Tubular breast cancer experience at Washington University: a review of the literature.

Our objective was to review the presentation and management of patients with tubular breast cancer treated at Barnes-Jewish Hospital and compare our findings with the available literature. Of 3908 cases of breast cancer treated at our institution between 1986 and 1995, the incidence of tubular breast cancer at Barnes-Jewish Hospital was 1.25%. We reviewed the breast cancer risks, initial presentation, treatment, and outcome of 39 women with 40 tubular breast cancers and compared our series with others in the literature. The mean patient age was 67 years, which is older than most other series. Twenty-nine of the 39 cancers (74%) were detected by screening mammography; the remainder presented with a palpable breast mass. The mean tumor size was 8 mm (range, 1-60 mm). Twenty-three of 25 tumors were ER+ (92%) and none had axillary nodal involvement. Bilateral breast cancer developed in 3 patients (8%). An additional 500 cases of tubular breast cancer have been described in the literature. When the component of the invasive tumor is > 75% tubular carcinoma, most patients present with early-stage disease that is ER+ in 47 of 56 tumors (84%). The natural history is indolent and metastases are rare. Bilateral breast cancer developed in 58 of the 540 cases (11%), 4 of which were tubular carcinomas. Local recurrences developed in 9 of 29 patients (31%) treated by excision alone. The role of tamoxifen has not been determined. Given the available data, the initial surgical staging and management of tubular carcinoma should be identical to other invasive histologies.

Cytology of angiosarcoma. Findings in fourteen fine-needle aspiration biopsy specimens and one pleural fluid specimen.

We report the cytologic features of 15 cases of angiosarcoma from various sites and include 14 fine-needle aspiration (FNA) biopsy specimens and 1 pleural fluid specimen. Six were initial diagnoses with histologic confirmation; an additional case in the liver was an initial diagnosis without tissue confirmation. One case represented lymph node metastasis from a primary prostatic epithelioid angiosarcoma. In 10 cases, immunohistochemical staining for factor VIII-related antigen, CD34, CD31, or Ulex europaeus agglutinin I was performed on the cytology or histology specimen. The aspirates varied in cellularity, and the degree of nuclear atypia ranged from relatively bland in a case of low-grade angiosarcoma of the prostate to highly pleomorphic in a lymph node metastasis from a facial cutaneous angiosarcoma. Vasoformative features such as intracellular RBCs, well-formed vessels, attempts at microacinar/lumen formation, and intracytoplasmic lumens were variably present. The background was bloody in all specimens, with necrosis in rare cases. This cytologic series emphasizes that the cytologic features are heterogeneous but that the diagnosis can be suggested by fine-needle aspiration (FNA) when vasoformative features are present. The diagnosis can be made conclusively by FNA with immunocytochemical confirmation of endothelial differentiation.

Interferon regulatory factor-1 and -2 expression in human melanoma specimens.

BACKGROUND: Interferon regulatory factor (IRF)-1 and IRF-2 are nuclear transcription factors that respond to interferon-gamma. IRF-1 acts as the effector arm of the interferon-gamma response in tumor cells, whereas IRF-2 binds to the same DNA consensus sequence and opposes IRF-1 activity. This effect is intact in human and murine tumor models, including melanomas; previous work in our laboratory demonstrated the tumor-suppressing activity of IRF-1 expression in in vivo models and the opposing effect of IRF-2. The expression of IRF-1 and -2 in human solid tumors had not been previously investigated. METHODS: Formalin-fixed, paraffin-embedded, archival tissue specimens from 38 human melanomas were obtained and stained with polyclonal anti-IRF-1 and anti-IRF-2 antibodies, using an avidin-biotin-peroxidase complex technique with epitope retrieval. RESULTS: Twenty-nine specimens showed granular cytoplasmic staining with the anti-IRF-1 or anti-IRF-2 antibodies. IRF-1 staining was correlated with less advanced disease. Superficial spreading and in situ lesions exhibited more frequent IRF-1 staining, compared with nodular or metastatic disease. Only more advanced lesions showed neither IRF-1 nor IRF-2 staining. CONCLUSIONS: Immunohistochemical staining of archival tissue identified IRF-1 and -2 in human melanomas; this had not been previously demonstrated. IRF-1 staining was correlated with the morphologic characteristics of less advanced disease. Tumor-suppressing effects of IRF-1 may account for the less aggressive biologic features of IRF-1-expressing melanomas, as we would predict from the experimental data.

Peripheral pulmonary adenocarcinomas with bronchioloalveolar features: immunophenotypes correlate with histologic patterns.

