Differentially expressed genes associated with high metabolic tumor volume served as diagnostic markers and potential therapeutic targets for pancreatic cancer.

BACKGROUND: The lack of distinct biomarkers for pancreatic cancer is a major cause of early-stage detection difficulty. The pancreatic cancer patient group with high metabolic tumor volume (MTV), one of the values measured from positron emission tomography-a confirmatory method and standard care for pancreatic cancer, showed a poorer prognosis than those with low MTV. Therefore, MTV-associated differentially expressed genes (DEGs) may be candidates for distinctive markers for pancreatic cancer. This study aimed to evaluate the possibility of MTV-related DEGs as markers or therapeutic targets for pancreatic cancer. METHODS: Tumor tissues and their normal counterparts were obtained from patients undergoing preoperative 18F-FDG PET/CT. The tissues were classified into MTV-low and MTV-high groups (7 for each) based on the MTV2.5 value of 4.5 (MTV-low: MTV2.5 < 4.5, MTV-high: MTV2.5 ≥ 4.5). Gene expression fold change was first calculated in cancer tissue compared to its normal counter and then compared between low and high MTV groups to obtain significant DEGs. To assess the suitability of the DEGs for clinical application, the correlation of the DEGs with tumor grades and clinical outcomes was analyzed in TCGA-PAAD, a large dataset without MTV information. RESULTS: Total RNA-sequencing (MTV RNA-Seq) revealed that 44 genes were upregulated and 56 were downregulated in the high MTV group. We selected the 29 genes matching MTV RNA-seq patterns in the TCGA-PAAD dataset, a large clinical dataset without MTV information, as MTV-associated genes (MAGs). In the analysis with the TCGA dataset, MAGs were significantly associated with patient survival, treatment outcomes, TCGA-PAAD-suggested markers, and CEACAM family proteins. Some MAGs showed an inverse correlation with miRNAs and were confirmed to be differentially expressed between normal and cancerous pancreatic tissues. Overexpression of KIF11 and RCC1 and underexpression of ADCY1 and SDK1 were detected in ~ 60% of grade 2 pancreatic cancer patients and associated with ~ 60% mortality in stages I and II. CONCLUSIONS: MAGs may serve as diagnostic markers and miRNA therapeutic targets for pancreatic cancer. Among the MAGs, KIF11, RCC1, ADCY, and SDK1 may be early diagnostic markers.

Artificial-Intelligence-Assisted Detection of Metastatic Colorectal Cancer Cells in Ascitic Fluid.

Ascites cytology is a cost-effective test for metastatic colorectal cancer (CRC) in the abdominal cavity. However, metastatic carcinoma of the peritoneum is difficult to diagnose based on biopsy findings, and ascitic aspiration cytology has a low sensitivity and specificity and a high inter-observer variability. The aim of the present study was to apply artificial intelligence (AI) to classify benign and malignant cells in ascites cytology patch images of metastatic CRC using a deep convolutional neural network. Datasets were collected from The OPEN AI Dataset Project, a nationwide cytology dataset for AI research. The numbers of patch images used for training, validation, and testing were 56,560, 7068, and 6534, respectively. We evaluated 1041 patch images of benign and metastatic CRC in the ascitic fluid to compare the performance of pathologists and an AI algorithm, and to examine whether the diagnostic accuracy of pathologists improved with the assistance of AI. This AI method showed an accuracy, a sensitivity, and a specificity of 93.74%, 87.76%, and 99.75%, respectively, for the differential diagnosis of malignant and benign ascites. The diagnostic accuracy and sensitivity of the pathologist with the assistance of the proposed AI method increased from 86.8% to 90.5% and from 73.3% to 79.3%, respectively. The proposed deep learning method may assist pathologists with different levels of experience in diagnosing metastatic CRC cells of ascites.

CD24 induced cellular quiescence-like state and chemoresistance in ovarian cancer cells via miR-130a/301a-dependent CDK19 downregulation.

