P1139: comparison of 6xr-chop21 versus 6xr-chop+2r for advanced-stage diffuse large b-cell lymphoma: a propensity score weighted population-based analysis

Topic: 19. Aggressive Non-Hodgkin lymphoma - Clinical Background: The standard first-line treatment for most patients with advanced-stage diffuse large B-cell lymphoma (DLBCL) is six cycles of rituximab, cyclophosphamide, doxorubicin, vincristine, and prednisone administered every 21 days (6xR-CHOP21). Until recently, Dutch treatment guidelines for DLBCL allowed for either 6xR-CHOP21 or 6xR-CHOP21 with two additional rituximab administrations (6xR-CHOP21 + 2R). As of 2021, based on the PETAL study, interim PET-guided treatment in DLBCL is recommended in the Netherlands, advising 6xR-CHOP21 + 2R only in the case of a positive interim PET. Aims: Since there is no randomized comparison between 6xR-CHOP21 and 6xR-CHOP21 + 2R in a pre-interim PET era, we aimed to assess the comparative effectiveness of these two treatment options in patients diagnosed with advanced-stage DLBCL in a pre-interim PET era in the Netherlands. Methods: From the nationwide, population-based Netherlands Cancer Registry, we identified adult (≥18 years) patients diagnosed with advanced-stage DLBCL (disease stages II-IV) in the Netherlands between 2014 and 2018 who completed 6xR-CHOP21±2R. Of note, the 6xR-CHOP21 group only included patients who did not die within 42 days after treatment completion to make treatment duration comparable. Event-free survival (EFS) was measured from the end of treatment (EOT) until progression, relapse, the start of second-line treatment, or death. Overall survival (OS) was measured from EOT until death. Since this study involves a non-randomized comparison using observational data, each patient has a different probability of receiving treatment (i.e., propensity score). Therefore, we weighted patients using stabilized inverse propensity scores to balance the patient characteristics between the treatment groups. The propensity score model was a logistic regression explaining treatment assignment according to the following characteristics: sex, the individual parameters of the International Prognostic Index (IPI), prior malignancy diagnosis, region of treatment, treatment at an academic center, and socioeconomic status. The difference in EFS and OS between the two treatment groups was calculated by Kaplan-Meier curves, log-rank tests, Cox proportional hazards models, and restricted mean survival time. We performed a stratified analysis to assess EFS and OS between the two treatment options across the four IPI risk groups. Results: Of the 1,626 included patients, most were male (55.9%), the median age was 70 years (interquartile range [IQR]; 63-76), and most were treated with 6xR-CHOP21 + 2R (54.7%). After weighing all patients according to their stabilized inverse propensity score, all characteristics were balanced between treatment groups (i.e., all standardized mean differences < 0.1). At a median follow-up of 4.5 years (IQR: 3.9-5.4 years), 6xR-CHOP21 + 2R was associated with better EFS (hazard ratio [HR], 0.75; 95% confidence interval [CI], 0.62-0.91; Panel A) and OS (HR, 0.74; 95% CI, 0.60-0.92; Panel B). Stratified analysis according to the IPI risk group revealed that this association was more pronounced with advancing IPI risk, particularly in patients with high-risk IPI (IPI score 4-5; Panel C-D). Summary/Conclusion: This propensity-weighted analysis using nationwide observational data covering a pre-interim PET era shows that 6xR-CHOP21 + 2R is associated with better EFS and OS than 6xR-CHOP21 in patients with advanced-stage DLBCL and a high-risk IPI score. Future population-based research is warranted to assess whether this association will be abrogated with PET-guided treatment intensification.Keywords: Population, Survival prediction, Treatment, DLBCL