Influence of Cellular, Microenvironmental, and Growth Parameters on Thermotolerance Kinetics in Vivo in Human Melanoma Xenografts

Abstract

The kinetics of thermotolerance in five human melanoma xenograft lines grown in BALB/c-nu/nu/BOM mice was studied in vivo. Local hyperthermia was given by immersing the tumor-bearing leg of the mice into a thermostatically regulated water bath. Specific growth delay was used as the end point for tumor response. Thermotolerance ratio (TTR), i.e., the ratio of the slopes of dose-response curves (specific growth delay versus heating time) for single-heated and preheated tumors, was used as a quantitative measure of thermotolerance. All melanoma lines developed thermotolerance; TTR reached a maximum (TTR_max) 16 to 24 h after the conditioning heat treatment and then decayed slowly. TTR_max and the time necessary to reach TTR_max tended to increase with increasing conditioning heat dose, whereas the half-time of thermotolerance decay did not change with the conditioning heat dose. The kinetics of thermotolerance differed considerably among the melanoma lines. After a conditioning heat treatment of 43.5°C for 30 min, TTR_max ranged from 2.3 ± 0.5 to 7.0 ± 1.2, the half-time of thermotolerance decay from 53 ± 13 h to 142 ± 30 h and the time necessary to reach complete decay of the thermotolerance from 5 days to more than 14 days. TTR_max showed no correlation to heat sensitivity or any known growth and microenvironmental parameter of the melanoma lines. On the other hand, TTR_max was positively correlated to TTR_max measured in vitro when cells from the melanomas were studied in soft agar. However, TTR_max in vivo was always somewhat lower than TTR_max in vitro. Consequently, the development of thermotolerance in the melanomas in vivo was governed mainly by the intrinsic ability of the tumor cells to develop thermotolerance and was just slightly modified by the tumor microenvironment. The rate of thermotolerance decay was independent of TTR_max. The half-times of thermotolerance decay in vivo were longer than, and not correlated to, those measured in vitro. However, the decay half-time in vivo tended to increase with increasing tumor volume-doubling time, and to decrease with increasing growth fraction and vascular density. There was no relationship between decay half-time and fraction of radiobiologically hypoxic cells. Consequently, the rate of thermotolerance decay in the melanomas in vivo was governed mainly by tumor growth parameters and not by intrinsic characteristics of the tumor cells. The considerable difference in the kinetics of thermotolerance observed among the melanoma lines suggests that fractionated hyperthermia cannot be expected to give optimum clinical results until individualized treatment regimens are being used. Thus, a simple and accurate predictive assay for thermotolerance is highly warranted. Intrinsic characteristics of the tumor cells as well as tumor growth properties have to be taken into consideration in the development of such predictive assays. The present results indicate that clinical treatment protocols probably should not prescribe more than one hyperthermic treatment per week.