AUGMENT: A Phase III Study of Lenalidomide Plus Rituximab Versus Placebo Plus Rituximab in Relapsed or Refractory Indolent Lymphoma.

PURPOSE: Patients with indolent non-Hodgkin lymphoma typically respond well to first-line immunochemotherapy. At relapse, single-agent rituximab is commonly administered. Data suggest the immunomodulatory agent lenalidomide could increase the activity of rituximab.
METHODS: A phase III, multicenter, randomized trial of lenalidomide plus rituximab versus placebo plus rituximab was conducted in patients with relapsed and/or refractory follicular or marginal zone lymphoma. Patients received lenalidomide or placebo for 12 cycles plus rituximab once per week for 4 weeks in cycle 1 and day 1 of cycles 2 through 5. The primary end point was progression-free survival per independent radiology review.
RESULTS: A total of 358 patients were randomly assigned to lenalidomide plus rituximab (n = 178) or placebo plus rituximab (n = 180). Infections (63% v 49%), neutropenia (58% v 23%), and cutaneous reactions (32% v 12%) were more common with lenalidomide plus rituximab. Grade 3 or 4 neutropenia (50% v 13%) and leukopenia (7% v 2%) were higher with lenalidomide plus rituximab; no other grade 3 or 4 adverse event differed by 5% or more between groups. Progression-free survival was significantly improved for lenalidomide plus rituximab versus placebo plus rituximab, with a hazard ratio of 0.46 (95% CI, 0.34 to 0.62; P < .001) and median duration of 39.4 months (95% CI, 22.9 months to not reached) versus 14.1 months (95% CI, 11.4 to 16.7 months), respectively.
CONCLUSION: Lenalidomide improved efficacy of rituximab in patients with recurrent indolent lymphoma, with an acceptable safety profile.

RCT Entities:   Population Interventions Outcomes

INTRODUCTION

Non-Hodgkin lymphomas (NHLs) are mostly of B-cell origin[1] and include low-grade, indolent histologies that usually respond to initial therapy but typically relapse.[1-4] The most common indolent NHL types, follicular lymphoma (FL) and marginal zone lymphoma (MZL), account for 22% and 7% of adult NHL, respectively.[5,6] Despite being distinct entities, recurrent FL and MZL are treated similarly.[7,8] Single-agent rituximab is approved by the US Food and Drug Administration and is commonly used as treatment of these patients.

Lenalidomide is an immunomodulatory (IMiD) drug that binds to the cereblon E3 ubiquitin ligase complex, resulting in ubiquitination of the transcription factors Aiolos and Ikaros, leading to antilymphoma effects.[9,10] Preclinically, lenalidomide restored the response of tumor-infiltrating lymphocytes in autologous T-cell conjugates[11] and increased natural killer cell count and function in peripheral blood and natural killer cell lines.[12,13] Adding lenalidomide to rituximab enhanced antibody-dependent cell-mediated cytotoxicity, immune synapse formation, monocyte-mediated killing, and direct cytotoxicity against FL cells.[11,14-16]

There are several treatment options, none considered curative, for patients with relapsed/refractory FL and MZL, including chemotherapy plus anti-CD20 monoclonal antibodies and targeted agents such as phosphatidylinositol 3-kinase inhibitors. Treatment choice is often based on duration of response to prior therapies, types of prior therapies, and patient comorbidities.[3,17] Rituximab monotherapy is a treatment option in patients who had previously responded to rituximab, on the basis of observations that frequent responses can occur with rituximab retreatment.[18,19] Rituximab monotherapy was commonly used in the second-line treatment of FL (25% to 47% of patients) according to studies in the United States and Europe.[20-22] Lenalidomide plus rituximab combination showed clinical activity in patients with previously treated indolent NHL in a key phase II study[23] and others[24,25] demonstrating overall response rates of 65% to 77%, complete response (CR) rates of 35% to 41%, and median progression-free survival (PFS)/time to progression of 1 to 2 years. Recently, the lenalidomide plus rituximab combination also showed clinical activity in a phase III study of advanced untreated FL.[26] The AUGMENT trial (ClinicalTrials.gov identifier: NCT01938001) prospectively compared efficacy and safety of lenalidomide plus rituximab to placebo plus rituximab (a standard of care, among several) in patients with relapsed or refractory indolent NHL who are appropriate for rituximab monotherapy (Appendix Table A1, online only).

