Harriette J. Kahn

Malignant myoepithelioma of the breast. An immunohistochemical study by light and electron microscopy.

Myoepitheliomas of the breast are extremely rare; only four cases have been reported to date. It is not clear whether these tumors behave in a benign or malignant fashion, and no metastatic spread has been documented. We report a myoepithelioma of the breast with metastatic spread to an axillary lymph node. Myoepithelial cell (MEC) differentiation in the tumor was characterized using electron microscopic (EM) criteria. Immunologic investigations at the LM and EM levels showed that the tumor cells were positive for S100 protein, actin, and epidermal cytokeratin; these findings are indicative of MEC differentiation. By immuno‐EM, cytokeratin filaments were present in a perinuclear location, while actin filaments were concentrated along the cell periphery. To the authors knowledge, this is the first report of a breast myoepithelioma that has metastasized and also the first report of the immunologic characterization of a myoepithelioma at the light and electron microscopic levels.

Clinical behavior of untreated axillary nodes after local treatment for primary breast cancer

AbstractBackground: The purpose of this study was to examine the rate of axillary failure in patients with primary breast cancer treated without axillary dissection or radiation and to determine what factors may be associated with axillary failure.nMethods: We studied 112 patients with invasive breast cancer treated for primary disease with breast-conserving surgery without axillary dissection or radiation to the breast or axilla, accrued between 1977 and 1986. Data for these patients were prospectively gathered for a research database and reviewed retrospectively to determine axillary failure. The effects of age, tumor size, estrogen receptor (ER) status, progesterone receptor (PgR) status, histologic grade, nuclear grade, and tumor emboli on time to axillary failure were examined.nResults: The median follow-up was 9.6 years. There were 26 axillary recurrences, resulting in a 10-year actuarial nodal control rate of 72%. Patients with nodal failure proceeded to axillary dissection with minimal morbidity. In both univariate and multivariate analyses, only tumor size was significantly associated with axillary failure (p=0.04 andp=0.06, respectively).nConclusions: This study demonstrates a significant effect of tumor size on axillary failure and a reasonable rate of local control in small tumors. Further research should examine the utility of axillary dissection in women with small breast cancers.

Immunohistochemical and electron microscopic assessment of childhood rhabdomyosarcoma. Increased frequency of diagnosis over routine histologic methods.

Histologic examination was carried out in 65 cases of childhood rhabdomyosarcoma (RMS), 53 embryonal, and 12 alveolar. Cross‐striations were seen on light microscopy in 12 (23%) embryonal and 4 (33%) alveolar tumors. The capacity of immunohistochemical staining (PAP technique) to increase diagnostic accuracy was assessed, using antibodies against myoglobin, the MM isoenzyme of creatine kinase, desmin, calcium magnesium‐dependent ATPase of sarcoplasmic reticulum and calsequestrin. Myoglobin was detected in 16 (30%) embryonal and eight (67%) alveolar RMS, higher numbers than obtained by viewing cross‐striations on light microscopy. The creatine kinase antibody was slightly better than the antibody to myoglobin and 15 of 25 (60%) embryonal RMS were positive when both specificities were used. The remaining three antibodies were less useful. Of 13 (two alveolar and 11 embryonal) RMS studied by electron microscopy, four showed cross‐striations, contained late myoblasts, and were positive for myoglobin. Three additional cases showed only late myoblasts and one of these was positive for myoglobin. Thus, 16 of 25 (64%) of the embryonal and seven of nine (78%) of the alveolar RMS showed either positive immunostaining or ultrastructural features of RMS. This study indicates that a combination of immunohistochemical staining, using antimyoglobin and anticreatine kinase (MM isoenzyme) antibodies, and electron microscopy are useful markers in the diagnosis of childhood RMS.

Characterization of the mononuclear infiltrate involved in regression of halo nevi

