Anthracycline-fludarabine-containing regimens with or without rituximab in the treatment of patients with advanced follicular lymphoma

Authors


  • The preliminary results of this study were presented at the American Society of Hematology annual meeting, San Francisco, California, December 6–9, 2008.

Abstract

BACKGROUND:

Recent experience has suggested that there has been a stepwise improvement in the survival outcomes of patients who have follicular lymphoma with the introduction of new treatment options. In the current study, the authors report the results of 2 subsequent phase 2 trials of 238 previously untreated patients.

METHODS:

In a trial of bleomycin, epidoxorubicin, cyclophosphamide, vincristine, and prednisone (BACOP) plus fludarabine, mitoxantrone, and dexamethasone (FND), 144 patients received 2 BACOP treatments followed by 4 FND treatments. In a trial of BACOP plus fludarabine and rituximab (FR), 94 patients received 3 BACOP treatments followed by 4 FR treatments.

RESULTS:

The complete remission (CR) rate for BACOP/FND was 62%. After a median follow-up of 60 months, the failure-free survival (FFS) and overall survival (OS) rates at 4 years were 53% and 77%, respectively. The CR rate for BACOP/FR was 79%. After a median follow-up of 36 months, the FFS and OS rates at 4 years were 56% and 97%, respectively, which were significant compared with the CR and OS rates achieved with BACOP/FND. Twenty-five of 42 bcl-2-positive patients attained a molecularly negative CR and had improved FFS. No significant differences were observed between the 2 trials in the percentage of infections or neutropenia.

CONCLUSIONS:

The CR and OS rates achieved with BACOP/FR were better, and overall toxicity did not increase. Furthermore, patients who received rituximab had a better FFS compared with patients who received chemotherapy alone. Finally, although conclusions between nonrandomized groups may depend on differences in observed and unobserved prognostic features, the current results suggested that the addition of rituximab to anthracycline-fludarabine–containing regimens have a favorable effect on the prognosis of patients with advanced follicular lymphoma. Cancer 2009. © 2009 American Cancer Society.

Patients with follicular lymphoma typically have an indolent course characterized by responsiveness to initial therapy, followed by disease recurrence, and subsequent remission of progressively shorter duration.1 Follicular lymphoma often is characterized by the translocation t(14,18), which results in the over-expression and activation of the bcl-2 gene and a subsequent inhibition of apoptosis.2 A management approach can include observation without therapy and various treatments, including alkylating agents, anthracycline- and fludarabine-based regimens, dose intensification with stem cell support, biologic therapies (such as interferons), and therapies that target the CD20 antigen.1, 3 Until a few years ago, there was a general consensus that the outcome of patients with low-grade lymphoma had not changed over the last 30 years.4 Contrasting experiences based on recent retrospective studies have been reported, supporting a stepwise improvement in survival outcomes with the introduction of new treatment options over the last few years.5 At the same time, phase 2 studies6, 7 and randomized trials have demonstrated that the addition of rituximab to the cyclophosphamide, doxorubicin, vincristine, and prednisone (CHOP) regimen,8, 9 or other anthracycline10-12 or nonanthracycline-containing regimens,13 resulted in higher complete remission (CR) rates and better progression-free survival rates. In the current study, we report the results of 2 subsequent phase 2 trials conducted by the Italian Group for Lymphoma Study (GISL) between 1997 and 2006 using CHOP-like plus fludarabine regimens with or without rituximab. These results confirm an improvement in the CR rate and better survival with the addition of rituximab.

