A deep learning method for predicting the origins of cervical lymph node metastatic cancer on digital pathological images.

The metastatic cancer of cervical lymph nodes presents complex shapes and poses significant challenges for doctors in determining its origin. We established a deep learning framework to predict the status of lymph nodes in patients with cervical lymphadenopathy (CLA) by hematoxylin and eosin (H&E) stained slides. This retrospective study utilized 1,036 cervical lymph node biopsy specimens at the First Affiliated Hospital of Sun Yat-Sen University (FAHSYSU). A multiple-instance learning algorithm designed for key region identification was applied, and cross-validation experiments were conducted in the dataset. Additionally, the model distinguished between primary lymphoma and metastatic cancer with high prediction accuracy. We also validated our model and other models on an external dataset. Our model showed better generalization and achieved the best results on both internal and external datasets. This algorithm offers an approach for evaluating cervical lymph node status before surgery, significantly aiding physicians in preoperative diagnosis and treatment planning.

Pathology diagnosis of intraoperative frozen thyroid lesions assisted by deep learning.

BACKGROUND: Thyroid cancer is a common thyroid malignancy. The majority of thyroid lesion needs intraoperative frozen pathology diagnosis, which provides important information for precision operation. As digital whole slide images (WSIs) develop, deep learning methods for histopathological classification of the thyroid gland (paraffin sections) have achieved outstanding results. Our current study is to clarify whether deep learning assists pathology diagnosis for intraoperative frozen thyroid lesions or not. METHODS: We propose an artificial intelligence-assisted diagnostic system for frozen thyroid lesions that applies prior knowledge in tandem with a dichotomous judgment of whether the lesion is cancerous or not and a quadratic judgment of the type of cancerous lesion to categorize the frozen thyroid lesions into five categories: papillary thyroid carcinoma, medullary thyroid carcinoma, anaplastic thyroid carcinoma, follicular thyroid tumor, and non-cancerous lesion. We obtained 4409 frozen digital pathology sections (WSI) of thyroid from the First Affiliated Hospital of Sun Yat-sen University (SYSUFH) to train and test the model, and the performance was validated by a six-fold cross validation, 101 papillary microcarcinoma sections of thyroid were used to validate the system's sensitivity, and 1388 WSIs of thyroid were used for the evaluation of the external dataset. The deep learning models were compared in terms of several metrics such as accuracy, F1 score, recall, precision and AUC (Area Under Curve). RESULTS: We developed the first deep learning-based frozen thyroid diagnostic classifier for histopathological WSI classification of papillary carcinoma, medullary carcinoma, follicular tumor, anaplastic carcinoma, and non-carcinoma lesion. On test slides, the system had an accuracy of 0.9459, a precision of 0.9475, and an AUC of 0.9955. In the papillary carcinoma test slides, the system was able to accurately predict even lesions as small as 2 mm in diameter. Tested with the acceleration component, the cut processing can be performed in 346.12 s and the visual inference prediction results can be obtained in 98.61 s, thus meeting the time requirements for intraoperative diagnosis. Our study employs a deep learning approach for high-precision classification of intraoperative frozen thyroid lesion distribution in the clinical setting, which has potential clinical implications for assisting pathologists and precision surgery of thyroid lesions.

Combining the tumor-stroma ratio with tumor-infiltrating lymphocytes improves the prediction of pathological complete response in breast cancer patients.

PURPOSE: The tumor-stroma ratio (TSR) is a common histological parameter that measures stromal abundance and is prognostic in breast cancer (BC). However, more evidence is needed on the predictive value of the TSR for the pathological complete response (pCR) to neoadjuvant chemotherapy (NAC). The purpose of this study was to determine the importance of the TSR in predicting pCR in NAC settings. METHOD: We evaluated the TSR on pretreatment biopsies of 912 BC patients from four independent Chinese hospitals and investigated the potential value of the TSR for predicting pCR. Meanwhile, stromal tumor-infiltrating lymphocytes (sTILs) were assessed, and we evaluated the predictive value of the combination of sTILs and TSR (TSRILs). RESULTS: Patients with low stroma showed a higher pCR rate than those with high stroma among the four independent hospitals, and in multivariate analysis, the TSR was proven to be an independent predictor for pCR to NAC with an odds ratio of 1.945 (95% CI 1.230-3.075, P = 0.004). Moreover, we found that TSRILs could improve the area under the curve (AUC) for predicting pCR from 0.750 to 0.785 (P = 0.039); especially in HER2-negative BCs, the inclusion of TSRILs increased the AUC from 0.801 to 0.835 in the discovery dataset (P = 0.048) and 0.734 to 0.801 in the validation dataset (P = 0.003). CONCLUSION: TSR and sTILs can be easily measured in pathological routines and provide predictive information without additional cost; with more evidence from clinical trials, TSRILs could be a candidate to better stratify patients in NAC settings.

