|
|
Background Sensitization of leukemic cells with hematopoietic growth factors may enhance the cytotoxicity of chemotherapy in acute myeloid leukemia (AML).
Methods In a multicenter randomized trial, we assigned patients (age range, 18 to 60 years) with newly diagnosed AML to receive cytarabine plus idarubicin (cycle 1) and cytarabine plus amsacrin (cycle 2) with granulocyte colony-stimulating factor (G-CSF) (321 patients) or without G-CSF (319). G-CSF was given concurrently with chemotherapy only. Idarubicin and amsacrin were given at the end of a cycle to allow the cell-cycle-dependent cytotoxicity of cytarabine in the context of G-CSF to have a greater effect. The effect of G-CSF on disease-free survival was assessed in all patients and in cytogenetically distinct prognostic subgroups.
Results After induction chemotherapy, the rates of response were not significantly different in the two groups. After a median follow-up of 55 months, patients in complete remission after induction chemotherapy plus G-CSF had a higher rate of disease-free survival than patients who did not receive G-CSF (42 percent vs. 33 percent at four years, P=0.02), owing to a reduced probability of relapse (relative risk, 0.77; 95 percent confidence interval, 0.61 to 0.99; P=0.04). G-CSF did not significantly improve overall survival (P=0.16). Although G-CSF did not improve the outcome in the subgroup with an unfavorable prognosis, the 72 percent of patients with standard-risk AML benefited from G-CSF therapy (overall survival at four years, 45 percent, as compared with 35 percent in the group that did not receive G-CSF [relative risk of death, 0.75; 95 percent confidence interval, 0.59 to 0.95; P=0.02]; disease-free survival, 45 percent vs. 33 percent [relative risk, 0.70]; 95 percent confidence interval, 0.55 to 0.90; P=0.006).
Conclusions Sensitization of leukemic cells with growth factors is a clinically applicable means of enhancing the efficacy of chemotherapy in patients with AML.
Hematopoietic growth factors stimulate AML cells in culture, activating metabolic processes and the cell cycle. In vitro, the simultaneous exposure of leukemic cells to chemotherapy and growth factors such as granulocyte–macrophage colony-stimulating factor (GM-CSF), granulocyte colony-stimulating factor (G-CSF), and interleukin-3, referred to as growth-factor priming, increases the susceptibility of the cells to killing by chemotherapy, especially by the cell-cycle–specific agent cytarabine.2,3,4,5,6,7,8,9,10,11 These observations suggest a novel therapeutic strategy for AML, but the value of such an approach has not been assessed clinically.
In previous trials of AML, G-CSF and GM-CSF have been widely used after chemotherapy to accelerate myeloid regeneration,12,13,14 but there is information only from uncontrolled studies15,16,17 and small, randomized studies18,19,20,21 about their use in growth-factor priming. We conducted a randomized trial to determine whether G-CSF given only during the first two induction cycles with cytarabine plus idarubicin and cytarabine plus amsacrin improves disease-free survival in adults with newly diagnosed AML by increasing the rate of complete response, reducing the relapse rate, or both. G-CSF was not given during the aplastic phase after chemotherapy. To avoid interference of the second chemotherapeutic agent with the cell-cycle–dependent synergy between cytarabine and G-CSF, idarubicin (first cycle) and amsacrin (second cycle) were given at the end of the cycles.
Methods
Patients
Previously untreated patients with a confirmed diagnosis of AML who were 18 to 60 years of age were eligible for the study. All subtypes of AML were included, except acute promyelocytic leukemia and blast crisis of chronic myeloid leukemia. Patients with another active cancer were not eligible, nor were patients with severe heart, lung, or neurologic disease. All patients were screened for eligibility before undergoing randomization.
The study was approved by the ethics committees of the participating institutions and was conducted in accordance with the Declaration of Helsinki. All participants gave their informed consent.
Risk Classification
At diagnosis, samples of bone marrow and blood were examined for cytogenetic abnormalities with the use of standard banding techniques and classified according to the International System for Human Cytogenetic Nomenclature.22 On the basis of the chromosomal analysis, patients were classified into three distinct prognostic categories: favorable risk, unfavorable risk, and standard risk.23,24,25 Favorable risk was defined by the presence of t(8;21)(q22;22), inv16(p13q22), or t(16;16)(p13q22) and a white-cell count of less than 20x103 per cubic millimeter at diagnosis.25 Unfavorable risk was defined by the presence of complex cytogenetic abnormalities (defined as at least four unrelated cytogenetic clones), monosomies, or deletions of chromosome 5 or 7 (5q–, 7q–, –5, or –7), abnormalities of the long arm of chromosome 3(q21;q26), t(6;9)(p23;q34), or abnormalities involving the long arm of chromosome 11(11q23). Leukemias that had occurred after chemotherapy or radiotherapy for a nonhematologic condition and leukemias that had occurred more than six months after a hematologic condition (secondary leukemias) were included in the unfavorable prognostic category. Patients who did not meet the criteria for favorable or unfavorable risk were classified as being at standard risk.
