PseudoCell: Hard Negative Mining as Pseudo Labeling for Deep Learning-Based Centroblast Cell Detection.

Background: Deep learning models for patch classification in whole-slide images (WSIs) have shown promise in assisting follicular lymphoma grading. However, these models often require pathologists to identify centroblasts and manually provide refined labels for model optimization. Objective: To address this limitation, we propose PseudoCell, an object detection framework for automated centroblast detection in WSI, eliminating the need for extensive pathologist's refined labels. Methods: PseudoCell leverages a combination of pathologist-provided centroblast labels and pseudo-negative labels generated from undersampled false-positive predictions based on cell morphology features. This approach reduces the reliance on time-consuming manual annotations. Results: Our framework significantly reduces the workload for pathologists by accurately identifying and narrowing down areas of interest containing centroblasts. Depending on the confidence threshold, PseudoCell can eliminate 58.18-99.35% of irrelevant tissue areas on WSI, streamlining the diagnostic process. Conclusion: This study presents PseudoCell as a practical and efficient prescreening method for centroblast detection, eliminating the need for refined labels from pathologists. The discussion section provides detailed guidance for implementing PseudoCell in clinical practice.

Public neoantigens in breast cancer immunotherapy (Review).

Among women globally, breast cancer is the most prevalent cancer and the leading cause of cancer­related death. Interestingly, though genetic mutations contribute to the disease, <15% of women diagnosed with breast cancer have a family history of the disease, suggesting a prevalence of sporadic genetic mutations in breast cancer development. In the rapidly rising field of cancer genomics, neoantigen­based immunotherapy has come to the fore. The investigation of novel proteins arising from unique somatic mutations or neoantigens have opened a new pathway for both individualized and public cancer treatments. Because they are shared among individuals with similar genetic changes, public neoantigens provide an opportunity for 'off­the­shelf' anticancer therapies, potentially extending the benefits to a wider patient group. The present review aimed to highlight the role of shared or public neoantigens as therapeutic targets for patients with breast cancer, emphasizing common hotspot mutations of certain genes identified in breast cancer. The clinical utilization of public neoantigen­based therapies for breast cancer treatment were also discussed.

The combined tumour-based Fascin/Snail and stromal periostin reveals the effective prognosis prediction in colorectal cancer patients.

Colorectal cancer (CRC) is the third most common malignancy cause of cancer-related mortality worldwide. Epithelial-mesenchymal transition (EMT) promotes cancer metastasis and a tumour-based Glasgow EMT score was associated with adverse clinical features and poor prognosis. In this study, the impact of using the established five tumour-based EMT markers consisting of E-cadherin (E-cad), ß-catenin (ß-cat), Snail, Zeb-1, and Fascin in combination with the stromal periostin (PN) on the prediction of CRC patients' prognosis were invesigated. Formalin-fixed paraffin-embedded tissues of 202 CRC patients were studies the expressions of E-cad, ß-cat, Snail, Zeb-1, Fascin, and PN by immunohistochemistry. Individually, cytoplasmic Fascin (Fc), cytoplasmic Snail (Sc), nuclear Snail (Sn), stromal Snail (Ss), and stromal PN (Ps) were significantly associated with reduced survival. A combination of Ps with Fc, Fs, and Sn was observed in 2 patterns including combined Fc, Fs, and Ps (FcFsPs) and Fc, Sn, and Ps (FcSnPs). These combinations enhanced the prognostic power compared to individual EMT markers and were independent prognostic markers. As the previously established scoring method required five markers and stringent criteria, its clinical use might be limited. Therefore, using these novel combined prognostic markers, either FcFsPs or FcSnPs, may be useful in predicting CRC patient outcomes.

Histopathological tumour microenvironment score independently predicts outcome in primary operable colorectal cancer.

Colorectal cancer (CRC) is a heterogenous malignancy and research is focused on identifying novel ways to subtype patients. In this study, a novel classification system, tumour microenvironment score (TMS), was devised based on Klintrup-Mäkinen grade (KMG), tumour stroma percentage (TSP), and tumour budding. TMS was performed using a haematoxylin and eosin (H&E)-stained section from retrospective CRC discovery and validation cohorts (n = 1,030, n = 787). TMS0 patients had high KMG, TMS1 were low for KMG, TSP, and budding, TMS2 were high for budding, or TSP and TMS3 were high for TSP and budding. Scores were assessed for association with survival and clinicopathological characteristics. Mutational landscaping and Templated Oligo-Sequencing (TempO-Seq) profiling were performed to establish differences in the underlying biology of TMS. TMS was independently prognostic in both cohorts (p < 0.001, p < 0.001), with TMS3 predictive of the shortest survival times. TMS3 was associated with adverse clinical features including sidedness, local and distant recurrence, higher T stage, higher N stage, and presence of margin involvement. Gene set enrichment analysis of TempO-Seq data showed higher expression of genes associated with hallmarks of cancer pathways including epithelial to mesenchymal transition (p < 0.001), IL2 STAT5 signalling (p = 0.007), and angiogenesis (p = 0.017) in TMS3. Additionally, enrichment of immunosuppressive immune signatures was associated with TMS3 classification. In conclusion, TMS represents a novel and clinically relevant method for subtyping CRC patients from a single H&E-stained tumour section.

BRCA1 deficiency enhances the aggressiveness of breast cancer cells expressing HPV16 oncoproteins.

