HYPERTHERMIA AND RADIATION-THERAPY OF LOCAL-REGIONAL RECURRENT BREAST-CANCER - PROGNOSTIC FACTORS FOR RESPONSE AND LOCAL-CONTROL OF DIFFUSE OR NODULAR TUMORS

Over the past decade, hyperthermia has been extensively studied as an adjuvant to radiation therapy in the management of local-regional metastases from adenocarcinoma of the breast. A retrospective review of our experience from July 1982 to January 1990 identified 241 fields in 89 patients which satisfied the following criteria: biopsy confirmation of recurrent or metastatic adenocarcinoma of the breast; involvement of the chest wall and/or regional lymph nodes with diffuse or nodular metastases; treatment which included radiation therapy and externally administered hyperthermia during which mechanically-mapped and/or multipoint normal tissue and intratumoral temperatures were monitored; and at least one follow-up evaluation at 3 weeks or more after completion of treatment. The majority of fields were in patients who had extensive prior treatment including radiation therapy (68%), chemotherapy (86%), and hormonal therapy (58%). Treatment consisted of radiation therapy (average dose: 39.88 Gy) and hyperthermia (1-12 treatments; average 3.12); concurrent chemotherapy or hormonal therapy were also administered in 3% and 32% of the fields, respectively. Parameters characterizing the initial breast cancer, the patient and tumor at the time of hyperthermia, and the treatment were studied in univariate and multivariate analyses with complete response rate at the time of maximum tumor regression and duration of local control as endpoints. The treatments were well tolerated with no life-threatening complications noted. The means for all fields of the mean minimum, mean maximum, and mean average measured intratumoral temperatures were 40.3-degrees-C, and 42.4-degrees-C, respectively. At 3 weeks following completion of radiation therapy, response rates were: complete response partial response (8%), no response (17%), and continuing regression (monotonic regression to < 50% of initial volume) was noted in 22% of the fields. At the time of maximum tumor regression local control was noted in 72% of the fields. Five parameters correlated with higher complete response in univariate and multivariate analysis: lower T-stage of the initial breast cancer; at the time of hyperthermia age < 50 years, Karnofsky status > 95%, and the absence of distant metastases; and the use of concurrent hormonal therapy. The absence of a family history of breast cancer and concurrent radiation dose less-than-or-equal-to 25 Gy significantly correlated with higher complete response in the univariate but not in the best multivariate models. In addition to the five parameters in the multivariate model that correlated with complete response, the seven covariate model that best correlated with duration of local control also included as favorable factors, an interval of at least 2 years between the initial breast cancer and the time of hyperthermia, and no history of prior chemotherapy at the time of hyperthermia. Other parameters significantly correlating with a longer duration of local control in univariate models included: an interval of greater-than-or-equal-to 1 year from initial breast cancer to first relapse, negative axillary nodes at the time of the initial breast cancer, higher mean minimum temperatures, and less than 20% mean % temperature < 41-degrees-C. Neither the number of hyperthermia treatments nor the tumor location showed a significant correlation with outcome. These pretreatment and treatment prognostic factors will aid in the design of subsequent trials for the management of local-regional recurrent metastatic breast cancers.