Molecular-based Classification Research Articles

Abstract PO3-07-04: Prediction of PAM50 molecular subtypes from H&E-stained breast cancer specimens using tumor microenvironment features and additive multiple instance learning models

Abstract Background: PAM50, a 50-gene signature, classifies breast cancers into one of five subtypes (basal, luminal A, luminal B, HER2-enriched, and normal-like), revealing information about underlying tumor biology, and has emerged as a key prognostic indicator influencing treatment decisions. There is growing interest in bridging the gap between expression-based metrics and histopathology, where immunohistochemistry (IHC) and sequencing-based approaches have been proposed for this purpose. However, hematoxylin and eosin (H&E)-stained slides are ubiquitously utilized by pathologists for cancer diagnosis, while IHC and sequencing-based approaches require additional tissue and specialized processing and/or analysis. Here, we describe a computer vision-based approach to predict PAM50 classification using H&E-stained whole slide images (WSIs). Methods: We obtained expression-based PAM50 subtype labels and corresponding H&E-stained WSIs for 961 breast carcinomas from the TCGA BRCA cohort. We used two separate machine learning (ML) approaches to predict PAM50 subtypes from WSIs. In the first approach, we deployed previously trained PathExplore models to extract quantitative human-interpretable features (HIFs) that summarize the TME. We subsequently trained random forest classification models on these HIFs to predict PAM50 subtypes. For the second approach, we developed additive multiple instance learning (aMIL) models. Additionally, we explored the effects of PAM50 subtype labeling and aggregation strategies beyond the 5-class approach. Our 3-class approach combines Luminal A and B, as seen in IHC efforts to increase agreement with PAM50 assays, while excluding Normal, a category containing few and heterogeneous samples. We also performed binary classification for each subtype in the 3-class model (e.g. luminal vs. other). Slides were split into training (60%), validation (20%), and test (20%) sets, stratified by PAM50 labels, and model performance was assessed using the area under the receiver operator curve (AUROC) metric on the held-out test set, using a one vs. rest approach for multi-class models. To establish a baseline for PAM50 prediction, we developed random forest classification models using only clinical covariates (tumor stage, histologic grade, histological subtype, and BRCA1/2 status). Results: We compared the performance of our two ML models (HIF and aMIL) to that of the baseline model, and we report the AUROC values in Table 1. These models both performed well in predicting Basal, Luminal A, Luminal B, and Luminal (A+B), while the model performance was less strong for predictions of the HER2 and Normal classifications. The three-class model showed improved performance of predicting Luminal classifications relative to the five-class model that separates Luminal A and B. Although simplifying classification problems to a binary use case typically provides improved performance, this phenomenon was not observed for any of the PAM50 subtypes. Conclusions: These results demonstrate that AI-powered digital pathology can accurately and reproducibly perform molecular-based classification tasks, such as predicting PAM50 classifications, using WSIs, suggesting a more efficient path toward clinically relevant breast cancer characterization. Table 1. Performance of all models in predicting PAM50 molecular subtypes. AUROC values are shown. Shaded cells represent the best test-set performance for each class (row). Citation Format: Maria Guramare, Syed Ashar Javed, Christian Kirkup, Dinkar Juyal, Jacqueline Brosnan-Cashman, Victoria Mountain, Ryan Leung, Bahar Rahsepar, John Abel, Amaro Taylor-Weiner, Jake Conway. Prediction of PAM50 molecular subtypes from H&E-stained breast cancer specimens using tumor microenvironment features and additive multiple instance learning models [abstract]. In: Proceedings of the 2023 San Antonio Breast Cancer Symposium; 2023 Dec 5-9; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2024;84(9 Suppl):Abstract nr PO3-07-04.

Molecular-based classification of endometrial carcinoma in Northern Thailand: impact on prognosis and potential for implementation in resource-limited settings

