Rapid artefact removal and H&E-stained tissue segmentation.

We present an innovative method for rapidly segmenting haematoxylin and eosin (H&E)-stained tissue in whole-slide images (WSIs) that eliminates a wide range of undesirable artefacts such as pen marks and scanning artefacts. Our method involves taking a single-channel representation of a low-magnification RGB overview of the WSI in which the pixel values are bimodally distributed such that H&E-stained tissue is easily distinguished from both background and a wide variety of artefacts. We demonstrate our method on 30 WSIs prepared from a wide range of institutions and WSI digital scanners, each containing substantial artefacts, and compare it to segmentations provided by Otsu thresholding and Histolab tissue segmentation and pen filtering tools. We found that our method segmented the tissue and fully removed all artefacts in 29 out of 30 WSIs, whereas Otsu thresholding failed to remove any artefacts, and the Histolab pen filtering tools only partially removed the pen marks. The beauty of our approach lies in its simplicity: manipulating RGB colour space and using Otsu thresholding allows for the segmentation of H&E-stained tissue and the rapid removal of artefacts without the need for machine learning or parameter tuning.

Stain normalization gives greater generalizability than stain jittering in neural network training for the classification of coeliac disease in duodenal biopsy whole slide images.

Around 1% of the population of the UK and North America have a diagnosis of coeliac disease (CD), due to a damaging immune response to the small intestine. Assessing whether a patient has CD relies primarily on the examination of a duodenal biopsy, an unavoidably subjective process with poor inter-observer concordance. Wei et al. [11] developed a neural network-based method for diagnosing CD using a dataset of duodenal biopsy whole slide images (WSIs). As all training and validation data came from one source, there was no guarantee that their results would generalize to WSIs obtained from different scanners and laboratories. In this study, the effects of applying stain normalization and jittering to the training data were compared. We trained a deep neural network on 331 WSIs obtained with a Ventana scanner (WSIs; CD: n=190; normal: n=141) to classify presence of CD. In order to test the effects of stain processing when validating on WSIs scanned on varying scanners and from varying laboratories, the neural network was validated on 4 datasets: WSIs of slides scanned on a Ventana scanner (WSIs; CD: n=48; normal: n=35), WSIs of the same slides rescanned on a Hamamatsu scanner (WSIs; CD: n=48; normal: n=35), WSIs of the same slides rescanned on an Aperio scanner (WSIs; CD: n=48; normal: n=35), and WSIs of different slides scanned on an Aperio scanner (WSIs; CD: n=38; normal: n=37). Without stain processing, the F1 scores of the neural network were 0.947, 0.619, 0.746, and 0.727 when validating on the Ventana validation WSIs, Hamamatsu and Aperio rescans of the Ventana validation WSIs, and Aperio WSIs from a different source respectively. With stain normalization, the performance of the neural network improved significantly with respective F1 scores 0.982, 0.943, 0.903, and 0.847. Stain jittering resulted in a better performance than stain normalization when validating on data from the same source F1 score 1.000, but resulted in poorer performance than stain normalization when validating on WSIs from different scanners (F1 scores 0.939, 0.814, and 0.747). This study shows the importance of stain processing, in particular stain normalization, when training machine learning models on duodenal biopsy WSIs to ensure generalizability between different scanners and laboratories.

Multiple-instance-learning-based detection of coeliac disease in histological whole-slide images.

