Monitoring early tumor response to vascular targeted therapy using super-resolution ultrasound imaging

The current standard for evaluating breast tumor response to neoadjuvant treatment remains the assessment of a tumor size change several weeks after therapy begins. However, the tumor microenvironment (including microvasculature) is known to change drastically before any detectable change in physical size manifests. To that end, we have developed a new high-resolution ultrasound (US) imaging modality in our laboratory, termed super-resolution ultrasound (SR-US), for improved visualization of the tumor angiogenic network. A clinical US scanner (Acuson Sequoia 512, Siemens Healthcare) equipped with a 15L8-S linear array transducer was used for our study. Operation of this system was done using a nonlinear harmonic imaging. A low transmit power (mechanical index <; 0.2) was used to minimize any MB destruction during imaging. After a slow injection of a MB contrast via a tail vein catheter placed in breast cancer-bearing mice (N = 10), each tumor was US imaged for 10 min at baseline and again at 1 and 2 h after dosing with an antiangiogenic drug (bevacizumab, Genentech Inc) or sham control drug. After collecting a stack of US images, images corrupted by respiratory motion were removed using curve fit-based filtering. A singular value decomposition (SVD)-based spatiotemporal filter was then used to localize individual MBs. Subsequently, a SR-US image (i.e., MB density map) was generated by mapping the cumulative MB localizations. SR-US images were reviewed and longitudinal changes in the tumor microvascular network in response to treatment were evaluated. Overall, an acute microvascular response to bevacizumab was found within 2 h of drug dosing. This observation was consistent with immunohistologic findings and suggests that in vivo SR-US imaging has enormous potential in monitoring the early tumor response to drug treatment.