Targeting placental growth factor/neuropilin 1 pathway inhibits growth and spread of medulloblastoma.

Medulloblastoma is the most common pediatric malignant brain tumor. Although current therapies improve survival, these regimens are highly toxic and are associated with significant morbidity. Here, we report that placental growth factor (PlGF) is expressed in the majority of medulloblastomas, independent of their subtype. Moreover, high expression of PlGF receptor neuropilin 1 (Nrp1) correlates with poor overall survival in patients. We demonstrate that PlGF and Nrp1 are required for the growth and spread of medulloblastoma: PlGF/Nrp1 blockade results in direct antitumor effects in vivo, resulting in medulloblastoma regression, decreased metastasis, and increased mouse survival. We reveal that PlGF is produced in the cerebellar stroma via tumor-derived Sonic hedgehog (Shh) and show that PlGF acts through Nrp1-and not vascular endothelial growth factor receptor 1-to promote tumor cell survival. This critical tumor-stroma interaction-mediated by Shh, PlGF, and Nrp1 across medulloblastoma subtypes-supports the development of therapies targeting PlGF/Nrp1 pathway.

Enhancing clinical genomic accuracy with panelGC: a novel metric and tool for quantifying and monitoring GC biases in hybridization capture panel sequencing.

Accurate assessment of fragment abundance within a genome is crucial in clinical genomics applications such as the analysis of copy number variation (CNV). However, this task is often hindered by biased coverage in regions with varying guanine-cytosine (GC) content. These biases are particularly exacerbated in hybridization capture sequencing due to GC effects on probe hybridization and polymerase chain reaction (PCR) amplification efficiency. Such GC content-associated variations can exert a negative impact on the fidelity of CNV calling within hybridization capture panels. In this report, we present panelGC, a novel metric, to quantify and monitor GC biases in hybridization capture sequencing data. We establish the efficacy of panelGC, demonstrating its proficiency in identifying and flagging potential procedural anomalies, even in situations where instrument and experimental monitoring data may not be readily accessible. Validation using real-world datasets demonstrates that panelGC enhances the quality control and reliability of hybridization capture panel sequencing.

SMARCB1 Loss in Poorly Differentiated Chordomas Drives Tumor Progression.

Poorly differentiated (PD) chordoma, a rare, aggressive tumor originating from notochordal tissue, shows loss of SMARCB1 expression, a core component of the Switch/Sucrose Non-Fermentable (SWI/SNF) chromatin remodeling complexes. To determine the impact of SMARCB1 re-expression on cell growth and gene expression, two SMARCB1-negative PD chordoma cell lines with an inducible SMARCB1 expression system were generated. After 72 hours of induction of SMARCB1, both SMARCB1-negative PD chordoma cell lines continued to proliferate. This result contrasted with those observed with SMARCB1-negative rhabdoid cell lines in which SMARCB1 re-expression caused the rapid inhibition of growth. We found that the lack of growth inhibition may arise from the loss of CDKN2A (p16INK4A) expression in PD chordoma cell lines. RNA-sequencing of cell lines after SMARCB1 re-expression showed a down-regulation for rRNA and RNA processing as well as metabolic processing and increased expression of genes involved in cell adhesion, cell migration, and development. Taken together, these data establish that SMARCB1 re-expression in PD chordomas alters the repertoire of SWI/SNF complexes, perhaps restoring those associated with cellular differentiation. These novel findings support a model in which SMARCB1 inactivation blocks the conversion of growth-promoting SWI/SNF complexes to differentiation-inducing ones, and they implicate SMARCB1 loss as a late event in tumorigenic progression. Importantly, the absence of growth inhibition after SMARCB1 restoration creates a unique opportunity to identify therapeutic vulnerabilities.

Amplification of the PLAG-family genes-PLAGL1 and PLAGL2-is a key feature of the novel tumor type CNS embryonal tumor with PLAGL amplification.

