Semi-supervised classification using sparse representation for cancer recurrence prediction

Compared with the phase angle (PA), the predictive ability of the standardized phase angle (SPA) in assessing nutritional status and clinical outcomes in cancer patients remains uncertain. This review aimed to assess (1) the relationship between SPA and nutritional status and clinical outcomes (including complications and survival) in cancer patients; (2) the predictive ability of SPA alone and in comparison with the predictive ability of PA; and (3) the cut-off value of SPA in cancer patients. Studies that addressed the relationship of SPA use to nutritional status, complications, and survival in cancer patients were searched and identified from six electronic databases (PubMed, Medline, CINAHL, Embase, Web of Science, and the Cochrane Library). The included studies were considered to meet the following criteria: English studies with original data that reflected the effects of SPA on nutritional status and clinical outcomes (including complications and survival) and reported a cut-off value of SPA in cancer patients aged ≥18. Thirteen studies that included a total of 2787 participants were evaluated. Five studies assessed the relationship between SPA and nutritional status, and four of them reported a positive relationship between SPA and nutritional status in cancer patients, even considering SPA as a predictor. Twelve studies assessed the relationship between SPA and clinical outcomes in cancer patients. Two-thirds of the studies that evaluated complications reported the predictive ability of SPA; 30% of survival studies reported a positive relationship, 40% reported SPA as a predictor, and 30% reported no relationship. The standard cut-off value for SPA has not yet been determined. Data from the selected studies suggest that SPA might be a predictor of nutritional status. Further studies are needed to determine the value of SPA in predicting nutritional status and clinical outcomes in cancer patients.

We propose a semisupervised sparse graph learning (SSGL) method for feature extraction of hyperspectral remote sensing imagery in this paper. The proposed SSGL method aims to build a semisupervised sparse graph that can maximize the class discrimination and preserve the local neighborhood information by combining labeled and unlabeled samples. In our semisupervised sparse graph, we connect labeled samples according to their label information and sparse representation, connect unlabeled samples by sparse representation which set unlabeled training samples set as dictionary. Moreover, by setting sparse connections between labeled sample and unlabeled samples, the label information can be well propagated from labeled samples to unlabeled samples. In this way, the similarity of data points can be well modelled in the spectral feature space, and the data distribution characteristics can be better expressed. Experimental results on real hyperspectral images (HSIs) demonstrate the advantages of our method compared to some related feature extraction methods.

Fast Laplacian twin support vector machine with active learning for pattern classification

A kernel-free Laplacian quadratic surface optimal margin distribution machine with application to credit risk assessment

Combinations of oxaliplatin combined with various schedules of antimetabolite 5- fluorouracil and leucovorin (FOLFOX) have been the first-line treatment for unresectable colorectal cancer (CRC) and survival benefits for the patients undergoing FOLFOX therapy were demonstrated. That is, the median duration of survival for the group treated with FOLFOX has been reported to be about 20 months, compared with 8 months for the best supportive care group. However, the rate of effect is not so high and has been reported to be approximately 50%. Molecular characterization using gene expression profiling will undoubtedly improve the prediction of treatment responses and ultimately the clinical outcome of cancer patients, which has never been accomplished by conventional clinicopathological classification. The aim of this study was to identify responders for FOLFOX therapy by applying Random Forests (RF) algorithm to gene expression data. Sixty-six unresectable colorectal cancer (CRC) patients were included, which were divided into training and test sets. Evaluated by computed tomography after 4 courses of FOLFOX therapy, 25 patients were diagnosed as responders and 25 patients as non-responders, respectively, in 50 training samples. Using RF, 22 of 25 responders (88.0% sensitivity) and 24 of 25 non-responders (96.0% specificity) were correctly classified, with an accuracy of 92.0% in out-of-bag cross-validation, and predictor genes were identified for FOLFOX therapy. Based on the predictor genes, proximities could clearly classify tumor samples into three subtypes, responders, non-responders, and outliers. After getting 5 outliers out of 50 training samples, RF was performed again with identification of predictor genes, such as RDX, SMURF2, and PEG10. In order to validate these predictor genes, we applied these decision tree analyses to independent 16 test samples, and achieved 75.9% sensitivity, 88.9% specificity, and 84.6% of classification accuracy. Based on these predictor genes, we performed Gene Set Enrichment Analysis and found activated pathways such as ‘hematopoietic stem cell genes’ pathway (including CIRBP and NPEPPS). Furthermore, we investigated survival benefit of the patients who were diagnosed as responders for FOLFOX therapy by RF algorithm and found the survival distribution for unresectable CRC outcome. Taken together, RF on gene expression data for unresectable CRC patients could effectively stratify responders for FOLFOX therapy with high accuracy rate, and pharmacogenomics in anticancer therapy is definitely the first step for personalized therapy. Citation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the 102nd Annual Meeting of the American Association for Cancer Research; 2011 Apr 2-6; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2011;71(8 Suppl):Abstract nr 4135. doi:10.1158/1538-7445.AM2011-4135