Study Design and Chemotherapy
Patients were enrolled and randomly assigned to receive G-CSF or no G-CSF during remission-induction cycles 1 and 2 (Figure 1). Cycle 1 consisted of cytarabine (200 mg per square meter of body-surface area given by continuous infusion on days 1 through 7) and idarubicin (12 mg per square meter given intravenously over a period of 5 to 10 minutes on days 6, 7, and 8). Cycle 2 consisted of cytarabine (1000 mg per square meter given intravenously over a period of 2 hours every 12 hours on days 1 through 6) and amsacrin (120 mg per square meter given intravenously over a 60-minute period on days 4, 5, and 6). G-CSF (lenograstim, Aventis) was given subcutaneously or intravenously in a dose of 150 µg per square meter per day beginning one day before chemotherapy (day 0) and continuing until the last day of cycles 1 and 2. The administration of G-CSF was postponed or interrupted in the event of leukocytosis (more than 30x103 leukocytes per cubic millimeter) until the white-cell count was below 20x103 per cubic millimeter. Patients with standard-risk or unfavorable-risk AML who were in complete remission after cycle 2 were randomly assigned to a third cycle of chemotherapy with etoposide and mitoxantrone or high-dose chemotherapy with busulfan and cyclophosphamide followed by autologous stem-cell transplantation. Allogeneic stem-cell transplantation was performed if a suitable donor was available and the patient was younger than 55 years of age. Patients with a favorable cytogenetic profile were also to receive the third cycle of chemotherapy with etoposide and mitoxantrone.
|
The primary objective of the study was to determine the effect of adding G-CSF to induction chemotherapy on the rate of response, disease-free survival, relapse-free survival, and overall survival. A secondary objective was to assess the relation between the defined prognostic subgroups and the outcome. A complete response was defined by cellular marrow with less than 5 percent blasts, no Auer rods, no evidence of extramedullary leukemia, and peripheral granulocyte and platelet counts of at least 1.0x103 per cubic millimeter and 100x103 per cubic millimeter, respectively. The time to hematopoietic recovery after both cycles 1 and 2 was measured from the first day of chemotherapy. Disease-free survival was measured from the time of the first complete remission to the date of relapse or death from any cause. Relapse was defined as a recurrence of leukemia after a first complete remission. Event-free survival was measured from the date of randomization to the date of failure to enter a complete remission (set as day 1), death, or relapse, whichever came first. Overall survival was measured from the date of randomization.
Random assignments were balanced with use of a minimization procedure with the hospital as a stratification factor. We planned to enroll 600 patients over a period of five years, with an additional follow-up of two years after the enrollment of the last patient. This number of patients would give the study a power of 78 percent to show an absolute increase of 10 percent in the rate of complete remission (from 70 percent to 80 percent) with the use of G-CSF; a power of 75 percent to show an absolute increase of 10 percent in the overall survival rate (from 35 percent to 45 percent) at three years, given a relative risk of death of 0.76 and with 375 expected deaths; and a power of 81 percent to show an absolute increase of 10 percent in long-term event-free survival (from 25 percent to 35 percent) at three years, given a relative risk of 0.76 and 423 expected events, with the use of two-sided tests and a 5 percent significance level. Within 3.5 years, 655 patients had been recruited, 640 of whom could be evaluated. As of August 2002, 445 events had occurred, as defined with respect to event-free survival, and 407 patients had died.
All analyses were conducted according to the intention-to-treat principle, but 14 ineligible patients (7 in each group) were excluded, as was 1 who was lost to follow-up on day 3 and whose data could therefore not be evaluated. Reasons for ineligibility were an incorrect diagnosis (lymphoid neoplasia) in eight patients and myelodysplasia in six. Logistic regression was used to analyze the effect of G-CSF on the rate of complete remission, whereas the log-rank test and Cox regression analysis were used to analyze the differences between the two groups with respect to overall survival, event-free survival, and disease-free survival. These analyses were done before and after adjustment for age, risk category, and transplantation status during a first complete remission (as a time-dependent covariate). Competing risk analysis was used to calculate the cumulative competing risks of treatment failure among patients with a complete response (defined as relapse after a complete remission and death during a first complete remission).
