CD30 expression helps to define anaplastic large-cell lymphoma, and it
is present in embryonal carcinoma, suggesting utility in defining emb
ryonal carcinoma as well. But the specificity of CD30 in nonhematopoie
tic tumors has not been completely characterized. To do this, we stain
ed 91 formalin-fixed, paraffin-embedded tumors for CD30 (Ber-H2). In a
ddition, six embryonal carcinomas were stained. Standard immunoperoxid
ase techniques with heat-induced epitope retrieval (HIER) and blocking
of endogenous avidin-binding activity were used. Tumor cells with sta
ining above background (1+ or greater) in a membranous pattern were ju
dged positive. Six (100%) of six embryonal carcinomas had strong CD30
expression in a unique membranous (''chicken-wire'') pattern. No stain
ing was identified in any of the other tumors studied: squamous cell c
arcinoma, 0/5; breast carcinoma, 0/5; gastric carcinoma, 0/6; pan crea
tic carcinoma. 0/5; colon carcinoma, 0/5; kidney carcinoma. 0/4; liver
carcinoma, 0/5; cholangiocarcinoma, 0/3; lung carcinoma, 0/5: urinary
bladder carcinoma, 0/5; ovarian carcinoma, 0/1; prostate carcinoma, 0
/5; endometrial carcinoma, 0/5; uterine cervical carcinoma, 0/1; papil
lary thyroid carcinoma, 0/5; follicular thyroid adenoma, 0/5; Leydig-c
ell tumor, Oil; malignant melanoma, 0/2; carcinoid, 0/6; mesothelioma,
0/3; fibrosarcoma, 0/3; synovial sarcoma, 0/1; medulloblastoma, 0/4;
and pinealblastoma, Oil. Positive cytoplasmic staining in normal cells
was identified only in plasma cells and occasional ''activated'' lymp
hoid cells in adjacent lymphoid tissues. Nonspecific granular cytoplas
mic staining was identified in many tumors and normal tissues before b
locking for endogenous avidin-binding activity. In summary, in this se
ries of 97 tumors, the membranous (chicken-wire) pattern of staining a
ppears specific and limited to embryonal carcinoma.