Rituximab Maintenance Treatment For a Maximum Of 5 Years In Follicular Lymphoma: Results Of The Randomized Phase III Trial SAKK 35/03

Abstract

BackgroundRituximab maintenance has been shown to be effective in patients with follicular lymphoma. The optimal duration of maintenance treatment remains unknown. Methods270 patients (median age 57 years: range 25-82) with either untreated, relapsed, stable or chemotherapy resistant follicular lymphoma of all grades were treated with 4 weekly doses of rituximab monotherapy (375 mg/m²). Patients responding to the induction treatment (partial or complete response) received rituximab (375 mg/ m²) maintenance and were randomized either to a short maintenance consisting of four administrations every two months (arm A), or a prolonged maintenance for a maximum of five years or until disease progression or unacceptable toxicity (arm B). The primary endpoint was event-free survival (EFS) from randomization. Sample size calculation allowed detecting a median EFS increase with prolonged maintenance from 2.5 to 4.5 years with 80% power. Secondary endpoints included progression-free survival (PFS), overall survival (OS) and objective response. Comparisons between the two treatment arms were performed using the log-rank test for survival endpoints. ResultsFrom October 2004 to November 2007 165 patients were randomized, 82 in arm A and 83 in arm B. 124 patients were chemotherapy-naïve. The median EFS is 3.4 years (95% CI 2.1-5.3) in arm A and 5.3 years (95% CI 3.5-NA) in arm B. Using the prespecified log-rank test this difference is statistically not significant (p=0.14). We observed an unexplained difference in disease progression and relapse during the first 8 months after randomization (3 in arm A vs. 10 in arm B) when treatment in both arms was the same which led to an early crossing of the EFS curves at 18 months. Considering only patients at risk after 8 months from randomization EFS in arm B is significantly prolonged compared to arm A (median EFS 7.1 (95% CI 4.4-NA) vs. 2.9 years (95% CI 1.8-4.8); log-rank test p=0.004). Median PFS is significantly longer in arm B (7.4 (95% CI 5.1-NA) vs. 3.5 years (95% CI 2.1-5.9); log-rank test p=0.04). There is no significant difference in OS and in the observed best response. Maintenance treatment was stopped due to unacceptable toxicity in no patient in arm A vs. 3 in arm B. Six subsequent cancers developed in arm A and 8 in arm B. One infection grade ≥3 was reported in arm A whereas seven infections grade ≥3 occurred in 5 patients in arm B. ConclusionsEFS, the primary endpoint was not met which is mainly due to the early separation of the survival curves favouring arm A, at a time when the treatment in both arms was the same. However, a retrospectively defined analysis considering only EFS events from the time when treatment was different in the two arms, shows a statistically significant increase in EFS with long-term maintenance compared to the 8 months maintenance regimen. Long-term rituximab maintenance also doubles the median PFS without leading to increased undue toxicity. Disclosures:Off Label Use: Rituximab for 5 years. Ghielmini:Roche: Consultancy, Honoraria, Research Funding, Speakers Bureau.