Targeted Ultrasound Contrast Imaging of Tumor Vasculature With Positively Charged Microbubbles

Abstract

Purpose: Molecular ultrasound imaging of tumor vasculature is being actively investigated with microbubble contrast agents targeted to neovasculature biomarkers. Yet, a universal method of targeting tumor vasculature independent of specific biomarkers, or in their absence, would be desirable. We report the use of electrostatic interaction to achieve adherence of microbubbles to tumor vasculature and resulting tumor delineation by ultrasound imaging.

Methods and materials: Microbubbles were prepared from decafluorobutane gas by amalgamation of aqueous micellar medium. Distearoyl phosphatidylcholine (DSPC) and polyethylene glycol (PEG)-stearate were used as microbubble shell-forming lipids; cationic lipid distearoyl trimethylammoniumpropane (DSTAP) was included to introduce positive electrostatic charge. Microbubbles were subjected to flotation in normal gravity, to remove larger particles. Murine colon adenocarcinoma tumor (MC38, J. Schlom, National Institutes of Health) was inoculated in the hind leg of C57BL/6 mice. Contrast ultrasound imaging was performed under isoflurane anesthesia, using a clinical imaging system in low power mode, with tissue signal suppression (contrast pulse sequencing, 7 MHz, 1 Hz; Mechanical Index, 0.2). The ultrasound probe was positioned to monitor the tumor and contralateral leg muscle; microbubble contrast signal was monitored for 30 minutes or more, after intravenous bolus administration of 2.10 microbubbles. Individual time point frames were extracted from ultrasound video recording and analyzed with ImageJ.

Results: Mean bubble diameter was ~1.6 to 2 μm; 99.9% were less than 5 μm, to prevent blocking blood flow in capillaries. For cationic DSTAP-carrying microbubbles, contrast signal was observed in the tumor beyond 30 minutes after injection. As the fraction of positively charged lipid in the bubble shell was increased, adherent contrast signal in the tumor also increased, but accumulation of DSTAP-microbubbles in the normal muscle increased as well. For bubbles with the highest positive charge tested, DSTAP-DSPC molar ratio 1:4, at 10 minutes after intravenous administration of microbubbles, the contrast signal difference between the tumor and normal muscle was 1.5 (P < 0.005). At 30 minutes, tumor/muscle contrast signal ratio improved and reached 2.1. For the DSTAP-DSPC 1:13 preparation, tumor/muscle signal ratio exceeded 3.6 at 10 minutes and reached 5.4 at 30 minutes. Microbubbles with DSTAP-DSPC ratio 1:22 were optimal for tumor targeting: at 10 minutes, tumor/muscle signal ratio was greater than 7 (P < 0.005); at 30 minutes, greater than 16 (P < 0.01), sufficient for tumor delineation.

Conclusions: Cationic microbubbles are easy to prepare. They selectively accumulate in the tumor vasculature after intravenous administration. These microbubbles provide target-to-control contrast ratio that can exceed an order of magnitude. Adherent microbubbles delineate the tumor mass at extended time points, at 30 minutes and beyond. This may allow for an extension of the contrast ultrasound examination time. Overall, positively charged microbubbles could become a universal ultrasound contrast agent for cancer imaging.

Figure 1.

Zeta potential of the lipid formulations, dependence on the molar ratio between cationic lipid DSTAP and zwitterionic DSPC (mean ± standard deviation).

Figure 2.

Ultrasound contrast signal dependence on the time interval from the intravenous administration bolus of 2.10⁷ cationic microbubbles (4:1 DSPC:DSTAP molar ratio) in tumor-bearing mice. Signal in the tumor tissue (triangles) and control contralateral leg muscle (circles). N=6. Data presented as mean pixel intensity ± standard deviation; *for 10 min time point, p < 0.005.

Figure 3.

Ultrasound contrast imaging of murine hind legs, 10–20–30 min after iv bolus of cationic microbubbles (13:1 DSPC:DSTAP molar ratio), signal from the tumor (red/dark) and control contralateral leg muscle (green/light). N=7. Data presented as mean pixel intensity ± standard deviation; *p < 0.01; **p < 0.05.

Figure 4.

Ultrasound contrast imaging of murine hind legs, 10–20–30 min after iv bolus of cationic microbubbles (22:1 DSPC:DSTAP molar ratio), signal from the tumor (red/dark) and control contralateral leg muscle (green/light). N=4. Data presented as mean pixel intensity ± standard deviation; *p < 0.005; **p <0.01; ***p < 0.02.