Chromogenic RNAscope dye and haematoxylin staining of cancer tissue facilitates diagnosis of the cancer type and subsequent treatment, and fits well into existing pathology workflows. However, manual quantification of the RNAscope transcripts (dots), which signify gene expression, is prohibitively time consuming. In addition, there is a lack of verified supporting methods for quantification and analysis. This paper investigates the usefulness of gray level texture features for automatically segmenting and classifying the positions of RNAscope transcripts from breast cancer tissue. Feature analysis showed that a small set of gray level features, including Gray Level Dependence Matrix and Neighbouring Gray Tone Difference Matrix features, were well suited for the task. The automated method performed similarly to expert annotators at identifying the positions of RNAscope transcripts, with an F1-score of 0.571 compared to the expert inter-rater F1-score of 0.596. These results demonstrate the potential of gray level texture features for automated quantification of RNAscope in the pathology workflow.
RNAscope staining of breast cancer tissue allows pathologists to deduce genetic characteristics of the cancer by inspection at the microscopic level, which can lead to better diagnosis and treatment. Chromogenic RNAscope staining is easy to fit into existing pathology workflows, but manually analyzing the resulting tissue samples is time consuming. There is also a lack of peer-reviewed, performant solutions for automated analysis of chromogenic RNAscope staining. This paper covers the development and optimization of a novel deep learning method focused on accurate segmentation of RNAscope dots (which signify gene expression) from breast cancer tissue. The deep learning network is convolutional and uses ConvNeXt as its backbone. The upscaling portions of the network use custom, heavily regularized blocks to prevent overfitting and early convergence on suboptimal solutions. The resulting network is modest in size for a segmentation network and able to function well with little training data. This deep learning network was also able to outperform manual expert annotation at finding the positions of RNAscope dots, having a final F_1 -score of 0.745. In comparison, the expert inter-rater F_1 -score was 0.596.
Background Metaplastic breast carcinoma encompasses a heterogeneous group of tumours with differentiation into squamous and/or spindle, chondroid, osseous or rhabdoid mesenchymal-looking elements. Emerging immunotherapies targeting Programmed Death Ligand 1 (PD-L1) and immune-suppressing T cells (Tregs) may benefit metaplastic breast cancer patients, which are typically chemo-resistant and do not express hormone therapy targets. Methods We evaluated the immunohistochemical expression of PD-L1 and FOXP3, and the extent of tumour infiltrating lymphocytes (TILs) in a large cohort of metaplastic breast cancers, with survival data. Results Metaplastic breast cancers were significantly enriched for PD-L1 positive tumour cells, compared to triple-negative ductal breast cancers ( P < 0.0001), while there was no significant difference in PD-L1 positive TILs. Metaplastic breast cancers were also significantly enriched for TILs expressing FOXP3, with FOXP3 positive intra-tumoural TILs (iTILs) associated with an adverse prognostic outcome ( P = 0.0226). Multivariate analysis identified FOXP3 iTILs expression status as an important independent prognostic factor for patient survival. Conclusions Our findings indicate the clinical significance and prognostic value of FOXP3, PD-1/PD-L1 checkpoint and TILs in metaplastic breast cancer and confirm that a subset of metaplastics may benefit from immune-based therapies.
To assist clinicians with treatment decisions for patients with breast cancer, we investigated whether the preoperative plasma levels of RNAs and the protein cancer antigen 15-3 were associated with prognosis. We identified a novel biomarker, microRNA-923, and combined this with cancer antigen 15-3 and clinicopathological features to build a multivariable model predictive of prognosis, irrespective of treatment, in 253 patients with breast cancer. Introduction: Circulating biomarkers have been increasingly used in the clinical management of breast cancer. The present study evaluated whether RNAs and a protein present in the plasma of patients with breast cancer might have utility as prognostic biomarkers complementary to existing clinical tests. Patients and Methods: We performed microarray profiling of small noncoding RNAs in plasma samples from 30 patients with breast cancer and 10 control individuals. Two small noncoding RNAs, including microRNA (miR)-923, were selected and quantified in plasma samples from an evaluation cohort of 253 patients with breast cancer, using droplet digital polymerase chain reaction. We also measured cancer antigen (CA) 15-3 protein levels in these samples. Cox regression survival analysis was used to determine which markers were associated with patient prognosis. Results: As independent markers of prognosis, the plasma levels of miR-923 and CA 15-3 at the time of surgery for breast cancer were significantly associated with prognosis, irrespective of treatment (Cox proportional hazards, P = 3.9 x 10(-3) and 1.9 x 10(-9), respectively). After building a multivariable model with standard clinical and pathological features, the addition of miR923 and CA 15-3 information into the model resulted in a significantly better predictor of disease recurrence in patients, irrespective of treatment, compared with the use of clinicopathological data alone (area under the curve at 3 years, 0.858 vs. 0.770 with clinicopathological markers only; P = .017). Conclusion: We propose that the plasma levels of miR-923 and CA 15-3, combined with standard clinicopathological predictors, could be used as a preoperative, noninvasive estimate of patient prognosis to identify which women might need more aggressive treatment or closer surveillance after surgery for breast cancer. (C) 2019 Elsevier Inc. All rights reserved.
