Dataset Information

Weighted Cox regression for the prediction of heterogeneous patient subgroups.

ABSTRACT:

Background

An important task in clinical medicine is the construction of risk prediction models for specific subgroups of patients based on high-dimensional molecular measurements such as gene expression data. Major objectives in modeling high-dimensional data are good prediction performance and feature selection to find a subset of predictors that are truly associated with a clinical outcome such as a time-to-event endpoint. In clinical practice, this task is challenging since patient cohorts are typically small and can be heterogeneous with regard to their relationship between predictors and outcome. When data of several subgroups of patients with the same or similar disease are available, it is tempting to combine them to increase sample size, such as in multicenter studies. However, heterogeneity between subgroups can lead to biased results and subgroup-specific effects may remain undetected.

Methods

For this situation, we propose a penalized Cox regression model with a weighted version of the Cox partial likelihood that includes patients of all subgroups but assigns them individual weights based on their subgroup affiliation. The weights are estimated from the data such that patients who are likely to belong to the subgroup of interest obtain higher weights in the subgroup-specific model.

Results

Our proposed approach is evaluated through simulations and application to real lung cancer cohorts, and compared to existing approaches. Simulation results demonstrate that our proposed model is superior to standard approaches in terms of prediction performance and variable selection accuracy when the sample size is small.

Conclusions

The results suggest that sharing information between subgroups by incorporating appropriate weights into the likelihood can increase power to identify the prognostic covariates and improve risk prediction.

SUBMITTER: Madjar K

PROVIDER: S-EPMC8650299 | biostudies-literature |

REPOSITORIES: biostudies-literature

ACCESS DATA

Similar Datasets

Project description:Purpose:This study focused on identification of long non-coding RNAs (lncRNAs) for prognosis prediction of glioblastoma (GBM) through weighted gene co-expression network analysis (WGCNA) and L1-penalized least absolute shrinkage and selection operator (LASSO) Cox proportional hazards (PH) model. Materials and methods:WGCNA was performed based on RNA expression profiles of GBM from Chinese Glioma Genome Atlas (CGGA), National Center for Biotechnology Information (NCBI) Gene Expression Omnibus (GEO), and the European Bioinformatics Institute ArrayExpress for the identification of GBM-related modules. Subsequently, prognostic lncRNAs were determined using LASSO Cox PH model, followed by constructing a risk scoring model based on these lncRNAs. The risk score was used to divide patients into high- and low-risk groups. Difference in survival between groups was analyzed using Kaplan-Meier survival analysis. IncRNA-mRNA networks were built for the prognostic lncRNAs, followed by pathway enrichment analysis for these networks. Results:This study identified eight preserved GBM-related modules, including 188 lncRNAs. Consequently, C20orf166-AS1, LINC00645, LBX2-AS1, LINC00565, LINC00641, and PRRT3-AS1 were identified by LASSO Cox PH model. A risk scoring model based on the lncRNAs was constructed that could divide patients into different risk groups with significantly different survival rates. Prognostic value of this six-lncRNA signature was validated in two independent sets. C20orf166-AS1 was associated with antigen processing and presentation and cell adhesion molecule pathways, involving nine common genes. LBX2-AS1, LINC00641, PRRT3-AS1, and LINC00565 were related to focal adhesion, extracellular matrix receptor interaction, and mitogen-activated protein kinase signaling pathways, which shared 12 common genes. Conclusion:This prognostic six-lncRNA signature may improve prognosis prediction of GBM. This study reveals many pathways and genes involved in the mechanisms behind these lncRNAs.

