Use of immunohistochemical procedures in diagnosing angiosarcoma. Evaluation of 98 cases
Masahiko Ohsawa, Norifumi Naka, Yasuhiko Tomita, Dan Kawamori, Hiroyuki Kanno, Katsuyuki Aozasa
Abstract
Masahiko Ohsawa, Norifumi Naka, Yasuhiko Tomita, Dan Kawamori, Hiroyuki Kanno, Katsuyuki Aozasa
Abstract
BACKGROUND: Differential diagnosis of angiosarcoma, predominantly showing a non- or poorly vasoformative proliferation from other types of sarcomas, poorly differentiated carcinomas, and amelanotic melanoma, is often problematic. METHODS: The use of antibodies directed against Factor VIII-related antigen (FVIIIRA), Ulex europaeus lectin type 1 (UEA-1), CD31, and vascular endothelial growth factor (VEGF) in the diagnosis of angiosarcoma was examined in 98 cases of autopsy-proven angiosarcoma diagnosed during 1974-1990 in a survey of 178 Japanese hospitals. Reactivity of angiosarcoma cells for epithelial membrane antigen, cytokeratin, and melanoma cell antigen (HMB45) also was examined. RESULTS: Histologic specimens were formed exclusively by vasoformative areas in 32 cases and combined vasoformative and varying extents of non- or poorly vasoformative areas in another 66 cases. In vasoformative areas, the proliferating cells showed a diffuse positive reaction in the cytoplasm and/or cell surface for anti-FVII-IRA in 82 (84%) of 98 cases, for anti-CD31 in 78 (80%), and for UEA-1 in 69 (70%). In non- or poorly vasoformative areas, the positivity rate for FVIIIRA, CD31, and UEA-1 was 29%, 62%, and 46%, respectively. A positive reaction was found for either one of three endothelial markers in the non- or poorly vasoformative areas of 57 cases (86%). Epithelial membrane antigen and anticytokeratin antibody were positive in 4 and 11 cases, respectively, in the vasoformative areas and in 3 and 14 cases, respectively, in non- or poorly vasoformative areas with a simultaneous positive reaction for either one of three endothelial cell markers. None of the proliferating cells showed a positive reactivity for HMB45. The positivity rates of the angiosarcoma cells for each marker were different according to the primary tumor sites. The angiosarcoma cells in non- or poorly vasoformative areas showed the lowest positivity rate for anti-FVIIIRA in the heart (9%) and for anti-CD31 in the extremities (17%) and the highest positivity rate for anticytokeratin in the trunk (60%). Ulex europaeus lectin type 1 had almost the same reactivity rate (30-56%) in every organ. Angiosarcoma cells in 13 (36%) of 36 biopsy specimens and 8 (14%) of 56 autopsy specimens were positive for the anti-VEGF antibody. CONCLUSION: These findings suggest that the combined use of endothelial cell markers including FVIIIRA, UEA-1, and CD31 is useful in the diagnosis of angiosarcoma, especially in cases exclusively with a non- or poorly vasoformative pattern.
OpenAlex reports 182 citations for this work. Citation counts describe recorded attention and do not establish research quality.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
BACKGROUND: Differential diagnosis of angiosarcoma, predominantly showing a non- or poorly vasoformative proliferation from other types of sarcomas, poorly differentiated carcinomas, and amelanotic melanoma, is often problematic. METHODS: The use of antibodies directed against Factor VIII-related antigen (FVIIIRA), Ulex europaeus lectin type 1 (UEA-1), CD31, and vascular endothelial growth factor (VEGF) in the diagnosis of angiosarcoma was examined in 98 cases of autopsy-proven angiosarcoma diagnosed during 1974-1990 in a survey of 178 Japanese hospitals. Reactivity of angiosarcoma cells for epithelial membrane antigen, cytokeratin, and melanoma cell antigen (HMB45) also was examined. RESULTS: Histologic specimens were formed exclusively by vasoformative areas in 32 cases and combined vasoformative and varying extents of non- or poorly vasoformative areas in another 66 cases. In vasoformative areas, the proliferating cells showed a diffuse positive reaction in the cytoplasm and/or cell surface for anti-FVII-IRA in 82 (84%) of 98 cases, for anti-CD31 in 78 (80%), and for UEA-1 in 69 (70%). In non- or poorly vasoformative areas, the positivity rate for FVIIIRA, CD31, and UEA-1 was 29%, 62%, and 46%, respectively. A positive reaction was found for either one of three endothelial markers in the non- or poorly vasoformative areas of 57 cases (86%). Epithelial membrane antigen and anticytokeratin antibody were positive in 4 and 11 cases, respectively, in the vasoformative areas and in 3 and 14 cases, respectively, in non- or poorly vasoformative areas with a simultaneous positive reaction for either one of three endothelial cell markers. None of the proliferating cells showed a positive reactivity for HMB45. The positivity rates of the angiosarcoma cells for each marker were different according to the primary tumor sites. The angiosarcoma cells in non- or poorly vasoformative areas showed the lowest positivity rate for anti-FVIIIRA in the heart (9%) and for anti-CD31 in the extremities (17%) and the highest positivity rate for anticytokeratin in the trunk (60%). Ulex europaeus lectin type 1 had almost the same reactivity rate (30-56%) in every organ. Angiosarcoma cells in 13 (36%) of 36 biopsy specimens and 8 (14%) of 56 autopsy specimens were positive for the anti-VEGF antibody. CONCLUSION: These findings suggest that the combined use of endothelial cell markers including FVIIIRA, UEA-1, and CD31 is useful in the diagnosis of angiosarcoma, especially in cases exclusively with a non- or poorly vasoformative pattern.
Key concepts: Angiosarcoma, Medicine, CD31, Hemangiosarcoma, Pathology, Hemangioendothelioma, Cytokeratin, Immunohistochemistry