Purpose: To evaluate the influence of pretreatment tumor temperatures and the temperature differential between treatment and pretreatment temperatures on local tumor control in patients who underwent combined radiation therapy and hyperthermia.
Methods and materials: Mapped intratumoral temperatures were measured immediately prior to and during hyperthermia in 138 hyperthermia fields among 59 patients with nodular (60 fields) or diffuse (78 fields) superficially-located tumors. In the nodular subgroup there were 40 fields with adenocarcinomas (31 breast, two prostate, seven other primary sites), six melanomas, nine squamous cell carcinomas, and five other histologies. The fields with diffuse tumor involvement consisted of 77 adenocarcinomas (67 breast, 10 other) and one melanoma. The maximum, minimum, and average temperatures were determined for both the pretreatment (pTmax, pTmin, pTave) and treatment (Tmax, Tmin, Tave) distributions and the differences, Dm = Tmin-pTmax, and Da = Tmin-pTave, computed. These quantities were averaged over treatments to produce the corresponding mean quantities for each hyperthermia field. Univariate and multivariate analyses were performed to determine treatment and pretreatment parameters which best correlated with the duration of local control.
Results: Pretreatment tumor temperatures were significantly lower than the oral temperatures with mean pTmax, mean pTmin, and mean pTave of 36.2 degrees C, 34.2 degrees C, and 35.4 degrees C, respectively. For the adenocarcinomas with diffuse involvement within the hyperthermia field, the covariates best correlating with local control duration on univariate analysis were concurrent radiation dose (p = 0.0026), Dm (p = 0.009), pTmax (p = 0.012) and Da (p = 0.036). Lower pTmax and larger Dm and Da were predictive for longer local control. In multivariate analyses, all thermal parameters lost power, however, the best model included Dm which was significant at the p = 0.040 level. For the nodular subgroup, nonthermal parameters and dichotomized thermal parameters were of prognostic significance for local control.
Conclusion: For fields diffusely involved with adenocarcinoma significant correlations with duration of local control have been demonstrated both for a) low pretreatment temperatures and b) large differentials between treatment and pretreatment intratumoral temperatures. These correlations were also found in a dichotomized description for fields with nodular tumors. The results support the concept that pretreatment hypothermic conditions can lead to an increase in thermal sensitization and may help explain the excellent clinical results noted in the treatment of superficial tumors with radiation and hyperthermia. Further exploitation of this approach by planned cooling of superficially-located recurrent tumors prior to hyperthermia treatment warrants investigation.