Abstract
Midline K27M gliomas are the most challenging type of gliomas in pediatric patients with very poor outcome (median survival, 9–12 months). Because their anatomic location, brainstem and thalamic gliomas are inoperable and hence no tissue is available to perform molecular analysis. Analysis of circulating biomarkers (liquid biopsies, including cfDNA) is a novel and promising, non-invasive alternative method for monitoring disease status that can tremendously facilitate biomarker screening for these pediatric patients. We have developed a robust cfDNA assay to detect genomic aberrations in plasma samples, despite genomic DNA contamination, which is a common issue in archival plasma samples. Our approach consisted of a modified droplet digital PCR (ddPCR) protocol that pre-selected the cfDNA molecules through the ligation of specific primers followed by PCR amplification. After pre-selection, cfDNA samples from 27 pediatric patients diagnosed with gliomas (17 diffuse intrinsic pontine glioma, 3 midline glioma, 7 high grade glioma) were screened using specific probes on a ddPCR platform to identify specific mutations in the H3F3A, PIK3CA and BRAF genes. Using this approach, we were able to identify low allele frequency mutations for H3F3A_K27M in 15 patients (62%; 11 with DIPG), PIK3CA_H1047R in 10 patients (37%), and BRAF_V600E with mutation allele frequencies ranging from 2% through 0.03% in 7 patients (26%). This very high sensitivity was achieved not only due to the nature of the ddPCR assay (1:20.000), but also due to the pre-selection of cfDNA molecules in our customized protocol. Since surgical resection is rarely an option for these patients, our non-invasive assay shows the feasibility of identifying these aberrations with a single blood sample. Additionally, because these mutations are associated with distinct prognosis and therapeutic implications, this cfDNA test can significantly impact clinical outcomes once implemented in a clinical lab to complement imaging for diagnosis and disease monitoring.