Cancer stem-like cell (CSC) is thought to be responsible for ovarian cancer recurrence. CD24 serves as a CSC marker for ovarian cancer and regulates the expression of miRNAs, which are regulators of CSC phenotypes. Therefore, CD24-regulated miRNAs may play roles in manifesting the CSC phenotypes in ovarian cancer cells. Our miRNA transcriptome analysis showed that 94 miRNAs were up or down-regulated in a CD24-high clone from an ovarian cancer patient compared to a CD24-low one. The CD24-dependent expression trend of the top 7 upregulated miRNAs (miR-199a-3p, 34c, 199a-5p, 130a, 301a, 214, 34b*) was confirmed in other 8 clones (4 clones for each group). CD24 overexpression upregulated the expression of miR-199a-3p, 34c, 199a-5p, 130a, 301a, 214, and 34b* in TOV112D (CD24-low) cells compared to the control, while CD24 knockdown downregulated the expression of miR-199a-3p, 199a-5p, 130a, 301a, and 34b* in OV90 (CD24-high) cells. miR-130a and 301a targeted CDK19, which induced a cellular quiescence-like state (increased G0/G1 phase cell population, decreased cell proliferation, decreased colony formation, and decreased RNA synthesis) and resistance to platinum-based chemotherapeutic agents. CD24 regulated the expression of miR-130a and 301a via STAT4 and YY1 phosphorylation mediated by Src and FAK. miR-130a and 301a were positively correlated in expression with CD24 in ovarian cancer patient tissues and negatively correlated with CDK19. Our results showed that CD24 expression may induce a cellular quiescence-like state and resistance to platinum-based chemotherapeutic agents in ovarian cancer via miR-130a and 301a upregulation. CD24-miR-130a/301a-CDK19 signaling axis could be a prognostic marker for or a potential therapeutic target against ovarian cancer recurrence.

Deciphering the Role of ERBB3 Isoforms in Renal Cell Carcinoma: A Comprehensive Genomic and Transcriptomic Analysis.

ERBB3, a key member of the receptor tyrosine kinase family, is implicated in the progression and development of various human cancers, affecting cellular proliferation and survival. This study investigated the expression of ERBB3 isoforms in renal clear cell carcinoma (RCC), utilizing data from 538 patients from The Cancer Genome Atlas (TCGA) Firehose Legacy dataset. Employing the SUPPA2 tool, the activity of 10 ERBB3 isoforms was examined, revealing distinct expression patterns in RCC. Isoforms uc001sjg.3 and uc001sjh.3 were found to have reduced activity in tumor tissues, while uc010sqb.2 and uc001sjl.3 demonstrated increased activity. These variations in isoform expression correlate with patient survival and tumor aggressiveness, indicating their complex role in RCC. The study, further, utilizes CIBERSORTx to analyze the association between ERBB3 isoforms and immune cell profiles in the tumor microenvironment. Concurrently, Gene Set Enrichment Analysis (GSEA) was applied, establishing a strong link between elevated levels of ERBB3 isoforms and critical oncogenic pathways, including DNA repair and androgen response. RT-PCR analysis targeting the exon 21-23 and exon 23 regions of ERBB3 confirmed its heightened expression in tumor tissues, underscoring the significance of alternative splicing and exon utilization in cancer development. These findings elucidate the diverse impacts of ERBB3 isoforms on RCC, suggesting their potential as diagnostic markers and therapeutic targets. This study emphasizes the need for further exploration into the specific roles of these isoforms, which could inform more personalized and effective treatment modalities for renal clear cell carcinoma.

Deep learning-based diagnosis of lung cancer using a nationwide respiratory cytology image set: improving accuracy and inter-observer variability.