Metaplastic breast carcinoma (MBC) is relatively rare but accounts for a significant proportion of global breast cancer mortality. This group is extremely heterogeneous and by definition exhibits metaplastic change to squamous and/or mesenchymal elements, including spindle, squamous, chondroid, osseous, and rhabdomyoid features. Clinically, patients are more likely to present with large primary tumours (higher stage), distant metastases, and overall, have shorter 5-year survival compared to invasive carcinomas of no special type. The current World Health Organisation (WHO) diagnostic classification for this cancer type is based purely on morphology - the biological basis and clinical relevance of its seven sub-categories are currently unclear. By establishing the Asia-Pacific MBC (AP-MBC) Consortium, we amassed a large series of MBCs (n = 347) and analysed the mutation profile of a subset, expression of 14 breast cancer biomarkers, and clinicopathological correlates, contextualising our findings within the WHO guidelines. The most significant indicators of poor prognosis were large tumour size (T3; p = 0.004), loss of cytokeratin expression (lack of staining with pan-cytokeratin AE1/3 antibody; p = 0.007), EGFR overexpression (p = 0.01), and for 'mixed' MBC, the presence of more than three distinct morphological entities (p = 0.007). Conversely, fewer morphological components and EGFR negativity were favourable indicators. Exome sequencing of 30 cases confirmed enrichment of TP53 and PTEN mutations, and intriguingly, concurrent mutations of TP53, PTEN, and PIK3CA. Mutations in neurofibromatosis-1 (NF1) were also overrepresented [16.7% MBCs compared to similar to 5% of breast cancers overall; enrichment p = 0.028; mutation significance p = 0.006 (OncodriveFM)], consistent with published case reports implicating germline NF1 mutations in MBC risk. Taken together, we propose a practically minor but clinically significant modification to the guidelines: all WHO_1 mixed-type tumours should have the number of morphologies present recorded, as a mechanism for refining prognosis, and that EGFR and pan-cytokeratin expression are important prognostic markers. Copyright (c) 2018 Pathological Society of Great Britain and Ireland. Published by John Wiley & Sons, Ltd.
We investigated the transcriptional pathways activated by estrogen receptor-cc and their relationship with patient survival after treatment with endocrine therapy. From this analysis, we propose that multimodal assessment Of breast tumors, using a combination of estrogen receptor-alpha status, estrogen receptor-alpha mRNA expression, and genomic indicators of estrogen pathway activity, could be useful for both research and treatment stratification.Background: Molecular markers have transformed our understanding of the heterogeneity of breast cancer and have allowed the identification of genomic profiles of estrogen receptor (ER)-alpha signaling. However, our understanding of the transcriptional profiles of ER signaling remains inadequate. Therefore, we sought to identify the genomic indicators of ER pathway activity that could supplement traditional immunohistochemical (IHC) assessments of ER status to better understand ER signaling in the breast tumors of individual patients. Materials and Methods: We reduced ESR1 (gene encoding the ER-alpha protein) mRNA levels using small interfering RNA in ER+ MCF7 breast cancer cells and assayed for transcriptional changes using Affymetrix HG U133 Plus 2.0 arrays. We also compared 1034 ER+ and ER breast tumors from publicly available microarray data. The principal components of ER activity generated from these analyses and from other published estrogen signatures were compared with ESR1 expression, ER-alpha IHC, and patient survival. Results: Genes differentially expressed in both analyses were associated with ER-alpha IHC and ESR1 mRNA expression. They were also significantly enriched for estrogen-driven molecular pathways associated with ESR1, cyclin D1 (CCND1), MYC (v-myc avian myelocytomatosis viral oncogene homolog), and NFKB (nuclear factor kappa B). Despite their differing constituent genes, the principal components generated from these new analyses and from previously published ER-associated gene lists were all associated with each other and with the survival of patients with breast cancer treated with endocrine therapies. Conclusion: A biomarker of ER-alpha pathway activity, generated using ESR1-responsive mRNAs in MCF7 cells, when used alongside ER-alpha IHC and ESR1 mRNA expression, could provide a method for further stratification of patients and add insight into ER pathway activity in these patients. (C) 2016 The Author(s). Published by Elsevier Inc.
