Development and validation of a prognostic model for HER2-low breast cancer to evaluate neoadjuvant therapy.

Background: Human epidermal growth factor receptor 2 (HER2) low breast cancer (BC) accounts for 30-51% of all BCs. How to precisely assess the response to neoadjuvant therapy in this heterogenous tumor is currently unanswered. With the advance in multi-omics, refining the molecular subtyping other than the current hormone receptor (HR)-based subtyping to guide the neoadjuvant therapy for HER2-low BC is potentially feasible. Methods: The messenger RNA (mRNA), clinical, and pathological data of all HER2-low BC patients (n=368) from the Neoadjuvant I-SPY2 Trial, were retrieved. Ninety-eight patients achieved pathological complete response (pCR) were randomly divided into the training and validation sets with 8:2 ratio. The non-pCR cases were corporated into the above datasets with 1:1 ratio. The rest non-pCR cases were served as the test set. Random forest (RF), support vector machine (SVM), and fully connected neural network (FCNN) were applied to establish a 1-dimensional (1D) model based on mRNA data. The method with best prediction value among the 3 models was selected for further modeling when combining pathological features. A new classification of deep learning (CDn) was proposed based on a multi-omics model. After identifying pCR-related features by the integral gradient and unsupervised hierarchical clustering method, the responses to neoadjuvant therapy associated with these features across different subgroups were analyzed. Results: Compared with the RF and SVM models, the FCNN model achieved the best performance [area under the curve (AUC): 0.89] based on the mRNA feature. By combining mRNA and pathological features, the FCNN model proposed 2 new subtypes including CD1 and CD0 for HER2-low BC. CD1 increased the sensitivity to predict pCR by 23.5% [to 87.8%; 95% confidence interval (CI): 78% to 94%] and improved the specificity to pCR by 12.2% (to 77.4%; 95% CI: 69% to 87%) when comparing with the current HR classification for HER2-low BC. Conclusions: The new typing method (CD1 and CD0) proposed in this study achieved excellent performance for predicting the pCR to neoadjuvant therapy in HER2-low BC. The patients who were not sensitive to neoadjuvant therapy according to multi-omics models might receive surgical treatment directly.

A multi-omics signature to predict the prognosis of invasive ductal carcinoma of the breast.

BACKGROUND: Precisely evaluating the prognosis of invasive ductal carcinoma (IDC) of the breast is challenging as most prognostic signatures use single-omics data based on gene or clinical information. METHODS: Whole-slide images (WSIs), transcriptome, and clinical data of breast IDC were collected from the Cancer Genome Atlas Database. The cancer-associated fibroblast (CAF) gene sets were downloaded from the Molecular Signatures Database. The WSI feature was extracted by artificial feature engineering. The CAF prognostic genes were determined by the Gene Set Enrichment Analysis, the Wilcoxon test, and univariate Cox regression. The IDC patients were divided into the training and test sets. The prognostic signatures based on WSIs, IDC-CAFs, bi-omics, and tri-omics were constructed using multivariate Cox regression. The samples were divided into low- and high-risk groups according to the median risk score. The Kaplan-Meier survival and receiver operating characteristic curves were applied to validate the prediction performance of the four signatures. RESULTS: In total, 508 IDC patients with complete data were included. The area under the curve (AUC) of single-omics signature based on WSI characteristics and CAFs was 0.765 and 0.775, whereas the AUC of bi-omics was 0.823. The tri-omics signature based on WSIs, CAFs, and lymph node status demonstrated the best predictive value with an AUC of 0.897. CONCLUSION: The multi-omics signature based on WSIs, CAFs, and clinical characteristics showed excellent prediction ability in breast IDC patients, whose risk factors can also provide a valuable diagnostic reference for the clinical course.

A cholesterogenic gene signature for predicting the prognosis of young breast cancer patients.