Deep learning (DL)-based image analysis has recently seen widespread application in digital pathology. Recent studies utilizing DL in cytopathology have shown promising results, however, the development of DL models for respiratory specimens is limited. In this study, we designed a DL model to improve lung cancer diagnosis accuracy using cytological images from the respiratory tract. This retrospective, multicenter study used digital cytology images of respiratory specimens from a quality-controlled national dataset collected from over 200 institutions. The image processing involves generating extended z-stack images to reduce the phase difference of cell clusters, color normalizing, and cropping image patches to 256 × 256 pixels. The accuracy of diagnosing lung cancer in humans from image patches before and after receiving AI assistance was compared. 30,590 image patches (1,273 whole slide images [WSIs]) were divided into 27,362 (1,146 WSIs) for training, 2,928 (126 WSIs) for validation, and 1,272 (1,272 WSIs) for testing. The Densenet121 model, which showed the best performance among six convolutional neural network models, was used for analysis. The results of sensitivity, specificity, and accuracy were 95.9%, 98.2%, and 96.9% respectively, outperforming the average of three experienced pathologists. The accuracy of pathologists after receiving AI assistance improved from 82.9% to 95.9%, and the inter-rater agreement of Fleiss' Kappa value was improved from 0.553 to 0.908. In conclusion, this study demonstrated that a DL model was effective in diagnosing lung cancer in respiratory cytology. By increasing diagnostic accuracy and reducing inter-observer variability, AI has the potential to enhance the diagnostic capabilities of pathologists.

MET overexpression in ovarian cancer via CD24-induced downregulation of miR-181a: A signalling for cellular quiescence-like state and chemoresistance in ovarian CSCs.

Increased expression of CD24 and MET, markers for cancer stem-like cells (CSCs), are each associated with ovarian cancer severity. However, whether CD24 and MET are co-expressed in ovarian CSCs and, if so, how they are related to CSC phenotype manifestation remains unknown. Our immunohistochemistry analysis showed that the co-expression of CD24 and MET was associated with poorer patient survival in ovarian cancer than those without. In addition, analyses using KM plotter and ROC plotter presented that the overexpression of CD24 or MET in ovarian cancer patients was associated with resistance to platinum-based chemotherapy. In our miRNA transcriptome and putative target genes analyses, miR-181a was downregulated in CD24-high ovarian cancer cells compared to CD24-low and predicted to bind to CD24 and MET 3'UTRs. In OV90 and SK-OV-3 cells, CD24 downregulated miR-181a expression by Src-mediated YY1 activation, leading to increased expression of MET. And, CD24 or MET knockdown or miR-181a overexpression inhibited the manifestation of CSC phenotypes, cellular quiescence-like state and chemoresistance, in OV90 and SK-OV-3 cells: increased colony formation, decreased G0/G1 phase cell population and increased sensitivity to Cisplatin and Carboplatin. Our findings suggest that CD24-miR-181a-MET may consist of a signalling route for ovarian CSCs, therefore being a combinatory set of markers and therapeutic targets for ovarian CSCs.

Beyond the Microscope: A Technological Overture for Cervical Cancer Detection.

Cervical cancer is a common and preventable disease that poses a significant threat to women's health and well-being. It is the fourth most prevalent cancer among women worldwide, with approximately 604,000 new cases and 342,000 deaths in 2020, according to the World Health Organization. Early detection and diagnosis of cervical cancer are crucial for reducing mortality and morbidity rates. The Papanicolaou smear test is a widely used screening method that involves the examination of cervical cells under a microscope to identify any abnormalities. However, this method is time-consuming, labor-intensive, subjective, and prone to human errors. Artificial intelligence techniques have emerged as a promising alternative to improve the accuracy and efficiency of Papanicolaou smear diagnosis. Artificial intelligence techniques can automatically analyze Papanicolaou smear images and classify them into normal or abnormal categories, as well as detect the severity and type of lesions. This paper provides a comprehensive review of the recent advances in artificial intelligence diagnostics of the Papanicolaou smear, focusing on the methods, datasets, performance metrics, and challenges. The paper also discusses the potential applications and future directions of artificial intelligence diagnostics of the Papanicolaou smear.

Deep Learning-Based Computational Cytopathologic Diagnosis of Metastatic Breast Carcinoma in Pleural Fluid.

