A randomized comparison of immediate versus delayed application of G-CSF in induction therapy for patients with acute myeloid leukemia unfit for intensive chemotherapy

Abstract

We randomized 66 elderly patients with AML unfit for conventional chemotherapy to receive GCSF from d6 or from d12 after induction-chemotherapy with cytarabine/idarubicin. There was no difference in duration of neutropenia (17 days vs. 19 days, p=0.67) or rate of complications. Delayed treatment can reduce the administration of G-CSF without adverse consequences.

Elderly AML-patients are prone to infection-related mortality. G-CSF might reduce the incidence of infections1–3 and shorten hospitalization.4, 5 However, the timing of G-CSF application is unclear. It is commonly given immediately after chemotherapy,4 but there are no randomized studies evaluating the time-point of application in AML after induction-therapy. We conducted a prospective study in elderly AML-patients unfit for standard induction-therapy with G-CSF application randomized to immediate versus delayed start. If patients responded, a second cycle of induction-therapy with the same dosage was given and GCSF applied in a cross-over design. The aim of the study was to determine the duration of neutropenia and the frequency of infections in both arms.

From November 1998 to March 2003, 66 patients (Table 1, Supplementary Data) were included in the study from 6 hospitals. Induction-chemotherapy consisted of cytarabine 100 mg/m as continuous intravenous infusion (civ) days 1–5 and idarubicin 5 mg/m civ days 1–5. For G-CSF, patients were randomized as follows: Group A received filgrastim (Amgen, Munich, Germany) from day 6 and group B from day 12 until the leukocyte count reached >4 G/L or a bone marrow puncture (BMP) confirmed blast-persistence. The dose was 480 μg/d if >75 kg and 300 μg/d if ≤75 kg. A control BMP was performed after regeneration of the peripheral blood count or on day 40. Complete remission (CR) and partial remission (PR) were defined according to standard criteria.

Table 1.Patient characteristics and eligibility criteria.

Table 2.Duration of G-CSF treatment, duration of neutropenia and infectious complications for the first cycle.

Statistical analyses were performed using SPSS (version 12.0 for Windows, Munich). ANOVA, Fisher’s exact, the χ and Mann-Whitney tests were used to evaluate any differences, a two-sided p value <0.05 was considered significant. Survival was analyzed using the Kaplan-Meier method, the log-rank test used to evaluate any differences. 95% confidence intervals were calculated using CIA. The study was powered to detect a 10% difference with 63 cycles included (power 80%, α=0.05). Compared with studies including fit elderly patients,6 few patients died early (n=14; 21%). Seventeen patients had blast-persistence (26%), 24 (36%) achieved a CR, 3 (5%) a CR with incomplete recovery of neutrophils/platelets, and 7 (11%) a PR after one cycle of therapy (one patient lost to followup). Thirty-three patients received another cycle of chemotherapy (16 patients with G-CSF from day 6 and 17 patients from day 12). Here, another 6 patients achieved a CR leading to an overall CR rate of 45%. The median overall survival for those in CR was 14 months (95%CI 11–17). There was no significant difference in response between the two groups. Duration of G-CSF application was reduced by five doses in the group with delayed administration (Table 2). This saves 770 € (German hospital pharmacy 2006 prices including 16% VAT) per patient. There was no difference in the duration of neutropenia between groups (median duration of neutropenia for all patients: 19 days, 95%CI 15–21). Twenty-five patients in group A and group B respectively (83% vs. 69%, p=0.22) had an episode of neutropenic fever with no significant difference in duration (Table 2). No significant differences were detected in the frequency of pneumonia or any chemotherapyinduced toxicity. Early deaths were comparable in both groups (group A: 8, group B: 6 patients, p=0.25). After the second cycle, duration of neutropenia and infectious complications were lower for both groups. Again, there was no significant difference between immediate and delayed application of G-CSF (Figure 1).

Figure 1.Duration of neutropenia after early or late G-CSF. Neutropenia is defined as leukocyte counts <1 G/L. A: cycle 1, B: cycle 2.

A pooled comparison of all patients receiving immediate G-CSF (n=46) versus delayed G-CSF (n=53) after the first or the second course of chemotherapy confirmed no difference: duration of neutropenia was 14 days (IQR 9–19 days) after immediate and 15 days (IQR 9–22 days) after delayed application of G-CSF (p=0.75). The number of febrile days was 2 (IQR 0–4 days) after immediate and 1 (IQR 0–4 days) after delayed treatment (p=0.77).

This study evaluated the timing of the start of G-CSF administration on duration of G-CSF use, duration of neutropenia and number of febrile days in AML patients unfit for conventional chemotherapy. Supportive therapy with growth factors seems useful in this high-risk group of patients who can still achieve a reasonable response with a low early death rate. In conclusion, administration of GCSF can be delayed to five days after the end of chemotherapy without a prolonged duration of neutropenia or other adverse effects.

Acknowledgments

we would also like to thank the nursing staff and the doctors involved in the treatment of the AML patients included in this study for providing excellent clinical care. In particular, we would like to thank Dr. Peter Ehscheidt, Dr. Eckhardt Jaeger and Professor Dr. Klaus Rieche for presenting patients and their help with study management

Footnotes

• Funding: this study was supported by unrestricted grants from the Leukämie-Initiative Bonn, AMGEN and Pharmacia.