Predictors of colorectal cancer survival using cox regression and random survival forests models based on gene expression data

A novel lipid metabolism-based risk model associated with immunosuppressive mechanisms in diffuse large B-cell lymphoma

BACKGROUND

The molecular diversity exhibited by diffuse large B-cell lymphoma (DLBCL) is a significant obstacle facing current precision therapies. However, scoring using the International Prognostic Index (IPI) is inadequate when fully predicting the development of DLBCL. Reprogramming lipid metabolism is crucial for DLBCL carcinogenesis and expansion, while a predictive approach derived from lipid metabolism-associated genes (LMAGs) has not yet been recognized for DLBCL.

METHODS

Gene expression profiles of DLBCL were generated using the Gene Expression Omnibus (GEO) and The Cancer Genome Atlas (TCGA) databases. The LASSO Cox regression was used to construct an effective predictive risk-scoring model for DLBCL patients. The Kaplan-Meier survival assessment was employed to compare a given risk score with the IPI score and its impact on the survival of DLBCL patients. Functional enrichment examination was performed utilizing the KEGG pathway. After identifying hub genes via single-sample GSEA (ssGSEA), immunohistochemical staining and immunofluorescence were performed on lymph node samples from control and DLBCL patients to confirm these identified genes.

RESULTS

Sixteen lipid metabolism- and survival-associated genes were identified to construct a prognostic risk-scoring approach. This model demonstrated robust performance over various datasets and emerged as an autonomous risk factor for predicting the development of DLBCL patients. The risk score could significantly distinguish the development of DLBCL patients from the low-risk and elevated-risk IPI classes. Results from the inhibitory immune-related pathways and lower immune scores suggested an immunosuppressive phenotype within the elevated-risk group. Three hub genes, MECR, ARSK, and RAN, were identified to be negatively correlated with activated CD8 T cells and natural killer T cells in the elevated-risk score class. Ultimately, it was determined that these three genes were expressed by lymphoma cells but not by T cells in clinical samples from DLBCL patients.

CONCLUSION

The risk level model derived from 16 lipid metabolism-associated genes represents a prognostic biomarker for DLBCL that is novel, robust, and may have an immunosuppressive role. It can compensate for the limitations of the IPI score in predicting overall survival and has potential clinical application value.

Use of survival support vector machine combined with random survival forest to predict the survival of nasopharyngeal carcinoma patients

Background

The Cox regression model is not sufficiently accurate to predict the survival prognosis of nasopharyngeal carcinoma (NPC) patients. It is impossible to calculate and rank the importance of impact factors due to the low predictive accuracy of the Cox regression model. So, we developed a system. Using the SEER (The Surveillance, Epidemiology, and End Results) database data on NPC patients, we proposed the use of random survival forest (RSF) and survival-support vector machine (SVM) from the machine learning methods to develop a survival prediction system specifically for NPC patients. This approach aimed to make up for the insufficiency of the Cox regression model. We also used the Cox regression model to validate the development of the nomogram and compared it with machine learning methods.

Methods

A total of 1,683 NPC patients were extracted from the SEER database from January 2010 to December 2015. We used R language for modeling work, established the nomogram of survival prognosis of NPC patients by Cox regression model, ranked the correlation of influencing factors by RSF model VIMP (variable important) method, developed a survival prognosis system for NPC patients based on survival-SVM, and used C-index for model evaluation and performance comparison.

Results

Although the Cox regression models can be developed to predict the prognosis of NPC patients, their accuracy was lower than that of machine learning methods. When we substituted the data for the Cox model, the C-index for the training set was only 0.740, and the C-index for the test set was 0.721. In contrast, the C index of the survival-SVM model was 0.785. The C-index of the RSF model was 0.729. The importance ranking of each variable could be obtained according to the VIMP method.

Conclusions

The prediction results from the Cox model are not as good as those of the RSF method and survival-SVM based on the machine learning method. For the survival prognosis of NPC patients, the machine learning method can be considered for clinical application.

Predicting Glaucoma Progression to Surgery with Artificial Intelligence Survival Models

Purpose

Prior artificial intelligence (AI) models for predicting glaucoma progression have used traditional classifiers that do not consider the longitudinal nature of patients' follow-up. In this study, we developed survival-based AI models for predicting glaucoma patients' progression to surgery, comparing performance of regression-, tree-, and deep learning-based approaches.

Design

Retrospective observational study.

Subjects

Patients with glaucoma seen at a single academic center from 2008 to 2020 identified from electronic health records (EHRs).

