A novel non-negative Bayesian stacking modeling method for Cancer survival prediction using high-dimensional omics data.

BACKGROUND: Survival prediction using high-dimensional molecular data is a hot topic in the field of genomics and precision medicine, especially for cancer studies. Considering that carcinogenesis has a pathway-based pathogenesis, developing models using such group structures is a closer mimic of disease progression and prognosis. Many approaches can be used to integrate group information; however, most of them are single-model methods, which may account for unstable prediction. METHODS: We introduced a novel survival stacking method that modeled using group structure information to improve the robustness of cancer survival prediction in the context of high-dimensional omics data. With a super learner, survival stacking combines the prediction from multiple sub-models that are independently trained using the features in pre-grouped biological pathways. In addition to a non-negative linear combination of sub-models, we extended the super learner to non-negative Bayesian hierarchical generalized linear model and artificial neural network. We compared the proposed modeling strategy with the widely used survival penalized method Lasso Cox and several group penalized methods, e.g., group Lasso Cox, via simulation study and real-world data application. RESULTS: The proposed survival stacking method showed superior and robust performance in terms of discrimination compared with single-model methods in case of high-noise simulated data and real-world data. The non-negative Bayesian stacking method can identify important biological signal pathways and genes that are associated with the prognosis of cancer. CONCLUSIONS: This study proposed a novel survival stacking strategy incorporating biological group information into the cancer prognosis models. Additionally, this study extended the super learner to non-negative Bayesian model and ANN, enriching the combination of sub-models. The proposed Bayesian stacking strategy exhibited favorable properties in the prediction and interpretation of complex survival data, which may aid in discovering cancer targets.

Impact of cladribine, cytarabine, and G-CSF (CLAG) as a bridging therapy prior to allogeneic hematopoietic stem cell transplantation in relapsed or refractory acute myeloid leukemia.

The combination of cladribine, cytarabine, and G-CSF (CLAG) has exhibited robust synergistic anti-leukemia activity as an induction therapy (IT) in acute myeloid leukemia (AML). However, the impact of CLAG as a bridging therapy (BT) administered between IT and allogeneic hematopoietic stem cell transplantation (allo-HSCT) for patients with relapsed or refractory (R/R) AML remains uncertain. In this retrospective study, we examined the efficacy of CLAG as a transitional strategy prior to allo-HSCT in R/R AML. We included 234 patients with R/R AML who received the modified busulfan plus cyclophosphamide conditioning regimen for allo-HSCT in our center during the past 6 years, performed a propensity-score matching analysis, partitioned them into four distinct cohorts, and further integrated them into the CLAG group and non-CLAG group based on response to IT and utilization of CLAG. Our cohorts encompassed 12 patients in Cohort A (modified composite complete remission (mCRc) after IT, CLAG), 31 in Cohort B (mCRc after IT, non-CLAG), 35 in Cohort C (non-complete remission (non-CR) after IT, CLAG), and 80 in Cohort D (non-CR after IT, non-CLAG). Intriguingly, among patients with non-CR status, the administration of CLAG correlated with a notably statistically diminished risk of relapse and improved survival at 2-year follow-up (Cohort C vs. Cohort D). Employing CLAG as a BT prior to allo-HSCT demonstrates substantial effectiveness, a relative degree of safety, and manageable toxicity in selected R/R AML cases.

A non-negative spike-and-slab lasso generalized linear stacking prediction modeling method for high-dimensional omics data.

