Library

Notes: Abstract: Standard practice in surgical pathology dictates that random sections from the four quadrants of the breast be taken in mastectomy specimens. These sections are obtained in addition to sampling of any grossly visible lesions within the breast specimen. While tradition dictates the submission of these sections, we are unaware of any study supporting their efficacy. We have investigated the utility and significance of these random sections in a series of 78 mastectomy specimens. This retrospective study identified mastectomy specimens from pathology files of Magee Woman's Hospital, the University of Pittsburgh, and the University of Utah School of Medicine between 1997 and 2000. Clinical data (palpable versus nonpalpable), radiographic features (mammographic diagnosis, presence of mass density and/or calcification), and pathologic features (size, histopathologic type, etc.) were studied. The histologic sections of the cases were reviewed and the random sections were specifically studied for pathologic findings. Diagnosis and clinically significant features obtained from examining these random sections, but not demonstrable in grossly selected sections, were tabulated. A total of 78 mastectomy specimens were analyzed. Diagnoses rendered were infiltrating ductal carcinoma (23), infiltrating ductal carcinoma with ductal carcinoma in situ (DCIS) (16), DCIS (25), infiltrating lobular carcinoma (4), biopsy cavity with no residual malignancy (4), infiltrating lobular carcinoma with lobular carcinoma in situ (3), invasive ductal and lobular carcinoma (1), adenoid cystic carcinoma (1), and atypical ductal hyperplasia (1). The number of random sections ranged from 2 to 17 (mean 9). Random sections provided additional information in 21 of 78 mastectomies (27%). The multifocal/multicentric nature of the lesion was diagnosed in 20 cases: DCIS (6), lobular carcinoma in situ and invasive (2), invasive ductal carcinoma (6), invasive and in situ ductal carcinoma (5), invasive lobular carcinoma (1), invasive ductal and lobular carcinoma (1). Additional findings include lymphovascular invasion (2 cases), atypical ductal hyperplasia (1), DCIS at the operative margin (1), DCIS within less than 1 mm of an operative margin (1), and atypical lobular hyperplasia (1). In the remaining 57 cases, random sections did not provide any additional information. Histologic examination of random sections from breast quadrants yielded important information about the presence of multifocality, multicentricity, vascular invasion, and margin involvement by carcinoma in only a minority of cases, many of which had a lobular morphology. 

Notes: Abstract: The assessment of steroid hormone receptors in resected breast carcinoma tissue is currently the standard of practice. The traditional method for assessment of receptor status is the ligand binding assay. More recently, immunohistochemistry (IHC) has become a popular method for such testing. Despite the widespread use of IHC and the availability of many antibodies, standardization of quantitative IHC for assessment of estrogen and progesterone receptors has not been achieved. While the College of American Pathologists (CAP) offers a Quality Assurance (QA) program for IHC quantitation of estrogen receptor (ER) and progesterone receptor (PgR), no universal standard is currently recognized in assessment of ER and PgR by IHC. We surveyed 300 laboratories within the United States for their current practices regarding the assessment of ER and PgR status in breast cancer tissue specimens. Eighty usable responses were received. Forty-nine (61%) laboratories performed the assay in-house, while the remainder sent the material out for assay. All responding laboratories performing their steroid receptor analysis in-house used the IHC technique. Forty-three (80%) laboratories answering the question on material accepted for analysis performed the assay only on paraffin-embedded material, three (6%) used either paraffin block or frozen material, and two (4%) used only frozen material. Eighty-eight percent of laboratories performing steroid receptor analysis in-house used a manual quantitation technique. Four (8%) used computer-assisted image analysis, and a single laboratory used laser scanning cytometry. Eight different antibodies were used among the 44 laboratories documenting the antibody supplier, and for any given commercially prepared antibody a wide variety of dilutions were used, with the exception of the standard solution used with the Ventana antibody. Of the laboratories using manual estimation techniques, 61% simply estimated the percentage of positive cells, 29% evaluated both the intensity of staining and percentage of nuclei staining, 6% used formal H-score analysis, 2% evaluated only intensity of nuclear staining, and 2% mainly counted the percentage of nuclei staining for ER but used a formal H score in the assessment of PgR. Cutoff points for the separation of positive and negative results varied widely, with some laboratories assessing any demonstrable positivity as a positive result, while others required as many as 19% of the nuclei to stain before a specimen was declared positive. Standardization techniques differed considerably among laboratories. Eighty-six percent used the CAP program for QA. While all laboratories utilized some form of intralaboratory control for assessment of ER and PgR, the nature of that control varied from laboratory to laboratory. Our survey indicates that a majority of laboratories perform their steroid hormone receptor analysis in-house using IHC. There is considerable variability in the antibodies utilized, the dilutions applied, and the quantitation method and level of expression used to dichotomize specimens into positive and negative groups. Finally, no universal control for interlaboratory standardization appears to exist.