Violations of proportional hazard assumption in Cox regression model of transcriptomic data in TCGA pan-cancer cohorts

Multisystem failure, tipping points, and risk of Alzheimer's disease

INTRODUCTION

Medical conditions including obesity, diabetes, hyperlipidemia, and depression significantly increased risk of Alzheimer's disease (AD). However, effect of their duration, influenced by non-modifiable factors like chromosomal sex and apolipoprotein E (APOE) genotype, remains unclear.

METHODS

Data from 5644 UKBiobank participants were analyzed using Cox regression model to identify critical tipping points based on age of onset, risk factor (RF) duration and their interaction with sex and APOE genotype.

RESULTS

Hypertension or diabetes before age 62 exerted greater AD risk than APOEε4 alone. Obesity before age 62 increased AD risk by 54%, with the risk nearly tripling between ages 62-72. Hyperlipidemia and depression were associated with age-independent risk increases of 33% and 69%, respectively. After age 72, APOEε4 became the dominant RF.

DISCUSSION

Duration of AD-risk-factors can have a greater impact than APOEε4. Identification of critical age-related tipping points highlights temporal dynamics of AD progression and role of multisystem failure in AD progression.

HIGHLIGHTS

AD risk factors impact AD onset, especially diagnosed between ages 62 and 72. Later diagnoses of hypertension, diabetes, and obesity delayed AD onset. Hyperlipidemia and depression increased AD risk by 33% and 69%, age-independent. APOEε4 carriers regardless of sex exhibited a higher risk increasing with age. Trajectories differed between APOEε4 carriers and non-carriers across sex.

Statistics Primer for Radiologists: Part 2-Advanced statistics for Enhancing Diagnostic Precision and Research Validity

Second part of this statistics primer focuses on advanced statistical concepts continuing on the foundation of basic statistics built from the first part of this primer. This advanced primer aims to delve deeper into essential statistical concepts beyond the basics, equipping the reader with the knowledge to effectively analyze complex data sets, explore correlations and causality, employ regression analysis techniques, interpret survival curves, and evaluate diagnostic tests rigorously. It primarily focuses on the statistical tests used to analyze the relationship between groups of variables (the statistical tests to analyze the difference between groups of variables was discussed in the part 1 of this series). Toward the end of the article concepts of survival curves and methods for assessing the diagnostic accuracy of tests are stressed upon.

Deep survival analysis using pseudo values and its application to predict the recurrence of stage IV colorectal cancer after tumor resection

An improved DeepSurv model is proposed for predicting the prognosis of colorectal cancer patients at stage IV. Our model, called as PseudoDeepSurv, is optimized by a novel loss function, which is the combination of the average negative log partial likelihood and the mean-squared error derived from the pseudo-observations approach. The public BioStudies dataset including 999 patients was utilized for performance evaluation. Our PseudoDeepSurv model produced a C-index of 0.684 and 0.633 on the training and testing dataset, respectively. While for the original DeepSurv model, the corresponding values are 0.671 and 0.618, respectively.

The PDZ domain of the E protein in SARS-CoV induces carcinogenesis and poor prognosis in LUAD

BACKGROUND

In both lung adenocarcinoma (LUAD) and severe acute respiratory syndrome (SARS), uncontrolled inflammation can be detected in lung tissue. The PDZ-binding motif (PBM) in the SARS-CoV-1 E protein has been demonstrated to be a virulence factor that induces a cytokine storm.

METHODS

To identify gene expression fluctuations induced by PBM, microarray sequencing data of lung tissue infected with wild-type (SARS-CoV-1-E-wt) or recombinant virus (SARS-CoV-1-E-mutPBM) were analyzed, followed by functional enrichment analysis. To understand the role of the screened genes in LUAD, overall survival and immune correlation were calculated.

RESULTS

A total of 12 genes might participate in the initial and developmental stages of LUAD through expression variation and mutation. Moreover, dysregulation of a total of 12 genes could lead to a poorer prognosis. In addition, the downregulation of MAMDC2 and ITGA8 by PBM could also affect patient prognosis. Although the conserved PBM (-D-L-L-V-) can be found at the end of the carboxyl terminus in multiple E proteins of coronaviruses, the specific function of each protein depends on the entire amino acid sequence.