BackgroundEndometrial carcinoma is molecularly categorized into four subgroups: polymerase-E exonuclease domain-mutant (POLE-mut), mismatch repair-deficient (MMR-d), p53-abnormal (p53-abn), and no specific molecular profile (NSMP). This classification scheme has been included into clinical recommendation for post-operative risk-based management, although there have been few Asian studies on this topic. The present study aimed to evaluate the prevalence and clinical outcomes of endometrial carcinoma using this classification in Northern Thailand and the feasibility of implementation in resource-limited settings.MethodsEndometrial carcinomas from hysterectomy specimens were classified using immunohistochemistry for MMR proteins and p53, as well as POLE mutation testing. Clinicopathological variables and outcomes were analyzed. The costs of the molecular information-based approach were compared to those incurred by the conventional approach (without molecular classification).ResultsOf 138 patients, 52.9% in the NSMP subgroup, 28.2% were in the MMR-d, 13.8% in the p53-abn, and 5.1% in the POLE-mut. After adjusting for other variables, patients with POLE-mut showed the most favorable outcomes, while those with p53-abn had the poorest survival. When estimating the costs for post-operative management, the use of molecular classification resulted in a 10% increase over the conventional approach. However, the cost increased only by 1% if only POLE testing was used to identify patients for treatment omission.ConclusionIn Northern Thailand, endometrial carcinoma had comparable subgroup distribution and prognostic implications to previous reports, supporting the implementation of management guidelines that incorporate molecular information. In resource-limited settings, at least POLE mutation testing in early-stage patients should be considered.

GlioPredictor: A Deep Learning Model for Identification of High-Risk Low-Grade Glioma toward Adjuvant Treatment Planning

High-risk low-grade glioma (LGG) patients are recommended to undergo adjuvant radiotherapy whereas watchful waiting is recommended for low-risk LGG patients per the latest NCCN guidelines. Based on Radiation Therapy Oncology Group (RTOG) 9802, high-risk features include age >40 or subtotal resection (STR). However, in the era of molecular-based classification for tumors of central neural system, current risk classification criteria based on gross disease and patient demographics may be outdated. Here, we aim to develop a molecular-based glioma risk classification system (GlioPredictor) that could potentially facilitate identification of high-risk LGG patients. A total of 507 LGG cases from The Cancer Genome Atlas-low grade glioma (TCGA-LGG), and 1,309 cases from AACR GENIE v13.0 datasets were studied for genetic disparities between IDH1-wildtype and mutated cohorts, and varying age groups. Through a feature selection technique using genomic profiling and correlation analyses, features such as mutation status, copy number variations (CNVs), among other clinicopathologic features prognostic of IDH1 mutation status were selected as potential inputs to train an artificial neural networks (ANNs) that could predict IDH1 mutation status. Model performance was assessed using the area under the receiver operating characteristic curve (AUC). Memorial Sloan Kettering (MSK) dataset (n = 404) for LGG was used to cross-validate the trained ANN. The optimized ANN model has 6 layers with 6 input nodes, 20 hidden nodes, and a binary output layer. The weights and biases of the hidden layers of the best-performing model were retrieved and reconstructed to yield the GlioPredictor score-the predicted risk of progression for IDH1-wildtype LGG. Over 81% of glioma patients age less than 40 have IDH1 mutation, as compared with 31% in those age above 60. Using age > 40 as a cutoff failed to identify high-risk IDH1-mutant LGG with early progression. IDH1 mutation is associated with decreased CNVs of EGFR (21 % vs. 3%), CDKN2A (20% vs. 6%) and PTEN (14% vs. 1.7%), and increased percentage of mutations for TP53 (15% vs. 63%), and ATRX (10% vs. 54%) (p<0.001). Using these molecular features, along with the patient's age, an ANN model with 6 layers and 20 hidden nodes can predict IDH1 mutation status with over 90% accuracy and AUC score over 0.91. We have developed an ANN model that is capable of learning the prognostic features of LGG associated with an IDH1-mutated LGG cohort and using the features to predict high-risk patients from the IDH1-wildtype cohort. This ANN model facilitates the selection of LGG patients who could benefit from immediate adjuvant radiotherapy. Future work includes the integration of image features to improve the prediction performance of the GlioPredictor system.

Mutational Landscape of ER Receptor Negative Endometrial Cancer Patients Categorized as No Specific Molecular Profile (NSMP)