MATERIALS AND METHODS

Patients and Entry Criteria

The 2 trials were of combined bleomycin, epidoxorubicin, cyclophosphamide, vincristine, and prednisone (BACOP) plus fludarabine, mitoxantrone, and dexamethasone (FND) and BACOP plus fludarabine and rituximab (FR) and were conducted at 15 GISL sites in Italy. The BACOP/FND study, which was designed in 1997, was based on the idea that combining a CHOP-like regimen with a fludarabine-based chemotherapy schema could increase the response rate and improve survival outcome. The BACOP/FR study was designed in 2003, when rituximab became available in Italy, through the National Health Service. The rationale for conducting the study was that, after a debulking with the CHOP-like regimen, the association of fludarabine and rituximab could clear bcl-2 better and destroy residual masses. The BACOP/FND study registered 144 patients between 1997 and 2002, and the BACOP/FR study registered 94 patients between 2003 and 2006. For both trials, eligible patients had histologically documented follicular lymphoma that was diagnosed according to Revised European-American Classification of Lymphoid Neoplasms/World Health Organization criteria. Patients were required to be previously untreated and to have Ann Arbor stage III or IV disease. Patients with stage II disease were eligible if they presented with B symptoms or bulky disease. Patients also had to have measurable disease, a need for therapy in the opinion of treating physicians, an expected survival >6 months, age 18 to 70 years, and an Eastern Cooperative Oncology Group performance status of 0 to 2. Patients with known human immunodeficiency virus or hepatitis B or C virus infection; central nervous system lymphoma; previous malignancies; or cardiac, renal, liver, or respiratory failure were excluded. Pregnant or lactating women and patients of childbearing potential were excluded unless birth control measures were used. The study complied with all requirements and was conducted in accordance with good clinical practice guidelines. The protocols were approved by the institutional review board at each participating center. Written informed consent was obtained from all patients.

Design of Studies and Treatment Schedules

Both trials were open-label, noncomparative, multicenter, phase 2 studies. Briefly, in the BACOP/FND trial, patients received BACOP chemotherapy comprised of 650 mg/m2 cyclophosphamide, 30 mg/m2 epidoxorubicin, and 1.4 mg/m2 vincristine on Days 1 and 8; 5 mg/m2 bleomycin on Days 15 and 22; and 100 mg/m2 prednisone daily on Days 15 through 28 every 4 weeks for 2 cycles. Then, the patients received FND chemotherapy comprised of 25 mg/m2 fludarabine daily on Days 1 through 3, 10 mg/m2 mitoxantrone on Day 1, and 20 mg dexamethasone daily on Days 1 through 5 every 4 weeks for 4 cycles. At the end of chemotherapy, patients who had attained a CR or a partial response (PR) were randomized either to undergo observation without therapy or to receive 3 MU/m2 interferon alpha (INF) every other day plus 20 mg dexamethasone orally for 4 days every 4 weeks for 6 months. In the BACOP/FR trial, patients received BACOP every 4 weeks for 3 cycles. Patients who attained a CR and bcl-2 clearance did not receive further therapy. Patients who attained a CR after 3 BACOP treatments without bcl-2 clearance, who attained a PR irrespective of bcl-2 status, or were bcl-2 negative at baseline received 4 sequential cycles of FR. The schema of chemotherapy with FR was comprised of 25 mg/m2 fludarabine daily for 3 consecutive days every 3 weeks for 4 cycles. A 375 mg/m2 dose of rituximab was administered by infusion on Day 1 of each cycle. INF was not used in the BACOP/FR trial. In both trials, at the end of chemotherapy, involved-field radiotherapy (IF-RT) (maximum dose, 3500 grays) was allowed to treat residual masses or at the site of previous bulky or extranodal disease at the discretion of the treating physician.

Statistical Analysis

In total, 205 patients were eligible for the protocols, and the intent-to-treat population included all eligible, enrolled patients. Clinical parameters were registered prospectively at the time of each patient's entry into the study. Overall survival (OS) was measured from the date of registration until the last follow-up or death from any cause. Failure-free survival (FFS) was calculated from the date of registration until disease progression, recurrence, or disease- or treatment-related death. Survival curves were estimated using the Kaplan-Meier method,14 and statistical comparisons between the baseline characteristics of the patients and survival estimates were made using the log-rank test.15 In addition, the Cox regression model16 was used to determine the effect of multiple variables on OS and FSS. The proportionality of risk was determined in all Cox regression analyses. Chi-square analysis, when appropriate, and the Fisher exact test were used to evaluate differences between baseline characteristics of the patients in categorized form and the clinical response and toxicity related to treatment. For continuous variables, the Mann-Whitney U test was used.17

The sample size was planned on the basis of the CR rate, whereas we did not plan a sample size to compare the 2 sequential trials for a specific endpoint. In all analyses, a 2-sided P value <.05 indicated a moderate strength of evidence against the null hypothesis. Also, this level of probability was helpful for detecting clinically useful findings. The Follicular Lymphoma International Prognostic Index (FLIPI)18 was calculated for 188 of the 205 assessable patients (92%).

