Thursday, April 23, 2009
CLL: differential diagnosis part II
B-PLL is a very rare disease and some people think that it doesn’t exist as a single entity. The picture shows what it looks like down the microscope and the definition says there have to be more than 55% prolymphocytes. It is said to be commoner in men and always associated with an enlarged spleen, usually without enlarged lymph nodes, and a very high white count. Immunophenotyping is uncertain. There seems to be confusion as to whether it is CD5 positive or not. In my experience all those cases that were CD5 positive also had the t(11;14) translocation, making them a blastic phase of mantle cell lymphoma. They all had a very bad prognosis. The other cases seem to be a hodge-podge of atypical B cell tumors. But none are transformations from CLL in my experience. I wouldn’t rule out that a transformed CLL with the t(14;19) might appear to be PLL, but I have never seen a case of this very rare tumor do this, and there is some reason to say that this is not really CLL at all.
Marginal zone lymphoma
There are several types of marginal zone lymphoma, but the one that is usually confused with CLL is splenic marginal zone lymphoma, SMZL (sometimes called splenic lymphoma with villous lymphocytes, SLVL). The picture shows why. Little projections are shown coming from the cell surface, particularly at the end where most of the cytoplasm is. However, these don't always show on many blood films and again it is the immunophenotyping that gives it away. SMZL is positive for the B cell markers, CD19, CD20, CD22, bright surface Ig and CD79b, and negative for the CLL markers CD5 and CD23. Of course there are always atypical cases that are weakly CD5 positive and even perhaps weakly CD23 positive, so these can be some confusion. Clinically, the patients have enlarged spleens without enlarged lymph nodes, and they usually have a very indolent disease which can be put into along term remission by splenectomy. Some patients have both CLL and SMZL together (don't ask me why) and careful immunophenotyping can recognise the two separate diseases.
A large proportion of follicular lymphoma patients have some minor involvement in teh blood. The number of cells seldom reaches the levels seen in CLL, and the disease is much more prominent in lymph nodes. The cells shown here have even less cytoplasm and the nucleus often shows a cleft (they are sometimes called 'baby-buttock cells). Despite the difference in shape to most CLL cells, cells with a cleaved nucleus can be seen in some cases (I had a patient with a very benign CLL whose clefted cells were very prominent throughout her disease). The majority of follicular lymphoma cells have the t(14:18) chromosomal translocation, but this can also be found in a small minority of true CLLs. The best distinction is again the immunophenotyping which shows the same pattern as in marginal zone lymphoma with the addition of positivity for CD10.
Polyclonal B cell lymphocytosis.
This is a very strange condition almost confined to female smokers. It is probably not a malignant condition, but I know of one unpublished case that underwent a Richter-like transformation. On the blood film the characteristic finding is of binucleate cells. Immunophenotyping shows that it is polyclonal with populations of cells staining for kappa and for lambda light chains.
I have already mentioned that sometimes immunophenotyping can recognise two populations of B cells - one from a CLL clone and one from a marginal zone lymphoma clone. Even more frequently, two CLL clones may be detected. This can be picked up on immunophenotyping (especially if one stains with anti-kappa and one with anti-lambda, but sometimes this does not pick it up and one has to resort to immunoglobulin gene sequencing to demonstrate that two different heavy chain genes are being used. Sometimes Richter's transformation occurs in a B cell that was not part often original CLL clone. It is not clear whether this was in an entirely innocent B cells that was attacked by a virus in the immunocompromised patient, or from a minor clone of CLL cells that had not been previously detected.
CLL can occur with other tumors of B cell origin which may cause confusion. Myeloma, Waldenstrom's macroglobulinemia and Hodgkin's disease are well recognised associations.
In conclusion, immunophenotyping is an essential part of the diagnostic work-up, and no-one should be satisfied with a diagnosis of CLL just made by looking down the microscope.