Key points are not available for this paper at this time.
BACKGROUND: The objective of this study was to evaluate noninvasive magnetic resonance (MR) thermography for the monitoring of regional hyperthermia (RHT) in patients with soft tissue sarcomas of the lower extremities and pelvis. METHODS: Noninvasive MR monitoring during RHT was performed in 9 patients who had high-risk soft tissue sarcomas of the lower extremities or pelvis during neoadjuvant chemotherapy plus RHT in the scope of the European Organization for Research and Treatment of Cancer 62961/European Society for Hyperthermic Oncology RHT-95 study. Anatomic and temperature-sensitive data sets were acquired every 10 minutes before and during RHT (using gradient-echo-sequences with variable echo times). MR temperature distributions were derived from the phase differences by using the proton-resonance frequency shift method. A phase convolution setting phase shifts to zero in the fat tissue was performed as a drift correction. The mean MR temperatures in the tumor and muscles and the index temperatures (e.g., T90, which covers 90% of the target volume) and thermal doses were determined and compared with pathohistologic responses and direct temperature measurements if available. RESULTS: Thirty of 72 MR-thermography data sets (>40% of heat sessions) were evaluable. A significant correlation was observed between pathohistologic response (defined as a necrosis rate >or=90%) and standardized thermal parameters, such as thermal dose cumulative equivalent minutes at 43 degrees C to 90% of the target volume (T90) (P = .050), mean T90 (P = .048), or T50 (P = .050). The correlation of 13 conventional temperature measurements performed in selected patients and sessions invasively in the tumor or noninvasively in rectum and bladder revealed an excellent correlation with MR temperatures (R2 = .96). CONCLUSIONS: Noninvasive MR thermography of soft tissue sarcoma was feasible and suitable for validating the quality of heating during RHT.
Gellermann et al. (Thu,) studied this question.