The role of pathologists is to provide diagnostic, prognostic and predictive data to enable clinical colleagues to manage patients optimally. Current histo/anatomical pathology is predominantly morphology-based, with the addition of biomarkers, applied largely through immunohistochemistry, fluorescence in-situ hybridization (FISH) and a limited range of polymerase chain reaction (PCR)-based molecular tests. The desire to apply genomics to the clinical care of patients has been facilitated by the human genome project and subsequently by high-throughput technologies known collectively as massive parallel sequencing (MPS, also referred to as next-generation sequencing, NGS). The use of MPS to identify mutations/variants and tissue RNA expression profiles for diagnosis, prognostication and targeted therapy stratification is now a reality in many clinical specialities. If histopathologists are considered experts in solid tumour pathology, MPS potentially falls within their scope; however, it challenges our predominant morphology-based paradigm. This review summarizes and comments on the current and future state of play of MPS for the practising histopathologist. It will focus on somatic mutations in solid tumours and will challenge histopathologists to take further leadership roles in this area.
Aims: The proliferation marker Ki67 has been extensively investigated as a prognostic factor in breast cancer, but has not gained widespread clinical acceptance. Phosphohistone H3 is a new immunohistochemical marker for quantifying mitoses; however, there is limited information on its prognostic value in breast cancer. In this study, we performed a head-to-head comparison of Ki67 and phosphohistone H3 to establish the marker with the greatest prognostic value.Methods and results: Tissue microarrays from 108 breast cancer patients were immunohistochemically stained for Ki67 and phosphohistone H3. Our results showed that phosphohistone H3 had a greater prognostic value than Ki67 in a multivariable model that adjusted for traditional prognostic variables in breast cancer. Phosphohistone H3 staining was a stronger predictor of survival at 5 years after diagnosis [hazard ratio (HR) 4.35, P < 10(-5)] than Ki67 (HR 2.44, P = 0.004), and better separated the risk of death in patients aged >45 years. Importantly, phosphohistone H3 consistently showed strong unequivocal staining, in contrast to the variable staining intensities associated with Ki67.Conclusions: Our study suggests that phosphohistone H3 staining is a stronger and more robust prognostic indicator than Ki67 staining in breast cancer patients, and has the potential for use in routine diagnostic laboratories.
Aims To determine whether patients diagnosed with breast cancer in 2012 received timely access to services at Christchurch Hospital when audited against Ministry of Health Tumour Standards of Service Provision (TS) (2013) and the Faster Cancer Treatment (FCT) indicators, and to discover factors which impeded patient pathways, and which would need to be addressed in order to meet the standards.Methods Data on referrals, dates and treatment for patients diagnosed with breast cancer at Christchurch Hospital was extracted from the Christchurch Breast Cancer Patient Register and other hospital databases.Results In 2012, 288 breast cancer patients were treated at Christchurch Hospital, 60% referred by general practitioners, and 40% via the national screening programme. Some 2013 Tumour Standards were achieved. The FCT indicator 1 (TS 2.4) and 3 (TS 2.5) were met, with 87% (>= 80%) receiving their first treatment within 62 days of referral, and 89% (>= 80%) within 31 days of decision-to-treat. However, FCT indicator 2 (TS 2.1), requiring first specialist assessment within 14 days of referral, was met in 61% (>= 90% required). Only 64% of women started adjuvant chemotherapy within 42 days of their surgery (TS 2.6, >= 90%).Conclusion The management of breast cancer patients by a multidisciplinary team is crucial to ensure patients receive timely and appropriate care. However, waiting for weekly multidisciplinary meetings and adequate anatomical pathology resource, together with other factors, were identified as delaying the patient pathway and solutions to resolve these are discussed.
