in both tumors and normal tissues. We conclude that neither SR-2508 nor benznidazole are superior to MISO in combination with CCNU.

Hypoxic cells in human tumours probably contribute to the failure of radiotherapy in some sites. Changes in the oxygen carrying capacity of the blood, such as in anaemia, have been shown to influence tumour response. The effect of acute and chronic changes in haematocrit on the radiosensitivity of three mouse tumours (EMT6, KHT and RIF-1) were studied. Alterations in haematocrit were achieved by bleeding followed by retransfusion. When radiation was preceded immediately by an acute reduction in haematocrit (anaemia), radiosensitivity was markedly reduced in each tumour. An acute rise in haematocrit (polycythaemia) increased or decreased X-ray sensitivity depending on its severity. The optimum haematocrit for maximum sensitivity was always found to be at a level 5-10 per cent above normal. When the time between induction of anaemia and irradiation was increased, simulating a progressively longer duration of anaemia, marked changes in radiosensitivity of all the tumours were observed. A short duration of anaemia resulted in a resistant tumour with each cell line, but the resistance was gradually lost as the anaemia was prolonged, even though no recovery in haematocrit occurred. The rate of recovery to normal radiosensitivity varied from 24 to 72 hours in the different tumours. Therefore, only haematocrit changes which occurred within 1-3 days of a dose of radiation affect the radiosensitivity of these tumours.

The calcium antagonists, verapamil, nifedipine, and flunarizine, were studied for their effects on blood flow and radiation response in SCCVII/St intradermal back tumors, over a dose range of 0.05-50 mg/kg. Verapamil, at low doses, increased tumor blood flow, as measured by relative fluorescence intensity of Ho33342 stain, and increased tumor radiosensitivity. However, at doses of 20 mg/kg and above, verapamil reduced Ho33342 fluorescence intensity, and increased tumor radioresistance. Nifedipine reduced tumor radiosensitivity and Ho33342 fluroscence intensity at doses above 1 mg/kg, but below this dose increased Ho33342 intensity was observed and a small radiosensitization was apparent. Flunarizine sensitized tumors to X rays at all doses tested, although increased Ho33342 intensity was seen only at 5 mg/kg. The relative affinities of these compounds for different sites within the host may explain the variations in blood flow and radiation sensitivity in this tumor system.

1. A study has been made of the tension responses and sarcomere length changes produced by servo-controlled stretches applied to isometrically contracting frog muscle. Sarcomere lengths were monitored by cine-photography of diffiraction spectra obtained by illuminating a small area of muscle with a laser. 2. The tension increment produced by a ramp-and-hold stretch of approximately 1 mm (ca. 4% of the muscle length) comprises three phases whose limits are defined by two points, S1 and S2, where the slope of the response decreases abruptly. S1 and S2 correspond to extensions of 0.13 and 1.2% of the muscle length. 3. Movements of the first order spectra relative to the zero order recorded during stretch reveal that S2 coincides with an abrupt elongation of the sarcomeres. This is termed sarcomere 'give' and it occurs when the filaments are displaced by 11-12 nm from their steady-state (isometric) position. 4. The stiffness of the sarcomeres, Es, up to S2 decreases with increasing sarcomere length. The maximum force sustained by the muscle at S2, PS2, also shows an inverse dependence on sarcomere length. Both Es and PS2 fall to zero at an extrapolated sarcomere spacing of 3.6-3.7 micrometer, coinciding with the length at which the actin and myosin filaments no longer overlap. 5. The ratio PS2/P0 (where P0 = maximum isometric tension) varies with temperature and speed of stretch. It increases with increasing speeds of stretch until a certain critical velocity, Vc, is reached, beyond which it is almost independent of any further increase. Vc has a positive temperature coefficient, increasing 5-6 in the range 0-30 degrees C (Q10 = 1.8). There is a positive correlation between the maximum speed of isotonic shortening (Vmax.) and Vc in different muscles. 6. Sarcomere 'give' during stretch is considered to be due to forcible detachment of cross-bridges between the actin and myosin filaments. This results in recoil of the extended series elastic elements in the muscle at the expense of the sarcomers. The amount of filament displacement required to induce sarcomere 'give' (11-12 nm) is thought to represent the range of movement over which a cross-bridge can remain attached to actin during a stretch.

