A moderate elevation of blood glucose level increases the effectiveness of thermoradiotherapy in a rat tumor model. I. Relative contributions of glucose and heating to tumor acidification.

PURPOSE: To establish dose-effect relationships for tumor acidification induced by heat and glucose as a basis for testing the value of adding glucose administration to combined heat and x-ray treatment at clinically achievable glucose and temperature levels. METHODS AND MATERIALS: Rhabdomyosarcoma BA1112 was grown s.c. in the upper leg of 16-20-week-old Wag/Rij rats. Animals were given 2 consecutive 100-min periods of saline (S) or glucose (G) infusion, while keeping tumor temperature at 37 degrees, 42 degrees, or 43 degrees C for 1 or 2 periods, in various combinations, each involving 6 animals. Glucose was infused i.v. as a 20% solution at 2.4-3 g/kg/h. Tumors were heated using 2,450-MHz electromagnetic radiation, and tumor pH was measured using a 0.7 mm fiberoptic probe. RESULTS: Mean overall baseline pH was 7.00 (SD 0.10). The change induced by G37G43 (i.e., glucose infusion for a full 200 min, first 100 min at 37 degrees C, final 100 min at 43 degrees C) was -0.48 +/- 0.03 (SEM) pH units, and -0.17 +/- 0.03 for S37S43. The effect of G37G42 was -0.37 +/- 0.03 pH units, compared with -0.08 +/- 0.02 for S37S42 and -0.28 +/- 0.04 for glucose alone (G37G37). Glucose was less effective when given after or fully parallel to heating: -0.21 +/- 0.02 pH units for S43G37 and -0.37 +/- 0.02 for G43G43. CONCLUSION: The glucose-induced tumor pH drop is much more pronounced than that induced by heat, both of which are dose dependent. The effects of glucose and heat seem additive if heating is started when glucose-induced acidification has reached its plateau level, but the overall effect is diminished if administration is fully simultaneous or in reversed order. Schedule G37G43 is optimal with respect to tumor acidification. Its predicted superiority in thermoradiotherapy as compared with S37S42, S37S43, and G37G42 treatment regimens was confirmed in a subsequent experimental tumor control study.

A moderate elevation of blood glucose level increases the effectiveness of thermoradiotherapy in a rat tumor model II. Improved tumor control at clinically achievable temperatures.

PURPOSE: To assess the therapeutic gain (at the TCD(50) level) that can be obtained by boosting thermoradiotherapy with intravenous glucose infusion at different temperatures. This completes our series of studies to determine the optimal conditions and the effectiveness of glucose administration at clinically achievable glucose levels and treatment temperatures. METHODS AND MATERIALS: Subcutaneous rat rhabdomyosarcoma BA1112 was irradiated with graded single doses of 300-kV X-rays (dose range 0-60 Gy). Fifteen minutes after irradiation, a 100-min intravenous infusion was started, consisting of either glucose (20% solution, 2.4-3 g/kg/h) or saline as a control. Then heat was applied to the tumors at 42 degrees C or 43 degrees C (water bath) during a subsequent 100-min period of infusion. Tumor control was scored as the absence of palpable growth at 100 days after treatment. RESULTS: Glucose infusion enhanced tumor control independent of temperature in the range 42-43 degrees C. At 42 degrees C, the TCD(50) for X-irradiation decreased by 5.9 Gy (SEM 1.8 Gy), from 41.6 (1.6) to 35.7 (1.5) Gy, and at 43 degrees C from 33.3 (1.6) to 27.3 (1.5) Gy, representing a glucose enhancement ratio of approximately 1.2. At doses corresponding to the TCD(50) at either 42 or 43 degrees C, the addition of glucose increased tumor control from 50% to 70%. An enhancement ratio of 2.1 was found for the combination of irradiation, glucose infusion, and heating at 43 degrees C, with respect to irradiation alone (TCD(50) 56.3 Gy, reanalyzed earlier data). The contribution of combined heat and glucose to tumor control represented an additive effect, probably on the hypoxic cell population. CONCLUSION: Moderate glucose administration (blood concentration 300 mg/100 mL) sizably improves experimental tumor control after combined X-irradiation and hyperthermia under clinically feasible conditions. Clinical treatment should benefit from this additional modality, in particular if unsatisfactory local control rates are due to insufficient heating. The therapeutic gain has to be evaluated further in clinical studies.

