Résumé
Abstract 270
Obinutuzumab (GA101), the first type II, glycoengineered, humanized anti-CD20 monoclonal antibody in clinical development, has shown single-agent activity in Phase I and II studies in follicular lymphoma (FL), but no studies have so far examined the safety and activity of GA101 in combination with chemotherapy, or compared GA101 dose levels in large cohorts. This study evaluates the feasibility, safety and efficacy of GA101 in combination with standard chemotherapy regimens for FL at two different doses of GA101.
Patients with relapsed or refractory FL (n=56) were stratified by chemotherapy regimen based upon prior treatment history (cyclophosphamide/doxorubicin/vincristine/prednisone [6–8 21 day cycles; n=28] or fludarabine/cyclophosphamide [4–6 28 day cycles; n=28]). Patients were then randomized to one of two GA101 dosing regimens: 1,600 mg on Days 1 and 8 of cycle 1 then 800 mg for subsequent cycles (1,600/800 mg) or 400 mg in all cycles (400/400 mg). These regimens represent a range of active doses in indolent lymphoma, based upon Phase I and II trials in which there were no dose-limiting toxicities. Responding patients were offered maintenance treatment for 2 years or until progression. The primary objective was safety, with response rate a secondary objective. Response was assessed at the end of induction using International Working Group response criteria (Cheson, et al. J Clin Oncol 1999), modified to classify unconfirmed complete response as partial response.
Baseline characteristics were similar for both groups (G-CHOP and G-FC, respectively): median age 62.5 and 61.0 years; low-risk FLIPI 29% in both groups; median prior treatments (range) 1 (1–3) and 2 (1–6); bone marrow involvement 25% and 26%; Ann Arbor stage III–IV at study entry 64% and 82%. All patients (28/28) in the G-CHOP arm and 22/28 patients in the G-FC arm completed treatment. Reasons for withdrawal (G-FC arm) were neutropenia (n=3), rash (n=1), infection (n=1) and insufficient response (n=1). The most common AEs in both groups were infusion-related reactions (all grades: 64% G-CHOP; 79% G-FC; Grade 3/4: 7% G-CHOP; 7% G-FC), mostly during the first infusion. Grade 3 or 4 neutropenia was reported in 39% of G-CHOP patients and 50% of G-FC patients. Of 190 cycles of G-CHOP delivered, 11 cycles (6%) were delayed in 8 patients for neutropenia or infection (6 cycles delayed by 1 week; 5 cycles delayed by 2 weeks). Dose of any CHOP component was reduced in 29% of patients, in 5 patients for neuropathy and in 1 patient each because of neutropenia, infection and allergic rhinitis. In the G-FC group, 14 of 135 delivered cycles (10%) were delayed in 10 patients for hematologic toxicity or infections (10 cycles delayed by 1 week; 4 cycles delayed by 2 weeks). Nine of these patients also had a dose reduction in both cytostatic components of the regimen with a further patient having a dose reduction only, for an overall dose reduction in 36% of patients. Three deaths were reported following G-FC induction treatment (progressive disease, n=1; underlying Parkinson's disease, n=1; and chronic obstructive pulmonary disease during maintenance, n=1), with none considered to be treatment-related. There was no evidence for increased toxicity with the 1,600/800 mg dose compared with the 400/400 mg dose of GA101. The overall response rate (ORR) at the end of induction was 96.4% in the G-CHOP group (39.3% complete response [CR]) and 92.9% in the G-FC group (50.0% CR) (Table). Data from the G-CHOP cohort were compared in a matched-pair analysis to the rituximab plus CHOP (R-CHOP) arm from study M39022 (EORTC 20981) in a similar patient population. Response rates to G-CHOP compared favorably with response rates to R-CHOP.Response rates at end of inductionResponse, n (%)G-CHOPG-FCOverall response27 (96.4)26 (92.9)Complete response11 (39.3)14 (50.0)Partial response16 (57.1)12 (42.9)Stable disease1 (3.6)0Progressive disease01 (3.6)Not assessed01 (3.6)
In conclusion, GA101 can be combined safely with chemotherapy regimens used in the treatment of FL, and demonstrates a high level of activity compared with historical controls. G-CHOP could be delivered at the protocol-specified 3-weekly interval in most patients. G-FC in a more heavily pretreated population showed worse tolerability. Following these promising results, GA101 will be studied in combination with CHOP and other chemotherapies in a randomized Phase III study against the standard of care, R-CHOP.