Combination chemoimmunotherapies with purine analogues are frequently used as initial therapy of patients with CLL. We investigated the efficacy and tolerability of the combination of FCR plus GM-CSF (sargramostim, Leukine) as initial therapy of patients with CLL based on the following rationale:

  1. high response rates of FCR in frontline CLL as previously reported;

  2. ability of GM-CSF to stimulate granulocyte and macrophage natural cytotoxicity and therefore to potentiate antibody-dependent cellular cytotoxicity (Voso MT, Haematol. 2002);

  3. ability of GM-CSF to increase surface CD20 expression on CLL cells making them a better target for rituximab (Venugopal P, Leuk. Res. 2000); and

  4. the potential to reduce the myelosuppressive effect of the FCR combination.

Patients with untreated CLL, who met NCI-WG criteria for initiation of therapy, were eligible for this trial. We limited accrual to patients with standard risk disease as determined by beta-2-microglobulin ≤ to 4 mg/dL. Patients received fludarabine 25 mg/m2 i.v. days 2–4, cyclosphosphamide 250 mg/m2 i.v. days 2–4, rituximab 375 mg/m2 i.v. day 1. GMCSF was given sc at 250 mcg/m2 on day -1 and days 5–11. For courses 2–6, FCR started day 1 together with rituximab 500 mg/m2, and GM-CSF was given sc at 250 mcg/m2 on day -1 and days 5–11 or until neutrophil recovery. Courses were repeated every 4–6 weeks. Patients were trained to self-administer GM-CSF during course 1 and therefore required to receive course 1 at our center. Local oncologists administered courses 2–6. All patients received standard anti-viral and PCP prophylaxis throughout the duration of therapy. Sixty patients were enrolled of whom 46 are evaluable for response and toxicity. The median age was 55 years (range 35–77). Twelve patients (20%) had Rai stage III–IV. Median beta-2-microglobulin was 2.8 mg/L (1.7–4.2). Seventeen patients (28%) had 11q deletion and 2 patients (3%) had 17p deletion by FISH. Unmutated IgVH occurred in 57% of the patients and ZAP-70 immunohistochemistry was positive in 53%. Thirty-three patients (72%) achieved CR, 5 patients (11%) nPR, and 8 patients (17%) PR, with an OR rate of 100%. Twenty-four patients (52%) had < 1% CD5/CD19+ cells in the marrow at the end of therapy. Twenty-four patients were evaluated for minimal residual disease (MRD) by PCR and 15 patients (63%) achieved MRD negativity at the end of therapy. Grade >/= 3 neutropenia occurred in 35 (76%) of the patients, thrombocytopenia in 8 (17%). Infectious episodes were seen in 5 patients (11%) including one patient who died after 4 cycles because of pneumonia. Nine patients (20%) discontinued GM-CSF after 1 (2 patients), 3 (3 patients) and 4 (4 patients) courses, because of pain/erythema at the injection site (7 patients), fever (1 patient) and syncope (1 patient). Thirty-nine patients (85%) completed all 6 planned courses and only 2 patients (4%) received less than 4 courses. In conclusion, FCR plus GM-CSF has a high CR rate in symptomatic frontline patients with CLL. This regimen has a favorable toxicity profile and was easily administered in the community setting. This study has completed accrual and time-to-event parameters and a comparison with FCR will be presented.

Disclosures: Ferrajoli:Bayer Healthcare: Honoraria, Research Funding; Genentech, Biogen idec: Honoraria. O’Brien:Celgene Corporation: Consultancy, Honoraria. Keating:Celgene Corporation: Consultancy, Honoraria. Off Label Use: The use of GM-CSF in CLL is considered off label.

Author notes

Corresponding author