Dedifferentiated melanoma (DM) and undifferentiated melanoma (UM) is defined as a primary or metastatic melanoma showing transition between conventional and undifferentiated components (DM) or lacking histologic and immunophenotypic features of melanoma altogether (UM). The latter is impossible to verify as melanoma by conventional diagnostic tools alone. We herein describe our experience with 35 unpublished cases to expand on their morphologic, phenotypic, and genotypic spectrum, along with a review of 50 previously reported cases (total: 85) to establish the diagnostic criteria. By definition, the dedifferentiated/undifferentiated component lacked expression of 5 routinely used melanoma markers (S100, SOX10, Melan-A, HMB45, Pan-melanoma). Initial diagnoses (known in 66 cases) were undifferentiated/unclassified pleomorphic sarcoma (n=30), unclassified epithelioid malignancy (n=7), pleomorphic rhabdomyosarcoma (n=5), other specific sarcoma types (n=6), poorly differentiated carcinoma (n=2), collision tumor (n=2), atypical fibroxanthoma (n=2), and reactive osteochondromatous lesion (n=1). In only 11 cases (16.6%) was a diagnosis of melanoma considered. Three main categories were identified: The largest group (n=56) comprised patients with a history of verified previous melanoma who presented with metastatic DM or UM. Axillary or inguinal lymph nodes, soft tissue, bone, and lung were mainly affected. A melanoma-compatible mutation was detected in 35 of 48 (73%) evaluable cases: BRAF (n=20; 40.8%), and NRAS (n=15; 30.6%). The second group (n=15) had clinicopathologic features similar to group 1, but a melanoma history was lacking. Axillary lymph nodes (n=6) was the major site in this group followed by the lung, soft tissue, and multiple site involvement. For this group, NRAS mutation was much more frequent (n=9; 60%) than BRAF (n=3; 20%) and NF1 (n=1; 6.6%). The third category (n=14) comprised primary DM (12) or UM (2). A melanoma-compatible mutation was detected in only 7 cases: BRAF (n=2), NF1 (n=2), NRAS (n=2), and KIT exon 11 (n=1). This extended follow-up study highlights the high phenotypic plasticity of DM/UM and indicates significant underrecognition of this aggressive disease among general surgical pathologists. The major clues to the diagnosis of DM and UM are: (1) presence of minimal differentiated clone in DM, (2) earlier history of melanoma, (3) undifferentiated histology that does not fit any defined entity, (4) locations at sites that are unusual for undifferentiated/unclassified pleomorphic sarcoma (axilla, inguinal, neck, digestive system, etc.), (5) unusual multifocal disease typical of melanoma spread, (6) detection of a melanoma-compatible gene mutation, and (7) absence of another genuine primary (eg, anaplastic carcinoma) in other organs.

Objective: Trefoil factor family peptide 3 (TFF3) has been shown to support catabolic functions in cases of osteoarthritis (OA). As in joint physiology and diseases such as OA, the synovial membrane (SM) of the joint capsule also plays a central role. We analyze the ability of SM to produce TFF compare healthy SM and its secretion product synovial fluid (SF) with SM and SF from patients suffering from OA or rheumatoid arthritis (RA). Methods: Real-time PCR and ELISA were used to measure the expression of TFFs in healthy SM and SM from patients suffering from OA or RA. For tissue localization, we investigated TFF1-3 in differently aged human SM of healthy donors by means of immunohistochemistry, real-time PCR and Western blot. Results: Only TFF3 but not TFF1 and -2 was expressed in SM from healthy donors as well as cases of OA or RA on protein and mRNA level. In contrast, all three TFFs were detected in all samples of SF on the protein level. No significant changes were observed for TFF1 at all. TFF2 was significantly upregulated in RA samples in comparison to OA samples. TFF3 protein was significantly downregulated in OA samples in comparison to healthy samples and cases of RA significantly upregulated compared to OA. In contrast, in SM TFF3 protein was not significantly regulated. Conclusion: The data demonstrate the production of TFF3 in SM. Unexpectedly, SF contains all three known TFF peptides. As neither articular cartilage nor SM produce TFF1 and TFF2, we speculate that these originate with high probability from blood serum.

No abstract