Severe Chronic Neutropenia in the Large Granular Lymphocyte Syndrome: Outcomes in Response to Granulocyte Colony Stimulating Factor (G-CSF) and Immunosuppressive Therapies

The large granular lymphocyte (LGL) syndrome and large granular lymphocytic leukemia (LGLL) are rare hematological disorders. Through the Severe Chronic Neutropenia International Registry (SCNIR) we have followed the clinical course and treatment responses of 27 LGL patients presenting with severe chronic neutropenia without evidence of leukemia, rheumatoid arthritis or autoimmune diseases. This report focuses on their treatment responses to G-CSF and outcomes with and without other immunosuppressive therapies.

Methods: Patients with severe chronic neutropenia and a diagnosis of LGL syndrome (at least 3 ANC values < 0.5 x 10⁹/L over a 3 month period) were enrollment in the SCNIR if they do not have concomitant systemic autoimmune disease, myeloid or lymphoid malignancy or drug-induced neutropenia.

Findings: The 27 LGL patients (8 males, median age 53; 19 females, median age 57) had chronic neutropenia for approximately 4 years (range 0.4 - 25.6 years) before enrollment. Diagnoses were primarily based on FACS analysis (12 bone marrow [50%]; 11 blood and bone marrow [46%]; 1 blood only [4%]). None of these patients presented with lymphocytosis, i.e., ALC > 5.0 x 10⁹/L. Before enrollment, 7 patients had anti-neutrophil antibody tests (3 positive, 4 negative), 4 patients had antinuclear antibody tests (2 positive, 2 negative), and ANCA was positive for the one patient tested. After enrollment, additional results were: 1 patient had a negative anti-neutrophil antibody, 8 patients had antinuclear antibody tests (2 positive, 6 negative), and ANCA was positive in two additional patients.

The diagnosis of LGL was based on increased LGL cells in the blood or bone marrow, most frequently increased CD 3+, CD 8+, CD 57+ cells. Increases in lymphocytes in the marrow, often in clusters were a common finding; lymphocytes in the marrow ranged from 6-79%. The range of CD 57+ in blood 11-19%, bone marrow 19-61%; CD 56+/NK cells in blood 0.06-14%, bone marrow <1-76.8%; increased CD 8+/CD 3+ in blood 1.5-90%, bone marrow <1-91%. Ten patients have gene rearrangement studies, 7 prior to enrollment (3 negative, 4 positive) and 3 patients had these studies after enrollment, all positive.

Because most patients had neutropenia, fevers and infections we evaluated the effectiveness and safety of G-CSF therapy. Treatments were G-CSF alone (13), G-CSF with prednisone (5), G-CSF with methotrexate and prednisone (4), G-CSF with methotrexate alone (2), G-CSF with cyclosporine (1), G-CSF with cyclosporine and prednisone (1), and G-CSF with prednisone, methotrexate and cyclosporine (1).

Prior to G-CSF treatment the median ANC was 0.54 x 10⁹/L, mean 1.48 +/- 0.35 SEM. The ALC median was 1.67 x 10⁹/L, mean 2.74 +/- 0.74. The G-CSF dosing median was 1.44 mcg/kg/day, mean 2.27 +/- 0.31. All responded to G-CSF alone or in combination therapy to achieve a mean ANC greater than 1.0 x 10⁹/L. After G-CSF treatment the median ANC was 2.23 x 10⁹/L, mean 3.78 +/- 0.40. The ALC median was 2.24 x 10⁹/L, mean 3.17 +/- 0.48. Time period for SCNIR observation median years is 9.6 years; mean 10.0 +/- 1.2, range 0.4 to 23.1 years. There was a total 270 years of observation in the SCNIR and 307 years of G-CSF administration in this group of 27 patients. The increase in neutrophils was consistently associated with fewer episodes of fever, infections and antibiotic use.

One of 27 patients had episodic thrombocytopenia (<50 x 10⁹/L) before G-CSF and continued while on G-CSF. Seven of 26 patients reported episodic thrombocytopenia after starting G-CSF. There were no report episodes of significant bleeding.

Fourteen of 27 are now living, ages 17 to 72, and all have continued on treatment. Over the 24 year observation period, 5 patients developed Hodgkin disease (1) or non-Hodgkin lymphoma (5); and 5/6 died. Five other LGL patients died of sepsis or pneumonia. Other deaths were attributed to a T-cell lymphoproliferative disorder (1), chronic pancreatitis (1) and lung cancer (1). There were no myeloid malignancies.

Conclusions: Our long term observations suggests that the combination of low dose G-CSF plus methotrexate or cyclosporine appears to be more effective than either drug alone, titrated each to minimal effective doses, and continued indefinitely. There was no increased frequency of opportunistic infections. The expected course is a long period of normal ANCs without other autoimmune diseases. Evolution to lymphoid, not myeloid, malignancies is the greatest concern.