Purpose: We aimed to establish a cholesterogenic gene signature to predict the prognosis of young breast cancer (BC) patients and then verified it using cell line experiments. Methods: In the bioinformatic section, transcriptional data and corresponding clinical data of young BC patients (age ≤ 45 years) were downloaded from The Cancer Genome Atlas (TCGA) database for training set. Differentially expressed genes (DEGs) were compared between tumour tissue (n = 183) and normal tissue (n = 30). By using univariate Cox regression and multi COX regression, a five-cholesterogenic-gene signature was established to predict prognosis. Subgroup analysis and external validations of GSE131769 from the Gene Expression Omnibus (GEO) were performed to verify the signature. Subsequently, in experiment part, cell experiments were performed to further verify the biological roles of the five cholesterogenic genes in BC. Results: In the bioinformatic section, a total of 97 upregulated genes and 124 downregulated cholesterogenic genes were screened as DEGs in the TCGA for training the model. A risk scoring signature contained five cholesterogenic genes (risk score = -1.169 × GRAMD1C -0.992 × NFKBIA + 0.432 × INHBA + 0.261 × CD24 -0.839 × ACSS2) was established, which could differentiate the prognosis of young BC patients between high-risk and low-risk group (<0.001). The prediction value of chelesterogenic gene signature in excellent with AUC was 0.810 in TCGA dataset. Then the prediction value of the signature was verified in GSE131769 with P = 0.033. In experiment part, although the downregulation of CD24, GRAMD1C and ACSS2 did not significantly affect cell viability, NFKBIA downregulation promoted the viability, colony forming ability and invasion capability of BC cells, while INHBA downregulation had the opposite effects. Conclusion: The five-cholesterogenic-gene signature had independent prognostic value and robust reliability in predicting the prognosis of young BC patients. The cell experiment results suggested that NFKBIA played a protective role, while INHBA played the pro-cancer role in breast cancer.

A Signature of Autophagy-Related Long Non-coding RNA to Predict the Prognosis of Breast Cancer.

Background: A surge in newly diagnosed breast cancer has overwhelmed the public health system worldwide. Joint effort had beed made to discover the genetic mechanism of these disease globally. Accumulated research has revealed autophagy may act as a vital part in the pathogenesis of breast cancer. Objective: Aim to construct a prognostic model based on autophagy-related lncRNAs and investigate their potential mechanisms in breast cancer. Methods: The transcriptome data and clinical information of patients with breast cancer were obtained from The Cancer Genome Atlas (TCGA) database. Autophagy-related genes were obtained from the Human Autophagy Database (HADb). Long non-coding RNAs (lncRNAs) related to autophagy were acquired through the Pearson correlation analysis. Univariate Cox regression analysis as well as the least absolute shrinkage and selection operator (LASSO) regression analysis were used to identify autophagy-related lncRNAs with prognostic value. We constructed a risk scoring model to assess the prognostic significance of the autophagy-related lncRNAs signatures. The nomogram was then established based on the risk score and clinical indicators. Through the calibration curve, the concordance index (C-index) and receiver operating characteristic (ROC) curve analysis were evaluated to obtain the model's predictive performance. Subgroup analysis was performed to evaluate the differential ability of the model. Subsequently, gene set enrichment analysis was conducted to investigate the potential functions of these lncRNAs. Results: We attained 1,164 breast cancer samples from the TCGA database and 231 autophagy-related genes from the HAD database. Through correlation analysis, 179 autophagy-related lncRNAs were finally identified. Univariate Cox regression analysis and LASSO regression analysis further screened 18 prognosis-associated lncRNAs. The risk scoring model was constructed to divide patients into high-risk and low-risk groups. It was found that the low-risk group had better overall survival (OS) than those of the high-risk group. Then, the nomogram model including age, tumor stage, TNM stage and risk score was established. The evaluation index (C-index: 0.78, 3-year OS AUC: 0.813 and 5-year OS AUC: 0.785) showed that the nomogram had excellent predictive power. Subgroup analysis showed there were difference in OS between high-risk and low-risk patients in different subgroups (stage I-II, ER positive, Her-2 negative and non-TNBC subgroups; all P < 0.05). According to the results of gene set enrichment analysis, these lncRNAs were involved in the regulation of multicellular organismal macromolecule metabolic process in multicellular organisms, nucleotide excision repair, oxidative phosphorylation, and TGF-ß signaling pathway. Conclusions: We identified 18 autophagy-related lncRNAs with prognostic value in breast cancer, which may regulate tumor growth and progression in multiple ways.

Prognostic model of invasive ductal carcinoma of the breast based on differentially expressed glycolysis-related genes.

