Characterization of Cell-Free DNA Size Distribution in Osteosarcoma Patients

Purpose: Cell-free DNA (cfDNA) analysis is a powerful tool for noninvasively predicting patient outcomes. We analyzed the size distribution of cfDNA and assessed its prognostic and diagnostic values in an osteosarcoma cohort. Experimental Design: The fragment size distribution and level of cfDNA were analyzed in 15 healthy donors and 50 patients with osteosarcoma using automated capillary electrophoresis. The prognostic performance of cfDNA size analysis was assessed using univariate and multivariable analyses. By performing wholegenome sequencing of matched cfDNA and osteosarcoma tissue samples, we investigated the correlation between the size and mutation profiles of cfDNA and the mutation concordance between Results: The size of cfDNA fragments in patients with osteosarcoma was significantly shorter than in healthy donors, with the integrative analysis of size distribution and level of cfDNA achieving a high specificity and sensitivity of 100%. The short cfDNA frag-ment (150-bp cut-off) was an independent prognostic predictor in this osteosarcoma cohort [HR, 9.03; 95% confidence interval (CI), 1.13-72.20; P 1/4 0.038]. Shortened cfDNA fragments were found to be a major source of mutations. Enrichment of cfDNA fragments with less than or equal to 150 bp by in silico size selection remarkedly improved the detection of copy-number variation signals up to 2.3-fold when compared with total cfDNA, with a higher concor-dance rate with matched osteosarcoma tissue. Conclusions: This finding demonstrated the potential of cfDNA size profiling in the stratification of poor prognostic patients with osteosarcoma. The short fragments of cfDNA are a promising source for boosting the detection of significant mutations in osteosarcoma. See related commentary by Weiser et al., p. 2017

Automated summary

This study analyzed the size distribution of cfDNA and its prognostic and diagnostic values in an osteosarcoma cohort. The results showed that the size of cfDNA fragments in osteosarcoma patients was significantly shorter than in healthy donors. Integrative analysis of size distribution and level of cfDNA achieved a high specificity and sensitivity of 100%. Shortened cfDNA fragments were found to be a major source of mutations.