Ultrasound-triggered therapeutic microbubbles enhance the efficacy of cytotoxic drugs by increasing circulation and tumor drug accumulation and limiting bioavailability and toxicity in normal tissues.

Most cancer patients receive chemotherapy at some stage of their treatment which makes improving the efficacy of cytotoxic drugs an ongoing and important goal. Despite large numbers of potent anti-cancer agents being developed, a major obstacle to clinical translation remains the inability to deliver therapeutic doses to a tumor without causing intolerable side effects. To address this problem, there has been intense interest in nanoformulations and targeted delivery to improve cancer outcomes. The aim of this work was to demonstrate how vascular endothelial growth factor receptor 2 (VEGFR2)-targeted, ultrasound-triggered delivery with therapeutic microbubbles (thMBs) could improve the therapeutic range of cytotoxic drugs. Methods: Using a microfluidic microbubble production platform, we generated thMBs comprising VEGFR2-targeted microbubbles with attached liposomal payloads for localised ultrasound-triggered delivery of irinotecan and SN38 in mouse models of colorectal cancer. Intravenous injection into tumor-bearing mice was used to examine targeting efficiency and tumor pharmacodynamics. High-frequency ultrasound and bioluminescent imaging were used to visualise microbubbles in real-time. Tandem mass spectrometry (LC-MS/MS) was used to quantitate intratumoral drug delivery and tissue biodistribution. Finally, 89Zr PET radiotracing was used to compare biodistribution and tumor accumulation of ultrasound-triggered SN38 thMBs with VEGFR2-targeted SN38 liposomes alone. Results: ThMBs specifically bound VEGFR2 in vitro and significantly improved tumor responses to low dose irinotecan and SN38 in human colorectal cancer xenografts. An ultrasound trigger was essential to achieve the selective effects of thMBs as without it, thMBs failed to extend intratumoral drug delivery or demonstrate enhanced tumor responses. Sensitive LC-MS/MS quantification of drugs and their metabolites demonstrated that thMBs extended drug exposure in tumors but limited exposure in healthy tissues, not exposed to ultrasound, by persistent encapsulation of drug prior to elimination. 89Zr PET radiotracing showed that the percentage injected dose in tumors achieved with thMBs was twice that of VEGFR2-targeted SN38 liposomes alone. Conclusions: thMBs provide a generic platform for the targeted, ultrasound-triggered delivery of cytotoxic drugs by enhancing tumor responses to low dose drug delivery via combined effects on circulation, tumor drug accumulation and exposure and altered metabolism in normal tissues.

Expanding 3D geometry for enhanced on-chip microbubble production and single step formation of liposome modified microbubbles.

Micron sized, lipid stabilized bubbles of gas are of interest as contrast agents for ultra-sound (US) imaging and increasingly as delivery vehicles for targeted, triggered, therapeutic delivery. Microfluidics provides a reproducible means for microbubble production and surface functionalisation. In this study, microbubbles are generated on chip using flow-focussing microfluidic devices that combine streams of gas and liquid through a nozzle a few microns wide and then subjecting the two phases to a downstream pressure drop. While microfluidics has successfully demonstrated the generation of monodisperse bubble populations, these approaches inherently produce low bubble counts. We introduce a new micro-spray flow regime that generates consistently high bubble concentrations that are more clinically relevant compared to traditional monodisperse bubble populations. Final bubble concentrations produced by the micro-spray regime were up to 10(10) bubbles mL(-1). The technique is shown to be highly reproducible and by using multiplexed chip arrays, the time taken to produce one millilitre of sample containing 10(10) bubbles mL(-1) was ∼10 min. Further, we also demonstrate that it is possible to attach liposomes, loaded with quantum dots (QDs) or fluorescein, in a single step during MBs formation.

High-frequency ultrasound for in vivo measurement of colon wall thickness in mice.

