Search

Objective: To investigate the relationship between the detection of prostatic intra-epithelial neoplasia (PIN) on initial prostate biopsy and subsequent invasive prostatic adenocarcinoma., Patients and Methods: Thirty-six men (mean age 67 years, range 52-82) with PIN underwent digital rectal examination (DRE), serum prostate-specific antigen (PSA) measurement and transrectal ultrasonography (TRUS) before the initial biopsy and documentation of PIN. They were followed up with serial PSA, TRUS and a repeat biopsy every 6 months until either invasive carcinoma was identified or 2 years had elapsed., Results: The initial biopsy revealed Grade I PIN in 33%, Grade II in 22%, and Grade III in 45% of the men. The repeat biopsy showed evidence of invasive carcinoma in 21 patients (58%; Group I), while 15 (42%) showed persistence of PIN (Group II). In Group I, 19 had had high-grade PIN (Grade II/III) on initial biopsy compared with one in Group II. The findings on DRE, and age, were no significantly different between groups. TRUS revealed a hypoechoic lesion in 15/21 patients in Group I compared with 7/15 patients in Group II. There was an increase in PSA level in 18 patients in Group I (from 8.4 to 11.6 ng/mL)., Conclusions: PIN and invasive adenocarcinoma of the prostate were closely associated, and the likelihood for coexistence was higher in patients with high-grade PIN, increasing PSA level or positive findings on TRUS. We recommend that all patients who show high-grade PIN on prostate biopsy be followed very closely with serial PSA measurements and repeat biopsies from both the area of PIN and other areas of the prostate.

Neovascularization of tumor tissue (tumor angiogenesis) is considered essential for tumor growth, proliferation and eventually metastasis. Microvessel density or count, a measure of tumor angiogenesis, correlates with clinical outcome in skin, breast, lung and prostate carcinomas. To determine whether an association of tumor angiogenesis and nodal metastasis exists in invasive bladder cancer, microvessel counts in 41 primary invasive stages (T2 to 4,NX,M0) bladder cancers were assessed. Microvessels were identified by immunostaining of endothelial cells for factor VIII-related antigen. Microvessels were scored in selected areas showing active neovascularization, either counting a 200 x field (0.74 mm.2) or by using a 10 x 10 square ocular grid (0.16 mm.2). The microvessel count correlated with the presence of occult lymph node metastases (p < 0.0001) by both techniques. The mean microvessel count in 27 patients without lymph node metastases was 56.2 microvessels per 200 x field (standard deviation [SD] 29.5, range 7 to 130) or 28.6 microvessels per grid (SD 14.4, range 4 to 65). The 14 patients with histologically proved lymph node metastases showed mean 138.1 microvessels per 200 x fields (SD 37.9, range 82 to 202) or 74.7 microvessels per grid (SD 14.4, range 43 to 115). Good correlation was noted between area and grid counting (r = 0.97). Tumor T stage, grade and the presence of vascular or lymphatic invasion did not correlate with the presence of lymph node metastases (p = 0.41, 0.59 and 0.26, respectively). Microvessel count may provide important information regarding the risk of occult metastasis in patients with invasive bladder carcinomas.

We report a 12-year-old male with acute post-streptococcal glomerulonephritis (APSGN) occurring 1 year after a cadaveric renal transplant. Although recurrent and de novo renal transplant glomerulonephritides have been well described in large series of adult and pediatric renal transplant recipients, post-infectious glomerulonephritis has been rarely reported, and APSGN has never been reported in either adult or pediatric transplant series. We speculate on the reasons for the lack of occurrence of APSGN in renal transplant recipients.

It has been estimated that approximately 1 in 9 women (living to age 85 years) in the United States will develop breast cancer at some time in their lives; hence the interpretation of breast biopsies has become a large and important component of the surgical pathologist's practice. Indeed, there is great pressure on surgical pathologists to make accurate diagnoses of breast lesions because of their high incidence, the increased public awareness of breast disease, the greater use of screening mammography to detect early carcinomas, the development of multiple therapeutic options (which is often determined by tumor pathology), and the harsh medical-legal climate. The focus of this article will be to define the clinicopathologic features of two malignant breast lesions that are particularly prone to simulate benign lesions, and thus pose important diagnostic problems to surgical pathologists. The two lesions include spindle-cell carcinoma and low-grade duct carcinoma in situ.

With greater public awareness and use of mammography, more breast carcinomas are detected early, before the axillary lymph nodes contain metastases. About 60% of patients will be node-negative; yet 20-30% of node-negative breast carcinoma patients will develop recurrent tumors, and they are at risk of dying of disease. These patients should be considered for adjuvant therapies, whereas the remaining 70-80% appear to be cured by surgery. Thus, accurate assessment of risk factors for recurrence provides valuable therapeutic information. This review focuses on recent reports that help identify subsets of patients at variable risk for recurrent disease.

Abundant evidence has shown that tumor growth and metastasis are dependent upon tumor angiogenesis (TA). TA refers to the growth of new vessels toward and within the tumor. Until TA occurs, tumors grow no larger than 2-4 mm in diameter. Also, TA is necessary at the beginning and at the end of the metastatic cascade of events. Thus, it seems reasonable that increasing intratumoral microvascular density (iMVD) might correlate with greater tumor aggressiveness, such as a higher frequency of metastases and/or decreased survival. Indeed, in 1991 my colleagues and I reported a statistically significant association between greater incidence of metastases in patients with breast carcinoma and increasing iMVD. Microvessel density was measured with a light microscope in a single area of invasive tumor (200x field or 0.74 mm2) representative of the highest microvessel density (neovascular "hot spot"). This was done after endothelial cells, lining the microvessels, had been highlighted with anti-factor VIII-related antigen/von Willebrand's factor (F8RA/vWF). Subsequent studies by other investigators, using either anti-F8RA/vWF or other relatively vessel-specific reagents such as anti-CD31, have shown that the association of greater tumor aggressiveness with increasing iMVD exists not only in breast carcinoma, but also in other solid tumors. This article reviews the methods of highlighting intratumoral vessels and describes the techniques for counting these vessels for assessing iMVD.

Recent evidence suggests that a tumor suppressor gene important in the progression of prostatic carcinoma may reside on chromosome 16q. The exact location and identity of this gene are unknown. We used fluorescence in situ hybridization in a novel manner to define more clearly the location of this gene. Region-specific chromosome 16 cosmid contig probes were hybridized directly to interphase prostatic carcinoma nuclei in order to measure physical deletion of chromosomal loci. Fifteen of 30 tumors (50%) showed evidence of physical deletion of 16q24. Other probes were used to test for regions of chromosome 16 deletions in the same specimens, and a map showing a region of common deletion was created. This map showed the proximal terminus of the region of common deletion to be located distal to 16q23.1. These data correlate well with loss of heterozygosity data in the literature and provide further evidence for the presence of a prostatic carcinoma tumor suppressor gene on chromosome 16q.