The endometrial cancer (EC) molecular classification introduced by The Cancer Genome Atlas (TCGA) has initiated molecular-based classification with clear prognostic value. Four major tumor subtypes exist: 1) TP53 mutation (p53abn); 2) POLE mutation (POLEmut); 3) mismatch repair protein deficient (MMRd); and 4) no specific molecular profile (NSMP). Of these subtypes, p53abn patients exhibited the worst prognosis. Recently, a prognostic refinement of NSMP EC using estrogen receptor (ER) status, from PORTEC-III trial data, showed that (ER-) NSMP patients had a strikingly poor prognosis, similar to p53abn patients. Currently, drivers of NSMP (ER-) disease are unknown and require genomic characterization. In this study, we perform an integrated molecular analysis to characterize this aggressive and poorly understood cohort. RNAseq and mutational data was downloaded from the TCGA, PanCancer Atlas, Uterine Endometrial Carcinoma dataset via the genomic data commons (GDC). All patients in this cohort demonstrated endometrioid histology. P53abn, and POLE mutated patients were filtered using mutational data. MMRd and NSMP ESR1 gene low expressors (ER Low) were filtered using Z-score cutoffs of RNAseq expression data. Survival analysis was performed using the Survival R-package. Gene expression testing was performed using the EdgeR R-package. Pathway Enrichment analysis was conducted using EnrichR software. Forty-two patients out of 529 samples met our GDC NSMP (ER low) filtering criteria (7%), similar to the 5% of PORTEC-III patients classified as NSMP (ER -). Strikingly 83% percent of GDC NSMP (ER low) patients harbored a mutation in the PI3K-AKT-mTOR signaling pathway, with 48% of GDC NSMP (ER low) patients carrying a mutation in the p110 alpha (PIK3CA) gene. Interestingly, of the patients with PIK3CA mutant tumors, 2/20 patients died in 5 years (10%), compared to 9/22 patients whose tumors were PIK3CA WT (41%) (p = 0.02). Among GDC NSMP (ER low) patients with PIK3CA WT tumors gene expression signatures were enriched for MYC target genes (p = 1.1e-51), DNA replication (p = 4.3e-30), cell cycle (p = 9.32e-20), and cell cycle checkpoints (p = 6.4e-20) pathways. Additionally, strongly in PIK3CA WT tumors are mitochondrial membrane proteins (p = 1.63e-37) and ribosomal proteins (2.9e-24). In this analysis, we show that GDC NSMP (ER low) tumors nearly all harbor a mutation in the PI3K-AKT-mTOR signaling pathway. Further we demonstrate that the most common gene mutation in the cohort, PIK3CA, is counterintuitively a marker of improved survival. Additionally, we show that within that this subpopulation, PIK3CA WT patients exhibit robustly upregulated gene expression programs dedicated to energy production, cell cycling, and division.

P03.03.A T2-FLUID-ATTENUATED INVERSION RECOVERY MISMATCH SIGN IN LOWER-GRADE GLIOMAS: CORRELATION WITH PATHOLOGICAL AND MOLECULAR FINDINGS

Abstract BACKGROUND After the new molecular-based classification was reported to be useful for predicting prognosis, the T2-fluid-attenuated inversion recovery (FLAIR) mismatch sign has gained interest as one of the promising methods for detecting lower-grade gliomas (LGGs) with isocitrate dehydrogenase (IDH) mutations and chromosome 1p/19q non-codeletion (IDH mut-Noncodel) with high specificity. Although all institutions could use T2-FLAIR mismatch sign without any obstacles, this sign was not completely helpful because of its low sensitivity. In this study, we attempted to uncover the mechanism of T2-FLAIR mismatch sign for clarifying the cause of this sign’s low sensitivity. MATERIAL AND METHODS In this study, data on 101 consecutive and unselected patients with LGGs (WHO grades 2 and 3) admitted to Miyazaki University Hospital between November 2005 and January 2021 were collected. The inclusion criteria were as follows: (1) obtained molecular information on IDH mutation and chromosome 1p/19q codeletion status and (2) obtained preoperative MRI data, including T2-weighted imaging and FLAIR sequences. In this study, 99 patients were enrolled and then classified into three groups based on IDH mutation and chromosome 1p/19q codeletion status: (1) IDH-mutant, 1p/19q-codeleted LGGs (IDH mut-codel); (2) IDH-mutant, 1p/19q-noncodeleted LGGs (IDH mut-Noncodel); and (3) IDH wild-type LGGs (IDH wt). RESULTS Among 99 patients with LGGs, 22 were T2-FLAIR mismatch sign-positive (22%), and this sign as a marker of IDH mut-Noncodel showed a sensitivity of 55.6% and specificity of 96.8%. Via pathological analyses, we could provide evidence that not only microcystic changes but the enlarged intercellular space was associated with T2-FLAIR mismatch sign (p =0.017). As per the molecular analyses, overexpression of mTOR-related genes (m-TOR, RICTOR) were detected as the molecular events correlated with T2-FLAIR mismatch sign (p = 0.020, 0.030. respectively). CONCLUSION We suggested that T2-FLAIR mismatch sign could pick up the IDH mut-Noncodel LGGs with enlarged intercellular space or that with overexpression of mTOR-related genes.

