Incremental value in outcome prediction with gene expression-based signatures in diffuse large B-cell lymphoma

Area of Science:

• Oncology
• Genomics
• Biostatistics

Background:

• Gene expression signatures are explored for predicting survival in diffuse large B-cell lymphoma (DLBCL).
• Current predictive models often rely on P values from multivariable Cox regression.
• The incremental value of existing gene signatures in DLBCL prognosis remains under-evaluated.

Purpose of the Study:

• To assess the added predictive value of three established gene expression signatures in DLBCL.
• To evaluate these signatures using advanced statistical methods for survival data.
• To compare the performance of gene signatures against clinical factors in DLBCL risk prediction.

Main Methods:

• Application of concordance measures for censored survival data.
• Analysis of two independent patient cohorts with DLBCL treated with CHOP or R-CHOP chemotherapy.
• Evaluation of three published gene expression-based signatures.

Main Results:

• Gene expression signatures demonstrated inferior predictive performance compared to established clinical factors.
• The incremental contribution of these signatures to survival risk assessment was minimal.
• Advanced statistical methods confirmed the limited added value of gene signatures.

Conclusions:

• Clinical factors are more critical than gene expression signatures for DLBCL survival prediction.
• Future DLBCL risk models require integrated approaches beyond gene expression alone.
• Gene expression studies in DLBCL still hold potential for understanding disease biology and identifying therapeutic targets.