p16 is overexpressed in uterine leiomyosarcomas compared with leiomyomas, benign leiomyoma variants and STUMPs, suggesting that p16 may be implicated in the pathogenesis of malignant uterine smooth muscle neoplasms. p16, in combination with p53 and MIB1, may be of value as an adjunct to morphological examination in the assessment of problematic uterine smooth muscle tumours, although further large-scale studies with follow-up are necessary to confirm this.

Ovarian small cell carcinoma of hypercalcemic type (OSCCHT) is a rare neoplasm with an aggressive behavior, broad differential diagnosis, and unknown histogenesis. To add to knowledge concerning the possible aid of immunohistochemistry in resolving problems in differential diagnosis and to further explore whether that modality points to any specific histogenesis, we undertook an immunohistochemical study of this neoplasm. Fifteen OSCCHTs (including four of the ''large cell" variant) were stained with a range of antibodies, some of which have not been investigated previously in this neoplasm. Cases were stained with AE1/3, EMA, BerEP4, CK5/6, calretinin, WT1, chromogranin, CD56, synaptophysin, CD99, NB84, desmin, S100, CD10, alpha inhibin, TTFI, and p53. Staining was classified as 0 (negative), 1+ (<5% cells positive), 2+ (5% to 25% cells positive), 3+ (26% to 50% cells positive), or 4+ (>50% cells positive). All cases were positive with p53 (two 1+, five 3+, eight 4+), 14 of 15 cases were positive with WT1 (one 1+, thirteen 4+), 14 of 15 with CD10 (three 1+, four 2+, two 3+, five 4+), 13 of 15 with EMA (three 1+, three 2+, two 3+, five 4+), 11 of 15 with calretinin (nine 1+, one 3+, one 4+), 9 of 15 with AE1/3 (eight 1+, one 2+), 4 of 15 with CD56 (one 1+, two 2+, one 4+), 3 of 15 with BerEP4 (two 2+, one 4+), 2 of 15 with synaptophysin (two 1+), and 1 of 15 with S100 (4+). All cases were negative with CK5/6, chromogranin, CD99, NB84, desmin, alpha inhibin, and TTF1. The only noticeable difference in the immunophenotype between typical OSCCHT and the large cell variant was that there was 4 +EMA positivity in three of four cases of large cell variant compared with two of 11 cases of typical OSCCHT. OSCCHT is characteristically positive with AE1/3, EMA, CD10, calretinin, WT1, and p53. Combined EMA and WT1 positivity, the latter usually intense and diffuse, may be of diagnostic value, inasmuch as only a few of the neoplasms in the differential diagnosis are positive with both antibodies. Negative staining with CD99, desmin, NB84, alpha-inhibin, and TTF1 may aid in the cases in which primitive neuroectodermal tumor, rhabdomyosarcoma, intraabdominal desmoplastic small round cell tumor, neuroblastoma, a sex cord-stromal tumor, and metastatic pulmonary small cell carcinoma are in the differential. Calretinin positivity precludes its use in the differential with granulosa cell tumors. The results of this investigation do not settle the issue of histogenesis, which remains enigmatic. The typical age distribution, follicle formation, and calretinin positivity are consistent with a sex cord origin. On the other hand, WT1 and EMA positivity and negative staining with alpha-inhibin would be unusual in a sex cord-stromal neoplasm and can be used as an argument for a surface epithelial origin. Germ cell and neuroendocrine origins seem highly unlikely.

The histological distinction between a primary endometrial and a primary endocervical adenocarcinoma is often difficult, especially in small biopsy specimens. A preoperative distinction is important because primary surgical management differs between the two tumors. Cases of primary endometrioid endometrial (n=30) and primary endocervical (n=26) adenocarcinoma of endocervical type were stained immunohistochemically with the monoclonal antibodies against carcinoembryonic antigen (CEA), vimentin, estrogen receptor (ER), and 34 beta E12. In all cases the origin of the adenocarcinoma was confirmed by examination of the definitive pathology specimen. There was diffuse positive nuclear staining for ER in 28 of 30 (93%) endometrial adenocarcinomas. ER was negative in 16 of 26 endocervical adenocarcinomas, and there was focal weak nuclear staining in the other cases. Vimentin was positive in 29 of 30 (97%) endometrial adenocarcinomas but in only 2 of 26 (8%) endocervical adenocarcinomas. CEA was positive in 25 of 26 (96%) endocervical adenocarcinomas, mostly with diffuse membranous and cytoplasmic staining. Positivity with CEA was present in 21 of 30 (70%) endometrial adenocarcinomas but was largely confined to squamoid areas with only 12 tumors exhibiting focal membranous staining of the glandular component. 34 beta E12 was diffusely positive in all except one cervical adenocarcinoma. In endometrial carcinomas, positivity was strongest in squamoid areas but there was positive staining, either focally or diffusely, of the glandular component in 27 cases. In summary, primary endometrioid endometrial adenocarcinomas are characterized by diffuse, strong, positive staining for vimentin and ER and negative or very focal, positive staining of the glandular component for CEA. In contrast, primary endocervical adenocarcinomas are characterized by CEA positivity, which is usually but not always diffuse, negativity for vimentin, and negativity or focal weak positivity for ER. 34 beta E12 is of no value in the distinction between endometrial and endocervical adenocarcinomas. A panel of immunohistochemical stains, comprising CEA, vimentin, and ER, generally allows confident preoperative distinction between a primary endometrial and endocervical adenocarcinoma.