BACKGROUND: High-dimensional omics data are increasingly utilized in clinical and public health research for disease risk prediction. Many previous sparse methods have been proposed that using prior knowledge, e.g., biological group structure information, to guide the model-building process. However, these methods are still based on a single model, offen leading to overconfident inferences and inferior generalization. RESULTS: We proposed a novel stacking strategy based on a non-negative spike-and-slab Lasso (nsslasso) generalized linear model (GLM) for disease risk prediction in the context of high-dimensional omics data. Briefly, we used prior biological knowledge to segment omics data into a set of sub-data. Each sub-model was trained separately using the features from the group via a proper base learner. Then, the predictions of sub-models were ensembled by a super learner using nsslasso GLM. The proposed method was compared to several competitors, such as the Lasso, grlasso, and gsslasso, using simulated data and two open-access breast cancer data. As a result, the proposed method showed robustly superior prediction performance to the optimal single-model method in high-noise simulated data and real-world data. Furthermore, compared to the traditional stacking method, the proposed nsslasso stacking method can efficiently handle redundant sub-models and identify important sub-models. CONCLUSIONS: The proposed nsslasso method demonstrated favorable predictive accuracy, stability, and biological interpretability. Additionally, the proposed method can also be used to detect new biomarkers and key group structures.

Development and validation of a cuproptosis-related prognostic model for acute myeloid leukemia patients using machine learning with stacking.

Our objective is to develop a prognostic model focused on cuproptosis, aimed at predicting overall survival (OS) outcomes among Acute myeloid leukemia (AML) patients. The model utilized machine learning algorithms incorporating stacking. The GSE37642 dataset was used as the training data, and the GSE12417 and TCGA-LAML cohorts were used as the validation data. Stacking was used to merge the three prediction models, subsequently using a random survival forests algorithm to refit the final model using the stacking linear predictor and clinical factors. The prediction model, featuring stacking linear predictor and clinical factors, achieved AUC values of 0.840, 0.876 and 0.892 at 1, 2 and 3 years within the GSE37642 dataset. In external validation dataset, the corresponding AUCs were 0.741, 0.754 and 0.783. The predictive performance of the model in the external dataset surpasses that of the model simply incorporates all predictors. Additionally, the final model exhibited good calibration accuracy. In conclusion, our findings indicate that the novel prediction model refines the prognostic prediction for AML patients, while the stacking strategy displays potential for model integration.

A clinical predictive model for pre-transplantation Klebsiella pneumoniae colonization and relevance for clinical outcomes in patients receiving allogeneic hematopoietic stem cell transplantation.

The purpose of this study is to establish a clinical prediction model to discriminate patients at high risk of Klebsiella pneumoniae (KP) colonization before allogeneic hematopoietic stem cell transplantation (allo-HSCT) and evaluate the impact of KP colonization on clinical outcomes after allo-HSCT. We retrospectively collected data from 2,157 consecutive patients receiving allo-HSCT between January 2018 and March 2022. KP colonization was defined as a positive test for KP from a pharyngeal or anal swab before allo-HSCT. Logistic regression was used to build a clinical prediction model. Cox regression analyses were performed to explore the effect of KP colonization on clinical outcomes. Among all the inpatients, 166 patients had KP colonization and 581 with no positive pathogenic finding before transplantation. Seven candidate predictors were entered into the final prediction model. The prediction model had an area under the curve of 0.775 (95% CI 0.723-0.828) in the derivation cohort and 0.846 (95% CI: 0.790-0.902) in the validation cohort. Statistically significantly different incidence rates were observed among patient groups with clinically predicted low, medium, and high risk for KP infection (P < 0.001). The presence of KP colonization delayed platelet engraftment (P < 0.001) and patients with KP colonization were more likely to develop KP bloodstream infections within 100 days after allo-HSCT (P < 0.0001). Patients with KP colonization had higher non-relapse mortality (P = 0.032), worse progression-free survival (P = 0.0027), and worse overall survival within 100 days after allo-HSCT (P = 0.013). Our findings suggest that increased awareness of risks associated with pre-transplantation bacterial colonization is warranted.IMPORTANCESeveral studies have identified that Klebsiella pneumoniae (KP) is among the most common and deadly pathogens for patients in hospital intensive care units and those receiving transplantation. However, there are currently no studies that evaluate the impact of KP colonization to patients undergoing allogeneic hematopoietic stem cell transplantation. Our results confirm that pre-existing KP colonization is relatively common in a hematology transplant ward setting and negatively affects post-transplantation prognosis. Our clinical prediction model for KP colonization can support early intervention in patients at high risk to avoid subsequent bloodstream infections and improve survival outcomes. Altogether, our data suggest that increased awareness of risks associated with pre-transplantation bacterial colonization is warranted. Future studies are needed to confirm these findings and to test early intervention strategies for patients at risk of complications from KP infection.