Binding of misonidazole (MISO) or a derivative to hypoxic cells in tumors has been proposed as a method for identifying tumors and measuring their level of hypoxia. We have recently shown that the hypoxic fraction of tumor cells can be altered over a wide range in vivo by acutely changing the hematocrit of the host animal by transfusion. The present study aimed to investigate the changes in binding by 14C MISO that accompanied this procedure. Tumor bearing mice were injected with 14C MISO, irradiated with a single dose of X rays (20 Gy) and their tumor excised and bisected. One half of each tumor was used to determine cell survival in vitro, the other was used for 14C scintillation counting. As previously described, tumor cell survival was dramatically increased in acutely anemic mice and this was accompanied by an increase in 14C MISO binding to the tumors. The relationship between clonogenic cell survival and binding was found to be linear on a log-log plot for each of the tumor lines studied, but the slopes of the lines were different tumor lines and generally steeper than the value of 1.0 expected for a 1:1 correspondence between cells binding radioactivity and radiobiological resistance. We attribute these differences to MISO binding to cells in the tumor which were not clonogenic.

1. The effects of single and double cycles of stretch and release on the tension response and relative sliding movement of the actin and myosin filaments in active frog's muscle were investigated. 2. The cross-bridges linking the filaments together are able to accommodate a greater range of filament displacement before becoming detached during a second cycle stretch, providing it commences without delay following the preceding release: sarcomere 'give' then occurs for displacements of around 18 nm, as compared with 12 nm for a first cycle stretch. It is postulated that the difference arises because the myosin heads adopt different 'preferred' positions in the isometric steady-state and at the end of a previous release. 3. Muscle length-tension loops were recorded and used to measure the energy absorbed when a muscle is subjected to cycles of stretch and release. The work absorbed per unit length change increases with increasing displacement of the cross-bridges from their initial (isometric) steady-state position, up to the point at which sarcomere 'give' occurs (S2); thereafter it remains constant. 4. More work is absorbed during the first cycle of a double stretch-release combination than during the second. The greater amount absorbed during the first cycle is associated with a correspondingly greater amount of filament sliding in the period following sarcomere 'give'. Sarcomere length-tension loops were constructed and these showed that not less than 80-85% of the work done on a muscle is absorbed by the sarcomeres themselves. 5. A greater amount of work is done on stretching up to (but not beyond) S2 during second cycle stretch as compared to a first. The difference amounts about 1 mJ.m-2 per half-sarcomere. 6. The results are compatible with the mechanism for force production proposed by Huxley & Simmons (1973), in which each myosin head generates force in a number of stepping movements, from one attached state to another. It is concluded that (a) during an unloaded isotonic contraction the working 'stroke' of the head would result in a 10-13 nm relative sliding movement of the filaments, and (b) the potential energy difference separating the two 'preferred' states is 6-9.6 kT per cross-bridge, or 3-4.8 kT per S-1 sub-units, assuming that each one interacts simultaneously with the actin filament.

We have investigated the ability of the antilipidaemia drugs clofibrate and bezafibrate, to reduce the binding affinity of haemoglobin for oxygen and to sensitize an experimental tumour (the SCCVII/St carcinoma) in the mouse to radiation. Clofibrate at a high dose (4.1 mmol/kg, p.o.) increased the P50 of the blood by about 10 mm Hg. Its effect on tumour radiosensitivity was dependent on tumour size. Highly significant sensitization, equivalent to a 40-fold reduction in the number of hypoxic cells, was seen in small tumours; but in large tumours there was much less effect. At a low dose, which is close to that currently used clinically (0.3 mmol/kg), clofibrate produced a small and barely significant increase in P50. The effect of low dose clofibrate on tumour radiosensitivity also depended on tumour size, small tumours (200 mg) being significantly sensitized, while no significant effect was seen in large tumours. Bezafibrate, at the low dose of 0.3 mmol/kg, gave a significant increase in P50 (by approximately equal to 8 mm Hg), but sensitization to radiation in small tumours was not impressive and not statistically significant. We must gain a better understanding of the mechanisms involved in these effects before applying this approach to clinical radiotherapy.