BACKGROUND: Invasive ductal carcinoma (IDC) is a common pathological type of breast cancer that is characterized by high malignancy and rapid progression. Upregulation of glycolysis is a hallmark of tumor growth, and correlates with the progression of breast cancer. We aimed to establish a model to predict the prognosis of patients with breast IDC based on differentially expressed glycolysis-related genes (DEGRGs). METHODS: Transcriptome data and clinical data of patients with breast IDC were from The Cancer Genome Atlas (TCGA). Glycolysis-related gene sets and pathways were from the Molecular Signatures Database (MSigDB). DEGRGs were identified by comparison of tumor tissues and adjacent normal tissues. Univariate Cox regression and least absolute shrinkage and selection operator (LASSO) regression were used to screen for DEGRGs with prognostic value. A risk-scoring model based on DEGRGs related to prognosis was constructed. Receiver operating characteristic (ROC) analysis and calculation of the area under the curve (AUC) were used to evaluate the performance of the model. The model was verified in different clinical subgroups using an external dataset (GSE131769). A nomogram that included clinical indicators and risk scores was established. Gene function enrichment analysis was performed, and a protein-protein interaction network was developed. RESULTS: We analyzed data from 772 tumors and 88 adjacent normal tissues from the TCGA database and identified 286 glycolysis-related genes from the MSigDB. There were 185 DEGRGs. Univariate Cox regression and LASSO regression indicated that 13 of these genes were related to prognosis. A risk-scoring model based on these 13 DEGRGs allowed classification of patients as high-risk or low-risk according to median score. The duration of overall survival (OS) was longer in the low-risk group (P < 0.001), and the AUC was 0.755 for 3-year OS and 0.726 for 5-year OS. The results were similar when using the GEO data set for external validation (AUC for 3-year OS: 0.731, AUC for 5-year OS: 0.728). Subgroup analysis showed there were significant differences in OS among high-risk and low-risk patients in different subgroups (T1-2, T3-4, N0, N1-3, M0, TNBC, non-TNBC; all P < 0.01). The C-index was 0.824, and the AUC was 0.842 for 3-year OS and 0.808 for 5-year OS from the nomogram. Functional enrichment analysis demonstrated the DEGRGs were mainly involved in regulating biological functions. CONCLUSIONS: Our prognostic model, based on 13 DEGRGs, had excellent performance in predicting the survival of patients with IDC of the breast. These DEGRGs appear to have important biological functions in the progression of this cancer.

Tumor Response to Irinotecan is Associated with IL-10 Expression Level in Metastatic Colorectal Cancer-Results from mCRC Biomarker Study.

PURPOSE: Metastatic colorectal cancer (mCRC) is a leading cause of cancer-related death. Resistance to chemotherapy is the main reason for the failure of the treatment of mCRC. IL-10 has been reported to decrease after surgery and increase after mCRC reoccurrence. The role of IL-10 in chemotherapy drug resistance of mCRC is not well elucidated. PATIENTS AND METHODS: The retrospective study recruited 264 mCRC patients between January 2012 and December 2016 (NCT03532711). All the enrolled patients received an oxaliplatin-containing or irinotecan-containing regimen. The expression level of IL-10 in 232 patients' plasma and 68 patients' tumor tissue was examined. The relationships between IL-10 and clinicopathological characteristics were analyzed. Kaplan-Meier method and Cox regression were used to evaluate the prognostic impact of IL-10. RESULTS: The median concentration of IL-10 was 7.60 pg/mL before treatment and 11.08 pg/mL after treatment, which suggested that IL-10 level was significantly increased by treatment with a chemotherapeutic regimen (p = 0.000). By utilizing univariate and multivariate Cox proportional hazard analyses, we found that low IL-10 level in plasma was significantly associated with improved overall survival (OS) of mCRC patients treated with irinotecan-containing regimen-with optimal cutoff value of 5.525pg/mL, respectively (p =0.002). In addition, the low IL-10 expression level in tumor tissue was significantly associated with the improved OS for the irinotecan-containing regimen (p = 0.023). CONCLUSION: Our study demonstrated that IL-10 could act as a prognostic biomarker for mCRC patients undergoing irinotecan-containing chemotherapy.

Decreased mean platelet volume predicts poor prognosis in metastatic colorectal cancer patients treated with first-line chemotherapy: results from mCRC biomarker study.

BACKGROUND: Metastatic colorectal cancer (mCRC) is a major cause of death of malignant tumor and the valuable prognostic biomarker for chemotherapy is crucial in decreasing mortality. Previous studies have proved the prognostic value of the mean platelet volume (MPV) in survival of primary operable CRC patients. However, the prognostic impact of MPV in mCRC is still unclear. In this study, we aimed to clarify the prognostic role of MPV in mCRC undergoing standard first-line chemotherapy. METHODS: From January 2012 to December 2016, we conducted a retrospective clinical study included 264 mCRC patients (NCT03532711). All the enrolled patients received the standard oxaliplatin-based or irinotecan-based chemotherapy. The association between the baseline MPV and clinicopathological features were examined. RESULTS: Univariate analysis revealed that decreased MPV, the platelet counts (PLT), platelet-to-lymphocyte ratio (PLR) and the platelet crit (PCT) were significantly associated with inferior overall survival (OS) (p < 0.05). On multivariate analysis, elevated PLR was significant prognostic factors for OS, with hazard ratios of (HR:1.006, 95% CI:1.001-1.011, p = 0.01) while MPV was not, respectively (p < 0.05). CONCLUSIONS: Our study demonstrated that the baseline MPV level may act as a predictive factor for survival in mCRC patients undergoing standard chemotherapy. TRIAL REGISTRATION: This study was retrospectively registered in date May the 20th 2018. The registration number (TRN) of this study was NCT03532711 .