Risk Factors for Carbapenem-Resistant Enterobacteriaceae Colonization and the Effect on Clinical Outcomes and Prognosis in Allogeneic Hematopoietic Stem Cell Transplanted Patients.

Purpose: The current study assesses which are the main risk factors, clinical outcome and prognosis following the colonization of CRE in patients that underwent allo-HSCT. Patients and Methods: A total of 343 patients subjected to allo-HSCT in the period comprised between June 2021 and June 2022 were enrolled in this retrospective study. The CRE colonization was diagnosed by clinical history and routine microbial culture of perirectal swab. In this regard, a clinical prediction model was designed based on independent risk factors underlying the pre-transplantation CRE colonization using a backward stepwise logistic regression, followed by the evaluation of its discrimination and calibration efficacies, along with clinical usefulness. Furthermore, univariate and multivariate Cox regression analyses were then conducted to assess the risk factors for post-transplantation clinical outcomes. Results: Out of 343 patients enrolled in this study, 135 (39.3%) reported CRE colonization. The independent risk factor variables for CRE colonization were incorporated into the nomogram to build a prediction model, which showed an area under the curve of 0.767 (95% CI: 0.716-0.818), and well-fitted calibration curves (χ2 = 1.737, P = 0.9788). The patients with CRE colonization reported a significantly lower platelet engraftment rate with a higher risk of post-transplantation BSI when compared with the non-CRE colonization group (P = 0.02 and P < 0.001; respectively). The non-relapse mortality (NRM) value was higher in the CRE patients (P < 0.05), consistently with a survival probability that was thus significantly lower for the same timeframe (P < 0.05). Conclusion: A reliable clinical prediction model for pre-transplantation CRE colonization was developed that demonstrated that the CRE colonization negatively affects platelet engraftment and survival outcomes following allo-HSCT.

High FAAP24 expression reveals poor prognosis and an immunosuppressive microenvironment shaping in AML.

BACKGROUND: As a core member of the FA complex, in the Fanconi anemia pathway, FAAP24 plays an important role in DNA damage repair. However, the association between FAAP24 and patient prognosis in AML and immune infiltration remains unclear. The purpose of this study was to explore its expression characteristics, immune infiltration pattern, prognostic value and biological function using TCGA-AML and to verify it in the Beat AML cohort. METHODS: In this study, we examined the expression and prognostic value of FAAP24 across cancers using data from TCGA, TARGET, GTEx, and GEPIA2. To further investigate the prognosis in AML, development and validation of a nomogram containing FAAP24 were performed. GO/KEGG, ssGSEA, GSVA and xCell were utilized to explore the functional enrichment and immunological features of FAAP24 in AML. Drug sensitivity analysis used data from the CellMiner website, and the results were confirmed in vitro. RESULTS: Integrated analysis of the TCGA, TARGET and GTEx databases showed that FAAP24 is upregulated in AML; meanwhile, high FAAP24 expression was associated with poor prognosis according to GEPIA2. Gene set enrichment analysis revealed that FAAP24 is implicated in pathways involved in DNA damage repair, the cell cycle and cancer. Components of the immune microenvironment using xCell indicate that FAAP24 shapes an immunosuppressive tumor microenvironment (TME) in AML, which helps to promote AML progression. Drug sensitivity analysis showed a significant correlation between high FAAP24 expression and chelerythrine resistance. In conclusion, FAAP24 could serve as a novel prognostic biomarker and play an immunomodulatory role in AML. CONCLUSIONS: In summary, FAAP24 is a promising prognostic biomarker in AML that requires further exploration and confirmation.