RESULTS

Demographics

Overall, the baseline clinical characteristics of the 205 patients were a median age of 55 years (range, 26-71 years), 55% of patients were men, 92% had Ann Arbor stage III and IV disease, 29% had a FLIPI score >2, and 26% had >1 extranodal site. In Table 1, the baseline parameters are listed by trial. The only clinical characteristic with a significant difference between the BACOP/FND trial and the BACOP/FR trial was the age of the treated patients. The P values related to histologic grade were not calculated because of the high percentage of unknown grade in the BACOP/FND protocol.

Table 1. Descriptive Characteristics of 205 Patients Enrolled in BACOP/FND and BACOP/FR Trials
CharacteristicPercentage of PatientsP
BACOP/FND, n=125BACOP/FR, n=80
  • BACOP indicates bleomycin, epidoxorubicin, cyclophosphamide, vincristine, and prednisone; FND, fludarabine, mitoxantrone, dexamethasone; FR, fludarabine and rituximab; NA, not available; AA, Ann Arbor; II bulk/B, with bulky disease or B symptoms; LDH, lactate dehydrogenase; ULN, upper limit of normality; Hb, hemoglobin; FLIPI, Follicular Lymphoma International Prognostic Index.

  • *

    Data on FLIPI were missing for 15 of 125 patients on the BACOP/FND trial and for 2 of 80 patients on the BACOP/FR trial (8.3%).

Age at diagnosis   
 >60 y, %3926.070
 Median (range), y56 (28-71)54 (26-70).038
Sex   
 Men5059.259
Histology grade   
 12348 
 22625 
 3721 
 NA446 
AA stage   
 II bulk/B88.683
 III2631 
 IV6661 
LDH >ULN2117.574
Hb <12 g/dL1914.344
>4 Lymph node areas3848.191
FLIPI*   
 0-12936.603
 24136 
 3-53028 
>1 Extranodal site2625.871

BACOP/FND

Overall, 144 patients were registered for the BACOP/FND trial. However, for the current analysis, 18 patients were excluded because they had either Ann Arbor stage II disease without B symptoms or bulky disease, and 1 patient was excluded because of a misdiagnosis. Therefore, 125 patients were considered for the current analysis. Treatment was discontinued early in 6 patients for the following reasons: 2 patients died from bronchopneumonia (1 after the first BACOP treatment and 1 after the first FND treatment), 1 patient died from myocardial infarction that was not treatment related in the opinion of the treating physicians after the first BACOP treatment, 2 patients refused to continue treatment after the first BACOP treatment, and 1 patient refused to begin therapy immediately after the informed consent had been obtained. Overall, 119 of 125 patients (93%) completed chemotherapy in 28 to 32 weeks.

The responses in the intent-to-treat analysis by phase of treatment are reported in Table 2. After IF-RT, the overall response rate for all 125 patients was 90%, and 78 patients (62%) attained a CR. A univariate analysis of baseline parameters versus CR indicated a significant negative impact (P = .023) for more than 1 extranodal site and a trend (P = .06) for high FLIPI scores (Table 3).

Table 2. Responses to Therapy in the Intent-to-treat Analysis by Phase of Treatment
ResponsePercentage of PatientsP
BACOP/FND, n=125BACOP/FR, n=80
  • BACOP indicates bleomycin, epidoxorubicin, cyclophosphamide, vincristine, and prednisone; FND, fludarabine, mitoxantrone, dexamethasone; FR, fludarabine and rituximab; CR, complete remission; PR, partial remission; SD-PD, stable disease-progressive disease; IF-RT: involved field radiotherapy.

  • *

    16 and 9% of patients received IF-RT in BACOP/FND and BACOP/FR, respectively.

