Role of CDKN2A deletion in grade 2/3 IDH-mutant astrocytomas: need for selective approach in resource-constrained settings

Abstract

OBJECTIVE The authors aimed to assess the frequency of homozygous CDKN2A deletion in isocitrate dehydrogenase (IDH)–mutant diffuse astrocytomas (grade 2/3) and to narrow down the clinicopathological indications in which the CDKN2A fluorescence in situ hybridization (FISH) assay is cost-effective in resource-constrained settings. METHODS IDH-mutant astrocytomas were analyzed for ATRX, p53, MIB1-LI, and p16 expression using immunohistochemistry. The FISH assay was used to evaluate CDKN2A deletion and 1p/19q codeletion. Survival outcomes were assessed according to the different molecular markers. RESULTS A total of 150 adult patients with IDH-mutant grade 2 (n = 95) and grade 3 (n = 55) astrocytomas (145 primary and 5 recurrent) were analyzed. Using a cutoff value of 30% for defining significant homozygous CDKN2A deletion, none of the grade 2 and 10.9% (6/55) of grade 3 astrocytomas showed this deletion (4 primary and 2 recurrent grade 3 tumors) and were reclassified as grade 4. This mutation was more frequent in recurrent (40%, 2/5) than primary (2.76%, 4/145) gliomas. Half (3/6, 50%) of the CDKN2A-deleted cases demonstrated poor outcomes; 2 of these cases experienced recurrence at 12 and 36 months after surgery, and 1 died at 5 months. The majority of CDKN2A-deleted cases showed marked cellularity (100%), pleomorphism (100%), brisk mitosis (83.3%), and tumor giant cell formation (83.4%). None of the cases with retained p16 expression harbored this deletion. Both overall survival (p = 0.039) and progression-free survival (p = 0.0045) were found to be worse in cases with p16 loss. Selectively performing CDKN2A FISH only in high-risk cases with histomorphological features of anaplasia, p16 loss, or recurrent tumors achieved a sensitivity and negative predictive value of 100%. This approach would have resulted in saving 41.1% of the original expenditure ($6900 US per 150 samples) and 27.6 person-minutes per sample without compromising the identification of deleted cases. CONCLUSIONS Homozygous CDKN2A deletion is conspicuously absent in grade 2 and rare in primary grade 3 IDH-mutant astrocytomas. The authors propose that restricting use of the FISH assay to cases showing histomorphological features of anaplasia, p16 loss, or recurrent tumors will help this platform to be utilized in the most cost-effective manner in resource-constrained settings.