The Immune system

B-Lymphocytes have one purpose in life, and that is to make antibodies. We are supplied with a full set of B cells, already pre-coded to make all the antibodies that we could possible need. But these B cells are in a sort of virgin condition until they meet their soul-mate - the germ (it's usually a germ, but it could be a different sort of foreign protein) that they are designed to react with.

That meeting starts a process that refines the immune process, so that the best fitting antibody that it is possible to make is produced. This refining takes place in a germinal center (and yes I'm using the American spelling because more people speak American than speak English). A germinal center is located in lymph nodes (or the spleen) and is sort of egg shaped. It contains T-cells, follicular dendritic cells, and B-cells. A complicated process takes place here which results in the shape of the antibody being made being subtlely altered so that it fits its target more snugly. This process involves changing the DNA code by 'somatic mutation' and altering the type of antibody from IgM (which is the type that resides on the surface of the lymphocyte, and although it can be secreted, it cannot escape from the blood stream into the tissues) into IgG which is the usual secreted form and which can recruit all the mechanisms needed to kill germs (an incidentally tumor cells, because rituximab is an IgG antibody).

This 'antibody refinement' is a normal process, absolutely necessary for the defence against germs. It is well controlled process that clearly has a mechanism for switching itself off when the immune response is done with.

When I discovered that more than half of CLL cases had these somatic mutations, my immediate thought was these cells must have passed through the germinal center to acquire them. This was heresy at the time, but it turns out to be true. The other half of CLL cases that had no mutations I thought must be tumors of virgin B cells, but this turns out not to be true. There is now incontrovertible evidence that these cells too have met the target, but this has not caused somatic mutations, or at least if it has it has only caused a couple.

We now know that there is another 'back-up' way of making antibody. Some people are born with a missing protein called CD40 ligand. These people are unable to make germinal centers. They can make antibody, however, and even small numbers of somatic mutations. So there is an 'extra-germinal center' way of encountering antigen, and our current thinking is that the unmutated subset represents tumors of cells that have met antigen in this way. This encounter takes place without the normal controls that would be there in a germinal center. And this is why, I believe, the cells behave in a more outrageous way. In my lectures I liken one circumstance (within the germinal center) to the marriage bed and the other to a good-time girl with less conventional habits.

As for trisomy 12, we have evidence that this is seen in both the mutated and unmutated subsets, but the prognosis is much better for the mutated subset. We have no evidence of a difference in response rates, however.

It's a bit more complicated than that. It is not that those without mutations lack germinal centers. It seems that there are two pathways to antibody production. The conventional way is for B cells to go through germinal centers and mutate. Some B-cells don't go that route and make antibodies without entering germinal centers and therefore don't mutate. This pathway was discovered in a rare group of patients who lack CD40 ligand and can't make germinal centers, but we all have it.

A tumor can arise in B-cells that arise from either pathway. There are some antibody targets that are so-called T-independent antigens. For example the cell walls of pneumococcus bacteria. Antibodies made against these don't need to go via germinal centers. The people who discovered the importance of CD38 in CLL, Nick Chorriazi and Malio Ferrarini, think that the unmutated type of CLL is a tumor of a B-cell participating in one of these T-independent immune responses. For a number of complicated reasons we disagree about this, but their explanation fits some of the facts.

There are ways of making these cells mutate in a test tube, but that wouldn't affect how the cells got where they are, or the particular nature of the tumor.

A ligand (sometimes pronounced to rhyme with fig-and, and sometimes pronounced with a long I as in like) is simply something that binds to a receptor. CD40 is a receptor on B-cells that when stimuated causes the B-cell to grow and divide. CD-40 ligand (now renamed CD154) is present on activated T-cells and it is this T-cell/B-cell interaction that causes B-cells to grow an divide in the germinal center.

CD40 ligand is being used in a form of immunotherapy against CLL by Tom Kipps's group in San Diego. The idea is that they genetically engineer the CLL cells by puting CD40 ligand ionto them. this makes them act as 'antigen-presenting cells' that can present characteristic (but unknown) CLl antigens to the patient's own T-cells an dtyhis will make ythem attack the CLL cells and kill them.

How would I cure CLL with 5 million pounds? It's not shortage of money that is holding us up. It takes time in a relatively slowly progressive disease. The first thing I'd do is recognise the natural history of CLL, and understand that many cases don't need curing. The problem with these is that some of them will run into problems with immunodeficiency, and therefore I would be looking at ways of restoring the immune system. The next thing is to stratify all clinical trials so we don't flatter ourselves with good results in patients that would do just as well with no treatment. Concentrating on the unmutated subset, I would investigate whether early treatment with all guns blazing (FCR followed by CAMPATH to clear up minimal residual disease) was indeed beneficial. I would follow up the significance of the proteins that are differently expressed between the mutated and unmutated subsets by microarrays.