Regional hyperthermia: new standard for soft-tissue sarcomas?

In this issue of The Lancet Oncology, Issels and colleagues1 describe a large, randomised phase 3 study assessing the effects of local hyperthermia in addition to chemotherapy and local therapy (surgery whenever possible with or without radiation) in a group of selected patients with intermediate or high-grade soft-tissue sarcomas greater than 5 cm, and deep.1 The trial was a multicentre study, although three institutions with major previous experience of using hyperthermia contributed 310 of the 341 patients, with the remainder distributed among another six centres. Including this small subset of patients had no effect on the results, and the reader is left to wonder whether the findings of the study could be extrapolated for widespread use or whether the technique should be limited to centres of expertise. If hyperthermia is to be considered a standard treatment, studies of more than 31 patients in six centres are needed. Any attempt to extrapolate the data to a typical group of soft-tissue sarcomas should be avoided. Issels and colleagues' series was heavily weighted towards retroperitoneal sarcomas, where local recurrence is the primary reason for treatment failure, and local progression is a major cause of death.2 None of the series referenced by Issels and colleagues contains such a preponderance of retroperitoneal tumours. The main local adjuvant treatment for soft-tissue sarcomas is radiation, which often cannot be administered to large retroperitoneal tumours, particularly postoperatively. The high local recurrence rate seen even in patients with extremity sarcomas might be due to the fact that radiation was given only to patients with R1 or R2 resections, while it would be standard practice for the tumours included in this study, at least in North America. The risk reduction from the addition of hyperthermia to radiation was less than 0—2, and was not statistically significant. Owing to the extremely high number of local recurrences in this series, the data on disease-free survival are superfluous, and could mislead those who would apply the data to other series where disease-free survival is highly influenced by freedom from metastasis. The authors present a survival analysis of patients who had sarcomas and received the induction therapy prescribed by the protocol unless they progressed before. That analysis shows different results from the intention-to-treat analysis presented in the body of the manuscript. I applaud this inclusion: it matters whether the patient had the disease under study and what treatment was given rather than what treatment …