Medulloblastoma

In the 5th edition of the WHO classification, medulloblastomas, which are representative pediatric brain tumors, are categorized into four groups: WNT, SHH-TP53 wild, SHH-TP53 mutant, and non-WNT/non-SHH, based on their molecular background. While the histopathological findings still hold importance in predicting prognosis, the histopathological classification is no longer utilized in this edition. SHH medulloblastomas are further subdivided into two groups based on the presence or absence of TP53 mutation, as their clinical characteristics and prognosis differ. Group 3 and Group 4 medulloblastomas, recognized as distinct molecular groups in clinical practice, are combined into a single group called "non-WNT/non-SHH", because they lack specific molecular pathway activation. Furthermore, based on methylation profiling, dividing SHH medulloblastoma into four subgroups and non-WNT/non-SHH medulloblastoma into eight subgroups was proposed. Understanding the unique clinical characteristics and prognosis associated with each group is crucial. However, it is important to acknowledge that our current understanding of prognosis is based on treatment approaches guided by clinical risk factors such as postoperative residual tumor volume and the presence of metastatic disease. This molecular-based classification holds promise in guiding the development of optimal treatment strategies for patients with medulloblastoma.

The Neurodevelopmental and Molecular Landscape of Medulloblastoma Subgroups: Current Targets and the Potential for Combined Therapies.

Medulloblastoma is the most common malignant pediatric brain tumor and is associated with significant morbidity and mortality in the pediatric population. Despite the use of multiple therapeutic approaches consisting of surgical resection, craniospinal irradiation, and multiagent chemotherapy, the prognosis of many patients with medulloblastoma remains dismal. Additionally, the high doses of radiation and the chemotherapeutic agents used are associated with significant short- and long-term complications and adverse effects, most notably neurocognitive delay. Hence, there is an urgent need for the development and clinical integration of targeted treatment regimens with greater efficacy and superior safety profiles. Since the adoption of the molecular-based classification of medulloblastoma into wingless (WNT) activated, sonic hedgehog (SHH) activated, group 3, and group 4, research efforts have been directed towards unraveling the genetic, epigenetic, transcriptomic, and proteomic profiles of each subtype. This review aims to delineate the progress that has been made in characterizing the neurodevelopmental and molecular features of each medulloblastoma subtype. It further delves into the implications that these characteristics have on the development of subgroup-specific targeted therapeutic agents. Furthermore, it highlights potential future avenues for combining multiple agents or strategies in order to obtain augmented effects and evade the development of treatment resistance in tumors.

Current status and future perspectives on the methods of prognosis evaluation for intrahepatic cholangiocarcinoma

Intrahepatic cholangiocarcinoma (ICC) is the second most common primary malignant tumor in the liver after hepatocellular carcinoma. Its incidence and mortality rates have increased worldwide in recent years. Surgical resection is the best treatment modality for ICC;however,the overall prognosis remains poor. Accurate evaluation of post operative prognosis allows personalized treatment and improved long-term outcomes of ICC. The American Joint Commission on Cancer TNM staging manual is the basis for the standardized diagnosis and treatment of ICC;however,the contents of stage T and stage N need to be improved. The nomogram model or scoring system established in the analysis of commonly used clinicopathological parameters can provide individualized prognostic evaluation and improve prediction accuracy;however,more studies are needed to validate the results before clinical use. Meanwhile,imaging features exhibit great potential to establish the post operative prognosis evaluation system for ICC. Molecular-based classification provides an accurate guarantee for prognostic assessment as well as selection of populations that are sensitive to targeted therapy or immunotherapy. Therefore,the establishment of a prognosis evaluation system,based on clinical and pathological characteristics and centered on the combination of multidisciplinary and multi-omics,will be conducive to improving the long-term outcomes of ICC after surgical resection in the context of big medical data.

A novel federated deep learning scheme for glioma and its subtype classification.