Halo nevi are characterized by progressive degeneration of nevus cells surrounded by a mononuclear cell infiltrate. We studied the morphological features of the nevus cells and the composition of the mononuclear cell infiltrate in 15 cases of halo nevi using immunohistochemical techniques and a battery of antibodies to different subsets of lymphocytes and histiocytes. Regression could be divided into four more or less identifiable stages, associated with different subsets of lymphocytes and monocyte‐macrophage lineage cells. Stage I (preregression): nests of unremarkable nevus cells were surrounded by a moderate number of T lymphocytes (relatively small percentage of helper/inducer T cells), occasional B cells and macrophages. Stage II (early regression): large number of T lymphocytes and FXIIIa‐positive cells were in close contact with nevus cell clusters which showed ragged edges. Lysozymepositive cells and epidermal Langerhans cells were mildly increased. Stage III (late regression): single nevomelanocytes showing mild atypia were present. Numerous T lymphocytes and macrophages positive for lysozyme, KP1 and/or FXIIIa were interspersed between the nevus cells. Increased numbers of epidermal Langerhans cells were present. Stage IV (complete regression): no nevus cells were observed and moderate numbers of T lymphocytes only remained. These results suggest that T cells, especially T‐suppressor cells, and different subsets of macrophages participate in the regression of the nevi.

Localized fibrous tumour of serosal surfaces. Immunohistochemical and ultrastructural evidence for a type of mesothelioma

It is uncertain whether localized lesions of serosal membranes have a kinship to mesotheliomas or are truly fibromatous in nature. Ultrastructural and immunohistochemical investigations were carried out on 12 localized benign and malignant pleural and peritoneal tumours from 10 patients. Electron microscopic findings, including the consistent and non-fibroblastic cellular organization of localized neoplasms, the presence of some form of intercellular junctions in 7 of 10 cases, basal lamina deposition in 3 cases, and polarized microvilli in one case indicated a form of mesothelial differentiation. Using monoclonal and polyclonal antibodies, positive immunostaining of tumour cells for cytokeratin peptides was detected in one case, while antibody to vimentin stained four cases. Light microscopic, ultrastructural and immunohistochemical features of one benign localized serosal tumour, with a unique blend of epithelial and spindle cells, provided further evidence for a histogenic link between localized serosal tumours and diffuse epithelial mesotheliomas. On the basis of the current findings and reports in the literature, it would appear that the majority of localized tumours of serosal membranes are a subset of mesothelioma, while a minority are fibromas.

Factors associated with local breast cancer recurrence after lumpectomy alone: Postmenopausal patients

Background: We have been following a cohort of patients who underwent a lumpectomy without receiving adjuvant radiotherapy or adjuvant systemic therapy. We now report the experience of a postmenopausal subgroup.Methods: The postmenopausal subgroup included 244 patients accrued between 1977 and 1986 and followed up. The end point was ipsilateral local breast cancer recurrence. The factors studied were the patient’s age in years; tumor size (in mm); nodal status (N-, Nx, N+); estrogen and progesterone receptor status (<10, ≥10 fmol/mg protein); presence or absence of lymphovascular/perineural invasion; presence or absence, and type, of DCIS (none, non-comedo, comedo); percentage of DCIS; histological grade (1,2,3); and nuclear grade (1,2,3). Univariate analyses consisted of Kaplan-Meier plots and the Wilcoxon (Peto-Prentice) test statistic; the multivariate analyses were step-wise Cox and log-normal regressions.Results: The median follow-up of those patients still alive was 9.1 years, and the overall relapse rate was 24% (59/244). The univariate results indicated that the characteristics of smaller tumor size, negative nodes, positive ER status, and no lymphovascular or perineural invasion were associated with significantly (P<.05) lower relapse. From the multivariate analyses, the factors lymphovascular or perineural invasion, age, and amount of DCIS were all significantly associated with local relapse with both Cox and log-normal regressions. Additionally, there was weak evidence of an association between ER (P=.08 in the Cox regression and in the log-normal) and nodal status (P=.09 in the log-normal regression) with local relapse. We also are able to define a low-risk subgroup (N-, age ≥65, no comedo, ER positive, no emboli) with a crude 10-year local recurrence rate of 9%.Conclusion: With longer follow-up, and for postmenopausal patients, there continues to be support for the theory that local relapse is affected by the factors lymphovascular or perineural invasion, age, amount of DCIS, ER, and nodal status. A low risk subgroup has been identified.

Categorization of pediatric neoplasms by immunostaining with antiprekeratin and antivimentin antisera.

Forty‐six tumors in children were examined using light microscopy and subsequently frozen sections were stained with antiprekeratin and antivimentin antisera, so that the tumors could be classified by tissue of origin. Except for two adrenal cortical carcinomas and four liver tumors, most epithelial neoplasms continued to produce prekeratin filaments, a characteristic of normal epithelial cells. Tumors and cells of epithelial origin did not produce vimentin filaments, whereas normal and neoplastic mesenchymal cells did. Tumors with both epithelial and mesenchymal components produced vimentin filaments in mesenchymal areas and prekeratin in epithelial areas. Tumors of lymphoid origin showed variable production of vimentin filaments, depending on the amount of cell cytoplasm, but did not contain prekeratin filaments. Of the neuroectodermal tumors, only the ganglioneuroma contained vimentin filaments and none contained prekeratin filaments. Thus, antibodies to both prekeratin and vimentin filaments are useful in diagnosing childhood neoplasms and studying their histogenesis.

