Thursday, November 8, 2007 - 2:40 PM
583g

Controlled Vessel Occlusion By Thermal And Ultrasound Responsive Microbubbles For Tumor Therapy

Praveena Mohan, Department of Chemical Engineering, University of Utah, 50 S. Central Campus Drive Rm. 3290 MEB, Salt Lake City, UT 84112, Natalya Rapoport, Bioengineering, University of Utah, 50 S. Central Campus Drive Rm. 3290 MEB, Salt Lake City, UT 84112, and Mikhail Skliar, Chemical Engineering, University of Utah, 50 S. Central Campus Drive Rm. 3290 MEB, Salt Lake City, UT 84112.

We studied the dynamic behavior of microbubbles, formed by phase transition of polymer-stabilized perfluorocarbon-based microdroplets of controlled 100nm -10μm size and customizable phase transition temperature. The bubbles' behavior as a function of temperature was characterized in microfluidic channels of varying diameters (100 – 320 μm). The suspension of microdroplets was injected into the channel placed on thermally controlled stage and the dynamics of the phase transition was observed under an optical microscope. The size of most microdroplets in our experiments was too small to visualize optically. Fig. 1a shows the formation of 100 μm bubbles at 34 ºC in a 100 μm channel. When the temperature was increased to 42 ºC, the microbubbles' diameter increased to 250 μm. Fig. 1b (left panel) shows the result after microdroplets of a different formulation, injected in a 320 μm channel, have transitioned into microbubbles with diameters between 25 and 75 μm. Upon further heating to 44 ºC the size of microbubbles increased to 280 μm (Fig. 1b, right). Further experiments showed that the behavior of microbubbles ranged from collapsing at higher temperatures to being stable and continuing to expand to 10 times or more their original size. It was found that stable bubbles have the tendency to aggregate. We, therefore, speculate that noninvasive, targeted heating using focused ultrasound or other means will lead to controlled embolization and hemostasis after systemic or localized introduction of microdroplets into patient's circulation. This hypothesis was confirmed in vitro, with Figure 1c showing the occlusion of a 100 μm channel caused by the aggregation of stable bubbles formed upon heating. In conjunction with these experiments, drug delivery using these microbubbles were also tested using suspended MDA 231 human breast cancer cells in microfluidic channels. Naturally fluorescing doxorubicin (Dox) was used as the model anti-cancer drug. Tumor cells were mixed with either free drug (0.1 mg/ml) or Dox of similar concentration encapsulated in the microbubbles. Images were taken 15 minutes after mixing (Fig. 2a and 2b for free Dox and Dox encapsulated microbubbles respectively). 1 minute of ultrasound (2 W/cm2 at 3 MHz) was applied to the channel containing microbubbles and images were taken again (Fig. 2c). After sonication, the fluorescence of tumor cells has increased and the background fluorescence of microbubbles decreased, indicating the drug release from the bubbles and its uptake by tumor cell, triggered by ultrasound. To confirm this result in vivo, we are currently conducting the experiments with zebrafish larvae as a perfusion model. We use fli:GFP zebrafish larva (5 -6 days old) since it has endothelial cells that produce green fluorescence protein (GFP), so the vasculature fluoresces green, making it easily to visualize. The ongoing experiments, which will be described during the meeting, involve injecting the prepared microbubbles into the selected vessel of the zebrafish, heating and capturing the perfusion modification with a microscope camera.