Design and Validation of a Gene-Targeted, Next-Generation Sequencing Panel for Routine Diagnosis in Gliomas

Abstract

The updated 2016 World Health Organization (WHO) classification system for gliomas integrates molecular alterations and histology to provide a greater diagnostic and prognostic utility than the previous, histology-based classification. The increasing number of markers that are tested in a correct diagnostic procedure makes gene-targeted, next-generation sequencing (NGS) a powerful tool in routine pathology practice. We designed a 14-gene NGS panel specifically aimed at the diagnosis of glioma, which allows simultaneous detection of mutations and copy number variations, including the 1p/19q-codeletion and Epidermal Growth Factor Receptor (EGFR) amplification. To validate this panel, we used reference mutated DNAs, nontumor and non-glioma samples, and 52 glioma samples that were previously characterized. The panel was then prospectively applied to 91 brain lesions. A specificity of 100% and sensitivity of 99.4% was achieved for mutation detection. Orthogonal methods, such as in situ hybridization and immunohistochemical techniques, were used for validation, which showed high concordance. The molecular alterations that were identified allowed diagnosis according to the updated WHO criteria, and helped in the differential diagnosis of difficult cases. This NGS panel is an accurate and sensitive method, which could replace multiple tests for the same sample. Moreover, it is a rapid and cost-effective approach that can be easily implemented in the routine diagnosis of gliomas.

Keywords: 1p/19q codeletion; glioma; molecular pathology; next-generation sequencing.

Figure 1

Analysis of the 1p/19q codeletion through the detection of single-nucleotide polymorphism (SNP)-based loss of heterozygosity (LOH), using the designed next-generation sequencing panel. (a) Representative 1p/19q non-codeleted sample. (b) Representative 1p/19q codeleted sample. For each sample and chromosome, the allelic frequency is represented according to the SNP location. The established ranges for homozygosity were defined between 0% and 5% (same allele as that of the reference genome) and between 90% and 100% (different allele than that of the reference genome). The range of heterozygosity was defined as being between 40% and 60% of allelic frequency (AF). The LOH for an SNP was defined as when the AF was outside of the established ranges for homozygosity or heterozygosity (gray areas). A sample was scored as having 1p/19q LOH when all 1p and 19q markers had LOH or were homozygous, and at least one marker in the opposite arm (1q and 19p) were heterozygous.

Figure 2

Genomic alterations identified in the whole series. Samples (columns) are grouped by 2016 World Health Organization (WHO) criteria. The WHO grade and somatic mutations are indicated by black squares.