Comprehensive analysis of ERCC3 prognosis value and ceRNA network in AML.

BACKGROUND: Acute myeloid leukemia (AML) is a hematological malignancy with high molecular and clinical heterogeneity, and is the most common type of acute leukemia in adults. Due to limited treatment options, AML is prone to relapse and has a poor prognosis. Excision repair cross-complementing 3 (ERCC3) is an important member of nucleotide excision repair (NER) that is overexpressed in types of solid cancers and potentially regarded as a prognostic factor. However, its role in AML remains unclear. The purpose of this study was to explore ERCC3 expression and functions in AML. METHODS: The Cancer Genome Atlas (TCGA) and GEO (Gene Expression Omnibus) were used to test the accuracy of ERCC3 expression levels for AML diagnosis. Using online databases and R packages, we also explored the signaling pathway, epigenetic regulation, infiltration of immune cells, clinical prognostic value, and ceRNA network in AML. RESULTS: Our results revealed that ERCC3 expression was increased in AML and that high ERCC3 expression had good value for disease-free survival and overall survival in AML patients who underwent allogeneic hematopoietic stem cell transplantation (allo-HSCT). We found that ERCC3 and co-expressed genes were mainly involved in chemical carcinogenesis/reactive oxygen species, ubiquitin-mediated protein degradation and oxidative phosphorylation. In addition, almost all the m6A-related coding genes (except GF2BP1) were positively associated with ERCC3 expression. We also constructed a ceRNA regulatory network containing ERCC3 in AML and identified 6 pairs of ceRNA networks, indicating that ERCC3 expression is regulated by a noncoding RNA system. CONCLUSION: This study demonstrated that ERCC3 was overexpressed in AML and that high ERCC3 expression can be considered a biomarker conducive to allo-HSCT in AML patients.

Diagnostic value of radiomics model based on gray-scale and contrast-enhanced ultrasound for inflammatory mass stage periductal mastitis/duct ectasia.

Objective: The present study aimed to investigate the clinical application value of the radiomics model based on gray-scale ultrasound (GSUS) and contrast-enhanced ultrasound (CEUS) images in the differentiation of inflammatory mass stage periductal mastitis/duct ectasia (IMSPDM/DE) and invasive ductal carcinoma (IDC). Methods: In this retrospective study, 254 patients (IMSPDM/DE: 129; IDC:125) were enrolled between January 2018 and December 2020 as a training cohort to develop the classification models. The radiomics features were extracted from the GSUS and CEUS images. The least absolute shrinkage and selection operator (LASSO) regression model was employed to select the corresponding features. Based on these selected features, logistic regression analysis was used to aid the construction of these three radiomics signatures (GSUS, CEUS and GSCEUS radiomics signature). In addition, 80 patients (IMSPDM/DE:40; IDC:40) were recruited between January 2021 and November 2021 and were used as the validation cohort. The best radiomics signature was selected. Based on the clinical parameters and the radiomics signature, a classification model was built. Finally, the classification model was assessed using nomogram and decision curve analyses. Results: Three radiomics signatures were able to differentiate IMSPDM/DE from IDC. The GSCEUS radiomics signature outperformed the other two radiomics signatures and the AUC, sensitivity, specificity, and accuracy were estimated to be 0.876, 0.756, 0.804, and 0.798 in the training cohort and 0.796, 0.675, 0.838 and 0.763 in the validation cohort, respectively. The lower patient age (p<0.001), higher neutrophil count (p<0.001), lack of pausimenia (p=0.023) and GSCEUS radiomics features (p<0.001) were independent risk factors of IMSPDM/DE. The classification model that included the clinical factors and the GSCEUS radiomics signature outperformed the GSCEUS radiomics signature alone (the AUC values of the training and validation cohorts were 0.962 and 0.891, respectively). The nomogram was applied to the validation cohort, reaching optimal discrimination, with an AUC value of 0.891, a sensitivity of 0.888, and a specificity of 0.750. Conclusions: The present study combined the clinical parameters with the GSCEUS radiomics signature and developed a nomogram. This GSCEUS radiomics-based classification model could be used to differentiate IMSPDM/DE from IDC in a non-invasive manner.

