the Departments of Pediatrics and Hematopathology, German Lymph Node Registry, University of Kiel, Kiel
http://www.100md.com
《临床肿瘤学》
the Departments of Leukemia, Hematopathology, and Biostatistics, The University of Texas M.D. Anderson Cancer Center, Houston, TX
ABSTRACT
PURPOSE: The efficacy, toxicity, and tolerability of chemoimmunotherapy with the combination of fludarabine, cyclophosphamide, and rituximab (FCR) were evaluated in previously treated patients with chronic lymphocytic leukemia (CLL). The purpose of this study was to improve the complete remission (CR) rate for previously treated patients and evaluate the quality of bone marrow response.
PATIENTS AND METHODS: One hundred seventy-seven previously treated patients with CLL were evaluated. Treatment consisted of rituximab 375 mg/m2 day 1 of course 1 and 500 mg/m2 day 1 of courses 2 to 6; fludarabine 25 mg/m2/d days 2 to 4 of course 1 and days 1 to 3 of courses 2 to 6; and cyclophosphamide 250 mg/m2/d days 2 to 4 of course 1 and days 1 to 3 of courses 2 to 6. Courses were repeated every 4 weeks.
RESULTS: CR was achieved in 25% of 177 patients, and nodular partial remission and partial remission were achieved in 16% and 32% of patients, respectively; the overall response rate was 73%. Twelve (32%) of 37 complete responders tested achieved molecular remission in bone marrow. Univariate and multivariate analyses were used to identify pretreatment patient characteristics associated with CR and overall remission, longer time to progression, and overall survival.
CONCLUSION: The FCR regimen was an active and well-tolerated treatment for previously treated patients with CLL. Myelosuppression was the most common toxicity. FCR induced the highest CR rate reported in a clinical trial of previously treated patients with CLL. Furthermore, molecular remissions were achieved in a third of patients achieving CR.
INTRODUCTION
Fludarabine is the most active single agent for previously treated patients with chronic lymphocytic leukemia (CLL). The combination of fludarabine, cyclophosphamide, and rituximab (FCR) was developed based on in vitro and in vivo data indicating complementary activity with combination. In vitro, fludarabine inhibits repair of cyclophosphamide-induced DNA interstrand cross links.1 In clinical trials, fludarabine combined with cyclophosphamide (FC) increased the remission rate and duration when compared with single-agent fludarabine in previously treated patients with CLL.2,3 Fludarabine can down-modulate the complement-resistance proteins CD46, CD55, and CD59 on leukemia cells, thereby potentially making cells more vulnerable to rituximab-induced complement-mediated lysis.4 In addition, in vitro, rituximab down-modulates interleukin-10 and bcl-2, making CLL B cells potentially more susceptible to fludarabine-induced apoptosis.5 In this study, previously treated patients with CLL received FCR as second-line or salvage treatment.
PATIENTS AND METHODS
Patient Group
November 1999 through December 2001, 177 previously treated patients with CLL were enrolled and treated on protocol. All patients provided informed consent according to institutional guidelines. Patients had received at least one prior treatment and had Rai stage III to IV disease or stage 0 to II disease with a National Cancer Institute Working Group (NCIWG) indication for treatment.6 Patients were required to have adequate performance status (Zubrod performance status 3) and kidney (serum creatinine < 2 mg/dL) and liver (total bilirubin < 2 mg/dL) function. Pretreatment evaluation consisted of a history and physical examination; laboratory studies including CBC and differential, serum electrolytes, renal and liver function tests, and serum beta2-microglobulin (2M). Patients underwent bone marrow aspiration with immunophenotyping and bone marrow biopsy. The majority of patients had standard metaphase karyotype analysis and/or molecular study of clonal immunoglobulin heavy-chain variable gene (IgVH). Neither fluorescence in situ hybridization for prevalent chromosome abnormalities nor sequencing of the IgVH to assess mutational status was performed. Other laboratory and radiographic studies were performed as clinically indicated.