In the present study we update previous results on the prognostic value of intratumoral microvessel density (IMD), determined immunocytochemically using the monoclonal antibody CD-31 and a standard streptavidin-immunoperoxidase technique, published in the J Clin Oncol 12:454-466, 1994. This study was undertaken in those 211 node-negative breast cancer (NNBC) cases of that series of which we had pathological material available to determine all the prognostic indicators. The median period of follow-up has been extended to 78 and 80 months for relapse-free survival (RFS) and overall survival (OS), respectively, and new biological indicators (i.e. Ki-67 labeling and 67 kDa laminin receptor expression) were included in the analysis. The main results obtained are: i) a confirmation that IMD is not associated with the other biological markers studied, i.e. expression of p53 protein, c-erbB-2 protein, 67 kDa laminin receptor, and cell kinetics; IMD was weakly associated only with histological grade (p = 0.053); ii) IMD remains a highly significant prognostic factor for RFS and OS (p < 0.0001 and p = 0.018, respectively) in univariate analysis; iii) in multivariate analysis on RFS, IMD (likelihood ratio test (LRT) = 30.16; p < 0.0001), 67 kDa laminin receptor (LRT = 9.80; p = 0.0017), the IMD/67 kDa laminin receptor interaction (LRT = 8.62; p = 0.0033), tumor size (LRT = 8.56; p = 0.0034), and p53 protein (LRT = 4.96; p = 0.025) are significant and independent prognostic indicators. For OS, only tumor size (LRT = 8.34; p = 0.0038), menopausal status (LRT = 5.16; p = 0.023), p53 protein (LRT = 4.37; p = 0.036), and IMD (LRT = 4.05; p = 0.044) retain a significant and independent prognostic value. The results of this study confirm the prognostic importance on RFS of the variables previously tested, but not of peritumoral lymphatic vessel invasion. A novel finding is that 67 kDa laminin receptor and the IMD/67 kDa laminin receptor interaction are also significant and independent variables. For OS, the results confirm that both IMD and tumor size are significant and independent variables. With prolonged follow-up the novel finding that emerges is the prognostic importance of menopausal status and p53 protein. This new information could be useful for a more accurate selection of high-risk NNBC patients who require careful follow-up and may benefit from adjuvant therapy.

A more accurate method of detecting nodal disease in squamous cell carcinoma of the tongue is needed so that treatment of the neck with its associated morbidity can safely be reserved for patients who actually have metastatic disease. Tumor angiogenesis and the expression of the p53 antigen--which have each been shown to be predictive of metastasis in breast and colon cancer, respectively--are examined for their ability to predict neck metastasis in tongue cancer. Fifty-seven patients with T1 and T2 squamous cell carcinoma of the oral tongue, whose neck disease was examined by dissection or by 2-year follow-up, were studied. Twenty-eight patients (49%) were node positive and 29 patients (51%) were node negative. The primary tumors were immunohistochemically stained for the p53 antigen and for factor VIII, which allowed the blood vessels within the tumor to be quantitated. The mean vessel counts per x200 high-power field were 59.8 and 61.5 for node-positive and node-negative patients, respectively (p = 0.8). Node-positive patients showed overexpression of p53 43% of the time, vs. 61% for node-negative patients (p = 0.17). Multivariate analysis confirmed that no difference in tumor angiogenesis or the expression of the p53 antigen was found between tumors that had metastasized and those that had not. Therefore neither tumor angiogenesis nor the p53 tumor marker is clinically useful in determining lymph node metastasis in these patients.

Tumor angiogenesis is essential for tumor growth and metastasis, and intratumoral microvessel density correlates with prognosis in breast carcinoma. Yet, how intratumoral microvessel density correlates with tumor cell and intratumoral endothelial cell proliferation remains incompletely understood. To this end, we stained 57 formalin-fixed, paraffin-embedded breast carcinomas with antibody MIB1 to determine tumor cell Ki67 labeling index and with anti-CD34 to observe microvessels. We correlated the tumor cell Ki67 labeling index and mitotic figure index with intratumoral microvessel density. Using a double labeling technique combining antibody MIB1 and anti-CD34, we measured intratumoral endothelial cell proliferation in 20 of these cases and correlated these findings with tumor cell Ki67 labeling index, mitotic figure index, and intratumoral microvessel density. The intratumoral Ki67-labeling index was 45-fold greater (P < 0.000001) than that of microvessels in adjacent benign breast. Yet, endothelial cell Ki67 labeling index did not correlate with intratumoral microvessel density, tumor cell Ki67 labeling index, or mitotic figure index nor did intratumoral microvessel density correlate with tumor cell Ki67 labeling index or mitotic figure index. These findings suggest that, although endothelial cells are actively proliferating within the tumor, intratumoral microvessel density and intratumoral endothelial cell proliferation are independent of each other and of tumor cell proliferation. Thus, intratumoral microvessel density, endothelial cell proliferation, and tumor cell proliferation may be regulated by separate mechanisms.

Ewing's sarcoma (ES) and peripheral neuroepithelioma (PN) are closely related tumors, and it can be difficult to distinguish them from other small-round-cell tumors (SRCTs). The glycoprotein p30/32MIC2 is highly, but not exclusively, expressed in both ES and PN. Although the monoclonal antibody (Mab) HBA71, which reacts with P30/32MIC2, has been reported to be relatively specific and highly sensitive for both neoplasms, it is not readily available. Yet, Mab O13 is commercially available, and it purportedly displays the same immunostaining characteristics as HBA71. Because O13 has not been studied extensively, we immunostained 21 ES/PNs and 147 other tumors or lesions that might show SRCT-like features with O13. The results were similar to those reported for HBA71. We found O13 to be 100% sensitive for ES/PN; and, no immunostaining was noted on the SRCTs often included in the differential diagnosis of ES/PN (i.e., conventional neuroblastoma, rhabdomyosarcoma, and non-lymphoblastic lymphomas). But, O13 immunoreacted with lymphoblastic lymphomas and some other tumors and normal tissues. Nonetheless, this nonspecific reactivity should not cause diagnostic problems, if an antibody panel containing anti-desmin and anti-leukocyte common antigen is used in conjunction with O13. We conclude that, within the proper diagnostic context, strong immunoreactivity of a SRCT tumor for O13 should be considered good evidence that the tumor is ES/PN.

Tumor proliferation is inversely associated with survival in patients with breast carcinoma. Although the proliferation marker Ki-67 antigen correlates with mitotic figure count (MFC) in breast carcinoma (number of mitotic figures per high-powered field), the more precise association of Ki-67 antigen expression with mitotic figure index (MFI) (number of mitotic figures per tumor cells) and histologic grade remains incompletely defined, especially when using the new "paraffin"-reactive monoclonal antibody MIBI to mark the Ki-67 antigen. To determine the association of Ki-67 antigen expression with mitotic figure content we immunostained 50 formalin-fixed, paraffin-embedded breast carcinomas with the MIB1 monoclonal antibody and correlated the results with MFC, MFI, and Scarff-Bloom-Richardson grade. The Ki-67 antigen-labeling index (K67LI) was the number of MIBI-positive cells/1,000 tumor cells, the MFI was the number of mitotic figures/1,000 tumor cells, and the MFC was the number of mitotic figures/10 high-powered fields. The K67LI correlated highly with the MFI (r = .76, P < .0001), MFC (r = .78, P < .0001), and Scarff-Bloom-Richardson grade (r = .73, P < .0001). In addition, analysis of variance showed that MFI was more precise than MFC in estimating K67LI and thus a more repeatable measure of tumor proliferation. Moreover, measurements made by MFI were as precise as those made by K67LI. Therefore, the correlation of K67LI with mitotic figure content was strong enough to suggest that a very carefully measured MFI provides an estimate of tumor growth fraction equivalent to the K67LI. However, which method is optimal for estimating tumor proliferation rate remains unclear.

