Artificial intelligence enhances whole-slide interpretation of PD-L1 CPS in triple-negative breast cancer: A multi-institutional ring study.

BACKGROUND AND AIMS: Evaluation of the programmed cell death ligand-1 (PD-L1) combined positive score (CPS) is vital to predict the efficacy of the immunotherapy in triple-negative breast cancer (TNBC), but pathologists show substantial variability in the consistency and accuracy of the interpretation. It is of great importance to establish an objective and effective method which is highly repeatable. METHODS: We proposed a model in a deep learning-based framework, which at the patch level incorporated cell analysis and tissue region analysis, followed by the whole-slide level fusion of patch results. Three rounds of ring studies (RSs) were conducted. Twenty-one pathologists of different levels from four institutions evaluated the PD-L1 CPS in TNBC specimens as continuous scores by visual assessment and our artificial intelligence (AI)-assisted method. RESULTS: In the visual assessment, the interpretation results of PD-L1 (Dako 22C3) CPS by different levels of pathologists have significant differences and showed weak consistency. Using AI-assisted interpretation, there were no significant differences between all pathologists (P = 0.43), and the intraclass correlation coefficient (ICC) value was increased from 0.618 [95% confidence interval (CI) = 0.524-0.719] to 0.931 (95% CI = 0.902-0.955). The accuracy of interpretation result is further improved to 0.919 (95% CI = 0.886-0.947). Acceptance of AI results by junior pathologists was the highest among all levels, and 80% of the AI results were accepted overall. CONCLUSION: With the help of the AI-assisted diagnostic method, different levels of pathologists achieved excellent consistency and repeatability in the interpretation of PD-L1 (Dako 22C3) CPS. Our AI-assisted diagnostic approach was proved to strengthen the consistency and repeatability in clinical practice.

Gastric mechanosensitive channel Piezo1 regulates ghrelin production and food intake.

Ghrelin, produced mainly by gastric X/A-like cells, triggers a hunger signal to the central nervous system to stimulate appetite. It remains unclear whether X/A-like cells sense gastric distention and thus regulate ghrelin production. Here we show that PIEZO1 expression in X/A-like cells decreases in patients with obesity when compared to controls, whereas it increases after sleeve gastrectomy. Male and female mice with specific loss of Piezo1 in X/A-like cells exhibit hyperghrelinaemia and hyperphagia and are more susceptible to overweight. These phenotypes are associated with impairment of the gastric CaMKKII/CaMKIV-mTOR signalling pathway. Activation of PIEZO1 by Yoda1 or gastric bead implantation inhibits ghrelin production, decreases energy intake and induces weight loss in mice. Inhibition of ghrelin production by Piezo1 through the CaMKKII/CaMKIV-mTOR pathway can be recapitulated in a ghrelin-producing cell line mHypoE-42. Our study reveals a mechanical regulation of ghrelin production and appetite by PIEZO1 of X/A-like cells, which suggests a promising target for anti-obesity therapy.

Dynamic DNA Methylation Regulates Season-Dependent Secondary Metabolism in the New Shoots of Tea Plants.

Plant secondary metabolites are critical quality-conferring compositions of plant-derived beverages, medicines, and industrial materials. The accumulations of secondary metabolites are highly variable among seasons; however, the underlying regulatory mechanism remains unclear, especially in epigenetic regulation. Here, we used tea plants to explore an important epigenetic mark DNA methylation (5mC)-mediated regulation of plant secondary metabolism in different seasons. Multiple omics analyses were performed on spring and summer new shoots. The results showed that flavonoids and theanine metabolism dominated in the metabolic response to seasons in the new shoots. In summer new shoots, the genes encoding DNA methyltransferases and demethylases were up-regulated, and the global CG and CHG methylation reduced and CHH methylation increased. 5mC methylation in promoter and gene body regions influenced the seasonal response of gene expression; the amplitude of 5mC methylation was highly correlated with that of gene transcriptions. These differentially methylated genes included those encoding enzymes and transcription factors which play important roles in flavonoid and theanine metabolic pathways. The regulatory role of 5mC methylation was further verified by applying a DNA methylation inhibitor. These findings highlight that dynamic DNA methylation plays an important role in seasonal-dependent secondary metabolism and provide new insights for improving tea quality.

Haem Oxygenase 1 is a potential target for creating etiolated/albino tea plants (Camellia sinensis) with high theanine accumulation.