Predicting neoadjuvant chemotherapy benefit using deep learning from stromal histology in breast cancer.

Neoadjuvant chemotherapy (NAC) is a standard treatment option for locally advanced breast cancer. However, not all patients benefit from NAC; some even obtain worse outcomes after therapy. Hence, predictors of treatment benefit are crucial for guiding clinical decision-making. Here, we investigated the predictive potential of breast cancer stromal histology via a deep learning (DL)-based approach and proposed the tumor-associated stroma score (TS-score) for predicting pathological complete response (pCR) to NAC with a multicenter dataset. The TS-score was demonstrated to be an independent predictor of pCR, and it not only outperformed the baseline variables and stromal tumor-infiltrating lymphocytes (sTILs) but also significantly improved the prediction performance of the baseline variable-based model. Furthermore, we discovered that unlike lymphocytes, collagen and fibroblasts in the stroma were likely associated with a poor response to NAC. The TS-score has the potential to better stratify breast cancer patients in NAC settings.

Tumor-Infiltrating Lymphocytes Improve Magee Equation-Based Prediction of Pathologic Complete Response in HR-Positive/HER2-Negative Breast Cancer.

OBJECTIVES: Magee equation 3 (ME3) is predictive of the pathologic complete response (pCR) to neoadjuvant chemotherapy (NAC) in patients with hormone receptor (HR)-positive, human epidermal growth factor receptor 2 (HER2)-negative breast cancer but with insufficient predictive performance. This study was designed to improve predictive ability by combining ME3 with additional clinicopathologic markers. METHODS: We retrospectively enrolled 460 patients with HR-positive/HER2-negative breast cancer from 2 centers. We obtained baseline characteristics, the ME3 score, and the number of stromal tumor-infiltrating lymphocytes (sTILs). After performing a logistic regression analysis, a predictive nomogram was built and validated externally. RESULTS: ME3 score (adjusted odds ratio [OR], 1.14 [95% confidence interval (CI), 1.10-1.17]; P < .001) and TILs (adjusted OR, 5.21 [95% CI, 3.33-8.14]; P < .001) were independently correlated with pCR. The nomogram (named ME3+) was established using ME3 and sTILs, and it demonstrated an area under the curve of 0.816 and 0.862 in internal and external validation, respectively, outperforming the ME3 score alone. sTILs and ME3 scores were also found to be positively correlated across the entire cohort (P < .001). CONCLUSIONS: The combination of sTILs and ME3 score potentially shows better performance for predicting pCR than ME3 alone. Larger validations are required for widespread application of ME3+ nomogram in NAC settings for HR-positive/HER2-negative breast cancer.

Clinical and genomic analyses of neuroendocrine neoplasms of the breast.

Breast neuroendocrine neoplasms (NENs) constitute a rare histologic subtype that includes both neuroendocrine tumors (NETs) and neuroendocrine carcinomas (NECs). In this study, we aimed to gain insight into the clinical and molecular characteristics of NENs of the breast. NEN and paired distant normal fresh tissues and clinicopathological data were obtained from 17 patients with NENs, and clinicopathological data were collected from 755 patients with invasive breast carcinomas of no special type (IBCs-NST). We compared the clinicopathological characteristics of NENs and IBCs-NST and performed whole-exome sequencing (WES) of both NEN and paired normal tissues. Compared with the IBC-NST patients, the NEN patients had a higher mean age, lower clinical stage, and lower pathological nodal (pN) stage (P < 0.001, P < 0.001, and P = 0.017, respectively). The most frequently mutated gene in NENs was KMT2C (3/17, 17.6%). NENs had copy number variations (CNVs) of 8q, 11q, and 17q amplification and 17q and 11q deletion and harbored the following specific genes related to tumorigenesis: (i) suppressor genes with loss of heterozygosity (LOH) such as ACE (2/17, 11.8%); (ii) tumor driver genes such as GATA3 (2/17, 11.8%); and (iii) susceptibility genes such as MAP3K4 (17/17, 100%) and PDE4DIP (17/17, 100%). The oncogenic/likely oncogenic mutations of NETs in PI3K pathway genes (50.0%, 18.2%; P < 0.001) and MAPK signaling pathway genes (83.3%, 18.2%; P = 0.035) affected higher proportions than those of NECs. In conclusion, this study provides certain clinical and molecular evidence supporting NENs as a distinct subtype of breast cancer and provides some potential molecular features for distinguishing NETs from NECs.