Purpose: To determine the absolute and relative value of microvessel density (MVD), p53 and c-erbB-2 protein expression, peritumoral lymphatic vessel invasion (PLVI), and conventional prognosticators in predicting relapse-free (RFS) and overall survival (OS) rates in patients with node-negative breast carcinoma (NNBC)., Patients and Methods: We monitored 254 consecutive patients with NNBC for a median of 62 months. Intratumoral MVD was measured after microvessels were immunostained using anti-CD31 antibody. p53 and c-erbB-2 protein and hormone receptors were also determined immunocytochemically. Results were analyzed by both univariate and multivariate statistical analysis., Results: Univariate analysis showed that MVD was significantly predictive of both RFS (odds ratio [OR], 8.30; P = .0001) and OS (OR, 4.50; P = .012) when tested as a continuous or dichotomous variable. Likewise, tumor size (OR, 3.16; P = .0012), PLVI (OR, 4.36; P = .0009), estrogen receptor (ER) status (OR, 2.35; P = .016), progesterone receptor (PR) status (OR, 2.00; P = .017), and expression of p53 protein (OR, 2.82; P = .004) were significantly associated with RFS. Tumor size (OR, 3.80; P = .0038) and expression of p53 protein (OR, 2.58; P = .024) were significantly associated with OS by univariate analysis. Multivariate analysis showed that MVD (P = .0004), p53 protein expression (P = .0063), tumor size (P = .0144), and PLVI (P = .0033) were all significant and independent prognostic factors for RFS. However, only tumor size (P = .004) and MVD (P = .047) were independent predictors for OS. c-erbB2 expression was not associated with outcome by either univariate or multivariate analysis., Conclusion: MVD, p53 expression, PLVI, and tumor size are independent prognostic indicators of recurrence, which are useful in selection of high-risk NNBC patients who may be eligible to receive adjuvant therapies.

Cervical dysplasia, or cervical intraepithelial neoplasia (CIN), is a premalignant precursor to cervical cancer. This study was designed to determine whether dysplastic lesions are angiogenic. Tissue sections from 23 surgical specimens were immunohistochemically stained for factor VIII antigen, a marker for endothelial cells. The results demonstrate that a region of neovascularization develops along the basement membrane subtending dysplastic epithelium when compared to adjacent normal epithelium. Comparison of microvessel counts underlying low grade lesions (condyloma and CIN I) with microvessel counts of CIN III lesions shows a statistically significant increase in the more advanced lesions. In a subset of the high grade lesions, large vascular structures are also noted in the upper layers of the epithelium, suggesting that a second stage of neovascularization consists of extension of stromal vascular papillae into the dysplastic lesions toward the surface of the epithelium. There is no statistical correlation between the amount of inflammation and the angiogenic ratio for each lesion, implying that angiogenesis is not secondary to the inflammatory response evoked by the lesion. The human papillomavirus type present in four CIN III lesions was determined by in situ hybridization; the amount of angiogenesis appears to be independent of the human papillomavirus type.

Epidermal growth factor receptor (EGFR) is a potentially useful new biological prognostic and predictive indicator in human breast cancer. Additional research on EGFR is warranted to enhance our information on: i) the method of choice for its detection and quality control issues; ii) its association with novel pathobiological markers of prognosis; iii) its prognostic value in multivariate analysis; and iv) its capability to predict response to hormone therapy and, in the future, to biological treatments using antibodies against the specific receptor or its ligands. In the present study we update previous data on EGFR status, determined immunocytochemically, by prolonging the period of observation up to 5 years and by including, in the multivariate analysis, several new biological indicators. The main results obtained are: i) EGFR is weakly associated with Ki-67 score (p = 0.073) and with p53 expression (p = 0.06); ii) EGFR is a significant indicator for recurrence (p < 0.01 and odds ratio of 2.82) but not for death (p = 0.27 and odds ratio of 1.49); iii) the prognostic power of EGFR is enhanced when combined with the knowledge of S-phase fraction; and iv) in multivariate analysis on relapse-free survival, EGFR and S-phase fraction (likelihood ratio test = 26.40; p < 0.01), c-erB-2 protein and p53 mutant protein expression (likelihood ratio test = 5.94; p = 0.05), cathepsin D (likelihood ratio test = 9.78; p < 0.01), and nodal status (likelihood ratio test = 7.32: p < 0.01) are significant and independent prognostic factors in early-stage breast carcinoma. This new information could be of help for a more rational approach in the use of EGFR as a marker in future clinical research.

Recent studies have shown that women with invasive breast carcinoma having high microvessel density (MVD) (a measure of tumor angiogenesis) or epidermal growth factor receptor (EGFR) expression have increased risk for metastasis and/or decreased survival. To determine if MVD and EGFR expression provide additive prognostic information, we analyzed these two prognostic indicators in the primary invasive breast carcinomas from 165 consecutive patients, who were followed for a median of 51 months. Univariate analysis showed a highly significant (p < 0.001) association of MVD with overall and relapse-free survival in all patients, including both node-negative and node-positive groups (see JNCI 84:1875-1887, 1992). For EGFR, although univariate analysis suggested that women with tumors showing EGFR expression relapsed earlier and, perhaps, died earlier, the differences were not statistically significant. Multivariate analysis, in contrast, revealed that determination of EGFR did provide significant additional information to that already provided by MVD for predicting relapse-free survival in all women (p = 0.001) and in the subset of node-positive women (p = 0.007). Among node-negative women, however, the contribution of EGFR expression in predicting relapse-free survival was not significant (p = 0.12). Likewise, EGFR measurement did not provide significant additional information beyond that of MVD alone for predicting overall survival. Thus, EGFR provides additional prognostic information to determination of MVD, but only for relapse-free survival in node-positive women with invasive breast carcinoma.

Squamous-cell carcinoma of the head and neck includes a heterogeneous group of tumours of the upper air and food passages for which prognosis is difficult to assess. In fact, patients in comparable stages may have diverse clinical courses and responses to similar treatments. In order to better define the prognosis of each patient there is therefore a need to identify novel biological markers which reflect more accurately growth rate, progression and metastatic potential of each tumour. We assessed whether metastases correlate with microvessel counts (i.e. intratumoral vascularity) using the CD-31 monoclonal antibody (MAb) and p53 mutant protein expression, determined in the primary by immunocytochemical methods in 70 patients with locally advanced head and neck cancer. Patients were treated with concurrent chemo-radiotherapy; 50 of these presented loco-regional node metastasis at diagnosis whereas 3 cases, initially node-negative, developed distant metastasis during the period of observation. No feature was predictive for objective response to treatment. The overall mean and median blood vessel density at "hot spots" was 37.42 and 36, respectively, and 57% of the tumours expressed p53 mutant proteins. These 2 biological markers were significantly associated. Patients with metastases (loco-regional and distant) had a significantly higher mean blood-vessel density than those without tumour spread. Also, patients with p53-positive (+/++) tumours had a significantly higher incidence of metastasis than those with negative ones. Multivariate analysis showed that both vascularity and stage, but not p53 expression, are significant and independent predictors of metastasis in this series.

