Multi-scale relational graph convolutional network for multiple instance learning in histopathology images.

Graph convolutional neural networks have shown significant potential in natural and histopathology images. However, their use has only been studied in a single magnification or multi-magnification with either homogeneous graphs or only different node types. In order to leverage the multi-magnification information and improve message passing with graph convolutional networks, we handle different embedding spaces at each magnification by introducing the Multi-Scale Relational Graph Convolutional Network (MS-RGCN) as a multiple instance learning method. We model histopathology image patches and their relation with neighboring patches and patches at other scales (i.e., magnifications) as a graph. We define separate message-passing neural networks based on node and edge types to pass the information between different magnification embedding spaces. We experiment on prostate cancer histopathology images to predict the grade groups based on the extracted features from patches. We also compare our MS-RGCN with multiple state-of-the-art methods with evaluations on several source and held-out datasets. Our method outperforms the state-of-the-art on all of the datasets and image types consisting of tissue microarrays, whole-mount slide regions, and whole-slide images. Through an ablation study, we test and show the value of the pertinent design features of the MS-RGCN.

Prostate Cancer Risk Stratification by Digital Histopathology and Deep Learning.

PURPOSE: Prostate cancer (PCa) represents a highly heterogeneous disease that requires tools to assess oncologic risk and guide patient management and treatment planning. Current models are based on various clinical and pathologic parameters including Gleason grading, which suffers from a high interobserver variability. In this study, we determine whether objective machine learning (ML)-driven histopathology image analysis would aid us in better risk stratification of PCa. MATERIALS AND METHODS: We propose a deep learning, histopathology image-based risk stratification model that combines clinicopathologic data along with hematoxylin and eosin- and Ki-67-stained histopathology images. We train and test our model, using a five-fold cross-validation strategy, on a data set from 502 treatment-naïve PCa patients who underwent radical prostatectomy (RP) between 2000 and 2012. RESULTS: We used the concordance index as a measure to evaluate the performance of various risk stratification models. Our risk stratification model on the basis of convolutional neural networks demonstrated superior performance compared with Gleason grading and the Cancer of the Prostate Risk Assessment Post-Surgical risk stratification models. Using our model, 3.9% of the low-risk patients were correctly reclassified to be high-risk and 21.3% of the high-risk patients were correctly reclassified as low-risk. CONCLUSION: These findings highlight the importance of ML as an objective tool for histopathology image assessment and patient risk stratification. With further validation on large cohorts, the digital pathology risk classification we propose may be helpful in guiding administration of adjuvant therapy including radiotherapy after RP.

A novel type-2 innate lymphoid cell-based immunotherapy for cancer.

Cell-based cancer immunotherapy has achieved significant advancements, providing a source of hope for cancer patients. Notwithstanding the considerable progress in cell-based immunotherapy, the persistently low response rates and the exorbitant costs associated with their implementation still present a formidable challenge in clinical settings. In the landscape of cell-based cancer immunotherapies, an uncharted territory involves Type 2 innate lymphoid cells (ILC2s) and interleukin-33 (IL-33) which promotes ILC2 functionality, recognized for their inherent ability to enhance immune responses. Recent discoveries regarding their role in actuating cytolytic T lymphocyte responses, including curbing tumor growth rates and hindering metastasis, have added a new dimension to our understanding of the IL-33/ILC2 axis. These recent insights may hold significant promise for ILC2 cell-based immunotherapy. Nevertheless, the prospect of adoptively transferring ILC2s to confer immune protection against tumors has yet to be investigated. The present study addresses this hypothesis, revealing that ILC2s isolated from the lungs of tumor-bearing mice, and tumor infiltrating ILC2s when adoptively transferred after tumor establishment at a ratio of one ILC2 per sixty tumor cells, leads to an influx of tumor infiltrating CD4+ and CD8+ T lymphocytes as well as tumor infiltrating eosinophils resulting in a remarkable reduction in tumor growth. Moreover, we find that post-adoptive transfer of ILC2s, the number of tumor infiltrating ILC2s is inversely proportional to tumor size. Finally, we find corollaries of the IL-33/ILC2 axis enhancing the infiltration of eosinophils in human prostate carcinomas patients' expressing high levels of IL-33 versus those expressing low levels of IL-33. Our results underscore the heightened efficacy of adoptively transferred ILC2s compared to alternative approaches, revealing an approximately one hundred fifty-fold superiority on a cell-per-cell basis over CAR T-cells in the specific targeting and elimination of tumors within the same experimental model. Overall, this study demonstrates the functional significance of ILC2s in cancer immunosurveillance and provides the proof of concept of the potential utility of ILC2 cell-based cancer immunotherapies.

