RATIONALE FOR USING INVASIVE THERMOMETRY FOR REGIONAL HYPERTHERMIA OFPELVIC TUMORS

Purpose: Invasive thermometry for regional hyperthermia Is time-consum ing, uncomfortable, and risky for the patient. We tried to estimate th e benefit/cost ratio of invasive thermometry in regional hyperthermia using the radiofrequency system BSD-2000. Methods and Materials: We ev aluated 182 patients with locally advanced pelvic tumors that underwen t regional hyperthermia. In every patient a tumor-related temperature measurement point was obtained either by invasive or minimally invasiv e catheter measurement tracks. In the earlier period for every patient an intratumoral measurement point was decided as obligatory and intra tumoral catheters were implanted intraoperatively, CT guided, or under fluoroscopy. In the later period, invasive thermometry often was avoi ded, if a measurement point in or near the tumor was reached by an end oluminally inserted catheter (rectal, vaginal, cervical, urethral, or vesical). For every patient side effects and complications referred to thermometry were evaluated and compared with the potential benefit of the invasively achieved temperature data. The suitability of endolumi mally registered temperatures is analyzed to estimate local feasibilit y (specific absorption rate achieved) and local effectiveness (thermal parameters correlated with response). Results: In 74 of 182 patients invasive thermometry was performed, at most CT-guided for soft tissue sarcomas and rectal recurrences. In 14 of 74 (19%) side effects such a s local inflammation, pain, or abscess formation occurred that enforce d removal of the catheter. However, local problems were strongly corre lated with the dwell time of the catheter and nearly never occurred fo r dwell times less than 5 days. Fortunately, no fatal complications (e .g., bleeding or perforation) occurred during or after implantation wh ich could be attributed to the invasive thermometry procedure. Endolum inal tumor-related temperature rises per time unit (to estimate power density) were correlated with intratumoral rises at the same patients (where both measurements were available). For a subgroup of patients p ooled in two Phase II studies with rectal (n = 37) and cervical (n = 1 8) carcinomas thermal parameters derived from endoluminal measurements were correlated with response or local control, resp. Conclusions: If a tumor-related endoluminal temperature measurement point is availabl e, additional invasive thermometry gives no further information to imp rove the power deposition pattern. For primary rectal and cervical can cer, and probably as well for prostate, bladder and anal cancer, endol uminal measurements are suitable to estimate local feasibility and eff ectiveness. Therefore, invasive thermometry is dispensable in the majo rity of patients. In some selected cases, temperature measurement in t he tumor center is required to estimate the maximum temperature. In th ose cases, dwell time of catheters should be minimized-and it should b e considered to perform invasive thermometry at the beginning (one or two heat treatments). (C) 1998 Elsevier Science Inc.