A Pleural effusion cytology is vital for treating metastatic breast cancer; however, concerns have arisen regarding the low accuracy and inter-observer variability in cytologic diagnosis. Although artificial intelligence-based image analysis has shown promise in cytopathology research, its application in diagnosing breast cancer in pleural fluid remains unexplored. To overcome these limitations, we evaluate the diagnostic accuracy of an artificial intelligence-based model using a large collection of cytopathological slides, to detect the malignant pleural effusion cytology associated with breast cancer. This study includes a total of 569 cytological slides of malignant pleural effusion of metastatic breast cancer from various institutions. We extracted 34,221 augmented image patches from whole-slide images and trained and validated a deep convolutional neural network model (DCNN) (Inception-ResNet-V2) with the images. Using this model, we classified 845 randomly selected patches, which were reviewed by three pathologists to compare their accuracy. The DCNN model outperforms the pathologists by demonstrating higher accuracy, sensitivity, and specificity compared to the pathologists (81.1% vs. 68.7%, 95.0% vs. 72.5%, and 98.6% vs. 88.9%, respectively). The pathologists reviewed the discordant cases of DCNN. After re-examination, the average accuracy, sensitivity, and specificity of the pathologists improved to 87.9, 80.2, and 95.7%, respectively. This study shows that DCNN can accurately diagnose malignant pleural effusion cytology in breast cancer and has the potential to support pathologists.

Densely Convolutional Spatial Attention Network for nuclei segmentation of histological images for computational pathology.

Introduction: Automatic nuclear segmentation in digital microscopic tissue images can aid pathologists to extract high-quality features for nuclear morphometrics and other analyses. However, image segmentation is a challenging task in medical image processing and analysis. This study aimed to develop a deep learning-based method for nuclei segmentation of histological images for computational pathology. Methods: The original U-Net model sometime has a caveat in exploring significant features. Herein, we present the Densely Convolutional Spatial Attention Network (DCSA-Net) model based on U-Net to perform the segmentation task. Furthermore, the developed model was tested on external multi-tissue dataset - MoNuSeg. To develop deep learning algorithms for well-segmenting nuclei, a large quantity of data are mandatory, which is expensive and less feasible. We collected hematoxylin and eosin-stained image data sets from two hospitals to train the model with a variety of nuclear appearances. Because of the limited number of annotated pathology images, we introduced a small publicly accessible data set of prostate cancer (PCa) with more than 16,000 labeled nuclei. Nevertheless, to construct our proposed model, we developed the DCSA module, an attention mechanism for capturing useful information from raw images. We also used several other artificial intelligence-based segmentation methods and tools to compare their results to our proposed technique. Results: To prioritize the performance of nuclei segmentation, we evaluated the model's outputs based on the Accuracy, Dice coefficient (DC), and Jaccard coefficient (JC) scores. The proposed technique outperformed the other methods and achieved superior nuclei segmentation with accuracy, DC, and JC of 96.4% (95% confidence interval [CI]: 96.2 - 96.6), 81.8 (95% CI: 80.8 - 83.0), and 69.3 (95% CI: 68.2 - 70.0), respectively, on the internal test data set. Conclusion: Our proposed method demonstrates superior performance in segmenting cell nuclei of histological images from internal and external datasets, and outperforms many standard segmentation algorithms used for comparative analysis.

Region Segmentation of Whole-Slide Images for Analyzing Histological Differentiation of Prostate Adenocarcinoma Using Ensemble EfficientNetB2 U-Net with Transfer Learning Mechanism.

Recent advances in computer-aided detection via deep learning (DL) now allow for prostate cancer to be detected automatically and recognized with extremely high accuracy, much like other medical diagnoses and prognoses. However, researchers are still limited by the Gleason scoring system. The histopathological analysis involved in assigning the appropriate score is a rigorous, time-consuming manual process that is constrained by the quality of the material and the pathologist's level of expertise. In this research, we implemented a DL model using transfer learning on a set of histopathological images to segment cancerous and noncancerous areas in whole-slide images (WSIs). In this approach, the proposed Ensemble U-net model was applied for the segmentation of stroma, cancerous, and benign areas. The WSI dataset of prostate cancer was collected from the Kaggle repository, which is publicly available online. A total of 1000 WSIs were used for region segmentation. From this, 8100 patch images were used for training, and 900 for testing. The proposed model demonstrated an average dice coefficient (DC), intersection over union (IoU), and Hausdorff distance of 0.891, 0.811, and 15.9, respectively, on the test set, with corresponding masks of patch images. The manipulation of the proposed segmentation model improves the ability of the pathologist to predict disease outcomes, thus enhancing treatment efficacy by isolating the cancerous regions in WSIs.