Methods

From the EHRs, we identified 361 baseline features, including demographics, eye examinations, diagnoses, and medications. We trained AI survival models to predict patients' progression to glaucoma surgery using the following: (1) a penalized Cox proportional hazards (CPH) model with principal component analysis (PCA); (2) random survival forests (RSFs); (3) gradient-boosting survival (GBS); and (4) a deep learning model (DeepSurv). The concordance index (C-index) and mean cumulative/dynamic area under the curve (mean AUC) were used to evaluate model performance on a held-out test set. Explainability was investigated using Shapley values for feature importance and visualization of model-predicted cumulative hazard curves for patients with different treatment trajectories.

Main Outcome Measures

Progression to glaucoma surgery.

Results

Of the 4512 patients with glaucoma, 748 underwent glaucoma surgery, with a median follow-up of 1038 days. The DeepSurv model performed best overall (C-index, 0.775; mean AUC, 0.802) among the models studied in this article (CPH with PCA: C-index, 0.745; mean AUC, 0.780; RSF: C-index, 0.766; mean AUC, 0.804; GBS: C-index, 0.764; mean AUC, 0.791). Predicted cumulative hazard curves demonstrate how models could distinguish between patient who underwent early surgery and patients who underwent surgery after > 3000 days of follow-up or no surgery.

Conclusions

Artificial intelligence survival models can predict progression to glaucoma surgery using structured data from EHRs. Tree-based and deep learning-based models performed better at predicting glaucoma progression to surgery than the CPH regression model, potentially because of their better suitability for high-dimensional data sets. Future work predicting ophthalmic outcomes should consider using tree-based and deep learning-based survival AI models. Additional research is needed to develop and evaluate more sophisticated deep learning survival models that can incorporate clinical notes or imaging.

Financial Disclosures

Proprietary or commercial disclosure may be found after the references.

Development and validation of a model for the prediction of disease-specific survival in patients with oral squamous cell carcinoma: based on random survival forest analysis

OBJECTIVE

To establish a model for predicting the disease-specific survival (DSS) of patients with oral squamous cell carcinoma (OSCC).

METHODS

Patients diagnosed with OSCC from the Surveillance, Epidemiology, and End Results (SEER) database were enrolled and randomly divided into development (n = 14,495) and internal validation cohort (n = 9625). Additionally, a cohort from a hospital located in Southeastern China was utilized for external validation (n = 582).

RESULTS

TNM stage, adjuvant treatment, surgery, tumor sites, age, grade, and gender were used for RSF model construction based on the development cohort. The effectiveness of the model was confirmed through time-dependent ROC curves in different cohorts. The risk score exhibited an almost exponential increase in the hazard ratio of death due to OSCC. In development, internal, and external validation cohorts, the prognosis was significantly worse for patients in groups with higher risk scores (all log-rank P < 0.05).

CONCLUSION

Based on RSF, a high-performance prediction model for OSCC prognosis was created and verified in this study.

Improved Survival Analyses Based on Characterized Time-Dependent Covariates to Predict Individual Chronic Kidney Disease Progression

Kidney diseases can cause severe morbidity, mortality, and health burden. Determining the risk factors associated with kidney damage and deterioration has become a priority for the prevention and treatment of kidney disease. This study followed 497 patients with stage 3-5 chronic kidney disease (CKD) who were treated at the ward of Taipei Veterans General Hospital from January 2006 to 2019 in Taiwan. The patients underwent 3-year-long follow-up sessions for clinical measurements, which occurred every 3 months. Three time-dependent survival models, namely the Cox proportional hazard model (Cox PHM), random survival forest (RSF), and an artificial neural network (ANN), were used to process patient demographics and laboratory data for predicting progression to renal failure, and important features for optimal prediction were evaluated. The individual prediction of CKD progression was validated using the Kaplan-Meier estimation method, based on patients' true outcomes during and beyond the study period. The results showed that the average concordance indexes for the cross-validation of the Cox PHM, ANN, and RSF models were 0.71, 0.72, and 0.89, respectively. RSF had the best predictive performances for CKD patients within the 3 years of follow-up sessions, with a sensitivity of 0.79 and specificity of 0.88. Creatinine, age, estimated glomerular filtration rate, and urine protein to creatinine ratio were useful factors for predicting the progression of CKD patients in the RSF model. These results may be helpful for instantaneous risk prediction at each follow-up session for CKD patients.

Recommendation for imaging follow-up strategy based on time-specific disease failure for nasopharyngeal carcinoma

BACKGROUND

To develop a common follow-up strategy for appropriate imaging examination at an appropriate time for nasopharyngeal carcinoma (NPC).

METHODS

Independent prognostic factors were identified by Cox regression analysis, and a nomogram model was developed. Random survival forest (RSF) model was constructed to depict probability of disease failure during a 5-year follow-up and establish a reasonable risk-based follow-up strategy.

RESULTS

The nomogram model finally categorized the patients into three risk groups. RSF model demonstrated distribution trends for local and regional recurrences, bone metastasis, liver metastasis, and lung metastasis of NPC. Adequate imaging at follow-up should be considered between 10 and 21 months for patients at moderate-risk of recurrence or metastasis and 7-36 months for those at high-risk.