An autoradiographic study of three corded mouse tumours is reported. The proliferation characteristics of both tumour cells and endothelial cells were studied. The doubling time of these three tumours differed by a factor of 2.6 but there was only a small difference in the intermitotic time. All three tumours showed a very high cell loss factor (approximately 0.80) and the differences in growth rate resulted mainly from differences in the growth fraction. The endothelial cell proliferation rates differed markedly in the three tumours, with labelling indices ranging from 18% in the faster tumours to 4.5% in the slowest. The potential doubling times for endothelium, calculated from these values, were much slower than the tumour cell cycle time or the tumour potential doubling time, but were two to four times faster than the volume doubling time of the tumour. It appears likely that the endothelial proliferation rate influences the growth fraction, but similar high cell loss factors can occur in tumours with a four-fold difference in endothelial cell production rates. Inadequate branching of blood vessels seems likely to be at least as important as inadequate production of endothelial cells. It is not possible to determine whether slow tumour cell production evokes a slower endothelial growth or vice versa.

Sensitization of hypoxic cells in tumors, by increasing their oxygen supply, has been attempted for at least 30 years. Only the use of hyperbaric oxygen has been shown unequivocally as a beneficial adjunct to radiotherapy; and even then, the number of sites sensitized is limited to head and neck and cervix. It is not clear whether this implies that all other tumors reoxygenate fully during treatment, or whether a better method would sensitize other sites. Nevertheless, the elimination of hypoxic cells is viewed by many as a worthy goal in radiobiology and many strategies have been tested in animal systems. These include: oxygen releasing chemicals, artificial oxygen carriers, inhibitors of oxygen consumption, blood flow modifiers, or the exploitation of tumor adaptation to altered oxygen availability. We must be aware that any procedure which improves tumor oxygenation will not only increase radiosensitivity, but will induce an adaptive response in the tumor such that, sensitization will be of limited duration. It is likely that in the apparent failure of measures to improve substantially the oxygen delivery to tumors, the elimination of most of the hypoxic cells, of the type accessible to them, may have been achieved. If, as has been suggested, there are two distinct types of hypoxic cells, a combination of more than one strategy may be necessary to achieve more substantial gains.

The nitroimidazole SR-2508 is currently being tested clinically as a radiosensitizer. Its relatively low toxicity allows it to be used at higher doses than misonidazole so that its potential as a chemosensitizer is also of considerable interest. Multiple injections of SR-2508 were given to SCC VII/St tumor-bearing mice to achieve a clinically realistic plasma concentration of approximately 300 micrograms/ml over 8 hrs. Single doses of melphalan (L-PAM) or cyclophosphamide (CY) were given at different times after the first SR 2508 injection. With L-PAM, a delay of at least 2 hr was necessary before enhancement of L-PAM cytotoxicity was observed. A similar result was obtained when a simulation was carried out with SCC VII/St tumor cells in vitro. Results with CY were less clear, although the most consistent enhancement was observed when a 4 to 8 hr interval elapsed between the beginning of SR 2508 exposure and the CY injection. In general, although precise timing was not essential for enhancement, an interval of at least 4 hr is recommended between the administration of SR 2508 and either alkylating agent. This is particularly important for L-PAM where no enhancement would be expected if the drugs were given simultaneously.