Characteristics of BRCA2 Mutated Prostate Cancer at Presentation.

Genetic alterations of DNA repair genes, particularly BRCA2 in patients with prostate cancer, are associated with aggressive behavior of the disease. It has reached consensus that somatic and germline tests are necessary when treating advanced prostate cancer patients. Yet, it is unclear whether the mutations are associated with any presenting clinical features. We assessed the incidences and characteristics of BRCA2 mutated cancers by targeted sequencing in 126 sets of advanced prostate cancer tissue sequencing data. At the time of diagnosis, cT3/4, N1 and M1 stages were 107 (85%), 54 (43%) and 35 (28%) samples, respectively. BRCA2 alterations of clinical significance by AMP/ASCO/CAP criteria were found in 19 of 126 samples (15.1%). The BRCA2 mutated cancer did not differ in the distributions of TNM stage, Gleason grade group or histological subtype compared to BRCA2 wild-type cancers. Yet, they had higher tumor mutation burden, and higher frequency of ATM and BRCA1 mutations (44% vs. 10%, p = 0.002 and 21% vs. 4%, p = 0.018, respectively). Of the metastatic subgroup (M1, n = 34), mean PSA was significantly lower in BRCA2 mutated cancers than wild-type (p = 0.018). In the non-metastatic subgroup (M0, n = 64), PSA was not significantly different (p = 0.425). A similar trend was noted in multiple metastatic prostate cancer public datasets. We conclude that BRCA2 mutated metastatic prostate cancers may present in an advanced stage with relatively low PSA.

Lysophosphatidic acid-induced amphiregulin secretion by cancer-associated fibroblasts augments cancer cell invasion.

Cancer-associated fibroblasts (CAFs) are key structural components of the tumor microenvironment and are closely associated with tumor invasion and metastasis. Lysophosphatidic acid (LPA) is a biolipid produced extracellularly and involved in tumorigenesis and metastasis. LPA has recently been implicated in the education and transdifferentiation of normal fibroblasts (NFs) into CAFs. However, little is known about the effects of LPA on CAFs and their participation in cancer cell invasion. In the present study, we identified a critical role of LPA-induced amphiregulin (AREG) secreted from CAFs in cancer invasiveness. CAFs secrete higher amounts of AREG than NFs, and LPA induces AREG expression in CAFs to augment their invasiveness. Strikingly, knocking out the AREG gene in CAFs attenuates cancer invasiveness and metastasis. Mechanistically, LPA induces Yes-associated protein (YAP) activation and Zinc finger E-box binding homeobox 1 (Zeb1) expression through the LPAR1 and LPAR3/Gi/Rho signaling axes, leading to AREG expression. Furthermore, we provide evidence that metformin, a biguanide derivative, significantly inhibits LPA-induced AREG expression in CAFs to attenuate cancer cell invasiveness. Collectively, the present data show that LPA induces AREG expression through YAP and Zeb1 in CAFs to promote cancer cell invasiveness, with the process being inhibited by metformin, providing potential biomarkers and therapeutic avenues to interdict cancer cell invasion.

p53 Immunohistochemistry and Mutation Types Mismatching in High-Grade Serous Ovarian Cancer.