Response after BACOP*   
 CR2626 
 PR6868 
 SD-PD22 
 Early withdrawal43 
Response after FND or FR   
 CR5772 
 PR3416 
 SD-PD52 
 Early withdrawal54 
Response after IF-RT*   
 Overall response9093.801
 CR6279.014
 PR2814 
 SD-PD52 
 Early withdrawal54 
Table 3. Association Between the Baseline Characteristics of Patients Enrolled in the BACOP/FND and BACOP/FR Trials and the Complete Response Rate at the End of Treatment
CharacteristicBACOP/ FND, %PBACOP/ FR, %P
  • BACOP indicates bleomycin, epidoxorubicin, cyclophosphamide, vincristine, and prednisone; FND, fludarabine, mitoxantrone, dexamethasone; FR, fludarabine and rituximab; AA, Ann Arbor; II bulk/B, with bulky disease or B symptoms; LDH, lactate dehydrogenase; ULN, upper limit of normality; Hb, hemoglobin; FLIPI, Follicular Lymphoma International Prognostic Index.

  • *

    Data on FLIPI were missing for 15 of 125 patients on the BACOP/FND trial and for 2 of 80 patients on the BACOP/FR trial.

Complete response rate62 79 
Sex    
 Men60.71377.782
 Women64 82 
Age at diagnosis, y    
 ≤6065.575781.00
 >6059 81 
AA stage    
 II bulk/B80.31833.017
 III-IV61 82 
LDH    
 ≤ULN67.46285.115
 >ULN57 62 
Hb, g/dL    
 ≥1267.23078.446
 <1252 92 
No. of lymph node areas    
 0-468.12881.785
 >453 76 
FLIPI*    
 0-175.06779.818
 269 79 
 3-548 86 
No. of extranodal sites    
 0-168.023781.00
 >145 80 

Overall, 119 patients were randomized to receive maintenance treatment: including 55 patients in the arm with dexamethasone and IFN and 64 patients in the control arm (observation). No significant differences were observed in FFS or OS. At the time of the last follow-up, 35 patients had died, and 5 patients had been lost to follow-up; whereas 85 patients remained alive, including 81 patients with an ongoing response and 4 patients with progressive disease. After a median follow-up of 60 months (range, 1-124 months), the FFS and OS rates were 53% and 77% at 4 years, respectively, and 45% and 72% at 5 years, respectively (Fig. 1). A univariate analysis of baseline parameters revealed a statistically negative impact on the FFS and OS rates for a high FLIPI score (with several factors determining the score) and for >1 extranodal site (Table 4). No impact of IF-RT on FFS and OS was observed.

Figure 1.

(A) Overall survival (OS) and (B) failure-free survival (FFS) in patients who received combined bleomycin, epidoxorubicin, cyclophosphamide, vincristine, and prednisone (BACOP) plus fludarabine, mitoxantrone, and dexamethasone (BACOP/FND) and patients who received combined BACOP plus fludarabine and rituximab (BACOP/FR) according to an intent-to-treat analysis.

Table 4. Association of Baseline Characteristics in Patients Enrolled on the BACOP/FND Trial With Overall and Failure-free Survival at 48 Months
CharacteristicOS, %PFFS, %P
  • OS indicates overall survival; FFS, failure-free survival; 95% CI, 95% confidence interval; AA, Ann Arbor; II bulk/B, with bulky disease or B symptoms; LDH, lactate dehydrogenase; ULN, upper limit of normality; Hb, hemoglobin; FLIPI, Follicular Lymphoma International Prognostic Index.

  • *

    Data were missing on 15 of 125 patients (12%).

Survival (95% CI)77 (68-84) 53 (43-62) 
Sex    
 Men77.86056.607
 Women78 54 
Age at diagnosis, y    
 ≤6086.01762.039
 >6064 44 
AA stage    
 II bulk/B89.68274.103
 III-IV76 54 
LDH    
 ≤ULN82.00262.011
 >ULN61 33 
Hb, g/dL    
 ≥1281.02360.075
 <1265 36 
No. of lymph node areas    
 0-482.23065.003
 >468 37 
FLIPI*    
 0-197.00287<.001
 278 54 
 3-557 28 
No. of extranodal sites    
 0-182.03559.026
 >162 43 
Maintenance therapy    
 Yes81.50157.850
 No83 57 

