Authors: (including presenting author): :
Lam RSL(1), Shek CM(1), Li SN(1), Leung CM(1), Chu HT(1)
Affiliation: :
(1) Molecular Laboratory, Department of Clinical Pathology, Hong Kong East Cluster
Keyword 1: :
CDKN2A/B homozygous deletion
Keyword 6: :
Copy number analysis
Introduction: :
The WHO Classification of CNS Tumors (5th Edition) emphasizes CDKN2A/B homozygous deletion (HD) as a critical prognostic marker for glioma classification. While next-generation sequencing (NGS) provides comprehensive molecular profiling, we observed significant diagnostic discordance with the widely adopted Vysis CDKN2A/CEP 9 FISH Probe Kit (Abbott Laboratories, USA). The long-spanning Vysis probe frequently missed microdeletions in the CDKN2A gene detected by NGS copy number analysis (CNA), creating diagnostic uncertainty and necessitating costly confirmatory testing with delayed turnaround time (TAT).
Objectives: :
We optimized a short-spanning SureFISH probe assay (Agilent Technologies Inc., USA) and developed an in-house digital PCR (dPCR) assay with algorithms specifically designed for CDKN2A/B CNA to resolve NGS-FISH discordance. For cases requiring only CDKN2A/B status, this streamlined workflow shortened TAT and reduced costs.
Methodology: :
We analyzed NGS CNA data from 26 CNS tumor specimens with paired NGS and Vysis FISH results. CDKN2A/B deletion detection by Vysis FISH was further compared to the optimized short-probe FISH assay, with both validated against NGS. Statistical analyses included diagnostic accuracy, Cohen's kappa, and McNemar's test. We developed an in-house CDKN2A/B dPCR assay with custom algorithms for precise CNA, validated against NGS and FISH.
Result & Outcome: :
Vysis long-probe FISH showed suboptimal concordance with NGS: 73.1% overall agreement (19/26), 56.3% positive percent agreement (PPA), 100% negative percent agreement (NPA) and Cohen's kappa of 0.47 (moderate agreement). The long-probe missed 43.8% (7/16) of NGS-detected deletions with significant negative bias (McNemar's p = 0.0156). In contrast, the optimized short-probe FISH achieved 100% concordance (26/26) with perfect PPA and NPA, and Cohen's kappa of 1.00, detecting all microdeletions missed by the long probe. Our in-house dPCR assay similarly achieved perfect concordance with 100% agreement (26/26), perfect PPA and NPA, and Cohen's kappa of 1.00. Implementing short-probe FISH and dPCR eliminated unnecessary NGS testing for cases requiring only CDKN2A/B status determination, lowering per-case costs by 90% and reducing TAT from approximately one month to 3–4 days while maintaining 100% diagnostic accuracy. Conclusion: Integrating short-spanning FISH probe and in-house dPCR assay into our glioma diagnostic service maintained perfect diagnostic accuracy while significantly reducing costs and TAT for CDKN2A/B status determination.