Deep learning (DL) has shown promising results in molecular-based classification of glioma subtypes from MR images. DL requires a large number of training data for achieving good generalization performance. Since brain tumor datasets are usually small in size, combination of such datasets from different hospitals are needed. Data privacy issue from hospitals often poses a constraint on such a practice. Federated learning (FL) has gained much attention lately as it trains a central DL model without requiring data sharing from different hospitals. We propose a novel 3D FL scheme for glioma and its molecular subtype classification. In the scheme, a slice-based DL classifier, EtFedDyn, is exploited which is an extension of FedDyn, with the key differences on using focal loss cost function to tackle severe class imbalances in the datasets, and on multi-stream network to exploit MRIs in different modalities. By combining EtFedDyn with domain mapping as the pre-processing and 3D scan-based post-processing, the proposed scheme makes 3D brain scan-based classification on datasets from different dataset owners. To examine whether the FL scheme could replace the central learning (CL) one, we then compare the classification performance between the proposed FL and the corresponding CL schemes. Furthermore, detailed empirical-based analysis were also conducted to exam the effect of using domain mapping, 3D scan-based post-processing, different cost functions and different FL schemes. Experiments were done on two case studies: classification of glioma subtypes (IDH mutation and wild-type on TCGA and US datasets in case A) and glioma grades (high/low grade glioma HGG and LGG on MICCAI dataset in case B). The proposed FL scheme has obtained good performance on the test sets (85.46%, 75.56%) for IDH subtypes and (89.28%, 90.72%) for glioma LGG/HGG all averaged on five runs. Comparing with the corresponding CL scheme, the drop in test accuracy from the proposed FL scheme is small (-1.17%, -0.83%), indicating its good potential to replace the CL scheme. Furthermore, the empirically tests have shown that an increased classification test accuracy by applying: domain mapping (0.4%, 1.85%) in case A; focal loss function (1.66%, 3.25%) in case A and (1.19%, 1.85%) in case B; 3D post-processing (2.11%, 2.23%) in case A and (1.81%, 2.39%) in case B and EtFedDyn over FedAvg classifier (1.05%, 1.55%) in case A and (1.23%, 1.81%) in case B with fast convergence, which all contributed to the improvement of overall performance in the proposed FL scheme. The proposed FL scheme is shown to be effective in predicting glioma and its subtypes by using MR images from test sets, with great potential of replacing the conventional CL approaches for training deep networks. This could help hospitals to maintain their data privacy, while using a federated trained classifier with nearly similar performance as that from a centrally trained one. Further detailed experiments have shown that different parts in the proposed 3D FL scheme, such as domain mapping (make datasets more uniform) and post-processing (scan-based classification), are essential.

808 Analysis of Medulloblastomas Subgroups in Pediatric Patients in Latin-America

INTRODUCTION: The 4 groups molecular-based classification of medulloblastomas is widely accepted. However, technology inaccessibility and low-quality formalin-fixed and paraffin-embedded (FFPE) samples, unable its application for most Latin-American cancer centers. METHODS: The goal of this study was to perform a feasibility analysis based on 18 mRNA expression by qPCR integrated with immunohistochemistry (IH) and fluorescent hybridization in situ (FISH) (c-MYC, n-MYC and monosomy 6) for molecular stratification to create an International Treatment Protocol at a low cost for low-income centers. RESULTS: Fifty-four FFPE samples were classified according to histologic and molecular aspects. Classic medulloblastoma was the commonest histology (53.7%), followed by desmoplastic/extensive nodularity (24.1%), NOS (16,7%) and anaplastic (5,6%). IH markers classified patients in three groups (WNT, SHH and non-WNT/non-SHH) in 98% of cases. PCR-based method confirmed results from IH in 81.5%. Additionally, it detected WNT activation in 2 more patients, previously classified as SHH. For both cases, the presence of monosomy 6 confirmed WNT subgroup. The integration of these three techniques resulted in the following frequencies: WNT (13.0%), SHH (38.9%), group 3 (9.3%), group 4 (20.3%) and non-WNT/non-SHH (18.5%). From 40 patients with clinical follow-up available, 3-year overall survival (n = 40) for low, intermediate and high-risk groups were 100%, 60% and 20%, respectively (p &lt; 0.05), based only in molecular criteria, which confirmed the prognostic importance of this method at a low cost. CONCLUSIONS: The prognosis of the treatment of medulloblastomas has changed after novel studies on the molecular subgroups and the introduction of target-based therapy. The current protocol will enable different centers to implement treatment based on the knowledge of molecular subgroups.

Molecular-based race classification of Pyrenophora tritici-repentis causing tan spot of wheat in Japan.

Tan spot, a foliar disease of Triticum spp. such as bread wheat (T. aestivum L.) and durum wheat (T. turgidum ssp. durum (Desf.) Husn.) caused by the filamentous fungus Pyrenophora tritici-repentis (Died.) Drechsler leads to serious losses of crop yield and quality in some areas in Japan. P. tritici-repentis is classified into eight races according to the combinations of three necrotrophic effectors, PtrToxA, PtrToxB, and PtrToxC encoded by ToxA, ToxB, and ToxC1, respectively. Race classification has been based on reaction of a differential variety to necrotrophic effectors, which is tested by inoculation. Recent identification of the Tox genes and development of specific DNA markers have enabled us to classify races of P. tritici-repentis collected in Japan by Tox gene genotyping. We found that 17 strains collected from Triticum spp. in Japan were mainly race 1 or 2, because they carried ToxA as a toxin gene by current race classification; wheat genotype tsn1 is resistant to ToxA. Establishment of wheat cultivars carrying tsn1 would be most effective for decreasing agronomic losses caused by the disease in Japan.

