Hyperthermia Extends Survival Rates in GMB

LOS ANGELES--In the first positive randomized US trial of hyperthermiain cancer, glioblastoma multiforme (GBM) patients who receivedhyperthermia in addition to bra-chytherapy survived significantlylonger than those who did not get the heat treatment, Penny K.Sneed, MD, said at the American Society for Therapeutic Radiologyand Oncology (ASTRO) meeting.

The patients who received hyper-thermia for 30 minutes beforeand after a brachytherapy boost survived a median of 85 weeksvs 76 weeks for those getting brachytherapy alone. When analyzedby intent to treat, the survival difference remained significant(80 weeks for hyper-thermia vs 76 weeks for controls).

The probability of 2-year survival was 31% for heat-treated patientsvs 14% for the control group.

The phase II/III study from the Departments of Radiation Oncologyand Neurological Surgery at the University of California, SanFrancisco, involved 112 adult patients who were eligible for brachytherapyafter surgery and conventional radiotherapy; eligible patientshad a unifocal, circumscribed, supratentorial glioblastoma of5 cm or less in diameter.

Three-three patients were not randomized, primarily due to tumorprogression during conventional radiotherapy, leaving 40 patientsrandomized to hyperther-mia and 39 to the control group. "Fora variety of reasons, not all of the randomized patients proceededwith treatment as per the protocol," Dr. Sneed said.

In the end, 33 of the no-heat patients had brachytherapy implants,as did 36 of the heat-treated patients; thus, the results wereanalyzed both by intention to treat and according to those whoactually received brachytherapy. Furthermore, four patients inthe hyperthermia arm did not receive heat treatments but wereanalyzed on the heat arm.

The median freedom from progression was 33 weeks for the brachytherapypatients in the no-heat arm versus 49 weeks for those in the heatarm, a significant difference. Although there were more toxicitieson the heat arm, most of these were mild neurologic toxicitiesthat were fully reversible, or partial seizures, she said.

"This is the first prospective randomized North Americantrial to show a benefit for hyperthermia, and we feel that thisis because we had adequate thermal dose, made possible by theprecise image-based treatment planning and catheter placementinherent in stereotactic technique, along with the lack of painsensation in the brain," Dr. Sneed said.

The median T90 thermal dose (the temperature attained or exceededby 90% of the tumor points) was 14.1 equivalent minutes, withtwo thirds of patients receiving a thermal dose of at least 5equivalent minutes. Previous investigators had estimated thata T90 thermal dose of at least 10 equivalent minutes would beneeded to conduct a meaningful phase III trial.

In the question and answer session, Dr. Sneed noted that her groupis planning a smaller phase II trial, in an attempt to optimizehyperthermia, first by giving longer treatment sessions and secondby heating before brachytherapy and at 72 hours in the middleof brachyther-apy, with an optional third treatment after brachytherapy.

"We'll heat for as long as 90 minutes as needed to get aT90 thermal dose of at least 20 to 50 equivalent minutes,"she commented.

Dr. Sneed believes that hyperthermia is promising for those patientswhose tumors are small enough to be accessible for brachytherapy,which represents about 30% of all glioblastoma multiforme patients.

She pointed out that hyperthermia is "still difficult todo, requires a lot of expertise, and is potentially dangerousin the brain, but if it is carefully controlled, I think it issafe for use in clinical trials, although not for general use."