Background
Phase III, open-label RCT (LNH-98.5; GELA). 399 patients aged 60–80 years with previously untreated DLBCL. The study established whether adding rituximab (anti-CD20 monoclonal antibody) to standard CHOP could improve outcomes in elderly DLBCL, a population with historically dismal prognosis on chemotherapy alone.
Interventions and follow up
Arm A: R-CHOP — rituximab 375 mg/m² day 1 + cyclophosphamide 750 mg/m² + doxorubicin 50 mg/m² + vincristine 1.4 mg/m² + prednisone 40 mg/m² days 1–5, q21d × 8 cycles (n=202)
Arm B: CHOP — cyclophosphamide, doxorubicin, vincristine, prednisone q21d × 8 cycles (n=197)
Primary endpoint: Complete response rate; secondary: EFS, PFS, OS
Median follow up: 10 years (Coiffier et al, JCO 2010)
Arm B: CHOP — cyclophosphamide, doxorubicin, vincristine, prednisone q21d × 8 cycles (n=197)
Primary endpoint: Complete response rate; secondary: EFS, PFS, OS
Median follow up: 10 years (Coiffier et al, JCO 2010)
Results
CR rate (R-CHOP vs. CHOP): 76% vs. 63%, P=.005
10-yr OS: 43.5% vs. 27.6%; absolute benefit ~16 pp
10-yr PFS: 36.5% vs. 20.1%
10-yr EFS: 33.8% vs. 19.8%
10-yr OS: 43.5% vs. 27.6%; absolute benefit ~16 pp
10-yr PFS: 36.5% vs. 20.1%
10-yr EFS: 33.8% vs. 19.8%
Adverse events
Cardiac: 6% (R-CHOP) vs. 9% (CHOP)
Infectious/hematologic (grade ≥3): infections 19% vs. 18%; neutropenia comparable between arms; infusion reactions (rituximab) 13% grade 1–2, no grade 4
Other: treatment-related deaths 3% both arms
Infectious/hematologic (grade ≥3): infections 19% vs. 18%; neutropenia comparable between arms; infusion reactions (rituximab) 13% grade 1–2, no grade 4
Other: treatment-related deaths 3% both arms
Conclusions
R-CHOP significantly improved CR rate, EFS, PFS, and OS compared to CHOP in elderly DLBCL, with 10-year OS improved from 27.6% to 43.5%. This trial established R-CHOP as the global standard of care for DLBCL — the most impactful advance in DLBCL treatment of the modern era.
Key Limitations
Single age group (60–80 years), limiting extrapolation to younger or very elderly patients. CHOP alone is no longer used, making the comparator outdated. 8-cycle R-CHOP became standard here, but RICOVER-60 later showed 6 cycles suffice in older patients. COO/cell-of-origin profiling was not performed — GCB vs. non-GCB outcomes were not distinguished. No PET staging used; modern staging and response standards differ. Cure is achieved in ~40% of elderly patients, with relapse risk persistent beyond 5 years.
Clinical Context
LNH-98.5 is the foundational trial establishing R-CHOP as the universal standard for frontline DLBCL worldwide. It drove FDA approval of rituximab for untreated DLBCL in 2006 and is the baseline comparator in virtually every subsequent frontline DLBCL trial. The 10-year follow-up (JCO 2010) confirmed durability of the OS benefit. Subsequent efforts to improve upon R-CHOP (GOYA, ROBUST, PHOENIX) largely failed in unselected populations, with the first successful improvement arriving with POLARIX (Pola-R-CHP, 2022).