The rates of hematologic recovery after cycles 1 and 2 were analyzed actuarially and compared with the use of the log-rank test. In these analyses, data on patients were censored at death or at the start of the next treatment, if hematologic recovery had not yet occurred. All P values reported are two-tailed.
Results
Characteristics of the Patients and Adherence to G-CSF Treatment
Between March 1995 and January 1999, 319 patients were assigned to induction chemotherapy without G-CSF, and 321 patients were assigned to chemotherapy combined with G-CSF. As of the time of the data analysis, the median follow-up was 55 months, and 90 percent of the patients had been followed for more than 40 months. Thirteen patients were lost to follow-up or were last seen more than one year before the analysis. Of these 13 patients, 7 had been followed for more than three years.
The two treatment groups were evenly matched with respect to various factors, including assignment to postinduction therapy (Table 1). As for the prognostic risk groups, most patients were in the standard-risk category, and approximately 20 percent were in the unfavorable-risk category (Table 1). Only about 7 percent of all the patients presented with prognostically favorable AML.
|
Response and Adverse Effects
The rates of complete remission were 83 percent in the group that did not receive G-CSF and 79 percent in the group that received G-CSF (P=0.24) (Table 2). In both groups, 73 percent of the complete remissions occurred after cycle 1. The rates of complete remission in the two groups did not differ significantly according to age or risk group. The frequencies of various grade 3 (severe) or grade 4 (very severe) adverse effects (according to World Health Organization criteria) after cycles 1 and 2 and the times to hematopoietic recovery after cycles 1 and 2 were similar in the two groups (Table 3). There were more deaths within 50 days after cycles 1 and 2 among patients who received G-CSF than among patients who did not receive G-CSF (Table 4).
|
|
|
Among patients who had a complete remission, the disease-free survival rate at four years was higher in the G-CSF group than in the group that did not receive G-CSF (42 percent vs. 33 percent; relative risk of relapse or death, 0.77; 95 percent confidence interval, 0.62 to 0.96; P=0.02) (Table 2 and Figure 2). This difference was related to a lower relapse rate in the G-CSF group (46 percent vs. 54 percent; relative risk, 0.77; 95 percent confidence interval, 0.61 to 0.99; P=0.04) (Table 2). At four years there were no statistically significant differences between the two groups in the rates of overall and event-free survival (Table 2 and Figure 2). The unadjusted Cox regression analysis and an analysis adjusted for age, risk category, and presence or absence of subsequent stem-cell transplantation during a first complete remission (time-dependent covariates) yielded similar estimates for the hazard rates associated with treatment results and P values (data not shown).
|
Among patients with standard-risk AML (72 percent of all patients), treatment with G-CSF reduced the probability of relapse and improved overall, event-free, and disease-free survival (Figure 3A and Figure 3B and Table 2). G-CSF did not, however, significantly affect overall, event-free, or disease-free survival among the 135 patients with unfavorable-risk AML (Figure 3C and Figure 3D and Table 2). In the 41 patients with favorable-risk AML, G-CSF priming had no effect.
|
Incubation of AML cells with G-CSF, GM-CSF, or interleukin-3 and the cell-cycle–dependent agent cytarabine increases intracellular levels of the active metabolite cytosine arabinoside triphosphate, incorporation of cytarabine into cellular DNA,5,7 and the killing of leukemic blasts and leukemic progenitor cells by the drug.2,3,4,5,6,7,8,9,10,11 We evaluated the clinical efficacy of growth-factor priming in patients with previously untreated AML. G-CSF was given beginning one day before the start of chemotherapy of cycles 1 and 2 and continued through the last day of induction cycles 1 and 2. Among patients who received G-CSF and who had a complete remission, the relapse rate was lower than that among patients in complete remission who did not receive G-CSF. Moreover, in the group of patients who had a complete remission, the disease-free survival rate at four years was 42 percent in the G-CSF group and 33 percent in the group that did not receive G-CSF (P=0.02). The difference in overall and event-free survival did not reach statistical significance. The significant difference in disease-free survival probably resulted from G-CSF–mediated activation of subpopulations of leukemic cells that were initially insensitive to cytosine arabinoside. Elimination of the primed cells may have reduced the frequency of relapse.