It has been postulated that prostatic intraepithelial neoplasia (PIN) is a precursor lesion to invasive carcinoma in the prostate, yet there has been little direct evidence for this relationship. Therefore, we studied the DNA content of the epithelial cells in PIN lesions with accompanying invasive carcinoma in seven cases. DNA content of nuclei in Feulgen-stained tissue sections was performed using the CAS200 System (Cell Analysis Systems Inc, Elmhurst, Ill), with normal and hyperplastic glands serving as controls for diploid DNA content. In four cases, the cells of both lesions (PIN and invasive carcinoma) contained only diploid DNA; in one case, an additional aneuploid tumor stem line was present only in the invasive component; and in two cases, both PIN and carcinoma cells contained aneuploid cells having similar DNA content. The concordance of DNA content between PIN and invasive carcinoma in these cases provides evidence that PIN is a precursor lesion for invasive carcinoma.

The first requirement of a new prognostic indicator is that it should possess a clear biological significance. Indeed, much evidence shows that tumor growth and metastasis depend on neovascularization. Tumor angiogenesis (TA) refers to the growth of new vessels toward and within the tumor; unless tumor neovascularization occurs, cell proliferation reaches a steady state, and the tumor grows no larger than about 2 mm greatest diameter. Moreover, for tumor cells to metastasize, they must gain access to the vasculature from the primary tumor, survive the circulation, localize in the target organ, and induce angiogenesis in that target organ. TA is necessary both at the beginning and at the end of the metastatic cascade of events. Recently, my colleagues and I showed that a statistically significant correlation exists between incidence of metastases and microvessel density (MVD) of primary invasive breast carcinomas. Now, subsequent studies have shown that the association of prognosis with MVD exists not only in breast carcinoma but also in non-small-cell lung carcinoma, prostate carcinoma, and head-and-neck carcinoma. This article reviews the concepts and mechanisms of TA, the evidence supporting its role in growth and metastasis of solid tumors, and how measuring MVD within invasive tumors correlates with factor VIII-related antigen, blood vessel.

Tumor proliferation is inversely associated with survival in patients with breast carcinoma. Labeling of tumor cells with bromodeoxyuridine (BRDU) correlates highly with that seen with [3H]thymidine, the current "gold standard" for measuring tumor S-phase. However, the relationship of BRDU labeling to mitotic figure content and tumor grade remains incompletely defined. To determine this, we labeled 55 breast carcinomas with BRDU in vivo and correlated the results with mitotic figure content. The BRDU labeling index was the number of BRDU-positive cells/2,000 tumor cells, the mitotic figure index was the number of mitotic figures per 1,000 tumor cells, and the mitotic figure count was the number of mitotic figures per 10 high-powered fields. BRDU labeling was also correlated with tumor grade (Scarff-Bloom-Richardson). The BRDU labeling index correlated highly with the mitotic figure index (r = 0.814, p = 7.0 x 10(-14)), mitotic figure count (r = 0.725, p = 6.0 x 10(-10)), and tumor grade (r = 0.68, p = 1.1 x 10(-8)). The correlation of BRDU labeling with mitotic figure content was strong enough to suggest that a very carefully measured mitotic figure index provides an estimate of tumor growth fraction equivalent to the BRDU labeling index. Also, analysis of variance showed that the mitotic figure index was twice as precise as the mitotic figure count in estimating BRDU labeling, and thus was a more accurate measure of tumor proliferation. Moreover, measurements made by the mitotic figure index were as precise as those made by BRDU labeling. However, which method is optimal for estimating tumor proliferation rate remains unclear. Further studies are indicated.

Tumor growth and metastasis require angiogenesis; and microvessel density, a measure of tumor angiogenesis, correlates with metastasis in breast and lung carcinoma. To determine how microvessel density correlated with metastasis in prostate carcinoma, we counted microvessels within the initial invasive carcinomas of 74 patients (29 with metastasis, 45 without). Microvessels were highlighted by immunostaining endothelial cells for factor VIII-related antigen. Without knowledge of the patient's cancer stage, microvessels were counted in a 200 field (0.739 mm2) in the most active areas of neovascularization. The mean microvessel count in tumors from patients with metastases was 76.8 microvessels per 200 field (median, 66; standard deviation, 44.6). The counts within carcinomas from patients without metastasis were significantly lower, 39.2 (median, 36; standard deviation, 18.6) (P < 0.0001). Microvessel counts increased with increasing Gleason's score (P < 0.0001), but this increase was present predominantly in the poorly differentiated tumors. Although Gleason's score also correlated with metastasis (P = 0.01), multivariate analysis showed that Gleason's score added no additional information to that provided by microvessel count alone. Assay of microvessel density within invasive tumors may prove valuable in selecting patients for aggressive adjuvant therapies in early prostate carcinoma.

A 47-year-old man presented with a soft tissue mass of the distal right thigh near the knee. The tumor was highly vascular with epithelioid tumor cells growing in a peritheliomatous pattern, suggesting a soft tissue glomus tumor. Yet many tumor cells contained hemosiderin pigment and formed papillary structures suggestive of pigmented villonodular synovitis. Tumor cells were cytologically bland, and there was minimal mitotic activity. The tumor cells were strongly immunoreactive for cytokeratin, however, and contained true desmosomes, gland lumina, microvilli, tonofilaments, and well-developed basal lamina. These findings plus the absence perinuclear aggregates of intermediate filaments rule out malignant rhabdoid tumor and epithelioid sarcoma. Also, magnetic resonance imaging revealed no other lesions to suggest metastatic carcinoma. Thus this tumor appears to be a predominantly epithelioid form of monophasic synovial sarcoma. Recognition of this variant of synovial sarcoma is important for prognostication and therapeutic decision making because some studies indicate that this variant of synovial sarcoma follows a relatively benign clinical course.

A 15-year-old girl presented with a small, indolent mass near the knee joint. Light microscopy revealed a peculiar myxoid plexiform tumor composed of cytologically bland cells. Interpretation of the initial biopsy material was controversial. Subsequent immunohistochemical studies revealed tumor cells to be strongly reactive for epithelial membrane antigen (EMA) and negative for S-100 protein. Ultrastructural studies revealed tumor cells with long, thin, bipolar cell processes and discontinuous basal laminae. They had no epithelial or histiocytoid features. Admixed among the tumor cells were Schwann cells, but they represented a rare and scattered component of the overall cell population. These features are most consistent with a so-called perineurioma and contrast with those of plexiform neurofibroma and traumatic neuroma (two lesions strongly positive for S-100 protein and showing a distinctive EMA immunoreactivity pattern with focal peripheral staining of neural bundles).

