The prediction of distant metastasis risk for male breast cancer patients based on an interpretable machine learning model.

OBJECTIVES: This research was designed to compare the ability of different machine learning (ML) models and nomogram to predict distant metastasis in male breast cancer (MBC) patients and to interpret the optimal ML model by SHapley Additive exPlanations (SHAP) framework. METHODS: Four powerful ML models were developed using data from male breast cancer (MBC) patients in the SEER database between 2010 and 2015 and MBC patients from our hospital between 2010 and 2020. The area under curve (AUC) and Brier score were used to assess the capacity of different models. The Delong test was applied to compare the performance of the models. Univariable and multivariable analysis were conducted using logistic regression. RESULTS: Of 2351 patients were analyzed; 168 (7.1%) had distant metastasis (M1); 117 (5.0%) had bone metastasis, and 71 (3.0%) had lung metastasis. The median age at diagnosis is 68.0 years old. Most patients did not receive radiotherapy (1723, 73.3%) or chemotherapy (1447, 61.5%). The XGB model was the best ML model for predicting M1 in MBC patients. It showed the largest AUC value in the tenfold cross validation (AUC:0.884; SD:0.02), training (AUC:0.907; 95% CI: 0.899-0.917), testing (AUC:0.827; 95% CI: 0.802-0.857) and external validation (AUC:0.754; 95% CI: 0.739-0.771) sets. It also showed powerful ability in the prediction of bone metastasis (AUC: 0.880, 95% CI: 0.856-0.903 in the training set; AUC: 0.823, 95% CI:0.790-0.848 in the test set; AUC: 0.747, 95% CI: 0.727-0.764 in the external validation set) and lung metastasis (AUC: 0.906, 95% CI: 0.877-0.928 in training set; AUC: 0.859, 95% CI: 0.816-0.891 in the test set; AUC: 0.756, 95% CI: 0.732-0.777 in the external validation set). The AUC value of the XGB model was larger than that of nomogram in the training (0.907 vs 0.802) and external validation (0.754 vs 0.706) sets. CONCLUSIONS: The XGB model is a better predictor of distant metastasis among MBC patients than other ML models and nomogram; furthermore, the XGB model is a powerful model for predicting bone and lung metastasis. Combining with SHAP values, it could help doctors intuitively understand the impact of each variable on outcome.

Correlation analysis between androgen receptor and the clinicopathological features and prognosis of mammary Paget's disease.

PURPOSE: Little is known about the prognostic value of androgen receptor (AR) status in mammary Paget's disease (MPD). The purpose of this study was to explore AR status and the distribution of molecular subtypes in MPD as well as the relationship between AR expression and clinicopathological factors and to evaluate its prognostic value. METHODS: We analyzed 170 MPD patients of varying subtypes. AR expression was verified by immunohistochemical staining, and the correlations between AR expression and clinicopathological characteristics and survival status were analyzed. We further investigated 91 MPD patients with invasive ductal carcinoma (MPD-IDC). RESULTS: AR was expressed in 55.3% of overall MPD patients, and 78.2% had the human epidermal growth factor receptor 2 (HER2) overexpression subtype. AR positivity was significantly correlated with BMI (P = 0.037) and pathological N stage (P = 0.023). Multivariate analysis indicated that pathological T stage and pathological N stage were independent prognostic factors for overall survival (OS). The positive AR group was significantly associated with better OS (P = 0.014). Among 91 MPD-IDC patients, AR was expressed in 56.0%, and 80.0% had the HER2 overexpression subtype. AR positivity was significantly correlated with pathological N stage (P = 0.033). Multivariate analysis indicated that AR and pathological T stage were independent prognostic factors for OS. Furthermore, AR positivity was significantly related to better OS (P = 0.005) in MPD-IDC patients as well as in patients with the HER2 overexpression subtype (P = 0.029). CONCLUSION: Our results confirmed that AR is a potential biomarker for evaluating the prognosis of patients.

circRNA-002178 act as a ceRNA to promote PDL1/PD1 expression in lung adenocarcinoma.

