73-Year-Old Male with Stage 0 Chronic Lymphocytic Leukemia with Skin Involvement and No Evidence of Mantle Cell Lymphoma

I have been following a 73-year-old male with Stage 0 chronic lymphocytic leukemia (CLL) for two years. He has a normal hemoglobin and platelet count with a white blood cell count of around 15,000 and cytogenetics showing 13q deletion and trisomy 12. He recently had two small asymptomatic skin lesions removed from his arm and back. These consisted of a clonal population of B cells that expressed Bcl-2 diffusely and were CD20-positive, but CD5- and CD43-negative. This was similar to his original flow cytometry from peripheral blood. This case was interpreted by a referral lab as B-cell lymphoma. The patient is completely asymptomatic with no evidence of liver or spleen enlargement or adenopathy on CT scans. I’m not sure how to put this all together; is this atypical CLL with skin involvement? There is no evidence of mantle cell lymphoma.

This case presents two unusual aspects of CLL. The first is that this patient has small skin lesions that appear to be deposits of CLL. The second unusual feature is that this patient’s cells are CD5-negative. Looking at the whole picture, I would agree that this is most likely atypical CLL with skin involvement.

I will discuss the phenotype first. CLL is a B-cell malignancy that also expresses CD5. The other well-known lymphoid malignancy that expresses CD5 in addition to B-cell markers is mantle cell lymphoma, which can present in a leukemic phase. Typically, the morphology of mantle cell is significantly different from CLL, although in less classical cases, there may be some blurring of the morphology. CD23 is a useful marker to differentiate them, as this antigen is present on CLL cells and not present on mantle cells. However, in this case we have a patient who appears to have CLL but is CD5-negative.

Other B-cell lymphoproliferative disorders are usually CD5-negative, and the ones that could potentially overlap with CLL are prolymphocytic leukemia (PLL), marginal-zone lymphoma/leukemia (MZL), and follicular lymphoma in a leukemic phase. Again, morphologies differ between all these diagnoses, but there can be overlap. The most common overlap is between CLL and MZL. That’s because classically the diagnosis of marginal zone is, as the name implies, based on pathologic examination of lymph nodes. Patients presenting with leukemic disease without significant adenopathy render this diagnosis more difficult to ascertain. A group of investigators in France evaluated 165 leukemic patients with B-cell lymphoproliferative disorders selected on the basis of a score of ≤3 in the Royal Marsden Hospital scoring system, which implies atypical cases that don’t fit clearly into one diagnostic entity.¹  Fifty-four cases were CD5-negative. Morphological and phenotypical examinations were able to reveal the diagnosis in 24 of these, including the most common diagnosis of MZL in 11 cases. Thirty CD5-negative cases still could not be fully characterized after morphological and phenotypical examination. Only five of these patients had superficial lymphadenopathy, and in three cases where lymph node biopsy was done, the diagnosis of MZL was established. All in all, 14 were diagnosed as MZL, which was the most common diagnosis, and eight patients remained “unclassified.” These are likely the patients we would call atypical CLL.

From a practical point of view, the good news is that the approach to MZL is similar to that of CLL — namely watch-and-wait in early-stage asymptomatic patients. One difference, however, is in the approach to treatment. MZL cells have very strong expression of CD20 (as opposed to CLL cells), and these patients have dramatic responses to single-agent rituximab. In patients who need treatment, I generally use 375 mg/m2 rituximab weekly for eight weeks and have had patients in complete remission for years after this single agent, whereas the efficacy of single-agent rituximab in CLL is much weaker. The reason I would diagnose CLL in this case is the presence of a 13q deletion and trisomy 12. These are well-known abnormalities seen in CLL, and, in fact, the presence of trisomy 12 is commonly associated with atypical features.

Skin involvement in CLL is quite rare. Probably the largest series to examine this came from a group of pathologists in Austria, who describe cutaneous infiltrates of CLL in 42 patients. The five-year survival of these patients was 66.6 percent, suggesting that they had a good, if not a more favorable, prognosis compared to those without skin involvement; there was no attempt to match them for comparable prognostic features to patients without skin involvement.^2,3 

I’ve been following some patients for years with skin involvement. In particular, I’m thinking of one who has small deposits on both earlobes and has very indolent CLL that’s never required treatment. If these lesions are in a problematic area, low-dose radiation might be an option, but generally this is not an indication for systemic therapy, and it can simply be monitored.