Beyond linearity and static risk: re-evaluating core prognostic factors in diffuse large B-cell lymphoma.

[BACKGROUND] This study challenges the static and linear assumptions of traditional prognostic models for diffuse large B-cell lymphoma (DLBCL), such as the International Prognostic Index (IPI).

[RESEARCH DESIGN AND METHODS] Analyzing 664 DLBCL patients treated with R-CHOP, the research explored the time-dependent and nonlinear effects of established risk factors.

[RESULTS] While the IPI maintained strong overall prognostic stability, its individual components exhibited dynamic behavior. ECOG performance score, beta-2 microglobulin (β2 M), and lactate dehydrogenase (LDH) were key predictors of early mortality, with their influence diminishing over time. Conversely, the importance of Ann Arbor stage and extranodal involvement grew, identifying them as markers of later risk. Notably, the study found a resurgence in the predictive power of β2 M after two years. Furthermore, restricted cubic spline modeling revealed significant nonlinear relationships between overall survival and age, LDH, and β2 M (all  < 0.001 for nonlinearity).

[CONCLUSIONS] The prognostic impact of baseline factors in DLBCL follows dynamic and nonlinear trajectories, challenging the one-size-fits-all approach of current risk scores. Future models should incorporate these temporal dynamics to provide a more accurate, personalized risk assessment.