Circular RNAs (circRNAs) have been identified play a vital role in various different types of cancer via sponging miRNAs (microRNAs). However, their role in lung adenocarcinoma (LUAD) remains largely unclear. In this study, we systematically characterized the circRNA expression profiles in the LUAD cancer tissues and paired adjacent non-cancerous tissues. Three circRNAs were found to be significantly upregulated. Among them, has-circRNA-002178 was further confirmed to be upregulated in the LUAD tissues, and LUAD cancer cells. Subsequently, we also found has-circRNA-002178 could enhance PDL1 expression via sponging miR-34 in cancer cells to induce T-cell exhaustion. More importantly, circRNA-002178 could be detected in exosomes of plasma from LUAD patients and could serve as biomarkers for LUAD early diagnosis. Finally, we found circRNA-002178 could be delivered into CD8+ T cells to induce PD1 expression via exosomes. Taken together, our study revealed that circRNA-002178 could act as a ceRNA to promote PDL1/PD1 expression in lung adenocarcinoma.

Long non-coding RNA RP11-468E2.5 curtails colorectal cancer cell proliferation and stimulates apoptosis via the JAK/STAT signaling pathway by targeting STAT5 and STAT6.

BACKGROUND: Long non-coding RNAs (lncRNAs) are tumor-associated biological molecules and have been found to be implicated in the progression of colorectal cancer (CRC). This study aims to examine the effects of lncRNA RP11-468E2.5 and its target genes (STAT5 and STAT6) on the biological activities of CRC cells via the Janus kinase-signal transducer and activator of transcription (JAK/STAT) signaling pathway. METHODS: We initially screened the GEO database for differentially expressed lncRNAs related to CRC and then made a prediction of the implicated target genes. Then we collected CRC tissues and adjacent normal tissues from 169 CRC patients. Human CRC HCT116 and SW480 cells were treated with small interference RNA (siRNA) against RP11-468E2.5, AG490 (an inhibitor of the JAK/STAT signaling pathway), or both in combination. Next, we measured the effects of RP11-468E2.5 treatment on cellular activities such as cell viability, cycle distribution and cell apoptosis, and studied interactions among RP11-468E2.5, STAT5/STAT6, and the JAK/STAT signaling pathway. Finally, an in vivo tumor formation assay was performed to observe the effect of RP11-468E2.5 on tumor growth. RESULTS: The CRC-related gene microarray data showed low expression of RP11-468E2.5 in CRC surgical specimens. However, RP11-468E2.5 was confirmed to target STAT5 and STAT6, which participate in the JAK/STAT signaling pathway. CRC tissues showed lower expression of RP11-468E2.5, higher expression of STAT5, STAT6 and of the cell cycle marker Cyclin D1 (CCND1), compared to the findings in adjacent normal tissues. The treatment of siRNA against RP11-468E2.5 increased expression of JAK2, STAT3, STAT5, STAT6, CCND1 and Bcl-2 along with the extent of STAT3, STAT5 and STAT6 phosphorylation, while lowering expression of P21 and P27. Treatment with AG490 exhibited approximately opposite effects, whereas siRNA against RP11-468E2.5 treatment stimulated CRC cell proliferation and reduced cell apoptosis, while promoting cell cycle entry; AG490 treatment reversed these results. CONCLUSIONS: Altogether, we conclude that up-regulation of RP11-468E2.5 inhibits the JAK/STAT signaling pathway by targeting STAT5 and STAT6, thereby suppressing cell proliferation and promoting cell apoptosis in CRC.

Overexpression of TNFAIP8 is associated with tumor aggressiveness and poor prognosis in patients with invasive ductal breast carcinoma.

Tumor necrosis factor α-induced protein 8 (TNFAIP8), a transcription factor nuclear factor κB-inducible, antiapoptotic and oncogenic molecule, is associated with prognosis of several human malignancies. However, the relationship between TNFAIP8 and the prognosis of the invasive ductal carcinoma (IDC) of the breast remains unclear. TNFAIP8 expression was evaluated using real-time polymerase chain reaction (PCR) and Western blot analysis in 20 fresh IDC tissues and immunohistochemical analysis in 351 paraffin-embedded IDC tissues. Real-time PCR and Western blot analysis demonstrated that both TNFAIP8 messenger RNA and protein were up-regulated in IDC tissues compared with the paired adjacent noncancerous tissues. Immunohistochemistry revealed that TNFAIP8 expression was significantly correlated with some clinicopathological factors, including axillary lymph node metastasis (P=.001), advanced TNM stage (P<.001), high histologic grade (P<.001), molecular subtype (P<.001), and postoperative recurrence (P<.001). Univariate and multivariate logistic regression analyses demonstrated that TNFAIP8 overexpression was strongly associated with axillary lymph node metastasis (odds ratio, 1.818; 95% confidence interval, 1.167-2.832; P=.008). Moreover, Kaplan-Meier analysis indicated that IDC patients with high TNFAIP8 expression had a shorter survival time than did those with low TNFAIP8 expression, and multivariate analysis indicated that TNFAIP8 was an independent prognostic factor for overall survival and disease-free survival in IDC (P=.041 and P=.020, respectively). Therefore, TNFAIP8 overexpression may contribute to tumor progression, and it may be a novel prognostic biomarker for the patients with IDC.