The study was undertaken to determine the clinico-pathobiological characteristics in a series of 49 patients who developed metachronous breast carcinoma. Possible differences between the two tumours of conventional clinico-pathological features and of some biological markers such as DNA ploidy, c-erbB-2 oncoprotein overexpression and tumour angiogenesis were evaluated. The McNemar's test for independence showed that all the characteristics analyzed between the two tumours, in the same case, were not significantly different. After a median follow-up time of 69 months the overall survival of the series was 87.5% and the only significant prognostic factor for clinical outcome was peritumoural lymphatic vessel invasion (PLVI). The median second tumour-free interval was of 32 months ( 13 to 160 months) and none of the variables analyzed on the first primary was predictive of the timing of appearance of the second tumour. To assess the association between the characteristics of the first tumour and the odds of developing a metachronous carcinoma a case-control analysis was conducted. For each woman of the present series who developed bilateral cancer (case) a woman who had unilateral breast cancer (control) was matched for the length of the follow-up. A log-logistic regression model for matched sets was also performed to assess the risk of developing the second tumour. Applying multivariate analysis we found that progesterone receptor (PgR) status was the most important prognostic factor for the odds of bilateral tumour (odds ratio 0.22, p=0.013) followed by histological grade (odds ratio 0.20, p=0.063) and presence of PLVI (odds ratio 3.13, p=0.067). These findings suggest that the knowledge on the initial primary of PgR, grading and PLVI could be important to assess the individual risk of developing metachronous breast cancer. The determination of these factors could improve our ability to identify subsets of patients operated for breast cancer with different risks for bilateral tumour, allowing for a better selection of those patients who need intensive surveillance of their contralateral breast, and eligible for chemoprevention.

We describe a unique case of pure squamous carcinoma of the larynx that developed a cervical lymph node metastasis showing rhabdomyosarcoma admixed with squamous carcinoma (that is, carcinosarcoma). The rhabdomyosarcoma showed foci immunoreactive to multiple cytokeratin monoclonal antibodies, as well as to markers for striated muscle, thus indicating true divergent epithelial and rhabdomyosarcomatous differentiation. Although the morphogenesis of carcinosarcomas remains controversial, the sequence of events for the current case favors sarcomatous transformation of the original carcinoma (that is, sarcomatous neometaplasia of the primary carcinoma clone). The possible contributory role of radiation therapy in this case in inducing such a change is noted.

Human mast cells (MCs) from multiple sites were studied to determine heterogeneity of expression of chymase, tryptase, and/or carboxypeptidase and the binding to avidin. Three immunophenotypes were found: MCs positive for tryptase (MCT), no immunodetectable chymase; MCs positive for tryptase and chymase (MCTC); and MCs positive for chymase (MCC), no immunodetectable tryptase. Chymase-positive MCs also bound avidin and contained carboxypeptidase. In breast skin and parenchyma and axillary lymph nodes > 95% were MCTC; a rare MC in skin and lymph nodes was MCT or MCC. In lung alveoli 91% of MCs were MCT, 8% were MCTC and 1% MCC. In bowel mucosa 58% of MCs were MCT, 35% were MCTC and 7% MCC, whereas in bowel submucosa 83% of MCs were MCTC and 17% MCC. Within esophageal epithelium 38% were MCT and 62% MCTC; whereas, in esophageal subepithelium all were MCTC. This study further documents site-dependent diversity among normal human MCs and recognizes a novel third immunophenotype rich in chymase and relatively deficient in tryptase, the MCC cell.

Background: Axillary lymph node status has been the most important prognostic factor in operable breast carcinoma, but it does not fully account for the varied disease outcome. More accurate prognostic indicators would help in selection of patients at high risk for disease recurrence and death who are candidates for systemic adjuvant therapy. Our recent findings indicated that microvessel density (count or grade) in invasive breast carcinoma (a measure of tumor angiogenesis) is associated with metastasis and thus may be a prognostic indicator., Purpose: This study was designed to further define the relationship of microvessel density with overall and relapse-free survival and with other reported prognostic indicators in breast carcinoma., Methods: In a prospective, blinded study of 165 consecutive patients, the microvessels within primary invasive breast carcinoma were highlighted by immunocytochemical staining to detect factor VIII-related antigen. Using light microscopy, we counted microvessels per 200x field in the most active areas of neovascularization and graded microvessel density. These findings were correlated, by univariate and multivariate analyses, with overall and relapse-free survival, axillary node status, and other prognostic indicators (median follow-up, 51 months)., Results: There was a highly significant (P < or = .001) association of microvessel density with overall survival and relapse-free survival in all patients, including node-negative and node-positive subsets. All patients with breast carcinomas having more than 100 microvessels per 200x field experienced tumor recurrence within 33 months of diagnosis, compared with less than 5% of the patients with breast carcinoma having 33 or fewer microvessels per 200x field. Moreover, microvessel density was the only statistically significant predictor of overall survival among node-negative women (P < .001). Only microvessel density (P < .001) and histologic grade (P = .04) showed statistically significant correlations with relapse-free survival in the node-negative subset., Conclusions: Microvessel density in the area of the most intense neovascularization in invasive breast carcinoma is an independent and highly significant prognostic indicator for overall and relapse-free survival in patients with early-stage breast carcinoma (I or II by International Union Against Cancer criteria)., Implications: Such an indicator would be useful in selection of those node-negative patients with breast carcinoma who are at high risk for having occult metastasis at presentation. These patients could then be given systemic adjuvant therapy.

Infiltrating lobular carcinoma (ILC) of the classic type is well recognized; less well appreciated is a group of variant forms of ILC, which includes solid, alveolar, mixed, apocrine, signet-ring, histiocytoid, and tubulolobular variants. In addition, Page et al (Diagnostic Histopathology of the Breast, Churchill Livingstone, 1987, pp 219-226) recently have emphasized the pleomorphic variant of ILC, which demonstrates the infiltrating pattern of classic ILC; however, the nuclei are more pleomorphic and have features that may overlap with those of infiltrating duct carcinoma. To determine whether pleomorphic ILC shows significant prognostic differences from classic ILC, we reviewed the clinical courses of 25 patients with classic ILC and compared them with 16 patients with pleomorphic ILC. The major determinant in placing ILC into the pleomorphic category was the presence of nuclei of nuclear grade 2 or 3. All classic ILCs had nuclei of grade 1 by the Scarff-Bloom-Richardson criteria. Survival to recurrence was significantly worse (P less than or equal to .05) for the patients with pleomorphic ILC at lengths of survival greater than 30 months. In addition, node-negative patients with pleomorphic ILC were four times more likely to experience recurrence than node-negative patients with classic ILC; those with positive nodes and pleomorphic histology were 30 times more likely to experience recurrence. Although there appeared to be a trend toward decreased overall length of survival for those patients with pleomorphic ILC when compared with patients with classic ILC, this difference was not statistically significant.

The factor(s) that causes excessive mast cell (MC) proliferation in indolent forms of mastocytosis is not known, nor is it known whether that proliferation is a regulated clonal expansion or merely a non-neoplastic hyperplasia. Human MCs display phenotypes that depend on the microenvironment. Thus, if the phenotype of MCs in mastocytosis lesions is determined to be abnormal for that tissue site (and therefore the MCs are refractory to microenvironmental signals) then a clonal process would be suggested. The authors determined the phenotypes of MCs from the lesional skin of 17 patients with indolent mastocytosis and the bone marrows of 9 patients. They compared them with the phenotypes of MCs from the lesional skin of 8 patients with various dermatitides, the skin of 2 patients with idiopathic anaphylaxis, and the breast skin of 15 control patients. MCs from all the skin specimens showed the characteristic skin MC phenotype, with predominantly scroll-poor granules by ultrastructure and containing tryptase and chymase by immunofluorescence detection (the MCTC immunophenotype). The skin MCs of each patient bound avidin and contained carboxypeptidase by immunofluorescence detection. MCs from the bone marrow of patients with indolent mastocytosis, the source of progenitors, also showed the scroll-poor and MCTC phenotypes. These findings do not support an unregulated clonal expansion in indolent forms of mastocytosis. They are consistent with a non-neoplastic hyperplasia or possibly a clonal process in which MCs remain responsive to microenvironmental regulation.

