Sonodynamic therapy is a promising new modality for cancer treatment based on the synergistic effect on cell killing by the combination of a drug (a sonosensitizer) and ultrasound. We have demonstrated sonodynamic toxicity of low concentrations (greater than or equal to 1 microM) of gallium porphyrin ATX-70 in human leukemia HL-525 cells exposed to 50 kHz ultrasound. The mechanism of this ATX-70 - dependent sonosensitization is unknown, but we have established the requirement of extracellular, rather than intracellular, localization of ATX-70 molecules for sonosensitization. Short-lived toxic intermediates produced from ATX-70 by ultrasound are implicated in the mechanism. We are currently exploring the possibility that pi-radicals are involved in the mechanism for sonosensitization by porphyrins. We have studied molecular mechanisms of sonochemical activation of new promising sonodynamic conpounds - organic azocompounds. These compounds were investigated previously as hyperthermia sensitizers, due to their ability to decompose thermally to yield reactive peroxyl radical intermediates capable of cell damage. Using electron paramagnetic resonance spectroscopy and the spin trap DMPO we have demonstrated formation of oxygen-centered radical intermediates from 2,2'azobis (N,N'-dimethyleneisobutyramidine) dihydrochloride (VA-044), 2-(carbomoylazo)-isobutyronitrile (V-30), and 2,2'azobis (2-amidinopropane-dihydrochloride (AAPH). While VA-044 and AAPH could also produce radicals at about 40 degrees C, radical decomposition of V-30 is ultrasound-specific in the range of physiological temperatures. Hence, V-30 appears to be a promising sonosensitizer, which could be decomposed site-specifically using a focused ultrasound beam, thus limiting undesirable damage to tissues outside the focused ultrasound beam.