Factors associated with local breast cancer recurrence after lumpectomy alone

AbstractBackground: The purpose was to determine the rate of local breast relapse in patients with breast cancer uniformly treated with partial mastectomy but without postoperative radiotherapy and without systemic adjuvant therapy. We also systematically examined the factors associated with local recurrence to determine whether a low-risk subgroup existed.nMethods: A retrospective review of a prospectively followed (median, 8 years) cohort of 293 patients was performed. The end-point was ipsilateral local breast cancer recurrence. The patients age, tumor size, nodal status, estrogen and progesterone receptor status, histology, and tumor and nuclear grade were studied, as were the presence and amount of carcinoma in situ and the presence of tumor emboli using univariate Kaplan-Meier and Cox step-wise multivariate analyses.nResults: The overall local relapse rate was 26% (77 recurrences). Univariate factors significantly associated with decreased local relapse included older age, negative nodes, small tumor size, positive estrogen receptor status, and absence of tumor emboli. Significant multivariate variables were age, nodal status, estrogen receptor status, absence of comedo carcinoma in situ, and tumor emboli. A low-risk subgroup of 66 patients was defined with a 6% 10-year local recurrence rate.nConclusion: Important patient and tumor variables associated with local breast cancer relapse after breast-conserving surgery can define a low-risk subgroup.

A comparison of all-subset Cox and accelerated failure time models with Cox step-wise regression for node-positive breast cancer

SummaryClinical studies usually employ Cox step-wise regression for multivariate investigations of prognostic factors. However, commercial packages now allow the consideration of accelerated failure time models (exponential, Weibull, log logistic, and log normal), if the underlying Cox assumption of proportional hazards is inappropriate. All-subset regressions are feasible for all these models.We studied a group of 378 node positive primary breast cancer patients accrued at the Henrietta Banting Breast Centre of Womens College Hospital, University of Toronto, between January 1, 1977, and December 31, 1986. 85% of these patients had complete prognostic factor data for multivariate analysis, and 96% of the patients were followed to 1990. There was evidence of marked departures from the proportional hazards assumption with two prognostic factors, number of positive nodes and adjuvant systemic therapy. The data strongly supported the log normal model. The all-subset regressions indicated that three models were similarly good. The variables 1) number of positive nodes, 2) tumour size, and 3) adjuvant systemic therapy were included in all three models along with one of three biochemical receptor variables 1) ER, 2) combined receptor (ER- PgR-; ER+ PgR-; ER- PgR+; ER+ PgR+; or 3) PgR.Better multivariate modeling was achieved by using quantitative prognostic factors, a check for appropriate underlying model-type, and all-subset variable selection. All-subset regressions should be considered for routine use with the many new prognostic factors currently under evaluation; it is very possible that there may not be a single model that is substantially better than others with the same number of variables.

Signet ring lymphoma of the skin: Ultrastructural and immunohistochemical features

Signet ring lymphoma is a recently described morphologic variant of B cell lymphoma. We are reporting the first case of this subtype of follicular lymphoma that was initially confined to the skin and remained so. Light microscopy showed the tumor to be composed of monomorphous plump epithelioid cells with abundant eosinophilic cytoplasm involving the dermis and subcutaneous fat. Cytoplasmic vacuoles were present within some of the tumor cells, suggesting the diagnosis of adenocarcinoma with a signet ring appearance. Some of the vacuoles were positive on periodic acid-Schiff with diastase staining, and no mucin was identified. Immunostaining showed the presence of intracytoplasmic IgG heavy chains and kappa light chains. Characteristic intracytoplasmic inclusions, which were formed by saccules and microvesicular bodies, were seen by electron microscopy. The diagnosis of signet ring lymphoma was made. No systemic involvement was found. Over a period of 7 years, multiple new skin lesions occurred, all of which responded to radiation therapy. This rare variant of follicular center cell lymphoma, which can occur in the skin, should not be confused with metastatic mucinous adenocarcinoma and malignant melanoma.