Pediatric central nervous system (CNS) tumors represent the most common cause of cancer-related death in children aged 0-14 years. They differ from their adult counterparts, showing extensive clinical and molecular heterogeneity as well as a challenging histopathological spectrum that often impairs accurate diagnosis. Here, we use DNA methylation-based CNS tumor classification in combination with copy number, RNA-seq, and ChIP-seq analysis to characterize a newly identified CNS tumor type. In addition, we report histology, patient characteristics, and survival data in this tumor type. We describe a biologically distinct pediatric CNS tumor type (n = 31 cases) that is characterized by focal high-level amplification and resultant overexpression of either PLAGL1 or PLAGL2, and an absence of recurrent genetic alterations characteristic of other pediatric CNS tumor types. Both genes act as transcription factors for a regulatory subset of imprinted genes (IGs), components of the Wnt/ß-Catenin pathway, and the potential drug targets RET and CYP2W1, which are also specifically overexpressed in this tumor type. A derived PLAGL-specific gene expression signature indicates dysregulation of imprinting control and differentiation/development. These tumors occurred throughout the neuroaxis including the cerebral hemispheres, cerebellum, and brainstem, and were predominantly composed of primitive embryonal-like cells lacking robust expression of markers of glial or neuronal differentiation (e.g., GFAP, OLIG2, and synaptophysin). Tumors with PLAGL1 amplification were typically diagnosed during adolescence (median age 10.5 years), whereas those with PLAGL2 amplification were diagnosed during early childhood (median age 2 years). The 10-year overall survival was 66% for PLAGL1-amplified tumors, 25% for PLAGL2-amplified tumors, 18% for male patients, and 82% for female patients. In summary, we describe a new type of biologically distinct CNS tumor characterized by PLAGL1/2 amplification that occurs predominantly in infants and toddlers (PLAGL2) or adolescents (PLAGL1) which we consider best classified as a CNS embryonal tumor and which is associated with intermediate survival. The cell of origin and optimal treatment strategies remain to be defined.

The utility of color normalization for AI-based diagnosis of hematoxylin and eosin-stained pathology images.

The color variation of hematoxylin and eosin (H&E)-stained tissues has presented a challenge for applications of artificial intelligence (AI) in digital pathology. Many color normalization algorithms have been developed in recent years in order to reduce the color variation between H&E images. However, previous efforts in benchmarking these algorithms have produced conflicting results and none have sufficiently assessed the efficacy of the various color normalization methods for improving diagnostic performance of AI systems. In this study, we systematically investigated eight color normalization algorithms for AI-based classification of H&E-stained histopathology slides, in the context of using images both from one center and from multiple centers. Our results show that color normalization does not consistently improve classification performance when both training and testing data are from a single center. However, using four multi-center datasets of two cancer types (ovarian and pleural) and objective functions, we show that color normalization can significantly improve the classification accuracy of images from external datasets (ovarian cancer: 0.25 AUC increase, p = 1.6 e-05; pleural cancer: 0.21 AUC increase, p = 1.4 e-10). Furthermore, we introduce a novel augmentation strategy by mixing color-normalized images using three easily accessible algorithms that consistently improves the diagnosis of test images from external centers, even when the individual normalization methods had varied results. We anticipate our study to be a starting point for reliable use of color normalization to improve AI-based, digital pathology-empowered diagnosis of cancers sourced from multiple centers. © 2021 The Pathological Society of Great Britain and Ireland. Published by John Wiley & Sons, Ltd.

Outcomes of a radiation sparing approach in medulloblastoma by subgroup in young children: an institutional review.

OBJECTIVE: To describe disease outcomes including overall survival and relapse patterns by subgroup in young pediatric patients treated for medulloblastoma with a radiation-sparing approach. METHODS: Retrospective analysis of clinical outcomes includes treatment, relapse, and salvage therapy and late effects in children treated for medulloblastoma with a radiation-sparing approach at British Columbia Children's Hospital (BCCH) between 2000 and 2020. RESULTS: There were 30 patients (median age 2.8 years, 60% male) treated for medulloblastoma with a radiation-sparing approach at BCCH. Subgroups included Sonic Hedgehog (SHH) (n = 14), group 3 (n = 7), group 4 (n = 6), and indeterminate status (n = 3). Three- and 5-year event-free survival (EFS) were 49.0% (30.2-65.4%) and 42.0% (24.2-58.9%) and overall survival (OS) 66.0% (95% CI 46.0-80.1%) and 62.5% (95% CI 42.5 and 77.2%), respectively, with a median follow-up of 9.5 years. Relapse occurred in 12/25 patients following a complete response, of whom six (group 4: n = 4; group 3: n = 1; unknown: n = 1) were successfully salvaged with craniospinal axis (CSA) RT and remain alive at a median follow-up of 7 years. Disease/treatment-related morbidity included endocrinopathies (n = 8), hearing loss n = 16), and neurocognitive abnormalities (n = 9). CONCLUSIONS: This radiation sparing treatment approach for young patients with medulloblastoma resulted in a durable cure in most patients with SHH subgroup medulloblastoma. In those patients with groups 3 and 4 medulloblastoma, relapse rates were high; however, most group 4 patients were salvaged with RT.