A 44-year-old woman with a long history of smoking presented with a single lung mass detected on routine chest radiographs. There were no other known primary tumors. Light microscopy showed a mucin-producing adenocarcinoma of bronchioloalveolar type. Furthermore, no gastrointestinal primary has been discovered 18 months after discovery of the lung tumor. Electron microscopy revealed that tumor cells had microvilli containing abundant microfilamentous, dense-core rootlets that showed long extensions into apical cytoplasm. The microvilli were capped by numerous glycocalyceal bodies. These findings are similar to the ultrastructural features previously described as being specific for gastrointestinal adenocarcinomas metastatic to lung. The current case, however, suggests that the findings of glycocalyceal bodies and microvillous dense-core rootlets are not helpful in distinguishing between primary and metastatic lung adenocarcinomas.

The mammographic and histologic findings were reviewed in 41 consecutive cases of isolated lobular carcinoma in situ (LCIS) unassociated with any malignant diagnosis. Thirty-one needle-directed breast biopsies were performed to evaluate clustered microcalcifications. In 24 of the 31 cases, the calcifications were found in areas of benign breast disease, with LCIS representing a separate process. In the few cases in which microcalcifications were seen in association with LCIS, a greater number of similar calcifications were present in adjacent benign disease. Soft-tissue abnormalities necessitating the performance of a biopsy represented benign foci, except in one patient with LCIS in and adjacent to a fibroadenoma. The authors conclude that LCIS has no characteristic mammographic features. LCIS is detected as an incidental finding at breast biopsy, with the mammographic abnormality predominantly reflecting a benign process.

Cytogenetic analysis of a low-grade metastatic endometrial stromal sarcoma in a 58-year-old woman revealed translocations involving both homologues of chromosome 7 with chromosomes 13 and 17, respectively, and an interstitial deletion of the long arm of chromosome 11. The karyotype of the tumor was 46,XX,t(7;13)(q11.1;p13),t(7;17)(p21;q12),del(11)(q13q21).

Chromosomal analysis of two pulmonary metastases from a 25-year-old male with alveolar soft-part sarcoma of the right lower extremity revealed multiple and complex chromosomal abnormalities. The rearrangements included deletions and translocations affecting chromosomes 1, 2, 3, 10, 11, 14, 15, and 16. Trisomy of chromosome 12 and loss of chromosome 17 was also observed.

A 46-year-old man presented with a cytologically bland testicular tumor composed of spindle cells that showed both epitheliallike (ie, true desmosomes and tonofilamentlike structures) and myogenous differentiation (ie, thin filaments with focal densities and alpha-smooth muscle actin immunoreactivity). Tumor cells were immunoreactive for vimentin and S-100 protein but negative for cytokeratin and desmin. Peritubular myoid cells are present in the normal testis; contain subplasmalemmal micropinocytotic vesicles; show thin filaments with focal densities; and are reactive with desmin, vimentin, and alpha-smooth muscle actin. They have no desmosomes and lie outside the basement membrane of the seminiferous tubules; thus they are not true myoepithelial cells (a cell type not present in the testis). Paradoxically, the current tumor appeared to show bidirectional differentiation, mimicking both a peritubular myoid spindle cell and an epitheliallike cell (possibly similar to the granulosa cell or rete testis epithelial cell). Although the findings suggest myoepithelial differentiation, the cytogenesis of this tumor remains uncertain.

A 57-year-old man with malaise, ascites, and abdominal pain was found to have a peritoneum studied with numerous, small nodular tumor masses. Light microscopy revealed an anaplastic malignant tumor of uncertain differentiation. Mucin stains were negative. Electron microscopy revealed pleomorphic tumor cells with diffusely distributed cytoplasmic tonofilaments and well-developed true desmosomes. No long, thin, branching microvilli were present, yet tumor cells were strongly positive for both callus keratin (polyclonal) and monoclonal cytokeratin (AE1/3) in a diffuse cytoplasmic distribution (a pattern corresponding to the diffuse cytoplasmic tonofilaments). Tumor cells were negative for Leu-M1 and carcinoembryonic antigen. The findings were most consistent with malignant mesothelioma, and additional questioning, after tissue diagnosis, revealed a work history of asbestos exposure.

Oncogenic or tumor-induced osteomalacia-rickets is a syndrome characterized by hypophosphatemia, renal phosphate wasting, and decreased serum 1,25-dihydroxyvitamin D3 levels. The tumors secrete a phosphaturic substance that causes total body phosphate depletion, leading to osteomalacia or rickets. Although the tumors are histologically polymorphous, personal review of 16 tumors documented to cause this syndrome revealed four morphologic patterns. The first contained 10 unique-appearing, mixed connective tissue tumors having variably prominent vessels, osteoclastlike giant cells, focal microcystic changes, dystrophic calcification, osseous metaplasia, and/or poorly developed cartilagelike areas. With one exception, all tumors of this group occurred in soft tissue and were benign. The single malignant tumor originated in bone, recurred locally, and metastasized to lung. The remaining tumors occurred in bone and showed benign clinical behavior. They resembled tumors known to occur in bone, that is osteoblastomalike (3 tumors), nonossifying fibromalike (2 tumors), and ossifying fibromalike (1 tumor).

A 66-year-old woman presented with a superior-anterior mediastinal mass that contained considerable calcification by computed tomography. Initial biopsy interpretation was inconclusive but suggested a giant cell sarcoma of soft parts because the tumor contained atypical epithelioid cells and osteoclastlike giant cells. After radiation and chemotherapy, the tumor was removed. It contained large epithelioid cells, pale chondroid areas, metaplastic bone, and osteoclastlike giant cells. Immunohistochemical and electron microscopic studies revealed the epithelioid cells to be factor VIII positive and to contain abundant intermediate filaments, micropinocytotic vesicles, Weibel-Palade bodies, basal lamina, and primitive cell-cell junctions. These epithelioid cells also stained for muscle-specific actin (HHF-35), yet were negative for desmin and cytokeratins (CAM 5.2 and AE1/3). The findings were most consistent with those reported to occur in epithelioid hemangioendothelioma.

A 62-year-old woman presented with an asymptomatic anterior mediastinal mass. Clinically considered to be a thymoma, the tumor was solid, firm, and composed of hypocellular dense collagen and cytologically bland, spindled fibroblastlike cells growing in a patternless pattern. There was no apparent connection to pleura or pericardium, yet the clinicopathologic features clearly fit with solitary fibrous tumor (SFT) of mediastinum. SFTs occur most commonly in pleura but have been reported in other locations, including the mediastinum, where aggressive behavior has been more common when these tumors are compared to those occurring in pleura. Although it is difficult to predict behavior for all cases of SFT occurring in the mediastinum by cytologic features alone, morphologic criteria for benign and malignant forms have been described. Roughly half the malignant forms will progress, yet the single most important indicator of clinical outcome is whether the tumor can be initially totally excised.

