P17.01 * CHROMOSOME ARM 9P LOSS OF HETEROZYGOSITY IS A MARKER OF SHORTER SURVIVAL IN 1P/19Q CO-DELETED ANAPLASTIC OLIGODENDROGLIOMA. A POLA NETWORK STUDY

Abstract

Anaplastic or WHO grade III oligodendrogliomas (AO) are a heterogeneous subgroup of diffuse glial tumors in adults. Three major histomolecular subtypes with clinical relevance have been individualized: (i) 1p/19q co-deleted AO -80% of cases-, (ii) non 1p/19q co-deleted and IDH mutated AO -10% of cases- and (iii) non 1p/19q co-deleted and IDH wild-typeAO -10% of cases-. 1p/19q co-deleted AO, and to a lesser extent IDH mutated AO, have better prognosis and better response to treatment with a median survival of more than 10 years, as shown in two phase III international trials. We have recently shown that loss of heterozygosity (LOH) in chromosome arm 9p, with copy number loss -CL LOH- or without -Copy Neutral LOH- is a frequent event in AO. Interestingly, 9p CN LOH, inducing gene under-expression, has biological significance. In the present study, we address the prognostic value of 9p loss in 1p/19q co-deleted AO. In the present study, 228 AO exhibiting 1p/19q co-deletion were included. All tumors were centrally reviewed in the setting of the French national network for high-grade oligodendroglial tumors (POLA network) that recruits prospectively newly diagnosed anaplastic oligodendroglial tumors. Tumor DNA was analyzed using high-resolution single nucleotide polymorphism array analysis. In the entire series, 70/228 (1/3) harbored 9p loss including CN-LOH, homozygous loss or CL-LOH. 9p loss was associated with a worse prognosis in 1p19q co-deleted AO in univariate analysis (3-years OS of 98% vs. 74%, p < 0.001) whereas PFS at three years was not significantly different (3-years PFS of 88% vs. 66% p = 0.1). After adjustment for age, KPS, and treatment, multivariate analysis demonstrated 9p loss to be an independent prognostic factor for OS, p-value = 0.03, HR = 4.9 [1.2-16], but not for PFS. 1p19q co-deleted AO harboring 9p loss have shorter overall survival compared to their non 9p-lost counterparts in this prospective cohort. Further studies are needed to validate our findings.