Fungal-derived cues promote ocular autoimmunity through a Dectin-2/Card9-mediated mechanism.

Uveitis (intraocular inflammation) is a leading cause of loss of vision. Although its aetiology is largely speculative, it is thought to arise from complex genetic-environmental interactions that break immune tolerance to generate eye-specific autoreactive T cells. Experimental autoimmune uveitis (EAU), induced by immunization with the ocular antigen, interphotoreceptor retinoid binding protein (IRBP), in combination with mycobacteria-containing complete Freund's adjuvant (CFA), has many clinical and histopathological features of human posterior uveitis. Studies in EAU have focused on defining pathogenic CD4+ T cell effector responses, such as those of T helper type 17 (Th17) cells, but the innate receptor pathways precipitating development of autoreactive, eye-specific T cells remain poorly defined. In this study, we found that fungal-derived antigens possess autoimmune uveitis-promoting function akin to CFA in conventional EAU. The capacity of commensal fungi such as Candida albicans or Saccharomyces cerevisae to promote IRBP-triggered EAU was mediated by Card9. Because Card9 is an essential signalling molecule of a subgroup of C-type lectin receptors (CLRs) important in host defence, we evaluated further the proximal Card9-activating CLRs. Using single receptor-deficient mice we identified Dectin-2, but not Mincle or Dectin-1, as a predominant mediator of fungal-promoted uveitis. Conversely, Dectin-2 activation by α-mannan reproduced the uveitic phenotype of EAU sufficiently, in a process mediated by the Card9-coupled signalling axis and interleukin (IL)-17 production. Taken together, this report relates the potential of the Dectin-2/Card9-coupled pathway in ocular autoimmunity. Not only does it contribute to understanding of how innate immune receptors orchestrate T cell-mediated autoimmunity, it also reveals a previously unappreciated ability of fungal-derived signals to promote autoimmunity.

Mesenchymal Stromal Cells are Readily Recoverable from Lung Tissue, but not the Alveolar Space, in Healthy Humans.

Stromal support is critical for lung homeostasis and the maintenance of an effective epithelial barrier. Despite this, previous studies have found a positive association between the number of mesenchymal stromal cells (MSCs) isolated from the alveolar compartment and human lung diseases associated with epithelial dysfunction. We hypothesised that bronchoalveolar lavage derived MSCs (BAL-MSCs) are dysfunctional and distinct from resident lung tissue MSCs (LT-MSCs). In this study, we comprehensively interrogated the phenotype and transcriptome of human BAL-MSCs and LT-MSCs. We found that MSCs were rarely recoverable from the alveolar space in healthy humans, but could be readily isolated from lung transplant recipients by bronchoalveolar lavage. BAL-MSCs exhibited a CD90Hi , CD73Hi , CD45Neg , CD105Lo immunophenotype and were bipotent, lacking adipogenic potential. In contrast, MSCs were readily recoverable from healthy human lung tissue and were CD90Hi or Lo , CD73Hi , CD45Neg , CD105Int and had full tri-lineage potential. Transcriptional profiling of the two populations confirmed their status as bona fide MSCs and revealed a high degree of similarity between each other and the archetypal bone-marrow MSC. 105 genes were differentially expressed; 76 of which were increased in BAL-MSCs including genes involved in fibroblast activation, extracellular matrix deposition and tissue remodelling. Finally, we found the fibroblast markers collagen 1A1 and α-smooth muscle actin were increased in BAL-MSCs. Our data suggests that in healthy humans, lung MSCs reside within the tissue, but in disease can differentiate to acquire a profibrotic phenotype and migrate from their in-tissue niche into the alveolar space. Stem Cells 2016;34:2548-2558.

Identification of a Sacral, Visceral Sensory Transcriptome in Embryonic and Adult Mice.

Visceral sensory neurons encode distinct sensations from healthy organs and initiate pain states that are resistant to common analgesics. Transcriptome analysis is transforming our understanding of sensory neuron subtypes but has generally focused on somatic sensory neurons or the total population of neurons in which visceral neurons form the minority. Our aim was to define transcripts specifically expressed by sacral visceral sensory neurons, as a step towards understanding the unique biology of these neurons and potentially leading to identification of new analgesic targets for pelvic visceral pain. Our strategy was to identify genes differentially expressed between sacral dorsal root ganglia (DRG) that include somatic neurons and sacral visceral neurons, and adjacent lumbar DRG that comprise exclusively of somatic sensory neurons. This was performed in adult and E18.5 male and female mice. By developing a method to restrict analyses to nociceptive Trpv1 neurons, a larger group of genes were detected as differentially expressed between spinal levels. We identified many novel genes that had not previously been associated with pelvic visceral sensation or nociception. Limited sex differences were detected across the transcriptome of sensory ganglia, but more were revealed in sacral levels and especially in Trpv1 nociceptive neurons. These data will facilitate development of new tools to modify mature and developing sensory neurons and nociceptive pathways.

