Roche’s GA101 (obinutuzumab): Engineering an antibody to beat Rituxan
Survival rates for people with B-cell driven blood cancers, such as non-Hodgkin’s Lymphoma and chronic lymphocytic leukemia, have vastly improved in the last decade thanks to the introduction of Rituxan, marketed by Biogen Idec and Genentech. But the drug, a chimeric monoclonal antibody targeting CD-20, a protein that sits on the surface of B-cells, has its limitations: not all patients respond at first, and others become resistant to the drug over time.
As a result, companies are tinkering with the sugar molecules that decorate antibodies in hopes of coming up with a drug that binds better to its target and, ultimately, is more effective at battling cancer. At the American Society of Clinical Oncology annual meeting, held earlier this year in Chicago, Roche offered Phase III data showing its glycoengineered antibody GA-101 worked better than Rituxan at delaying the progression of CLL. If all goes well with FDA, the drug could be approved by the end of the year.
Although the CD20 antigen is expressed on both normal and malignant cells, it has proven to be a useful target therapeutically. Rituximab, ofatumumab and most of the anti-CD20 antibodies in earlier development are Type I monoclonal antibodies, which means that they have good complement-dependent cytotoxicity (CDC) and Ab-dependent cell mediated cytotoxicity (ADCC), but are weak inducers of direct cell death.
In contrast to Type I monoclonal antibodies, next generation monoclonals are increasingly Type II, such as GA101 (obinutuzumab) in CLL and NHL and mogamulizumab (anti-CCR4), for T-cell leukemias and lymphomas. They have little CDC activity, but are much more effective at inducing ADCC and also direct cell death, at least based on in vitro studies performed to date.
How does glycoengineering make a difference?
Glycoengineering is the term used to refer to manipulation of sugar molecules to improve the binding of monoclonal antibodies with immune effector cells, thereby increasing ADCC.
Obinutuzumab is a very different molecule from rituximab, in that it is a novel compound in its own right (originally developed by scientists at Glycart before being bought by Genentech). It is not a biosimilar of rituximab. It is also a glycoengineered molecule designed specifically to improve efficacy through greater affinity to the Fc receptor, thereby increasing ADCC activity.
The overall intent with the development of obinutuzumab was to significantly improve efficacy over rituximab and Type I monoclonal antibodies in B-cell malignancies using glycoengineering techniques.
At the recent ASCO annual meeting, data from a phase III trial was presented to evaluate rituximab or obinutuzumab in combination with the chemotherapy chlorambucil versus chlorambucil alone in newly diagnosed CLL. Patients elderly and had co-existing co-morbidities, excluding them from standard chemotherapy with fludarabine and cyclophosphamide (FC).
This two part trial sought to compare both combinations to the chemotherapy initially, and then against each other in a head-to-head comparison once the survival data matured in the second phase. Data from the first phase of the study was reported at this meeting.
What did the results show?
When looking at the response rates, both obinutuzumab and rituximab combinations had a higher overall response rate (ORR) than chemotherapy alone (75.5% and 65.9% vs. 30.2% and 30.0%). Importantly, the combinations had a great proportion of complete responses (CR) i.e. 22.2% and 8.3% compared to 0% in the chlorambucil arms.
Minimal residual disease (MRD), a measure of the number of leukemia cells remaining in the blood, was 31.1% in the peripheral blood of the obinutuzumab combination compared with 0% in the chemotherapy arm. Corresponding values in the rituximab and chlorambucil arms were 2.0% and 0%, respectively.
Median progression-free survival (PFS) i.e. the length of time during which people lived without their disease worsening for the obinutuzumab plus chlorambucil arm were impressively higher than chemotherapy alone. PFS was more than doubled (23 months compared to 10.9 months, HR=0.14, p <.0001) when compared to chlorambucil alone. The corresponding outcome data for the rituximab combination were 15.7 versus 10.8 months for chlorambucil alone (HR=0.32, p <.0001).
Since ASCO, Roche have announced that the FDA granted Priority Review for obinutuzumab in CLL (in addition to the Breakthrough Designation already received in May, when the company filed a new drug application for obinutuzumab), meaning that the PDUFA date is set as December 20th. In addition, the Data Monitoring Committee decided that the interim data analysis was sufficient to meet the primary endpoint of the trial, ahead of schedule. The data confirms that obinutuzumab was superior to rituximab in terms of the disease worsening (PFS). The full data will be presented at ASH in December, when overall survival data (ie did the patients live longer) may be available.
The adverse event profiles were slightly different between the monoclonal antibodies. Patients in the obinutuzumab arm experienced more infusion site reactions, and a slightly higher degree of myelosuppression (thrombocytopenia and neutropenia), but lower infection rates.
The study demonstrated that both obinutuzumab and rituximab were more beneficial to elderly patients living with CLL and co-existing medical conditions than chemotherapy alone. The final head to head analysis of the two combinations will be available once the second stage of the study has mature data. Based on the progress to date, the signs are very encouraging that the chemical engineering behind the development of obinutuzumab may potentially have produced a superior compound to rituximab for treatment of B-cell malignancies.
Should the mature outcome data show a positive survival advantage in obinutuzumab’s favour over rituximab, we may well see similar glycoengineering techniques applied to other monoclonal antibodies in the near future, potentially leading to further improvement in outcomes.
Comments are closed.