Treatment and Monitoring
Fludarabine 25 mg/m2 and cyclophosphamide 250 mg/m2 were administered on days 2 to 4 of course 1 and on days 1 to 3 of courses 2 to 6, and intravenous rituximab 375 mg/m2 was administered on day 1 of course 1, and rituximab 500 mg/m2 was administered on day 1 of courses 2 to 6. Before rituximab, patients received oral acetaminophen 650 mg and diphenhydramine 25 to 50 mg. Before chemotherapy, patients received intravenous ondansetron 24 mg. At the discretion of the treating physician, patients received prophylaxis for Pneumocystis carinii pneumonia with trimethoprim-sulfamethoxazole or equivalent and prophylaxis for herpes virus with valacyclovir 500 mg daily during treatment. Courses were administered every 4 weeks as permitted by recovery of neutrophil and platelet counts. Patients who achieved a stable partial response or who had continued response after their first three courses received a total of six courses. Patients with no response or progressive disease after one to three courses were considered as having failed treatment and were taken off treatment and observed for progression and survival.
Monitoring of patients consisted of CBC with differential every 1 to 2 week for the first course and every 2 to 4 weeks during therapy thereafter. Physical examination and serum chemistries, including kidney and liver function tests, were performed before each treatment course. Dose adjustments were made if pneumonia, sepsis, or other life-threatening infection occurred or if the neutrophil or platelet count did not return to pretreatment level by day 35 of the course. Nonhematologic toxicity of grade 3 or 4 by National Cancer Institutes Common Toxicity Criteria (http://ctep.cancer.gov/reporting/ctc.html) required dose reductions to level –1 or –2, respectively. Dose level –1 was fludarabine 25 mg/m2 daily for 3 days and cyclophosphamide 200 mg/m2 daily for 3 days, and dose level –2 was fludarabine 20 mg/m2 and cyclophosphamide 200 mg/m2 daily for 3 days. There were no dose adjustments for rituximab. Patients were evaluated after their third and sixth course with physical examination, CBC with differential, serum chemistries, and bone marrow biopsy and aspiration with immunophenotyping and, in some cases, IgVH molecular studies.
Response Criteria
Patients were evaluated for response using the NCIWG criteria.7 Molecular monitoring for residual disease was performed using a polymerase chain reaction (PCR)-based ligase assay for patient-specific clonal IgVH (Manshouri et al, submitted for publication). Briefly, genomic DNA was isolated from leukemia cells, and the IgVH CDR3 region was amplified by PCR using FAM-labeled consensus sequence primers. The sequences of the FR3 and J primers are 5'-ACACGGCCGTGTATTACTGT-3' and 5'-GTGACCAGGGTNCCTTGGCCCCAG-3', respectively. The resulting unique PCR product was used to generate patient-specific CDR3 segments by digestion with HaeIII restriction enzyme. The digested fragments, including the FAM-labeled fragment, were isolated by gel electrophoresis and subsequently used for a ligase reaction. The RAS oncogene was amplified with each IgVH reaction using fluorescent primers as previously described.8 The amplification products of the RAS oncogene were used to verify the integrity of the DNA and to normalize the levels of CDR3 amplification products in each sample. The ligase reaction was performed using standard reaction. The products were quantified using the ABI Prism 310 or 3100 genetic analyzers (Applied Biosystems, Foster City, CA). Peak areas were measured using GeneScan analysis software (Applied Biosystems). The ratio of IgVH to RAS (PCR ratio) was calculated and used to quantify the levels of residual disease.9,10 A ratio of 0.001 to 0.10 was considered to be low positive ratios, and higher ratios were considered to be high positive ratios. This assay was able to detect one leukemia cell in 105 mononuclear cells, as determined by cell-mixing studies.
Statistical Considerations
The goal of this study was to improve the complete remission (CR) rate, and the statistical design evaluated both response and morbidity according to the method of Bryant and Day.11 On the basis of a historical CR rate of 12% for FC,2 the desirable CR rate was 25%, with an acceptable toxicity rate being 20%; a 40% toxicity rate was unacceptable.
Associations between categoric patient characteristics and CR or overall response (OR) were evaluated by 2 analysis.12 The Wilcoxon two-sample test was used to evaluate associations between continuous patient characteristics and CR or OR. For multivariable analysis, logistic regression analysis was the statistical methodology used for examining the relationship of a binary dependent variable (CR v nodular partial remission [NPR], partial remission [PR], and no response, or OR [CR + NPR + PR] v no response) with one or more independent variables (for example, age, sex, and so on). The independent variables could be continuous or categoric. There were more than 10 potentially important independent variables in this study; therefore, a stepwise (backward) procedure was carried out once a full model was fitted.
