Background
15-year follow-up analysis of SWOG S0016, a phase III multicenter randomized trial. 531 patients with previously untreated advanced-stage follicular lymphoma enrolled 2001–2008 across US academic and community sites. This analysis assessed long-term outcomes and applied cure modeling to determine whether functional cure is achievable with CHOP-based chemoimmunotherapy. Median follow-up 15.5 years.
Interventions and follow up
Arm A: R-CHOP (rituximab + cyclophosphamide, hydroxydaunorubicin, vincristine, prednisone), 6 cycles
Arm B: CHOP followed by iodine-131 tositumomab radioimmunotherapy (CHOP-RIT)
Primary endpoint: PFS and OS (long-term)
mFollow up: 15.5 years
Arm B: CHOP followed by iodine-131 tositumomab radioimmunotherapy (CHOP-RIT)
Primary endpoint: PFS and OS (long-term)
mFollow up: 15.5 years
Results
15-yr OS: 70% overall, no significant difference between arms
15-yr PFS: 40% overall; CHOP-RIT 47% vs R-CHOP 34%, P=.004
Cure rate (modeling): 42% overall — highest in low-FLIPI, normal beta-2 microglobulin
Annual relapse rate: 6.8% (years 0–5) declining to 0.6% (years 15–20)
15-yr PFS: 40% overall; CHOP-RIT 47% vs R-CHOP 34%, P=.004
Cure rate (modeling): 42% overall — highest in low-FLIPI, normal beta-2 microglobulin
Annual relapse rate: 6.8% (years 0–5) declining to 0.6% (years 15–20)
Adverse events
Late toxicity: Long-term follow-up identified late toxicities from radioimmunotherapy in the CHOP-RIT arm, including secondary malignancies
New events: No new unexpected late events reported with extended follow-up
New events: No new unexpected late events reported with extended follow-up
Conclusions
Approximately 42% of patients with advanced-stage FL treated with CHOP-based chemoimmunotherapy achieved functional cure, with relapse rates declining markedly after year 5. This represents a paradigm shift: FL should no longer be universally characterized as incurable.
Key Limitations
CHOP-RIT used tositumomab (radioimmunotherapy), which is no longer commercially available in the US, limiting comparability to current regimens. R-CHOP without maintenance rituximab was used — most current protocols include rituximab maintenance, which likely improves cure rates. Cure modeling is a statistical framework with inherent assumptions. Contemporary therapies (obinutuzumab, lenalidomide, PI3K inhibitors, CAR-T) were not available.
Clinical Context
These 15-year data establish that a substantial minority of FL patients achieve durable remission indistinguishable from cure with chemoimmunotherapy. Low-risk patients (low FLIPI, normal beta-2M) have the highest cure rates, supporting an individualized approach to treatment intensity and expectation-setting. The results inform patient counseling and support continued development of curative strategies.