Predictive markers in breast cancer--the present.

Breast cancer is a heterogeneous disease and there is a continual drive to identify markers that will aid in predicting prognosis and response to therapy. To date, relatively few markers have established prognostic power. Oestrogen receptor (ER) is probably the most powerful predictive marker in breast cancer management, both in determining prognosis and in predicting response to hormone therapies. Progesterone receptor (PR) is also a widely used marker, although its value is less well established. HER-2 status has also become a routine prognostic and predictive factor in breast cancer. Given the importance of these biological markers in patient management, it is essential that assays are robust and quality controlled, and that interpretation is standardized. Furthermore, it is important to be aware of the limitations in their predictive power, and how this may be refined through addition of further biological markers. The aim of this review is to provide an overview of the established role of ER, PR and HER-2 in patient management, the current standards for assessing these markers, as well as highlighting the controversies that still surround their use and methods of assessment.

Methylene blue dye versus combined dye-radioactive tracer technique for sentinel lymph node localisation in early breast cancer.

AIM: The study compared the accuracy and success rate of two techniques, methylene blue alone versus combined methylene blue and radioactive colloid in sentinel lymph node localisation in the management early breast cancer. METHODS: Three hundred and twenty-nine patients with tumours less than 2 cm on ultrasound assessment were prospectively evaluated. One hundred and seventy-three patients (Group A) underwent sentinel lymph node localisation using 1 ml of 1% methylene blue. A combined technique of both methylene blue and radioactive colloid was used in 156 patients (Group B). Application of both was subdermal and subareolar. Sentinel lymph nodes were examined by standard microscopy. Patients underwent breast conservation surgery or mastectomy and sentinel node guided four node axillary sampling+/-clearance. RESULTS: In Group A, the sentinel lymph node identification rate was 96.5%. The negative predictive value was 96.3%, with false negative of 3.7% and accuracy of 87.4%. In group B the identification rate for sentinel lymph node was 98.7%, with false negative of 4.1%, negative predictive value of 96%, and accuracy of 83.8%. CONCLUSION: Sentinel lymph node localisation using methylene blue or combined dye and radioactive tracer technique predicts the axillary lymph node status in early breast cancer with comparable success rates, accuracy and false negative rates. The combined technique facilitates quicker identification of sentinel lymph node; however the dye technique alone can be used successfully in centres without nuclear medicine facilities.

Prognostic impact of macrometastasis linear size in sentinel node biopsy for breast carcinoma.

AIM: The aim of the present study was to evaluate the risk of axillary non-sentinel lymph-node metastases (ALN) in breast cancer patients presenting macrometastasis (Mac-m) in the sentinel lymph node (SN). MATERIALS AND METHODS: A retrospective series of 1464 breast cancers from patients who underwent ALN dissection following the diagnosis of Mac-m in the sentinel node (SN) was studied. In all the cases the MAC-m linear size was evaluated and correlated with presence or absence of non-SN ALN metastases. RESULTS: Non-SN metastases were detected in 644∖1464 cases (43.98%). The risk of further axillary metastases ranged from 20.2% (37/183) in cases with Mac-m between 2 and 2.9 mm, to 65.3% (262/401) in cases with Mac-m measuring > 10 mm. The risk of non-SN ALN metastases showed a 3% increase, parallel to each mm increment in SN metastasis size. The data evaluated with the receiver operating characteristic (ROC) curve showed that the Mac-m could be subdivided according to a new cut-off of 7 mm. pT1 tumours, with Mac-m < 7 mm had a risk of non-SN ALN metastases of <30%. Furthermore 109/127 of these (85.8%) had 3 or less non-SN ALN -metastases. CONCLUSIONS: The present data give a detailed description on the risk of non-SN ALN involvement, that may be useful in the evaluation of breast cancer patients. It is suggested that a Mac-m size of <7 mm is related to a low residual axillary disease burden in breast cancer patients with small (pT1) tumours.