Theanine content is highly correlated with sensory quality and health benefits of tea infusion. The tender shoots of etiolated and albino tea plants contain higher theanine than the normal green tea plants and are valuable materials for high quality green tea processing. However, why these etiolated or albino tea plants can highly accumulate theanine is largely unknown. In this study, we observed an Arabidopsis etiolated mutant hy1-100 (mutation in Haem Oxygenase 1, HO1) that accumulated higher levels of glutamine (an analog of theanine). We therefore identified CsHO1 in tea plants and found CsHO1 is conserved in amino acid sequences and subcellular localization with its homologs in other plants. Importantly, CsHO1 expression in the new shoots was much lower in an etiolated tea plants 'Huangkui' and an albino tea plant 'Huangshan Baicha' than that in normal green tea plants. The expression levels of CsHO1 were negatively correlated with theanine contents in these green, etiolated and albino shoots. Moreover, CsHO1 expression levels in various organs and different time points were also negatively correlated with theanine accumulation. The hy1-100 was hypersensitive to high levels of theanine and accumulated more theanine under theanine feeding, and these phenotypes were rescued by the expression of CsHO1 in this mutant. Transient knockdown CsHO1 expression in the new shoots of tea plant using antisense oligonucleotides (asODN) increased theanine accumulation. Collectively, these results demonstrated CsHO1 negatively regulates theanine accumulation in tea plants, and that low expression CsHO1 likely contributes to the theanine accumulation in etiolated/albino tea plants.

Multi-center study on predicting breast cancer lymph node status from core needle biopsy specimens using multi-modal and multi-instance deep learning.

The objective of our study is to develop a deep learning model based on clinicopathological data and digital pathological image of core needle biopsy specimens for predicting breast cancer lymph node metastasis. We collected 3701 patients from the Fourth Hospital of Hebei Medical University and 190 patients from four medical centers in Hebei Province. Integrating clinicopathological data and image features build multi-modal and multi-instance (MMMI) deep learning model to obtain the final prediction. For predicting with or without lymph node metastasis, the AUC was 0.770, 0.709, 0.809 based on the clinicopathological features, WSI and MMMI, respectively. For predicting four classification of lymph node status (no metastasis, isolated tumor cells (ITCs), micrometastasis, and macrometastasis), the prediction based on clinicopathological features, WSI and MMMI were compared. The AUC for no metastasis was 0.770, 0.709, 0.809, respectively; ITCs were 0.619, 0.531, 0.634, respectively; micrometastasis were 0.636, 0.617, 0.691, respectively; and macrometastasis were 0.748, 0.691, 0.758, respectively. The MMMI model achieved the highest prediction accuracy. For prediction of different molecular types of breast cancer, MMMI demonstrated a better prediction accuracy for any type of lymph node status, especially in the molecular type of triple negative breast cancer (TNBC). In the external validation sets, MMMI also showed better prediction accuracy in the four classification, with AUC of 0.725, 0.757, 0.525, and 0.708, respectively. Finally, we developed a breast cancer lymph node metastasis prediction model based on a MMMI model. Through all cases tests, the results showed that the overall prediction ability was high.

High Dynamic Range Dual-Modal White Light Imaging Improves the Accuracy of Tumor Bed Sampling After Neoadjuvant Therapy for Breast Cancer.

OBJECTIVES: Accurate evaluation of residual cancer burden remains challenging because of the lack of appropriate techniques for tumor bed sampling. This study evaluated the application of a white light imaging system to help pathologists differentiate the components and location of tumor bed in specimens. METHODS: The high dynamic range dual-mode white light imaging (HDR-DWI) system was developed to capture antiglare reflection and multiexposure HDR transmission images. It was tested in 60 specimens of modified radical mastectomy after neoadjuvant therapy. We observed the differential transmittance among tumor tissue, fibrosis tissue, and adipose tissue. RESULTS: The sensitivity and specificity of HDR-DWI were compared with x-ray or visual examination to determine whether HDR-DWI was superior in identifying tumor beds. We found that tumor tissue had lower transmittance (0.12 ± 0.03) than fibers (0.15 ± 0.04) and fats (0.27 ± 0.07) (P < .01). CONCLUSIONS: HDR-DWI was more sensitive in identifying fiber and tumor tissues than cabinet x-ray and visual observation (P < .01). In addition, HDR-DWI could identify more fibrosis areas than the currently used whole slide imaging did in 12 samples (12/60). We have determined that HDR-DWI can provide more in-depth tumor bed information than x-ray and visual examination do, which will help prevent diagnostic errors in tumor bed sampling.