PDIA2 has a dual function in promoting androgen deprivation therapy induced venous thrombosis events and castrate resistant prostate cancer progression.

Androgen deprivation therapy (ADT) is the first line of treatment for metastatic prostate cancer (PCa) that effectively delays the tumor progression. However, it also increases the risk of venous thrombosis event (VTE) in patients, a leading cause of mortality. How a pro-thrombotic cascade is induced by ADT remains poorly understood. Here, we report that protein disulfide isomerase A2 (PDIA2) is upregulated in PCa cells to promote VTE formation and enhance PCa cells resistant to ADT. Using various in vitro and in vivo models, we demonstrated a dual function of PDIA2 that enhances tumor-mediated pro-coagulation activity via tumor-derived extracellular vehicles (EVs). It also stimulates PCa cell proliferation, colony formation, and xenograft growth androgen-independently. Mechanistically, PDIA2 activates the tissue factor (TF) on EVs through its isomerase activity, which subsequently triggers a pro-thrombotic cascade in the blood. Additionally, TF-containing EVs can activate the Src kinase inside PCa cells to enhance the AR signaling ligand independently. Androgen deprivation does not alter PDIA2 expression in PCa cells but enhances PDIA2 translocation to the cell membrane and EVs via suppressing the clathrin-dependent endocytic process. Co-recruitment of AR and FOXA1 to the PDIA2 promoter is required for PDIA2 transcription under androgen-deprived conditions. Importantly, blocking PDIA2 isomerase activity suppresses the pro-coagulation activity of patient plasma, PCa cell, and xenograft samples as well as castrate-resistant PCa xenograft growth. These results demonstrate that PDIA2 promotes VTE and tumor progression via activating TF from tumor-derived EVs. They rationalize pharmacological inhibition of PDIA2 to suppress ADT-induced VTE and castrate-resistant tumor progression.

DPYSL5 is highly expressed in treatment-induced neuroendocrine prostate cancer and promotes lineage plasticity via EZH2/PRC2.

Treatment-induced neuroendocrine prostate cancer (t-NEPC) is a lethal subtype of castration-resistant prostate cancer resistant to androgen receptor (AR) inhibitors. Our study unveils that AR suppresses the neuronal development protein dihydropyrimidinase-related protein 5 (DPYSL5), providing a mechanism for neuroendocrine transformation under androgen deprivation therapy. Our unique CRPC-NEPC cohort, comprising 135 patient tumor samples, including 55 t-NEPC patient samples, exhibits a high expression of DPYSL5 in t-NEPC patient tumors. DPYSL5 correlates with neuroendocrine-related markers and inversely with AR and PSA. DPYSL5 overexpression in prostate cancer cells induces a neuron-like phenotype, enhances invasion, proliferation, and upregulates stemness and neuroendocrine-related markers. Mechanistically, DPYSL5 promotes prostate cancer cell plasticity via EZH2-mediated PRC2 activation. Depletion of DPYSL5 decreases proliferation, induces G1 phase cell cycle arrest, reverses neuroendocrine phenotype, and upregulates luminal genes. In conclusion, DPYSL5 plays a critical role in regulating prostate cancer cell plasticity, and we propose the AR/DPYSL5/EZH2/PRC2 axis as a driver of t-NEPC progression.