Consistency in recognizing microinvasion in breast carcinomas is improved by immunohistochemistry for myoepithelial markers.

Microinvasion is the smallest morphologically identifiable stage of invasion. Its presence and distinction from in situ carcinoma may have therapeutic implications, and clinical staging also requires the recognition of this phenomenon. Microinvasion is established on the basis of several morphological criteria, which may be difficult and not perfectly reproducible among pathologists. The aim of this study was to assess the consistency of diagnosing microinvasion in the breast on traditional haematoxylin and eosin (HE) stained slides and to evaluate whether immunohistochemistry (IHC) for myoepithelial markers could improve this. Digital images were generated from representative areas of 50 cases stained with HE and IHC for myoepithelial markers. Cases were specifically selected from the spectrum of in situ to microinvasive cancers. Twenty-eight dedicated breast pathologists assessed these cases at different magnifications through a web-based platform in two rounds: first HE only and after a washout period by both HE and IHC. Consistency in the recognition of microinvasion significantly improved with the use of IHC. Concordance rates increased from 0.85 to 0.96, kappa from 0.5 to 0.85, the number of cases with 100% agreement rose from 9/50 to 25/50 with IHC and the certainty of diagnosis also increased. The use of IHC markedly improves the consistency of identifying microinvasion. This corroborates previous recommendations to use IHC for myoepithelial markers to clarify cases where uncertainty exists about the presence of microinvasion. Microinvasive carcinoma is a rare entity, and seeking a second opinion may avoid overdiagnosis.

Characterization of ovine growth hormone (oGH) and ovine placental lactogen (oPL) binding to fetal and adult hepatic tissue in sheep: evidence that oGH and oPL interact with a common receptor.

The GH receptor (GHR) plays a key role in postnatal growth regulation. Although plasma concentrations of GH are high during fetal life, its role during fetal development is not well understood. Recent data suggest that GHR are present in fetal hepatic tissue as early as 51 days gestation. However, the levels of GHR expression are markedly lower in fetal hepatic tissue compared to postnatal values, and there are conflicting data suggesting that ovine placental lactogen (oPL) and oGH share a common receptor. Given the uncertainty about whether oPL acts via the oGHR or a distinct receptor, we performed ligand binding and affinity cross-linking studies on hepatic microsomal membranes from adult castrated male, pregnant female, and fetal sheep. Ligand binding assays at a constant concentration of membranes showed that [125I]oPL yielded consistently higher (P < 0.001) specific binding (59.5 +/- 6.4%, 30.5 +/- 5.7%, and 7.6 +/- 2.4% for castrated male, pregnant female, and fetal sheep, respectively) compared to [125I]oGH (17.8 +/- 4.7%, 5.0 +/- 1.6%, and 1.2 +/- 0.4% for castrated male, pregnant female, and fetal sheep, respectively). Cross-reactivity studies showed that unlabeled oPL was consistently more potent than unlabeled oGH in displacing either of the labeled ligands. The dissociation constant (Kd) for oPL binding ranged from 0.16-0.40 nM and was not changed by solubilization with Triton X-100. Equilibrium binding analysis for oGH showed lower affinity for hepatic microsomal membranes (Kd, 1.7-3.2 nM) in each of the three groups of animals. Affinity cross-linking of microsomal membranes from castrated male and pregnant female sheep liver showed four major cross-linked complexes with both [125I]oPL and [125I]oGH, with mol wt of 150, 97, 75, and 60 kilodaltons. All four bands were identified with both ligands. Unlabeled oPL showed markedly higher potency than unlabeled oGH in reducing the signal of the [125I]oPL cross-linked complexes, whereas unlabeled oGH and oPL showed comparable potencies in reducing the signal of the [125I]oGH complexes. Immunoprecipitation of detergent-solubilized hepatic microsomal membranes from pregnant and fetal sheep using a panel of monoclonal antibodies raised against the extracellular region of the rabbit GHR showed potent immunological recognition of the [125I]oPL-receptor complexes. We suggest that oGH and oPL bind to a common or a related receptor protein(s). It is possible that differences in receptor dimerization or association with other membrane proteins are the basis of the differences in affinity and biological actions of the two hormones.

Quality assurance in breast cancer screening cytology: a review of the literature and a report on the U.K. national cytology scheme.