CONCLUSIONS

In summary, PBM containing the SARS-CoV-1 E protein promoted the carcinogenesis of LUAD by dysregulating important gene expression profiles and subsequently influencing the immune response and overall prognosis.

Associations between fruit and vegetable intakes and incident depression in middle-aged and older adults from 10 diverse international longitudinal cohorts

BACKGROUND

Emerging observational evidence supports a role for higher fruit and vegetable intake in protecting against the development of depression. However, there is a scarcity of research in older adults or in low- to middle-income countries (LMICs).

METHODS

Participants were 7801 community-based adults (mean age 68.6 ± 8.0 years, 55.8 % female) without depression, from 10 diverse cohorts, including four cohorts from LMICs. Fruit and vegetable intake was self-reported via comprehensive food frequency questionnaire, short food questionnaire or diet history. Depressive symptoms were assessed using validated measures, and depression defined applying validated cut-offs. The associations between baseline fruit and vegetable intakes and incident depression over a follow-up period of three to nine years were examined using Cox regression. Analyses were performed by cohort with results meta-analysed.

RESULTS

There were 1630 cases of incident depression (21 % of participants) over 40,258 person-years of follow-up. Higher intake of fruit was associated with a lower risk of incident depression (HR 0.87, 95%CI [0.77, 0.99], I2 = 4 %). No association was found between vegetable intake and incident depression (HR 0.93, 95%CI [0.84, 1.04], I2 = 0 %).

LIMITATIONS

Diverse measures used across the different cohorts and the modest sample size of our study compared with prior studies may have prevented an association being detected for vegetable intake.

CONCLUSIONS

Our study supports a role for fruit, but not vegetable intake in protecting against depression. Research investigating different types of fruits and vegetables using standardised measures in larger cohorts of older adults from low- and middle-income countries is warranted.

18F-Fluorodeoxyglucose Positron Emission Tomography-Based Risk Score Model for Prediction of Five-Year Survival Outcome after Curative Resection of Non-Small-Cell Lung Cancer

The aim of our retrospective study is to develop and assess an imaging-based model utilizing 18F-FDG PET parameters for predicting the five-year survival in non-small-cell lung cancer (NSCLC) patients after curative surgery. A total of 361 NSCLC patients who underwent curative surgery were assigned to the training set (n = 253) and the test set (n = 108). The LASSO regression model was used to construct a PET-based risk score for predicting five-year survival. A hybrid model that combined the PET-based risk score and clinical variables was developed using multivariate logistic regression analysis. The predictive performance was determined by the area under the curve (AUC). The individual features with the best predictive performances were co-occurrence_contrast (AUC = 0.675) and SUL peak (AUC = 0.671). The PET-based risk score was identified as an independent predictor after adjusting for clinical variables (OR 5.231, 95% CI 1.987-6.932; p = 0.009). The hybrid model, which integrated clinical variables, significantly outperformed the PET-based risk score alone in predictive accuracy (AUC = 0.771 vs. 0.696, p = 0.022), a finding that was consistent in the test set. The PET-based risk score, especially when integrated with clinical variables, demonstrates good predictive ability for five-year survival in NSCLC patients following curative surgery.

T cell-mediated tumor killing sensitivity gene signature-based prognostic score for acute myeloid leukemia

BACKGROUND AND OBJECTIVE

Acute myeloid leukemia (AML) is an aggressive, heterogenous hematopoetic malignancies with poor long-term prognosis. T-cell mediated tumor killing plays a key role in tumor immunity. Here, we explored the prognostic performance and functional significance of a T-cell mediated tumor killing sensitivity gene (GSTTK)-based prognostic score (TTKPI).