Methylation of Cancer-Stem-Cell-Associated Wnt Target Genes Predicts Poor Prognosis in Colorectal Cancer Patients

With the increasing data dimensionality, feature selection has become a fundamental task to deal with high-dimensional data. Semi-supervised feature selection focuses on the problem of how to learn a relevant feature subset in the case of abundant unlabeled data with few labeled data. In recent years, many semi-supervised feature selection algorithms have been proposed. However, these algorithms are implemented by separating the processes of feature selection and classifier training, such that they cannot simultaneously select features and learn a classifier with the selected features. Moreover, they ignore the difference of reliability inside unlabeled samples and directly use them in the training stage, which might cause performance degradation. In this paper, we propose a joint semi-supervised feature selection and classification algorithm (JSFS) which adopts a Bayesian approach to automatically select the relevant features and simultaneously learn a classifier. Instead of using all unlabeled samples indiscriminately, JSFS associates each unlabeled sample with a self-adjusting weight to distinguish the difference between them, which can effectively eliminate the irrelevant unlabeled samples via introducing a left-truncated Gaussian prior. Experiments on various datasets demonstrate the effectiveness and superiority of JSFS.

Safe intuitionistic fuzzy twin support vector machine for semi-supervised learning

418P Analysis of the efficacy of cytokine-induced killer cell therapy in adjuvant treatment of colorectal cancer

Introduction: Under surgical trauma, the pathological amplification and activation of neutrophils with potent immunosuppressive functions, also known as polymorphonuclear myeloid-derived suppressive cells (PMN-MDSCs), can promote metastasis by forming neutrophil extracellular traps (NETs) that are web-like structures consisting of nuclear DNA, histones, and granule proteins. Preoperative exercise (PEx) reduces postoperative complications with improved clinical outcomes in cancer patients undergoing surgeries. are. Our preliminary data show that PEx suppresses liver ischemia/reperfusion (I/R)-induced colorectal cancer (CRC) liver metastasis. Our recent studies have shown I/R-induced NETs promote liver CRC metastasis by capturing tumor cells and preventing infiltration and cytotoxicity of CD8+ T cells in tumor microenvironment (TME). We hypothesize that PEx remodels the TME by suppressing PMN-MDSCs formation of NETs. Methods: Male mice were subjected to exercise for 4 weeks, before MC38 colorectal cancer cells were injected through portal vein, and then subjected to 70% partial liver warm I/R as surgical stress. The single-cell RNA sequencing (scRNA-seq) was performed on CD45+leukocytes from TME 3 weeks after surgery. Immunofluorescence staining and Western Blot were used to determine NETs formation in the TME. Flow cytometry was used to determine the infiltration of PMN-MDSCs and CD8+ T lymphocytes in the tumors. Results: Our scRNA-seq data reveal that PEx leads to a remarkable transcriptomic shift in most immune cells types, such as PMN-MDSCs, M-MDSCs, T cells, Kupffer cells, B cells, NK and NKT cells, in the TME. PEx decreases the percentages of PMN-MDSCs in the TME and inhibits theNETs formation by PMN-MDSCs. scRNA-seq shows a significant increase of the chemokines CXCL10 in PEx-trained PMN-MDSCs obtained 3 weeks after I/R-induced hepatic metastasis, compared with sedentary controls. In addition, PEx increases the number of CD8+ T lymphocytes infiltrating the TME. Summary: Our study shows that PEx downregulates NETs formation and increases the CXCL10 expression in PMN-MDSCs, thus promoting the recruitment more cytotoxic CD8 T lymphocytes to suppress tumor metastasis. Citation Format: Xiang Cheng, Hongji Zhang, Allan Tsung, Hai Huang. Preoperative exercise inhibits hepatic metastasis by suppressing PMN-MDSC formation of NETs [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2022; 2022 Apr 8-13. Philadelphia (PA): AACR; Cancer Res 2022;82(12_Suppl):Abstract nr 2544.

The tumor suppressor p53, encoded by the TP53 gene, is mutated or nullified in nearly 50% of human cancers. It has long been debated whether TP53 mutations can be utilized as a biomarker to predict clinical outcomes of cancer patients. In this study, we applied computational methods to calculate p53 deficiency scores (PDSs) that reflect the inactivation of the p53 pathway, instead of TP53 mutation status. Compared to TP53 mutation status, the p53 deficiency gene signature is a powerful predictor of overall survival and drug sensitivity in a variety of cancer types and treatments. Interestingly, the PDSs predicted clinical outcomes more accurately than drug sensitivity in cell lines, suggesting that tumor heterogeneity and/or tumor microenvironment may play an important role in predicting clinical outcomes using p53 deficiency gene signatures.

