SOX10 and S100 in the Diagnosis of Soft Tissue Neoplasms.
Jason R Karamchandani, Torsten Nielsen, Matt Van de Rijn, Robert B West. Stanford University Medical Center, CA; University of British Columbia, Vancouver, Canada
Background: SOX10 is involved in neural crest differentiation, and consistently expressed in melanocytic and schwann cells (Am J Surg Pathol 2008;32:1291–1298). S100 is commonly employed by pathologists to determine if a tumor has neural or melanocytic differentiation, though S100 also stains numerous benign and malignant soft tissue tumors and tissues that are not thought to arise from neural crest tissue (cartilage, dendritic cells, etc.). We compared SOX10 and S100 protein expression in soft tissue tumors to determine which antibody is more specific in identifying soft tissue tumors of neural crest origin.
Design: We compared the results of SOX10 and S100 protein expression in 989 specimens, which included 71 types of neoplasm (predominantly soft tissue), and 33 normal tissues. The majority of the samples were analyzed on tissue microarrays. We also evaluated 20 full cross sections.
Results: Both SOX10 and S100 reliably stained schwannomas (50 cases), neurofibromas (55 cases), and staining was similar in cases of malignant peripheral nerve sheath tumors (78 cases). In 7 of 9 cases of desmoplastic melanoma, SOX10 showed a greater intensity of staining than S100. SOX10 also showed greater intensity of staining in 4/7 cases of clear cell sarcoma, 3 of which were S100 negative. We identified 5 non-neural, non-melanocytic sarcoma types in which a subset of cases showed strong S100 expression but where no SOX10 expression was seen: synovial sarcoma (12/79), Ewings sarcoma (3/14), rhabdomyosarcoma (4/17), chondrosarcoma (3/4), and extraskeletal myxoid chondrosarcoma (5/11). We also studied 550 cases representing 38 benign and malignant soft tissue neoplasms, including adipocytic, smooth mucle, skeletal muscle, vascular, and fibrohistiocytic tumors, as well as several reactive pathologies and tumors of uncertain differentiation. None of these samples showed SOX10 positivity. 20 of these cases showed spurious S100 staining.
Conclusions: SOX10 is a reliable marker of neural crest differentiation, and is consistently expressed in schwannian and melanocytic tumors. In our series, SOX10 proved superior to S100 in the detection of desmoplastic melanoma, and clear cell sarcoma. We did not see SOX10 expression in cases of synovial sarcoma, Ewings sarcoma, rhabdomyosarcoma, chondrosarcoma, or extra-skeletal myxoid chondrosaroma, all of which occasionally demonstrated strong S100 positivity. Compared to S100, SOX10 shows increased specificity for tumors of neural crest origin, and should be used in the place of S100 in soft tissue tumor diagnosis.
Category: Bone & Soft Tissue
Monday, February 28, 2011 9:30 AM
Poster Session I Stowell-Orbison/Surgical Pathology/Autopsy Awards Poster Session # 11, Monday Morning