Low-Level MDM2 Amplification by FISH: An Institutional Experience With a Diagnostic Dilemma

Background: MD M2 (murine double minute-2) amplification via fluorescence in-situ hybridization (FISH) is the gold standard test used for confirming the diagnosis of atypical lipomatous tumor/well differentiated liposarcoma and dedifferentiated liposarcoma. It is also used as a screening test in high grade spindle cell or pleomorphic neoplasms. MDM2 FISH is considered positive for amplification when the MDM2/CEP12 ratio is greater than 2; however, a ratio between 2 and 3 is considered a "low-level" amplification and raises the possibility of a false positive result; thus, posing a diagnostic dilemma. Another molecular modality, next generation sequencing (NGS) assay, can help in such situations in confirming or excluding the M D M 2 amplification status. Confronted by a number of neoplastic specimens at our institution with "low-level" MDM2 amplification via FISH, we aimed to assess the specificity of fluorescence in situ hybridization (FISH) in such tumors by comparing their NGS assay results and determine an accurate MDM2 amplification status that further aids in diagnosis. Methods: Tumors with "low-level" MDM2 amplification via FISH, defined as MDM2/CEP12 ratio between 2 and 3, and harboring a high grade and/or pleomorphic morphology were retrospectively retrieved from our institutional archives from the last five years. The retrieved specimens were evaluated for concordant retrospective Oncomine v3 analysis. Oncomine v3 is an institutional NGS assay that covers 161 genes and assesses for DNA mutations, RNA fusions, and copy number alterations including MDM2 gene gain or amplification. The tumors with Oncomine v3 results were compared and the FISH specificity was calculated. Results: Twenty-seven high grade and/or pleomorphic tumors with "low-level" MDM2 amplification were retrieved. Eight out of twenty-seven tumors had Oncomine v3 performed on them. The tumors correlated to neoplasms from different lineage types including undifferentiated melanoma, sarcomatoid squamous cell carcinoma, leiomyosarcoma, myxofibrosarcoma, undifferentiated pleomorphic sarcoma, and high-grade poorly differentiated pleomorphic neoplasm. All 8 tumors had a low-level MD M2 amplification ratio ranging between 2.09 and 2.84. Seven out of eight had no MD M2 copy number alteration. One only (leiomyosarcoma) demonstrated MD M2 copy number gain (∼5 copies) that did not qualify as amplification due to the "6 copy number" gain cutoff. TP53, CDKN2A/B, PIKC3, and PTEN alterations were the most common genetic aberrations detected by Oncomine v3. Conclusion: We demonstrated the absence of MDM2 amplification via Oncomine in all our 8 "low-level" MDM2 FISH amplification specimens confirming the FISH results as "false positive" with a corresponding FISH specificity rate of 0%. Laboratory measures and utilizing NGS assay when needed, could be implemented when encountering such problematic "low-level" MDM2 amplification specimens to avoid misdiagnosis and misuse of targeted therapy. Future studies are needed to better characterize and investigate such findings.