BACOP/FR

Overall, 94 patients were registered for the BACOP/FR trial. However, 5 patients were excluded because they had either Ann Arbor stage II disease without B symptoms or bulky disease. In addition, 9 patients were excluded for the following reasons: Two patients were not eligible because of misdiagnoses, the bone marrow in 6 patients was not evaluated for bcl-2 rearrangements, and 1 patient refused to start treatment immediately after the informed consent had been obtained. Therefore, 80 patients were considered in the current analysis. Treatment was discontinued early in 8 patients for the following reasons: Two patients were lost at follow-up after 1 and 3 BACOP treatments, and 3 patients had grade 3 or 4 toxicity (1 cardiotoxicity and 2 infections) after 1 to 3 BACOP treatments. All 3 patients attained a CR after salvage treatment, and 3 patients had stable or progressive disease after 3 BACOP treatments and then attained a CR after autologous bone marrow transplantation. Overall, 72 of 80 patients (90%) completed chemotherapy, with or without immunotherapy, in 14 to 30 weeks.

The response rates in the intent-to-treat analysis by phase of treatment are reported in Table 2. After IF-RT, the overall response rate for all patients was 93%, and 63 patients (79%) attained a CR. Univariate analysis of baseline variables versus CR revealed a significant negative impact (P = .02) of either Ann Arbor stage II disease with B symptoms or bulky disease (Table 3). At the time of the last follow-up, 3 patients had died, and 3 patients had been lost at follow-up; whereas 60 patients remained alive with an ongoing response, and 14 patients had progressive disease. After a median follow-up of 36 months (range, 2-56 months), the FFS and OS rates at 4 years were 56% and 97%, respectively (Fig. 1). Univariate analysis revealed a statistically negative impact on FFS of either Ann Arbor stage II disease with B symptoms or bulky disease and a trend for men. A significant negative impact on OS was observed for either Ann Arbor stage II disease with B symptoms or bulky disease, for high levels of lactate dehydrogenase, and for a hemoglobin level <12 g/dL (Table 5). No impact of IF-RT on FFS or OS was observed.

Table 5. Association of Baseline Characteristics in Patients Enrolled on the BACOP/FR Trial With Overall and Failure-free Survival at 48 Months
CharacteristicOS, %PFFS, %P
  • OS indicates overall survival; FFS, failure-free survival; 95% CI, 95% confidence interval; AA, Ann Arbor; II bulk/B, with bulky disease or B symptoms; LDH, lactate dehydrogenase; ULN, upper limit of normality; Hb, hemoglobin; FLIPI, Follicular Lymphoma International Prognostic Index.

  • *

    Data were missing on 2 of 80 patients (2.5%).

Survival (95% CI)97 (90-99) 56 (43-68) 
Sex    
 Men95.16653.061
 Women100 78 
Age at diagnosis, y    
 ≤6096.27266.827
 >60100 60 
AA stage    
 II bulk/B75.00950.010
 III-IV98 66 
LDH    
 ≤ULN100.01364.296
 >ULN92 62 
Hb, g/dL    
 ≥1298.04366.085
 <1291 45 
No. of lymph node areas    
 0-497.66767.443
 >497 59 
FLIPI*    
 0-196.20377.556
 2100 57 
 3-5100 64 
No. of extranodal sites    
 0-198.12369.048
 >195 50 

Polymerase Chain Reaction Assay for bcl-2

Bone marrow aspirates that were obtained before treatment were positive for bcl-2 in 42 of 80 patients on the BACOP/FR trial (53%). Of those 42 patients, 3 patients went off protocol during BACOP treatment, whereas 5 patients (12%) attained a molecularly negative CR. Of the remaining 34 patients, 6 patients in CR were molecularly positive, and 28 patients were in PR, including 22 bcl-2-positive patients and 6 bcl-2-negative patients. Those 34 patients received 4 cycles of FR. After this additional chemoimmunotherapy, 20 patients were in molecularly negative CR, 7 patients were in molecularly positive CR, 2 patients in CR were not evaluated for bcl-2 status, 3 patients were in molecularly negative PR, 1 patient was in molecularly positive PR, and 1 patient in PR was not evaluated for bcl-2 status. In addition, 5 patients were CR negative after 3 BACOP treatments, and 25 of 42 patients who were bcl-2-positive at baseline (60%) attained a molecularly negative CR after chemoimmunotherapy. No significant differences were noted in the rates of CR, FFS, and OS when we compared patients who had positive and negative baseline bcl-2 status. Considering bcl-2 status at the end of treatment, we observed a trend toward an improved FFS rate for patients who became bcl-2 negative (P = .06). In addition, bcl-2-positive patients had a hazard ratio (HR) of 2.35 (95% confidence interval, 0.96-5.83) compared with bcl-2-negative patients, confirming the trend for better FFS in bcl-2-negative patients after treatment.

