W. E. Barry1, G. Asuelime1, L. Wang2, E. S. Kim1 1Children’s Hospital Los Angeles,Pediatric Surgery,Los Angeles, CA, USA 2Children’s Hospital Los Angeles,Pathology,Los Angeles, CA, USA
Introduction: We have previously described a small subpopulation of neuroblastoma (NB) cancer stem cells (CSCs) that express the granulocyte colony stimulating factor (G-CSF) receptor. In previous studies, the ligand G-CSF leads to increased primary tumor growth and increased metastasis in mice, but the impact of G-CSF on established metastatic disease is not fully understood. Despite these concerns, G-CSF is commonly administered to children with high-risk metastatic NB. We therefore hypothesize that G-CSF will lead to a more aggressive phenotype in vitro and decreased survival in a murine NB metastatic model of minimal residual disease (MRD).
Methods: Metastatic cell lines from the liver and bone marrow of NSG mice were derived after orthotopic implantation of the human parental NB cell lines (CHLA-255 and SH-SY5Y). DIMSCAN cell viability, colony formation and Matrigel invasion assays were performed using these metastatic cell lines with and without the addition of G-CSF. In vivo, 1 million cells (SH-SY5Y) were injected into the renal capsule of 14 NSG mice. The primary tumor was resected on post-injection day 7 and the mice were randomly assigned to either receive treatment of G-CSF (n=7) or no treatment (n=7). Mice receiving G-CSF were given 250ug/kg/day intraperitoneal for 14 days. Metastatic disease was assessed with bioluminescent imaging and confirmed with histopathology. Survival was analyzed using log-rank analysis, metastasis with Fisher’s exact test, and in vitro studies with Student’s t-test.
Results: In vitro, increasing doses of G-CSF had no significant impact on cell viability by DIMSCAN assay, nor did G-CSF increase the invasive capacity of metastatic NB cells by matrigel invasion assay. Using a colony formation assay, the addition of G-CSF was found to increase colony counts in parental neuroblastoma cells (SH-SY5Y) and bone marrow metastatic cells compared to control (p≤0.01). In vivo, the burden of metastatic disease at week 4 by bioluminescent imaging, as well as the post-necropsy burden of liver disease were increased in mice treated with G-CSF, however this did not reach statistical significance (p=0.2 for both). Overall survival was significantly decreased in the mice treated with G-CSF (Figure).
Conclusions: In vitro, G-CSF does not appear to impact cell viability or invasiveness of metastatic NB cells but may have a modest effect on anchorage-independent growth, a hallmark of carcinogenesis. In vivo, G-CSF leads to decreased survival in a metastatic neuroblastoma mouse model of minimal residual disease. Further studies are necessary to validate and fully elucidate the potential etiologies of these findings to determine the safety of its use in patients with high-risk NB.