The impact of Rituximab administered before transplantation in patients undergoing allogeneic hematopoietic stem cell transplantation: A real-world study.

Rituximab is used to eliminate B cells as a chimeric monoclonal antibody directed against CD20, a B-cell antigen expressed on B cells. To explore the impact of rituximab administered before transplantation, we implemented a retrospective, monocentric study and utilized real-world data collected at our center between January 2018 and December 2020, and then followed until December 2021. Based on whether a dose of 375mg/m2 rituximab was used at least once within two weeks before transplantation, patients undergoing allo-HSCT were classified into two groups: rituximab (N=176) and non-rituximab (N=344) group. Amongst all the patients, the application of rituximab decreased EBV reactivation (P<0.01) and rituximab was an independent factor in the prevention of EBV reactivation by both univariate and multivariate analyses (HR 0.56, 95%CI 0.33-0.97, P=0.04). In AML patients, there were significant differences in the cumulative incidence of aGVHD between the two groups (P=0.04). Our data showed that rituximab was association with a decreased incidence of aGVHD in AML patients according to both univariate and multivariate analyses. There was no difference between the two groups in other sets of populations. Thus, our study indicated that rituximab administered before transplantation may help prevent EBV reactivation in all allo-HSCT patients, as well as prevent aGVHD in AML patients after allo-HSCT.

Development and Validation of Prognostic Model for Lung Adenocarcinoma Patients Based on m6A Methylation Related Transcriptomics.

Existing studies suggest that m6A methylation is closely related to the prognosis of cancer. We developed three prognostic models based on m6A-related transcriptomics in lung adenocarcinoma patients and performed external validations. The TCGA-LUAD cohort served as the derivation cohort and six GEO data sets as external validation cohorts. The first model (mRNA model) was developed based on m6A-related mRNA. LASSO and stepwise regression were used to screen genes and the prognostic model was developed from multivariate Cox regression model. The second model (lncRNA model) was constructed based on m6A related lncRNAs. The four steps of random survival forest, LASSO, best subset selection and stepwise regression were used to screen genes and develop a Cox regression prognostic model. The third model combined the risk scores of the first two models with clinical variable. Variables were screened by stepwise regression. The mRNA model included 11 predictors. The internal validation C index was 0.736. The lncRNA model has 15 predictors. The internal validation C index was 0.707. The third model combined the risk scores of the first two models with tumor stage. The internal validation C index was 0.794. In validation sets, all C-indexes of models were about 0.6, and three models had good calibration accuracy. Freely online calculator on the web at https://lhj0520.shinyapps.io/LUAD_prediction_model/.

Relationship of Oropharyngeal Colonization Microorganisms to Clinical Outcomes within 100 Days after Allogeneic Hematopoietic Stem Cell Transplantation.