BACKGROUND INFORMATION: The precise etiology of breast cancer is not completely understood, although women with BRCA1 gene mutations have a significantly increased risk of developing the disease. In addition, sporadic breast cancer is frequently associated with decreased BRCA1 gene expression. Growing evidence of Human papillomaviruses (HPVs) infections in breast tumors has raised the possibility of the involvement of HPVs in the pathogenesis of breast cancer. We investigated whether the effects of HPV oncoproteins E6 and E7 were influenced by the expression levels of BRCA1. HPV16E6E7 (prototype or E6D25E/E7N29S Asian variant type) were stably expressed in MDA-MB231 breast cancer cells, wild type for BRCA1, or with BRCA1 knocked down. RESULTS: Expression of HPV16E6E7 oncogenes did not affect BRCA1 levels and the abundance of HPV16E6E7 was not altered by BRCA1 knockdown. BRCA1 levels did not alter HPV16E6E7-dependent degradation of G1-S cell cycle proteins p53 and pRb. However, we found that the expression of G2-M cell cycle protein cyclin B1 enhanced by HPV16E6E7 was impacted by BRCA1 levels. Especially, we found the correlation between BRCA1 and cyclin B1 expression and this was also confirmed in breast cancer samples from a Thai cohort. We further demonstrated that the combination of HPV oncoproteins and low levels of BRCA1 protein appears to enhance proliferation and invasion. Transactivation activities of HPV16E6E7 on genes regulating cell proliferation and invasion (TGF-ß and vimentin) were significantly increased in BRCA1-deficient cells. CONCLUSIONS: Our results indicate that a deficiency of BRCA1 promotes the transactivation activity of HPV16E6E7 leading to increase of cell proliferation and invasion. SIGNIFICANCE: HPV infection appears to have the potential to enhance the aggressiveness of breast cancers, especially those deficient in BRCA1.

Stimulating T cell responses against patient-derived breast cancer cells with neoantigen peptide-loaded peripheral blood mononuclear cells.

Breast cancer stands as a formidable global health challenge for women. While neoantigens exhibit efficacy in activating T cells specific to cancer and instigating anti-tumor immune responses, the accuracy of neoantigen prediction remains suboptimal. In this study, we identified neoantigens from the patient-derived breast cancer cells, PC-B-142CA and PC-B-148CA cells, utilizing whole-genome and RNA sequencing. The pVAC-Seq pipeline was employed, with minor modification incorporating criteria (1) binding affinity of mutant (MT) peptide with HLA (IC50 MT) ≤ 500 nm in 3 of 5 algorithms and (2) IC50 wild type (WT)/MT > 1. Sequencing results unveiled 2513 and 3490 somatic mutations, and 646 and 652 non-synonymous mutations in PC-B-142CA and PC-B-148CA, respectively. We selected the top 3 neoantigens to perform molecular dynamic simulation and synthesized 9-12 amino acid neoantigen peptides, which were then pulsed onto healthy donor peripheral blood mononuclear cells (PBMCs). Results demonstrated that T cells activated by ADGRL1E274K, PARP1E619K, and SEC14L2R43Q peptides identified from PC-B-142CA exhibited significantly increased production of interferon-gamma (IFN-γ), while PARP1E619K and SEC14L2R43Q peptides induced the expression of CD107a on T cells. The % tumor cell lysis was notably enhanced by T cells activated with MT peptides across all three healthy donors. Moreover, ALKBH6V83M and GAAI823T peptides from PC-B-148CA remarkably stimulated IFN-γ- and CD107a-positive T cells, displaying high cell-killing activity against target cancer cells. In summary, our findings underscore the successful identification of neoantigens with anti-tumor T cell functions and highlight the potential of personalized neoantigens as a promising avenue for breast cancer treatment.

Targeting MYC at the intersection between cancer metabolism and oncoimmunology.

MYC activation is a known hallmark of cancer as it governs the gene targets involved in various facets of cancer progression. Of interest, MYC governs oncometabolism through the interactions with its partners and cofactors, as well as cancer immunity via its gene targets. Recent investigations have taken interest in characterizing these interactions through multi-Omic approaches, to better understand the vastness of the MYC network. Of the several gene targets of MYC involved in either oncometabolism or oncoimmunology, few of them overlap in function. Prominent interactions have been observed with MYC and HIF-1α, in promoting glucose and glutamine metabolism and activation of antigen presentation on regulatory T cells, and its subsequent metabolic reprogramming. This review explores existing knowledge of the role of MYC in oncometabolism and oncoimmunology. It also unravels how MYC governs transcription and influences cellular metabolism to facilitate the induction of pro- or anti-tumoral immunity. Moreover, considering the significant roles MYC holds in cancer development, the present study discusses effective direct or indirect therapeutic strategies to combat MYC-driven cancer progression.

Detection of centroblast cells in H&E stained whole slide image based on object detection.

Introduction: Detection and counting of Centroblast cells (CB) in hematoxylin & eosin (H&E) stained whole slide image (WSI) is an important workflow in grading Lymphoma. Each high power field (HPF) patch of a WSI is inspected for the number of CB cells and compared with the World Health Organization (WHO) guideline that organizes lymphoma into 3 grades. Spotting and counting CBs is time-consuming and labor intensive. Moreover, there is often disagreement between different readers, and even a single reader may not be able to perform consistently due to many factors. Method: We propose an artificial intelligence system that can scan patches from a WSI and detect CBs automatically. The AI system works on the principle of object detection, where the CB is the single class of object of interest. We trained the AI model on 1,669 example instances of CBs that originate from WSI of 5 different patients. The data was split 80%/20% for training and validation respectively. Result: The best performance was from YOLOv5x6 model that used the preprocessed CB dataset achieved precision of 0.808, recall of 0.776, mAP at 0.5 IoU of 0.800 and overall mAP of 0.647. Discussion: The results show that centroblast cells can be detected in WSI with relatively high precision and recall.