Mouse models are becoming increasingly important in the study of molecular mechanisms of colorectal disease and in the development of novel therapeutics. To enhance this phase of preclinical research, cost-effective, easy to use noninvasive imaging is required to detect and monitor changes in the colon wall associated with disease pathology. This study investigated the feasibility of using 40-MHz (high frequency) B-mode ultrasound (HF-US) to image the normal mouse colon and measure its thickness in vivo by establishing a robust imaging protocol and conducting a blinded comparison of colon wall thickness (CWT) measurement between and within operators. The in vivo and ex vivo appearance of mouse colon under HF-US revealed distinct patterns. Colon wall thickness was reproducibly and accurately measured using HF-US compared with histology measurement. The technique was more sensitive in detecting changes in CWT in distal than proximal colon as it showed the highest level of inter- and intraoperator reproducibility. Using the protocol described, it is possible to detect changes in thickness of 0.09 mm and 0.25 mm in distal and proximal colon, respectively. In conclusion, HF-US provides an easy to use and noninvasive method to perform anatomical investigations of mouse colon and to monitor changes in CWT.

Assessment of ultrasound monitor image display performance.

The display monitor on an ultrasound scanner is used to make primary diagnoses. In this study, 31 ultrasound systems were assessed against current American Association of Physicists in Medicine (AAPM) display standards. Measurements of peak levels (L(max) and L(min)) were generated. Ambient light, L(amb) (cd/m(2)) and room illuminance, L(x) (Lux) were measured. Luminance ratio was calculated (LR' = (L(max)+L(amb))/(L(min)+L(amb))). Initially, only 8/31 systems (26%) passed all the criteria. After adjustment, a further 7/31 (23%) passed making a total of 15/31 passes (48%). A total of 16/31 (52%) were considered overall fails: three due to poor room lighting, 14 due to poor monitor performance. Considering errors this could be as low as 6/31 (19%). Although further work is required to confirm the applicability of these results, it is of concern that three-quarters of ultrasound scanners could be suboptimally adjusted with 19%-55% unable to pass the AAPM criteria. The impact of this on clinical practice is unknown but there is clearly a need to review display quality assurance on ultrasound scanners.

Measurement of hepatic arterial flow using phase contrast magnetic resonance imaging.

Flow measurements through the hepatic artery were performed using retrospectively gated phase contrast magnetic resonance imaging in 22 patients. In 13 patients where three consecutive measurements of hepatic arterial flow were made, the average intra-subject coefficient of variation was 10.0% (range 2.7-21.2%). The measured blood flow was significantly higher in the eight patients where it was expected that there would be a high hepatic arterial flow compared to nine patients where it was expected that there would be a normal hepatic arterial flow based on the patients pathology (435 ml min(-1) versus 235 ml min(-1), difference = 200 ml min(-1), 95% confidence intervals on difference = 73-327 ml min(-1), p < 0.05, independent t-test). Phase correction was performed by fitting a quadratic surface to stationary tissue. The average blood flow correction due to phase correction was 6.3% (20.9 ml min(-1)) with a range of 0.8% (1 ml min(-1)) to 15.0% (70.0 ml min(-1)). Flow measurements through the hepatic artery can be performed, but care must be taken to minimize pixel size and localize accurately.

The temperature dependence of the speed of sound in bovine bone marrow at 750 kHz.

We present values for the speed of sound (SOS) in bovine bone marrow as a function of temperature between 17 degrees C and 44 degrees C. The measurements were made using a time-of-flight approach on a volume of roughly 10 mL, at 750 kHz. The equipment was validated using both distilled water and castor oil. The results show a linear response with SOS changing from 1456.23 ms(-1) at 17 degrees C to 1342.40 ms(-1) at 44 degrees C. The mean value at 37 degrees C was (1371.91 ms(-1)). The temperature coefficient of the SOS was found to be -4.21 +/- 0.19 ms(-1) degrees C(-1). This was well fitted to a least squares model with R2 = 0.88.