We previously reported molecular karyotype analysis of invasive breast tumour core needle biopsies by comparative genomic hybridization (CGH) and fluorescence in situ hybridization (FISH) (Walker et al, Genes Chromosomes Cancer, 2008 May; 47(5):405-17). That study identified frequently recurring gains and losses involving chromosome bands 8q22 and 8p21, respectively. Moreover, these data highlighted an association between 8q22 gain and typically aggressive grade 3 tumors. Here we validate and extend our previous investigations through FISH analysis of tumor touch imprints prepared from excised breast tumor specimens. Compared to post-surgical tumor excisions, core needle biopsies are known to be histologically less precise when predicting tumor grade. Therefore investigating these chromosomal aberrations in tumor samples that offer more reliable pathological assessment is likely to give a better overall indication of association. A series of 60 breast tumors were screened for genomic copy number changes at 8q22 and 8p21 by dual-color FISH. Results confirm previous findings that 8p loss (39%) and 8q gain (74%) occur frequently in invasive breast cancer. Both absolute quantification of 8q22 gain across the sample cohort, and a separate relative assessment by 8q22: 8p21 copy number ratio, showed that the incidence of 8q22 gain significantly increased with grade (p = 0.004, absolute and p = 0.02, relative). In contrast, no association was found between 8p21 loss and tumor grade. These findings support the notion that 8q22 is a region of interest for invasive breast cancer pathogenesis, potentially harboring one or more genes that, when amplified, precipitate the molecular events that define high tumor grade.
Aim The status of excision margins is a well recognised predictive factor for likelihood of recurrence following breast conserving surgery (BCS). Radial margins may be examined by perpendicular sections or shaving. The aim of this audit was to determine what, if any, additional information shave margins provide. Methods Three hundred consecutive BCS specimens were reviewed, with shave margins having been performed in 209. Histological involvement of radial and shave margins for invasive carcinoma, ductal carcinoma in situ and pleomorphic lobular carcinoma in situ was recorded, as were findings in any further re-excision specimens. Results In 13% of cases shave margins only were positive, and 57% of these cases had residual invasive or in situ disease in re-excision specimens. In 8% of cases where both the radial and shave margins were involved, the positive shave margin was different to the radial and would have led to a more extensive re-excision. Therefore shave margins gave additional information to clinicians and impacted on surgical management in 21% of cases. Conclusions Shave margins from BCS specimens appear to identify margin involvement in an increased proportion of patients who would otherwise be left with residual disease. We suggest that these findings support the routine use of shave margins despite current Royal College of Pathologists of Australasia (RCPA) recommendations. The status of excision margins is a well recognised predictive factor for likelihood of recurrence following breast conserving surgery (BCS). Radial margins may be examined by perpendicular sections or shaving. The aim of this audit was to determine what, if any, additional information shave margins provide. Three hundred consecutive BCS specimens were reviewed, with shave margins having been performed in 209. Histological involvement of radial and shave margins for invasive carcinoma, ductal carcinoma in situ and pleomorphic lobular carcinoma in situ was recorded, as were findings in any further re-excision specimens. In 13% of cases shave margins only were positive, and 57% of these cases had residual invasive or in situ disease in re-excision specimens. In 8% of cases where both the radial and shave margins were involved, the positive shave margin was different to the radial and would have led to a more extensive re-excision. Therefore shave margins gave additional information to clinicians and impacted on surgical management in 21% of cases. Shave margins from BCS specimens appear to identify margin involvement in an increased proportion of patients who would otherwise be left with residual disease. We suggest that these findings support the routine use of shave margins despite current Royal College of Pathologists of Australasia (RCPA) recommendations.
Aims The aim of this article is to present the first year's findings of the Christchurch Breast Cancer Patient Register (CBCR) to establish the incidence and management of breast cancer in the Canterbury region.Methods CBCR commenced recruitment of breast cancer patients in Canterbury from June 2009. Ethical approval was granted by regional ethics committees to collect data. Patient data is recorded onto the database once informed consent is obtained.Results A total of 337 patients (including one male) consented. At diagnosis, 231 (68.5%) were aged 50 years or over. 48 (14.2%) patients had carcinoma in situ with no invasive component. 289 (85.8%) patients had invasive carcinoma with 47.4% undergoing mastectomy and 44.6% breast conserving surgery whereas 8% had no primary surgery. Nodes were positive in 102 (38.8%), and the predominant tumour type was Ductal NST (no special type) in 68.9% (199) of patients with invasive carcinoma. Additional data incorporating ethnicity, oncology, additional surgical management and pathological variables are also presented in detail.Conclusion Findings on 337 patients recruited and recorded on CBCR database in the first year are discussed. Due to the short follow up, outcome data is not analysed.