High-grade serous carcinoma (HGSCa) of the ovary is featured by TP53 gene mutation. Missense or nonsense mutation types accompany most cases of HGSCa that correlate well with immunohistochemical (IHC) staining results-an all (missense) or none (nonsense) pattern. However, some IHCs produce subclonal or mosaic patterns from which TP53 mutation types, including the wild type of the gene, cannot be clearly deduced. We analyzed a total of 236 cases of ovarian HGSCa and tumors of other histology by matching the results of p53 IHC staining and targeted next-generation sequencing (TruSight Tumor 170 panel). Ambiguous IHCs that do not belong to the conventional "all or none" groups were reviewed to distinguish the true wild type (WT) from potentially pathogenic subclonal or mosaic patterns. There were about 9% of sequencing-IHC mismatching cases, which were enriched by the p53 c-terminal encoding nuclear localization signal and oligomerization domain, in which the subcellular locations of p53 protein were affected. Indeed, mutations in the oligomerization domain of the p53 protein frequently revealed an unmatched signal or cytosolic staining (L289Ffs*57 (Ins), and R342*). We conclude that both mutation types and IHC patterns of p53 are important sources of information to provide a precise diagnosis of HGSCa.

Differences in genomic profile of high-grade urothelial carcinoma according to tumor location.

OBJECTIVES: To establish targeted therapies based on the molecular landscape in upper urinary tract urothelial carcinoma (UTUC), we tried to investigate the molecular characteristics of UTUC compared with those of bladder urothelial carcinoma (BLUC) by next-generation sequencing (NGS). MATERIALS AND METHODS: We selected 71 high-grade infiltrating urothelial carcinoma tissue specimens from 33 UTUC and 38 BLUC patients. NGS analysis was performed with the Illumina TruShigt Oncology-500 panel. RESULTS: Both UTUC and BLUC showed similar clinicopathologic characteristics, as well as morphologic similarities. The median tumor mutation burden (TMB) of all cases was 7.8 mutations/Mb. The majority of alterations were missense mutations. TP53 (40/71, 56.3%), KDM6A (30/71, 42.3%), and TERT promoter mutations (23/71, 32.4%) were observed regardless of tumor location. Compared with UTUC, BLUC showed frequent mutations in several genes: ARID1A (P = 0.001), ASXL1 (P = 0.017), ERBB3 (P = 0.005), PRKDC (P = 0.004) and RB1 (P = 0.041). On the contrary, copy number loss of FGFR3 was observed more in UTUC than BLUC (P = 0.018). Also, 6 cases showed oncogenic fusions: 3 cases with FGFR2 fusion in UTUC and 3 cases with FGFR3-TACC3 fusion in BLUC. CONCLUSION: Despite the small cohort size, we identified genetic differences between UTUC and BLUC in Korean patients by NGS. An understanding of the comprehensive molecular characteristics of UTUC and BLUC may be helpful in detecting candidates for targeted therapy.

Androgen-Independent Prostate Cancer Is Sensitive to CDC42-PAK7 Kinase Inhibition.

Prostate cancer is a common form of cancer in men, and androgen-deprivation therapy (ADT) is often used as a first-line treatment. However, some patients develop resistance to ADT, and their disease is called castration-resistant prostate cancer (CRPC). Identifying potential therapeutic targets for this aggressive subtype of prostate cancer is crucial. In this study, we show that statins can selectively inhibit the growth of these CRPC tumors that have lost their androgen receptor (AR) and have overexpressed the RNA-binding protein QKI. We found that the repression of microRNA-200 by QKI overexpression promotes the rise of AR-low mesenchymal-like CRPC cells. Using in silico drug/gene perturbation combined screening, we discovered that QKI-overexpressing cancer cells are selectively vulnerable to CDC42-PAK7 inhibition by statins. We also confirmed that PAK7 overexpression is present in prostate cancer that coexists with hyperlipidemia. Our results demonstrate a previously unseen mechanism of action for statins in these QKI-expressing AR-lost CRPCs. This may explain the clinical benefits of the drug and support the development of a biology-driven drug-repurposing clinical trial. This is an important finding that could help improve treatment options for patients with this aggressive form of prostate cancer.

Metabolic tumour volume on 18F-FDG PET/CT predicts extended pathological T stages in patients with renal cell carcinoma at staging.