We present a multiple-instance-learning-based scheme for detecting coeliac disease, an autoimmune disorder affecting the intestine, in histological whole-slide images (WSIs) of duodenal biopsies. We train our model to detect 2 distinct classes, normal tissue and coeliac disease, on the patch-level, and in turn leverage slide-level classifications. Using 5-fold cross-validation in a training set of 1841 (1163 normal; 680 coeliac disease) WSIs, our model classifies slides as normal with accuracy (96.7±0.6)%, precision (98.0±1.7)%, and recall (96.8±2.5)%, and as coeliac disease with accuracy (96.7±0.5)%, precision (94.9±3.7)%, and recall (96.5±2.9)% where the error bars are the cross-validation standard deviation. We apply our model to 2 test sets: one containing 191 WSIs (126 normal; 65 coeliac) from the same sources as the training data, and another from a completely independent source, containing 34 WSIs (17 normal; 17 coeliac), obtained with a scanner model not represented in the training data. Using the same-source test data, our model classifies slides as normal with accuracy 96.5%, precision 98.4% and recall 96.1%, and positive for coeliac disease with accuracy 96.5%, precision 93.5%, and recall 97.3%. Using the different-source test data the model classifies slides as normal with accuracy 94.1% (32/34), precision 89.5%, and recall 100%, and as positive for coeliac disease with accuracy 94.1%, precision 100%, and recall 88.2%. We discuss generalising our approach to screen for a range of pathologies.

Correction: S1PR1 drives a feedforward signalling loop to regulate BATF3 and the transcriptional programme of Hodgkin lymphoma cells.

An amendment to this paper has been published and can be accessed via a link at the top of the paper.

Cutaneous mastocytosis in human immunodeficiency virus: an unfortunate coincidence?

We report the case of a young man who developed cutaneous mastocytosis on the trunk and limbs after an episode of severe sunburn. He was diagnosed with human immunodeficiency virus (HIV) shortly afterwards. We discuss this interesting and previously unreported temporal relationship between the development of mastocytosis and recent HIV infection, speculating on the potential pathogenetic mechanisms of association.

S1PR1 drives a feedforward signalling loop to regulate BATF3 and the transcriptional programme of Hodgkin lymphoma cells.

The Hodgkin/Reed-Sternberg cells of classical Hodgkin lymphoma (HL) are characterised by the aberrant activation of multiple signalling pathways. Here we show that a subset of HL displays altered expression of sphingosine-1-phosphate (S1P) receptors (S1PR)s. S1P activates phosphatidylinositide 3-kinase (PI3-K) in these cells that is mediated by the increased expression of S1PR1 and the decreased expression of S1PR2. We also showed that genes regulated by the PI3-K signalling pathway in HL cell lines significantly overlap with the transcriptional programme of primary HRS cells. Genes upregulated by the PI3-K pathway included the basic leucine zipper transcription factor, ATF-like 3 (BATF3), which is normally associated with the development of dendritic cells. Immunohistochemistry confirmed that BATF3 was expressed in HRS cells of most HL cases. In contrast, in normal lymphoid tissues, BATF3 expression was confined to a small fraction of CD30-positive immunoblasts. Knockdown of BATF3 in HL cell lines revealed that BATF3 contributed to the transcriptional programme of primary HRS cells, including the upregulation of S1PR1. Our data suggest that disruption of this potentially oncogenic feedforward S1P signalling loop could provide novel therapeutic opportunities for patients with HL.

Epithelial cyst of the cardiac papillary muscle: case report and review of the literature.

BACKGROUND: Reports of endodermal heterotopia (previously known as inclusion cysts) in cardiac atria are rare and there is only a single previous case report of endodermal heterotopia in a cardiac papillary muscle. AIM AND METHODS: A cyst in a cardiac papillary muscle was identified during the autopsy of an 87-year-old man who had died from an unrelated myocardial infarction. The cyst was examined histologically and mucin staining and immunostaining were carried out. RESULTS: We report a unilocular cyst in a cardiac papillary muscle, which is lined by low cuboidal, pseudostratified and occasionally ciliated respiratory-type epithelium, surrounded by a layer of smooth muscle. The immunohistochemical features (MNF116+, cytokeratin (CK)7+, CK8+, CK18+, CK19+, epithelial membrane antigen positive, scattered cells positive for neuroendocrine markers) suggest that this is an endodermal heterotopia. Immunostaining of positive thyroid transcription factor-1 provides evidence for bronchogenic differentiation. DISCUSSION: The differential diagnoses of cystic structures in cardiac papillary muscle and the origin and importance of endodermal heterotopias are discussed.