Comparison of the tumor immune microenvironment phenotypes in different breast cancers after neoadjuvant therapy.

Neoadjuvant therapy (NAT) treats early-stage breast cancers, especially triple-negative breast cancers (TNBCs). NAT improves pathological complete response (pCR) rates for different breast cancer patients. Recently, immune checkpoint inhibitors that target programmed death 1 (PD-1) or programmed death ligand 1 (PD-L1) in combination with NAT have shown antitumor activity in patients with early breast cancer. However, the tumor immune microenvironment (TME) in different subtypes of breast cancers, like TNBC, hormone receptor-positive (HR+), and human epidermal growth factor receptor 2 amplified (HER2+) and its changes by NAT remain to be fully characterized. We analyzed pre-NAT tumor biopsies from TNBC (n = 27), HR+ (n = 24), and HER2+ (n = 30) breast cancer patients who received NAT, followed by surgery. The different immune makers (PD-1, PD-L1, CD3, and CD8) of tumor-infiltrating lymphocytes (TILs) were identified with immunofluorescence-based microenvironment analysis. TILs within cancer parenchyma (iTILs) and in cancer stroma (sTILs) were counted separately. We found that PD-L1+ cells in tumor and stroma were significantly higher in TNBC patients than in others. PD-L1+ sTILs were significantly higher in pCR than in non-pCR patients of all the subtypes. The infiltration scores of B-cell memory, T-cell CD4+ memory activated, T-cell follicular helper, and Macrophage M0 and M1 were relatively higher in TNBC patients, indicating immunoreactive TME in TNBC. Analysis of TCGA-BRCA RNA-seq indicated that PD-L1 was highly expressed in TNBC patients compared with HR+ and HER2+ patients. Higher PD-L1 expression in TNBC patients was associated with significantly longer overall survival (OS). Our results demonstrated that PD-L1 expression level of iTILs and sTILs is highest in TNBC among breast cancers. TNBC patients had significantly different immunoreactive TME compared with HR+ and HER2+ patients, suggesting potentially favorable outcomes for immunotherapy in these patients. Also, PD-L1+ could be a powerful predictor of pCR in TNBC patients after NAT.

Reconstructing virtual large slides can improve the accuracy and consistency of tumor bed evaluation for breast cancer after neoadjuvant therapy.

BACKGROUND: To explore whether the "WSI Stitcher", a program we developed for reconstructing virtual large slide through whole slide imaging fragments stitching, can improve the efficiency and consistency of pathologists in evaluating the tumor bed after neoadjuvant treatment of breast cancer compared with the conventional methods through stack splicing of physical slides. METHODS: This study analyzed the advantages of using software-assisted methods to evaluate the tumor bed after neoadjuvant treatment of breast cancer. This new method is to use "WSI Stitcher" to stitch all the WSI fragments together to reconstruct a virtual large slide and evaluate the tumor bed with the help of the built-in ruler and tumor proportion calculation functions. RESULTS: Compared with the conventional method, the evaluation time of the software-assisted method was shortened by 35%(P < 0.001). In the process of tumor bed assessment after neoadjuvant treatment of breast cancer, the software-assisted method has higher intraclass correlation coefficient when measuring the length (0.994 versus 0.934), width (0.992 versus 0.927), percentage of residual tumor cells (0.947 versus 0.878), percentage of carcinoma in situ (0.983 versus 0.881) and RCB index(0.997 versus 0.772). The software-assisted method has higher kappa values when evaluating tumor staging(0.901 versus 0.687) and RCB grading (0.963 versus 0.857). CONCLUSION: The "WSI Stitcher" is an effective tool to help pathologists with the assessment of breast cancer after neoadjuvant treatment.

Development and Validation of a Novel Model for Predicting Prognosis of Non-PCR Patients After Neoadjuvant Therapy for Breast Cancer.

