Fax: (713) 794-4297
Fludarabine and mitoxantrone for patients with chronic lymphocytic leukemia
Article first published online: 6 MAY 2004
Copyright © 2004 American Cancer Society
Volume 100, Issue 12, pages 2583–2591, 15 June 2004
How to Cite
Tsimberidou, A. M., Keating, M. J., Giles, F. J., Wierda, W. G., Ferrajoli, A., Lerner, S., Beran, M., Andreeff, M., Kantarjian, H. M. and O'Brien, S. (2004), Fludarabine and mitoxantrone for patients with chronic lymphocytic leukemia. Cancer, 100: 2583–2591. doi: 10.1002/cncr.20264
- Issue published online: 2 JUN 2004
- Article first published online: 6 MAY 2004
- Manuscript Accepted: 10 MAR 2004
- Manuscript Revised: 2 FEB 2004
- Manuscript Received: 23 OCT 2003
- chronic lymphocytic leukemia;
The objective of the current study was to assess the efficacy of combination therapy with fludarabine and mitoxantrone in patients with B-cell chronic lymphocytic leukemia (CLL).
Eighty-eight patients were treated with fludarabine 30 mg/m2 intravenously daily for 3 days and mitoxantrone 10 mg/m2 on Day 1 (FN). Patients were divided into four groups based on expected response to single-agent fludarabine. These four groups included previously untreated patients, patients who previously were treated with alkylating agents, patients who were successfully treated with alkylating agents and fludarabine but who developed recurrent disease, and patients whose disease was refractory to fludarabine with or without alkylating agents.
The overall response rate was 66%. The response rates were 83% in previously untreated patients, 87% in patients previously treated with alkylating agents, 50% in patients whose disease was not refractory to fludarabine at the start of therapy, and 25% in patients whose disease was refractory to fludarabine. The complete remission (CR) rate was 20% for previously untreated patients, which was not significantly different from the CR rate for a group of historical control patients who were treated with single-agent fludarabine. The median follow-up was 8 years for surviving patients. The median progression free survival was 24 months for all patients and 34 months for previously untreated patients. The median overall survival was 40 months, and the median survival of previously untreated patients was 88 months. The most common toxicities were myelosuppression and infection. Eleven patients (12.5%) developed a second malignancy after a median of 62 months.
The FN regimen did not have a significant advantage over fludarabine alone in the treatment of patients with CLL. Cancer 2004. © 2004 American Cancer Society.
The introduction of the purine analog fludarabine into the treatment of chronic lymphocytic leukemia (CLL) has resulted in substantial clinical benefit. Single-agent fludarabine induces responses in approximately 70–80% of previously untreated patients and in 45–55% of patients with recurrent CLL.1–5 However, complete remission (CR) rates are low, and this has led to the use of fludarabine in combination with other active agents.6–12
The combination of fludarabine and prednisone was evaluated both in previously untreated patients and in patients with recurrent CLL.13 The addition of prednisone did not improve overall response rates or CR rates but increased the incidence of atypical infections.13
A Phase I study of FND (fludarabine, mitoxantrone [Novantrone; OSI Pharmaceuticals, Melville, NY], and dexamethasone) in 21 patients with recurrent indolent lymphoma resulted in an overall response rate of 71% and a CR rate of 43%.14 A subsequent Phase II study of FND in 51 patients with recurrent low-grade lymphoma yielded an overall response rate of 94% and a CR rate of 47%.15
In vitro studies have demonstrated synergism of mitoxantrone with fludarabine in inducing cytotoxicity in B-cell CLL.16–18 The cytotoxic effect of mitoxantrone in B-CLL cells appears to be a result of apoptosis through the induction of DNA fragmentation and the proteolytic cleavage of poly(ADP-ribose) polymerase, a marker for the activation of caspases.18
Based on the substantial efficacy of the FND combination in patients with indolent lymphoma and the in vitro data suggesting synergistic activity, a Phase II trial was performed to assess this regimen in patients with CLL. Dexamethasone was excluded from the combination because of the lack of enhanced response and the significant incidence of atypical infections seen with the combination of fludarabine and prednisone. The current report summarizes the results of that trial after 8 years of follow-up.