Assessment of the Cardiff nephrectomy cut-up protocol with total blocking of the renal sinus: effect on tumour staging and practical issues.

AIM: To evaluate the effects on detection of vascular invasion and workload of a new standard dissection protocol for examining nephrectomy specimens for renal cell carcinoma. METHODS: Using 192 consecutive renal cell carcinoma nephrectomy specimens, the incidence of vascular invasion and number of tissue blocks per tumour were compared before and after introduction of the new protocol. RESULTS: The Cardiff protocol increased the percentage of tumours staged as T3b (renal sinus or hilar vein invasion) from 37.7% to 55.7% cases (p<0.001), with an increase from 9.1% to 21.7% of those staged as T3b due to renal sinus vein invasion alone (p<0.01). A small, but significant, permanent increase in workload was observed from an average of 11.7 to 13.4 blocks per case (p<0.001). CONCLUSIONS: This protocol is suitable for use in routine practice to evaluate pathological prognostic determinants important for clinical management, while causing only a small increase in workload.

Dendritic and mast cell involvement in the inflammatory response to primary malignant bone tumours.

BACKGROUND: A chronic inflammatory cell infiltrate is commonly seen in response to primary malignant tumours of bone. This is known to contain tumour-associated macrophages (TAMs) and lymphocytes; dendritic cells (DCs) and mast cells (MCs) have also been identified but whether these and other inflammatory cells are seen commonly in specific types of bone sarcoma is uncertain. METHODS: In this study we determined the nature of the inflammatory cell infiltrate in 56 primary bone sarcomas. Immunohistochemistry using monoclonal antibodies was employed to assess semiquantitatively CD45+ leukocyte infiltration and the extent of the DC, MC, TAM and T and B lymphocyte infiltrate. RESULTS: The extent of the inflammatory infiltrate in individual sarcomas was very variable. A moderate or heavy leukocyte infiltrate was more commonly seen in conventional high-grade osteosarcoma, undifferentiated pleomorphic sarcoma and giant cell tumour of bone (GCTB) than in Ewing sarcoma, chordoma and chondrosarcoma. CD14+/CD68+ TAMs and CD3+ T lymphocytes were the major components of the inflammatory cell response but (DC-SIGN/CD11c+) DCs were also commonly noted when there was a significant TAM and T lymphocyte infiltrate. MCs were identified mainly at the periphery of sarcomas, including the osteolytic tumour-bone interface. DISCUSSION: Our findings indicate that, although variable, some malignant bone tumours (e.g. osteosarcoma, GCTB) are more commonly associated with a pronounced inflammatory cell infiltrate than others (e.g. chondrosarcoma. Ewing sarcoma); the infiltrate is composed mainly of TAMs but includes a significant DC, T lymphocyte and MC infiltrate. CONCLUSION: Tumours that contain a heavy inflammatory cell response, which includes DCs, TAMs and T lymphocytes, may be more amenable to immunomodulatory therapy. MCs are present mainly at the tumour edge and are likely to contribute to osteolysis and tumour invasion.

FOXP1 suppresses immune response signatures and MHC class II expression in activated B-cell-like diffuse large B-cell lymphomas.