Malignant transformation of adult-onset pilocytic astrocytoma to diffuse leptomeningeal glioneuronal tumor within the thoracic spine: a case report and review of the literature.

We describe a 31-year-old male who presented with progressive myelopathy from a thoracic pilocytic astrocytoma (PA). Following multiple recurrences and resections, 10 years after his index surgery, pathology revealed diffuse leptomeningeal glioneuronal tumor (DLGNT) with high-grade features. We discuss his clinical course, management, histopathological findings, and present a comprehensive review of spinal PA undergoing malignant transformation in adults and adult-onset spinal DLGNT. To our knowledge, we present the first reported case of adult-onset spinal PA malignant transformation to DLGNT. Our case adds to the paucity of clinical data characterizing such transformations and highlights the importance of developing novel management paradigms.

Amelanotic melanocytoma of the cervicomedullary junction.

A 64-year-old female with a 1-year history of gait difficulties and right-sided dysesthesias was found to have an intra-axial left-sided exophytic cervicomedullary enhancing mass. Microscopic, immunohistochemical, and ultrastructural findings demonstrated an amelanotic melanocytic neoplasm. Next-generation sequencing with a targeted exomic oncopanel identified a variant of functional significance in the GNA11 gene, thus confirming the diagnosis of a primary amelanotic melanocytoma. The crucial distinction from a melanoma was possible by correlating all of these studies that utilize different technologies.

Deep-learning based classification distinguishes sarcomatoid malignant mesotheliomas from benign spindle cell mesothelial proliferations.

Sarcomatoid mesothelioma is an aggressive malignancy that can be challenging to distinguish from benign spindle cell mesothelial proliferations based on biopsy, and this distinction is crucial to patient treatment and prognosis. A novel deep learning based classifier may be able to aid pathologists in making this critical diagnostic distinction. SpindleMesoNET was trained on cases of malignant sarcomatoid mesothelioma and benign spindle cell mesothelial proliferations. Performance was assessed through cross-validation on the training set, on an independent set of challenging cases referred for expert opinion ('referral' test set), and on an externally stained set from outside institutions ('externally stained' test set). SpindleMesoNET predicted the benign or malignant status of cases with AUC's of 0.932, 0.925, and 0.989 on the cross-validation, referral and external test sets, respectively. The accuracy of SpindleMesoNET on the referral set cases (92.5%) was comparable to the average accuracy of 3 experienced pathologists on the same slide set (91.7%). We conclude that SpindleMesoNET can accurately distinguish sarcomatoid mesothelioma from benign spindle cell mesothelial proliferations. A deep learning system of this type holds potential for future use as an ancillary test in diagnostic pathology.

Targeted RNA expression profiling identifies high-grade endometrial stromal sarcoma as a clinically relevant molecular subtype of uterine sarcoma.