Germline variants such as BRCA1/2 play an important role in tumorigenesis and clinical outcomes of cancer patients. However, only a small fraction (i.e., 5–10%) of inherited variants has been associated with clinical outcomes (e.g., BRCA1/2, APC, TP53, PTEN and so on). The challenge remains in using these inherited germline variants to predict clinical outcomes of cancer patient population. In an attempt to solve this issue, we applied our recently developed algorithm, eTumorMetastasis, which constructs predictive models, on exome sequencing data to ER+ breast (n = 755) cancer patients. Gene signatures derived from the genes containing functionally germline variants significantly distinguished recurred and non-recurred patients in two ER+ breast cancer independent cohorts (n = 200 and 295, P = 1.4 × 10−3). Furthermore, we compared our results with the widely known Oncotype DX test (i.e., Oncotype DX breast cancer recurrence score) and outperformed prediction for both high- and low-risk groups. Finally, we found that recurred patients possessed a higher rate of germline variants. In addition, the inherited germline variants from these gene signatures were predominately enriched in T cell function, antigen presentation, and cytokine interactions, likely impairing the adaptive and innate immune response thus favoring a pro-tumorigenic environment. Hence, germline genomic information could be used for developing non-invasive genomic tests for predicting patients’ outcomes in breast cancer.

Background:Previous studies indicate that endothelial injury, as demonstrated by the presence of circulating endothelial cells (CECs), may predict clinical outcome in cancer patients. In addition, soluble CD146 (sCD146) may reflect activation of angiogenesis. However, no study has investigated their combined clinical value in patients undergoing resection for non-small cell lung cancer (NSCLC).Methods:Data were collected from preoperative blood samples from 74 patients who underwent resection for NSCLC. Circulating endothelial cells were defined, using the CellSearch Assay, as CD146+CD105+CD45−DAPI+. In parallel, sCD146 was quantified using an ELISA immunoassay. These experiments were also performed on a group of 20 patients with small-cell lung cancer, 60 healthy individuals and 23 patients with chronic obstructive pulmonary disease.Results:The CEC count and the plasma level of sCD146 were significantly higher in NSCLC patients than in the sub-groups of controls (P<0.001). Moreover, an increased CEC count was associated with higher levels of sCD146 (P=0.010). Both high CEC count and high sCD146 plasma level at baseline significantly correlated with shorter progression-free survival (P<0.001, respectively) and overall survival (P=0.005; P=0.009) of NSCLC patients.Conclusions:The present study provides supportive evidence to show that both a high CEC count and a high sCD146 level at baseline correlate with poor prognosis and may be useful for the prediction of clinical outcome in patients undergoing surgery for NSCLC.

Sarcopenia and frailty status have been shown to be associated with cancer patients' survival. Low alanine aminotransferase (ALT) blood activity was previously shown to be associated with frailty and poor clinical outcomes in cancer patients. This association was not addressed in head and neck cancer patients. This was a retrospective study, investigating the possible association between pre-operative low ALT blood levels and clinical outcomes in patients undergoing total laryngectomy due to malignancy. We did a retrospective analysis using the largest national Israeli HMO (Clalit health maintenance organization) database, including the demographic, clinical background, laboratory and clinical outcomes of head and neck cancer patients who underwent total laryngectomy. A total of 427 patients were included. Patients were divided into two groups according to their ALT values: low ALT (0-10 IU/L, N = 100) and normal ALT (11-40 IU/L, N = 327). Low ALT levels were associated with significantly longer duration of in-hospital stay (26 ± 17 vs. 23 ± 16 days, respectively; p < 0.001), higher probability of 1-month, post-discharge unplanned visits to the emergency department (RR = 1.66; p < 0.05) and worse 1-year survival rate (67% vs. 77%, respectively; p < 0.05). Low, pre-operative ALT levels are associated with worse prognosis for total laryngectomy patients. ALT levels may be used for better, personalised medicine by surgeons.

Skin toxicity is a frequent complication of anti-epidermal growth factor receptor therapy, which can be an obstacle in maintaining the dose intensity and may negatively impact on the clinical outcome of cancer patients. Skin lesions depend on the disruption of the keratinocyte development pathways and no treatment is clearly effective in resolving the cutaneous alterations frequently found during anti-epidermal growth factor receptor therapy. Among systemic treatments, oral tetracycline proved to be useful in preventing skin manifestations. We describe the case of a patient affected by metastatic colorectal cancer, for whom a combination of chemotherapy and cetuximab was used as second-line treatment. The patient developed a symptomatic papulopustular skin rash that disappeared completely after a twice-daily application of a hydrating and moisturizing cream, mainly consisting of a mixture of paraffin, silicone compounds, and macrogol. The marked cutaneous amelioration allowed the patient to continue cetuximab without any further symptoms and was associated with a partial radiological response.