Abstract PR001: Deciphering clonal evolution under chemotherapy in high-risk neuroblastoma using patient derived models

Abstract Neuroblastoma is a pediatric tumor originating from the developing sympathetic nervous system, most often arising in the adrenal medulla in children below 5 years of age. After diagnosis, children with high-risk neuroblastoma are treated with COJEC, which is an aggressive chemotherapeutic treatment protocol encompassing five different chemotherapeutic agents. This is followed by surgical resection of the remaining tumor tissue. The aim of this project was to decipher how the subclonal landscape in the tumor changes during treatment with COJEC. To do this, three different patient derived xenograft (PDX) models were used by implanting dissociated tumor organoids subcutaneously in nude and NSG mice. The mice were subsequently randomized into either obtaining a mock treatment with saline solution (n=16), or treatment with the COJEC protocol (n=18). In PDX1 (3 controls, 4 treated) and in PDX2 (3 controls, 3 treated), progression or moderate response during treatment was seen respectively. In PDX3 (10 controls, 9 treated who went through surgical removal of remaining tumor tissue, 5 treated mice that didn’t undergo tumor resection) late relapses were seen in two of the mice who underwent surgery as well as two mice who did not. In the control group all tumors progressed. SNP-array analysis and RNA-sequencing was conducted on the parental sample as well as the tumors after treatment and relapses, in addition to single cell whole genome sequencing (scWGS) on a subset of samples. Array analysis was followed by computation of the proportion of cells in each sample harboring each genetic alteration, allowing phylogenetic reconstruction with DEVOLUTION. This revealed extensive convergent evolution of specific chromosomal regions across the PDX models. In PDX3 PTPRD deletions (also seen in PDX1) were found in all cells across samples before treatment. Deletions of the second PTPRD allele were detected in a total of six unique ways. Four of these were only seen in the relapses, while the remaining two were found after surgical resection. Other genes affected by convergent deletions were e.g., MACROD2 and LSAMP that were convergent in all three PDX models. High RNA expression of the genes encompassed by the affected chromosomal regions correlated with good treatment response both in the PDX models and patients, but not the copy number. Surprisingly, no significant increase in the number of genetic alterations was found after compared to before treatment. Phylogenetic reconstruction was conducted on scWGS-data for PDX1 and 3. PDX1 displayed a vast genetic diversity across tumors. In PDX3 parallel evolution of unique 1pq+ and 17q+ alterations and genetic bottlenecks with a selective sweep of 1pq/17q++ were seen in several tumors. In conclusion, both COJEC treatment and surgical resection induced bottlenecks and selective sweeps in the subclonal landscape and extensive convergent evolution was seen across tumors. Combined with the expression results, it supports the relevance of these specific genes in the evolution of neuroblastoma under treatment pressure. Citation Format: Natalie Andersson, Adriana Mañas Nuñez, Kristina Aaltonen, Karin Hansson, Alexandra Seger, Katarzyna Radke, Javanshir Esfandyari, Daniel Bexell, David Gisselsson. Deciphering clonal evolution under chemotherapy in high-risk neuroblastoma using patient derived models [abstract]. In: Proceedings of the AACR Special Conference on the Evolutionary Dynamics in Carcinogenesis and Response to Therapy; 2022 Mar 14-17. Philadelphia (PA): AACR; Cancer Res 2022;82(10 Suppl):Abstract nr PR001.