The ability of colony-stimulating factors to activate AML cells has been directly demonstrated in vivo: injection of G-CSF or GM-CSF 18 to 72 hours before the beginning of chemotherapy drives AML cells into the cell cycle.26,27 This effect is consistent with the notion that G-CSF receptors, because of their high binding affinity, require minimal levels of ligands for activation.28
From our data, we cannot determine whether the sensitization effect was mediated by increasing the efficacy of cytarabine, idarubicin and amsacrin, or the combined chemotherapeutic agents we used. In any case, the efficacy of a chemotherapy regimen that included cytarabine at doses of 200 mg per square meter as well as 1000 mg per square meter was enhanced by the addition of G-CSF priming. Studies of the dose effect of cytarabine have shown that doses of 3 g per square meter29,30 were more effective than doses of 200 or 400 mg per square meter in preventing relapse but did not result in an increased rate of remission. Similarly, in this study, only the duration of remission, not the number of remissions, changed as the result of G-CSF sensitization.
We used a dose of 1 g of cytarabine per square meter in cycle 2. The comparative effect of a dose of 1 g per square meter and a dose of 3 g per square meter has not been established in AML therapy. It would be of interest to know whether G-CSF priming would have a similarly positive effect on the probability of relapse in regimens containing a dose of 3 g of cytarabine per square meter.
The fact that our results do not suggest a benefit of G-CSF priming in patients with favorable-risk AML might relate to the small numbers of cases, or it might indicate that the dose of 1 g of cytarabine per square meter was optimal in terms of its ability to kill neoplastic cells in this subgroup. After the two induction cycles with or without G-CSF, approximately one third of patients received a third cycle of chemotherapy and another third went on to high-dose chemotherapy followed by stem-cell transplantation. It is unlikely that the postinduction treatment influenced the outcome of G-CSF treatment. The two groups were evenly matched in terms of assignment to postinduction therapy. Besides, Cox regression analysis with autologous and allogeneic transplantation during a first remission as time-dependent covariates yielded results similar to those of the unadjusted analysis.
The fact that more deaths occurred during induction among patients who received G-CSF may explain the slightly reduced rate of complete remission in this group. These deaths had several causes and were thus not due to a common problem. However, because of a reduction in the incidence of later deaths, the overall death rate among patients in the G-CSF group was lower than that among those treated with chemotherapy alone.
Previous studies of G-CSF and GM-CSF in AML have been almost entirely confined to the ability of these growth factors to accelerate hematopoietic recovery and reduce morbidity and mortality due to infection after chemotherapy. The efficacy of these agents in modulating chemotherapy has, however, been evaluated in controlled18,19,20,21 and uncontrolled15,16,17 studies involving limited numbers of patients. In two relatively large, randomized studies, GM-CSF was administered concomitantly with and after chemotherapy.31,32 These studies involved older patients, most of whom had AML with an unfavorable prognosis. One of these studies reported a higher rate of disease-free survival among the patients who received GM-CSF than among those who did not receive GM-CSF,32 but it was not possible to distinguish the effect of priming from the effect of enhanced hematopoietic recovery. By contrast, our study selectively focused on the effect of growth-factor priming in AML and was conducted in young and middle-aged adults with previously untreated leukemia.
We found that G-CSF improved overall and disease-free survival in the group with standard-risk AML. There were too few patients in the group with a favorable prognosis to allow a meaningful analysis. There was no indication that G-CSF priming improved the outcome among patients with chemotherapy-refractory, unfavorable-risk AML. This lack of benefit explains why overall survival was not significantly better in the G-CSF group as a whole.
Additional studies of G-CSF priming in specific subgroups of patients and regimens of combination therapy seem warranted. The results of our study provide proof of the principle that chemotherapy and sensitization of leukemia cells by hematopoietic growth factors is a plausible strategy for reducing the risk of relapse in patients with AML.
Local data management was supported by the Dutch Cancer Society Queen Wilhelmina Fund. Aventis provided an unrestricted grant for central data management to the Dutch–Belgian Hemato-Oncology (HOVON) Cooperative Group and provided lenograstim free of charge.
We are indebted to Janine Vrij for secretarial assistance.
Source Information
From the Department of Hematology (B.L., P.S., G.G.) and the HOVON Data Center and Department of Statistics (W.P.), Erasmus University Medical Center, Rotterdam, the Netherlands; the Department of Hematology and Oncology, Johannes Gutenberg–University Hospital, Mainz, Germany (M.T.); Onkozentrum, Hirslanden Klinik Im Park, Zurich, Switzerland (J.G.); the Department of Hematology, University Medical Center, Utrecht, the Netherlands (L.V.); the Institute of Medical Oncology, University and Inselspital, Berne, Switzerland (M.F.); the Department of Internal Oncology and Hematology, University Hospital, Maastricht, the Netherlands (H.S.); the Department of Hematology, Cliniques Universitaires Saint-Luc, Brussels, Belgium (A.F.); the Division of Hematology, University Hospital, Lausanne, Switzerland (T.K.); the Division of Hematology, Kantonsspital, Basel, Switzerland (A.G.); the Department of Hematology, University Hospital, Groningen, the Netherlands (S.D.); the Department of Hematology, Free University Medical Center, Amsterdam (P.H.); and the Department of Hematology, Hospital Gasthuisberg, Leuven, Belgium (M.B.).