Deep learning-based predictive biomarker of pathological complete response to neoadjuvant chemotherapy from histological images in breast cancer.

BACKGROUND: Pathological complete response (pCR) is considered a surrogate endpoint for favorable survival in breast cancer patients treated with neoadjuvant chemotherapy (NAC). Predictive biomarkers of treatment response are crucial for guiding treatment decisions. With the hypothesis that histological information on tumor biopsy images could predict NAC response in breast cancer, we proposed a novel deep learning (DL)-based biomarker that predicts pCR from images of hematoxylin and eosin (H&E)-stained tissue and evaluated its predictive performance. METHODS: In total, 540 breast cancer patients receiving standard NAC were enrolled. Based on H&E-stained images, DL methods were employed to automatically identify tumor epithelium and predict pCR by scoring the identified tumor epithelium to produce a histopathological biomarker, the pCR-score. The predictive performance of the pCR-score was assessed and compared with that of conventional biomarkers including stromal tumor-infiltrating lymphocytes (sTILs) and subtype. RESULTS: The pCR-score derived from H&E staining achieved an area under the curve (AUC) of 0.847 in predicting pCR directly, and achieved accuracy, F1 score, and AUC of 0.853, 0.503, and 0.822 processed by the logistic regression method, respectively, higher than either sTILs or subtype; a prediction model of pCR constructed by integrating sTILs, subtype and pCR-score yielded a mean AUC of 0.890, outperforming the baseline sTIL-subtype model by 0.051 (0.839, P = 0.001). CONCLUSION: The DL-based pCR-score from histological images is predictive of pCR better than sTILs and subtype, and holds the great potentials for a more accurate stratification of patients for NAC.

Enzyme-free and DNA-based universal platform for the construction of various logic devices based on graphene oxide and G-quadruplex.

In the fields of biocomputing and biomolecular, DNA molecules are applicable to be regarded as data of logical computing platform that uses elaborate logic gates to perform a variety of tasks. Graphene oxide (GO) is a type of novel nanomaterial, which brings new research focus to materials science and biosensors due to its special selectivity and excellent quenching ability. G-quadruplex as a unique DNA structure stimulates the intelligent application of DNA assembly on the strength of its exceptional binding activity. In this paper, we report a universal logic device assisted with GO and G-quadruplex under an enzyme-free condition. Integrated with the quenching ability of GO to the TAMRA (fluorophore, Carboxytetramethylrhodamine) and the enhancement of fluorescence intensity produced by the peculiar binding of G-quadruplex to the NMM (N-methylmesoporphyrin IX), a series of basic binary logic gates (AND. OR. INHIBIT. XOR) have been designed and verified through biological experiments. Given the modularity and programmability of this strategy, two advanced logic gates (half adder and half subtractor) were realized on the basis of the same work platform. The fluorescence signals generated from different input combinations possessed satisfactory results, which provided proof of feasibility. We believe that the proposed universal logical platform that operates at the nanoscale is expected to be utilized for future applications in molecular computing as well as disease diagnosis.

Experimental study of land subsidence in response to groundwater withdrawal and recharge in Changping District of Beijing.

Over exploitation of groundwater in Changping District of Beijing city has caused serious land subsidence in the past decades. In recent years, the operation of the South-to-North Water Transfer Project has reduced the land subsidence rate. In this paper, Experimental tests are performed using the GDS Consolidation Testing System to characterize the compression and rebound of soils at depths of less than 100 m caused by groundwater withdrawal and recharge in Changping District. The results indicate that the compressible layers are the main contributors to land subsidence. The first compressible layer experiences greater deformation and more considerable hysteresis than the other compressible layers with the same decrease in the pore water pressure. Therefore, the exploitation of the adjacent aquifer should be controlled in the future. The deformation in the second and third compressible layers is a gradual and long-term process with little rebound; therefore, the subsidence should be seriously addressed when the groundwater in the two compressible layers is exploited on a large scale. In the same compressible layer, silty clay is more compressible and hysteretic than silt. For the same soil sample, the deformation rate decreases gradually as the pore water pressure decreases, whereas the creep deformation shows an overall increasing trend. A parameter named the subsidence index Cw is proposed in this paper to describe the soil compressibility during groundwater withdrawal. All the soil samples are characterized by elastic-plastic deformation, and the shallow soil samples with less pore water pressure decrease are more likely to rebound.