TABLE A1.

List of AUGMENT Trial Investigators

METHODS

Patients

Eligible patients had MZL or FL (grades 1 to 3a) requiring treatment per investigator assessment; at least one prior chemotherapy, immunotherapy, or chemoimmunotherapy and two or more previous doses of rituximab; and relapsed, refractory, or progressive disease and not rituximab-refractory disease. Patients with neuropathy grade greater than one were excluded. Additional eligibility criteria are in the Appendix (online only).

Trial Design and Treatment

Patients were randomly assigned (1:1 ratio) to lenalidomide plus rituximab (lenalidomide plus rituximab group) or placebo plus rituximab (placebo plus rituximab group). Random assignment was stratified according to previous rituximab treatment (yes or no), time since last therapy (≤ 2 years v > 2 years), and histology (FL v MZL). Prior induction and maintenance treatment were considered one treatment line.

Treatment continued for 12 cycles or relapse, progressive disease, withdrawal of consent, or unacceptable toxicity. Lenalidomide plus rituximab dosing included oral lenalidomide 20 mg daily (10 mg for creatinine clearance 30 to 59 mL/min) on days 1 to 21 plus intravenous rituximab 375 mg/m2 days 1, 8, 15, and 22 of cycle 1 and day 1 of cycles 2 to 5 every 28 days. Placebo plus rituximab was administered similarly. The rituximab regimen was selected using efficacy and statistical assumptions from the LYM-3001 trial (ClinicalTrials.gov identifier: NCT00312845) in similar patients.[18] Rationale for lenalidomide and rituximab dosing schedules are detailed in the Appendix. On treatment completion or discontinuation, patients were observed for progression, subsequent therapies, response to next therapies, and second malignancies for up to 5 years after the last patient was randomly assigned (Appendix).

Dose adjustments of lenalidomide were planned to manage toxicity (Appendix). In patients with thromboembolism risk, prophylactic anticoagulation or antiplatelet therapy at investigator discretion was recommended. Rituximab dose was not reduced, but if discontinued for toxicity, lenalidomide plus placebo continued per protocol (Data Supplement). Growth factor use was allowed per ASCO/European Society for Medical Oncology guidelines.[27,28]

This study was performed following Good Clinical Practice per International Conference on Harmonization Guideline E6 requirements under ethical principles in the Declaration of Helsinki. Study conduct followed guidance of each site’s institutional review board, independent ethics committee, and regulatory authorities. All patients provided written informed consent before trial enrollment.

Efficacy and Safety Assessments

Primary efficacy analyses were conducted in the intention-to-treat population, defined as all patients randomly assigned, regardless of receiving trial treatment. The primary end point was PFS assessed by the independent review committee (IRC) per 2007 International Working Group criteria,[29] without positron emission tomographic imaging. Secondary end points included overall response rate, CR, duration of response, overall survival (OS), event-free survival, and time to next antilymphoma treatment. Time to next chemotherapy treatment and histologic transformation were exploratory end points.

Response and progression outcomes were assessed by a blinded, independent central review using 2007 International Working Group criteria on the basis of computed axial tomography/magnetic resonance imaging scans. Patients with gastric mucosa–associated lymphoid tissue lymphoma underwent endoscopy for response evaluation. Bone marrow biopsy was required to confirm CR.

The safety analysis population included all patients receiving at least one dose of study medication. Adverse event classification used National Cancer Institute Common Terminology Criteria for Adverse Events version 4.03, except for tumor flare (graded by National Cancer Institute Common Terminology Criteria for Adverse Events version 3.0) and tumor lysis (assessed per Cairo-Bishop Criteria).[30]

Statistical Analyses

The primary objective was to compare efficacy in the lenalidomide plus rituximab group versus the placebo plus rituximab group on the basis of the primary end point of PFS at one-sided α = 0.025 level. We hypothesized that median PFS for lenalidomide plus rituximab would be 17.6 months, and 11 months for placebo plus rituximab, on the basis of previous results of rituximab monotherapy (Appendix).[18] For 90% power to detect this difference with one-sided α = 0.025 and one interim analysis for the futility at 50% information time, a total of 193 PFS events were required.