The FOXP1 (forkhead box P1) transcription factor is a marker of poor prognosis in diffuse large B-cell lymphoma (DLBCL). Here microarray analysis of FOXP1-silenced DLBCL cell lines identified differential regulation of immune response signatures and major histocompatibility complex class II (MHC II) genes as some of the most significant differences between germinal center B-cell (GCB)-like DLBCL with full-length FOXP1 protein expression versus activated B-cell (ABC)-like DLBCL expressing predominantly short FOXP1 isoforms. In an independent primary DLBCL microarray data set, multiple MHC II genes, including human leukocyte antigen DR alpha chain (HLA-DRA), were inversely correlated with FOXP1 transcript expression (P<0.05). FOXP1 knockdown in ABC-DLBCL cells led to increased cell-surface expression of HLA-DRA and CD74. In R-CHOP (rituximab, cyclophosphamide, doxorubicin, vincristine and prednisone)-treated DLBCL patients (n=150), reduced HLA-DRA (<90% frequency) expression correlated with inferior overall survival (P=0.0003) and progression-free survival (P=0.0012) and with non-GCB subtype stratified by the Hans, Choi or Visco-Young algorithms (all P<0.01). In non-GCB DLBCL cases with <90% HLA-DRA, there was an inverse correlation with the frequency (P=0.0456) and intensity (P=0.0349) of FOXP1 expression. We propose that FOXP1 represents a novel regulator of genes targeted by the class II MHC transactivator CIITA (MHC II and CD74) and therapeutically targeting the FOXP1 pathway may improve antigen presentation and immune surveillance in high-risk DLBCL patients.

Expression of DC-SIGN in human foreskin may facilitate sexual transmission of HIV.

This study demonstrates that the human immunodeficiency virus (HIV) binding C-type lectin DC-SIGN is coexpressed with CD4 and CCR5 on dendritic cells/macrophages in human foreskin. It is hypothesised that DC-SIGN may contribute to the sexual transmission of HIV in the foreskin, by enabling infection of permissive cells in cis and/or in trans.

HCV and HIV binding lectin, DC-SIGNR, is expressed at all stages of HCV induced liver disease.

The process by which hepatitis C virus (HCV) enters cells and the reason for its hepatotropism remain obscure. Recently, the human immunodeficiency virus (HIV) binding lectins, DC-SIGN and DC-SIGNR, were shown to bind HCV. This article reports the expression of DC-SIGN and DC-SIGNR in HCV related liver disease and discusses whether these lectins, in particular DC-SIGNR, are responsible for HCV hepatotropism.

Reciprocal expression of the endocytic protein HIP1R and its repressor FOXP1 predicts outcome in R-CHOP-treated diffuse large B-cell lymphoma patients.

We previously identified autoantibodies to the endocytic-associated protein Huntingtin-interacting protein 1-related (HIP1R) in diffuse large B-cell lymphoma (DLBCL) patients. HIP1R regulates internalization of cell surface receptors via endocytosis, a process relevant to many therapeutic strategies including CD20 targeting with rituximab. In this study, we characterized HIP1R expression patterns, investigated a mechanism of transcriptional regulation and its clinical relevance in DLBCL patients treated with immunochemotherapy (rituximab, cyclophosphamide, doxorubicin, vincristine and prednisone, R-CHOP). HIP1R was preferentially expressed in germinal center B-cell-like DLBCL (P<0.0001) and inversely correlated with the activated B-cell-like DLBCL (ABC-DLBCL) associated transcription factor, Forkhead box P1 (FOXP1). HIP1R was confirmed as a direct FOXP1 target gene in ABC-DLBCL by FOXP1-targeted silencing and chromatin immunoprecipitation. Lower HIP1R protein expression (≤ 10% tumoral positivity) significantly correlated with inferior overall survival (OS, P=0.0003) and progression-free survival (PFS, P=0.0148) in R-CHOP-treated DLBCL patients (n=157). Reciprocal expression with ≥ 70% FOXP1 positivity defined FOXP1(hi)/HIP1R(lo) patients with particularly poor outcome (OS, P=0.0001; PFS, P=0.0016). In an independent R-CHOP-treated DLBCL (n=233) microarray data set, patients with transcript expression in lower quartile HIP1R and FOXP1(hi)/HIP1R(lo) subgroups exhibited worse OS, P=0.0044 and P=0.0004, respectively. HIP1R repression by FOXP1 is strongly associated with poor outcome, thus further understanding of FOXP1-HIP1R and/or endocytic signaling pathways might give rise to novel therapeutic options for DLBCL.

