Abstract Background: Low grade gliomas (LGGs) account for 1/3 of pediatric brain tumors. Despite favorable outcomes with current therapeutic strategies, some tumors still recur following total resection. New clinically-relevant pediatric LGG (PLGG) animal models that replicate this subset of aggressive tumors are therefore needed. Methods: Fresh tumor tissues from 36 PLGGs were collected. Self-renewal capacity was examined in vitro in 11 tumors with a neurosphere assay. Putative glioma stem cells (CD133+ and/or CD15+ cells) were analyzed in 23 PLGGs by flow cytometry (FCM). Formation of orthotopic xenograft tumors was examined by direct implantation of 25 PLGG tumor cells into the matched locations in the brains of SCID mice followed by serial tumor sub-transplantations. In addition to histo-pathological characterizations, BRAF V600E mutation was validated with pyrosequencing, and CDKN2A deletion with FISH. Results: The yield of viable tumor cells was low. Neurosphere formation was absent in 10/11 PLGGs after 37.6 ± 15.2 days. FCM analysis of 20 pilocytic astrocytomas and 2 gangliogliomas confirmed the presence of low levels (<1.5%) of CD133+ (CD133+/CD15− and CD133+/CD15+) cells, and relatively high fractions of CD133−/CD15+ cells (12.9 ± 20.1% in pilocytic astrocytomas and 67.9 ± 26.1% in gangliogliomas). Unlike 21 pilocytic astrocytomas, 1 grade II astrocytoma and 2 gangliogliomas that did not formed xenograft tumors after 219.9 ± 70.4 days of observations, a pleomorphic xanthoastrocytoma (PXA) (WHO grade II) formed intra-cerebral (IC) xenografts that have been serially subtransplanted 3 times in vivo. Similar to the patient tumor, xenografts from this model (IC-3635PXA) exhibited low cell proliferation (Ki67+ cells <10%), medium microvessel density, homozygous CDKN2A deletion and relatively low abundance of CD133+ (<2%). The GFAP+ and CD15+ cells in xenografts, however, were slightly reduced while the vimentin positive cells dramatically increased from 50% to 100% compared to patient tumor. The allele frequency of BRAF V600E mutation increased from 28% in patient tumor to 33%, 70% and 67% in passage I, II and III xenografts, respectively, and to 69% and 67% in cells grown in FBS-based media and serum-free media, respectively. All these biological changes were accompanied by extensive single cell invasion along blood vessels into normal mouse brains. Conclusion: This study demonstrated that the tumor take rate of PLGGs was low despite the presence of putative glioma stem cells. Nonetheless, we have successfully established a PXA xenograft mouse model and xenograft-derived 3635PXA cell lines which replicate key features of the original patient tumor and demonstrate both enrichment of the BRAF V600E mutant allele frequency and active invasive growth in mouse brains. This novel model system should facilitate biological studies and pre-clinical drug screenings targeting the BRAF V600E mutation in pediatric PXA. Citation Format: Mari Kogiso, Lin Qi, Frank K. Braun, Xiumei Zhao, Zhigang Liu, Holly B. Lindsay, Yuchen Du, Huiyuan Zhang, Sibo Zhao, Ching-Hua Liu, Vidya P. Mehta, Xiaoyun Shen, Jeffrey Murray, Laszlo Perlaky, Jack Su, Patricia Baxter, Adekunle Adesina, Donald W. Parsons, Xinyan Lu, Murali Chintagumpala, Xiao-Nan Li. Development of patient-derived orthotopic xenograft mouse models of pediatric low grade gliomas. [abstract]. In: Proceedings of the 107th Annual Meeting of the American Association for Cancer Research; 2016 Apr 16-20; New Orleans, LA. Philadelphia (PA): AACR; Cancer Res 2016;76(14 Suppl):Abstract nr 2455.