An Enzyme- and Label-Free Fluorescence Aptasensor for Detection of Thrombin Based on Graphene Oxide and G-Quadruplex.

An enzyme- and label-free aptamer-based assay is described for the determination of thrombin. A DNA strand (S) consisting of two parts was designed, where the first (Sa) is the thrombin-binding aptamer and the second (Se) is a G-quadruplex. In the absence of thrombin, Sa is readily adsorbed by graphene oxide (GO), which has a preference for ss-DNA rather than for ds-DNA. Upon the addition of the N-methyl-mesoporphyrin IX (NMM), its fluorescence (with excitation/emission at 399/610 nm) is quenched by GO. In contrast, in the presence of thrombin, the aptamer will bind thrombin, and thus, be separated from GO. As a result, fluorescence will be enhanced. The increase is linear in the 0.37 µM to 50 µM thrombin concentration range, and the detection limit is 0.37 nM. The method is highly selective over other proteins, cost-effective, and simple. In our perception, it represents a universal detection scheme that may be applied to other targets according to the proper choice of the aptamer sequence and formation of a suitable aptamer-target pair.

Clinicopathologic characteristics of HER2-positive pure mucinous breast carcinoma: a systematic investigation into an unusual tumor.

Pure mucinous breast carcinoma (PMBC) accounts for approximately 2% of all breast carcinoma. Overexpression or amplification of human epidermal growth factor receptor 2 (HER2) is rarely observed in PMBC. We retrieved 119 PMBCs, which included 12 HER2-positive PMBCs and 107 HER2-negative PMBCs, to compare the clinicopathologic features between HER2-positive and HER2-negative neoplasms. The assessed parameters included patient age, menstruation, laterality, tumor size, lymph node status, tumor-node-metastasis (TNM) stage, nuclear grade, receptor status, treatment and prognostic features. HER2-positive PMBCs represented approximately 10.1% of the PMBCs examined. HER2-positive PMBCs showed more frequent lymph node metastasis (P=0.038), a significantly higher clinical TNM stage (P<0.001) and nuclear grade (P<0.001), lower estrogen receptor (ER) and progesterone receptor (PR) expression and higher Ki67 expression than the HER2-negative group (P=0.011, P=0.005, and P=0.001, respectively). HER2-positive PMBCs (untreated with HER2-targeted therapy) had a significantly lower overall survival (OS) rate than HER2-negative PMBCs (P=0.005). Nodal metastasis, higher TNM stage and nuclear grade were identified as factors that result in poorer OS of patients with PMBCs (P<0.001, P=0.016, P<0.001, and P<0.001, respectively). Univariate and multivariate Cox analyses confirmed that HER2 status was an independent prognostic factor for PMBCs (P=0.003 and P=0.012, respectively). HER2-positive PMBC is a rare subtype of breast carcinoma with aggressive biological behavior. It is important to identify tumors with these aggressive clinical behaviors and manage them differently. To the best of our knowledge, this study represents the first systematic investigation of the clinicopathologic features of HER2-positive PMBCs.

Mammary epithelioid myofibroblastoma mimicking invasive carcinoma: case report and literature review.

Myofibroblastoma (MFB) of the breast is a rare benign neoplasm, which exhibits several morphologic variants and presents diagnostic dilemmas for pathologists. Here, we describe a case of a 42-year-old female patient diagnosed as epithelioid MFB. This painless tumor was well-circumscribed and found in the left breast for three months. Histologically, this tumor was predominantly composed of epithelioid cells, which arranged as single cells or small clusters, and formed a cellular nodule. Tumor stroma was collagenized, with scattered myxoid areas. This case was misinterpreted as invasive lobular carcinoma in the original diagnosis. Immunohistochemical profile demonstrated positivity for desmin, SMA, calponin, CD34 and hormone receptors, whereas pan-CK, CK7, CK8, CK34bE12, CK5/6, EMA, p63 and S-100 were negative, confirming the diagnosis of epithelioid MFB. Awareness of this unusual variant and careful integration of clinicopathologic findings would be critical to diagnosis this challenging lesion and avoid potential diagnostic pitfalls.