CD4 expression by mast cells in mastocytosis: a case report.

A 59-year-old man presented with a five-year history of an asymptomatic widespread truncal rash. Cutaneous biopsies showed mastocytosis, which was shown to be CD4-positive immunohistochemically. This potential diagnostic pitfall in the discrimination of mastocytosis from a variety of other haematopoietic neoplasms, particularly mycosis fungoides, is discussed.

DC-SIGN association with the Th2 environment of lepromatous lesions: cause or effect?

The clinical spectrum of leprosy is related to patients' immune responses. Non-responsiveness towards Mycobacterium leprae (ML) seems to correlate with a Th2 cytokine profile. The reason for such a polarized immune response remains unclear. The C-type lectin, DC-SIGN, expressed by subsets of dendritic cells (DCs) and macrophages, has previously been associated with Th2 responses. Here we show abundant DC-SIGN expression in lepromatous but not borderline tuberculoid leprosy, in both HIV-positive and HIV-negative patients. Moreover, we demonstrate that DC-SIGN can act as an entry receptor for ML, as it does for M. tuberculosis, through the cell wall component lipoarabinomannan. DC-SIGN is expressed on virtually all ML-containing cells, providing further evidence for its role as a receptor. DC-SIGN may therefore be induced on macrophages in lepromatous leprosy and may then contribute to mycobacterial entry into these cells.

Use of novel monoclonal antibodies to determine the expression and distribution of the hypoxia regulatory factors PHD-1, PHD-2, PHD-3 and FIH in normal and neoplastic human tissues.

AIMS: The cellular response to hypoxia includes the hypoxia inducible factor (HIF)-induced transcription of genes involved in diverse processes such as glycolysis, angiogenesis and the growth of experimental tumours. Regulation of the level of hypoxia inducible factors 1alpha and 2alpha (HIF-1alpha and HIF-2alpha) is a primary determinant of HIF activity. Recent biochemical and candidate gene approach studies have led to the discovery of three HIF-regulatory prolyl hydroxylases, PHD-1, -2 and -3 and an asparaginyl hydroxylase, also known as FIH (factor inhibiting HIF). In this study, we raised and characterized monoclonal antibodies against PHD-1, PHD-2, PHD-3 and FIH. METHODS AND RESULTS: Immunohistochemistry of normal tissues with these monoclonal antibodies demonstrated a wide distribution in epithelial cells, stromal cells and leucocytes, with cytoplasmic staining predominating over nuclear staining. A preliminary study of tumours showed variable staining in tumour, stromal and inflammatory cells. While all tumour types showed some positive staining with each antibody, the overall pattern suggested a slight decrease in the amount of staining seen with PHD-1, -2 and -3 and an increase in FIH staining in neoplasia compared with corresponding normal tissues. CONCLUSIONS: These monoclonal antibodies will allow further larger scale studies to determine the significance of PHD and FIH expression in neoplasia.

DC-SIGN; a related gene, DC-SIGNR; and CD23 form a cluster on 19p13.

DC-SIGN is a C-type lectin, expressed on a dendritic cell subset. It is able to bind ICAM3 and HIV gp120 in a calcium-dependent manner. Here we report the genomic organization of DC-SIGN and map it to chromosome 19p13 adjacent to the C-type lectin CD23 (FcepsilonRII). We also report a novel, closely linked gene, DC-SIGNR, which shows 73% identity to DC-SIGN at the nucleic acid level and a similar genomic organization. Proteins encoded by both genes have tracts of repeats of 23 aa, predicted to form a coiled coil neck region. They also possess motifs that are known to bind mannose in a calcium-dependent fashion. We show concomitant expression of the two genes in endometrium, placenta, and stimulated KG1 cells (phenotypically similar to monocyte-derived dendritic cells). The existence of a DC-SIGN-related gene calls for reinterpretation of the HIV data to consider possible DC-SIGN/DC-SIGNR hetero-oligomerization.