AIMS The aim of this article is to present the first year's findings of the Christchurch Breast Cancer Patient Register (CBCR) to establish the incidence and management of breast cancer in the Canterbury region. METHODS CBCR commenced recruitment of breast cancer patients in Canterbury from June 2009. Ethical approval was granted by regional ethics committees to collect data. Patient data is recorded onto the database once informed consent is obtained. RESULTS A total of 337 patients (including one male) consented. At diagnosis, 231 (68.5%) were aged 50 years or over. 48 (14.2%) patients had carcinoma in situ with no invasive component. 289 (85.8%) patients had invasive carcinoma with 47.4% undergoing mastectomy and 44.6% breast conserving surgery whereas 8% had no primary surgery. Nodes were positive in 102 (38.8%), and the predominant tumour type was Ductal NST (no special type) in 68.9% (199) of patients with invasive carcinoma. Additional data incorporating ethnicity, oncology, additional surgical management and pathological variables are also presented in detail. CONCLUSION Findings on 337 patients recruited and recorded on CBCR database in the first year are discussed. Due to the short follow up, outcome data is not analysed.
We performed immunohistochemical analysis of 3 cancer stem cell-related markers (CD44(+)/CD24(-/low), aldehyde dehydrogenase [ALDH]-1, CD133) in 94 invasive ductal carcinomas and assessed relationships with markers of hypoxia (carbonic anhydrase IX [CAIX]), tumor microvessel density (CD31), and clinicopathologic variables. Overall, 10% of tumors were CD44(+)/CD24(-/low), 13% were ALDH-1(+), 25% were CD133(+), 35% were immunonegative, and 1 tumor was immunopositive for all 3 markers. Associated ductal carcinoma in situ (DCIS) was present in 48% of tumors. Marker immunopositivity was detected in DCIS in 13% (CD44(+)/CD24(-/low)), 7% (ALDH-1(+)), and 32% (CD133(+)) of these tumors and was more likely present in DCIS when also detected in the invasive compartment (P = .03, P = .001, and P = .009, respectively). CD44(+)/CD24(-/low) cells were more common in progesterone receptor-negative tumors (P < .01), and ALDH-1(+) cells were more common in estrogen receptor-negative tumors (P < .01). CD133(+) cells were more common in patients younger than 50 years (P < .05) and in high grade (P < .01), localized (P < .05), and estrogen receptor-negative (P < .001), progesterone receptor-negative (P = .02), and triple-negative breast cancers (P < .001). CD44(+)/CD24(-/low) (P = .06) and CD133(+) (P = .02) tumor cells were more common in CAIX(+) versus CAIX(-) tumors, whereas ALDH-1(+) tumors had a higher mean microvessel density than did ALDH-1(-) tumors (P = .002). No significant relationships were observed between the markers studied and survival for 5 years. Our study demonstrated the presence of cancer stem cell marker-positive tumor cells in DCIS as well as invasive breast cancer and showed that CD44(+)/CD24(-/low) and CD133(+) cells were more frequently observed in hypoxic regions of tumor, whereas ALDH-1(+) cells more commonly colocalized to tumors with high microvessel density.
BACKGROUND:New Zealand Māori have a poorer outcome from breast cancer than non-Māori, yet prognostic data are sparse. The objective of this study was to quantify levels of prognostic factors in a cohort of self-declared Māori and European breast cancer patients from Christchurch, New Zealand.METHODS AND RESULTS:Clinicopathological and survival data from 337 consecutive breast cancer patients (27 Māori, 310 European) were evaluated. Fewer tumours were high grade in Māori women than European women (p = 0.027). No significant ethnic differences were detected for node status, tumour type, tumour size, human epidermal growth factor receptor, oestrogen and progesterone receptor (ER/PR) status, or survival.In addition, tumour and serum samples from a sub-cohort of 14 Māori matched to 14 NZ European patients were analyzed by immunohistochemistry and enzyme linked immunosorbent assay for molecular prognostic factors. Significant correlations were detected between increased grade and increased levels of hypoxia inducible factor-1 (HIF-1α), glucose transporter-1 (GLUT-1), microvessel density (MVD) and cytokeratins CK5/6 (p < 0.05). High nodal status correlated with reduced carbonic anhydrase IX (CA-IX). Negative ER/PR status correlated with increased GLUT-1, CA-IX and MVD. Within the molecular factors, increased HIF-1α correlated with raised GLUT-1, MVD and CK5/6, and CK5/6 with GLUT-1 and MVD (p < 0.05). The small number of patients in this sub-cohort limited discrimination of ethnic differences.CONCLUSIONS:In this Christchurch cohort of breast cancer patients, Māori women were no more likely than European women to have pathological or molecular factors predictive of poor prognosis. These data contrast with data from the North Island NZ, and suggest potential regional differences.