MATERIALS AND METHODS
Patients were entered into the study after written informed consent was obtained according to institutional guidelines. Entry criteria included 1) confirmation of the diagnosis of CLL with an absolute peripheral lymphocytosis > 5 × 109/L for > 60 days and evidence of monoclonality on 2-parameter flow cytometry; 2) adequate renal function (creatinine levels < 2 mg/dL) and hepatic function (bilirubin levels < 2 mg/dL), unless these abnormalities were due to organ infiltration by lymphocytes; 3) Zubrod performance status ≤ 3; and 4) age ≥ 16 years. Previously treated or untreated patients with Rai Stage III and IV disease were eligible. Patients with Rai Stage I–II disease were eligible if they had evidence of active disease as defined by the National Cancer Institute (NCI)-sponsored Working Group19 (i.e., an increase in symptoms related to leukemia, including weight loss of ≥ 10% over a 6-month period, body temperature of 38 °C without evidence of infection, worsening anemia or thrombocytopenia, progressive lymphocytosis with a rapid lymphocyte doubling time, marked hypogammaglobulinemia or paraproteinemia, or extreme fatigue). Patients with fludarabine-refractory disease, defined as those who did not experience a response to fludarabine alone or who developed recurrent disease within 6 months of experiencing a response to fludarabine, were included. Patients who had positive direct antiglobulin (Coombs) test results were not excluded from the trial. Patients underwent pretreatment evaluation, including acquisition of medical history, physical examination, complete blood counts, differential and platelet counts, a biochemical assay panel (including liver and renal function studies), serum β2-microglobulin and serum immunoglobulin measurements, bone marrow aspiration and biopsy, and acquisition of marrow samples for immunophenotyping.
The regimen of fludarabine and mitoxantrone (FN) included intravenous fludarabine 30 mg/m2 over 30 minutes on Days 1–3 and intravenous mitoxantrone 10 mg/m2 over 15 minutes on Day 1 administered 4 hours after completion of the first dose of fludarabine. FN therapy was repeated every 4 weeks for a maximum of 6 courses. Patients were reevaluated after 3 courses and 6 courses; those who achieved partial remission (PR) or who demonstrated a continued response at 3 months were considered for 3 additional courses. Patients who did not achieve at least a PR after six cycles of FN were withdrawn from the study.
If Grade 1 toxicity occurred, then the dose could be escalated to Level +1 (fludarabine 25 mg/m2 on Days 1–4 and mitoxantrone 10 mg/m2 on Day 1) on subsequent cycles. Patients were treated at Level −1 (fludarabine 30 mg/m2 on Days 1–3 and mitoxantrone 8 mg/m2 on Day 1) if they developed pneumonia, septicemia, or another life-threatening infection during any course; if platelet recovery to pretreatment levels exceeded 35 days; or if Grade 3–4 nonhematologic toxicity occurred.
The response criteria were those defined by the NCI-sponsored Working Group.19 A CR required the disappearance of all palpable disease; the normalization of blood counts with neutrophil counts > 1.5 × 109/L, platelet counts > 100 × 109/L, hemoglobin levels > 11 g/dL, bone marrow aspirate lymphocytes < 30%, and no evidence of disease on bone marrow biopsy. A nodular PR (nPR) met the same criteria as a CR, with the exception that lymphoid nodules could be seen on bone marrow biopsy. A PR required a reduction ≥ 50% in palpable disease as well as ≥ 1 of the remaining features: neutrophil counts ≥ 1.5 × 109/L or 50% improvement over baseline, platelet counts > 100 × 109/L or 50% improvement over baseline, and hemoglobin levels > 11.0 g/dL or 50% improvement over baseline without transfusions. No bone marrow evaluation was required for determination of PR. Computerized tomography (CT) scans were not required for evaluation of response. After the completion of therapy, patients were reevaluated at 3-month intervals via acquisition of history, physical examination, and blood counts. Bone marrow examinations were performed every 6 months. Toxicities were graded according to the NCI Common Toxicity Criteria Version 2.0.20
Cutoff points for quantitative variables were either those that defined abnormal levels or others that were commonly used. The chi-square test was used to analyze the difference between two categoric variables. Distributions of survival and time to disease progression were estimated using the method of Kaplan and Meier.21 Survival intervals were measured from the first day of chemotherapy until death; deaths due to all causes were included. The time to disease progression was measured from the start of chemotherapy until the detection of recurrent disease.