The distribution of time to progression and survival were estimated by the Kaplan-Meier method.13 The Kaplan-Meier curves for the categoric variables were plotted for disease-free and overall survival. The P values for testing the differences between subgroups and levels for each variable were calculated by the log-rank test.14 Time intervals were measured from the first day of treatment until progression, relapse, or death, and deaths from all causes were included.
The Cox proportional hazards model was used to perform survival regression analysis and, thereby, examine the relationship of survival time with independent variables.15 Additionally, the final Cox model for time to disease progression was fitted.
RESULTS
Patient Characteristics
Patient characteristics are listed in Table 1. Five patients (3%) had low-risk disease, 84 (47%) had intermediate-risk disease, and 88 (50%) had high-risk disease using the modified Rai staging criteria.16 Immunostaining and flow cytometry analysis of bone marrow lymphocytes demonstrated that 107 (68%) of 158 patients had 30% of lymphocytes expressing CD38; and 32% of patients had more than 30% CD38-positive lymphocytes. Metaphase karyotype analysis on bone marrow aspirate was obtained in 129 patients. Three patients (2%) had 13q deletion as the sole abnormality, 84 patients (65%) had diploid karyotype, and 42 patients (33%) had complex abnormalities, 11q deletion, 17p deletion, or trisomy 12 as sole abnormalities or as part of a complex abnormal karyotype.
The median number of prior treatments was two (range, one to 10 treatments). Thirty-two patients (18%) had received only alkylating agent therapies; 145 patients (82%) had previously received fludarabine alone or in combination, 108 patients (61%) had fludarabine-sensitive disease, and 37 patients (21%) were refractory to fludarabine. Fludarabine refractoriness was defined as failure to achieve at least a PR with the last fludarabine-based treatment or progression within 6 months of treatment. Twenty-two patients (12%) had previously received rituximab, either alone or in combination; and seven patients received rituximab as their only prior treatment. Thirty-four patients (19%) were previously treated with FC; 30 of these patients had fludarabine-sensitive disease, and four patients were refractory to fludarabine.
Response
CR was achieved in 45 patients (25%), NPR was achieved in 28 patients (16%), and PR was achieved in 57 patients (32%), for an OR rate of 73%. Responses by prior treatment are listed in Table 2. CR and OR rates were comparable for patients previously treated with an alkylating agent versus patients previously treated with fludarabine with or without cyclophosphamide.
Univariate Analysis and Multivariate Analysis
Pretreatment characteristics were evaluated for correlations with response (CR and OR) in univariate analysis of categoric variables and continuous variables using 2 (Table 3) and Wilcoxon two-sample (Table 4) analysis, respectively. CR was significantly (P < .05) associated with younger age (mean, 55 years), earlier Rai stage (P = .01), fewer number of prior treatments (P < .001), remission with last fludarabine-based regimen (P = .002), lower 2M ( 4 mg/L; P < .002), higher hemoglobin (mean, 13 g/dL), higher platelet count (mean, 150 k/μL), and higher albumin (mean, 4.1 g/dL). A greater likelihood for CR was associated with diploid cytogenetics or 13q deletion as the sole karyotype abnormality (P = .083) versus abnormal cytogenetics. No statistically significant association was noted between CR and sex, pretreatment WBC count, absolute lymphocyte count, creatinine, percentage of lymphocytes or prolymphocytes in bone marrow, or presence of less than 20% to 30% CD38-positive lymphocytes in bone marrow. OR was significantly (P < .05) associated with earlier Rai stage (P < .001), remission with last fludarabine-based treatment (P = .039), lower 2M ( 4 mg/L; P < .02), higher hemoglobin (mean, 12.4 g/dL) and platelet count (mean, 137 k/μL), diploid cytogenetics or 13q deletion as the sole karyotype abnormality (P = .001), and presence of less than 30% CD38-positive lymphocytes in bone marrow (P < .001). No statistically significant association was noted between OR and age, sex, number of prior treatments, albumin, WBC count, absolute lymphocyte count, creatinine, or percentage of lymphocytes or prolymphocytes present in bone marrow.