High-grade endometrial stromal sarcoma (HGESS) may harbor YWHAE-NUTM2A/B fusion, ZC3H7B-BCOR fusion, and BCOR internal tandem duplication (ITD). NTRK3 upregulation and pan-Trk expression were reported in soft tissue lesions that share similar morphology and genetic abnormalities. To confirm these findings in HGESS, differential expression analysis was performed at gene level comparing 11 HGESS with 48 other uterine sarcomas, including 9 low-grade endometrial stromal sarcomas, 23 undifferentiated uterine sarcomas, and 16 leiomyosarcomas, using targeted RNA sequencing data. Pan-Trk immunohistochemistry was performed on 35 HGESS, including 10 tumors with RNA expression data, with genotypes previously confirmed by targeted RNA sequencing, fluorescence in situ hybridization, and/or genomic PCR. Unsupervised hierarchical clustering of the top 25% of differentially expressed probes identified three molecular groups: (1) high NTRK3, FGFR3, RET, BCOR, GLI1, and PTCH1 and low ESR1 expression; (2) low NTRK3, FGFR3, RET, BCOR, GLI1, and PTCH1 and high ESR1 expression; and (3) low NTRK3, FGFR3, RET, BCOR, GLI1, PTCH1, and ESR1 expression. Among HGESS, 64% of tumors clustered in group 1, while 27% clustered in group 2. Cytoplasmic and/or nuclear pan-Trk staining of variable extent and intensity was seen in 91% of HGESS regardless of cyclin D1 and/or BCOR positivity. ER and PR expression was seen in 44% of HGESS despite ESR1 downregulation. Two patients with ER and PR positive but ESR1 downregulated stage I HGESS were treated with endocrine therapy, and both recurred at 12 and 36 months after primary resection. By RNA expression, HGESS appear homogenous and distinct from other uterine sarcomas by activation of kinases, including NTRK3, and sonic hedgehog pathway genes along with downregulation of ESR1. Most HGESS demonstrate pan-Trk staining which may serve as a diagnostic biomarker. ESR1 downregulation is seen in some HGESS that express ER and PR which raises implications in the utility of endocrine therapy in these patients.

Novel findings and expansion of phenotype in a mosaic RASopathy caused by somatic KRAS variants.

Mosaic KRAS variants and other RASopathy genes cause oculoectodermal, encephalo-cranio-cutaneous lipomatosis, and Schimmelpenning-Feuerstein-Mims syndromes, and a spectrum of vascular malformations, overgrowth and other associated anomalies, the latter of which are only recently being characterized. We describe eight individuals in total (six unreported cases and two previously reported cases) with somatic KRAS variants and variably associated features. Given the findings of somatic overgrowth (in seven individuals) and vascular or lymphatic malformations (in eight individuals), we suggest mosaic RASopathies (mosaic KRAS variants) be considered in the differential diagnosis for individuals presenting with asymmetric overgrowth and lymphatic or vascular anomalies. We expand the association with embryonal tumors, including the third report of embryonal rhabdomyosarcoma, as well as novel findings of Wilms tumor and nephroblastomatosis in two individuals. Rare or novel findings in our series include the presence of epilepsy, polycystic kidneys, and T-cell deficiency in one individual, and multifocal lytic bone lesions in two individuals. Finally, we describe the first use of targeted therapy with a MEK inhibitor for an individual with a mosaic KRAS variant. The purposes of this report are to expand the phenotypic spectrum of mosaic KRAS-related disorders, and to propose possible mechanisms of pathogenesis, and surveillance of its associated findings.

Making heads or tails - the emergence of capicua (CIC) as an important multifunctional tumour suppressor.

Capicua, encoded by the gene CIC, is an evolutionarily conserved high-mobility group-box transcription factor downstream of the receptor tyrosine kinase and mitogen-activated protein kinase pathways. It was initially discovered and studied in Drosophila. Recurrent mutations in CIC were first identified in oligodendroglioma, a subtype of low-grade glioma. Subsequent studies have identified CIC aberrations in multiple types of cancer and have established CIC as a potent tumour suppressor involved in regulating pathways related to cell growth and proliferation, invasion and treatment resistance. The most well-known and studied targets of mammalian CIC are the oncogenic E-Twenty Six transcription factors ETV1/4/5, which have been found to be elevated in cancers with CIC aberrations. Here, we review the role of CIC in normal mammalian development, oncogenesis and tumour progression, and the functional interactors that mediate them. © 2020 Pathological Society of Great Britain and Ireland. Published by John Wiley & Sons, Ltd.

Synthesis of diagnostic quality cancer pathology images by generative adversarial networks.