Sample Size Determination
Previously untreated patients
The rate of response to fludarabine among previously untreated patients was 70–80%. However, the CR rate was only 33%. The objective of the current study was to increase the CR rate to ≥ 55%. Fourteen patients were to be entered in the first stage, and the study would be terminated if the CR rate was < 35.7% (5 of 14 patients). Otherwise, 22 additional patients, for a total of 36 patients, would be entered. Therapy would not be recommended for further study at the end of the second stage if the CR rate was < 42% (15 of 36 patients). This design assumes type 1 and type 2 error rates of 10% and has an average sample size of 25 patients and a 49% probability of early termination if the true response rate is 33%.
Patients previously treated with alkylating agents
Twenty-two patients could be entered in the first stage and the study terminated if the response rate was < 31.8% (7 of 22 patients). Otherwise, 24 additional patients, for a total of 46 patients, would be entered. Therapy would not be recommended for further study at the end of the second stage if the response rate was < 37% (17 of 46 patients). This design assumes type 1 and type 2 error rates of 10% and has an average sample size of 30 patients and a 67% probability of early termination if the true response rate is 30%.
Patients previously treated with fludarabine
The expected response rate for patients who previously had received fludarabine was 20%. The objective of the study was to identify a regimen with a response rate > 40%. Seventeen patients could be entered in the first stage, and the study would be terminated if the response rate was < 17.6% (3 of 17 patients). Otherwise, 20 additional patients, for a total of 37 patients, could be entered. Therapy would not be recommended for further study at the end of the second stage if the response rate was < 27% (10 of 37 patients). This design assumes type 1 and type 2 error rates of 10% and has an average sample size of 26 patients and a 55% probability of early termination if the true response rate is 20%.
Patients who experienced failure after receiving alkylating agents and fludarabine
The study would be halted for lack of response if 0 of 14 responses were observed or if < 2 responses were observed among the first 25 patients. This rule yields a 0.03 probability of early study termination if the true response rate is 0.2 and provides 0.64 and 0.97 probabilities of early study termination if the true response rates are 0.05 and 0.1, respectively.
Patients were divided into four groups with different expected responses to single-agent fludarabine: 1) previously untreated patients; 2) patients who were treated previously with alkylating agents either alone or in combination (these patients had not received therapy with nucleoside analogs); 3) patients who had received therapy with alkylating agents and fludarabine and had initially responded to fludarabine therapy but then developed recurrent disease; and 4) patients who had previously experienced failure after receiving fludarabine-based treatment along with one alkylating agent. Eighty-nine patients were enrolled, and 88 patients were evaluable. One patient was excluded because he withdrew consent and was not treated.
Patient characteristics are detailed in Table 1. The median age was 59 years (range, 21–78 years). Thirty-six patients (41%) had Rai Stage III–IV disease, and 35 patients (40%) were previously untreated. Among previously treated patients, the median number of previous regimens was 2 (range, 1–7). Among the 53 previously treated patients with CLL, 15 patients (28%) had disease that was refractory to alkylating agents, and 12 patients (23%) had disease that was refractory to fludarabine. The median β2-microglobulin level was 3.6 mg/L (range, 1.5–10.8 mg/L).
|Characteristic||No. of patients (%)|
|> 60||38 (43)|
|No. of prior regimens|
|≥ 4||12 (14)|
|Disease refractory to|
|Alkylating agents||15/53a (28)|
|β2-microglobulin concentration (mg/L)|
|> 3.0||45/70 (64)|
|> 4.0||27/70 (39)|
Rates of response to FN according to previous therapy are summarized in Table 2. The overall response rate was 66%, with 12% of patients achieving CR. Previously untreated patients had an overall response rate of 83%. The CR rate was 20% for previously untreated patients, which was not significantly different from the CR rate observed with single-agent fludarabine. The CR and overall response rates were 7% and 87%, respectively, for patients who were treated previously with alkylating agents; 12% and 50%, respectively, for patients whose disease was not refractory to fludarabine at the start of therapy; and 0% and 25%, respectively, for patients whose disease was refractory to fludarabine.
