Large granular lymphocytic (LGL) leukemia is an uncommon disease, characterized by a clonal proliferation of mature, post-thymic T-cells, typically CD3+, CD4-, CD8+, CD16+, CD57+ phenotype, representing constitutively active T-cells Less commonly, LGL leukemia is derived from CD3-, CD56+ natural killer (NK) cells. Clonal T-LGLs escape apoptosis by failure to respond to the Fas/Fas ligand (FasL) pathway. Activating mutations in the STAT3gene occur frequently in LGL leukemia, and may play a role in pathogenesis. Autoimmune disorders are frequently associated with LGL leukemia (∼1/3 present with rheumatoid arthritis).

The association between LGL leukemia and B-cell lymphoproliferative disorders has been reported, often with low-grade histologies, but is deemed uncommon and the pathogenesis is not well established. We have analyzed a series of patients (pts) diagnosed with both LGL leukemia or expansion and clonal B-cell disorders.

Patients and methods

Pts with NK or T-LGL leukemia or expansion who were evaluated at Fox Chase Cancer Center or the Cleveland Clinic Taussig Cancer Institute were reviewed, after Institutional Review Board approval. Inclusion criteria were age ≥ 18 yrs and diagnosis of both LGL and B-cell lymphoproliferative disorder.


One hundred and twenty six pts with a diagnosis of T-LGL leukemia, NK-LGL leukemia or T-LGL expansion were identified. Of these, 44 (34.9%) pts were diagnosed with a clonal B-cell disorder. Twenty-six pts (20.6%) were diagnosed with a clonal B-cell disorder concomitantly with or shortly after the LGL diagnosis, 15 of whom presented with monoclonal gammopathy of unknown significance (MGUS) as their B-cell disorder, 9 with monoclonal B-cell lymphocytosis (MBL), 5 of whom also had monoclonal gammopathy. Eighteen pts (14.2%) had a previous diagnosis of clonal B-cell disorder, including diffuse large B cell lymphoma (DLBCL) (N= 6), CLL (N = 3), mantle cell lymphoma (N=3), multiple myeloma (N = 2), Hodgkin lymphoma (N = 2), Burkitt lymphoma (N = 1) and hairy cell leukemia (N = 1). Fifteen pts (11.9%) received treatment prior to the diagnosis of LGL, 10 of them (7.9%) with regimens including rituximab. The median time from completion of last treatment with rituximab to diagnosis of LGL disorder was 33 months. An additional patient with prior DLBCL was diagnosed with LGL shortly after receiving an oral SYK inhibitor.

Two illustrative patients had unexpectedly prolonged remissions of their B cell disorder. A 66 years old man with multiple myeloma who achieved complete remission (CR) after 8 months of bortezomib therapy was then diagnosed with T-LGL, and his myeloma is in ongoing remission now 5 years after T-LGL diagnosis without further therapy. A 66 years old woman with relapsed DLBCL treated with 2ndline immunochemotherapy (R-ICE) for 3 cycles developed lymphocytosis and was diagnosed with T-LGL. With no further therapy, DLBCL is in ongoing remission now 5 years after diagnosis of T-LGL.


We report a large series of patients with both clonal B-cell disorders and LGL. Where diagnosis of B-cell disorder and LGL are concurrent, we hypothesize an underlying immune dysregulation leading to both B-cell and T-cell proliferations. Where B-cell disorder precedes LGL, we hypothesize that the underlying disease and/or its treatment creates the environment for LGL, either directly allowing LGL expansion or permitting persistence of antigens that drive LGL expansion. Most pts with antecedent B lymphoma received rituximab (R), with a median time from R-treatment to LGL diagnosis of 33 months. Late onset neutropenia (LON) has been linked to bone marrow expansion of LGL in patients treated with rituximab, suggesting a possible pathogenetic role in our cases as well. Further, in some pts primary B-cell malignancies unexpectedly entered prolonged remission after T-LGL developed, suggesting a possible anti-B cell immune component of LGL. Further studies are warranted.


No relevant conflicts of interest to declare.

Author notes


Asterisk with author names denotes non-ASH members.