The National Breast Screening Programme in the United Kingdom has had an external quality assessment (EQA) scheme for breast screening histopathology since 1990. Recently, it was decided, by the Cytology sub-group of the National Co-ordinating Committee for Breast Screening Pathology, to institute two forms of cytology quality assurance. An EQA scheme is planned with circulation of slides to pathologists, but this involves extra time and effort from the participants at a time when general pathology workloads are high. Because of this, a computer routine has been written to analyse the data already present within the National Breast Screening Computer Systems, to enable the calculation of sensitivity and specificity of fine needle aspiration, correlating the cytology results with subsequent histology or follow-up mammography for lesions where no biopsy is performed. This routine uses standardised terminology and calculations and, therefore, inter-unit comparisons can be made. Where problems are identified within a unit, the Quality Assurance team can investigate the cause and institute appropriate measures to correct the problem. This article details the procedures involved in this audit and reviews the literature, recalculating the parameters in a standard manner for a number of publications. The results of the cytology quality assurance routine from seven screening units in one health region in the U.K. are presented and the measures taken to improve the level of service are discussed.

Causes of inconsistency in diagnosing and classifying intraductal proliferations of the breast. European Commission Working Group on Breast Screening Pathology.

It is now widely recognised that classifying ductal carcinoma in situ (DCIS) of the breast and diagnosing atypical ductal hyperplasia are associated with significant interobserver variation. Two possible reasons for this inconsistency are differences in the interpretation of specified histological features and field selection where morphology is heterogeneous. In order to investigate the relative contribution of these two factors to inconsistent interpretation of intraductal proliferations, histological sections of 32 lesions were sent to 23 European pathologists followed 3 years later by images of small parts of these sections. Kappa statistics for diagnosing hyperplasia of usual type, atypical ductal hyperplasia and ductal carcinoma in situ were 0.54, 0.35 and 0.78 for sections and 0.47, 0.29 and 0.78 for images, respectively, showing that most of the inconsistency is due to differences in morphological interpretation. Improvements can thus be expected only if diagnostic criteria or methodology are changed. In contrast, kappa for classifying DCIS by growth pattern was very low at 0.23 for sections and better at 0.47 for images, reflecting the widely recognised variation in the growth pattern of DCIS. Higher kappa statistics were obtained when any mention of an individual growth pattern was included in that category, thus allowing multiple categories per case; but kappa was still higher for images than sections. Classifying DCIS by nuclear grade gave kappa values of 0.36 for sections and 0.49 for images, indicating that intralesional heterogeneity has hitherto been underestimated as a cause of inconsistency in classifying DCIS by this method. More rigorous assessment of the proportions of the different nuclear grades present could lead to an improvement in consistency.

Pathological work-up of sentinel lymph nodes in breast cancer. Review of current data to be considered for the formulation of guidelines.

Controversies and inconsistencies regarding the pathological work-up of sentinel lymph nodes (SNs) led the European Working Group for Breast Screening Pathology (EWGBSP) to review published data and current evidence that can promote the formulation of European guidelines for the pathological work-up of SNs. After an evaluation of the accuracy of SN biopsy as a staging procedure, the yields of different sectioning methods and the immunohistochemical detection of metastatic cells are reviewed. Currently published data do not allow the significance of micrometastases or isolated tumour cells to be established, but it is suggested that approximately 18% of the cases may be associated with further nodal (non-SN) metastases, i.e. approximately 2% of all patients initially staged by SN biopsy. The methods for the intraoperative and molecular assessment of SNs are also surveyed.

Platelet-derived growth factor expression in phyllodes tumors and fibroadenomas of the breast.

The development of normal breast tissue and the pathogenesis of various tumors are influenced by growth factor-mediated epithelial-stromal interactions. Similar interactions may occur in fibroepithelial breast tumors. We have studied the expression of platelet-derived growth factor (PDGF) and PDGF beta receptor (PDGFRbeta) in 46 phyllodes tumors (18 benign, 15 borderline, 13 malignant), 11 fibroadenomas, and 6 samples of normal breast. There was neoplastic stromal cell positivity for PDGFRbeta in almost 50% of phyllodes tumors and for PDGF in 24%. Both were associated with prominent nuclear pleomorphism (P<.01), PDGF with high grade (P<.01), and a higher mean Ki-67 labeling index (P = .013), and PDGFRbeta with conspicuous stromal overgrowth (P<.01). Co-positivity for stromal PDGF and PDGFRbeta was found in 15% of phyllodes tumors, and for epithelial PDGF and stromal PDGFRbeta in 43%. Both types of co-positivity were associated with prominent nuclear pleomorphism and the latter type with conspicuous stromal overgrowth (all P<.01). Follow-up of 41 phyllodes tumors showed that disease-related death was associated with established histologic features of malignancy including mitotic count, stromal overgrowth, an infiltrative tumor margin, and nuclear pleomorphism. In addition, stromal PDGFRbeta positivity (P =.013) and epithelial PDGF/stromal PDGFRbeta co-positivity (P =.0075) were associated with disease-related death. Stromal PDGF and PDGFRbeta expression in fibroadenomas was less common and less extensive (P<.05) than in phyllodes tumors. The results suggest that PDGF influences the pathogenesis of fibroepithelial breast tumors and that PDGF-dependent paracrine and autocrine mechanisms may operate. Also, it is possible that assessment of PDGF and PDGFRbeta expression could contribute to the management of these tumors in the future.

