Wednesday, April 22, 2009

CLL differential diagnosis

The diagnosis of CLL is easy. The cells are very characteristic on the blood film. They have dark staining nuclei that takes up most of the cell with a thin rim of cytoplasm surrounding. Very little nuclear detail can be seen and the cells are about the same side as normal lymphocytes - indeed it can be quite difficult to distinguish the leukemia cells from the normal lymphocytes that we know are there.

I have already written at length about MBL, but it comes into the differential diagnosis. The distinction is made on the B-cell count in the blood. If it equal to or greater than 5000 per microliter it is CLL; if it is less than 5000 per microliter it is MBL, unless there is evidence of lymph node enlargement, when the diagnosis is small lymphocytic lymphoma (SLL). It is easy enough to feel in the neck, armpits (ugh!) and groin (double ugh!) for lymph nodes, but the lymph nodes might be at the back of the tummy unreachable by the hand and only diagnosable by ultrasound or CT. Now MBL is hundreds of times commoner than CLL and you can't do a CT scan on every case of MBL on the remote prospect that it might be SLL, so one would only do this if there were some other reason to suspect SLL, like weight loss, fever, night sweats or severe fatigue.

The definitive diagnosis of CLL is by immunophenotyping. This is done with a flow cytometer which is able to measure how well certain antibodies react with the cells and therefore whether the cells have certain antigens on them.

The most important antigens are: CD19 - which is present on all B cells so it establishes that we are not dealing with a T cell tumor. Don't laugh, some cutaneous T cell tumors and even some T-PLLs can resemble CLL down the microscope and I have been fooled and so have many others.
CD5 - which is present on T cells and surprisingly on two types of B cell tumors: CLL and mantle cell lymphoma? So this is pretty good at separating CLL from most other B cell tumors.
CD23 - which is present on some B cells, but of B cell tumors it is pretty well confined to CLL. Unfortunately, it decays on storage so it isn't very good for going back to stored material that was frozen down before we knew about CD23.
Surface Ig - which is present on all B cells but strangely on CLL cells there is only about 10% of the normal amount. This is very helpful in distinguishing between CLL and mantle cell lymphoma where the surface Ig is if anything more than normal.
CD79a - this is one of the accessory molecules that attaches itself to surface Ig and forms part of the B cell receptor. Like surface Ig it is usually at a very low level on the surface and sometimes cannot be detected at all. This is quite unlike the situation in other B cell lymphomas, and especially in the case of mantle cell lymphoma.
CD20 - this is the rituximab target. This is less easy to detect on CLL cells than on other lymphomas, which may be why rituximab works less well in CLL than in other lymphomas. It is probably there all right, but we think that the lipid conformation of the cell membrane hides it.
FMC7 - this is part of the CD20 molecule and it is the part that is hidden in CLL so CLL cells are usually negative or very weak for FMC7.

Atypical CLL

This was originally a diagnosis made on the blood film. You can see from this picture that there are some larger cells with more cytoplasm here and that they have a pale circle in the nucleus (called a nucleolus). These are prolymphocytes. when there are a lot of these (over 15%) or a lot of other cells with notches in their nuclei (sometimes called baby bottom cells) the diagnosis was called atypical CLL. Nowadays, the term is usually applied if the flow cytometry result is not really typical. The surface Ig and CD79a might be a bit more obvious than usual or CD20 might be brighter and FMC7 moderately strong. CD5 might be a bit weak and sometimes CD22 comes up even though it is usually negative.

Atypical CLL is usually just CLL with slightly atypical markers. Often the reason is that the chromosomal pattern is trisomy 12 or del 11q. In these the prognosis belongs to the chromosomes. The most atypical patterns are seen with the very rare t(14;19) translocations.

CLL/PLL is another form of CLL but one that is much understood. The second picture shows what it looks like - so it is a kind of atypical CLL. In CLL you are allowed to have up to 15% prolymphocytes present. If there are between 15% and 55% then the diagnosis is CLL/PLL. More than 55% means a diagnosis of B-PLL (B-cell prolymphocytic leukemia)

Many people think that finding CLL/PLL means that the CLL is turning into PLL. This simply is not so. CLL never turns into PLL. So what does it mean if you find CLL/PLL? Contrary to the common perception, in the original study of 55 cases of CLL/PLL that defined the condition, half showed a stable picture without a progressive increase in prolymphocytes. The prognosis of this group was similar to that of stable CLL without prolymphocytes. In one third of cases the increase in prolymphocytes was unsustained and probably represented a reaction to an infection or vaccination. In only 18% was there a definite progression towards a more malignant phase of the disease. It is this last 18% in which the term 'prolymphocytoid' (not 'prolymphocytic') transformation can be applied. It is often associated with the acquisition of a malign chromosomal change like del 11q or del 17p.

Mantle cell lymphoma

This is a much more malignant disease than CLL and in this picture the cells look quite different. It is the other B-cell tumor that is CD5 positive, but the other markers are quite different and it has some special features of its own. It virtually always carries the chromosomal translocation t(11;14) which engineers the appearance of the protein cyclin D1 in the cell nucleus. About 80% of mantle cell lymphomas have tumor cells in the bone marrow and these often appear in the blood. Mantle cell lymphoma is a great mimic and the cells can look like CLL cells of PLL cells or even marginal zone lymphoma in the spleen. It is always important to investigate cases with atypical markers to se if they might not be mantle cell lymphoma cells. Interestingly some of the cases that resemble CLL are very indolent and may behave more like CLL than mantle cell lymphoma.

This is already a very long blog, so I will finish it off in a second post later today or tomorrow.


Anonymous said...

Hasn't CLL/PLL been done away with in the new WHO?

Terry Hamblin said...

CLL/PLL was never a separate diagnosis, but was a description of some cases of atypical CLL (which still has a passing reference in the CLL article in WHO). The definitive work on CLL/PLL came from Junia Melo and David Galton at the Hammersmith 20 years ago and there has been nothing definitive since then. The new WHO makes it clear that CLL transforming to PLL is exceedingly rare (I would say it never occurs). CLL/PLL is not a very useful distinction except to disparage the idea that CLL transforms to PLL.