The distribution for PFS and other time-to-event data were estimated using the Kaplan-Meier procedure.[31] A hazard ratio (HR) with a two-sided 95% CI was estimated using the stratified Cox proportional hazard regression model. For binary types of secondary efficacy end points, the number and percentage of patients were tabulated by treatment group, a stratified Cochran-Mantel-Haenszel test adjusting for possible confounding effects of the stratification factors was performed, and a P value was provided. Prespecified subgroup analysis of PFS was also performed and HR estimated from Cox proportional hazard regression model; P value was generated using Fisher’s exact t test for binary end points and log-rank test for time to event end points.

RESULTS

Patients and Trial Treatment

From February 13, 2014 through January 26, 2017, 358 patients enrolled from 97 centers in 15 countries were randomly assigned 1:1 to lenalidomide plus rituximab (n = 178) or placebo plus rituximab (n = 180; Fig 1). Baseline characteristics were similar (Table 1). Median age was 63 years (range, 26 to 88 years); 295 patients (82%) had FL, and 63 (18%) had MZL. Overall, 51% had high tumor burden per Groupe d’Etude des Lymphomes Folliculaires (GELF) criteria. The median number of prior therapies was one (range, one to 12), and 86 patients (24%) had three or more prior systemic treatments. Relapse or progression within 2 years of initial diagnosis had occurred in 117 patients (33%), and 56 (16%) were refractory to their last regimen. Four patients had been lost to follow-up.

FIG 1.

Lenalidomide plus rituximab and placebo plus rituximab group CONSORT diagram (flow of patients from enrollment to analysis). AE, adverse event; ITT, intention to treat.

TABLE 1.

Baseline Demographic and Disease Characteristics (ITT Population)*

In the safety population, the full treatment course was completed in 71% of patients with lenalidomide plus rituximab and 61% with placebo plus rituximab. Adverse events leading to dose interruptions were more common with lenalidomide plus rituximab than lenalidomide plus placebo (64% v 26%), reductions (26% v 3%), and discontinuations (9% v 5%). Neutropenia was the most common adverse event leading to reduction (18%) and interruption (39%) of lenalidomide. Dose modifications successfully addressed neutropenia, with only five patients (3%) discontinuing lenalidomide because of neutropenia. Disease progression led to treatment discontinuations in 21 patients (12%) and 54 patients (30%) in the lenalidomide plus rituximab and placebo plus rituximab groups, respectively. Type 1 error was not controlled for secondary and exploratory end points, and P values were descriptive in nature.

Efficacy

Primary end point.

At the final analysis, median follow-up was 28.3 months, and 185 events total (progression or death) were assessed by IRC (200 events per investigator assessment) before censoring. The primary end point of PFS assessed by IRC was significantly superior in the lenalidomide plus rituximab group (HR, 0.46; 95% CI, 0.34 to 0.62; P < .001; Fig 2A; Table 2). Median PFS assessed by IRC was 39.4 months (95% CI, 22.9 months to not reached) with lenalidomide plus rituximab versus 14.1 months (95% CI, 11.4 to 16.7 months) with placebo plus rituximab. PFS assessed by investigator also showed superiority with lenalidomide plus rituximab versus placebo plus rituximab (HR, 0.51; 95% CI, 0.38 to 0.69; P < .0001; the median PFS was 25.3 months; 95% CI, 21.2 months to not reached v 14.3 months; 95% CI, 12.4 to 17.7 months; Appendix Fig A1, online only; Table 2). PFS probability at 2 years also favored lenalidomide plus rituximab (Table 2). Post hoc subgroup analyses for PFS on the basis of prior rituximab plus bendamustine or plus cyclophosphamide, doxorubicin, vincristine, and prednisone showed results consistent with those of the overall population (Appendix Table A2, online only).

