Obinutuzumab (GA101) in Combination with CHOP (Cyclophosphamide, Doxorubicin, Vincristine and Prednisone) or Bendamustine for the First-Line Treatment of Follicular Non-Hodgkin Lymphoma: Final Results from the Maintenance Phase of the Phase Ib GAUDI Study

Obinutuzumab (GA101; G) is an anti-CD20 monoclonal antibody with activity in relapsed/refractory follicular lymphoma (FL) as a single agent and in combination with chemotherapy. G-chemotherapy induction followed by G-maintenance has not been evaluated for untreated FL. The open-label, randomized, phase Ib GAUDI study (NCT00825149) investigated safety and efficacy of G + CHOP (G-CHOP) or bendamustine (G-B) as first-line treatment for FL. We report data from patients (pts) who received G-maintenance after responding to G-chemotherapy induction.

Pts with treatment-naïve CD20+ B-cell FL and ≥1 bi-dimensionally measurable lesion (CT scan; largest dimension >1.5cm) were allocated on a per-center basis to receive induction G IV 1000mg + standard CHOP (3-weekly, 6–8 cycles) or B IV 90mg/m²(4-weekly, 4–6 cycles). Induction responders received G-maintenance every 3 months for 2 years or until disease progression (PD). Pts completing maintenance were followed for a further 2 years, until PD or start of new anti-lymphoma therapy. Anti-infective prophylaxis was used at investigator discretion. The primary objective was safety. Secondary objectives included progression-free survival (PFS) and response rates.

The overall safety population comprised 81 pts (G-B: n=41; G-CHOP: n=40). Baseline characteristics (age, sex, FLIPI status, bone marrow involvement, bulky disease [lesion ≥7cm], time from diagnosis) were balanced between arms. Median observation time from study start was 31 months (G-B) and 33 months (G-CHOP).

The maintenance safety population comprised 72 pts (n=36 from each induction therapy arm). Most pts completed maintenance (G-B: 81%; G-CHOP: 72%). There were 17 discontinuations (24%; G-B: n=7; G-CHOP: n=10), due to an adverse event (AE)/intercurrent illness (n=9), insufficient therapeutic response (n=5), administrative/other (n=2) and death (n=1).

During 2 years’ maintenance most pts had AEs: G-B, 100% (44% grade ≥3); G-CHOP, 78% (31% grade ≥3). The most common AE (all grades) was cough (G-B: 17%; G-CHOP: 11%). The grade ≥3 AEs mainly reflected infections and cytopenia. Six G-B pts (17%) experienced 7 grade ≥3 infections; 4 were considered treatment related (genital infection, oral herpes, pneumonia klebsiella, neutropenic infection). Five G-CHOP pts (14%) had one grade ≥3 infection each; 4 were considered treatment related (viral meningitis, respiratory tract infection [RTI], bacterial pneumonia [2 events]). Six G-B pts (17%) experienced 10 grade ≥3 cytopenia AEs; 7 were considered treatment related (anemia, febrile neutropenia, pancytopenia, neutropenia [2 events], thrombocytopenia [2 events]). No G-CHOP pt experienced a grade ≥3 cytopenia AE. AEs led to dose delays in 17% (G-B) and 6% (G-CHOP) of pts. Three pts (G-B: n=1; G-CHOP: n=2) had treatment-related AEs during, or within 24 hours of, an infusion (all grade 1–2). Two deaths (both G-CHOP) occurred during maintenance or maintenance follow-up; 1 due to PD and 1 due to a G-related AE (RTI leading to fatal lactic acidosis).

At the end of maintenance, all pts with data available (G-B: n=41; G-CHOP: n=39) had experienced B-cell depletion (<0.07x10⁹ cells/L). At data cut-off, only 22 pts (28%) had a B-cell assessment within 6–9 months of follow-up after the end of maintenance treatment; all remained B-cell depleted. Six of 12 pts (G-B: n=2; G-CHOP: n=4) assessed between 9 and 24 months after the end of maintenance recovered (>0.07x10⁹cells/L). Median IgG levels remained within normal range during maintenance.

In the overall safety population, complete response (CR) rate (based on CT scan rather than PET) as best overall response increased from end of induction (G-B: 37%; G-CHOP: 35%) to end of maintenance (G-B: 61%; G-CHOP: 70%). PFS rate at 32 months after first study drug was 92% (G-B) and 84% (G-CHOP). Median PFS was not reached; 10 pts (G-B: n=4; G-CHOP: n=6) had PD, including one transformation to diffuse large B-cell lymphoma.

G-maintenance after G-chemotherapy induction was associated with a high CR rate in pts with previously untreated FL. Opportunistic infections occurred infrequently. Clinically relevant neutropenia was experienced by 14% of pts who received G-B induction but was not observed in G-CHOP pts. A phase III trial (GALLIUM) is investigating G versus rituximab in chemoimmunotherapy induction followed by immunotherapy maintenance in pts with untreated indolent non-Hodgkin lymphoma.