DNA copy number changes correlate with clinical behavior in melanocytic neoplasms: proposal of an algorithmic approach.

ABSTRACT

Increasingly, molecular methods are being utilized in the workup of melanocytic neoplasms. To this end, we sought to correlate data from a single nucleotide polymorphism (SNP) array platform based on molecular inversion probes with clinical data. Copy number variation (CNV) data were obtained on 95 melanocytic tumors (6 ordinary nevi, 15 atypical nevi, 34 ambiguous neoplasms, and 40 melanomas) from 92 patients. The average number of significant CNVs was 0 for nevi, 0.6 for atypical nevi (range 0-3), 2.8 for ambiguous neoplasms (range 0-17), and 18.1 for melanomas (range 0-69). Clinical follow-up data were available in 57 of 95 lesions (56 of 92 patients). Tumors from patients with adverse events demonstrated an average number of CNVs of 24.5 (range 6-69) as compared with 7.9 (range 0-35) among tumors without an associated adverse event (p ≤ 0.001). No adverse events were observed in nevi including atypical nevi. Adverse events were found in 2 of 19 ambiguous neoplasms and 10 of 32 melanomas with follow up. In these two latter groups of neoplasms, the correlation between adverse events and the average number of CNVs remained statistically significant even when controlled for Breslow depth (21.5 versus 8.7, p value = 0.036). No neoplasm with adverse events had ≤3 CNVs. These results provide further evidence that SNP array testing for CNVs may be helpful in the classification and prognostication of ambiguous neoplasms. Based on these results, an algorithmic approach to challenging melanocytic neoplasms using CNV data is suggested, using as cutoff of >3 CNVs with some caveats, as the threshold for a positive result. Future clinical validation, using a larger cohort of relevant tumors, will be necessary.