OBJECTIVES

This monocenter, observational secondary data use retrospective study aimed to describe the choice of first-line (1L) treatment of diffuse large B-cell lymphoma (DLBCL) in a real-world setting.

METHODS

Data of 213 newly diagnosed DLBCL patients who begun 1L at IRST between 2011 and 2017 were collected, grouping patients according to 1L treatment: 1) R-CHOP schema and 2) other schemas (OTH). Baseline characteristics, Overall Survival (OS), Progression free survival (PFS), Objective Response Rate (ORR) were presented; a propensity score inverse probability weighting was applied for adjusting for study groups imbalances.

RESULTS

52.6% of patients received R-CHOP and 47.4% OTH. R-CHOP patients were younger (62 yrs vs 78; p<0.00001) and 33.7% of them had International Prognostic Score (IPI) >2 (p=0.0002). R-CHOP patients had a significantly higher OS (p<0.0001). From Cox regression analysis higher age, ECOG≥2, Stage III-IV, extranodal sites ≥2 and IPI>2 resulted associated with an increased risk of death and receiving R-CHOP treatment was a protective factor (also confirmed in the multivariable model). A propensity score inverse probability weighting applied to the multivariable-Cox-model confirmed R-CHOP treatment protective effect (HR 0.41; 95%CI [0.28;0.59]). R-CHOP treatment delayed median PFS by more than 6 years (91.1 months vs 17.5). ORR were obtained in 96.3% of R-CHOP and in 88.9% of OTH patients (p=0.045).

CONCLUSIONS

This study confirmed in our experience the great efficacy of R-CHOP as 1L treatment, also after having applied a propensity score adjustment, to reduce imbalances between groups.