Little is known about oropharyngeal colonization microorganisms in patients during allogeneic hematopoietic stem cell transplantation (allo-HSCT), and updated epidemiologic investigations are advisable. This study aimed to characterize oropharyngeal colonization microorganisms in patients during allo-HSCT and confirm whether they were related to clinical outcomes. This retrospective, matched case-control study included 1267 consecutive patients undergoing allo-HSCT between January 2018 and December 2020 at our institution. Patients with oropharyngeal colonization microorganisms were those with a positive throat swab before or on the day of transplantation without the occurrence of any symptoms of infection. Propensity score matching was used. Characteristics of oropharyngeal colonization microorganisms were evaluated among patients in the transplant medicine wards and compared with clinical outcomes within 100 days in positive and negative colonization groups. A total of 127 patients had oropharyngeal colonization microorganisms before or on the day of transplantation. Using propensity score matching, we matched the 127 patients in the positive colonization group with 508 patients in the negative colonization group at a 1:4 ratio (total of 635 cases). None of the differences in clinical traits between the 2 groups remained significant. Among the 127 patients with oropharyngeal colonization microorganisms, 90 patients suffered from the documented infection subsequently, and the others were asymptomatic. A total of 82 single gram-negative bacteria were identified in 127 isolates. There were no differences between the positive and negative colonization groups in the occurrence of oral mucositis, Epstein-Barr virus, or acute graft-versus-host disease and relapse within 100 days. However, the rate of neutrophil or platelet recovery was significantly lower in the positive colonization group compared with the negative colonization group (hazard ratio [HR], .71; 95% confidence interval [CI], .59 to .84; P < .001; HR .69; 95% CI, .58 to .83; P = .003; separately). The risk of bloodstream infection was higher in the positive colonization group compared with the negative colonization group (HR, 6.09; 95% CI, 3.16 to 11.75; P < .001). The continency rate between the bacteria isolated from the blood samples and oropharyngeal colonization microorganisms among the patients with positive results was 73.3%. Patients in the positive colonization group were more vulnerable to cytomegalovirus infection compared with the negative colonization group (HR, 1.41; 95% CI, 1.00 to 1.99; P = .049). The nonrelapse mortality at day +100 was higher in the positive colonization group (HR, 3.46; 95% CI, 1.69 to 7.08; P < .001). The survival probability within 100 days was significantly lower in the positive colonization group (HR, 3.38; 95% CI, 1.78 to 6.41; P < .001). Our data show that the presence of oropharyngeal colonization microorganisms is related to clinical outcomes, and that oropharyngeal microorganism monitoring may be useful during allo-HSCT.

Development and validation of an immune-related gene signature for predicting the radiosensitivity of lower-grade gliomas.

Radiotherapy is an important treatment modality for lower-grade gliomas (LGGs) patients. This analysis was conducted to develop an immune-related radiosensitivity gene signature to predict the survival of LGGs patients who received radiotherapy. The clinical and RNA sequencing data of LGGs were obtained from The Cancer Genome Atlas (TCGA) and the Chinese Glioma Genome Atlas (CGGA). Lasso regression analyses were used to construct a 21-gene signature to identify the LGGs patients who could benefit from radiotherapy. Based on this radiosensitivity signature, patients were classified into a radiosensitive (RS) group and a radioresistant (RR) group. According to the Kaplan-Meier analysis results of the TCGA dataset and the two CGGA validation datasets, the RS group had a higher overall survival rate than that of the RR group. This gene signature was RT-specific and an independent prognostic indicator. The nomogram model performed well in predicting 3-, and 5-year survival of LGGs patients after radiotherapy by this gene signature and other clinical factors (age, sex, grade, IDH mutations, 1p/19q codeletion). In summary, this signature is a powerful supplement to the prognostic factors of LGGs patients with radiotherapy and may provide an opportunity to incorporate individual tumor biology into clinical decision making in radiation oncology.

Integration of immune and hypoxia gene signatures improves the prediction of radiosensitivity in breast cancer.