CONCLUSIONS

The temporal distribution of incidence rates of recurrence or metastasis varied among different risk groups. We recommend implementing a focused and targeted imaging surveillance intervention at appropriate times to improve its efficiency and reduce costs.

Human Colonoid-Myofibroblast Coculture for Study of Apical Na+/H+ Exchangers of the Lower Cryptal Neck Region

Cation and anion transport in the colonocyte apical membrane is highly spatially organized along the cryptal axis. Because of lack of experimental accessibility, information about the functionality of ion transporters in the colonocyte apical membrane in the lower part of the crypt is scarce. The aim of this study was to establish an in vitro model of the colonic lower crypt compartment, which expresses the transit amplifying/progenitor (TA/PE) cells, with accessibility of the apical membrane for functional study of lower crypt-expressed Na⁺/H⁺ exchangers (NHEs). Colonic crypts and myofibroblasts were isolated from human transverse colonic biopsies, expanded as three-dimensional (3D) colonoids and myofibroblast monolayers, and characterized. Filter-grown colonic myofibroblast-colonic epithelial cell (CM-CE) cocultures (myofibroblasts on the bottom of the transwell and colonocytes on the filter) were established. The expression pattern for ion transport/junctional/stem cell markers of the CM-CE monolayers was compared with that of nondifferentiated (EM) and differentiated (DM) colonoid monolayers. Fluorometric pH_i measurements were performed to characterize apical NHEs. CM-CE cocultures displayed a rapid increase in transepithelial electrical resistance (TEER), paralleled by downregulation of claudin-2. They maintained proliferative activity and an expression pattern resembling TA/PE cells. The CM-CE monolayers displayed high apical Na⁺/H⁺ exchange activity, mediated to >80% by NHE2. Human colonoid-myofibroblast cocultures allow the study of ion transporters that are expressed in the apical membrane of the nondifferentiated colonocytes of the cryptal neck region. The NHE2 isoform is the predominant apical Na⁺/H⁺ exchanger in this epithelial compartment.

Deep learning survival model for colorectal cancer patients (DeepCRC) with Asian clinical data compared with different theories

INTRODUCTION

Colorectal cancer (CRC) is the third most common cancer. Precise prediction of CRC patients' overall survival (OS) probability could offer advice on its treatment. Neural network (NN) is the first-class algorithm, but a consensus on which NN survival models are better has not been established yet. A predictive model on CRC using Asian data is also lacking.

METHODS

We conducted 8 NN survival models of CRC (n = 416) with different theories and compared them using Asian data.

RESULTS

DeepSurv performed best with a C-index value of 0.8300 in the training cohort and 0.7681 in the test cohort.

CONCLUSIONS

The deep learning survival model for CRC patients (DeepCRC) could predict CRC's OS accurately.

Identification of a novel lipid metabolism-related gene signature for predicting colorectal cancer survival

Colorectal cancer (CRC) is a common malignant tumor worldwide. Lipid metabolism is a prerequisite for the growth, proliferation and invasion of cancer cells. However, the lipid metabolism-related gene signature and its underlying molecular mechanisms remain unclear. The aim of this study was to establish a lipid metabolism signature risk model for survival prediction in CRC and to investigate the effect of gene signature on the immune microenvironment. Lipid metabolism-mediated genes (LMGs) were obtained from the Molecular Signatures Database. The consensus molecular subtypes were established using "ConsensusClusterPlus" based on LMGs and the cancer genome atlas (TCGA) data. The risk model was established using univariate and multivariate Cox regression with TCGA database and independently validated in the international cancer genome consortium (ICGC) datasets. Immune infiltration in the risk model was developed using CIBERSORT and xCell analyses. A total of 267 differentially expressed genes (DEGs) were identified between subtype 1 and subtype 2 from consensus molecular subtypes, including 153 upregulated DEGs and 114 downregulated DEGs. 21 DEGs associated with overall survival (OS) were selected using univariate Cox regression analysis. Furthermore, a prognostic risk model was constructed using the risk coefficients and gene expression of eleven-gene signature. Patients with a high-risk score had poorer OS compared with patients in the low-risk score group (p = 3.36e-07) in the TCGA cohort and the validationdatasets (p = 4.03e-05). Analysis of immune infiltration identified multiple T cells were associated with better prognosis in the low-risk group, including Th2 cells (p = 0.0208), regulatory T cells (p = 0.0425), and gammadelta T cells (p = 0.0112). A nomogram integrating the risk model and clinical characteristics was further developed to predict the prognosis of patients with CRC. In conclusion, our study revealed that the expression of lipid-metabolism genes were correlated with the immune microenvironment. The eleven-gene signature might be useful for prediction the prognosis of CRC patients.