Temporary vascular occlusion and glucose: effects on tumour and normal tissue pH in animal experiments.

The relationship between duration of a period of vascular occlusion and magnitude of pH decrease in tumour and normal tissue was investigated in rats. To acidify tissue pH further, moderate dose glucose (2.4-3.0 g.kg(-1).hr(-1)) was administered intravenously through a catheter positioned in a tail vein, immediately after the clamp was released. This sequence of pH modifying modalities was chosen since it is employed in clinical regional isolation perfusion for recurrence of malignant melanoma of the limbs. Tumour pH in rat rhabdomyosarcoma BA1112 decreased more than normal tissue pH under 10, 20, 30 or 60 min of temporary vascular occlusion. Administration of glucose following any period of clamping always decreased tumour pH further. The largest pH decrease (0.29 pH units) was obtained after 30 min of clamping followed by 60 min glucose and 60 min saline infusion. In the clinic the combination of a maximum of 30 min of clamping followed by moderate dose glucose infusion, which can decrease tumour pH effectively, can be easily achieved in the setting of regional isolation perfusion. It can be used for treatment modalities that are known to be enhanced at lowered tissue pH, such as hyperthermia and certain chemotherapeutic drugs. These results form the basis for studying the therapeutic gain which can be obtained with this model.

Dose-effect relationships for glucose-induced tumour acidification and its erythrocyte flux.

Preclinical investigations were performed with glucose administration in WAG/Rij rats carrying the rhabdomyosarcoma BA1112 in two sites per animal: one in the subcutis of the flank (for pH measurements in the tumour tissue) and one in the transparent "sandwich" chamber for measuring the erythrocyte flux in the tumour tissue as an indication for changes in tumour blood flow. A glucose solution (20%) was slowly infused intravenously in a range of dose levels, similar to those reported for inducing long-term hyperglycaemia in man. The eventual aim of such investigations is to sensitise tumours for hyperthermic treatment. This approach is not new, but the present experiments were performed with the aim to explore the level of the minimal amount of glucose which would nonetheless yield a likely therapeutic effect. Endpoints in this study were the blood glucose level and pH and the relative erythrocyte flux in the tumour tissue. Obviously, as one would expect, many significant changes in the various parameters were found as a response to administration of glucose. However, the changes in the blood glucose level, the induced decrease in tumour pH and the influence of the tumour volume did not show a well-defined relationship which was reliable enough to predict the exact influence of the various parameters on the magnitude of the desired changes in individual animals and/or tumours. This was probably caused by interfering differences in physiological feedback mechanisms. Nonetheless, the data indicate that the optimal effect was not obtained with the highest treatment level, but with moderate doses of glucose, i.e. 2.4-3 g.kg/h which induced a satisfactory tumour acidification of 0.25 pH units. This may turn out to a clinically useful pH drop for enhancing the cytotoxic effect of hyperthermia. The erythrocyte flux through the tumour tissue does not appear to be influenced to a sizeable extent by such a treatment.

Differences in the response of the microcirculation to hyperthermia in five different tumours.

The response of the microcirculation in five different tumours, growing in 'sandwich' observation chambers in the back of the rat, to hyperthermia was investigated. The tumours investigated encompassed three human xenografted tumours, of which two were carcinomata of the colon and one of the lung, and two isologous rat tumours, the Rhabdomyosarcoma BA1112 and a rat mammary carcinoma. It was concluded (1) that the various tumours required significantly different exposure times for inducing 50% stoppage of the tumour microcirculation (ST50). This seems to indicate that differences in the characteristics of the tumour cells are more important for causing microcirculatory stoppage than is the sensitivity of the cells of the blood vessels. (2) An increase in surface (i.e. volume) was observed in all four tumours examined for this phenomenon. The rate of increase (usually 1-2% per hour at 42.5 degrees C) was, however, significantly different between the various tumours. This rate was higher exposure temperatures (43 and 43.5 degrees C), but this was only investigated for the Rhabdomyosarcoma BA1112. Extensive statistical analysis of this phenomenon of volume increase could not demonstrate a correlation with any of the circulation parameters. (3) The relative velocity of the erythrocytes in selected capillaries in the tumours decreases as a result of the hyperthermic treatment, and is probably related to the tumour-specific ST50. (4) A human colon carcinoma xenograft, one of the tumours investigated, showed strong fluctuations in the parameter 'erythrocyte velocity'. The appearance of such fluctuations did not seem to influence the heat-induced stoppage of the circulation. Probably the phenomenon of fluctuations in the velocities of the erythrocytes in the tumour capillaries is a tumour-specific phenomenon.