In these preliminary experiments, we have found enhanced cell killing by the bifunctional alkylating agent L-phenylalanine mustard (L-PAM) in the presence of inhibitors of poly (ADP-ribose) polymerase (ADPRP) in vitro. In vivo enhancement of the tumoricidal effects of L-PAM was observed with the ADPRP inhibitor nicotinamide (1000 mg/kg), although enhanced myelosuppression was also demonstrated. Nicotinamide also increased the plasma elimination half-life of L-PAM by a factor of at least 2. This alteration of L-PAm pharmacokinetics makes it difficult to assess the role that ADPRP inhibition plays in the enhancement of L-PAM tumor cell killing in vivo.

Experiments were carried out to determine whether the enhancement of alkylating-agent cytotoxicity seen after large single doses of misonidazole (MISO) in mouse tumours can also be achieved by prolonged exposure to low MISO levels similar to those which can be tolerated clinically. The level in mouse blood plasma could be maintained at about 100 micrograms/ml for 7 h by injecting small doses of MISO every 1/2 h. The effect of this treatment in combination with cyclophosphamide (CY) or melphalan (L-PAM) was studied in the RIF-1 tumour, using regrowth delay and cell-survival cloning assays. In each case, prolonged exposure to low levels of MISO gave enhancement ratios very close to those obtained with a large single dose. ERs of 1.6-2.0 were obtained with CY and 1.8-2.2 with L-PAM over the range of alkylating-agent doses used. In experiments with CY the response of 2 normal-tissue systems, marrow and WBC count, was also studied. No significant enhancement of CY damage occurred in either case. In the L-PAM experiments the LD50/30 and WBC counts were determined as normal-tissue end points. Multiple MISO had no effect. Our results show that levels of MISO which can be achieved safely in man yield good enhancement of the tumour cytotoxicity of 2 widely used chemotherapeutic agents without increasing the damage to normal tissues.

The proliferation response of the mouse bladder was investigated, using continuous labelling with tritiated thymidine, at various times after a single dose of radiation. Bladder epithelial and vascular endothelial cells were studied. The cell turnover rate in unirradiated epithelium and endothelium was found to be extremely slow (in excess of 1 year). Irradiation with a single dose of 25 Gy resulted in compensatory proliferation of the epithelium but the response was not initiated for many months. At 3 months after irradiation there was little difference from the control proliferation rate, but from 6 to 22 months after irradiation (the end of the study) there was a period of sustained rapid proliferation with the cell turnover time reduced to approximately 1 week. The increase in proliferative activity observed at 22 months was found to be dose-dependent. Endothelial cells in the blood vessels of the submucosa also showed an increased turnover rate after irradiation and the timing this response was found to be similar to that of the epithelium. The onset of compensatory proliferation in both cell types was found to coincide with marked histological and functional changes in the bladder. In this slowly proliferating tissue, the onset of rapid compensatory proliferation after irradiation is delayed and occurs at the time that functional impairment is observed. This supports the postulate that proliferation is unlikely to contribute much to the sparing effect of prolonged fractionated radiotherapy in slowly dividing tissues.

The effect of several purine nucleoside analogs on the cytotoxicity of melphalan (L-PAM) in the RIF-1 tumor in vivo was investigated. All the analogs tested--3'-deoxyguanosine (3'-dG), 3'-deoxyadenosine (3'-dA), and N6-butyryl-3'-deoxyadenosine (N6-BC)--at a dose of 100 mg/kg, enhanced the tumor cell killing by L-PAM as measured by an in vivo/in vitro procedure. This enhancement was maximal when the drugs were given simultaneously. The mean enhancement ratios (ER's) determined from the L-PAM dose response curves were 1.4 for 3'-dG, 1.3 for 3'-dA, and 1.4 for N6-BC. A similar modification of the L-PAM-induced depression of white blood cell counts was not obtained. A large single dose of L-PAM (8 mg/kg) produced a transient drop in mouse body temperature. The analog 3'-dG (100 mg/kg) increased and prolonged this hypothermic effect. In addition 3'-dG also delayed the clearance of L-PAM from the plasma of C3H mice, such that the half-life of the chemotherapeutic agent was extended from 28 minutes to 35 minutes. The enhancement of the efficacy of L-PAM by these analogs probably results from this pharmacokinetic effect.

