Variant allele fraction or copy-neutral loss of heterozygosity? A comparison of testing platforms in the classification of myeloid neoplasia.

Myelodysplastic syndrome (MDS) and myelodysplastic/myeloproliferative neoplasm (MDS/MPN) classification requires integration of mutational and cytogenetic data. MDS with biallelic TP53 inactivation (biTP53) is an established category, and MDS and MDS/MPN with biallelic TET2 (biTET2) inactivation is an emerging one. Biallelic genetic inactivation is established by both a mutation and copy loss or copy-neutral loss of heterozygosity (CNLOH) of the other allele, or it can be inferred by identifying multiple or single mutations at a sufficiently high variant allele fraction (VAF). The purpose of this study was to determine whether CNLOH genotyping data is needed to assign patients to biTP53 or biTET2 categories. We studied 157 patients with MDS or MDS/MPN who had sequencing, karyotype, and microarray performed at our institution and assigned patients to biTP53 and biTET2 categories using thresholds established in prior studies. We identified 24 biTP53 and 27 biTET2 patients. In the biTP53 group, 8 patients had > 1 mutation, 9 had a single mutation with 17p loss identified by karyotype and microarray, 2 had a single mutation with 17p loss identified only by microarray, and 3 had a single mutation and 17p CNLOH. All patients with 17p CNLOH had TP53 mutant VAF > 55%. In the biTET2 group, 24 patients had > 1 TET2 mutation with VAFs summing to > 50% and 3 had 4q CNLOH. All patients with 4q CNLOH had TET2 mutant VAF > 50%. In this cohort, CNLOH in 17p and 4q by microarray did not provide information in addition to that provided by inferring the allelic status of TP53 and TET2 using the VAF. This supports that platforms such as optical genome mapping that do not readily detect CNLOH would, in conjunction with sequencing, be adequate to identify MDS and MDS/MPN patients in the biTP53 and biTET2 categories in the great majority of cases.