Robust cytotoxic T cell infiltration has proven to be difficult to achieve in solid tumors. We set out to develop a flexible protocol to efficiently transfect tumor and stromal cells to produce immune-activating cytokines, and thus enhance T cell infiltration while debulking tumor mass. By combining ultrasound with tumor-targeted microbubbles, membrane pores are created and facilitate a controllable and local transfection. Here, we applied a substantially lower transmission frequency (250 kHz) than applied previously. The resulting microbubble oscillation was significantly enhanced, reaching an effective expansion ratio of 35 for a peak negative pressure of 500 kPa in vitro. Combining low-frequency ultrasound with tumor-targeted microbubbles and a DNA plasmid construct, 20% of tumor cells remained viable, and ∼20% of these remaining cells were transfected with a reporter gene both in vitro and in vivo. The majority of cells transfected in vivo were mucin 1+/CD45−tumor cells. Tumor and stromal cells were then transfected with plasmid DNA encoding IFN-β, producing 150 pg/106cells in vitro, a 150-fold increase compared to no-ultrasound or no-plasmid controls and a 50-fold increase compared to treatment with targeted microbubbles and ultrasound (without IFN-β). This enhancement in secretion exceeds previously reported fourfold to fivefold increases with other in vitro treatments. Combined with intraperitoneal administration of checkpoint inhibition, a single application of IFN-β plasmid transfection reduced tumor growth in vivo and recruited efficacious immune cells at both the local and distant tumor sites.

Both adjuvants and focal ablation can alter the local innate immune system and trigger a highly effective systemic response. Our goal is to determine the impact of these treatments on directly treated and distant disease and the mechanisms for the enhanced response obtained by combinatorial treatments.Methods: We combined RNA-sequencing, flow cytometry and TCR-sequencing to dissect the impact of immunotherapy and of immunotherapy combined with ablation on local and systemic immune components.Results: With administration of a toll-like receptor agonist agonist (CpG) alone or CpG combined with same-site ablation, we found dramatic differences between the local and distant tumor environments, where the directly treated tumors were skewed to high expression of F4/80, Cd11b and Tnf and the distant tumors to enhanced Cd11c, Cd3 and Ifng. When ablation was added to immunotherapy, 100% (n=20/20) of directly treated tumors and 90% (n=18/20) of distant tumors were responsive. Comparing the combined ablation-immunotherapy treatment to immunotherapy alone, we find three major mechanistic differences. First, while ablation alone enhanced intratumoral antigen cross-presentation (up to ~8% of CD45+ cells), systemic cross-presentation of tumor antigen remained low. Combining same-site ablation with CpG amplified cross-presentation in the draining lymph node (~16% of CD45+ cells) compared to the ablation-only (~0.1% of CD45+ cells) and immunotherapy-only cohorts (~10% of CD45+ cells). Macrophages and DCs process and present this antigen to CD8+ T-cells, increasing the number of unique T-cell receptor rearrangements in distant tumors. Second, type I interferon (IFN) release from tumor cells increased with the ablation-immunotherapy treatment as compared with ablation or immunotherapy alone. Type I IFN release is synergistic with toll-like receptor activation in enhancing cytokine and chemokine expression. Expression of genes associated with T-cell activation and stimulation (Eomes, Prf1 and Icos) was 27, 56 and 89-fold higher with ablation-immunotherapy treatment as compared to the no-treatment controls (and 12, 32 and 60-fold higher for immunotherapy-only treatment as compared to the no-treatment controls). Third, we found that the ablation-immunotherapy treatment polarized macrophages and dendritic cells towards a CD169 subset systemically, where CD169+ macrophages are an IFN-enhanced subpopulation associated with dead-cell antigen presentation.Conclusion: While the local and distant responses are distinct, CpG combined with ablative focal therapy drives a highly effective systemic immune response.

Microbubble contrast agents are widely used in ultrasound imaging and therapy, typically with transmission center frequencies in the MHz range. Currently, an ultrasound center frequency near 250 kHz is proposed for clinical trials in which ultrasound combined with microbubble contrast agents is applied to open the blood brain barrier, since at this low frequency focusing through the human skull to a predetermined location can be performed with reduced distortion and attenuation compared to higher frequencies. However, the microbubble vibrational response has not yet been carefully evaluated at this low frequency (an order of magnitude below the resonance frequency of these contrast agents). In the past, it was assumed that encapsulated microbubble expansion is maximized near the resonance frequency and monotonically decreases with decreasing frequency. Our results indicated that microbubble expansion was enhanced for 250 kHz transmission as compared with the 1 MHz center frequency. Following 250 kHz insonation, microbubble expansion increased nonlinearly with increasing ultrasonic pressure, and was accurately predicted by either the modified Rayleigh–Plesset equation for a clean bubble or the Marmottant model of a lipid-shelled microbubble. The expansion ratio reached 30-fold with 250 kHz at a peak negative pressure of 400 kPa, as compared to a measured expansion ratio of 1.6 fold for 1 MHz transmission at a similar peak negative pressure. Further, the range of peak negative pressure yielding stable cavitation in vitro was narrow (~100 kPa) for the 250 kHz transmission frequency. Blood brain barrier opening using in vivo transcranial ultrasound in mice followed the same trend as the in vitro experiments, and the pressure range for safe and effective treatment was 75–150 kPa. For pressures above 150 kPa, inertial cavitation and hemorrhage occurred. Therefore, we conclude that (1) at this low frequency, and for the large oscillations, lipid-shelled microbubbles can be approximately modeled as clean gas microbubbles and (2) the development of safe and successful protocols for therapeutic delivery to the brain utilizing 250 kHz or a similar center frequency requires consideration of the narrow pressure window between stable and inertial cavitation.