Immunity and hypoxia are two important factors that affect the response of cancer patients to radiotherapy. At the same time, considering the limited predictive value of a single predictive model and the uncertainty of grouping patients near the cutoff value, we developed and validated a combined model based on immune- and hypoxia-related gene expression profiles to predict the radiosensitivity of breast cancer patients. This study was based on breast cancer data from The Cancer Genome Atlas (TCGA). Spike-and-slab Lasso regression analysis was performed to select three immune-related genes and develop a radiosensitivity model. Lasso Cox regression modeling selected 11 hypoxia-related genes for development of radiosensitivity model. Three independent datasets (Molecular Taxonomy of Breast Cancer International Consortium [METABRIC], E-TABM-158, GSE103746) were used to validate the predictive value of radiosensitivity signatures. In the TCGA dataset, the 10-year survival probabilities of the immune radioresistant (IRR) and hypoxia radioresistant (HRR) groups were 0.189 (0.037, 0.973) and 0.477 (0.293, 0.776), respectively. The 10-year survival probabilities of the immune radiosensitive (IRS) and hypoxia radiosensitive (HRS) groups were 0.778 (0.676, 0.895) and 0.824 (0.723, 0.939), respectively. Based on these two gene signatures, we further constructed a combined model and divided all patients into three groups (IRS/HRS, mixed, IRR/HRR). We identified the IRS/HRS patients most likely to benefit from radiotherapy; the 10-year survival probability was 0.886 (0.806, 0.976). The 10-year survival probability of the IRR/HRR group was 0. In conclusion, a combined model integrating immune- and hypoxia-related gene signatures could effectively predict the radiosensitivity of breast cancer and more accurately identify radiosensitive and radioresistant patients than a single model.

A Radiosensitivity Prediction Model Developed Based on Weighted Correlation Network Analysis of Hypoxia Genes for Lower-Grade Glioma.

Background and Purpose: Hypoxia is one of the basic characteristics of the physical microenvironment of solid tumors. The relationship between radiotherapy and hypoxia is complex. However, there is no radiosensitivity prediction model based on hypoxia genes. We attempted to construct a radiosensitivity prediction model developed based on hypoxia genes for lower-grade glioma (LGG) by using weighted correlation network analysis (WGCNA) and least absolute shrinkage and selection operator (Lasso). Methods: In this research, radiotherapy-related module genes were selected after WGCNA. Then, Lasso was performed to select genes in patients who received radiotherapy. Finally, 12 genes (AGK, ETV4, PARD6A, PTP4A2, RIOK3, SIGMAR1, SLC34A2, SMURF1, STK33, TCEAL1, TFPI, and UROS) were included in the model. A radiosensitivity-related risk score model was established based on the overall rate of The Cancer Genome Atlas (TCGA) dataset in patients who received radiotherapy. The model was validated in TCGA dataset and two Chinese Glioma Genome Atlas (CGGA) datasets. A novel nomogram was developed to predict the overall survival of LGG patients. Results: We developed and verified a radiosensitivity-related risk score model based on hypoxia genes. The radiosensitivity-related risk score served as an independent prognostic indicator. This radiosensitivity-related risk score model has prognostic prediction ability. Moreover, a nomogram integrating risk score with age and tumor grade was established to perform better for predicting 1-, 3-, and 5-year survival rates. Conclusions: We developed and validated a radiosensitivity prediction model that can be used by clinicians and researchers to predict patient survival rates and achieve personalized treatment of LGG.

Robust radiosensitization of hemoglobin-curcumin nanoparticles suppresses hypoxic hepatocellular carcinoma.

BACKGROUND: Radioresistance inducing by hypoxic microenvironment of hepatocellular carcinoma is a major obstacle to clinical radiotherapy. Advanced nanomedicine provides an alternative to alleviate the hypoxia extent of solid tumor, even to achieve effective synergistic treatment when combined with chemotherapy or radiotherapy. RESULTS: Herein, we developed a self-assembled nanoparticle based on hemoglobin and curcumin for photoacoustic imaging and radiotherapy of hypoxic hepatocellular carcinoma. The fabricated nanoparticles inhibited hepatoma migration and vascular mimics, and enhanced the radiosensitivity of hypoxic hepatoma cells in vitro via repressing cell proliferation and DNA damage repair, as well as inducing apoptosis. Benefit from oxygen-carrying hemoglobin combined with polyphenolic curcumin, the nanoparticles also effectively enhanced the photoacoustic contrast and the efficacy of radiotherapy for hepatocellular carcinoma in vivo. CONCLUSIONS: Together, the current study offered a radiosensitization platform for optimizing the efficacy of nanomedicines on hypoxic radioresistant tumor.