Ultrastructural and immunocytochemical study of a case of invasive cribriform breast carcinoma.

A case of recurrent invasive cribriform carcinoma of the breast mimicked the histological and mucin staining characteristics of adenoid cystic carcinoma. The diagnosis was based on negative immunocytochemical staining for laminin and ultrastructural evidence of luminal differentiation by cells lining the cystic spaces. Accurate characterisation of this type of breast tumour can be facilitated by retrospective immunocytochemical or ultrastructural examination, or both.

Immunohistochemical localisation of androgen receptor in apocrine metaplasia and apocrine adenosis of the breast: relation to oestrogen and progesterone receptors.

AIM: To investigate the receptor status of the sex steroid hormones in apocrine metaplasia of the breast. METHODS: 82 cases of apocrine metaplasia, including 18 of the rare lesion apocrine adenosis, were studied immunohistochemically for the expression of androgen receptor, oestrogen receptor, and progesterone receptor proteins on formalin fixed, paraffin embedded tissue sections. The standard avidin biotin complex (ABC) technique was followed and appropriate positive and negative controls were used. RESULTS: All the studied cases (82/82) were positive for androgen receptor, but were negative for oestrogen receptor and progesterone receptor. CONCLUSIONS: Apocrine metaplastic epithelium, unlike the normal breast epithelium, is responsive to androgens, through androgen receptors, rather than to the female sex hormones. This may have clinical implications.

Ultrastructural analysis of fine needle aspirates from benign breast lesions.

Thirty six fine needle aspirates from various types of benign breast lesions were examined by electron microscopy and correlated with their cytological appearances. In all cases the parenchyma consisted of clumps of cohesive cells with the ultrastructural features of epithelial cells. In many cases, particularly from fibroadenomas, the parenchyma consisted of single layers of polarised epithelial cells showing lumen formation. Similar arrays of apocrine epithelial cells were observed in 60% of fibrocystic lesions. The more solid clumps from hyperplastic lesions consisted of epithelial cells of variable shape and electron density with disorganised lumen formation. Irrespective of the type of lesion, the epithelial cells were not normally subtended by myoepithelial cells or basal lamina. The extraction process seems to result in a shearing between the epithelium and basal lamina with lysis of the myoepithelial cells. Most naked nuclei probably result from lysed myoepithelial cells.

Argentaffin and argyrophil reactions and serotonin content of endocrine tumours.

Sixty carcinoid tumours were tested in a retrospective study with an immunoperoxidase technique using a monoclonal antibody against serotonin immunoreactive sites, with argyrophil staining using the Grimelius technique, and with argentaffin staining using the Masson-Fontana technique. A good correlation between all three techniques in the diagnosis of ileal carcinoid tumour was found, but the immunoperoxidase technique showed greater sensitivity than the Masson-Fontana technique and greater specificity than the Grimelius technique in the diagnosis of foregut and hindgut carcinoid tumours. The immunoperoxidase technique with a monoclonal antibody against serotonin immunoreactive sites (YC5/45) is recommended as a sensitive and specific test for carcinoid tumours. The reactions in other endocrine tumours are also included.

Androgen receptor expression in ductal carcinoma in situ of the breast: relation to oestrogen and progesterone receptors.

