Stanford University institutional report. Phase I evaluation of equipment for hyperthermia treatment of cancer.

From September 16, 1981, through April 4, 1986, a total of 21 radiative electromagnetic (microwave and radiofrequency), ultrasound and interstitial radio-frequency hyperthermia applicators and three types of thermometry systems underwent extensive phantom and clinical testing at Stanford University. A total of 996 treatment sessions involving 268 separate treatment fields in 131 patients was performed. Thermal profiles were obtained in 847 of these treatment sessions by multipoint and/or mapping techniques involving mechanical translation. The ability of these devices to heat superficial, eccentrically located and deep-seated tumours at the major anatomical locations is evaluated and the temperature distributions, acute and subacute toxicities, and chronic complications compared. Average measured tumour temperatures between 42 degrees C and 43 degrees C were obtained with many of the devices used for superficial heating; average tumour temperatures of 39.6 degrees C to 42.1 degrees C were achieved with the three deep-heating devices. When compared to the goal of obtaining minimum tumour temperatures of 43.0 degrees C, all devices performed poorly. Only 14 per cent (118/847) of treatments with measured thermal profiles achieved minimum intratumoural temperatures of 41 degrees C. Fifty-six per cent of all treatments had associated acute toxicity; 14 per cent of all treatments necessitated power reduction resulting in maximum steady-state temperatures of less than 42.5 degrees C. Direct comparisons between two or more devices utilized to treat the same field were made in 67 instances, including 19 treatment fields in which two or more devices were compared at the same treatment session. The analyses from direct comparisons consistently showed that the static spiral and larger area scanning spiral applicators resulted in more favourable temperature distributions. Three fibreoptic thermometry systems (Luxtron single channel, four channel and eight channel multiple [four] probe array), the BSD Bowman thermistor system and a thermocouple system were evaluated with respect to accuracy, stability and artifacts. The clinical reliability, durability, and patient tolerance of the thermometry systems were investigated. The BSD Bowman and third generation Luxtron systems were found clinically useful, with the former meeting all of our established criteria.