Inhibition of C-C Motif Chemokine Ligand 2 with Chemotherapy Prolongs Survival in a Minimal Residual Disease Mouse Model of Metastatic Neuroblastoma

Background: While 80% of children with high-risk neuroblastoma (NB) achieve remission with multi-modal therapy, the majority will succumb to recurrent metastatic disease. C-C motif chemokine ligand 2 (CCL2) is known to attract macrophages and monocytes to sites of inflammation, including solid tumors. These tumor-associated macrophages (TAMs) have been found to facilitate metastatic disease in several human cancers. We sought to characterize the in vitro effect of CCL2 on neuroblastoma cells and hypothesized that inhibiting CCL2 in combination with chemotherapy would prolong survival in a metastatic mouse model of neuroblastoma.

Methods: Human NB cell lines (CHLA-255, NGP, SHSY5Y, KCNR) were utilized. NB cells were co-cultured with human monocytes, and supernatant CCL2 protein levels were measured by ELISA. NB cell proliferation assays were performed with varying concentrations of CCL2. We tracked NB cell migration corresponding to an increasing gradient of CCL2 + anti-CCL2 antibody by in vitro chemotaxis assay with live cell microscopy. In vivo, a NB tumor resection model of minimal residual disease (MRD) was established by intrarenally injecting NB cells in NSG mice (CHLA-255 n=40, NGP n=40 and COG-415 PDX n=44), followed by xenograft tumor resection 1 week later. Mice were divided into four cohorts: untreated control, anti-CCL2 antibody alone (Carlumab, Janssen Pharmaceuticals, NJ,USA), etoposide alone, and anti-CCL2 antibody+etoposide. CHLA-255 xenograft tumors from mice treated + anti-CCL2 antibody underwent FACS to determine the presence of TAMs. Student t-test, linear regression, and ANOVA were used.

Results: CCL2 supernatant protein levels from NB cells (CHLA-255, NGP, KCNR, SHSY5Y) co-cultured with monocytes were significantly higher than in monocyte or NB supernatant alone (p < 0.05). Monocytes mixed with NB supernatant led to significantly higher CCL2 levels, compared to NB cells mixed with monocyte supernatant, NB supernatant alone, and monocyte supernatant alone (p < 0.05). As expected, CCL2 does not increase NB cell proliferation (CHLA-255, NGP, SHSY5Y). Interestingly, NB cells migrate toward higher CCL2 concentration, suggesting CCL2 induces NB cells to recruit monocytes/TAMs and stimulate a positive feedback. This chemotaxis was abrogated by anti-CCL2 antibody (Figure 1). In vivo, anti-CCL2 antibody+etoposide significantly prolonged survival in three cell lines (CHLA-255, NGP, COG-415 PDX) of our MRD mouse model of recurrent metastatic disease (Figure 2). CHLA-255 xenograft tumors from mice treated with anti-CCL2 antibody demonstrated a trend towards decreased intratumoral TAMs compared to untreated tumors (p=0.255).

Conclusion: Our in vitro studies show that CCL2 production by monocytes is augmented by neuroblastoma cells, and that increasing concentrations of CCL2 leads to increased migration of NB cells, suggesting a positive feedback loop for metastatic disease in NB. In vivo, anti-CCL2 antibody combined with chemotherapy significantly increases survival in our MRD mouse model of recurrent metastatic disease. Inhibition of CCL2 may represent a novel treatment strategy to target recurrent metastatic disease in neuroblastoma.