AIMS: Ductal carcinoma in situ (DCIS) of the breast has been diagnosed increasingly since the advent of mammographic screening. In contrast to the situation in invasive breast carcinoma, there are no reports on androgen receptor (AR) status in DCIS and few reports on oestrogen (ER) and progesterone (PR) receptors. METHODS: AR expression was examined in 57 cases of DCIS of the breast and correlated to the degree of differentiation and ER/PR status using immunohistochemical methods. RESULTS: AR positivity was noted in 19 of the cases, whereas the other 38 cases were negative. There was no significant association between AR expression and the degree of differentiation of DCIS; three of the 13 well differentiated DCIS cases, 10 of the 19 intermediately differentiated cases, and six of the 25 poorly differentiated cases were positive (p = 0.093). However, a strong association was shown between the expression of ER (p < 0.0001) and PR (p = 0.002) and the degree of differentiation of DCIS. In addition, no significant association was found between the expression of AR and the expression of ER (p = 0.26) or PR (p = 0.57) in DCIS of the breast. CONCLUSIONS: A large number of cases of DCIS of the breast express AR and this may be associated with apocrine differentiation, which may impact on accurate typing of DCIS. Moreover, the expression of AR (but not ER or PR) in DCIS does not appear to be associated with the degree of differentiation.

Laminin and fibronectin in adenoid cystic carcinoma.

The distribution of fibronectin and laminin was examined by immunohistochemistry in 11 adenoid cystic breast carcinomas, six adenoid cystic carcinomas of mouth and salivary gland, and six cribriform ductal breast carcinomas. Both proteins were present lining cystic lumina and around tumour islands in all the adenoid cystic breast carcinomas and in five of six salivary gland tumours. Abundant laminin and fibronectin were dispersed among adenoid cystic tumour cells arranged in sheets. One adenoid cystic carcinoma from buccal mucosa showed a transition from a cribriform tumour positive for both fibronectin and laminin to a cribriform tumour negative for fibronectin and laminin to undifferentiated carcinoma. Fibronectin and laminin seemed to disappear simultaneously from tumour cell surfaces. Another adenoid cystic carcinoma from buccal mucosa was negative for fibronectin and laminin from the time of initial biopsy. This was the only tumour that gave rise to disseminated metastases, resulting in the death of the patient within two years of surgery. In cribriform invasive ductal breast carcinomas the linings of cystic lumina were always negative for fibronectin and laminin. Varying quantities were present at the tumour boundaries. We suggest that staining for fibronectin and laminin may be a valuable aid to the diagnosis of adenoid cystic carcinomas and that the absence of these proteins may have important prognostic implications.

Apocrine adenosis: a precursor of aggressive breast cancer?

AIM: To investigate overexpression of c-erbB2, expression of the p53 protein product and proliferation rates in benign breast lesions with specific reference to apocrine adenosis. METHODS: Twenty one cases of apocrine adenosis were stained with monoclonal antibodies to p185, the protein product of the c-erbB2 oncogene, the protein product of the p53 tumour suppressor gene and to the cell cycle related protein Ki67. Three cases were associated with concomitant ductal carcinoma in situ of large cell type and two were associated with invasive tubular or cribriform carcinoma. RESULTS: Twelve (57.1%) cases showed membrane staining for c-erbB2 oncoprotein of apocrine cells within sclerosing adenosis and six (28.6%) had occasional p53 protein positive cells. One case not associated with carcinoma showed extensive staining of apocrine metaplasia outside the area of apocrine adenosis. The proliferation rate, as measured by Ki67 staining, was increased in some of the lesions and all lesions showed at least some of the cells to be in the cell cycle. CONCLUSIONS: The expression of abnormal oncogene products and increased proliferation in some of these apocrine lesions questions the supposed degenerative nature of the atypia seen in such cases and suggests that there may be an association between these lesions and large cell ductal carcinoma in situ and hence invasive carcinoma.

Spherical connective tissue inclusions in epithelial hyperplasia of the breast ("collagenous spherulosis").

Partial myoepithelial differentiation is common in simple epithelial hyperplasia (epitheliosis) of the breast but functional myoepithelial differentiation with basement membrane production is exceedingly rare. A peculiar change of hyaline globules within benign epithelial hyperplasia has been recognised before as "collagenous spherulosis" and type IV collagen has been shown by immunohistochemistry. Another seven cases are described which show the presence of laminin and collagens IV and III within the proliferation. Electron microscopy examination of two cases using material retrieved from the wax block showed varying degrees of myoepithelial differentiation of the cells immediately surrounding the spherules and basal lamina material, including mature collagen fibrils in one case. The degree of myoepithelial differentiation of the cells surrounding the spherules seemed to correlate with the differing types and amounts of extracellular matrix in the spherule. Histopathologists should be aware of this rare change as it may be misinterpreted as in situ carcinoma.