Treatment with single agent rituximab is associated with overall response (OR) rate of 15–35% in previously treated patients and up to 44% in untreated patients with CLL. The majority of the responses are partial (PR). GM-CSF was found to increase surface CD20 expression in vitro and potentiate ADCC and therefore improve the efficacy of rituximab. The combination of rituximab and GM-CSF has been investigated in follicular lymphomas and found to be associated with significant responses (

McLaughlin et al.
Ann. Oncol
16
:
v68
,
2005
). Therefore, we designed a phase II study of rituximab plus GM-CSF in CLL. Patients were eligible for this study if they belonged to one of the following: Group 1: previously untreated patients, Rai stage 0-II and b2M >3 mg/mL, B symptoms or significant fatigue. Group 2: previously untreated patients ≥ 70 years with indication for treatment (NCI-WG criteria). Group 3: previously treated patients with evidence of active disease (NCI-WG criteria). All patients received 4 weekly infusions of rituximab 375 mg/m2 plus GM-CSF 250 mcg sc three times a week for 8 weeks. Prior to study entry patients were evaluated for genomic aberration (FISH), IgVH mutation status, ZAP70 expression, and number of CD20 antigen sites. Changes in self-reported fatigue were measured using the FACT-An fatigue scale. Eighty-two patients have been enrolled at this time and 55 are evaluable for response and toxicity. Responses (NCI-WG):

PtsCR (%)NPR (%)*PR (%)OR (%)
*Nodular PR:residual lymphoid nodules present 
Group 1 9   7  7 (78) 
Group 2 20 3 2 12  17 (85) 
Group 3 26 2 3 9  14 (54) 
OVERALL 55 5 (9) 5 (9) 28 (51)  38 (69) 
PtsCR (%)NPR (%)*PR (%)OR (%)
*Nodular PR:residual lymphoid nodules present 
Group 1 9   7  7 (78) 
Group 2 20 3 2 12  17 (85) 
Group 3 26 2 3 9  14 (54) 
OVERALL 55 5 (9) 5 (9) 28 (51)  38 (69) 
View Large

Toxicities attributable to the administration of GM-CSF consisted in Grade 1 injection site reactions in 25% and Grade 1–2 bone pain in 13%. Grade 3 thromobocytopenia was observed in 2% of the patients and Grade 3–4 neutropenia in 4%. Grade 3 infections were observed in 9% of the patients.

The following characteristics were correlated with responses.

CharacteristicPatientsCR%OR%
FISH    
11q-/17p-11  1 (9) 4 (36) 
13q-/+12/normal 38 4 (11) 32 (84) p<.001 
IgVH    
Unmutated 23 1 (4) 11 (48) 
Mutated 28 4 (14) 23 (82) p=.01 
ZAP70    
Positive 30 1 (3) 17 (57) 
Negative 23 4 (17) 17 (74) p=NS 
CD20 sites/cell    
≤10,000 20  10 (50) 
>10,000 16 3 (19) 13 (81) p=.05 
CharacteristicPatientsCR%OR%
FISH    
11q-/17p-11  1 (9) 4 (36) 
13q-/+12/normal 38 4 (11) 32 (84) p<.001 
IgVH    
Unmutated 23 1 (4) 11 (48) 
Mutated 28 4 (14) 23 (82) p=.01 
ZAP70    
Positive 30 1 (3) 17 (57) 
Negative 23 4 (17) 17 (74) p=NS 
CD20 sites/cell    
≤10,000 20  10 (50) 
>10,000 16 3 (19) 13 (81) p=.05 
View Large

Treatment with rituximab plus GM-CSF was associated with a reduction of self-reported fatigue (median score reduction 7 points) in two thirds of the patients.

In conclusion, the combination of rituximab plus GM-CSF is well tolerated and associated with an encouragingly high response rate. Favorable genomic profile, mutated IgVH status and high level of CD20 expression predicted for response to this combination. Ongoing correlative studies aim to clarify the mechanism of action of this combination. Accrual to this study is ongoing.

Topics:
chronic b-cell leukemias, chronic lymphocytic leukemia, granulocyte-macrophage colony-stimulating factor, recombinant granulocyte-macrophage colony-stimulating factors, rituximab, cd20 antigens, fatigue, toxic effect, zap-70 kinase, bone pain

Author notes

Corresponding author

2005, The American Society of Hematology
2005
Sign in via your Institution