Assessment of Measures to Reduce Infusion-Related Reactions in Patients with Chronic Lymphocytic Leukemia Treated with Obinutuzumab in the GREEN Study

Background: Obinutuzumab (GA101; GAZYVA/GAZYVARO; G) is a glycoengineered type II anti-CD20 antibody with significant activity in chronic lymphocytic leukemia (CLL). In the CLL11 study, infusion-related reactions (IRRs) were more common and more severe with G plus chlorambucil (Clb) than with rituximab plus Clb in CLL patients (pts) with coexisting conditions, and seen most frequently after the first G infusion. GREEN (NCT01905943) is an ongoing Phase 3b study investigating the safety and efficacy of G alone or in combination with chemotherapy in CLL pts. The primary endpoint is safety (adverse event [AE] profile). We report an exploratory analysis of the effectiveness of three different approaches to reduce IRRs during or after G administration.

Methods: Enrolled pts were aged ≥18 years (y) with documented CLL, an ECOG PS of 0-2, and adequate hematologic function. Pts received G 1000mg i.v., alone or with chemotherapy, on Day (D) 1, 8, and 15 of cycle (C) 1, and D1 of C2-6 (six 28-day cycles), with the C1D1 dose administered over 2 days as follows: Cohort 1, 25mg + 975mg infused at 12.5mg/hour (h) and 25mg/h, respectively; Cohort 2, 100mg (25mg/h) and 900mg (50-400mg/h) with oral dexamethasone 20mg or equivalent 12h pre-dose; Cohort 3 (recruited after independent analysis of Cohorts 1 and 2), doses and infusion rates as in Cohort 1 with pre-medication scheme as in Cohort 2. All three cohorts also received i.v. corticosteroids 1h pre-dose on D1. Administered regimens included G alone for any pt or G combined with chemotherapy: fludarabine and cyclophosphamide for fit pts only (CIRS ≤6 and CrCl ≥70ml/min); Clb for unfit pts only (CIRS >6 and/or CrCl <70ml/min); or bendamustine (B) for any pt. Pts refractory to previous G monotherapy had to receive G with chemotherapy. IRRs were defined broadly as any AE occurring during or within 24h of G infusion and considered related to G. Pts with high tumor burden (bulky lymphoadenopathy ≥10cm, or ≥5cm and <10cm with peripheral blood lymphocyte count ≥25x10⁹/l) were considered high risk for tumor lysis syndrome (TLS). The data cut-off for the analysis was March 25, 2016.

Results: A total of 622 first-line (1L) pts were analyzed; 609 were evaluated by Cohort (1, 238; 2, 228; 3, 143). Median age was 66.0y in Cohort 1 (61.8% at-risk for TLS), 64.5y in Cohort 2 (54.4% at-risk), and 64.0y in Cohort 3 (49.7% at-risk); approximately 54% of pts in each cohort were fit. Baseline demographic and disease characteristics were similar across cohorts. Median time on G treatment was 20.4 weeks (wks) in Cohort 1, 20.3 wks in Cohort 2, and 8.3 wks in Cohort 3; 62.6%, 46.5%, and 53.1% of pts, respectively, received G-B. Median observation time was 23.4, 15.5, and 3.4 wks, respectively. The frequency of all IRRs was similar across cohorts, but grade 3-4 IRRs, serious IRRs, and IRRs leading to discontinuation were more frequent in Cohort 3 than Cohorts 1 and 2 (Table). There were no grade 5 IRRs. All 44 TLS cases (36 classified as IRRs) were grade 3-5 and 23 were serious (5 [2.1%], 5 [2.2%] and 8 [5.6%] in Cohorts 1, 2, and 3, respectively); 29/44 (66%) were identified from laboratory findings alone. Of grade 3-4 and serious IRRs, TLS was the most frequent AE in 1L pts. TLS (as an IRR) was more frequent in Cohort 3 (9.8%) than in Cohorts 1 (6.7%) or 2 (2.6%), but cytokine-related IRRs, e.g. chills, gastrointestinal events (nausea and vomiting), and blood pressure events (hypertension and hypotension) were less frequent in Cohort 3 than Cohorts 1 and 2 (Table). In Cohort 3, most grade 3-4 IRRs (73%) occurred >5h after the C1D1 dose; in Cohorts 1 and 2, most occurred within 5h of dosing (60% and 58%, respectively). Treatment-emergent AEs were reported by 95.0% of all enrolled 1L pts (n=622; grade ≥3 AEs, 71.5%; serious AEs, 46.3%; AEs leading to G discontinuation, 13.5%). Of 29 deaths (4.7%) in 1L pts, 9 (1.4%) were related to G.

Conclusions: The frequency of all IRRs was similar across the three cohorts, but the highest frequencies of grade ≥3 and serious IRRs, most notably TLS, were seen in Cohort 3 (most stringent IRR reduction measures). In contrast, rates of some cytokine-related IRRs were lowest in this cohort. Sequential recruitment may have resulted in over-reporting of IRRs, particularly TLS, in Cohort 3: updated definitions of patients at risk of TLS and additional risk mitigation measures were communicated to investigators in January 2016 in the early stages of recruitment to this cohort.

View largeDownload slide

View largeDownload slide

Close modal