|Prior treatment||No. of patients||CR (%)||Nodular PR (%)||PR (%)||Overall response (%)|
|None||35||7 (20)||17 (49)||5 (14)||29 (83)|
|Alkylating agents||15||1 (7)||3 (20)||9 (60)||13 (87)|
|Fludarabine—sensitive||26||3 (12)||2 (8)||8 (30)||13 (50)|
|Fludarabine—refractory||12||0 (0)||0 (0)||3 (25)||3 (25)|
|Total||88||11 (12)||22 (25)||25 (28)||58 (66)|
A total of 367 courses of FN were administered. The median number of courses administered to all patients was 5 (range, 1–12 courses), with a median of 6 courses administered to responders (range, 2–12 courses) and a median of 3 courses administered to nonresponders (range, 1–6 courses). Overall, eight patients received more than six cycles of treatment. All 8 were treated with additional cycles off-study, because they had either a PR (n = 7) or evidence of residual disease on flow cytometry (n = 1) and the treating physician elected to continue treatment with the goal of maximizing response. However, no change in response was noted in these patients in subsequent cycles compared with the assessment after six cycles. In Figures 1 and 2, response rates achieved with FN are compared with those achieved with other regimens used at The University of Texas M. D. Anderson Cancer Center (Houston, TX) in previously untreated and previously treated patients with CLL.
The most common toxicities were myelosuppression and infection. Table 3 shows the percentages of patients who developed Grade 3–4 toxicity at any time during treatment. Patients were included if they developed Grade 3–4 myelosuppression at any time during therapy. There were no differences in the incidence of Grade 3–4 neutropenia or Grade 3–4 thrombocytopenia between untreated patients and previously treated patients.
|Grade 3–4 toxicity||Patients (n = 88)||Cycles (n = 367)|
|No. (%)||No. (%)|
|Neutropenia||65 (74)||156 (43)|
|Thrombocytopenia||20 (23)||46 (13)|
During treatment, infectious complications were noted in 25% of patients, including 12.5% who developed Grade 3–4 infections. Pneumonia occurred in 5.7% of patients, sepsis occurred in 2.3% of patients, and pneumonia and sepsis together occurred in 4.5% of patients. Fever of unknown origin, which frequently was associated with neutropenia and necessitated hospitalization, was observed in 17% of patients. Minor infections occurred in 25% of patients. Herpes zoster virus infections occurred in 4.5% of patients, and reactivation of herpes simplex occurred in 6.8% of patients. One patient developed respiratory syncytial virus infection. Six percent of patients developed hemolytic anemia. One patient developed progressive multifocal leukoencephalopathy 7 years after the completion of FN therapy; this patient subsequently received fludarabine and cyclophosphamide (FC) as well as alemtuzumab therapy.
Although ondasentron prophylaxis was administered to all patients, 30% of patients experienced nausea or emesis. Fourteen percent of patients had fatigue, 6% of patients had urticarial rashes, 2% of patients had diarrhea, and 3% of patients had bone pain. Other toxicities included mucositis (n = 2), weight loss (n = 2), alopecia (n = 2), anorexia (n = 2), dyspepsia (n = 2), asthmatic reaction (n = 1), muscle pain (n = 1), and syncope (n = 1).
Eleven patients (12.5%) developed a second malignancy after a median of 62 months (range, 6–98 months). Six patients had squamous cell carcinoma of the skin, one patient developed prostate carcinoma, three patients were diagnosed with lung carcinoma (small cell carcinoma in one patient, nonsmall cell carcinoma in one patient, and adenocarcinoma in one patient), and one patient had vaginal squamous cell carcinoma.
One male patient (1.13%) age 57 years developed refractory anemia with excess blasts in transformation (RAEB-t) 6.5 years after the completion of FN therapy. He achieved a PR following FN treatment, and 1.5 years later, he was treated with FC. Cytogenetic analysis at the time of the diagnosis of RAEB-t showed hypodiploid clones with del(5)(q12), −7, del(9)(q21), and +2mar. The patient underwent allogeneic stem cell transplantation from his human leukocyte antigen–identical sister; fludarabine and melphalan were used as the conditioning regimen. The patient achieved a CR but developed recurrent disease after 2 months. Gemtuzumab ozogamicin and, subsequently, fludarabine, cytarabine, and donor lymphocyte infusion were administered as salvage treatment, but the patient died of Aspergillus pneumonia without evidence of remission 5 months after the diagnosis of RAEB-t.