METHODS

Publicly available transcriptomic data for AML were obtained from TCGA and NCBI-GEO. GSTTK were identified from the TISIDB database. Signature GSTTK for AML were identified by differential expression analysis, COX proportional hazards and LASSO regression analysis and a comprehensive TTKPI score was constructed. Prognostic performance of the TTKPI was examined using Kaplan-Meier survival analysis, Receiver operating curves, and nomogram analysis. Association of TTKPI with clinical phenotypes, tumor immune cell infiltration patterns, checkpoint expression patterns were analysed. Drug docking was used to identify important candidate drugs based on the TTKPI-component genes.

RESULTS

From 401 differentially expressed GSTTK in AML, 24 genes were identified as signature genes and used to construct the TTKPI score. High-TTKPI risk score predicted worse survival and good prognostic accuracy with AUC values ranging from 75 to 96%. Higher TTKPI scores were associated with older age and cancer stage, which showed improved prognostic performance when combined with TTKPI. High TTKPI was associated with lower naïve CD4 T cell and follicular helper T cell infiltrates and higher M2 macrophages/monocyte infiltration. Distinct patterns of immune checkpoint expression corresponded with TTKPI score groups. Three agents; DB11791 (Capmatinib), DB12886 (GSK-1521498) and DB14773 (Lifirafenib) were identified as candidates for AML.

CONCLUSION

A T-cell mediated killing sensitivity gene-based prognostic score TTKPI showed good accuracy in predicting survival in AML. TTKPI corresponded to functional and immunological features of the tumor microenvironment including checkpoint expression patterns and should be investigated for precision medicine approaches.

Integrated analysis of single-cell sequencing and weighted co-expression network identifies a novel signature based on cellular senescence-related genes to predict prognosis in glioblastoma

BACKGROUND

Glioblastoma (GBM) is a highly aggressive cancer with heavy mortality rates and poor prognosis. Cellular senescence exerts a pivotal influence on the development and progression of various cancers. However, the underlying effect of cellular senescence on the outcomes of patients with GBM remains to be elucidated.

METHODS

Transcriptome RNA sequencing data with clinical information and single-cell sequencing data of GBM cases were obtained from CGGA, TCGA, and GEO (GSE84465) databases respectively. Single-sample gene set enrichment analysis (ssGSEA) analysis was utilized to calculate the cellular senescence score. WGCNA analysis was employed to ascertain the key gene modules and identify differentially expressed genes (DEGs) associated with the cellular senescence score in GBM. The prognostic senescence-related risk model was developed by least absolute shrinkage and selection operator (LASSO) regression analyses. The immune infiltration level was calculated by microenvironment cell populations counter (MCPcounter), ssGSEA, and xCell algorithms. Potential anti-cancer small molecular compounds of GBM were estimated by "oncoPredict" R package.

RESULTS

A total of 150 DEGs were selected from the pink module through WGCNA analysis. The risk-scoring model was constructed based on 5 cell senescence-associated genes (CCDC151, DRC1, C2orf73, CCDC13, and WDR63). Patients in low-risk group had a better prognostic value compared to those in high-risk group. The nomogram exhibited excellent predictive performance in assessing the survival outcomes of patients with GBM. Top 30 potential anti-cancer small molecular compounds with higher drug sensitivity scores were predicted.

CONCLUSION

Cellular senescence-related genes and clusters in GBM have the potential to provide valuable insights in prognosis and guide clinical decisions.

DoSurvive: A webtool for investigating the prognostic power of a single or combined cancer biomarker

We present DoSurvive, a user-friendly survival analysis web tool and a cancer prognostic biomarker centered database. DoSurvive is the first database that allows users to perform multivariant survival analysis for cancers with customized gene/patient list. DoSurvive offers three survival analysis methods, Log rank test, Cox regression and accelerated failure time model (AFT), for users to analyze five types of quantitative features (mRNA, miRNA, lncRNA, protein and methylation of CpG islands) with four survival types, i.e. overall survival, disease-specific survival, disease-free interval, and progression-free interval, in 33 cancer types. Notably, the implemented AFT model provides an alternative method for genes/features which failed the proportional hazard assumption in Cox regression. With the unprecedented number of survival models implemented and high flexibility in analysis, DoSurvive is a unique platform for the identification of clinically relevant targets for cancer researcher and practitioners. DoSurvive is freely available at http://dosurvive.lab.nycu.edu.tw/.

