BACKGROUND: Dermatofibrosarcoma protuberans (DFSP) typically affects the dermis and subcutaneous tissue. The subcutaneous variant is rare. OBJECTIVE: We sought to characterize the subcutaneous DFSP (SC-DFSP) variant and compare it with cutaneous DFSP (C-DFSP). METHODS: This work was a retrospective study of DFSP treated in our institution. RESULTS: Of 124 cases of DFSP, 18 were SC-DFSP (14.5%). Except for the deep location, the pathologic and genetic features were indistinguishable from the C-DFSP variant. Histologically, of 18 SC-DFSP cases, 13 were classic DFSP, 3 fibrosarcomatous DFSP (FS-DFSP), 1 Bednar tumor, and 1 giant-cell fibroblastoma. All tumors expressed CD34 and the COL1A1-PDGFB fusion transcripts. In our series, higher proportions of SC-DFSP tumors (61%) than C-DFSP tumors (8.5%) were located on the head (P < .001). Of the 20 DFSP tumors on the head (16.1%), 11 were SC-DFSP and 9 were C-DFSP. In addition, half the SC-DFSP tumors affected muscle or periosteum, compared with a quarter of the C-DFSP tumors (P = .009). SC-DFSP needed a higher number of Mohs stages than did C-DFSP (P = .009). Median follow-up time was 63 months, and 2 FS-DFSP tumors recurred (1 SC-DFSP, 1 C-DFSP). LIMITATIONS: Limitations include the retrospective aspect of the study. CONCLUSIONS: Most DFSP tumors involving the head were subcutaneous and required more complex surgery. Dermatologists should be aware of this atypical presentation, especially in lesions involving the head.
BACKGROUND:Dermatofibrosarcoma protuberans (DFSP) typically affects the dermis and subcutaneous tissue. The subcutaneous variant is rare. OBJECTIVE: We sought to characterize the subcutaneous DFSP (SC-DFSP) variant and compare it with cutaneous DFSP (C-DFSP). METHODS: This work was a retrospective study of DFSP treated in our institution. RESULTS: Of 124 cases of DFSP, 18 were SC-DFSP (14.5%). Except for the deep location, the pathologic and genetic features were indistinguishable from the C-DFSP variant. Histologically, of 18 SC-DFSP cases, 13 were classic DFSP, 3 fibrosarcomatous DFSP (FS-DFSP), 1 Bednar tumor, and 1 giant-cell fibroblastoma. All tumors expressed CD34 and the COL1A1-PDGFB fusion transcripts. In our series, higher proportions of SC-DFSP tumors (61%) than C-DFSP tumors (8.5%) were located on the head (P < .001). Of the 20 DFSP tumors on the head (16.1%), 11 were SC-DFSP and 9 were C-DFSP. In addition, half the SC-DFSP tumors affected muscle or periosteum, compared with a quarter of the C-DFSP tumors (P = .009). SC-DFSP needed a higher number of Mohs stages than did C-DFSP (P = .009). Median follow-up time was 63 months, and 2 FS-DFSP tumors recurred (1 SC-DFSP, 1 C-DFSP). LIMITATIONS: Limitations include the retrospective aspect of the study. CONCLUSIONS: Most DFSP tumors involving the head were subcutaneous and required more complex surgery. Dermatologists should be aware of this atypical presentation, especially in lesions involving the head.