LONG-TERM OUTCOMES, SURVIVAL PREDICTORS AND UP-FRONT TREATMENT PATTERNS IN 568 ELDERLY DIFFUSE LARGE B-CELL LYMPHOMA PATIENTS FROM LATAM: IS R-CHOP STILL THE FIRST CHOICE FOR TREATING PATIENTS OLDER THAN 60 YEARS?

Diffuse large B-cell lymphoma (DLBCL), the most common non-Hodgkin lymphoma subtype, mainly affects those over 60. Elderly patients show high-risk features, EBV infection, chemoresistance, lower drug tolerability, and poorer outcomes. Excluding new chemo-free treatments, cornerstone of elderly DLBCL is immunochemotherapy based on anthracyclines. Therefore, in addition to full-dose R-CHOP regimen, attenuated protocols, such as R-miniCHOP (containing doxorubicin 25/sqm) and Elderly-R-miniCHOP (rituximab 375 mg/sqm, cyclophosphamide 400 mg/sqm, doxorubicin 25 mg/sqm, vincristine 1 mg fixed dose, and prednisone 40 mg/sqm) have emerged for this particularly fragile population. This study aims to assess long-term outcomes, determine survival predictors, and compare responses and toxicities between different R-CHOP-like protocols in a large real-world cohort. Retrospective, observational, and single-center study involving 568 DLBCL, NOS patients aged ≥ 60 years, treated at USP, Brazil, between January 2009 and December 2023. The median age at diagnosis was 70 years (60-97) and 52.3% were female. Comorbidities were prevalent, including 27.8% of immobility, 28.1% of malnutrition and 25% of polypharmacy. Advanced clinical stage (III/IV) was observed in 79% of cases, 50.4% had bulk disease ≥ 7 cm and 70.3% had IPI ≥ 3. Among 518 (91.1%) cases effectively treated, 66.2% received full-dose R-CHOP, 18.7% R-miniCHOP, 11.4% Elderly-R-miniCHOP, and 3.7% anthracycline-free protocols. ORR for the whole cohort was 90.1% (95% CI: 86.9-92.6%), with 79.9% (95% CI: 75.8-83.5%) achieving CR. ORR was 92.4% for R-CHOP, 89% for R-miniCHOP, and 75.7% for Elderly-R-miniCHOP, p = 0.011. Overall mortality (p = 0.010) and induction mortality (p = 0.002) rates were significantly lower in patients treated with full-dose R-CHOP. With a median follow-up of 72.3 months, estimated 5-year OS and EFS were 51.3% (95% CI: 47.1-56%) and 47.5% (95% CI: 43.3-52.2%), respectively. The estimated 5-year OS and EFS were 61.7% (95% CI: 56.5-67.5%) and 57.7% (95% CI: 52.4-63.5%) for full-dose R-CHOP, 47.2% (95% CI: 37.1-59.9%) and 42.1% (95% CI: 32.4-54.8%) for R-miniCHOP, and 36.1% (95% CI: 24.1-54%) and 33.7% (95% CI: 22.5-50.4%) for Elderly-R-miniCHOP, p < 0.0001 for both outcomes. Additionally, very-elderly (≥80 years) and frail (KPS<50) patients had markedly decreased OS and EFS compared to younger groups (60-79 years) and fit cases (KPS 50-70 and >70), p < 0.0001. Despite having promoted a better outcome, the full- dose R-CHOP regimen was associated with higher rates of neutropenia (p < 0.0001), not translated into an increase in treatment discontinuation rates (p = 0.155) or microbiologically documented infections (p = 0.063). In multivariate analysis age ≥ 70 years, congestive heart failure, KPS < 70, involvement ≥ 2 extranodal areas, Elderly-IPI int- high/high-risk, increased LDH and β2-microglobulin values were independent predictors for decreased survival. Although a significant proportion of DLBCL patients older than 60 years are considered ineligible for full-dose anthracycline-based regimens, attenuated immunochemotherapeutic protocols provided markedly inferior long-term outcomes in our large and long-term cohort. Therefore, we concluded that outside clinical trials settings, full-dose R-CHOP regimen is nowadays still considered as first-choice therapy for up-front management of elderly DLBCL, being associated with greater efficacy, better outcomes and a relatively safe adverse event profile.