RBC autoimmunization and the development of autoimmune hemolytic anemia should be recognized as a complication of allogeneic blood transfusion. The need for additional blood transfusion was successfully avoided in one patient by treatment with recombinant human EPO and corticosteroid therapy. Once RBC autoimmunization is identified, subsequent management should incorporate a strategy that minimizes subsequent exposure to allogeneic blood.

WT1 diffusely stains most ovarian serous carcinomas; reactivity of uterine papillary serous carcinomas has not been evaluated. We studied WT1 expression in 13 International Federation of Gynecology and Obstetrics stage 1 and 5 stage 3 or 4 uterine papillary serous carcinomas without ovarian metastases and compared their reactivity with the WT1 staining of 30 ovarian serous carcinomas. WT1 reactivity was evaluated with the C19 and 6F-H2 antibody clones. All 18 uterine papillary serous carcinomas were nonreactive for WT1. The nonovarian metastases of the 5 high-stage uterine papillary serous carcinomas also were nonreactive for WT1. In contrast, 29 (97%) of 30 ovarian serous carcinomas were reactive for WT1. WT1 reactivity in an unknown primary serous carcinoma would suggest it is from a nonuterine site. The mechanisms underlying these findings are unknown. They raise the possibility of genetic differences between the 2 morphologically similar neoplasms.

We investigated whether a panel of antibodies including WT1 could separate pancreaticobiliary and ovarian carcinomas by staining 64 pancreaticobiliary adenocarcinomas, 41 ovarian serous carcinomas, and 12 primary ovarian mucinous neoplasms with WT1, cytokeratin (CK) 17, CK20, carcinoembryonic antigen (CEA), and CA-125. Moderate or strong intensity reactivity in more than 25% of cells was a positive result. Of the ovarian serous carcinomas, 38 (93%) were WT1 reactive and 22 (54%) WT1 positive, 9 (22%) had CK20 reactivity, and 3 (7%) were CK20 positive in fewer than 50% of cells. All were CK17 or CEA nonreactive. Of the ovarian mucinous neoplasms, all were WT1 and CK17 nonreactive and 11 (92%) were CEA reactive, 8 (67%) CEA positive, 10 (83%) CK20 reactive, and 6 (50%) CK20 positive. Of the pancreaticobiliary adenocarcinomas, 19 (30%) were CK20 positive, 27 (42%) CK17 positive, and 52 (81%) CEA positive. All were WT1 nonreactive. A panel including WT1, CK17, CK20, and CEA is useful to distinguish pancreaticobiliary and ovarian serous carcinomas. Extensive CK17 reactivity is supportive of a pancreaticobiliary adenocarcinoma when the differential diagnosis includes ovarian mucinous neoplasm. None of the antibodies positively identified ovarian mucinous neoplasms.

No abstract

Kaposi sarcoma (KS) is a low-grade vascular neoplasm associated with human herpesvirus-8 (HHV-8) infection. Clinically, lesions commence as blue-red macules that may develop into plaques and eventually into nodules. The histologic appearance spans a broad spectrum and varies with the stage of the lesion. At each stage, KS has significant morphologic overlap with other vasoproliferative lesions. Recently, we encountered 6 KS tumors that histologically mimicked pyogenic granuloma (PG), a common benign vascular tumor of the skin that usually does not figure in the histologic differential diagnosis of KS. We stained 6 PG-like KS and 28 PGs with a mouse monoclonal antibody (13B10) against HHV-8 latent nuclear antigen-1 (LNA-1) to determine the utility of immunoperoxidase staining in distinguishing KS from PG. All 6 PG-like KS demonstrated nuclear staining for HHV-8 LNA-1. No staining was identified in any of the 28 PGs. Histologic criteria often used to differentiate between these two entities were not helpful in our cases. The only distinguishing feature was the presence or absence of HHV-8 LNA-1 staining. The presence of HHV-8 LNA-1 nuclear staining seems to be a specific marker for KS when comparing PGs and PG-like KS. Immunoperoxidase staining for HHV-8 LNA-1 is a useful diagnostic tool in this setting.