Exploring Prognosis, Tumor Microenvironment and Tumor Immune Infiltration in Hepatocellular Carcinoma Based on ATF/CREB Transcription Factor Family Gene-Related Model

Introduction

Hepatocellular carcinoma (HCC) is the most common form of primary liver cancer. It is the fourth leading cause of cancer-related death worldwide. Deregulation of the ATF/CREB family is associated with the progression of metabolic homeostasis and cancer. Because the liver plays a central role in metabolic homeostasis, it is critical to assess the predictive value of the ATF/CREB family in the diagnosis and prognosis of HCC.

Methods

Using data from The Cancer Genome Atlas (TCGA), this research evaluated the expression, copy number variations, and frequency of somatic mutations of 21 genes in the ATF/CREB family in HCC. A prognostic model based on the ATF/CREB gene family was developed via Lasso and Cox regression analyses, with the TCGA cohort serving as the training dataset and the International Cancer Genome Consortium (ICGC) cohort serving as the validation set. Kaplan-Meier and receiver operating characteristic analyses verified the accuracy of the prognostic model. Furthermore, the association among the prognostic model, immune checkpoints, and immune cells was examined.

Results

High-risk patients exhibited an unfavorable outcome as opposed to those in the low-risk category. Multivariate Cox analysis revealed that the risk score calculated based on the prognostic model was an independent prognostic factor for HCC. Analysis of immune mechanisms revealed that the risk score had a positive link to the expression of immune checkpoints, particularly CD274, PDCD1, LAG3, and CTLA4. Differences in immune cells and immune-associated roles were found between the high- and low-risk patients, as determined by single-sample gene set enrichment analysis. The core genes ATF1, CREB1, and CREB3 in the prognostic model were shown to be upregulated in HCC tissues as opposed to adjoining normal tissues, and the 10-year overall survival (OS) rate was worse among patients with elevated expression levels of ATF1, CREB1, and CREB3. Elevated expression levels of ATF1, CREB1, and CREB3 in HCC tissues were confirmed by qRT-PCR and immunohistochemistry studies.

Conclusion

According to the results of our training set and test set, the risk model based on the six ATF/CREB gene signatures predicting prognosis has certain predictive accuracy in predicting the survival of HCC patients. This study provides novel insights into the individualized treatment of patients with HCC.

A nomogram model to predict prognosis of patients with hepatoblastoma

BACKGROUND

Hepatoblastomas (HBs) are malignant liver tumors that most commonly develop in pediatric patients. Microvascular invasion may be a prognosis factor for patients with HBs. This study aimed to construct a model to predict the survival outcome in HBs.

METHODS

We retrospectively analyzed the clinical data of 311 patients with HBs who underwent surgical resection at our institution between June 2014 and August 2021. First, patients were divided into two groups: those who had pathologic microvascular invasion (n = 146) and those who did not (n = 165). Propensity score-matched (PSM) analysis was carried out between the two groups. The preoperative parameters and overall survival (OS) rate were compared between the two groups. Second, all 311 patients were randomly divided into the training and validation cohort in a ratio of 4:1. A nomogram was created in the training cohort to visualize the prediction of OS. Moreover, the validation cohort was used for validation.

RESULTS

Multivariate analysis suggested that age, histology type, microvascular invasion, multifocality, distant metastasis, and macrovascular involvement are independent prognostic factors for HBs. The nomogram showed good predictive ability in the training and validation cohorts with a C-index of 0.878 (95% CI, 0.831-0.925) and 0.847 (95% CI, 0.757-0.937), respectively. The calibration curve indicated good agreement between the prediction and observation for one-, two-, and three-year OS probabilities.

CONCLUSION

By combining preoperative imaging results and other clinical data, we established a nomogram to predict OS probability for patients with HB, which could be a potential tool to guide personalized treatment.