(P 045) DETECTION OF CELL-FREE TUMOR DNA IN PATIENTS WITH NEWLY DIAGNOSED PEDIATRIC BONE AND SOFT TISSUE SARCOMAS

Objective: PlaPlasma cell-free DNA (cfDNA) is increasingly being utilized to diagnose, monitor disease progression, inform therapy, and further understand the etiology of solid cancers in adults. Reports of the utility of cfDNA in pediatric cancer patients with solid tumors are limited. Here, we explored the feasibility of detecting cancer-derived circulating tumor DNA (ctDNA) as a potential approach for diagnosing and monitoring pediatric sarcomas using Low-Pass Whole Genome Sequencing (LP-WGS).

Methods: Blood samples were collected at baseline, at the time of disease evaluations, and prior to and after local control therapy (surgery and/or radiation) from patients with pediatric bone and soft tissue sarcomas after written informed consent. cfDNA was extracted with the MagMax cell-free total nucleic acid extraction kit from plasma isolated from 4-6ml peripheral blood collected in EDTA tubes. LP-WGS libraries were constructed with the xGen Prism DNA Library Prep Kit using 5ng of cfDNA and sequenced on an Illumina NextSeq 500 platform to an average depth of coverage of 4.4x. Copy number alterations (CNAs) were called using the ichorCNA R package. The CNAs identified using cfDNA samples were compared with those identified by chromosomal microarray analysis (CMA) of the patient’s tumor. Genome-wide similarity of copy number between CMA and LP-WGS was calculated using Pearson’s correlation. Correlations between patient and tumor characteristics and the detection of ctDNA were analyzed using the Kruskal Wallis and Fisher’s exact test, as appropriate.

Results: We report our preliminary findings on the baseline samples collected between July 2019 and May 2022 from 31 newly diagnosed patients [20 females, median age 12 years (range, 0.4 – 19)] with pediatric bone and soft tissue sarcomas, including 9 (29%) with osteosarcoma, 13 (42%) with Ewing sarcoma and 9 (29%) with soft tissue sarcomas. Using our LP-WGS assay, ctDNA was detected in 18 (67%) of the 27 samples obtained from patients prior to any therapy. Of these 27 patients, 18 had localized disease, the median maximum tumor dimension (MTD) of the primary was 7.4 cm (range 2.7 – 20.3), and 4 of 26 and 11 of 20 had elevated alkaline phosphatase (ALP) and lactate dehydrogenase (LDH), respectively. Except for MTD of the primary tumor (p-value 0.021), tumor burden as assessed by stage (localized vs. metastatic) and elevated ALP or LDH did not significantly correlate with detection of ctDNA. Age and tumor type also did not significantly correlate with the detection of ctDNA. CMA was available from the tumor of 26 patients (21 with detectable abnormalities). The CNA profiles detected in plasma using LP-WGS (N = 11/16, 69%) were very similar to those of the tumors observed by CMA and were reflected by high Pearson’s correlation scores for each case (r2 > 0.7). For the remaining 5 abnormal LP-WGS profiles, there were CNAs in the tumor that were not identified by LP-WGS (0.3 < r2 < 0.7) due to low tumor fraction in plasma in 4 cases and a complex karyotype in one case. Two samples did not have tumor data available for comparison.

Conclusion: These preliminary data suggest that our LP-WGS assay detects ctDNA in approximately two-thirds of baseline samples from patients with newly diagnosed pediatric sarcoma and were similar to patterns of aberrations in the patient’s tumor. The detection of ctDNA may be correlated with MTD of the primary and independent of tumor stage. Longer follow-up is needed to assess correlation with outcomes. Ongoing efforts to validate assays to detect specific fusion transcripts and mutations common in pediatric sarcomas will likely improve ctDNA detection rates.