Abnormal expression of p190RhoGAP in colorectal cancer patients with poor survival.

PURPOSE: This study aims to investigate the expression and clinical significance of p190RhoGAP, a member of the RhoGAP family, in colorectal cancer (CRC). METHODS: The expression p190RhoGAP was detected by RT-PCR, western blot (WB) and immunohistochemistry (IHC) in 14 paired CRCs and matched non-cancerous mucosal tissues. The protein content of p190RhoGAP was identified in 114 CRCs by IHC. In addition, the association of the expression of p190RhoGAP with carcinogenesis, distant metastasis and prognosis was further evaluated. RESULTS: In 14 paired fresh tissues, the mRNA (P<0.0001) and protein (P = 0.003) expression levels of p190RhoGAP were significantly higher in primary CRCs than in paired non-cancerous mucosal tissues; and was consistent with WB results. The expression of p190RhoGAP increased from normal mucosa to adenoma, and became even greater in primary carcinoma (P = 0.001). The expression level of p190RhoGAP was highest in liver metastasis compared to primary carcinoma (P = 0.028). The incidence of p190RhoGAP expression-positive cases was 58.77% in 114 CRC tissues. Furthermore, the enhanced expression of p190RhoGAP was significantly associated with shorter disease-specific survival (P<0.001) and shorter disease-free survival (P<0.001). Cox regression analysis indicated that p190RhoGAP was an independent prognostic parameter for CRC. CONCLUSION: p190RhoGAP may be an independent predictive factor for the prognosis of CRC, and the abnormal expression of p190RhoGAP may play a crucial role in colorectal carcinogenesis and distant metastasis.

Identification of miRNA-miRNA synergistic relationships in colorectal cancer.

PURPOSE: A large list of microRNAs (miRNAs) which may play important roles in the tumorigenesis of colorectal cancer (CRC) were generated recently. The purpose of our study is to analyze the synergistic regulations among miRNAs which were differentially expressed in tumorigenesis of CRC. METHODS: In this study, we downloaded the miRNA microarray and gene expression microarray of CRC from Gene Expression Omnibus (GEO), and identified the differentially expressed miRNAs (DE-miRNAs) and genes (DEGs) in cancer tissues compared with normal controls. In addition, we investigated the complex synergistic relationships among DE-miRNAs and constructed a miRNA-miRNA synergistic network by assembling all synergistic pairs. RESULTS: A total of 32 DE-miRNAs formed 8 functional modules in the synergistic network. These miRNAs in functional modules could independently implement specific functions as a whole in CRC, such as endocytosis, cell cycle, and p53 signaling pathway. CONCLUSIONS: The network allows for an in-depth analysis of individual miRNAs in the context of their synergistic surroundings.

[Study on retrograde metastasis rule of middle-low rectal cancer].

OBJECTIVE: To investigate the resection range of mesorectum and rectum below the inferior margin of tumor for the total mesorectum excision (TME) in middle-low rectal cancer. METHODS: Sixty patients were enrolled in the study. After TME operation, serial 5 mm interval sections were made in specimens of middle-low rectal cancer. The retrograde metastasis of rectal cancer was observed by routine HE staining. RESULTS: The phenomena of retrograde metastasis in mesorectum were found in 15 cases, and the distance of retrograde metastasis was 0.5-4.0(2.47+/-1.06) cm, which was correlated with Dukes stage, lymph node metastasis and histological differentiation. The retrograde metastases in bowel were found in 11 cases, and the distance of retrograde metastasis was 0.5-4.0 (1.64+/-1.16) cm, which was correlated with histological differentiation. CONCLUSIONS: The distal mesorectum should be resected at least 4 cm when TME is carried out, and the distal bowel at least 2.5 cm. More than 5 cm mesorectum and bowel should be resected when advanced Dukes stage, extensive lymph node metastasis and poor histological differentiation occurred.