Project description:BackgroundClinical prediction rules (CPRs) should be externally validated by independent researchers. Although there are many cardiovascular CPRs, most have not been externally validated. It is not known why some CPRs are externally validated by independent researchers and others are not.MethodsWe analyzed cardiovascular risk CPRs included in a systematic review. Independent external validations were identified by forward citation searches of derivation studies. Time between the publication of a cardiovascular CPR and the first independent external validation was calculated. We assessed Kaplan-Meier estimates of the probability to have an independent external validation. Using univariable Cox regression, we explored whether characteristics of derivation (design, location, sample size, number of predictors, presentation format, validation in derivation), reporting (participants, predictors, outcomes, performance measure, information for risk calculation), and publication (journal impact factor) are associated with time to the first independent external validation.ResultsOf 125 cardiovascular risk CPRs, 29 had an independent external validation. The median follow-up was 118 months (95% CI, 99-130). The 25th percentile of event time was 122 months (95% CI, 91-299). Cardiovascular risk CPRs from the USA were 4.15 times (95% CI, 1.89-9.13) more likely to have an independent external validation. Increasing the sample size of derivation by ten times was associated with a 2.32-fold (95% CI, 1.37-3.91) increase in the probability of having an independent external validation. CPRs presented with an internal validation tend to get an independent external validation sooner (HR = 1.73, 95% CI, 0.77-3.93). CPRs reporting all the information necessary for calculating individual risk were 2.65 (95% CI, 1.01-6.96) times more likely to have an independent external validation. Publishing a cardiovascular risk CPR in a journal that has one unit higher impact factor was associated with a 6% (95% CI, 3-9) higher likelihood of an independent external validation.ConclusionsThe probability for cardiovascular risk CPRs to get an independent external validation was low even many years after their derivations. Authors of new cardiovascular risk CPRs should consider using adequate sample size, conducting an internal validation, and reporting all the information needed for risk calculation as these features were associated with an independent external validation.

Project description:BackgroundDNA microarray studies identified gene expression signatures predictive of metastatic relapse in early breast cancer. Standard feature selection procedures applied to reduce the set of predictive genes did not take into account the correlation between genes. In this paper, we studied the performances of three high-dimensional regression methods - CoxBoost, LASSO (Least Absolute Shrinkage and Selection Operator), and Elastic net - to identify prognostic signatures in patients with early breast cancer.MethodsWe analyzed three public retrospective datasets, including a total of 384 patients with axillary lymph node-negative breast cancer. The Amsterdam van't Veer's training set of 78 patients was used to determine the optimal gene sets and classifiers using sensitivity thresholds resulting in mis-classification of no more than 10% of the poor-prognosis group. To ensure the comparability between different methods, an automatic selection procedure was used to determine the number of genes included in each model. The van de Vijver's and Desmedt's datasets were used as validation sets to evaluate separately the prognostic performances of our classifiers. The results were compared to the original Amsterdam 70-gene classifier.ResultsThe automatic selection procedure reduced the number of predictive genes up to a minimum of six genes. In the two validation sets, the three models (Elastic net, LASSO, and CoxBoost) led to the definition of genomic classifiers predicting the 5-year metastatic status with similar performances, with respective 59, 56, and 54% accuracy, 83, 75, and 83% sensitivity, and 53, 52, and 48% specificity in the Desmedt's dataset. In comparison, the Amsterdam 70-gene signature showed 45% accuracy, 97% sensitivity, and 34% specificity. The gene overlap and the classification concordance between the three classifiers were high. All the classifiers added significant prognostic information to that provided by the traditional prognostic factors and showed a very high overlap with respect to gene ontologies (GOs) associated with genes overexpressed in the predicted poor-prognosis vs. good-prognosis classes and centred on cell proliferation. Interestingly, all classifiers reported high sensitivity to predict the 4-year status of metastatic disease.ConclusionsHigh-dimensional regression methods are attractive in prognostic studies because finding a small subset of genes may facilitate the transfer to the clinic, and also because they strengthen the robustness of the model by limiting the selection of false-positive predictive genes. With only six genes, the CoxBoost classifier predicted the 4-year status of metastatic disease with 93% sensitivity. Selecting a few genes related to ontologies other than cell proliferation might further improve the overall sensitivity performance.