Background: Microwave ablation is a promising technology for cancer treatment that has several theoretical benefits over radiofrequency ablation, including more heat generation, less susceptibility to changes in tissue properties and better multiple-applicator support. Single antennas have limited power handling ability, so multiple antennas are often used to create very large zones of ablation. Constructively phased arrays enhance multiple-antenna performance but may require precise antenna positioning with complex, and expensive, adaptive phase control to account for tissue property changes during ablation. As an alternative, we propose temporally switching high powers between antennas to create large and reproducible zones of ablation.

Methods: We used 2D finite-element computer simulations to model microwave heating in tissue using: 1) a single, high-power antenna, 2) 2-4 antennas in phased-array mode and 3) 2-4 antennas in switched mode. The single antenna was assumed to be 15ga (140W) while the other antennas were assumed to be 18-20ga (30-60W). In switched mode, input power to each antenna was N-times the power rating with a duty cycle of 1/N to keep average power below rating. We modeled several different antenna spacings (0.5-5.0cm), switching times (1-10s) and tissue properties. Isotherms and ablation areas were used to compare groups.
Results: Switching between antennas delivers more power to the tissue, which results in larger ablations than those created with single antennas (22-30cm^2 versus 22cm^2) but with reduced invasiveness. Switching also creates more consistent ablations than phasing when spacing or perfusion changes (Figures 1-2). Temperature-dependent changes reduced ablation area less than 10% with switching compared to 20-60% with phasing. Results did not depend significantly on switching time.

Conclusions: Temporal switching is simpler, less expensive and more robust than phase control, and is more effective and potentially less invasive than high-power single antennas for microwave ablation.