Nuclear matrix proteins (NMP) are nonhistone proteins found in the nucleus of many eukaryotic cells. Furthermore certain NMPs are reported to be cell-type specific and expressed differentially by malignant cells. To study the specificity of NM-200.4 (an antibody reactive to NMPs extracted from cultured breast carcinoma cells of the T-47D line), cancers and benign tissues from multiple body sites were surveyed. All 17 breast carcinomas showed strong reactivity to tumor cell nuclei. Also nuclei from one of two lung carcinomas, a papillary thyroid carcinoma, an ovarian fibroma, and a lymphoma were strongly reactive. One leiomyosarcoma and a dermoid cyst were negative. Although 1 benign breast with duct hyperplasia showed moderate reactivity, only 1 of 10 benign breast biopsies without hyperplasia showed reactivity. Three of 4 skin biopsies, 2 liver biopsies, 6 of 9 kidney biopsies, and 5 of 10 gastrointestinal mucosal biopsies showed reactivity in benign nuclei. It is concluded that, although breast carcinoma nuclei showed the most consistent reactivity for NM-200.4, both benign and malignant nuclei from other body sites also show reactivity.

Cytogenetic analysis of two pulmonary chondroid hamartomas and nine breast adenofibromas revealed clonal chromosome aberrations in both hamartomas and in four breast tumors. To determine lineage of the cells with chromosome aberrations, a combined immunohistochemical/cytogenetic approach was developed that enabled simultaneous ascertainment of cytogenetic aberrations and immunohistochemical features in individual cells. Immunohistochemical/cytogenetic evaluation of one hamartoma and two adenofibromas demonstrated that neoplastic proliferation, in each case, was confined to the mesenchymal (stromal) component, whereas epithelial cells appeared to be reactive. Cytogenetically abnormal short-term cultures of the remaining hamartoma and another of the breast adenofibromas were composed entirely of mesenchymal elements, indicating mesenchymal clonality in those tumors as well. Our findings support redesignation of pulmonary chondroid hamartomas as 'pulmonary chondromas' and suggest that carcinomas developing within fibroadenomas arise from reactive epithelial proliferation. Combined immunohistochemical/cytogenetic analysis might be useful in the development of novel therapeutic approaches that selectively target neoplastic populations within solid tumors.

Granulomatous prostatitis and poorly differentiated prostate carcinoma can mimic each other both clinically and histologically. To develop criteria useful in resolving problem cases, the authors compared the reactivities of these conditions (nine cases of granulomatous prostatitis and six cases of poorly differentiated carcinoma) with a panel of antibodies to cytokeratin (AE1/3), prostate-specific antigen (PSA), prostatic acid phosphatase (PAP), lysozyme, antimacrophage M, and leukocyte common antigen (LCA). In granulomatous prostatitis, histiocytes were not immunoreactive for PAP, PSA, or cytokeratin; however, histiocytes reacted to lysozyme in nine of nine cases, antimacrophage M in seven of nine cases, and LCA in one of nine cases. Tumor cells from all six carcinoma cases reacted with PAP, PSA, and cytokeratin; all failed to react with lysozyme, LCA, and antimacrophage M. The authors conclude that granulomatous prostatitis and poorly differentiated carcinoma can be reliably distinguished with immunohistochemical methods.

This article reviews the ultrastructural and immunohistochemical features of normal human mast cells (MC) at multiple tissue sites. Current literature indicates that granules containing discrete scrolls (scroll-rich morphology) are frequent in MC from bowel mucosa and lung, locations where the majority of MC show only tryptase immunoreactivity (MCT). In contrast, most MC from skin, breast parenchyma, axillary lymph nodes, and bowel submucosa are characterized by scroll-poor morphology (that is, granules are rimmed by incomplete scrolls forming parallel lamellae and containing central, amorphous granular material or grating/lattice-like structures) and show both tryptase and chymase immunoreactivity (MCTC). MC having granules with both scroll-rich and scroll-poor features can occur in all tissue sites, and an occasional MC, especially in lung and bowel, may show only chymase immunoreactivity (MCC). Chymase immunoreactivity in MC also is closely associated with avidin binding and carboxypeptidase reactivity. We conclude that there is ultrastructural and immunophenotypic diversity among normal human MC, although certain forms predominate in specific tissue environments. In skin, breast tissue, axillary lymph nodes, and bowel submucosa MC tend to have scroll-poor granules and stain for avidin, chymase, tryptase, and carboxypeptidase, whereas, in lung and bowel mucosa MC granules tend to be scroll-rich and stain only for tryptase with currently available reagents.

This article reviews the ultrastructural and immunohistochemical features of normal human mast cells (MC) at multiple tissue sites. Current literature indicates that granules containing discrete scrolls (scroll-rich morphology) are frequent in MC from bowel mucosa and lung, locations where the majority of MC show only tryptase immunoreactivity (MC(T)). In contrast, most MC from skin, breast parenchyma, axillary lymph nodes, and bowel submucosa are characterized by scroll-poor morphology (that is, granules are rimmed by incomplete scrolls forming parallel lamellae and containing central, amorphous granular material or grating/lattice-like structures) and show both tryptase and chymase immunoreactivity (MC(TC)). MC having granules with both scroll-rich and scroll-poor features can occur in all tissue sites, and an occasional MC, especially in lung and bowel, may show only chymase immunoreactivity (MC(C)). Chymase immunoreactivity in MC also is closely associated with avidin binding and carboxypeptidase reactivity. We conclude that there is ultrastructural and immunophenotypic diversity among normal human MC, although certain forms predominate in specific tissue environments. In skin, breast tissue, axillary lymph nodes, and bowel submucosa MC tend to have scroll-poor granules and stain for avidin, chymase, tryptase, and carboxypeptidase, whereas, in lung and bowel mucosa MC granules tend to be scroll-rich and stain only for tryptase with currently available reagents.

Background: Malignant soft-tissue tumors often present substantial diagnostic challenges. Chromosome aberrations that might be diagnostic have been identified in some types of soft-tissue tumors, but the overall frequency and diagnostic relevance of these aberrations have not been established., Methods: We attempted to determine the karyotypes of a series of 62 consecutive, unselected malignant spindle-cell or small round-cell soft-tissue tumors (from 46 adults and 16 children) after direct harvesting of cells or short-term culture. All tumors were examined independently by immunohistochemical staining in addition to routine light-microscopical evaluation, and all but two tumors were examined by electron microscopy., Results: Metaphases were obtained from 61 of the 62 tumors, and clonal chromosome aberrations were identified in 55 (89 percent). In the six tumors that yielded metaphases but lacked apparent clonal aberrations, the normal metaphases were found to originate from non-neoplastic stromal elements within the tumor specimens. Thus, all tumors in which karyotyping was successful contained clonal chromosome aberrations. Forty of 62 tumors (65 percent) contained clonal chromosome aberrations that either suggested or confirmed a specific diagnosis; in 15 of these tumors (24 percent of all tumors), the aberrations were important in establishing the final diagnosis. Cytogenetic analyses were particularly informative about small round-cell tumors from children: 8 of 14 round-cell tumors contained diagnostically important chromosome aberrations. Using the combined approaches of light and electron microscopy, immunohistochemistry, and cytogenetics, we established an unambiguous diagnosis for 60 of 62 tumors., Conclusions: Cytogenetic analyses reveal clonal chromosome aberrations in virtually all malignant soft-tissue tumors. These clonal chromosome aberrations, particularly in small round-cell tumors in children, often have diagnostic relevance.