We have developed exchange transfusion methods to alter the hematocrit of tumour-bearing mice. The effects of anaemia and its correction by blood transfusion on the radiosensitivity of two mouse tumours (SCCVII/St and RIF-1) were studied using excision, in vivo/in vitro assay. An acute reduction in haematocrit caused a high degree of radioresistance equivalent to an increase in the hypoxic fractions by factors of 10 (SCCVII/St) and 30 (RIF-1). As the duration of the anaemia was prolonged, radioresistance was lost until within about 6 h normal radiosensitivity was observed even though the anaemia persisted. The restoration of the normal haematocrit by red blood cell transfusion after 24 h of anaemia caused increased radiosensitivity equivalent to a reduction in the hypoxic fraction by factors of 5 (SCCVII/St) and 10 (RIF-1), but again the effect was transient and normal radiosensitivity was re-established within 24-48 h of retransfusion. Measurements of 14C misonidazole (MISO) binding to RIF-1 tumours after these procedures indicated changes in the number of hypoxic cells which were qualitatively almost identical to those using the cell survival endpoint, leading us to believe that changes in oxygenation were responsible for the altered radiosensitivity. We feel that transfusion procedures could be used to advantage in the radiotherapy of some cancers.

The effect of cisplatin on the cytotoxicity of alkylating agents in the RIF-1 tumor both in vivo and in vitro was investigated. A single dose of cisplatin (1 mg/kg) enhanced the in vivo tumor killing by melphalan (L-PAM; 8 mg/kg). The effect was maximal when cisplatin was given between 3 hours before and 1 hour after injecting L-PAM; the enhancement was lost as the time interval increased. Similar results were obtained with cyclophosphamide (CYT; 75 mg/kg) and CCNU (50 mg/kg) although the enhancement was much less than that seen with L-PAM. The enhancement by cisplatin was constant at all L-PAM doses as measured by both cloning assay and regrowth delay. Measurements of white blood cell counts four days after drug administration suggests that a therapeutic gain can be achieved. Exponential and plateau phase RIF-1 cells grown in monolayer culture were also exposed to various L-PAM doses. A 1 hour pre-exposure to cisplatin (0.5 micrograms/ml) under aerobic conditions increased the cell killing by L-PAM. The results are discussed with reference to the possible reasons for the therapeutic gain.

The proliferation pattern of a transplantable mouse mammary carcinoma has been studied in relation to its macroscopic and microscopic structure. No significant differences were seen in the labelling or mitotic indices or in the percentage labelled mitoses curves for the peripheral 2.0 mm rim or for the central tumour core. When these parameters were scored for cells classified according to their position in relation to capillaries or to necrotic regions, marked differences were observed in all the parameters. Higher labelling and mitotic indices and higher grain counts were seen adjacent to the capillaries. These appear to result from a shorter cell cycle duration and a higher growth fraction. The variation in cell cycle is mainly due to a change in the duration of G1.

The effect of misonidazole (MISO) on the cytotoxicity of cyclophosphamide (CY) was investigated in the mouse. The response of the RIF-1 tumour was measured by growth delay and by cell survival in a cloning assay. MISO enhanced the cytotoxicity of CY. For single treatment, enhancement was maximal when MISO was given 30 min to 2 h before CY. The enhancement ratio (i.e. the dose of CY alone divided by the dose of CY with MISO required to cause the same response) increased with increasing dose of MISO up to 250 mg/kg, but decreased with increasing dose of CY above 50 mg/kg. For 5 daily treatments, enhancement increased with CY dose up to approximately 25 mg/kg/injection. Survival of marrow stem cells was measured using the spleen-colony assay. MISO did not enhance significantly the cytotoxicity of CY at doses under 100 mg/kg. Enhancement was seen at higher doses, but the effect was less than in tumours. CY reduced the number of circulating white blood cells. Neutrophils were most severely depleted. The WBC count was slightly lower when CY was given in combination with MISO than after CY alone, but the effect could be accounted for by direct MISO cytotoxicity. These experiences suggest that a therapeutic gain may be achieved if MISO is combined with doses of CY in the clinical range. From experiments performed to investigated the possible mechanisms involved, we conclude that for the RIF-1 tumour the major effect of MISO is to inhibit the repair from CY-induced potentially lethal damage.