Logistical regression multivariate analysis was performed to identify characteristics (both categoric and continuous) associated with CR and OR (Table 5). The following pretreatment variables were analyzed in the model: age, sex, Rai stage, refractoriness to fludarabine, number of prior treatments, WBC count, absolute lymphocyte count, hemoglobin, platelet count, creatinine, 2M, albumin, percentage of CD38-positive lymphocytes in bone marrow, and abnormal karyotype. Characteristics that were not significant were dropped from the model in a stepwise fashion; the final model contained only statistically significant characteristics. In the final model, CR was associated with higher pretreatment platelet count, fewer number of prior treatments, and lower pretreatment 2M. In the final model, OR was associated with higher pretreatment hemoglobin, and treatment failure (nonresponders) was associated with fludarabine refractoriness and higher pretreatment creatinine. There was no correlation between advanced age and higher creatinine (data not shown).
Quality of Remission
Molecular evidence of residual disease was assessed by performing PCR to detect the IgVH gene of the malignant clone in the bone marrow of 37 patients achieving CR; in 12 patients (32%), the clonal population was not detected (Table 6). PCR was performed in 16 patients with NPR, and two patients (13%) were negative. Of the 22 patients with PRs who were tested, two patients (9%) were negative for the presence of the malignant population. Two-color immunostaining and flow cytometry analysis for CD5-positive and CD19-positive lymphocytes was also performed on bone marrow in some patients (Table 6).
Disease Progression and Survival
The current median follow-up time for all patients is 28 months (range, 1 to 50 months), and for surviving patients, it is 35 months (range, 14 to 50 months). The Kaplan-Meier estimates of time to progression for responding patients and overall survival for all patients are shown in Figure 1. The median time to progression was 28 months (range, 3 to 49 months). Moreover, the median time to progression for patients achieving CR, NPR, and PR was 39, 33, and 15 months, respectively (Fig 2). Time to progression by bone marrow PCR status at end of treatment is shown in Figure 3. There is not yet a statistical difference in time to progression by bone marrow PCR status in patients with available data; however, follow-up continues.
The multivariate Cox proportional hazards models was used to explore correlations between pretreatment characteristics and time to progression for the 129 responders and survival for the entire 177 patients (Table 7). The following variables were analyzed in the model: age, Rai stage, response to FCR, refractoriness to fludarabine, number of prior treatments, WBC count, absolute lymphocyte count, hemoglobin, platelet count, creatinine, 2M, albumin, percentage of CD38-positive cells in bone marrow, percentage of lymphocytes and CD5-positive and CD19-positive cells in bone marrow, and abnormal karyotype. In the Cox model for time to progression, there were three significant independent variables; these variables were 2M, response to FCR, and cytogenetics (Table 7). Time to treatment failure was significantly shorter for patients with increased 2M (hazard ratio, 1.1). The time to progression was significantly shorter for patients not achieving a remission versus responders (CR, NPR, or PR; hazard ratio, 7.8). Patients with diploid karyotype or 13q deletion as a sole abnormality had longer time to progression than patients with karyotype abnormalities.
The Kaplan-Meier estimate of overall survival is shown in Figure 1. The estimated median survival for the 177 patients treated on this study is 42 months (range, 1 to 50 months). Survival by response to FCR is shown in Figure 4. The estimated median survival times for all responders, complete responders, and nodular partial responders have not been reached but are estimated to be 45 and 30 months, respectively. The estimated median survival time for the partial responders is 39 months; for nonresponders, it is 11 months (Fig 4). In the Cox model for survival, there were three significant independent variables; these were 2M, response to FCR, and cytogenetics (Table 7). The hazard function for risk for death increased with higher levels of 2M, failure to achieve a CR, NPR, or PR, and unfavorable karyotype (Table 7).
Toxicity
All six intended courses of treatment were administered to 81 (46%) of 177 patients, and 68 of these patients received full-dose treatment. Sixty-six patients (37%) received one to three courses, 16 patients (9%) received four courses, and 14 patients (8%) received five courses of FCR. Myelosuppression was the most common reason for discontinuing treatment before completing six courses and occurred in 46 patients (26%). Sixteen patients (9%) stopped treatment because of disease resistance, and 11 patients (6%) stopped because of infection. Other reasons for early treatment cessation included death (n = 8), physician preference (n = 4), stem-cell transplantation (n = 2), and Richter’s transformation (n = 2). Dose reduction occurred in 25 patients, and four patients had two dose reductions. The primary reason for dose reduction was myelosuppression.
