Obinutuzumab and CHOP combination for advanced DLBCL: Results from the phase II GATHER trial

On 2 October 2018, Jeff P. Sharman from the Willamette Valley Cancer Institute and Research Center, Oregon, USA and colleagues, published in Leukemia & Lymphoma the final analysis of the phase II trial GATHER (NCT01414855). In this open-label, multicenter study, the efficacy and safety of obinutuzumab (RO5072759 [GA101]) in combination with cyclophosphamide, doxorubicin, vincristine, prednisone (CHOP) chemotherapy in patients with advanced diffuse large B-cell lymphoma (DLBCL) was evaluated.

DLBCL is one of the most common lymphomas and with the current standard of care (rituximab and CHOP), approximately 60% of patients are cured or improve. Nevertheless, there is still 40% of patients who remain uncured. Obinutuzumab is a glycoengineered, humanized, type 2 anti-CD20 monoclonal antibody with high anti-tumor activity. In this phase II trial, the safety and efficacy of obinutuzumab in combination with CHOP (G-CHOP) was investigated, with a minimum of three years follow-up, in newly-diagnosed advanced DLBCL patients. The primary outcomes of the study were overall response rate (ORR) and complete response (CR) rates calculated by investigator assessment, while secondary outcomes included independent review of ORR and CR, as well as progression-free survival (PFS) and overall survival (OS). The importance of the cell of origin (COO) on the efficacy and safety of a shorter duration infusion of obinutuzumab was also evaluated in this study.

Study design

• N = 87 previously untreated patients (≥ 18 years old) with CD20-positive advanced DLBCL, completed all planned therapy in 25 US centers
• Key patient inclusion criteria:
  • Ann Arbor stage II with bulky disease (mass > 7.5 cm), stage III, or stage IV
  • International Prognostic Index (IPI): 1 with bulky disease, or 2−5
  • Eastern Cooperative Oncology Group (ECOG) performance status: 0−2
• Treatment plan:
  • Patients received eight 21-day cycles (10 doses) of 1000 mg obinutuzumab intravenously (IV) on Day 1 of cycles 1–8 with additional doses on Day 8 and Day 15 of cycle 1, in combination with six cycles of standard CHOP (C: 750 mg/m² IV Day 1; H: 50 mg/m² IV Day 1; O: 1.4 mg/m² that could be capped at 2 mg IV Day 1; and P: 100 mg orally Days 1–5)
• For prophylaxis against infusion-related reactions (IRR), acetaminophen 650−1000 mg and antihistamines e.g. diphenhydramine 50−100 mg) were administered orally 30 to 60 min before each obinutuzumab infusion. The use of prophylactic corticosteroids was considered for high-risk IRR patients (administered in n = 71 patients)
• For neutropenia prophylaxis, granulocyte-colony stimulating factor (filgrastim) was administered to patients ≥ 60 years old or with comorbidities and was strongly recommended for all patients in cycle 1 based on the incidence of neutropenia in previous studies
• Primary analysis cut-off date was April 2013, while the final analysis cut-off date was December 2016 with a data snapshot date of March 2017

Key findings 

• Investigator-assessed ORR and CR rate were 82.0% and 55.0%, respectively. Independently reviewed ORR and CR rate were similar at 75.0% and 58.0%, respectively
• At final analysis, median observation time for time-to-event endpoints was 50.7 months (range, 3.5–63.2 months)
• Investigator-assessed PFS values were not affected by IPI:
  • Median PFS was 48.3 months (95% CI, 46.1–58.2)
  • First-year PFS was 84.2% (95% CI, 76.9–91.6)
  • Two-year PFS was 76.3% (95% CI, 67.7–85.0)
  • Three-year PFS was 71.6% (95% CI, 62.3–80.9)
• Independent review PFS values were similar to the ones from the investigator assessment:
  • Median PFS was 46.1 months (95% CI, 46.1–58.2)
  • Three-year PFS was 66.6% (95% CI, 57.0–76.3)
• Two-year OS was 87.5% (95% CI, 80.9–94.1)
• Three-year OS was 85.3% (95 CI, 78.2–92.4)
• Median OS was not reached

COO analysis

• COO analysis was performed for n = 78 patients, of those 58% (n =45) had germinal center B-cell (GCB) DLBCL, 24% (n =19) had activated B-cell (ABC) DLBCL and 18% (n =14) were unclassified
• Response rates and PFS by COO:
  • ORR: GCB, 84.4%; ABC, 78.9% and unclassified, 71.4%
  • CR rate: GCB, 62.2%; ABC, 42.1% and unclassified, 35.7%
  • Three-year PFS: GCB, 75.9% (95% CI, 62.9−0) and ABC, 66.5% (95% CI, 44.5−88.6)

Safety 

• All patients experienced at least one adverse event (AE)
• Fifty-one patients had neutropenia: one Grade 1, one Grade 2, 16 Grade 3 and 33 Grade 4
• Hypogammaglobulinemia was commonly reported in these patients: with IgM depletion happening in 30% of patients, IgA depletion in 9% and IgG depletion in 8%
• The most commonly reported serious AEs (SAEs) were infections (19%), then blood and lymphatic system disorders (17%) and febrile neutropenia despite 85% of patients receiving filgrastim
• IRRs were commonly associated with obinutuzumab treatment. Overall, n = 69 patients (69%) syexperienced 244 IRRs. The majority occurred in cycle 1 (187/244) and were predominantly Grade 2 (129/244). One Grade 4 IRR was reported and there were no Grade 5 IRRs

The phase II GATHER trial demonstrated that G-CHOP is a safe first-line treatment with good efficacy for patients with advanced DLBCL. The authors stated that despite the fact that initial obinutuzumab administration was frequently associated with IRRs, following the first infusion IRRs were uncommon, mild and manageable.