PURPOSE: Pathologic complete response (pCR) after neoadjuvant therapy is an important indicator of long-term prognosis and the primary endpoint of many neoadjuvant studies. For breast cancer patients who do not achieve pCR, prognostic indicators related to prognosis are particularly important. This study is constructing a prediction model with more accurate and reliable prediction results by combining multiple clinicopathological factors, so as to provide a more accurate decision-making basis for subsequent clinical treatment. PATIENTS AND METHODS: In this study, 1,009 cases of invasive breast cancer and surgically resected after neoadjuvant therapy from 2010 to 2017. All indicators in this trial were interpreted in a double-blind manner by two pathologists with at least 10 years of experience, including histological grading, Tils, ER, PR, HER2, and Ki67. The prediction model used R language to calculate the calibration degree and ROC curve of the prediction model in the training set and validation set. RESULTS: Through univariate survival analysis, the results showed histological grade (P=0.037), clinical stage (P<0.001), HER2 (P=0.044), RCB class (P<0.001), Tils (P<0.001), lymph node status (P =0.049), MP grade (P=0.013) are related to OS in non-PCR patients after neoadjuvant. Data were analyzed by substituting in a multivariate analysis, and the results were that clinical stage, HER2, RCB grading, and Tils grading were correlated with OS in non-PCR patients after neoadjuvant therapy for breast cancer. Among all cases in the training set, the prediction model predicted that the 3-year survival AUC value was 0.95 and 5-year survival AUC value was 0.79, and the RCB classification of 3-year survival and 5-year survival were 0.70 and 0.67, respectively, which proved that the prediction model could predict the OS of non-PCR patients after neoadjuvant therapy for breast cancer more accurately than the RCB classification, and showed the same results in HR, HER2+, and TN classifications. It also showed the same results in validation set. CONCLUSION: These data indicate that the predicted values of the prediction model developed in this study match the actual survival rates without underestimating the mortality risk and have a relatively accurate prediction effect.

[AI-assisted Prediction of Lymph Node Metastasis of Breast Cancer: Current and Prospective Research].

One of the most important application of artificial intelligence (AI) in pathology is prediction, using morphological features, of patient prognosis and response to specific treatments. As one of the most common kinds of malignancies in the world and the crucial important cause of death due to malignant tumor among women, breast cancer has become the center of attention in clinical services. Axillary lymph node metastasis is an important prognostic factor in breast cancer. The accuracy of the assessment of axillary lymph node metastasis bears heavily on clinical diagnosis and treatment. At present, based on the principle of non-invasive procedures, many studies have been done to develop models that can be used to predict sentinel lymph node metastasis of breast cancer. However, different clinical and pathological parameters are used in these predictive models. How to analyze the clinical and pathological data of breast cancer patients in a more comprehensive way and how to establish a prediction model with better precision have become the future direction of development. In this paper, we describe the research progress of AI in pathology and the current status of its use in breast cancer research. We have conducted in-depth reflection and looked into the future of ways to predict effectively breast cancer lymph node metastasis and to establish more accurate and effective deep-learning algorithm based on AI assistance so as to continuously improve the diagnosis and treatment of breast cancer.

Insertion of a solo LTR retrotransposon associates with spur mutations in 'Red Delicious' apple (Malus × domestica).

KEY MESSAGE: Insertion of a solo LTR, which possesses strong bidirectional, stem-specific promoter activities, is associated with the evolution of a dwarfing apple spur mutation. Spur mutations in apple scions revolutionized global apple production. Since long terminal repeat (LTR) retrotransposons are tightly related to natural mutations, inter-retrotransposon-amplified polymorphism technique and genome walking were used to find sequences in the apple genome based on these LTRs. In 'Red Delicious' spur mutants, a novel, 2190-bp insertion was identified as a spur-specific, solo LTR (sLTR) located at the 1038th nucleotide of another sLTR, which was 1536 bp in length. This insertion-within-an-insertion was localized within a preexisting Gypsy-50 retrotransposon at position 3,762,767 on chromosome 4. The analysis of transcriptional activity of the two sLTRs (the 2190- and 1536-bp inserts) indicated that the 2190-bp sLTR is a promoter, capable of bidirectional transcription. GUS expression in the 2190-bp-sense and 2190-bp-antisense transgenic lines was prominent in stems. In contrast, no promoter activity from either the sense or the antisense strand of the 1536-bp sLTR was detected. From ~150 kb of DNA on each side of the 2190 bp, sLTR insertion site, corresponding to 300 kb of the 'Golden Delicious' genome, 23 genes were predicted. Ten genes had predicted functions that could affect shoot development. This first report, of a sLTR insertion associated with the evolution of apple spur mutation, will facilitate apple breeding, cloning of spur-related genes, and discovery of mechanisms behind dwarf habit.