FIG 2.

Progression-free survival (PFS) and overall survival (OS) as assessed by independent review committee in the intention-to-treat population: (A) progression-free survival; (B) overall survival.

TABLE 2.

Efficacy Outcomes (ITT Population)

FIG A1.

Progression-free survival (PFS) as assessed by investigator in the intention-to-treat population.

TABLE A2.

Post hoc Subgroup Analyses for Progression-Free Survival on the Basis of Prior Treatment as Assessed by IRC

Secondary and exploratory end points.

Among patients in the lenalidomide plus rituximab versus placebo plus rituximab groups, the rate of best overall response assessed by IRC was 78% (95% CI, 71% to 83%) versus 53% (95% CI, 46% to 61%; P < .0001), with 34% (95% CI, 27% to 41%) versus 18% (95% CI, 13% to 25%) of patients achieving CR (P = .001; Table 2), with similar investigator-assessed results. Other secondary and exploratory time-to-event end points assessed by IRC also showed superior results with lenalidomide plus rituximab—response duration (Appendix Fig A2, online only; Table 2), event-free survival (Appendix Fig A3, online only; Table 2), time to next antilymphoma treatment (Appendix Fig A4, online only; Table 2), time to next chemotherapy treatment (Appendix Fig A5, online only; Table 2), and PFS on next antilymphoma treatment (Appendix Fig A6, online only; Table 2).

FIG A2.

Duration of response as assessed by independent review committee in the intention-to-treat population.

FIG A3.

Event-free survival as assessed by independent review committee in the intention-to-treat population.

FIG A4.

Time to next antilymphoma treatment in the intention-to-treat population.

FIG A5.

Time to next antilymphoma chemotherapy in the intention-to-treat population.

FIG A6.

Progression-free survival incorporating next antilymphoma treatment (PFS2) in the intention-to-treat population.

Overall survival results are maturing, with an HR of 0.61 (95% CI, 0.33 to 1.13; Fig 2B; Table 2). Numerically fewer deaths in treated patients have been observed with lenalidomide plus rituximab versus placebo plus rituximab (15 v 26), although the trial was not powered to detect OS differences. Estimated 2-year survival in the lenalidomide plus rituximab group was 93% (95% CI, 87% to 96%) versus 87% (95% CI, 81% to 92%) in the placebo plus rituximab group. Survival for FL and MZL subgroups are shown in Appendix Figures A7 and A8 (online only). In patients with FL, OS results favored lenalidomide plus rituximab (HR, 0.45; 95% CI, 0.22 to 0.92; P = .02; Appendix Table A3, online only), with 11 deaths with lenalidomide plus rituximab versus 24 with placebo plus rituximab. No difference was seen between groups in patients with MZL (HR, 2.89; 95% CI, 0.56 to 14.92), but with few events—five with lenalidomide plus rituximab versus two with placebo plus rituximab (Appendix Table A4, online only). Two patients with MZL in the lenalidomide plus rituximab arm died early after random assignment (3 days [did not receive any study treatment] and 13 days). Estimated 2-year survival probability in FL was 95% (95% CI, 90% to 98%) with lenalidomide plus rituximab and 86% (95% CI, 79% to 91%) with placebo plus rituximab (Appendix Table A3). Estimated 2-year survival in MZL was 82% (95% CI, 61% to 92%) with lenalidomide plus rituximab and 94% (95% CI, 77% to 98%) with placebo plus rituximab (Appendix Table A4).

FIG A7.

Overall survival (OS) in patients with follicular lymphoma in the intention-to-treat population.

FIG A8.

Overall survival (OS) in patients with marginal zone lymphoma in the intention-to-treat population.

TABLE A3.

Efficacy in Patients with Follicular Lymphoma

TABLE A4.

Efficacy in Patients With Marginal Zone Lymphoma

Histologic transformation occurred in two patients (incidence per 100 person-years: 0.5%) with lenalidomide plus rituximab and 10 patients (incidence per 100 person-years: 2.5%) with placebo plus rituximab (Table 3). After transformation, one patient in the lenalidomide plus rituximab group and six in the placebo plus rituximab group died.