Background: Experimental evidence suggests that the growth of a tumor beyond a certain size requires angiogenesis, which may also permit metastasis. To investigate how tumor angiogenesis correlates with metastases in breast carcinoma, we counted microvessels (capillaries and venules) and graded the density of microvessels within the initial invasive carcinomas of 49 patients (30 with metastases and 19 without)., Methods: Using light microscopy, we highlighted the vessels by staining their endothelial cells immunocytochemically for factor VIII. The microvessels were carefully counted (per 200x field), and their density was graded (1 to 4+), in the most active areas of neovascularization, without knowledge of the outcome in the patient, the presence or absence of metastases, or any other pertinent variable., Results: Both microvessel counts and density grades correlated with metastatic disease. The mean (+/- SD) count and grade in the patients with metastases were 101 +/- 49.3 and 2.95 +/- 1.00 vessels, respectively. The corresponding values in the patients without metastases were significantly lower--45 +/- 21.1 and 1.38 +/- 0.82 (P = 0.003 and P less than or equal to 0.001, respectively). For each 10-microvessel increase in the count per 200x field, there was a 1.59-fold increase in the risk of metastasis (95 percent confidence interval, 1.19 to 2.12; P = 0.003). The microvessel count and density grade also correlated with distant metastases. For each 10-microvessel increase in the vessel count per 200x field, there was a 1.17-fold increase in the risk of distant metastasis (95 percent confidence interval, 1.02 to 1.34; P = 0.029)., Conclusions: The number of microvessels per 200x field in the areas of most intensive neovascularization in an invasive breast carcinoma may be an independent predictor of metastatic disease either in axillary lymph nodes or at distant sites (or both). Assessment of tumor angiogenesis may therefore prove valuable in selecting patients with early breast carcinoma for aggressive therapy.

Nucleolar organizer regions (NOR) are DNA loops encoding ribosomal RNA production. Detectable by the argyrophilia (AgNOR) of associated proteins, AgNOR numbers correlate with growth fraction and may have diagnostic and prognostic utility in human tumors. Because nucleolar size is important in the diagnosis of prostate carcinoma, we compared AgNOR counts in benign, atypical, and malignant prostate lesions and correlated them with nucleolar diameter. We counted AgNOR in benign prostatic hyperplasia, atypical adenomatous hyperplasia, intraductal dysplasias, and carcinomas of various Gleason grades. The mean AgNOR count per nucleus for benign prostatic hyperplasia nuclei was 4.51; for atypical adenomatous hyperplasia, 5.64; for intraductal dysplasias, 7.35; for carcinoma of Gleason grades 2 + 3, 8.87; and for carcinoma Gleason grades 4 + 5, 10.42. Counts in the carcinomas and intraductal dysplasias were significantly different from those of benign prostatic hyperplasia. Mean AgNOR counts of the carcinomas and intraductal dysplasias did not overlap with those of benign prostatic hyperplasia, suggesting that intraductal dysplasia shares more with carcinoma than benign prostatic hyperplasia. Nucleolar diameters increased from benign prostatic hyperplasia to atypical adenomatous hyperplasia, intraductal dysplasias, and the carcinomas, correlating with increasing AgNOR counts. Nucleolar diameters in the carcinomas were significantly different than those of benign prostatic hyperplasia; those in intraductal dysplasias were not. Our findings suggest that AgNOR counts are superior to nucleolar diameters alone in separating intraductal dysplasias from benign prostatic hyperplasia but provide no additional information in diagnosing carcinoma.

The ultrastructural features of 502 mast cells, including 34,187 granules, were studied from seven human tissue sites (lung, skin, colon, stomach, small bowel, breast parenchyma, and axillary lymph nodes). Granule areas and substructure were detailed according to the microenvironment (lung alveoli, lung bronchi, bowel mucosa, bowel submucosa, breast skin, breast parenchyma, and axillary lymph nodes). Although completely closed, discrete scrolls were present in some mast cell granules at all tissue locations, they were present more frequently in granules from bowel mucosa and lung (scroll-rich morphology). In contrast, most of the mast cell granules from skin, breast parenchyma, axillary lymph nodes, and bowel submucosa were rimmed by incomplete scrolls forming parallel lamellae; centrally, amorphous granular material and/or grating/lattice-like structures occurred (scroll-poor morphology). In breast, the latter granules had a mean granule area almost twice that of all other sites. Individual mast cells having granules with both scroll-rich and scroll-poor features were common in all tissue sites and microenvironments. In fact, 7.8% of mast cells had granules showing at least one completely closed, discrete scroll and grating/lattice-like structures, sometimes within the same granule. These observations indicate that mixed forms occur and that there is considerable morphologic diversity between the predominant mast cell found in mucosa and lung (scroll-rich morphology) and the predominant mast cell found in skin, breast parenchyma, axillary lymph nodes, and bowel submucosa (scroll-poor morphology).

Biopsies of lesional and nonlesional skin from 14 patients with localized cutaneous or associated systemic mastocytosis were examined by ultrastructural and immunohistochemical techniques. Mast cells within lesions of the dermis were highly variable between patients with regard to cell number and extent of degranulation, although lesional sites consistently contained more mast cells than did nonlesional sites. Two mast cell patterns were identified based upon granule morphology. In biopsies from 8 patients, the majority of granules contained electron-dense amorphous zones; crystalline lattices; and indistinct, incomplete solid scrolls forming parallel lamellae. In biopsies from 6 patients, in addition to these granules, there were also granules composed of electron-dense amorphous zones, reticulated matrices, and/or distinct scrolls with lucent cores interrupted by dense spheres. The granule morphology for the first group (N = 8) was identical with that seen in the preponderant type of skin mast cell of 6 normal control subjects, whereas the granule morphology of the second group (N = 6) displayed an abnormal ultrastructural phenotype for skin that included granule types normally found not only in skin but also in intestinal lamina propria and lung. For individual patients, the patterns of granule ultrastructure were consistent between clinically nonlesional and lesional skin. A minority of cells in both patient groups appeared primitive ultrastructurally, exhibiting rudimentary, Golgi-associated progranules; monocyte-like morphologic characteristics; and mitotic activity. Moreover, when mast cells in lesional skin were screened for a limited panel of surface antigens, they displayed common patterns of reactivity (M718+, HLA-DR/DQ+, CD4+), and in a selected case, immunoelectron microscopy confirmed the presence of these antigens on mast cell plasma membranes. Dermal mast cells from normal donors (N = 6) lack these epitopes. These observations suggest that infiltrates in cutaneous mastocytosis may exhibit phenotypic characteristics not only of cutaneous mast cells, but in some patients also of mucosal mast cells. In either circumstance, the mast cells may display antigenic determinants common to monocyte/macrophages. Concordance of granule phenotype between lesional and clinically uninvolved skin of individual patients furthers the notion that even localized mastocytosis reflects covertly defective systemic mast cell homeostasis.