Panelsequenzierung in der Kopf-Hals-Onkologie am Beispiel von sinunasalen und Speicheldrüsenmalignomen

BackgroundAs a result of the increasing implementation of innovative molecular technologies during the last two decades, the pathology of salivary gland carcinomas and sinonasal carcinomas has witnessed a tremendous revolution.AimWith the introduction of molecular tumor boards as a central opportunity exploring the possibility of actionable molecular targets in head and neck oncology, the question has been frequently raised as to which entities/patients might represent good or best candidates for OncoPanel sequencing in the context of these regularly implemented molecular tumor boards. The main aim of this review is to address this question considering recent developments and in context of the current World Health Organization (WHO) classification.Materials and methodsThis review is based on a selective literature search (PubMed) using the terms salivary gland malignancy, sinonasal carcinoma and targeted therapies as well as the authors’ own data and observations.Results and conclusionAs a result of several recent seminal studies on salivary and sinonasal carcinomas, some old concepts have been reconceptualized, some new concepts introduced, and some old, established entities became better defined molecularly. These achievements have influenced not only diagnostics (they enable a much refined, mostly molecular-based classification of many entities that were previously unclassified) but they also have a significant therapeutic impact (they enable better histology-tailored therapeutic stratification, at the same time some of them represent valuable biomarkers predicting response to targeted therapies).

Olfactory Neuroblastoma: Morphological Reappraisal and Molecular Insights with Quantum Leap in Clinical Perspectives.

The purpose of review is to provide a comprehensive review of the literature focusing on the recent advances in the diagnosis, molecular underpinning, and targeted therapy of olfactory neuroblastoma (ONB). Studies focused on the molecular fingerprinting of ONB are critical to engage new promising treatment strategies. Molecular-based subtype classifications have been proposed (basal-like ONB and neural-like ONB) but are not widely used. The rationale for implementation of DNA methylation analysis and IDH2 sequencing in routine work-up for ONB is gaining recognition. Expression of somatostatin receptors (SSTR) in ONB open new avenues for both, diagnostic (especially metastatic disease) and new treatment protocols with somatostatin analogs. Olfactory carcinoma is proposed as a unifying diagnostic terminology pertinent to epithelial divergent differentiation in olfactory neuroblastoma. Molecular (genetic and epigenetic) efforts on olfactory neuroblastoma are promising; however further refinement is needed for employment of these biomarkers as clinical standard of care. Ongoing and future multi-institutional collaborative studies will contribute to further understanding of ONB biology and aid the development of targeted treatments for this disease.

Correlative Assessment of p53 Immunostaining Patterns and TP53 Mutation Status by Next-Generation Sequencing in High-Grade Endometrial Carcinomas.

TP53 mutations are frequently identified in the copy number-high molecular subgroup of endometrial carcinomas (ECs). P53 immunohistochemistry (IHC) is a widely used surrogate marker reflecting the mutational status of TP53 , and recent reports have shown ~95% concordance between the two methods in ECs. While these results are promising, studies evaluating the correlation between different p53 IHC staining patterns and comprehensive next-generation sequencing results are still limited. We compared the p53 IHC staining patterns, scored as wild-type, diffuse nuclear overexpression, null/complete absence, and cytoplasmic, to next-generation sequencing results reported by FoundationOneCDx in 43 high-grade ECs: 20 serous ECs, 9 mixed ECs with a serous component, 4 carcinosarcomas with a serous component, and 10 grade 3 endometrioid ECs. The concordance of p53 IHC and TP53 mutation status was 100% (43/43) overall, including 100% (33/33) concordance in tumors with a serous component and 100% (10/10) in endometrioid ECs. Among the 35 tumors with aberrant p53 expression the most commonly observed pattern was diffuse nuclear overexpression seen in 69% (24/35), followed by cytoplasmic staining in 17% (6/35), and complete absence of staining (null) in 14% (5/35) of tumors. Of the 6 tumors with cytoplasmic staining, 4 corresponded to missense mutations within the DNA binding domain (V157F in 2 tumors, and S127P and R280S, in 2 tumor each), while 2 corresponded to nonsense mutations in the tetramerization domain (p.E339*). Our results further support that p53 IHC can serve as an accurate predictor of TP53 alterations in ECs to aid the molecular-based tumor classification and the distinction between tumor histotypes, both of which play an important role in the assessment of clinical prognosis and therapeutic decision making. In addition, our data suggest, that the type and position of TP53 mutation may not directly correlate with the observed p53 IHC pattern in all tumors, and that there may be alternative mechanisms for cytoplasmic localization (other than mutations involving the nuclear localization domain), possibly due to conformational changes or posttranslational modifications of the aberrant p53 protein.

