Kerr, Darcy A.

Context.— Detecting copy number variations (CNVs) at certain loci can aid in the diagnosis of histologically ambiguous melanocytic neoplasms. Droplet digital polymerase chain reaction (ddPCR) is a rapid, automated, and inexpensive method for CNV detection in other cancers, but not yet melanoma. Objective.— To evaluate the performance of a 4-gene ddPCR panel that simultaneously tests for ras responsive binding element protein 1 (RREB1) gain; cyclin-dependent kinase inhibitor 2A (CDKN2A) loss; MYC proto-oncogene, bHLH transcription factor (MYC) gain; and MYB proto-oncogene, transcription factor (MYB) loss in melanocytic neoplasms. Design.— One hundred sixty-four formalin-fixed, paraffin-embedded skin samples were used to develop the assay, of which 65 were used to evaluate its performance. Chromosomal microarray analysis (CMA) data were used as the gold standard. Results.— ddPCR demonstrated high concordance with CMA in detecting RREB1 gain (sensitivity, 86.7%; specificity, 88.9%), CDKN2A loss (sensitivity, 80%; specificity, 100%), MYC gain (sensitivity, 70%; specificity, 100%), and MYB loss (sensitivity, 71.4%; specificity, 100%). When one CNV was required to designate the test as positive, the 4-gene ddPCR panel distinguished nevi from melanomas with a sensitivity of 78.4% and a specificity of 71.4%. For reference, CMA had a sensitivity of 86.2% and a specificity of 78.6%. Our data also revealed interesting relationships with histology, namely (1) a positive correlation between RREB1 ddPCR copy number and degree of tumor progression; (2) a statistically significant correlation between MYC gain and nodular growth; and (3) a statistically significant correlation between MYB loss and a sheetlike pattern of growth. Conclusions.— With further validation, ddPCR may aid both in our understanding of melanomagenesis and in the diagnosis of challenging melanocytic neoplasms.

Giant cell tumor of bone (GCT) is a benign locally aggressive neoplasm composed of mononuclear cells admixed with innumerable osteoclast-type giant cells. H3F3A gene mutations producing mutant histone protein product H3.3 have been identified in 96% of GCT; mutant H3.3 is reliably demonstrated by immunohistochemistry. GCT may contain woven bone and rarely, neoplastic cartilage nodules which causes diagnostic challenges with aggressive neoplasms such as osteosarcoma. We describe the features of GCT with cartilage matrix and report the next-generation sequencing findings in a subset of tumors. Seventeen cases of GCT with cartilage matrix form the cohort: 7 males and 10 females, 13 to 55 (mean: 25) years old. Tumors involved the fibula (6), femur (6), and patella, tibia, humerus, S1, and scapula (1 case each). Tumors were radiolucent, circumscribed, lytic, and expansile. All contained classic GCT, foci of cartilage matrix, and trabeculae of woven bone. Immunohistochemistry showed diffuse staining for H3.3 in 9/9 cases and 1 case was positive for S100 and SOX9 in the cartilage areas. Next-generation sequencing showed a mutation in the H3F3A gene in 6/6 cases. On follow-up, 2 patients who underwent resection showed no disease after 12, and 7 months, respectively. Three patients had recurrences 10, 12, and 27 months after curettage; there were no metastases. GCT with cartilage matrix is uncommon. The cartilage matrix is associated with woven bone suggesting the neoplastic cells may differentiate into chondrocyte-like and osteoblast-like cells. Recognition of this neoplasm is important to prevent misdiagnosis and overtreatment of affected patients.