Abstract 2731: Identifying circulating tumor DNA in patients with osteosarcoma

Introduction: Osteosarcoma (OS) is the most common primary malignant bone tumor and primarily affects children and adolescents. During and after treatment there is no non-invasive test to assess disease response and early relapse. We hypothesize that circulating tumor DNA (ctDNA) can be used to assess tumor burden, evaluate response to treatment, and monitor for recurrence in OS. We effectively tested this hypothesis on a mouse model and have utilized our methods on a first cohort of human patients. Materials and Methods: Twenty blood samples with matched primary tumor material were obtained from 10 patients with OS who were at various stages of treatment or post treatment. Cell free DNA was isolated from blood collected in Cell-Free DNA blood collection tubes using the QIamp Circulating Nucleic Acid kit. We performed massively parallel sequencing (MPS) with the Illumina HiSeq 2500 in the Epigenomics Shared Facility (ESF) to a depth of at least 800x using custom designed probes (Roche NimbleGen SeqCap EZ) capturing genes commonly altered in OS including TP53, RB1, ATRX, DLG2, MET, PTEN, and SLC19A1. We aligned the fastq files with BWA, called the variants using Varscan and GATK HaplotypeCaller, and annotated the variants with SnpEff. Summary of Data: Using an initial threshold of 2% of variants, ctDNA was identified in two of the patients in our study. The first patient’s sample, collected at the completion of therapy, contained a missense mutation in the coding exon of DLG2 at position chr11 84027990 (2.07% of reads). This patient has no radiographic evidence of recurrent disease 12 months after completion of treatment. In a second patient with 4 samples drawn, a mutation was found in the intron region of RB1 at position chr13 48986157. The variant was 0.17% of reads upon completion of therapy, undectable at 3 months, 2.16% of reads at 6 months, and 1.5% of reads at 9 months. The patient is at high-risk for recurrence but is currently disease-free. Sequencing and analysis of serially collected patient samples, including from 3 patients with recurrent disease, and primary tumor is ongoing. Conclusions: We have identified pathogenic variants in the cell free DNA of patients with no radiologic evidence of OS but who are at high risk for relapse. This is the first step to utilizing a non-invasive test to assess tumor burden, response to treatment, and likelihood of recurrence. As part of future work, we will lower the threshold for calling variants and quanti