Deep learning-based computer vision methods have recently made remarkable breakthroughs in the analysis and classification of cancer pathology images. However, there has been relatively little investigation of the utility of deep neural networks to synthesize medical images. In this study, we evaluated the efficacy of generative adversarial networks to synthesize high-resolution pathology images of 10 histological types of cancer, including five cancer types from The Cancer Genome Atlas and the five major histological subtypes of ovarian carcinoma. The quality of these images was assessed using a comprehensive survey of board-certified pathologists (n = 9) and pathology trainees (n = 6). Our results show that the real and synthetic images are classified by histotype with comparable accuracies and the synthetic images are visually indistinguishable from real images. Furthermore, we trained deep convolutional neural networks to diagnose the different cancer types and determined that the synthetic images perform as well as additional real images when used to supplement a small training set. These findings have important applications in proficiency testing of medical practitioners and quality assurance in clinical laboratories. Furthermore, training of computer-aided diagnostic systems can benefit from synthetic images where labeled datasets are limited (e.g. rare cancers). We have created a publicly available website where clinicians and researchers can attempt questions from the image survey (http://gan.aimlab.ca/). © 2020 Pathological Society of Great Britain and Ireland. Published by John Wiley & Sons, Ltd.

Texture analysis of T2-weighted MRI predicts SDH mutation in paraganglioma.

PURPOSE: Texture analysis can quantify sophisticated imaging characteristics. We hypothesized that 2D textures computed with T2-weighted and post-contrast T1-weighted MRI can predict succinate dehydrogenase (SDH) mutation status in head and neck paragangliomas. METHODS: Our retrospective study included 21 patients (1 to 4 tumors/patient) with 24 pathologically proven paragangliomas in the head and neck. Fourteen lesions (58%) were SDH mutation-positive. All patients underwent T2-weighted and post-contrast T1-weighted MRI sequences. Three 2D texture features of dependence non-uniformity normalized (DNN), small dependence high gray level emphasis (SDHGLE), and small dependence low gray level emphasis (SDLGLE) were calculated. Computed textures between SDH mutants and non-mutants were compared using Mann-Whitney U test. Area under the receiver operating characteristic (AUROC) curve was used to quantify the predictive power of each texture. RESULTS: Only T2-based SDLGLE was statistically significant (p = 0.048), and AUROC was 0.71. Diagnostic accuracy was 70.8%. CONCLUSION: 2D texture parameter of T2-based SDLGLE predicts SDH mutation in head and neck paragangliomas. This noninvasive technique can potentially facilitate further genetic workup.

Histologic Correlates of Molecular Group 4 Pediatric Medulloblastoma: A Retrospective Canadian Review.

INTRODUCTION: The World Health Organization currently classifies medulloblastoma (MB) into four molecular groups (WNT, SHH, Group 3 and Group 4) and four histologic subtypes (classic, desmoplastic nodular, MB with extensive nodularity, and large cell/anaplastic). "Classic" MB is the most frequent histology, but unfortunately it does not predict molecular group or patient outcome. While MB may exhibit additional histologic features outside of the traditional WHO subtypes, the clinical significance of such features, in a molecular context, is unclear. METHODS: The clinicopathologic features of 120 pediatric MB were reviewed in the context of NanoString molecular grouping. Each case was evaluated for five ancillary histologic features, including: nodularity without desmoplasia (i.e., "biphasic", B-MB), rhythmic palisades, and focal anaplasia. Molecular and histological features were statistically correlated to clinical outcome using Chi-square, log-rank, and multivariate Cox regression analysis. RESULTS: While B-MB (N = 32) and rhythmic palisades (N = 12) were enriched amongst non-WNT/SHH MB (especially Group 4), they were not statistically associated with outcome. In contrast, focal anaplasia (N = 12) was not associated with any molecular group, but did predict unfavorable outcome. CONCLUSION: These data nominate B-MB as a surrogate marker of Groups 3 and particularly 4 MB, which may earmark a clinically significant subset of cases.

TRIM25 promotes Capicua degradation independently of ERK in the absence of ATXN1L.

