A Phase I/II Study of RO5072759 (GA101) in Patients with Relapsed/Refractory CD20⁺ Malignant Disease

RO5072759 (GA101) is the first humanized and glycoengineered monoclonal anti-CD20 antibody to enter clinical trials. Glycoengineeering results in a significantly increased antibody-dependent cytotoxicity (ADCC) compared to rituximab as shown in in vitro models. Additionally, RO5072759 binds with high affinity to a type II epitope on CD20 and is characterized by reduced complement-dependent cytotoxicity (CDC) and strongly enhanced direct cell death compared to type I antibodies. In preclinical in vivo lymphoma models RO5072759 has shown superior efficacy compared to rituximab.

In this Phase I/IIa study RO5072759 was administered as a single agent to patients with CD20⁺ malignant disease for whom no therapy of higher priority was available. The aim of which was to determine the safety and tolerability of RO5072759, any dose-limiting toxicity (DLT), its pharmacokinetics and to establish the recommended phase II dose. Patients were treated with RO5072759 by intravenous infusion (premedication with acetaminophen and anti-histamines) administered as a flat dose on days 1, 8 and 22 and subsequently every 3 weeks for a total of 9 infusions. The dose of the first infusion was 50% that of subsequent infusions. The dose was escalated based on the safety in a 3+3 design.

Since September 2007 twenty four patients have been treated with RO5072759 at doses from 50 mg to 2000 mg. Complete data, presented herein, is available on the first 12 patients, median age 59 yrs (39–83), from the first 4 cohorts (50 mg–800 mg). Most patients had follicular NHL (9) others include DLBCL (1) CLL (1) Waldenstrom’s macroglobulinemia (1). All were previously exposed to rituximab and had received a median of 4 (range 1–7) prior regimens. RO5072759 was well tolerated with no DLTs observed. The most common adverse events are Grade 1 or 2 (CTCAE V3.0) infusion related reactions, characterized by fever, chills, hypo/hypertension, nausea and vomiting limited in the main, to the first infusion. These responded well to slowing or interruption (5 pts) of the infusion and steroids (1 pt). No tumor lysis syndrome has been observed. Only 6 minor infections (4 upper respiratory track, 1 urinary track and 1 oral herpes) were observed to-date. Measurement of plasma cytokines and complement during and immediately after the first infusion showed an increase in IL6 and IL8 with a smaller increase in IL10 and TNFα. No change in complement fractions (C3, C3a, C4a, C5, C5a, Bb) was observed. Concurrent to cytokine increase was apparent a decrease in T-cell subsets (CD3, CD4 and CD8 subsets) and NK cell counts in the peripheral blood. This decrease, as with the cytokine increase, was transient and levels were at or near baseline by day 8. No similar changes, for the majority of patients, were seen with subsequent infusions. Circulating B-cell (CD19) depletion occurs rapidly and is sustained. The pharmacokinetics of RO5072759 are broadly similar to those of rituximab and show a dose-dependent increase in exposure, but with significant inter- and intra-patient variability. Time dependent clearance was noted which is consistent with a reduction in target-mediated antibody clearance with increasing duration of treatment. To-date 7 of 12 patients have responded by day 85 to treatment [3 CR (25%), 4 PR (33%) ORR (58%)], 1 patient has SD at 8 months, 3 patients have PD (2 have since died) and 1 patient died from an event unrelated to treatment (cerebrovascular accident). Responses have occurred at all dose levels.

In conclusion, RO5072759 is a novel ADCC-enhanced type II anti-CD20 antibody that has shown a similar safety profile to rituximab and promising efficacy in this difficult-to-treat patient population. Dose finding is continuing and data from all 24 patients will be presented.