PS1262 POST HOC ANALYSIS OF THE AUGMENT PHASE III RANDOMIZED STUDY OF LENALIDOMIDE PLUS RITUXIMAB (R2) VS RITUXIMAB/PLACEBO IN PATIENTS WITH RELAPSED/REFRACTORY MARGINAL ZONE LYMPHOMA

Background: Lenalidomide plus rituximab (R 2 ) has established efficacy and a tolerable safety profile in patients with previously untreated and relapsed/refractory (R/R) indolent NHL. This includes patients with marginal zone lymphoma (MZL) who are generally treated in a similar manner to follicular lymphoma (FL). Aims: Primary results for the AUGMENT study in R/R FL grade 1–3a and MZL showed significantly improved progression‐free survival (PFS) for R 2 over R/placebo in the overall population and in the subgroup analyses, with the exception of the MZL subgroup. Given the small sample size for MZL patients, post hoc analyses of baseline characteristics and univariate/multivariate analyses of PFS were conducted to examine possible explanations for these results. Methods: The phase III AUGMENT study randomized patients 1:1 to R 2 (lenalidomide PO 20 mg/day (d), d1–21/28 X12 cycles plus rituximab IV 375 mg/m 2 weekly in cycle 1 and d1 of cycles 2–5) or rituximab‐placebo (R/placebo; same dosing schedule). PFS by 2007 IWG (without PET) was the primary endpoint. Post hoc Cox regression models were used for univariate analysis of one risk factor and multivariate analyses including treatment arm and significant risk factors ( P  < 0.05) from the univariate analyses. Results: A total of 63/358 (18%) MZL patients were part of the overall study population, including n = 14/16 MALT, n = 9/6 splenic, and n = 8/10 nodal MZL subtypes for respective R 2 and R/placebo arms. Numerical differences between treatment arms for baseline demographics and disease characteristics for MZL patients favoring the R/placebo arm were: fewer R 2 patients had an ECOG score of 0 (55% R 2 vs 72% R/placebo), and more R 2 (vs R/placebo) patients were older (≥65 y: 68% vs 59%; ≥70 y: 42% vs 38%), were refractory to the last prior regimen (13% vs 3%), had Ann Arbor stage IV disease at enrollment (65% vs 41%), high‐risk MALT‐IPI score (50% vs 19%), elevated LDH (29% vs 19%), B symptoms (13% vs 3%), and high tumor burden per GELF criteria (65% vs 56%). The most common grade 3/4 AEs for MZL patients receiving R 2 vs R/placebo, respectively, were neutropenia (47% vs 16%), pneumonia (3% vs 13%), and leukopenia (10% vs 0). A total of 5 deaths occurred in the R 2 arm (most associated with progression of disease; 2 deaths occurred early at 3 and 13 days after randomization [one prior to receiving R 2 and one 2 days after treatment initiation]) and 2 in the R/placebo arm. Despite worse prognostic features, best response was improved with R 2 (vs R/placebo) with ORR = 65% vs 44% ( P  = 0.13) and CR = 29% vs 13% ( P  = 0.13), although neither were statistically significant. Improved response rates did not translate into a survival advantage for R 2 . Median PFS for MZL patients was 20.2 mo R 2 vs 25.2 mo R/placebo (HR = 1.00; 95% CI, 0.47–2.13; P  = 1.0). Univariate analyses showed that several baseline factors were prognostic for MZL patients ( P  < 0.05): Ann Arbor stage IV, elevated LDH, and unfit for chemotherapy (Table). Multivariate analyses adjusting for the imbalance in these 3 significant prognostic factors showed an adjusted PFS HR of 0.51 (95% CI, 0.20–1.28) favoring the R 2 arm, similar to the PFS HR in the overall population (HR = 0.46; 95% CI, 0.34–0.63). The univariate and multivariate analyses suggested that the imbalance in baseline prognostic factors impacted PFS results for MZL patients. Summary/Conclusion: Overall, these findings suggest that the PFS results in MZL patients were negatively impacted by the imbalance in baseline prognostic factors and more aggressive disease in the R 2 over the R/placebo arm.