BACKGROUND: Aberrations in Capicua (CIC) have recently been implicated as a negative prognostic factor in a multitude of cancer types through the derepression of targets downstream of the mitogen-activated protein kinase (MAPK) signaling cascade, such as oncogenic E26 transformation-specific (ETS) transcription factors. The Ataxin-family protein ATXN1L has previously been reported to interact with CIC in both developmental and disease contexts to facilitate the repression of CIC target genes and promote the post-translational stability of CIC. However, little is known about the mechanisms at the base of ATXN1L-mediated CIC post-translational stability. RESULTS: Functional in vitro studies utilizing ATXN1LKO human cell lines revealed that loss of ATXN1L leads to the accumulation of polyubiquitinated CIC protein, promoting its degradation through the proteasome. Although transcriptomic signatures of ATXN1LKO cell lines indicated upregulation of the mitogen-activated protein kinase pathway, ERK activity was found to contribute to CIC function but not stability. Degradation of CIC protein following loss of ATXN1L was instead observed to be mediated by the E3 ubiquitin ligase TRIM25 which was further validated using glioma-derived cell lines and the TCGA breast carcinoma and liver hepatocellular carcinoma cohorts. CONCLUSIONS: The post-translational regulation of CIC through ATXN1L and TRIM25 independent of ERK activity suggests that the regulation of CIC stability and function is more intricate than previously appreciated and involves several independent pathways. As CIC status has become a prognostic factor in several cancer types, further knowledge into the mechanisms which govern CIC stability and function may prove useful for future therapeutic approaches.

Differential expression and prognostic relevance of autophagy-related markers ATG4B, GABARAP, and LC3B in breast cancer.

PURPOSE: Previous studies indicate that breast cancer molecular subtypes differ with respect to their dependency on autophagy, but our knowledge of the differential expression and prognostic significance of autophagy-related biomarkers in breast cancer is limited. METHODS: Immunohistochemistry (IHC) was performed on tissue microarrays from a large population of 3992 breast cancer patients divided into training and validation cohorts. Consensus staining scores were used to evaluate the expression levels of autophagy proteins LC3B, ATG4B, and GABARAP and determine the associations with clinicopathological variables and molecular biomarkers. Survival analyses were performed using the Kaplan-Meier function and Cox proportional hazards regression models. RESULTS: We found subtype-specific expression differences for ATG4B, with its expression lowest in basal-like breast cancer and highest in Luminal A, but there were no significant associations with patient prognosis. LC3B and GABARAP levels were highest in basal-like breast cancers, and high levels were associated with worse outcomes across all subtypes (DSS; GABARAP: HR 1.43, LC3B puncta: HR 1.43). High ATG4B levels were associated with ER, PR, and BCL2 positivity, while high LC3B and GABARAP levels were associated with ER, PR, and BCL2 negativity, as well as EGFR, HER2, HER3, CA-IX, PD-L1 positivity, and high Ki67 index (p < 0.05 for all associations). Exploratory multi-marker analysis indicated that the combination of ATG4B and GABARAP with LC3B could be useful for further stratifying patient outcomes. CONCLUSIONS: ATG4B levels varied across breast cancer subtypes but did not show prognostic significance. High LC3B expression and high GABARAP expression were both associated with poor prognosis and with clinicopathological characteristics of aggressive disease phenotypes in all breast cancer subtypes.

High-grade transformation of low-grade endometrial stromal sarcomas lacking YWHAE and BCOR genetic abnormalities.

High-grade histologic transformation of low-grade endometrial stromal sarcoma (LGESS) is rare. Here, we describe the clinicopathologic features and gene fusion status of 12 cases (11 primary uterine corpus and 1 primary vaginal), 11 diagnosed prospectively from 2016, and 1 retrospectively collected. Targeted RNA sequencing and/or fluorescence in situ hybridization was employed in all cases. High-grade transformation was seen at the time of initial diagnosis in eight patients and at the time of recurrence in four patients, 4-11 years after initial diagnosis of LGESS. High-grade morphology consisted of generally uniform population of round to epithelioid cells with enlarged nuclei one to two times larger than a lymphocyte, visible nucleoli, and increased mitotic index (range, 6-30; median, 16 per 10 high-power fields); there was often an associated sclerotic and/or myxoid stroma. Estrogen receptor, progesterone receptor, and CD10 expression was absent or significantly decreased (compared with the low-grade component) in the high-grade foci of five tumors. One tumor demonstrated positive (diffuse and strong) cyclin D1 and BCOR staining. p53 staining was wild type in both components of all eight tumors tested. JAZF1-SUZ12 (n = 6), JAZF1-PHF1 (n = 3), EPC1-PHF1, (n = 1), or BRD8-PHF1 (n = 1) fusions were detected in 11 tumors; no fusions were found in one by targeted RNA sequencing. Patients presented with FIGO stages I (n = 4), II (n = 4), III (n = 1), and IV disease (n = 2). Median overall survival calculated from the time of histologic transformation was 22 months (range, 8 months to 8 years) with five patients who died of disease 8-18 months after transformation. High-grade transformation may occur in LGESS with JAZF1 and PHF1 rearrangements at the time of or years after initial diagnosis. Such high-grade transformation is characterized by nuclear enlargement, prominent nucleoli, and increased mitotic index compared with typical LGESS. Histologic high-grade transformation may herald aggressive behavior.