We evaluated the predictive value of 18F-fluorodeoxyglucose (FDG) uptake on positron emission tomography/CT (PET/CT) for extended pathological T (pT) stages (≥ pT3a) in Renal cell carcinoma (RCC) patients at staging. Thirty-eight RCC patients who underwent 18F-FDG PET/CT at staging, followed by radical nephrectomy between September 2016 and September 2018, were included in this prospective study. Patients were classified into two groups (limited pT stage: stage T1/2, n = 17; extended pT stage: T3/4, n = 21). Univariate and multivariate logistic regression analyses were performed to identify clinicopathological and metabolic variables to predict extended pT stages. 18F-FDG metabolic parameters were compared in relation to International Society of Urological Pathology (ISUP) grade and lymphovascular invasion (LVI). In univariate analysis, maximum standardised uptake value, metabolic tumour volume (MTV), and ISUP grade were significant. In multivariate analysis, MTV was the only significant factor of extended pT stages. With a cut-off MTV of 21.2, an area under the curve was 0.944, which was higher than 0.824 for clinical T stages (p = 0.037). In addition, high MTV, but not tumour size, was significantly correlated with aggressive pathologic features (ISUP grade and LVI). High glycolytic tumour volume on 18F-FDG PET/CT in RCC patients at staging is predictive of extended pT stages which could aid decision-making regarding the best type of surgery.

Revisiting extraprostatic extension based on invasion depth and number for new algorithm for substaging of pT3a prostate cancer.

Extraprostatic extension (EPE) is a factor in determining pT3a stage in prostate cancer. However, the only distinction in EPE is whether it is focal or non-focal, causing diagnostic and prognostic ambiguity. We substaged pT3a malignancies using classification of EPE to improve personalized prognostication. We evaluated 465 radical prostatectomy specimens with a digital image analyzer by measuring the number, radial distance and two-dimensional square area of the EPE. The most significant cut-off value was proposed as an algorithm for the pT3a substaging system to predict biochemical recurrence (BCR). A combination of the radial distance and the number of EPEs predicted BCR the most effectively. The optimal cut-off criteria were 0.75 mm and 2 mm in radial distance and multifocal EPE (hazard ratio: 2.526, C-index 0.656). The pT3a was subdivided into pT3a1, < 0.75 mm and any number of EPEs; pT3a2, 0.75-2 mm and one EPE; and pT3a3, > 2 mm and any number of EPEs or 0.75-2 mm and ≥ 2 EPEs. This combined tier was highly significant in the prediction of BCR-free survival. The combination of radial distance and number of EPEs could be used to subdivide pT3a prostate cancer and may aid in the prediction of BCR.

Immunohistochemical and genetic characteristics of HPV-associated endocervical carcinoma with an invasive stratified mucin-producing carcinoma (ISMC) component.

Invasive stratified mucin-producing carcinoma (ISMC) is a recently described entity of human papillomavirus (HPV)-associated endocervical adenocarcinoma with phenotypic plasticity and aggressive clinical behavior. To identify the cell of origin of ISMC, we investigated the immunohistochemical expression of cervical epithelial cell markers (CK7, PAX8, CK5/6, p63, and CK17), stemness markers (ALDH1 and Nanog), and epithelial-mesenchymal transition (EMT) markers (Snail, Twist, and E-cadherin) in 10 pure and mixed type ISMCs with at least 10% of ISMC component in the entire tumor, seven usual type endocervical adenocarcinomas (UEAs), and seven squamous cell carcinomas (SCCs). In addition, targeted sequencing was performed in 10 ISMCs. ISMC was significantly associated with larger tumor size (p = 0.011), more frequent lymphovascular invasion and lymph node metastasis (p < 0.001), higher FIGO stage (p = 0.022), and a tendency for worse clinical outcomes (p = 0.056) compared to other HPV-associated subtypes. ISMC showed negative or borderline positivity for PAX8, CK5/6, and p63, which were distinct from UEA and SCC (p < 0.01). Compared to UEA and SCC, ISMC showed higher expression for ALDH1 (p = 0.119 for UEA and p = 0.009 for SCC), Snail (p = 0.036), and Twist (p = 0.119), and tended to show decreased E-cadherin expression (p = 0.083). In next-generation sequencing analysis, ISMC exhibited frequent STK11, MET, FANCA, and PALB2 mutations compared to conventional cervical carcinomas, and genes related to EMT and stemness were frequently altered. EMT-prone and stemness characteristics and peripheral expression of reserve cell and EMT markers of ISMC suggest its cervical reserve cell origin. We recommend PAX8, CK5/6, and p63 as diagnostic triple biomarkers for ISMC. These findings highlight the distinct biological basis of ISMC.

