Return to CAP Home
Printable Version

  Anatomic Abstracts





cap today

May 2001

Reliability and reproducibility of estrogen and progesterone receptor immunohistochemistry
Estrogen receptor and progesterone receptor immunohistochemistry are widely performed and, at least in the case of estrogen receptor, can play an important role in patient care. The degree of interlaboratory variation and the technical factors that affect this variability have not been widely investigated. The authors studied the reliability of the immunohistochemical assay for estrogen receptor and progesterone receptor using data from 105 laboratories participating in the United Kingdom External Quality Assessment Scheme for Immunohistochemistry during a two-year period. Sixty-six of these laboratories participated in a continuous fashion, but reliable assays were identified in only 24 (36 percent). "Reliable" was defined as appropriate staining in at least seven of the eight assessment opportunities. The most significant problem encountered was inadequate assay sensitivity due to weak staining. The authors identified poor microwave antigen retrieval as the major culprit and determined that the most likely explanation for poor antigen retrieval was inadequate heating. This contention was supported by the fact that those laboratories that used pressure cooker techniques for antigen retrieval, which achieve a higher temperature (115ºC or higher), performed better than those laboratories that used microwave technology (100ºC). Moreover, increasing the length of time for antigen retrieval for those using microwaves significantly improved results. Other factors, including the choice of primary antibody and detection system and automated versus manual techniques, appeared to be less problematic. The role of fixation, while considered to be important, was not addressed. The study focused solely on technical variables and did not address interpretation of the slides. It is noteworthy that this program helped those participating laboratories with unacceptable results to improve.

Rhodes A, Jasani B, Balaton AJ, et al. Study of interlaboratory reliability and reproducibility of estrogen and progesterone receptor assays in Europe. Documentation of poor reliability and identification of insufficient microwave antigen retrieval time as a major contributory element of unreliable assays. Am J Clin Pathol. 2001;115:44-58.

Reprints: Dr. A. Rhodes, UK NEQAS-ICC Manager, Dept. of Histopathology, University College London Medical School, Rockefeller Bldg., University St., London WC1E 6JJ, England

The border between borderline and malignant endometrioid tumors of the ovary
A spectrum exists amongst endometrioid tumors of the ovary that ranges from benign to malignant in a manner analogous to the benign, borderline, and malignant designations for the other ovarian epithelial neoplasms. Criteria to separate atypical proliferative (borderline) endometrioid tumors from well-differentiated endometrioid carcinoma are not widely agreed upon. The authors attempted to develop such criteria using 56 endometrioid tumors and focusing on such factors as invasion and cytologic atypia. Clearly benign and moderate to poorly differentiated malignant neoplasms, as well as adenosarcomas, were excluded from the study, as were cases with synchronous endometrial primaries. Thirty-three tumors were characterized as atypical proliferative, three as atypical proliferative with intraepithelial carcinoma, five as microinvasive carcinoma, and 15 as invasive carcinoma. All tumors were stage I at the time of evaluation. The authors discussed the growth pattern of the various tumors as well as the patterns of stromal invasion. Confluent glandular growth was thought to represent the most common pattern of stromal invasion and thus to be the best criteria for the diagnosis of carcinoma. Of the 21 patients with clinical followup, 20 were alive with no evidence of disease at a mean followup of 47 months. Those 20 included six with invasive carcinoma, two with microinvasion, one with intraepithelial carcinoma, and 11 with atypical proliferative tumors. One patient with carcinoma suffered a recurrence at 46 months but was alive 40 months after resection of her recurrent tumor. Considering the favorable prognosis of this group of stage I patients, the authors suggested it would be best to divide these tumors into two categories-atypical proliferative tumors and well-differentiated carcinoma. The authors noted that features such as cytologic atypia and intraepithelial and microinvasive carcinoma did not adversely effect prognosis and that the spectrum of tumors studied reflected various stages of endometrioid carcinognesis in the ovary.

Bell KA, Kurman RJ. A clinicopathologic analysis of atypical proliferative (borderline) tumors and well-differentiated endometrioid adenocarcinomas of the ovary. Am J Surg Pathol. 2000;24:1465-1479.

Reprints: Dr. Robert J. Kurman, Dept. of Pathology, The Johns Hopkins Hospital, 600 N. Wolfe St., Baltimore, MD 21287;

Distinguishing gastrointestinal stromal tumor from intra-abdominal fibromatosis of the bowel wall
Intra-abdominal fibromatosis may secondarily involve or arise from the intestinal wall and may mimic a gastrointestinal stromal tumor. Some features of IAF, such as large size, increased mitotic rate, and infiltration of adjacent structures, can cause it to be confused with gastrointestinal stromal tumor. IAF, a benign tumor, may be locally aggressive but does not metastasize. In contrast, gastrointestinal stromal tumors, or GIST, can have malignant potential. The authors conducted a study to evaluate the gross, histologic, immunophenotypic, ultrastructural, and clinical features of IAF with prominent involvement of the bowel wall (n=13) as well as GIST of the small intestine, colon, or mesentery (n=35), to highlight differences between these entities. The authors noted that an organoid cellular arrangement of tumor cells and pushing margins was a feature of GIST but was not seen in IAF. Other features that may be present in GIST but were not observed in IAF in this study included hemorrhage, necrosis and cystic change (gross findings), presence of epithelioid cells, and skenoid fibers. An extensively collagenous stroma with fine collagen fibrils between individual cells within fascicles and infiltrative margins characterized IAF. Immunohistochemical studies revealed the two lesions had overlapping features, with the exception of CD34 and S-100 staining, which was noted in some cases of GIST but not in IAF. The clinical outcome of GIST underscores the importance of separating these two entities, reported the authors, who noted that eight of 29 patients in this study who had followup died of disease.

Yantiss RK, Spiro IJ, Compton CC, et al. Gastrointestinal stromal tumors versus intra-abdominal fibromatosis of the bowel wall. A clinically important differential diagnosis. Am J Surg Pathol. 2000;24:947-957.

Reprints: Dr. Andrew E. Rosenberg, Dept. of Pathology, Massachusetts General Hospital, 55 Fruit St., Boston, MA 02114

Histological identification of H. pylori using various staining methods
Helicobacter pylori is a commonly sought pathogen in gastric biopsy material. Choosing an appropriate stain to help identify the organism may be probelmatic because one must balance sensitivity, specificity, reproducibility, ease of performance, and cost. This study compared four staining methods used on histological sections from 63 paired gastric biopsies. The stains were a modified McMullen's, silver stain (HpSS), modified Giemsa stain, and immunohistochemical stain for H. pylori antigen. Thirty cases were known to be positive via laboratory tests, such as the rapid biopsy urease test, urea breath test, culture, serology, and previous histology. Thirty cases were known to be negative, and three cases had discordant results using a combination of the tests. Interobserver agreement was 98 percent for the antibody test, 90 percent for McMullen's, 87 percent for Giemsa, and 85 percent for HpSS. The modified Giemsa stain identified organisms in all 30 of the known positive cases. After considering such factors as reproducibility, relative cost, and staining time, the Giemsa stain was determined to be the method of choice.

Rotimi O, Cairns A, Gray S, et al. Histological identification of Helicobacter pylori: comparison of staining methods. J Clin Pathol. 2000; 53:756-759.

Reprints: M. F. Dixon at