TABLE 3.

Treatment-Emergent Adverse Events in the Safety Population (≥ 10% of patients)

Subgroup analyses.

PFS improvements were consistent with those of the overall population in all prespecified subgroups except the MZL subgroup (Fig 3). In this subset (n = 63), PFS was not significantly different between the two groups, with an unstratified HR of 1.00 (95% CI, 0.47 to 2.13), and the wide confidence intervals imply the data are uninformative. PFS results in the MZL subgroup are difficult to interpret because of the small sample size and imbalance in baseline prognostic factors (Appendix Table A5, online only).

FIG 3.

Forest plot: subgroup analyses of progression-free survival. GELF, Groupe d’Etude des Lymphomes Folliculaires; HR, hazard ratio.

TABLE A5.

Baseline Demographic and Disease Characteristics (ITT Population) by Histology

Safety

The safety population included 176 patients who received lenalidomide plus rituximab and 180 patients who received placebo plus rituximab. Of these, 174 patients who received lenalidomide plus rituximab (99%) and 173 who received placebo plus rituximab (96%) had any-grade adverse events within 28 days after last dose. Most common adverse events are in Table 3. Adverse events of any grade that occurred more frequently (≥ 10% difference) with lenalidomide plus rituximab versus placebo plus rituximab included neutropenia (58% v 22%), constipation (26% v 14%), leukopenia (20% v 9%), anemia (16% v 4%), thrombocytopenia (15% v 4%), and tumor flare (11% v 1%), respectively. More patients who received lenalidomide plus rituximab (69%) had at least one grade 3 or 4 adverse event compared with placebo plus rituximab (32%). The increased rates of grade 3 or 4 adverse events with lenalidomide plus rituximab were attributable primarily to increased grade 3 or 4 neutropenia (50% v 13%) and leukopenia (7% v 2%). No other grade 3 or 4 adverse events occurred in 5% or more patients in either group. Febrile neutropenia occurred in 3% of patients with lenalidomide plus rituximab versus 1% with placebo plus rituximab. Growth factors were administered to 36% of the lenalidomide plus rituximab group versus 12% in the placebo plus rituximab group. Neutropenia was primarily managed through dose interruptions and/or reductions and growth-factor support. Only five patients had neutropenia leading to lenalidomide discontinuation. All incidences of grade 3 or 4 neutropenia in the lenalidomide plus rituximab group recovered to grade 1 or less, with a median time of 9.0 days. Efficacy was similar regardless of occurrence of grade 3 or 4 neutropenia in the lenalidomide plus rituximab group on the basis of post hoc analysis (Appendix Table A6, online only).

TABLE A6.

Efficacy by Occurrence of Grade 3 or 4 Neutropenia in the Safety Population

Fatal adverse events (grade 5) occurred in four patients (1%), two in each group (lenalidomide plus rituximab: arrhythmia and cardiopulmonary failure; placebo plus rituximab: general physical health deterioration and pneumonia). Among treated patients, 15 deaths (9%) occurred in the lenalidomide plus rituximab group (five deaths attributed to lymphoma) versus 26 (14%) with placebo plus rituximab (18 deaths attributed to lymphoma).

With lenalidomide plus rituximab, 45 patients (26%) had at least one serious adverse event versus 25 (14%) with placebo plus rituximab (Appendix Table A7, online only). Pneumonia was the most common serious adverse event, occurring in five patients (3%) in each group. Deep vein thrombosis occurred in three patients (2%) in the lenalidomide plus rituximab and one patient (1%) in the placebo plus rituximab groups; only one incidence (lenalidomide plus rituximab group) was a serious adverse event. Antiplatelet and anticoagulant medication use is listed in Appendix Table A8 (online only). Second primary cancers were reported in six patients (3%) in the lenalidomide plus rituximab group and 10 patients (6%) with placebo plus rituximab (Appendix Table A9, online only). Two patients died of second primary cancers (both placebo plus rituximab arm).

TABLE A7.

Serious Adverse Events Occurring in Three or More Patients in Either Group in the Safety Population

TABLE A8.