It has been recognized for some time that alterations in the affinity of Hb for oxygen could offer a means of improving oxygen delivery to tumors and achieving radiosensitization. Three antilipidemic drugs, clofibrate, bezafibrate, and gemfibrozil, two of which had previously been shown to reduce Hb/O2 affinity in vitro, were tested in mice for their ability to affect Hb/O2 affinity and to alter the radiosensitivity of the RIF-1 sarcoma. Each of the drugs produced a significant increase in the P50 of the blood, from a mean control value of 45 mm Hg to 55, 74, and 51 mm Hg after a dose of 1 g/kg of clofibrate, bezafibrate, and gemfibrozil, respectively. However, they had very different effects on the radiosensitivity of the RIF-1 tumor. When the mice breathed air at the time of irradiation, clofibrate produced a marked sensitization equivalent at the optimum time to a 20-fold reduction in hypoxic fraction; bezafibrate gave a lower sensitization equivalent to a 4-fold reduction, while gemfibrozil caused dramatic radioresistance equivalent to a 10-fold increase in hypoxic fraction. When the mice were given 95% O2/5% CO2 to breathe at the time of irradiation to ensure complete Hb saturation in the lungs, a large increase in the sensitization by bezafibrate was seen, but there was only a small change with clofibrate. We conclude that drugs which reduce Hb/O2 affinity could have a role in sensitizing tumors to radiation.

The effect of inhibitors of nuclear ADP-ribosyl transferase (ADPRT) on the cytotoxicity of melphalan (L-PAM) in the RIF-1 tumour in vivo was investigated. A large single dose of nicotinamide (1000 mg kg-1) enhanced the tumour cell killing by L-PAM as measured by tumour cell survival. This enhancement was maximum when nicotinamide was administered within 1 h before injecting the L-PAM. When given at this time, the nicotinamide had a dose-modifying effect on all L-PAM doses tested, giving rise to a mean enhancement ratio (ER) of 2.2. Nicotinamide did not appear to inhibit the recovery from L-PAM induced potentially lethal damage. L-PAM (6 mg kg-1) produced a transient drop in mouse body temperature. This effect was both increased and prolonged by nicotinamide. In addition the inhibitor also delayed the clearance of L-PAM from the plasma of C3H mice, such that the half-life of the chemotherapeutic agent was extended from 41 min to 143 min. The effect of combining L-PAM with nicotinamide doses below 1000 mg kg-1 was also investigated. The results showed that as the nicotinamide dose was decreased, the enhancement of the effects on body temperature, pharmacokinetics and white blood cell counts were reduced. However, a concomitant loss in the enhancement of tumour cell killing was also observed. Similar results were obtained using 3-aminobenzamide, a more efficient inhibitor of ADPRT.

The ability of a range of 2-nitroimidazoles with similar electron affinities but widely differing partition coefficients (P) to enhance the cytotoxicity of cyclophosphamide (CY) in mouse tumour and normal tissues was investigated. In a preliminary study large single doses of benznidazole (BENZ), misonidazole (MISO), desmethylmisonidazole (DMM), and SR-2508 were found to give similar enhancement of the RIF-1 and SCC VII/St tumours. SR-2555 was less effective. A direct comparison was made between MISO and SR-2508 using prolonged, low-level drug exposures, achieved by multiple injections. The enhancement of CY cytotoxicity achieved in the two tumour systems (RIF-1 and SCC VII/St) was found to be similar for a given blood sensitizer concentration. In the normal tissue assays (white blood cell count, bone marrow CFU-S and testis spermatogonia) neither MISO nor SR-2508 produced significant enhancement of CY cytotoxicity, so that the therapeutic gains achieved at a given blood concentration of sensitizer were similar for SR-2508 and MISO. The main advantage of SR-2508, however, will probably lie in its lower toxicity, permitting higher blood levels to be achieved. However, the slope of the dose response curves are rather shallow so we would not predict a dramatically increased benefit.