Survival and Progression-Free Survival
The median follow-up duration for surviving patients was 8 years (range, 0.8–9.3 years). Sixty-six patients (75%) died during that time. The median overall survival was 3.3 years (range, 0.1–9.3 years), and the 5-year survival rate was 43%.
The median survival of previously untreated patients was 7.3 years (Fig. 3). The median survival was 31 months for patients who were treated previously with alkylating agents, 36 months for patients who were treated with alkylating agents and fludarabine, and 9 months for patients who had disease that was refractory to fludarabine (Fig. 3).
Forty-seven of 58 responders developed disease progression. The median progression-free survival was 2 years. The median time to disease progression for previously untreated patients was 34 months (Fig. 4). The median time to disease progression was 16 months for patients who were treated previously with alkylating agents, 17 months for patients who were treated previously with alkylating agents and fludarabine, and 5 months for patients whose disease was refractory to fludarabine (Fig. 4). The median progression-free survival durations according to response to fludarabine and mitoxantrone were 34 months for patients who achieved a CR, 28 months for patients who achieved an nPR, and 16 months for patients who achieved a PR (Fig. 5).
In the current study, 88 patients with B-cell CLL were treated with the FN regimen. The overall response rate was 66%. For previously untreated patients, the overall response rate was 83% (CR rate, 20%), and the median progression-free survival was 34 months. Response and progression-free survival rates, as expected, were inferior among patients who previously had been exposed to therapy. FN was well tolerated, and the most common treatment-associated toxicities were myelosuppression and infection.
When it is used as initial treatment, fludarabine yields high rates of response and prolonged progression-free survival in patients with CLL,1, 2, 22 but it does not prolong overall survival.2, 22 To increase response rates and improve outcome, fludarabine has been combined with other active agents.7–9, 23–29 The current study was based on the favorable results of the FND regimen in patients with indolent lymphoma and the results of in vitro studies demonstrating synergism of mitoxantrone with fludarabine in inducing cytotoxicity in B-cell CLL.16–18
The overall response rate (83%) and the CR rate (20%) achieved with FN in the current study did not differ from the rates achieved with single-agent fludarabine (63–74% and ∼20%, respectively) in previously untreated patients with CLL.1, 2, 4, 22, 30 Data emerging from trials involving the combined FC regimen appear more favorable.7, 9, 25, 31, 32 FC resulted in overall response rates of 88% (CR rate, 35%) in previously untreated patients (Fig. 1) and ≥ 85% (CR rate, 12–15%) in patients who did not previously have fludarabine-refractory disease.9 In previously untreated patients who received FC, the median progression-free survival had not been reached at 41 months; in contrast, the median progression-free survival among previously untreated patients was 34 months in the current study of FN.9
In two recent studies, fludarabine-based therapy was combined with rituximab in patients with previously untreated CLL.28, 29 A 3-drug regimen of fludarabine, cyclophosphamide, and rituximab in 202 patients produced CR, nPR, and PR rates of 68%, 18%, and 14%, respectively, and produced significant rates of molecular remission (Fig. 1).28 In a randomized trial that compared fludarabine with concurrent or sequential rituximab in patients with previously untreated CLL, the overall response rates were 90% (CR rate, 47%) for the concurrent regimen, compared with 77% (CR rate, 28%) for the sequential regimen (P = 0.048).29
FN therapy did not increase the response rates compared with the response rates achieved using single-agent fludarabine (30–35%) in patients who had experienced responses to fludarabine and then developed recurrent disease (Fig. 2). Although response rates appeared to be somewhat higher (87%) than those reported (45–65%) for patients who previously had been treated with alkylating agents, the significance of this finding was limited by the small number of patients.13, 22, 33–36