Efficient feature extraction from highly sparse binary genotype data for cancer prognosis prediction using an auto-encoder.

Genomics involving tens of thousands of genes is a complex system determining phenotype. An interesting and vital issue is how to integrate highly sparse genetic genomics data with a mass of minor effects into a prediction model for improving prediction power. We find that the deep learning method can work well to extract features by transforming highly sparse dichotomous data to lower-dimensional continuous data in a non-linear way. This may provide benefits in risk prediction-associated genotype data. We developed a multi-stage strategy to extract information from highly sparse binary genotype data and applied it for cancer prognosis. Specifically, we first reduced the size of binary biomarkers via a univariable regression model to a moderate size. Then, a trainable auto-encoder was used to learn compact features from the reduced data. Next, we performed a LASSO problem process to select the optimal combination of extracted features. Lastly, we applied such feature combination to real cancer prognostic models and evaluated the raw predictive effect of the models. The results indicated that these compressed transformation features could better improve the model's original predictive performance and might avoid an overfitting problem. This idea may be enlightening for everyone involved in cancer research, risk reduction, treatment, and patient care via integrating genomics data.

FEN1 Status and Its Correlation with Clinicopathologic Characteristic in Colorectal Cancer.

OBJECTIVE: The goal of this study was to investigate the status of FEN1 in colorectal cancer (CRC) and determine the potential correlation between FEN1 expression level and clinicopathological parameters in CRC patients. METHODS: Expression of FEN1 in CRC tissue on tissue microarray was detected using immunohistochemistry (IHC). The relationship between FEN1 expression status and clinicopathologic characteristics of CRC was analyzed by the Chi-square test. The survival data of TCGA Colon Cancer (COAD) were obtained from ucsc xena browser (https://xenabrowser.net/). Patients were separated into higher and lower expression groups by median FEN1 expression. The association with prognosis of CRC patients was determined by Kaplan-Meier survival analysis with Log-rank test. RESULTS: FEN1expression level and cellular localization had wide variability among different individuals; we classified the staining results into four types: both positive in nucleus and cytoplasm, both negative in nucleus and cytoplasm, only positive in the nucleus, only positive in the cytoplasm. Moreover, FEN1 expression status only correlated with patient's metastasis status, and the patients in the NLCL group showed more risk of cancer cell metastasis. CONCLUSION: Our results indicate that FEN1 expression level and cellular localization had wide variability in CRC and is not a promising biomarker in CRC.

Explore association of genes in PDL1/PD1 pathway to radiotherapy survival benefit based on interaction model strategy.

PURPOSE: To explore the association of genes in "PD-L1 expression and PD-1 check point pathway in cancer" to radiotherapy survival benefit. METHODS AND MATERIALS: Gene expression data and clinical information of cancers were downloaded from TCGA. Radiotherapy survival benefit was defined based on interaction model. Fast backward multivariate Cox regression was performed using stacking multiple interpolation data to identify radio-sensitive (RS) genes. RESULTS: Among the 73 genes in PD-L1/PD-1 pathway, we identified 24 RS genes in BRCA data set, 25 RS genes in STAD data set and 20 RS genes in HNSC data set, with some crossover genes. Theoretically, there are two types of RS genes. The expression level of Type I RS genes did not affect patients' overall survival (OS), but when receiving radiotherapy, patients with different expression level of Type I RS genes had varied survival benefit. Oppositely, Type II RS genes affected patients' OS. And when receiving radiotherapy, those with lower OS could benefit a lot. Type II RS genes in BRCA had strong positive correlation and closely biological interactions. When performing cluster analysis using these related Type II RS genes, patients could be divided into RS group and non-RS group in BRCA and METABRIC data sets. CONCLUSIONS: Our study explored potential radio-sensitive biomarkers of several main cancer types in an important tumor immune checkpoint pathway and revealed a strong association between this pathway and radiotherapy survival benefit. New types of RS genes could be identified based on expanded definition to RS genes.