Engineered Donor Marrow Macrophages Phagocytose Cancer Cells and Aggressively Shrink Solid Tumor Xenografts Compared to Tumor Associated Macrophages

Cell-based immunotherapies such as those based on engineered T-cells appear safe and often effective against liquid tumors.  In solid tumors, macrophages are typically abundant, but the density of tumor associated macrophages (TAMs) correlates with poor clinical outcomes as they promote tumor growth and immunosuppression. In our studies, less differentiated donor marrow phagocytes are engineered in order to target tumors and selectively phagocytose cancer cells. Xenograft tumors were first made on the flanks of NSG mice using a tdTomato human lung carcinoma cell line (A549). Systemic injections of anti-human IgG (anti-hum) with large tumors (~70 mm²) showed no effect on tumor growth. However, systemic injection of bone marrow from donor NSG mice together with biweekly anti-hum treatments effectively stopped growth of the solid tumors. Replacing anti-hum with a non-specific antibody had no effect on tumor growth. Based on tdTomato signal intensity within macrophages isolated from tumors, 5-10-fold more donor macrophages are phagocytic compared to resident TAMs (2-3% are phagocytic). Since cancer cells express on their surface ‘self’ markers that limit the phagocytosis of these cells, we inhibited the ‘self’ receptors on the injected donor phagocytes prior to systemic injection of the donor marrow. This combination of ‘self’-receptor inhibition with anti-hum causes a rapid decrease in tumor burden, shrinking tumors by ~40% in just 10 days compared to a similar growth of untreated tumors in the same time period. The anti-hum injection was again necessary as injection of a non-specific antibody failed to affect tumor growth. Tumor analysis showed that >65% of macrophages that were ‘self’-receptor inhibited had phagocytosed the tdTomato A549 cells, which is ~20-fold greater than resident macrophages. Importantly, these cell therapy treatments appear safe with no significant decreases in hematocrit or platelets, which is unlike the anemia that has been reported upon systemic injection of ‘self’ inhibitors. Our results thus suggest that therapies based on engineered macrophages can be safe and effective against solid tumors if three requirements are met: a phagocytic phenotype, target opsonization, and inhibition of ‘self’ signaling. Future experiments will involve characterizing macrophage phenotypes of donor versus resident macrophages as well as efforts to maintain the necessary phagocytic phenotype.