Rituximab-associated toxicities were associated with the first infusion. One hundred eleven patients (63%) had adverse events associated with the first infusion of rituximab, and all events were grade 1 or 2, with the exception of one episode of grade 3 shortness of breath, and all were self-limited. There were no grade 4 toxicities associated with any rituximab infusion. Of the grade 1 to 2 toxicities, fever occurred in 65 patients (37%), chills occurred in 89 patients (50%), hypotension occurred in 23 patients (13%), nausea and vomiting occurred in 22 patients (12%), and shortness of breath occurred in five patients (2%). Rash, fatigue, headache, and arthralgias were all grade 1 and occurred in less than 5% of patients.
The most common toxicity with FCR was neutropenia, with grade 3 and 4 neutropenia noted in 21% and 41% of 529 assessable courses. Fifteen percent and 66% of patients experienced one or more episode of grade 3 or 4 neutropenia, respectively, as their most severe grade of toxicity during treatment. Thrombocytopenia occurred in 17% of 539 assessable courses (grade 3 in 10% and grade 4 in 7%). Sixteen percent and 18% of patients experienced one or more episode of grade 3 or 4 thrombocytopenia, respectively, as their most severe grade of toxicity during treatment. Anemia (grade 3 or 4) was observed in 24% of patients during this study.
Major infections, defined as sepsis, pneumonia, or infection requiring hospitalization, occurred in 34 (5%) of 745 assessable courses or 16% of treated patients. Fever of unknown origin occurred in 78 (10%) of 745 assessable courses or 31% of treated patients, and minor infections occurred in 59 (8%) of 745 assessable courses or 18% or treated patients. Minor infections were defined as upper respiratory tract infection, bronchitis, cellulitis, and so on. One patient was removed from study because of cytomegalovirus pneumonitis. Herpes simplex and zoster were associated with four (1%) and six (1%) of 745 assessable courses, respectively. The frequency of major infections, herpes reactivation, and minor infections was similar for fludarabine-refractory patients compared with fludarabine-sensitive patients.
We defined late neutropenia as grade 3 or 4 neutropenia that occurred more than 30 days after the last course of FCR in patients who had recovered their absolute neutrophil count to greater than 1,000/μL. Late neutropenia occurred in five of 45 CR patients. It was transient in all cases, and bone marrow biopsy at the time of neutropenia showed no disease in any of these patients. The grade 4 events were isolated events; one occurred on day 82 and the other occurred on day 321 after the last course of FCR. One of the grade 3 events lasted 3 weeks; the others were isolated events. There were five other late neutropenic events in CR patients; however, at the time of the event, all of these patients had lymphoid aggregates or overt disease relapse in bone marrow.
To date, one patient developed acute myelogenous leukemia (French-American-British classification M5) 2 months after his fourth course of FCR. This patient had previously received single-agent fludarabine, and there were no cytogenetic abnormalities associated with his acute myelogenous leukemia. Four patients have developed myelodysplastic syndrome (1.5 years, 1 year, 9 months, and 3 months after five or six courses of FCR). Three of these patients had cytogenetic abnormalities (–5,–7 in two patients and –7 in one patient). Two of these patients were previously treated with single-agent purine analog, and two were previously treated with purine analog combined with alkylating agent.
DISCUSSION
The ultimate goal for treatment of CLL is cure; however, this has not yet been demonstrated with standard-dose chemotherapy. Historically, responders to treatment live longer and fare better, particularly complete responders. Recently, more sensitive assays, such as flow cytometry and PCR, have been developed to detect residual disease. The results of these assays have not yet been incorporated into formal response criteria. However, early data suggest that negative results with these more sensitive tests may predict longer remission duration and improved survival. The objective in developing new treatments for CLL is to achieve the highest CR rate and the highest quality CR as assessed by sensitive tests such as molecular and flow cytometry assay.
Typically, the CR and OR rates achieved with fludarabine as a salvage treatment were 5% to 10% and 40% to 60%, respectively. The median time to progression was 12 to 13 months, and median survival was 20 to 25 months.17-19 The combination of FC produced CR and OR rates of 10% to 15% and 60% to 70%, respectively, with a median response duration of 20 to 25 months in previously treated patients with CLL.2,3 This report of 177 previously treated patients with CLL receiving FCR is the largest series of such patients treated with a single regimen reported in the literature. The patient characteristics were similar to the characteristics of patients treated on historic single-agent and combination trials. Both CR and OR rates were higher than with single-agent fludarabine or FC. Furthermore, the median time to progression for responding patients was longer (28 months). More importantly, the estimated median overall survival for all patients (42 months) was longer than in historical trials. There are potential limitations regarding the comparability to previously reported single-center phase II studies. These limitations relate, in part, to patient characteristics, treatment, and post-trial therapy. We performed extensive evaluation to identify pretreatment characteristics correlated with response and survival and have reported these. This analysis is intended to provide information that may be applied to patients seen in practice.