Validation of a breast cancer nomogram to predict lymphedema in a Chinese population.

BACKGROUND: Upper arm lymphedema (LE) is a common complication after axillary lymph node dissection (ALND) in breast cancer patients. This retrospective cohort study aimed to validate a published nomogram to predict the risk of LE in the Chinese breast cancer patients. METHODS: A total of 409 breast cancer patients who underwent breast cancer surgery and ALND (level I and II) were identified. Cox regression analysis was used to identify the risk factors for LE. The nomogram predictive of LE of breast cancer was evaluated by receiver-operating curve analysis, calibration plots, and Kaplan-Meier analysis in our study population. RESULTS: With a median follow-up of 68 months, the 5-year cumulative incidence of LE was 22.3%. Higher body mass index (hazard ratio [HR] = 1.06, 95% CI: 1.00-1.13), neoadjuvant chemotherapy (HR = 3.76, 95% CI: 2.29-6.20), larger extend of axillary surgery (level I/II/III versus level I/II: HR = 2.39, 95% CI: 1.30-4.37), and radiotherapy (HR = 4.90, 95% CI: 1.90-12.5) were independently associated with LE. The AUC value of the nomogram was 0.706 (95% CI: 0.648-0.752). A high-risk subgroup of patients defined by nomogram had significantly higher cumulative risk of LE than those in the low-risk subgroups (P < 0.01). The calibration plots revealed that the nomogram was well calibrated (Hosmer-Lemeshow test, P = 0.0634). CONCLUSIONS: The nomogram to predict the risk of LE in breast cancer patients with ALND has been validated to be discriminative and accurate. More studies are needed to evaluate the impact of other factors (lifestyle, behaviors, and so forth) on the performance of the nomogram.

Association of single nucleotide polymorphisms in MTHFR and ABCG2 with the different efficacy of first-line chemotherapy in metastatic colorectal cancer.

Either oxaliplatin- or irinotecan-containing regimen could receive a good effectiveness in patients with metastatic colorectal cancer as the first-line chemotherapy, but not all patients would benefit from the treatment they have received. This study was to investigate the role of single nucleotide polymorphisms (SNPs) of methylenetetrahydrofolate reductase (MTHFR) and ATP-binding cassette sub-family G member 2 (ABCG2) in selecting the most appropriate treatment for individual patients. Ninety-two metastatic colorectal cancer patients treated with first-line 5-fluoropyrimidine (5-FU), leucovorin, and oxaliplatin (FOLFOX), capecitabine, and oxaliplatin (XELOX) and sixty-two patients receiving 5-FU, leucovorin, and irinotecan (FOLFIRI) were reviewed. The SNPs of MTHFR and ABCG2 were detected using gene sequencing method after DNA PCR amplification, which was extracted from peripheral blood karyocytes. Clinical characteristics and gene polymorphisms were evaluated in univariate and multivariate analysis as predictive factors for response rate (RR) and progression-free survival (PFS). In patients bearing 2-4 genotypes of MTHFR 677C/C, MTHFR 1298 A/C or C/C, ABCG2 34G/G, and ABCG2 421C/A or A/A, those who received oxaliplatin-based chemotherapy achieved a higher RR (41.7 vs. 18.8 %, P = 0.027) and longer median PFS (mPFS) than irinotecan-based therapy [8.9 vs. 7.1 m, FOLFIRI: hazard ratio (HR) = 1.722, 95 % confidence interval (CI) 1.026-2.892, P = 0.040, compared with FOLFOX/XELOX]; on the contrary, patients carrying 0 or 1 above genotype exhibited better outcomes after receiving FOLFIRI chemotherapy (mPFS: 9.3 vs. 6.4 m, FOLFIRI: HR = 0.422, 95 % CI 0.205-0.870, P = 0.019, compared with FOLFOX/XELOX). Combination of SNPs with MTHFR and ABCG2 may play a role in helping clinicians to select first-line chemotherapy for patients with metastatic colorectal cancer.