MONITORING OF TISSUE TEMPERATURE DURING HYPERTHERMIA THERAPY *

If hyperthermic therapy is to progress beyond merely the recording of anecdotal case histories to a reliable means of treating cancer, it is essential that temperature distributions be determined as a function of time throughout the treated field of the patient. This is necessary for both localized and whole-body hyperthermia. The thermal dose, which is some sort of integrated function of the measured parameters, time and temperature, is not addressed here. Our problem of determining the temperature distribution is complicated by the fact that heterogeneous tissues have differing power absorption parameters, such as resistivity or ultrasonic attenuation, and differing rates of cooling depending upon the distribution of blood perfusion or the proximity of the tissue element to the surface. It is probable that the power absorption parameters are different for tumors than for surrounding normal tissue and that cooling rates due to heat exchange with the blood are less for tumors than for the surrounding normal tissue. Certainly, the latter is true for the necrotic core of large tumors and is plausible in the general case. Furthermore, neither the physiological response of tissue to heating, such as vasodilatation, nor the pathological effects of heating, such as possible destruction of blood vessels, are fully understood. Both affect the thermal distributions that result. Detailed mathematical models similar to those available for radiation dosimetry are plagued by both the complexity of the heterogeneity of the tissues and by the lack of data upon which to base the models. Thus, the temperature distribution as a function of time at every point within the field cannot be calculated because the parameters needed in mathematical interpolation functions are not known. Neither can the temperature the measured at a sufficient number of points because the clinical trauma would be too great and the perturbation of the field would be excessive. Our purpose here is to indicate the techniques that we use for estimating the temperature distributions. We do not attempt a comprehensive review. The first section of the paper deals with thermometer calibrations and quality assurance of these calibrations. The second section gives two clinical cases, which are used to demonstrate our approach to thermal dosimetry and to indicate its features and limitations. In the third section, thermal measurements required in certain biological experiments are illustrated. We conclude the paper by suggesting areas of research that should lead to improved thermal dosimetry in the near future.