Toxicity

Toxicity was evaluable in 193 patients. The most commonly noted toxicities were infections, which were observed in 3.6% of patients, and hematologic side effects, such as neutropenia, thrombocytopenia, and anemia, which were reported in 3.1%, 2.6%, and 2.6% of patients, respectively. Overall, the hematologic toxicities were transient and reversible. However, a treatment delay was necessary in patients who suffered grade 3 and 4 neutropenia. In these patients, we observed a median treatment delay of approximately 30 days in both trials. No significant differences were observed in the percentage of grade 3 and 4 side effects for the 2 groups of patients.

Comparison Between the Results of the 2 Trials

The comparison between our 2 phase 2, sequential trials is limited by the retrospective nature of the study, and the results have to be evaluated carefully. There was a significant difference between the observed CR rate for BACOP/FND and BACOP/FR (P = .014). A significant difference also was maintained after adjusting for age. No differences were noted in the FFS rate. The OS at 4 years was significantly different (P = .002) when it was evaluated with the Kaplan-Meier estimate and the log-rank test. In addition, by using Cox analysis before and after and adjusting for age and FLIPI score, we observed an increased HR for BACOP/FND (P = .009). Hematologic side effects and infections were more frequent in the BACOP/FND trial compared with the BACOP/FR trial, and 2 patients died as a consequence of pulmonary infection in the BACOP/FND cohort. However, we did not observe any significant differences regarding toxicity between the 2 trials.

DISCUSSION

The results of retrospective and controlled studies suggest that the addition of rituximab to several different chemotherapies induces an improvement in the survival outcome of patients with advanced follicular lymphoma. In the current study, we report the results from 2 subsequent phase 2 studies using similar chemotherapy schemes with or without the addition of rituximab. To assess whether the improvements in CR and survival observed in patients who received BACOP/FR were a direct result of the addition of rituximab, as opposed to a variation in the patient population, we analyzed the baseline clinical features in the 2 cohorts. We did not observe differences in the FLIPI score or in other prognostic variables in the 2 studies. The only parameter with a significant difference was age, but a Cox analysis that was adjusted for age and FLIPI score revealed an increased HR for BACOP/FND. Although conclusions between nonrandomized groups may depend on differences in observed and unobserved prognostic features, we are confident that the statistical analysis of our results suggests that the addition of rituximab has a favorable effect on the prognosis of advanced follicular lymphoma. The results obtained in the BACOP/FR trial were better in terms of response and OS compared with the results from the BACOP/FND trial.

It is interesting to note that approximately 26% of patients attained a molecularly negative CR (12%) or PR (14%) after only 3 BACOP treatments but before the addition of rituximab. These results agree with others.19 The initial chemotherapy most likely is able to select patients who have disease that is particularly sensitive to chemotherapy or patients with low numbers of bcl-2-positive cells in the bone marrow.20 Furthermore, by using a Cox regression analysis that was adjusted for FLIPI score, patients who received rituximab had better FFS compared with all other patients enrolled in both trials who received chemotherapy only, as indicated by a decreased HR (0.57; 95% confidence interval, 0.34-0.96; P = .035). In conclusion, our data indicate that the response rates and survival outcomes improved significantly with the addition of rituximab to anthracycline-fludarabine–containing chemotherapy regimens in patients with advanced follicular lymphoma, whereas overall toxicity does not increase but remains transient and tolerable.

Acknowledgements

We thank the Italian Group for Lymphoma Study (GISL) trial office staff and, in particular, Raffaella Marcheselli and Caterina Mammi for central data management.

Conflict of Interest Disclosures

Supported by the Associazione Angela Serra per la Ricerca sul Cancro.

Ancillary