Toward CD47 Blockade Immunotherapy in Brain Metastases with MRI-Guided Focused Ultrasound Blood Brain Barrier Opening

Introduction: Brain metastases (BMs) are one of the most common complications of cancer and the most common brain tumor in adults. Breast cancer BMs (BCBMs) rank as the second most common cause of BMs, arising in approximately 5-20% of breast cancer patients. With the rising trend in breast cancer incidence worldwide – and among younger women in particular – the need for more effective, less invasive, less toxic therapies that preserve long-term quality of life has never been greater. Focused ultrasound (FUS) delivers on this need as a non-ionizing, non-invasive, and tunable strategy for highly localized acoustic energy deposition into brain tumors. Low-intensity FUS in combination with i.v. microbubbles can safely, transiently, and repeatedly open the blood brain/tumor barriers (BBB/BTB) - distinct locoregional barriers that restrict transport and distribution of therapeutics within brain tumors. In preclinical and clinical practice, a common strategy for maintaining safety of FUS BBB/BTB opening procedures is real-time, closed-loop control of microbubble oscillation via passive cavitation detection (PCD). While FUS-mediated BBB/BTB opening (BBB/BTBO) has been demonstrated to enable transport of diverse pharmacologic agents into brain tumors, our understanding of the relationship between acoustic exposure conditions and delivery of large, complex biologics (e.g. monoclonal antibodies) remains limited. In this study, we investigate this in the context of FUS BBB/BTBO regimens performed with or without real-time PCD control. We establish feasibility of FUS BBB/BTBO regimens and their capacity to augment tumor-targeted antibody delivery in immunocompetent murine BCBM models (Fig 1A).

Materials and

Methods: To establish orthotopic BCBMs, parental or brain-tropic E0771 mammary carcinoma cells were stereotactically implanted into the right striatum of C57/Bl6 mice. BCBM growth kinetics were validated via serial contrast-enhanced MRI. For each study, mice were randomized into groups on the basis of contrast-enhancing tumor volume-matching via pre-treatment MRI. FUS BBB/BTBO regimens were established using a 9.4T MRI-guided FUS system (1.15 MHz, 0.35 MPa, 0.5% duty cycle, 2 min period) with integrated passive cavitation feedback control. BCBM-bearing mice received concomitant i.v. microbubbles (1 × 10 /g B.W.) and fluorescently-labeled monoclonal antibodies prior to sonication. We tested both non-specific IgG and a model antibody targeting CD47, an innate immune checkpoint upregulated within the local BCBM microenvironment. FUS was applied with either a fixed peak negative pressure or with real-time PCD feedback control averaging on a similar pressure over the sonication period – the latter of which mirrors standard practice for clinical FUS BBB opening applications designed to safely and reversibly disrupt the blood brain barrier without damage to brain tissue. Using a custom pipeline developed in Matlab, spectral analysis was performed on frequency-transformed acoustic emission recordings at each sonication target. For each mouse, PCD feedback was then resolved into a harmonic cavitation dose (HCD) representative of a stable (linear) microbubble oscillation regime and inertial cavitation dose (ICD) representative of a inertial (non-linear) microbubble oscillation regime. For antibody delivery studies, antibody penetrance was quantified via ex vivo epifluorescence imaging of perfused whole brains.

Results, Conclusions, and Discussions: Engraftment of E0771 BCBM tumors was confirmed using T1 RARE MR imaging and a multi-target sonication regimen exhibited effective BBB/BTBO in BCBMs. A significantly greater volume and intensity of contrast enhancement was observed on post-contrast coronal and sagittal T1 MRI in E0771-bearing mice (Fig 1B). Notably, this sonication pattern was designed to achieve barrier opening not only directly within the MR-visible tumor bulk (BTB), but also throughout surrounding margins protected by intact BBB. FUS was deployed at either a fixed peak negative pressure (“0.35 MPa (Fixed)”) or with a standardized PCD feedback-controlled pressure calibration approach (“PCD Monitoring '' / “PCD”) (Fig 1C). While both regimens conferred clear evidence of BBB/BTBO on MRI, retrospective PCD analysis revealed a significant increase in HCD but no change in ICD in the fixed pressure regimen relative to PCD feedback-controlled regimen (Fig 1D) – suggesting noteworthy differences in stable cavitation-driven mechanical energy deposition but no overt differences in safety profile. Interestingly, this finding was consequential for large molecule delivery. FUS BBB/BTBO applied at a fixed pressure significantly enhanced penetration of both IgG and αCD47 into BCBMs relative to sham controls, in which antibody penetrance was poor at baseline. PCD-controlled FUS BBB/BTBO displayed a trend toward improved IgG delivery and significantly improved αCD47 delivery in BCBMs; however, in both contexts, fixed pressure outperformed PCD-controlled sonications (Fig 1E-H). Our results confirm the feasibility of performing FUS BBB/BTBO in murine E0771 BCBMs and demonstrate that FUS can significantly enhance the penetration of antibodies (IgG and αCD47) in this setting. However, our findings also highlight the impact of acoustic exposure conditions on antibody delivery and highlight critical considerations for PCD feedback control of FUS regimens intended to preserve safety while maximizing antibody delivery. We expect that these insights will offer timely and important considerations for neuro-oncology clinical trials investigating FUS for antibody delivery. Ongoing preclinical investigations are evaluating the therapeutic impact of rational FUS BBB/BTBO and αCD47 combinations in BCBMs.

Acknowledgements (Optional): This work was supported by NIH DP5OD031846 and T32CA009109, as well as the Wallace H. Coulter Foundation for Translational Research.