Experimental hyperthermic treatment of a human colon carcinoma xenograft. The thermal sensitivity of the tumour microcirculation.

The effect of hyperthermia on the microcirculation of a human colon tumour growing in a 'sandwich' observation chamber in immune-suppressed rats was investigated. Evaluation of the effect was based on microscopic observation, measurement of the relative flow rate of the blood in the capillaries of the tumour and photographic recording. The results indicated that moderate hyperthermia (3 h at 42.5 degrees C) has a destructive effect on the microcirculation of the tumour, followed the next day by severe necrosis. These results indicate that this human colon carcinoma xenograft has--for the endpoints that were investigated--a heat sensitivity that is comparable with rodent tumours.

Destruction of rat mammary tumor and normal tissue microcirculation by hematoporphyrin derivative photoradiation observed in vivo in sandwich observation chambers.

The effect of hematoporphyrin derivative photoradiation on tumor and normal tissue microcirculation was studied microscopically in vivo on rats with mammary carcinomas transplanted into subcutis in transparent observation chambers. One day after i.p. injection of hematoporphyrin derivative (15 mg/kg), chambers were exposed to red light (632 +/- 2 nm, eight light dose values, 0 to 270 J/cm2). After an initial blanching (ischemia) of the tumor accompanied by apparent vasoconstriction, reperfusion was observed with a slowing down of the tumor circulation, vasodilatation, and eventually a complete stasis, together with diffuse hemorrhages and subsequent necrosis. Besides, in large normal tissue vessels, platelet aggregates were observed, but no hemorrhage. Tumor regrowth occurred unless the tumor circulation and the adjacent normal tissue circulation were both destroyed. Tumor cell viability after treatment was assessed by transplanting the tumor from the chamber into the flank of the same animal. Even after a combined porphyrin and light dose 4 times the lethal dose for all tissues in the chamber, five of five transplanted tumors did regrow. This leads to the conclusion that, in our model system, tumor cell death after photoradiation occurs secondary to destruction of the microcirculation. In order to obtain additional information on normal tissue damage, rat ears were also irradiated. For the same light dose, the biological effect was only slightly larger than that of the normal tissue in the observation chambers, even though the measured ratio of porphyrin concentrations in ears and normal tissue in the chambers (subcutis) was about six.

Hyperthermia-induced alteration in erythrocyte velocity in tumors.

The effect of a hyperthermic treatment on the microcirculation of the experimental tumor Rhabdomyosarcoma BA1112 was investigated. The tumor was grown in 'sandwich' chambers in the rat skin. The erythrocyte velocity in selected capillaries was used to assess changes in the microcirculation. For this purpose, a simplified comparative method in which the (adjusted) velocity of a train of light dots indicated the velocity in the capillary was employed. During hyperthermic treatment (n = 16) consisting of an exposure of 180 minutes at 43 degrees C, it was apparent that a slight increase in erythrocyte velocity first occurred. This was followed, however, by a decrease and eventually a complete standstill.

Tumour pH in human mammary carcinoma.

Human tumour pH was investigated using the new Philips C902S tissue pH electrode. In 22 mammary carcinomas a pH of 7.29 +/- 0.050 (S.E.M.) was observed, whereas in the human subcutis this value was 7.63 +/- 0.034. Tumour pH in some experimental rat tumours was slightly lower than in humans, 7.15 +/- 0.029 in rhabdomyosarcoma BA1112 and 7.07 +/- 0.024 in a small group of other miscellaneous rat tumours. Rat skeletal muscle was found to have a pH of 7.59 +/- 0.070. It is concluded that a small but highly significant (P less than 0.0001) pH difference exists between human mammary carcinoma and human subcutis. This difference is smaller than expected on the basis of animal studies.