On May 4th this year the UKCLL Forum held a meeting in London. The topic was "Monoclonal Antibodies in the Treatment of CLL" I have written a report of the meeting and the first part is reproduced here.
Monoclonal antibodies were one of the first attempts at targeted therapy. Two monoclonal antibodies, rituximab (anti-CD20) and Alemtuzumab (anti-CD52) are currently widely used in the treatment of CLL. Lumiliximab (anti-CD23), and HuMax (a fully humanised anti-CD20) are in clinical trials. Although the response to single agent Rituximab in previously treated patients is only 13%, in untreated patients it is 51%. The response in untreated patients to single agent Alemtuzumab is even better at 87% and patients whose disease shows deletions of p53 respond well even though this defect defeats most other therapeutic agents, but Alemtuzumab also kills T cells and engenders immunodeficiency. Used in combination with chemotherapy, the response to rituximab approaches 100%, though the use of chemotherapy detracts from the specificity of the targeting.
According to Professor Martin Glennie, University of Southampton, CD20 is in many respects the ideal target. Compared with most other antigens on the surface of CLL cells antibodies that target it are much more effective killers. It is believed that rituximab kills in three different ways: 1] antibody dependent cellular cytotoxicity (ADCC) in which macrophages, NK cells and possibly granulocytes recognise the Fc portion of cell-bound antibody through their FC receptors, and this invokes killing and ingestion of the CLL cell; 2] complement dependent killing (CDC) – either tumbling the complement cascade right up to the membrane-attack-complex that punches a hole in the CLL membrane or, more likely, opsonizing the cell so that a derivative of the third component of complement, C3b, sits on the cell surface to be recognized by the C3b receptor of macrophages; 3] direct killing by cross-linking adjacent CD20 molecules into lipid rafts which sends a signal through the cell to self-destruct, either through apoptosis or through some other, as yet unclear, process. It is probably true that all three mechanisms play a part in real life, but experiments have revealed two quite distinct types of anti-CD20s. Most are rituximab-like and kill by ADCC and CDC – these are known as type I anti-CD20s – type II antibodies like B1 the original antibody produced in Boston, and the one used in Bexxar, kill by a mixture of ADCC and direct killing. The acute reactions seen after rituximab are cause by the release of C3a, an anaphylactoid, after the fixation of complement, but these also act as chemo-attractants for effector cells.
Dr Glennie also spoke about HuMax, a fully humanized anti-CD20 developed in his lab and currently undergoing clinical trials. This is a type I antibody, but it has certain advantages over rituximab. It is directed against a slightly different portion of the CD20 molecule from rituximab, and a slower off-rate (it stays on the cell longer) so that it is more effective at killing cells with lower quantities of CD20 on the surface.
Dr Josee Golay from Bergamo, Italy stressed the importance of complement in rituximab killing. In animal models lacking complement no killing occurred. Other models emphasize the importance of effector cells, both NK cells and macrophages being required. She went on to stress the importance of the interaction of various killing mechanisms. Fixation of anti-CD20 leads to the upregulation of the chemokines CCL3 and CCL4, which attract and activate a number of cell types including mononuclear phagocytes, NK cells, T cells and granulocytes through interaction with their CCR1 and CCR5 receptors. Which killing mechanisms are predominant depends on which tumor is being attacked and in which body compartment – blood, marrow, spleen or lymph nodes.
Dr Ron Taylor from the University of Virginia, Charlottesville also stressed the importance of complement fixation in rituximab-induced killing and added the information that complement could be consumed. Post-rituximab-infusion sera lacks complement activity, principally through consumption of the C2 component of complement. Although this might be overcome by the infusion of fresh plasma or of recombinant C2, a more important barrier to successful immunotherapy was also revealed. Antigen shaving is another mechanism for antigenic modulation. Antigenic modulation was seen as a potential barrier to antibody therapy in the 1980s. Fixation of antibody led to movement of the antigen in the lipid cell membrane – patching and capping – which in turn led to pinocytosis (internalization) of the antigen/antibody complex so that the cell became naked of antigen. One of the reasons that CD20 was chose was that it is not subject to antigenic modulation. However Dr Taylor has shown that after exposure to quite small amounts of antibody CLL cells are cleared from the circulation, to shortly reappear, this time bereft of surface CD20. The antigen had not been internalized, but phagacytosed by macrophages. It appears that the capacity of macrophages to ingest antibody coated tumor cells is soon exhausted and thereafter as the antibody binds to the macrophage Fc receptor, the most that can be accomplished is the stripping of the immune complex from the tumor cell – antigen shaving – and its ingestion by the phagocytic cell.
The choice of 375 mg per square meter as the dose of rituximab was quite arbitrary, dependent on the amount of antibody available and the number of patients needed to be treated. It may be that most of this antibody is wasted. Dr Taylor suggested that smaller doses given fairly frequently might be more effective. A dose between 20 and 60 mg was suggested, given once or twice a week
"Antigen shaving" is an interesting finding.
ReplyDeleteRituxan has been used in varying dosages, and it has generally been thought that higher doses provide superior responses. (See 2005 ASH abstract 5034 "High Dose of Rituximab in B-Cell Malignancies" and 2003 ASH abstract 5164 "High Dose Rituximab in Chronic Lymphocytic Leukemia"). It has been suggested that since CD20 is found in the serum, and may 'sop up' Rituxan, higher doses would overcome the loss due to free-floating CD20.
Rituximab 'maintenance' is an occasionally-tried therapy, though most people ultimately fail, and require alternative drugs.
This idea should be tested in a clinical trial.
The idea that complement can be depleted in rituxan treatment is also not new. There was a paper appearing at the 2005 ASH meeting entitled "Successful Induction of a Rapid Improvement of Both Clinical
and Laboratory Parameters in a Patient with Advanced CLL by Combining
Rituximab with Fresh Frozen Plasma as a Source for Complement. A
Novel Therapeutic Approach?" It was abstract number 5030.
The finding that CD20 can be found in the serum has not been confirmed and most workers now believe that this finding was in error. The paper on very high doses of rituximab giving higher response rate was raised at the meeting, and Dr Taylor told us that he found that paper unconvincing.
ReplyDeleteA clinical trial of low dose rituximab has been suggested though one could hardly expect it to be supported by the pharmaceutical companies. Perhaps third party funders would be interested?
In 1980 my colleagues and I demonstrated that treatment of CLL with anti-idiotype consumed complement and improvement could be achieved by infusions of fresh frozen plasma, so as you say the idea is not new. I presented these data in my talk at the meeting.
Here's a quote from another blog:
ReplyDelete"Dr. Weirda mentioned that by increasing the Rituxan dosage by 10 times, the response rates go up to 80%."
I think there is more than just one source for the idea that rituxan gives more complete reactions the higher the dose.
One of the original papers on Rituxan showed that response rates were poor in CLL, and responses went up when dosages were increased.
UCSD (Kipps) has a trial open (or soon to open) that has steroids at high doses and rituxan at higher doses, in hopes of getting a better result.
Is he just plain wrong?