Patients who were previously treated with FC were included in this trial. Thirty-four such patients were treated; 30 were fludarabine sensitive, and four were fludarabine refractory. The CR and OR rates for these patients were similar to those for the entire group and to those for patients previously treated with only an alkylating agent (Table 2). This suggests that prior treatment with FC does not have a detrimental impact on response to FCR. Patients who are refractory to fludarabine typically have low response rates with fludarabine-based regimens. Consistent with this, the CR and OR rates in the fludarabine-refractory patients were 6% and 58%, respectively. Treatment of fludarabine-refractory patients with the monoclonal antibody alemtuzumab produced CR and OR rates of 2% and 33%, respectively.20,21 The results with the FCR regimen seem comparable.
The NCIWG criteria for evaluating remission incorporate blood count with differential, morphologic evaluation of the bone marrow, and physical examination for adenopathy and organomegaly, without radiographic evaluation. Radiographic examination by computed tomography of chest and/or abdomen was not routinely performed and not required in this study to confirm response. Although responses tend to be uniform across nodal sites, it is conceivable but unlikely that there was residual intrathoracic or intra-abdominal lymphadenopathy or splenomegaly in a fraction of the responders. Flow cytometry and molecular techniques to detect malignant clonal cells are now being applied to patients as measures of the quality of remission. The importance of molecular remission was recently reported in a clinical trial of fludarabine combined with cyclophosphamide and mitoxantrone in previously treated patients with CLL.22 With fludarabine, cyclophosphamide, and mitoxantrone, 50% of 60 treated patients achieved CR, and 10 of 25 complete responders who were tested were negative for residual disease by PCR in bone marrow. This correlated with longer disease-free and overall survival. With FCR, 32% of 37 complete responders were negative by PCR for residual disease in bone marrow, and this was correlated with longer progression-free survival. Further follow-up is needed to fully assess a potential survival benefit in these patients.
Identifying prognostic features for response, duration of response, and survival is important because it aids in counseling patients before treatment and in stratifying patients in clinical trials. In addition, insight into the biology of the disease and the mechanism of action of therapeutic agents may be gained. Age was correlated with the CR rate but not the OR rate. Also, advanced age correlated with fewer number of courses of FCR received. Patients of advanced age may tolerate less chemotherapy, possibly because of attrition of normal bone marrow elements. Alternatively, older patients may have disease that is intrinsically resistant to treatment. Both hemoglobin and platelet count correlated with CR and OR rates; lower hemoglobin and platelet count predicted for lower CR and OR rates. Lower hemoglobin and platelet count may reflect more extensive bone marrow infiltration and/or resistance to treatment in cells residing in bone marrow. The use of growth factors during treatment may benefit such patients and merits evaluation.
The log-rank test demonstrated that lower 2M, higher albumin, early Rai stage, and response (CR and OR) to FCR correlated with longer survival. For both longer time to progression and overall survival, multivariate Cox model analysis demonstrated a correlation with lower serum 2M, diploid cytogenetics or 13q deletion as a sole chromosomal abnormality, and response (CR, NPR, or PR) to FCR. Lower 2M was also associated with likelihood for CR and OR, and 4.0 mg/L (normal range, 0.6 to 2.0 mg/dL) was a significant cut point. The biologic basis of this correlation is unclear, but the association has repeatedly been demonstrated in CLL as well as other B-cell malignancies such as multiple myeloma and lymphoma. The importance of 2M as a prognostic factor for response, time to progression, and survival was intriguing. The source of the soluble 2M is most likely the leukemia cells; however, this has not been definitively demonstrated. 2M is expressed on the surface of nucleated cells as a heterodimer with HLA class I. It is a well-established prognostic factor in CLL and other B-cell neoplasms. 2M is cleared from circulation by the kidneys and may be elevated in patients with renal insufficiency. Patients enrolled onto this study were required to have a creatinine of 2.0 mg/dL. Therefore, creatinine was not likely the major factor influencing 2M levels. Consistent with other studies, abnormal cytogenetics predicted for failure to achieve CR and OR and were associated with shorter time to progression and shorter survival.