Peripheral pulmonary adenocarcinomas (PPAs) often demonstrate a bronchioloalveolar component, with or without glandular differentiation. PPAs can be nondescript, mucinous, or show features of Type II pneumocytes. Particularly, mucinous lung carcinomas can resemble gastrointestinal metastases. Previous reports suggested that patterns of keratins 7 (K7) and 20 (K20) differ in pulmonary tumors versus enteric metastases. These studies, however, often failed to specify the precise morphotypes of PPA. Thus, we undertook this evaluation of PPAs with different histologic images. Thirty-nine cases were retrieved from institutional files; all were confirmed as primary tumors by clinicopathologic and radiographic review. Cases were classified as Type I (mucinous) bronchioloalveolar carcinoma (BAC1); Type II (nonmucinous) bronchioloalveolar carcinoma (BAC2); conventional PPA with BAC1-like areas (PPA1); or conventional PPA with BAC2-like foci (PPA2). Immunostains were performed for K7, K20, carcinoembryonic antigen, CA19-9, tumor-associated glycoprotein-72, surfactant apoprotein-A, and the c-erbB-2 peptide. BAC1 and PPA1 failed to express surfactant apoprotein-A, and BAC2 also consistently lacked K20, whereas 28% of PPA2 lesions were labeled for K20. All of the other determinants, however, were seen in variable proportions in each subgroup of PPA. Primary BAC1 and PPA1 resembled enteric adenocarcinomas immunophenotypically; on the other hand, BAC2 demonstrated a pattern of protein expression similar to that of Type II pneumocytes. PPA2s are a diverse group of neoplasms, and a subset of PPA2 does show K20 reactivity, as would be expected in metastatic enteric carcinomas. Thus, immunohistochemical data on PPAs must be interpreted carefully and only in clinicopathologic context. With respect specifically to primary pulmonary mucinous tumors, there still seems to be no uniformly reliably marker that will always allow the exclusion of metastatic enteric tumors.

Small (1.5 cm or less) liver metastases: US-guided biopsy.

PURPOSE: To determine the techniques used for and the success of ultrasound (US)-guided biopsy of hepatic metastases 1.5 cm in diameter or smaller. MATERIALS AND METHODS: A computer search of radiology reports identified 29 patients who underwent US-guided biopsy of 30 hepatic masses 1.5 cm in diameter or smaller suspected to be metastases. All 30 lesions were sampled for biopsy with the free-hand technique. Parameters assessed were lesion size and location, needle size, transducer type, number of passes made, cytologic or histologic analysis, and final histologic diagnosis. Biopsies were considered successful if a positive histologic diagnosis of metastasis was made. RESULTS: The mean lesion diameter was 1.3 cm (range, 0.9-1.5 cm). Lesion depth was 3-9 cm (mean, 5 cm). Twenty biopsies were performed with a 22-gauge aspirating needle and analyzed cytologically. An average of 2.7 passes were made per lesion. Phased-array sector transducers were used in 23 lesions. In 28 (93%) lesions and 28 (96%) patients, an adequate specimen was obtained to establish the histologic diagnosis of metastatic disease. CONCLUSION: US appears to be an effective guidance technique for biopsy of small liver metastases.

The expression of trophoblastic cell markers by lung carcinomas.

The authors studied the expression of trophoblastic cell markers in lung carcinomas cells by immunoperoxidase staining using antibodies against three trophoblastic glycoproteins (human chorionic gonadotropin [hCG]; human placental lactogen [hPL]; pregnancy-specific beta 1-glycoprotein [SP-1]). One hundred five tissue sections from 44 lung carcinomas of various histological types were examined for positive staining with three antibodies. Hematoxylin-eosin-stained sections from the same tissue blocks were examined for the presence of tumor giant cells. The aim was to study the relationship between tumor giant cells and the trophoblastic glycoprotein expression in lung carcinomas. Small cell carcinoma (SCC) did not show any positive reaction with all three markers. Squamous cell carcinoma (SQCC) showed positive staining with hCG in 21% of cases, hPL in 28%, and SP-1 in 64%. Adenocarcinoma showed positive staining with hCG in 60% of cases, hPL in 10%, and SP-1 in 80%. Large cell carcinoma (LCC) showed positive staining with hCG in 93% of cases, hPL in 56%, and SP-1 in 93%. The positive reaction did not appear to be restricted to nor associated with the tumor giant cells. It was concluded that these trophoblastic cell markers are expressed in various types of lung carcinomas and that they are not associated with certain histological types or with tumor giant cells.

Cytologic characterization of bronchial epithelial changes in small cell carcinoma of the lung.

Cytologic specimens obtained by fiberoptic bronchoscopy were examined in 111 cases; the diagnosis of small cell carcinoma of the lung was established with this procedure. The findings of cytologic examination were compared to those of histologic examination of the biopsy specimens obtained simultaneously. Our aim was to evaluate other cell components in the specimen besides small cell carcinoma. The majority of the cases showed morphologic changes in the bronchial epithelium, which ranged from benign squamous metaplasia (59%) to atypical squamous metaplasia (8%) and squamous cell carcinoma (8%). In 6% of cases a large cell carcinoma component was found in addition to the small cell carcinoma. The findings show a good correlation with those of histologic examination of the biopsy specimens, indicating that cytologic examination is an excellent procedure for detecting concomitant changes in the bronchial epithelium in the setting of small cell carcinoma of the lung.