A continuous labelling technique was employed to study the effects of external beta-radiation on the proliferation of endothelial cells and smooth muscle cells in the mesenteric arterioles of mice. Labelled and non-labelled cells of either type were determined by autoradiographic techniques in control animals and at different times (3, 12 and 48 weeks) after single doses of 20 and 45 Gy (2000 and 4500 rads). The fraction of cells labelled, even after 7 days of repeated injections was very low in all instances. Calculations showed very long turnover times for the two cell populations in control animals (greater than 2 years for endothelium and greater than 3 years for smooth muscle). After 20 and 45 Gy, no significant increase in endothelial proliferation was seen except at 3 weeks. No significant increase in labelling was observed in smooth muscle at any time after irradiation. These labelling data have been compared with the pattern of cell depletion of the irradiated endothelium. It was concluded that the depletion was much earlier than expected for a slowly proliferating tissue, if all the cells were cycling very slowly. Such an early depletion is, however, consistent with cell death resulting from a small proportion of the cells having a short cell cycle. The recovery of the endothelial cell numbers between 9 and 12 months was not accompanied by a rise in the fraction of labelled cells. Its is suggested that repopulation may occur from outside the treated area.

The effect of prolonged exposure to low misonidazole (MISO) levels on the cytotoxicity of three alkylating agents was studied in mouse tumors. A concentration of 100 micrograms/ml was maintained in the plasma for 7 hr by multiple injections of MISO. Cyclophosphamide (CYC). Melphalan (L-PAM), or CCNU were given after 4 hrs of MISO exposure. In each case, prolonged low level MISO exposure enhanced tumor response as measured by regrowth delay or a cloning assay. The effect of this treatment was also studied in several normal tissues: bone marrow, white blood cell counts, and spermatogonia. In none of these was any enhancement seen after prolonged MISO exposure. These encouraging results show that clinically relevant exposures to MISO can greatly improve tumor response to alkylating agents without increased normal tissue toxicity.

Anemia may often become a problem in the treatment of the cancer patient. There are insufficient clinical data to assess the overall importance of anemia in radiotherapy, but there is clear evidence that uncorrected anemia is detrimental to local tumor control in some sites. There may be situations, however, when the transfused, previously anemic patient is at an advantage. These patients have shown a dramatically better response than non-anemic patients when radiotherapy for cancer of the cervix was given in hyperbaric oxygen. Animal experiments suggest that adaptive processes may be responsible for this effect. There is an important difference between acute and chronic anemia in their influence on the radiosensitivity of mouse tumors; while acute anemia consistently causes radioresistance, this effect is lost as the duration of the anemia prior to irradiation is prolonged. This would suggest that anemia per se should not cause tumor radioresistance in the chronically anemic patient. Blood transfusion in previously anemic animals has been shown to produce a markedly increased tumor radiosensitivity, but again this is only transient and sensitivity returns to normal when the interval between transfusion and irradiation is extended to 24 hrs. The mechanisms responsible for tumor adaptation to anemia and blood transfusion are not known, but there is evidence that changes in diffusion distances occur within tumors in response to alterations in oxygen availability and that changes in blood chemistry through the 2,3-DPG system may alter the release of oxygen to the tissues. These are complex processes and it remains to be determined what influence they have in the treatment of human cancer. However, the animal data suggest a clear benefit of blood transfusion to restore the hemoglobin level in radiotherapy, but they also emphasize the need to irradiate immediately so that adaptive mechanisms cannot erode the effect.