Interpretable multimodal deep learning for real-time pan-tissue pan-disease pathology search on social media.

Pathologists are responsible for rapidly providing a diagnosis on critical health issues. Challenging cases benefit from additional opinions of pathologist colleagues. In addition to on-site colleagues, there is an active worldwide community of pathologists on social media for complementary opinions. Such access to pathologists worldwide has the capacity to improve diagnostic accuracy and generate broader consensus on next steps in patient care. From Twitter we curate 13,626 images from 6,351 tweets from 25 pathologists from 13 countries. We supplement the Twitter data with 113,161 images from 1,074,484 PubMed articles. We develop machine learning and deep learning models to (i) accurately identify histopathology stains, (ii) discriminate between tissues, and (iii) differentiate disease states. Area Under Receiver Operating Characteristic (AUROC) is 0.805-0.996 for these tasks. We repurpose the disease classifier to search for similar disease states given an image and clinical covariates. We report precision@k = 1 = 0.7618 ± 0.0018 (chance 0.397 ± 0.004, mean ±stdev ). The classifiers find that texture and tissue are important clinico-visual features of disease. Deep features trained only on natural images (e.g., cats and dogs) substantially improved search performance, while pathology-specific deep features and cell nuclei features further improved search to a lesser extent. We implement a social media bot (@pathobot on Twitter) to use the trained classifiers to aid pathologists in obtaining real-time feedback on challenging cases. If a social media post containing pathology text and images mentions the bot, the bot generates quantitative predictions of disease state (normal/artifact/infection/injury/nontumor, preneoplastic/benign/low-grade-malignant-potential, or malignant) and lists similar cases across social media and PubMed. Our project has become a globally distributed expert system that facilitates pathological diagnosis and brings expertise to underserved regions or hospitals with less expertise in a particular disease. This is the first pan-tissue pan-disease (i.e., from infection to malignancy) method for prediction and search on social media, and the first pathology study prospectively tested in public on social media. We will share data through http://pathobotology.org . We expect our project to cultivate a more connected world of physicians and improve patient care worldwide.

Improved structural variant interpretation for hereditary cancer susceptibility using long-read sequencing.

PURPOSE: Structural variants (SVs) may be an underestimated cause of hereditary cancer syndromes given the current limitations of short-read next-generation sequencing. Here we investigated the utility of long-read sequencing in resolving germline SVs in cancer susceptibility genes detected through short-read genome sequencing. METHODS: Known or suspected deleterious germline SVs were identified using Illumina genome sequencing across a cohort of 669 advanced cancer patients with paired tumor genome and transcriptome sequencing. Candidate SVs were subsequently assessed by Oxford Nanopore long-read sequencing. RESULTS: Nanopore sequencing confirmed eight simple pathogenic or likely pathogenic SVs, resolving three additional variants whose impact could not be fully elucidated through short-read sequencing. A recurrent sequencing artifact on chromosome 16p13 and one complex rearrangement on chromosome 5q35 were subsequently classified as likely benign, obviating the need for further clinical assessment. Variant configuration was further resolved in one case with a complex pathogenic rearrangement affecting TSC2. CONCLUSION: Our findings demonstrate that long-read sequencing can improve the validation, resolution, and classification of germline SVs. This has important implications for return of results, cascade carrier testing, cancer screening, and prophylactic interventions.