On-Study Use of Antiplatelet or Anticoagulant Concomitant Medications in the Safety Population

TABLE A9.

Second Primary Malignancies in the Safety Population

DISCUSSION

The AUGMENT study met its primary end point; lenalidomide plus rituximab demonstrated statistically significant and clinically relevant superiority in PFS over placebo plus rituximab in patients with relapsed or refractory indolent FL/MZL who are considered appropriate for rituximab monotherapy. Lenalidomide plus rituximab, administered over 1 year, reduced risk of progression by 54% (HR, 0.46; 95% CI, 0.34 to 0.62; P < .0001) and increased median PFS by more than 2 years compared with rituximab monotherapy. Efficacy of the combination was also reflected by improvements in secondary and exploratory end points—response rates, response duration, time to next treatment, and time to next chemotherapy. Furthermore, PFS improved in all prespecified subgroups (prior rituximab, age, time since last therapy, sex, race, region, number of prior antilymphoma regimens, stage, refractory to last regimen, high tumor burden, chemoresistant, and unfit for chemotherapy [defined as age ≥ 70 years, or age 60 to 69 years and creatinine clearance < 60 mL/min or Eastern Cooperative Oncology Group performance status ≥ 2]) except for the MZL subgroup. No difference in PFS was observed between treatment groups in the MZL subgroup. Although this could relate to lack of effect in this subset, the small number of patients and imbalance in prognostic factors between arms limit interpretation for this histology.

One limitation of this study was the differences in median PFS seen when assessed by IRC (39.4 months) compared with investigator assessment (25.3 months); however, HRs and P values were similar between assessments, indicating consistency in PFS results between IRC and investigator assessments. Another limitation was that longer treatment duration occurred in the lenalidomide plus rituximab group; however, this likely did not influence efficacy results. Notably, separation of PFS Kaplan-Meier curves began early, and PFS benefits and durability of responses persisted beyond 1 year of study treatment duration and were consistently observed throughout the study follow-up period. In addition, schedule of rituximab (eight doses given over 5- to 28-day cycles) is not likely to have influenced results; several studies showed benefit of extended treatment with rituximab (ie, beyond standard 4 weekly infusions) in a manner that is independent of the number or schedule of extended rituximab dosing (ie, four or more doses every 1, 2, 3, or 6 months).[18,32-35] Importantly, the control group performed as expected on the basis of historical rituximab data, with results similar to the 11-month estimate used in the statistical plan.[18,19,36]

Current treatment options for recurrent FL include rituximab monotherapy, bendamustine, or other chemotherapy alone or with obinutuzumab[37] or rituximab.[38] In addition, the phosphatidylinositol 3-kinase inhibitors idelalisib[39] and copanlisib are approved for patients who are more heavily pretreated than the patients in the AUGMENT study.[40] For many patients, the efficacy/toxicity tradeoffs of rituximab versus chemotherapy-based or kinase inhibitor regimens are major issues. There is clear rationale for adding lenalidomide to rituximab in relapsed patients with FL,[26] and in a previous study in recurrent FL, lenalidomide plus rituximab was superior to lenalidomide alone.[23] Our data demonstrate that lenalidomide plus rituximab offers clinically meaningful efficacy advantages over single-agent rituximab in the context of the safety profile. Adverse events were more common in the lenalidomide plus rituximab group, largely due to higher rates of grade 3 or 4 neutropenia, which was successfully managed in most patients through dose modifications and growth factors. Neutropenia was predictable and manageable; the modest increase in infections and cutaneous reactions must be considered in light of the significantly greater efficacy of the lenalidomide plus rituximab combination. In fact, more patients in the lenalidomide plus rituximab group completed therapy than those in the placebo plus rituximab group (71% v 61%) because of a lower rate of progression. The magnitude of efficacy differences between the two treatments is clinically meaningful and suggests that lenalidomide plus rituximab should replace rituximab monotherapy as a standard of care for patients with relapsed or refractory indolent NHL.

TABLE A10.

Prior Systemic Anticancer Therapies by Regimen

TABLE A11.

Response to Subsequent Treatment