Address reprint requests to Dr. Löwenberg at Erasmus University Medical Center, Department of Hematology, P.O. Box 2040, 3000 CA Rotterdam, the Netherlands, or at b.lowenberg{at}erasmusmc.nl.
References
-D-arabinofuranosylcytosine (ara-C) on acute myeloblastic leukaemia (AML) cells. Br J Haematol 1988;70:121-123. [Medline]
The following centers and persons participated in the study: Study coordinators — B. Löwenberg, M.A. Boogaerts, and J. Gmür; Statistician — W. van Putten (Dutch–Belgian Hemato-Oncology [HOVON] Cooperative Group Data Center, Rotterdam, the Netherlands); Cytogenetic review — A. Hagemeijer (Leuven, Belgium), S.L. Bhola (Leiden, the Netherlands), and M. Jotterand-Bellomo (Lausanne, Switzerland); Cytology review committee — M.B. van 't Veer (Rotterdam, the Netherlands), J. Fehr (Zurich, Switzerland); Central data management — A.J.M. Meurisse, M. van Os (HOVON Cooperative Group Data Center) and B. Rufener (Swiss Group for Clinical Cancer Research Trial Office Center, Bern, Switzerland); Participating centers (and investigators) — the Netherlands: Free University Medical Center, Amsterdam (P.C. Huijgens, G.J. Ossenkoppele); University Hospital, Groningen (S.M.J.G. Daenen, E. Vellenga); University Hospital, Utrecht (A.W. Dekker, L.F. Verdonck); Erasmus Medical Center and Daniel den Hoed Cancer Center, Rotterdam (B. Löwenberg, P. Sonneveld, G.E. de Greef); University Hospital, Maastricht (H.C. Schouten); University Hospital, Amsterdam (J. van der Lelie); Leijenburg Hospital, The Hague (P.W. Wijermans); Hospital Eemland, Amersfoort (S. Wittebol); Medical Center Twente, Enschede (M.R. Schaafsma); Sophia Hospital, Zwolle (M. van Marwijk Kooy); Antonius Hospital, Nieuwegein (D.H. Biesma); Antoni van Leeuwenhoekhuis, Amsterdam (J.W. Baars); Diaconessen Hospital, Meppel (H. de Korte); Hospital Reinier de Graaf, Delft (E. Maartense); Medical Center, Leeuwarden (P. Joosten); Catharina Hospital, Eindhoven (W.G. Peters); and Hospital Gooi Noord, Blaricum (H.P. Muller); Belgium: Hospital Gasthuisberg, Leuven (M.A. Boogaerts, G. Verhoef); Cliniques Universitaires Saint-Luc, Brussels (A. Ferrant); Cliniques Universitaires de Mont-Godinne, Yvoir (A. Bosly); and Hôpital de Jolimont, Haine-St. Paul (A. Delannoy); Germany: Johannes Gutenberg-University Hospital, Mainz (M. Theobald, J. Beck); and Nordwest Hospital, Frankfurt am Main (A. Knuth); Switzerland: University Hospital, Zurich (J. Gmür, E. Jacky); University Hospital, Bern (M.F. Fey, A. Tobler); University Hospital, Lausanne (T. Kovacsovics); University Hospital, Basel (A. Gratwohl, A. Tichelli); Kantonsspital, St. Gallen (U. Hess); Hôpital Cantonal Universitaire, Geneva (B. Chapuis); University Hospital, Neuchâtel (D. Piguet); Hospital St. Giovanni, Bellinzona (L. Leoncini); Kantonsspital, Winterthur (T. Kroner); and Kantonsspital, Arau (M. Wernli).
| |||||||||||||||||||||||||||||||||||||||||||||||||||||
Related Letters:
G-CSF Priming in Acute Myelogenous Leukemia
Murashige N., Kami M., Takaue Y., Löwenberg B.
Extract |
Full Text |
PDF
N Engl J Med 2003;
349:2071-2072, Nov 20, 2003.
Correspondence
This article has been cited by other articles:
HOME | SUBSCRIBE | SEARCH | CURRENT ISSUE | PAST ISSUES | COLLECTIONS | PRIVACY | HELP | beta.nejm.org Comments and questions? Please contact us. The New England Journal of Medicine is owned, published, and copyrighted © 2008 Massachusetts Medical Society. All rights reserved. |
Previous