Prognostic impact of MYD88 and CXCR4 mutations assessed by droplet digital polymerase chain reaction in IgM monoclonal gammopathy of undetermined significance and smouldering Waldenström macroglobulinaemia.

Waldenström macroglobulinaemia (WM) is characterized by recurrent somatic mutations in MYD88 and CXCR4 genes. However, limitations arise when analysing these mutations in IgM monoclonal gammopathy of undetermined significance (MGUS) or smouldering WM (SWM) given the lower tumour load. Here, we used droplet digital polymerase chain reaction (ddPCR) to analyse MYD88 L265P and CXCR4 S338* mutations (C1013G and C1013A) in unsorted bone marrow (BM) or cell-free DNA (cfDNA) samples from 101 IgM MGUS and 69 SWM patients. ddPCR was more sensitive to assess MYD88 L265P compared to allele-specific PCR, especially in IgM MGUS (64% vs 39%). MYD88 mutation burden correlated with other laboratory biomarkers, particularly BM infiltration (r=0.8; p < 0.001). CXCR4 C1013G was analysed in MYD88-mutated samples with available genomic DNA and was detected in 19/54 (35%) and 18/42 (43%) IgM MGUS and SWM cases respectively, also showing correlation with BM involvement (r=0.9; p < 0.001). ddPCR also detected 8 (38%) and 10 (63%) MYD88-mutated cfDNA samples in IgM MGUS and SWM respectively. Moreover, high BM mutation burden (≥8% MYD88 and ≥2% CXCR4) was associated with an increased risk of progression to symptomatic WM. We show the clinical applicability of ddPCR to assess MYD88 and CXCR4 in IgM MGUS and SWM and provide a molecular-based risk classification.

The evolving role of morphology in endometrial cancer diagnostics: From histopathology and molecular testing towards integrative data analysis by deep learning

Endometrial cancer (EC) diagnostics is evolving into a system in which molecular aspects are increasingly important. The traditional histological subtype-driven classification has shifted to a molecular-based classification that stratifies EC into DNA polymerase epsilon mutated (POLEmut), mismatch repair deficient (MMRd), and p53 abnormal (p53abn), and the remaining EC as no specific molecular profile (NSMP). The molecular EC classification has been implemented in the World Health Organization 2020 classification and the 2021 European treatment guidelines, as it serves as a better basis for patient management. As a result, the integration of the molecular class with histopathological variables has become a critical focus of recent EC research. Pathologists have observed and described several morphological characteristics in association with specific genomic alterations, but these appear insufficient to accurately classify patients according to molecular subgroups. This requires pathologists to rely on molecular ancillary tests in routine workup. In this new era, it has become increasingly challenging to assign clinically relevant weights to histological and molecular features on an individual patient basis. Deep learning (DL) technology opens new options for the integrative analysis of multi-modal image and molecular datasets with clinical outcomes. Proof-of-concept studies in other cancers showed promising accuracy in predicting molecular alterations from H&E-stained tumor slide images. This suggests that some morphological characteristics that are associated with molecular alterations could be identified in EC, too, expanding the current understanding of the molecular-driven EC classification. Here in this review, we report the morphological characteristics of the molecular EC classification currently identified in the literature. Given the new challenges in EC diagnostics, this review discusses, therefore, the potential supportive role that DL could have, by providing an outlook on all relevant studies using DL on histopathology images in various cancer types with a focus on EC. Finally, we touch upon how DL might shape the management of future EC patients.

Glioneuronal and neuronal tumors: A review with special reference to the new types in the WHO CNS5 classification.

Glioneuronal and neuronal tumors (GNTs) are slow-growing lower-grade neuroepithelial tumors with mature neuronal and, less consistently, glial differentiation. Their identification has relied solely on histological proof of neuronal differentiation, which was considered to represent the well-differentiated nature of GNTs. However, after discovering the genetic alterations in GNTs, particularly those in the MAP-kinase pathway, it became evident that histological diagnoses are not always concurrent with genetic alterations and vice versa. Furthermore, since several inhibitors mediating the MAP-kinase pathway are available, at least for clinical trials, molecular-based classification is now warranted. Thus, the upcoming WHO Classification of Central Nervous System Tumors, 5th edition (WHO5CNS) applied DNA methylation profiling to segregate low-grade neuroepithelial tumors. This review gives an overview of the pathological features of GNTs with particular reference to the newly listed tumor types in WHO5CNS. The knowledge and awareness of each tumor type are essential to make a correct diagnosis and avoid unnecessary radical resection and chemoradiotherapy, as GNTs are relatively indolent and have a prolonged clinical course. In addition, being distinctive in location, age group, and histology, the integration of clinicopathological information will help identify relevant tumor types of GNTs without genetic testing, even in resource-limited settings.

