Autoimmune cytopenias are amongst the more troublesome aspects of the management of patients with B-chronic lymphocytic leukaemia (B-CLL) (Hamblin, 2001). Impaired immunity is well recognised in this condition (Bartik et al, 1998) and this baseline anergy may be further impaired by treatment with fludarabine or other immunosuppressive agents. Impaired T-cell surveillance allows the emergence of aberrant B-cell clones that may produce autoantibodies against elements, such as red cells (Mynt et al, 1995; Maclean et al, 1996) and platelets (Leach et al, 2000). We report three cases where antibody proven autoimmune neutropenia followed treatment for B-CLL.
Case 1 was a 66-year-old female presenting with Binet Stage C CLL. She received four cycles of oral fludarabine 24 mg/m2 and cyclophosphamide 150 mg/m2 (FC) for 5 days. Post-treatment blood counts showed lymphopenia and persistent thrombocytopenia. The patient returned a month later with severe mouth ulceration and marked neutropenia (neutrophil count 0·2 × 109/l). The neutrophil count recovered after a prolonged course of filgrastim. She proceeded to further treatment with a combination of Rituximab 375 mg/m2 dl, cyclophosphamide 750 mg/m2dl and dexamethasone 12 mg/day (RCD) for 6 days, 4-weekly for four cycles, resulting in normalisation of the blood count.
Case 2 involved a 63-year-old man, who presented 5 years post diagnosis of Binet Stage A CLL, with B symptoms and splenomegaly. He commenced treatment with chlorambucil 10 mg/m2 for 7 days, repeated 4-weekly. Immediately prior to his third cycle his neutrophil count was 0·4 × 109/l and responded to filgrastim. He completed eight cycles of chlorambucil with intermittent filgrastim support, resulting in a normal blood count at completion.
Case 3 was a 49-year-old woman, who relapsed 11 years post initial diagnosis with progressive Stage A disease. Previous therapy included chlorambucil, fludarabine and CHOP (cyclophosphamide, doxorubicin, vincristine and prednisone). She was treated with six cycles of FC, with the addition of Rituximab (375 mg/m2) after the first course. Following the fourth cycle, delayed neutrophil recovery was noted and cyclophosphamide was omitted from subsequent courses. Neutropenia recurred 3 months later (neutrophil count 0·2 × 109/l), and persisted over the next 4 months. The patient received a reduced intensity conditioned allogeneic stem cell transplant from a fully matched unrelated male donor. Neutropenia has not recurred in the post-transplant period of 13 months.
In all cases, trephine biopsies demonstrated only low-level disease infiltrate at the time of neutropenia and samples were sent to the granulocyte immunology reference laboratory, Bristol, UK for investigation of granulocyte antibodies. Serum samples were analysed using the Granulocyte Chemiluminescence Test (GCLT), the Granulocyte Immunofluorescence Test (GIFT) and a panel of granulocytes typed for human neutrophil antigens (HNA)-1a, -1b, -1c, -2a, -3a and -4a. The direct and ‘indirect auto’ Immunofluorescence tests for granulocyte-bound immunoglobulin IgG and IgM were performed to determine if the serum antibodies were autoimmune in nature (Table I).
|Case||Sample taken during||Direct GIFT||‘Indirect auto’ GIFT||GCLT|
|1||Acute phase||ND||Positive||Positive 3/3 donors|
|Remission phase||Positive (IgG)||Negative||Negative|
|2||Acute phase||ND||Positive||Positive 3/3 donors|
|Remission phase||Positive (IgG)||*||Negative|
In cases 1 and 2, neutrophil antibodies were detected in acute phase serum but not in remission. ‘Indirect auto’ immunofluorescence tests confirmed the presence of autoantibodies in the acute phase serum but not in remission phase serum (Table I). Unfortunately, these tests were unable to confirm the autoimmune nature of antibodies in case 3 due to high levels of granulocyte membrane-associated immunoglobulins. In each case with antibody detection, the antibodies reacted independently of HNA-1a, -1b, -1c, -2a, -3a and -4a.
Only one suspected case of therapy-related autoimmune neutropenia has been previously reported (Stern et al, 1999). There are no data on the frequency of anti-neutrophil antibodies in CLL patients without neutropenia. These cases demonstrate that neutropenia in patients with CLL is not always related to chemotherapy or marrow infiltration. Autoimmune neutropenia is a previously unrecognised complication following therapy for CLL, and may be effectively treated with a Rituximab-containing regime.