Prostate epithelial genes define therapy-relevant prostate cancer molecular subtype.

BACKGROUND AND OBJECTIVES: Transcriptomic landscape of prostate cancer (PCa) shows multidimensional variability, potentially arising from the cell-of-origin, reflected in serum markers, and most importantly related to drug sensitivities. For example, Aggressive Variant Prostate Cancer (AVPC) presents low PSA per tumor burden, and characterized by de novo resistance to androgen receptor signaling inhibitors (ARIs). Understanding PCa transcriptomic complexity can provide biological insight and therapeutic guidance. However, unsupervised clustering analysis is hindered by potential confounding factors such as stromal contamination and stress-related material degradation. MATERIALS AND METHODS: To focus on prostate epithelial cell-relevant heterogeneity, we defined 1,629 genes expressed by prostate epithelial cells by analyzing publicly available bulk and single- cell RNA sequencing data. Consensus clustering and CIBERSORT deconvolution were used for class discovery and proportion estimate analysis. The Cancer Genome Atlas Prostate Adenocarcinoma dataset served as a training set. The resulting clusters were analyzed in association with clinical, pathologic, and genomic characteristics and impact on survival. Serum markers PSA and PAP was analyzed to predict response to docetaxel chemotherapy in metastatic setting. RESULTS: We identified two luminal subtypes and two aggressive variant subtypes of PCa: luminal A (Adipogenic/AR-active/PSA-high) (30.0%); luminal S (Secretory/PAP-high) (26.0%); AVPC-I (Immune-infiltrative) (14.7%), AVPC-M (Myc-active) (4.2%), and mixed (25.0%). AVPC-I and AVPC-M subtypes predicted to be resistant to ARI and have low PSA per tumor burden. Luminal A and AVPC-M predicted to be resistant to docetaxel and have high PSA/PAP Ratio. Metastatic PCa patients with high PSA/PAP ratio (>20) had significantly shorter progression-free survival than those with low ratio (≤20) following docetaxel chemotherapy. CONCLUSION: We propose four prostate adenocarcinoma subtypes with distinct transcriptomic, genomic, and pathologic characteristics. PSA/PAP ratio in advanced cancer may aid in determining which patients would benefit from maximized androgen receptor inhibition or early use of antimicrotubule agents.

Artificial Intelligence Techniques for Prostate Cancer Detection through Dual-Channel Tissue Feature Engineering.

The optimal diagnostic and treatment strategies for prostate cancer (PCa) are constantly changing. Given the importance of accurate diagnosis, texture analysis of stained prostate tissues is important for automatic PCa detection. We used artificial intelligence (AI) techniques to classify dual-channel tissue features extracted from Hematoxylin and Eosin (H&E) tissue images, respectively. Tissue feature engineering was performed to extract first-order statistic (FOS)-based textural features from each stained channel, and cancer classification between benign and malignant was carried out based on important features. Recursive feature elimination (RFE) and one-way analysis of variance (ANOVA) methods were used to identify significant features, which provided the best five features out of the extracted six features. The AI techniques used in this study for binary classification (benign vs. malignant and low-grade vs. high-grade) were support vector machine (SVM), logistic regression (LR), bagging tree, boosting tree, and dual-channel bidirectional long short-term memory (DC-BiLSTM) network. Further, a comparative analysis was carried out between the AI algorithms. Two different datasets were used for PCa classification. Out of these, the first dataset (private) was used for training and testing the AI models and the second dataset (public) was used only for testing to evaluate model performance. The automatic AI classification system performed well and showed satisfactory results according to the hypothesis of this study.