The clinical use of misonidazole as a hypoxic cell radiosensitizer is at present limited by its neurotoxicity at high doses (Urtasun et al., 1977; Dische et al., 1977). An in vivo neurological end point, viz. measurement of nerve conduction velocity, has been developed to examine sensitizer action. Conduction velocity in mice was measured as a function of time after a single dose of misonidazole and as a function of drug dose. Doses greater than 0.33 mg/g produced significant transient reductions in velocity. The time course of the reduction in velocity closely followed the uptake/excretion profile of misonidazole from blood serum.

Dose responses for the effects of four calcium antagonists on SCCVII/St back and leg tumour sensitivity to 20 Gy X-rays, and on tumour perfusion indicated by relative uptake of 86rubidium, were determined. The effects of these agents on 86rubidium uptake in some normal tissues were also investigated. Flunarizine sensitized tumours to X-rays over the dose range 0.005-500 mg/kg given i.p., but only small increases in tumour perfusion were seen at doses of 0.05-5 mg/kg. Verapamil increased tumour radioresistance at doses of 20 mg/kg and above, and reduced tumour perfusion at 50 mg/kg. Below 10 mg/kg, verapamil sensitized the tumours to X-rays, but produced little or no increase in tumour perfusion. Nifedipine at 10 mg/kg and above produced very radioresistant tumours, with correspondingly large reductions in tumour perfusion. At doses below 0.5 mg/kg sensitization was seen, but no increased tumour perfusion. Diltiazem at 50 mg/kg also increased tumour radioresistance with a reduction in tumour perfusion, and at lower doses sensitized tumours to X-rays, with small increases in tumour perfusion. The effects of these agents on normal tissues were varied and difficult to interpret. The similarity between the tumour radiation responses to the four calcium antagonists suggests that they may be acting upon the tumour vasculature in a manner independent of the systemic circulation.

The effect of the nitroimidazoles misonidazole, Ro-05-9963, RGW-608 and metronidazole on nerve conduction velocity (NCV) were measured in the anaesthetized mouse. The compounds were administered by i.p. injection either as a single dose of 1 mg/g (only 0.5 mg/g for RGW-608) or in 36 fractions of 0.15 mg/g over 18 days (only 4 fractions in 2 days for RGW-608). After single doses a reduction in nerve conduction velocity was seen with all the compounds except metronidazole, which had no significant effect. During chronic exposure, a reduction in NCV occurred towards the end of the course of injections. All compounds produced an effect, although RGW-608 was the most neurotoxic, giving the largest reduction in NCV after only 4 injections. After the end of chronic exposure to misonidazole, Ro-05-9963 and metronidazole, recovery to normal took 2-3 weeks.

We have studied the influence of the antilipidemia drug, clofibrate, and several structurally related analogs on the binding affinity of hemoglobin for oxygen and the radiation sensitivity of the SCCVII/St carcinoma in the mouse. Several compounds in this class reduced hemoglobin affinity in vivo; and two of these, ML1024 (etophyline clofibrate) and ML1037, were at least as effective as clofibrate at reducing hemoglobin affinity and much less toxic. When given orally at a dose of 4.1 m mole/Kg, 1/2-2 hrs before 20 Gy X rays, clofibrate gave radiosensitization in the SCCVII/St tumor equivalent to a 40-fold reduction in hypoxic fraction. ML1024 and ML1037 at a dose (3.0 m mole/Kg), which had a similar effect on hemoglobin, gave much less sensitization of the tumor. Only ML1024 produced a statistically significant effect, equivalent to a four-fold reduction in hypoxic fraction. We conclude that there are several clofibrate analogs which in relation to their toxicity are much better hemoglobin modifiers than the parent compound. They do not, however, show the same radiosensitizing effects, leading us to believe that mechanisms other than changes in hemoglobin/oxygen binding must also be involved.