Breast Cancer Molecular Subtype Prediction on Pathological Images with Discriminative Patch Selection and Multi-Instance Learning.

Molecular subtypes of breast cancer are important references to personalized clinical treatment. For cost and labor savings, only one of the patient's paraffin blocks is usually selected for subsequent immunohistochemistry (IHC) to obtain molecular subtypes. Inevitable block sampling error is risky due to the tumor heterogeneity and could result in a delay in treatment. Molecular subtype prediction from conventional H&E pathological whole slide images (WSI) using the AI method is useful and critical to assist pathologists to pre-screen proper paraffin block for IHC. It is a challenging task since only WSI-level labels of molecular subtypes from IHC can be obtained without detailed local region information. Gigapixel WSIs are divided into a huge amount of patches to be computationally feasible for deep learning, while with coarse slide-level labels, patch-based methods may suffer from abundant noise patches, such as folds, overstained regions, or non-tumor tissues. A weakly supervised learning framework based on discriminative patch selection and multi-instance learning was proposed for breast cancer molecular subtype prediction from H&E WSIs. Firstly, co-teaching strategy using two networks was adopted to learn molecular subtype representations and filter out some noise patches. Then, a balanced sampling strategy was used to handle the imbalance in subtypes in the dataset. In addition, a noise patch filtering algorithm that used local outlier factor based on cluster centers was proposed to further select discriminative patches. Finally, a loss function integrating local patch with global slide constraint information was used to fine-tune MIL framework on obtained discriminative patches and further improve the prediction performance of molecular subtyping. The experimental results confirmed the effectiveness of the proposed AI method and our models outperformed even senior pathologists, which has the potential to assist pathologists to pre-screen paraffin blocks for IHC in clinic.

Overexpression of NDUFA4L2 is associated with poor prognosis in patients with colorectal cancer.

BACKGROUND: NDUFA4L2 (NADH dehydrogenase (ubiquinone) 1 alpha subcomplex, 4-like 2, also called NADH-ubiquinone oxidoreductase MLRQ subunit homologue) was clearly enriched in the mitochondrial fraction under hypoxic conditions, and immunofluorescence showed a clear colocalization of NDUFA4L2 and cytochrome c in some tumour cells. However, little study has investigated its prognostic value in colorectal cancer (CRC). METHODS: In our study, mRNA-NDUFA4L2 and protein expression were analysed in 150 cases of CRC and adjacent normal tissues using immunohistochemistry, semi-quantitative reverse transcriptase-polymerase chain reaction. The correlation between NDUFA4L2 expression and clinicopathological factors was evaluated by the Chi-square test. Overall survival of patients was analysed by the Kaplan-Meier method. RESULTS: NDUFA4L2 overexpression was observed in 84% (126/150) of CRC tissues, but only in 24.7% (37/150) of adjacent normal tissues (P < 0.05). Semi-quantitative reverse transcriptase-polymerase chain reaction showed average mRNA expression levels to be 23.34 ± 1.356 and 4.34 ± 1.132 for CRC tissue and adjacent normal tissue (P < 0.05). Statistical analysis showed a significant correlation of NDUFA4L2 expression with histological grade, Dukes' stages, lymph node metastasis and liver metastasis. More importantly, multivariate analysis indicated that overexpression of NDUFA4L2 was an independent prognostic factor for CRC patients (P = 0.002). NDUFA4L2-negative patients had a higher tumour-free/overall survival rate than patients with high NDUFA4L2 expression (P = 0.001 and 0.002, respectively). CONCLUSIONS: Our data suggest that NDUFA4L2 overexpression is associated with tumour progression and a poor prognosis in CRC patients.

Decreased SCF/c-kit signaling pathway contributes to loss of interstitial cells of Cajal in gallstone disease.

Cholecystolithiasis is a common disease, and gallbladder dysmotility is considered as a pivotal pathogenesis. Interstitial cells of Cajal (ICCs) serve as pacemakers and mediators of neuromuscular transmission for gastrointestinal motility. Reduction of ICCs has been reported in gallstone diseases. However, there are no reasonable mechanisms for the cholecystolithiasis-associated loss of ICCs in humans. Stem cell factor (SCF) and its ligand c-kit are essential for normal development and survival of ICCs. To date, little is known about the SCF/c-kit signaling pathway in gallstone diseases. The purpose of this study was to investigate the role of the SCF/c-kit signaling pathway in the loss of ICCs in cholecystolithiasis. Data from 18 patients with gallstones and 14 individuals without gallstones were compared. The gallbladder contractility was assessed by measuring the gallbladder ejection fraction (GEF) ultrasonographically. Tissues samples were obtained during surgery, changes of ICC quantities were analyzed by immunohistochemistry, and the mRNA and protein expression of SCF and c-kit were detected by Real-Time PCR and Western-blot analysis. Compared with the controls, the GEF was significantly reduced in the gallstone group, and decreased number of ICCs was present obviously in the gallstone group. Furthermore, the mRNA and protein expression of SCF and c-kit were significantly attenuated in the gallstone group. These data indicate that gallbladder motility may be affected by reduction of ICCs in gallstone disease. Additionally, the decreased of SCF/c-kit signaling pathway play an important role in the loss of ICCs.