Genetic analysis of breast tumor core needle biopsies may provide early diagnostic information that is useful to direct subsequent clinical management. We report metaphase comparative genomic hybridization (CGH) analysis of 42 core needle biopsies prospectively sampled from invasive ductal breast carcinomas of 41 patients, and show that recurrent chromosomal copy number changes are associated with histologically defined tumor features. As far as is known, this is the first time CGH profiles from diagnostic breast tumor core needle biopsies have been reported in association with pathological data in such detail. A comparison of Grade 1 ( n = 7), Grade 2 ( n = 16), and Grade 3 tumors ( n = 19) revealed a chromosomal copy number gain involving 8q22 that was associated with higher grade (1/7 vs. 16/19, respectively) and therefore altered cell differentiation status. CGH results were validated using tumor touch imprints prepared from a subgroup from the same sample set ( n = 18) by fluorescence in situ hybridization (FISH) and two probes selected from regions of interest within 8p21 and 8q22. Overall concordance between CGH and FISH for 8p21 and 8q22 imbalances was 9/18 (50%) and 12/18 (67%), respectively. When FISH and CGH data were combined, and tumors classified according to better defined resected tumor histology available for 34cases, 19/19 Grade 3 tumors showed 8q22 gain compared with 0/6 Grade 1 tumors and 4/9 Grade 2 tumors. Further comprehensive studies are required to verify the biological significance of this association and its potential to facilitate early clinicopathological assessment of breast cancer. This article contains Supplementary Material available at http://www.interscience.wiley.com/jpages/1045‐2257/suppmat . © 2008 Wiley‐Liss, Inc.
Invasive ductal carcinomas of the breast (IDC) are routinely assessed on hematoxylin and eosin stained paraffin sections, with limited use of immunohistochemistry (IHC). Most IDC are regarded as a single diagnostic entity, IDC of no special type (IDC-NST), which is subdivided further only by grading. However, recent research suggests that there is high clinical relevance in differentiating IDC subtypes. Here, we ascertain whether tumor histology alone can predict basal or luminal cell phenotype in high-grade IDC-NST, and whether IHC and molecular characteristics are associated with the observed morphologies. A total of 29 grade 3 IDC-NST samples were studied, 10 tumors from a selected pilot cohort A and 19 tumors from an unselected validation cohort B. Along with histopathological assessment, the expression of ESR1, PGR, ERBB2 (HER-2), the basal/myoepithelial marker TP73L (p63), cytokeratins 5/6 (KRT5/6) and 14 (KRT14), and the luminal-specific cytokeratins 8/18 (KRT 8/18) and 19 (KRT19) was assessed by IHC. Hierarchical cluster analysis of clinicopathological variables and, separately, microarray expression profiles showed that the phenotypically distinctive basaloid and luminal tumors of cohort A fell into two main groups, defined by heterogeneous or uniformly positive expression of KRT8/18. The 38 genes differentially expressed between these two classes included ERBB2, KRT8, and six other genes previously associated with ERBB2-positive or luminal phenotypes. Tumor histology was not predictive for validation cohort B, but quantitative real-time polymerase chain reaction (qRT-PCR) analysis revealed two molecularly defined clusters that again aligned with the KRT8/18 staining phenotypes. Metaphase comparative genomic hybridization revealed 10q, 16q, and 20q copy-number imbalances that associated recurrently with KRT8/18 staining patterns.
Tumours of perivascular epithelioid cells (PEComas) are being increasingly reported at visceral and somatic sites. Both benign and malignant variants have been identified, although clinical follow-up is often limited, which prevents meaningful predictions of behavior. We report the case of a malignant soft tissue PEComa with histologically confirmed regional lymph node metastases and radiologically confirmed pulmonary metastases. The light microscopic appearance and immunohistochemical profile (HMB-45, smooth muscle actin positive) and electron microscopic appearance support perivascular epithelioid cell differentiation.