T2-fluid-attenuated inversion recovery mismatch sign in lower grade gliomas: correlation with pathological and molecular findings.

After the new molecular-based classification was reported to be useful for predicting prognosis, the T2-fluid-attenuated inversion recovery (FLAIR) mismatch sign has gained interest as one of the promising methods for detecting lower grade gliomas (LGGs) with isocitrate dehydrogenase (IDH) mutations and chromosome 1p/19q non-codeletion (IDH mut-Noncodel) with high specificity. Although all institutions could use T2-FLAIR mismatch sign without any obstacles, this sign was not completely helpful because of its low sensitivity. In this study, we attempted to uncover the mechanism of T2-FLAIR mismatch sign for clarifying the cause of this sign's low sensitivity. Among 99 patients with LGGs, 22 were T2-FLAIR mismatch sign-positive (22%), and this sign as a marker of IDH mut-Noncodel showed a sensitivity of 55.6% and specificity of 96.8%. Via pathological analyses, we could provide evidence that not only microcystic changes but the enlarged intercellular space was associated with T2-FLAIR mismatch sign (p = 0.017). As per the molecular analyses, overexpression of mTOR-related genes (m-TOR, RICTOR) were detected as the molecular events correlated with T2-FLAIR mismatch sign (p = 0.020, 0.030. respectively). Taken together, we suggested that T2-FLAIR mismatch sign could pick up the IDH mut-Noncodel LGGs with enlarged intercellular space or that with overexpression of mTOR-related genes.

Novel approaches to the management of patients with 5-15 brain metastases: a narrative review.

The management of metastatic disease has been greatly influenced by molecular-based tumor classification and associated therapeutic targets, leading to a significant improvement in survival in many cases. This improvement, in both progression free survival and overall survival, has led to an increased incidence of brain metastases (BM) in a population with systemically well controlled disease or patients with promising therapeutic options available. Within this review, we discuss the paradigm of treatment for 5 to 15 BM, and how the treatment has evolved away from short-term palliation towards providing long term intracranial control. A review of literature pertaining to treatment of multiple BM was performed. We searched in PubMed to identify literature on treatment of multiple brain metastases. Only English literature published until February 1st, 2022 was reviewed. The management of 5-15 BM include multi-modality treatment pathways that are tailored towards each individual's primary cancer and burden of disease. Surgical resection of a dominant metastasis is still reserved for large symptomatic lesions, and is combined with post-operative local disease control. Overall, there is a shift away from whole brain radiation therapy (WBRT) due to side effect profile towards stereotactic radiosurgery (SRS). However, advances in WBRT continue to be studied, as well as the use of immunotherapy, targetable mutations, and synergistic effects between SRS and targeted therapies. The use of SRS to treat 5 to 15 BM is an increasingly acceptable and well-regarded practice, along with a combinatorial approach taking into account systemic options during all treatment timepoints.

Does clustering of DNA barcodes agree with botanical classification directly at high taxonomic levels? Trees in French Guiana as a case study.

Characterizing biodiversity is one of the main challenges for the coming decades. Most diversity has not been morphologically described, and barcoding is now complementing morphological-based taxonomy to further develop inventories. Both approaches have been cross-validated at the level of species and OTUs. However, many known species are not listed in reference databases. One path to speed up inventories using barcoding is to directly identify individuals at coarser taxonomic levels. We therefore studied in barcoding of plants whether morphological-based and molecular-based approaches are in agreement at genus, family and order levels. We used Agglomerative Hierarchical Clustering (with Ward, Complete and Single Linkage) and Stochastic Block Models (SBM), with two dissimilarity measures (Smith-Waterman scores, kmers). The agreement between morphological-based and molecular-based classifications ranges in most of the cases from good to very good at taxonomic levels above species, even though it decreases when taxonomic levels increase, or when using the tetramer-based distance. Agreement is correlated with the entropy of morphological-based classification and with the ratio of the mean within- and mean between-groups dissimilarities. The Ward method globally leads to the best agreement, whereas Single Linkage can show poor behaviours. SBM provides a useful tool to test whether or not the dissimilarities are structured by the botanical groups. These results suggest that automatic clustering and group identification at taxonomic levels above species are possible in barcoding.