The results of the current study are compatible with previously published data on fludarabine- and anthracycline-based combination regimens in patients with refractory or recurrent CLL.6, 37, 38 In a German study, fludarabine and epirubicin were administered to 38 patients with CLL (25 previously untreated patients and 15 patients with first recurrences); the response rate among patients with recurrent disease was 62% (CR rate, 15%), compared with 92% (CR rate, 40%) among previously untreated patients.37 The Italian group showed that combination therapy with fludarabine, cytarabine, mitoxantrone, and dexamethasone had activity in 31 patients with refractory or recurrent CLL.6 A combination regimen containing fludarabine, cyclophosphamide, and mitoxantrone yielded a response rate of 78% (CR rate, 50%) in patients with resistant or recurrent CLL.38
The results achieved with FN in patients with CLL appear inferior to those achieved with FN in previously untreated patients with indolent lymphoma.39–41 Velasquez et al. found that FN was associated with a response rate of 94% (CR rate, 44%) and, at 4 years, reported high rates of overall survival and failure-free survival (88% and 38%, respectively).40 In patients with recurrent, indolent lymphoma, FN also is associated with higher response rates compared with the response rates observed in the current study in patients with recurrent CLL.42, 43
Similarly, FND is effective against low-grade lymphoma in both previously untreated and previously treated patients14, 15, 44, 45 and induces high response rates in previously untreated patients with indolent Stage IV lymphoma.44 These findings raise the question of the significance of dexamethasone in the FND regimen. In the CLL trial, this agent was excluded from the treatment regimen based on our previous experience with a regimen containing fludarabine and prednisone, in which prednisone was found to increase the number of atypical infections without enhancing efficacy in patients with CLL.13
In terms of tolerability, the results of the current study are consistent with those of other trials that evaluated fludarabine-based therapy, indicating that myelosuppression and infection are the most common toxicities.1, 2, 9, 11, 25 Second malignancies occurred in 12.5% of patients who were treated with the FN regimen. Although three patients developed lung carcinoma, most second malignancies were localized to the skin. Second malignancies occur with increased frequency in patients with CLL and may be secondary to immune defects associated both with the underlying disease and with the consequences of therapy.46–50 The occurrence of second malignancies after treatment with nucleoside analogs in patients with chronic lymphoid malignancies has been reported previously; these agents did not appear to increase the risk of second malignancies beyond the risk already associated with the primary diagnosis.51
In conclusion, the FN regimen induced responses in 83% of patients with B-cell CLL who receiving it as first-line therapy. However, CRs occurred in only a minority of patients. The addition of mitoxantrone to fludarabine does not significantly increase the CR rate beyond what is achieved with single-agent fludarabine.
- 12Fractionated cyclophosphamide, vincristine, liposomal daunorubicin, and dexamethasone plus rituximab and granulocyte-macrophage-colony stimulating factor (GM-CSF) alternating with methotrexate and cytarabine plus rituximab and GM-CSF in patients with Richter syndrome or fludarabine-refractory chronic lymphocytic leukemia. Cancer. 2003; 97: 1711–1720., , , et al.
- 20National Cancer Institute Cancer Therapy Evaluation Program. Revised common toxicity criteria. Version 2.0 [monograph online]. Available from URL: http://ctep.info.nih.gov/reporting/CTC-3.html[accessed March 2003].
- 28A high proportion of true complete remission can be obtained with a fludarabine, cyclophosphamide, rituximab combination (FCR) in chronic lymphocytic leukemia [abstract]. Proc Am Soc Clin Oncol. 2003; 22: 569., , , et al.
- 29Randomized Phase 2 study of fludarabine with concurrent versus sequential treatment with rituximab in symptomatic, untreated patients with B-cell chronic lymphocytic leukemia: results from Cancer and Leukemia Group B 9712 (CALGB 9712). Blood. 2003; 101: 6–14., , , et al.
- 31Fludarabine and cyclophosphamide achieves high complete response rate in patients with previously untreated chronic lymphocytic leukemia. ECOG 1997 [abstract]. Blood. 2001; 98: 632a., , , et al.
- 32Oral fludarabine and cyclophosphamide in previously untreated CLL: preliminary data on 59 patients [abstract]. Blood. 2001; 98: 772a., , , et al.
- 50Prognostic factors in chronic lymphocytic leukaemia: the importance of age, sex and response to treatment in survival. A report from the MRC CLL 1 trial. MRC Working Party on Leukaemia in Adults. Br J Haematol. 1989; 72: 141–149., , .