Malakoplakia of liver associated with a perforated colonic diverticulum. A case report and review of the literature.

Malakoplakia of the liver is rare, only three cases having been documented in the world literature. Described here is the first case of malakoplakia of the liver as a complication of a perforated colonic diverticulum.

Bronchial epithelial changes associated with small cell carcinoma of the lung.

From 1976 to 1991, 151 cases of small cell carcinoma of the lung were diagnosed by fiberoptic bronchoscopic biopsy at our institution. One hundred twenty-eight of 151 cases provided suitable material for the examination of the morphologic changes in the bronchial surface epithelium. Thirty-seven percent of the cases showed normal bronchial epithelium, 47% showed benign squamous metaplasia, 9% showed atypical squamous metaplasia, and 5% showed squamous cell carcinoma in situ. Immunohistochemical examination for bombesin and epidermal growth factor was performed on selected biopsy specimens. The biopsy specimens chosen for immunohistochemistry included 20 specimens that showed normal bronchial epithelium, 20 specimens with benign squamous metaplasia, 12 specimens with atypical squamous metaplasia, and seven specimens with squamous cell carcinoma in situ. All the specimens showed positive staining with anti-bombesin. With anti-epidermal growth factor 10% of biopsy specimens with normal epithelium showed positive staining. The positive reaction increased from 25% for biopsy specimens with benign squamous metaplasia to 58% for biopsy specimens with atypical squamous metaplasia and to 71% for biopsy specimens with carcinoma in situ. These findings suggest a connection between epidermal growth factor production by small cell carcinoma of the lung cells and changes in the bronchial surface epithelium.

Dissociation between plasma renin and plasma aldosterone induced by dietary glycine hydrochloride.

We evaluated the effects of selective dietary chloride loading (without sodium) on plasma renin activity (PRA) and plasma aldosterone in the sodium-deprived Sprague-Dawley rat. Three groups of animals were fed one of the following diets for 13 days: 1) low NaCl; 2) high NaCl; or 3) low sodium, high chloride, provided as glycine hydrochloride. Compared with NaCl-deprived animals, PRA and plasma aldosterone were lower (P less than 0.01) in animals fed low sodium high chloride, whereas aldosterone in animals fed glycine hydrochloride was higher (P less than 0.01) than that of NaCl-deprived animals. In contrast, plasma concentrations of corticosterone and 18-hydroxycorticosterone were not increased by selective chloride loading. Glycine chloride-fed animals were acidotic and had elevated plasma concentrations of potassium and ionized calcium. Thus stimulation of aldosterone by selective chloride loading is not related to PRA or ACTH but may be due to a direct effect of acidosis or an indirect effect of acidosis on potassium and/or calcium. Additionally, selective chloride loading appears to stimulate the conversion of 18-hydroxycorticosterone to aldosterone.

Dissociation of hyperreninemia and renal prostaglandin synthesis in the adrenalectomized rat.

The aim of this study was to determine whether hyperreninemia in the adrenalectomized (ADX) rat is dependent on renal prostaglandin synthesis, as has been suggested for two other hyperreninemic conditions, Bartter's syndrome and chronic liver disease. Plasma renin concentration (PRC) in anesthetized, ADX rats was significantly increased (delta +480%; p less than 0.001) compared to sham-operated controls. In vivo, indomethacin (10 mg/kg i.v.) significantly reduced PRC of anesthetized, ADX rats after both 45 min (delta -34%; p less than 0.05) and 90 min (delta -47%; p less than 0.05). In vitro renin release from renal cortical slices of ADX rats was also significantly greater (delta +130%; p less than 0.05) than from sham-operated control cortical slices. Renin release from cortical slices of ADX rats given dexamethasone (10 micrograms/kg/day) for 4 days prior to sacrifice did not differ from sham-operated control values. Prostaglandin E2 (PGE2) release from cortical slices of ADX rats did not differ significantly from controls. However, PGE2 synthesis in glomeruli microdissected from ADX rats was significantly increased (delta +110%; p less than 0.001) compared to controls. PGE2 synthesis in glomeruli of dexamethasone-treated ADX rats remained significantly elevated compared to controls. Ibuprofen (10(-6) M) decreased PGE2 synthesis in cortical slices by 80%. However, prostaglandin synthesis inhibition had no effect on renin release from either ADX or control renal cortical slices. These results suggest that despite increased glomerular synthesis, prostaglandins do not directly influence renin release in the ADX rat.