The toxicities for FCR were typical and expected for such a group of previously treated patients.2,3 Infusion-related toxicities with rituximab were similar in frequency and severity to the toxicities seen in other studies. FC has been evaluated in previously treated patients with CLL.2,3 Myelosuppression was also the most frequent toxicity in these studies. In one previous study of 101 previously treated patients, one or more cycle of treatment was associated with grade 3 and 4 neutropenia in 75% and 48% of patients, respectively. Of note, these patients received higher doses of fludarabine (30 mg/m2 daily for 3 days) and cyclophosphamide (300 mg/m2 daily for 3 days).2 In this previous trial, infections were more frequent, with sepsis or pneumonia in 25% of patients and fever of unknown origin in 25% of patients. This may be related to differences in use of prophylactic antibiotics. The incidence of grade 3 or 4 thrombocytopenia in patients treated with FC was 19%, which is similar to the incidence noted in this study with FCR. Thus, the incidence of neutropenia, thrombocytopenia, and infection was similar for patients who received FCR compared with previously treated patients in a previous trial who received FC, indicating that the addition of rituximab did not significantly increase toxicity. Late neutropenia was transient and rare in the absence of relapsed or residual disease in the bone marrow.
FCR is a well-tolerated regimen with significant activity, even for patients who have received a substantial amount of prior therapy. Work continues in developing new combination regimens with complementary agents, with the overall objective to improve CR rate and quality of CR.
Authors’ Disclosures of Potential Conflicts of Interest
The following authors or their immediate family members have indicated a financial interest. No conflict exists for drugs or devices used in a study if they are not being evaluated as part of the investigation. Performed contract work within the last 2 years: William Wierda, Berlex Laboratories Inc, Xycte Therapies Inc, Ligand Pharmaceuticals.
NOTES
Authors’ disclosures of potential conflicts of interest are found at the end of this article.
REFERENCES
Yamauchi T, Nowak BJ, Keating MJ, et al: DNA repair initiated in chronic lymphocytic leukemia lymphocytes by 4-hydroperoxycyclophosphamide is inhibited by fludarabine and clofarabine. Clin Cancer Res 7:3580-3589, 2001
O’Brien SM, Kantarjian HM, Cortes J, et al: Results of the fludarabine and cyclophosphamide combination regimen in chronic lymphocytic leukemia. J Clin Oncol 19:1414-1420, 2001
Hallek M, Schmitt B, Wilhelm M, et al: Fludarabine plus cyclophosphamide is an efficient treatment for advanced chronic lymphocytic leukaemia (CLL): Results of a phase II study of the German CLL Study Group. Br J Haematol 114:342-348, 2001
Di Gaetano N, Xiao Y, Erba E, et al: Synergism between fludarabine and rituximab revealed in a follicular lymphoma cell line resistant to the cytotoxic activity of either drug alone. Br J Haematol 114:800-809, 2001
Alas S, Emmanouilides C, Bonavida B: Inhibition of interleukin 10 by rituximab results in down-regulation of bcl-2 and sensitization of B-cell non-Hodgkin’s lymphoma to apoptosis. Clin Cancer Res 7:709-723, 2001
Cheson BD, Bennett JM, Grever M, et al: National Cancer Institute-sponsored Working Group guidelines for chronic lymphocytic leukemia: Revised guidelines for diagnosis and treatment. Blood 87:4990-4997, 1996
Cheson BD, Bennett JM, Rai KR, et al: Guidelines for clinical protocols for chronic lymphocytic leukemia: Recommendations of the National Cancer Institute-sponsored Working Group. Am J Hematol 29:152-163, 1988
Slavickova A, Ullmannova V, Klener P: Optimized multiplex IgH/ras PCR: A tool for quantitative monitoring of B-lymphoproliferative disorders. Biotechniques 28:716-718, 720-721, 2000
Albitar M, Wu WI, Feltz E, et al: Simplified reverse dot blot analyses for detecting of ras oncogene mutations. Mol Diagn 2:169-176, 1997
Schultze JL, Donovan JW, Gribben JG: Minimal residual disease detection after myeloablative chemotherapy in chronic lymphatic leukemia. J Mol Med 77:259-265, 1999
Bryant J, Day R: Incorporating toxicity considerations into the design of two-stage phase II clinical trials. Biometrics 51:1372-1383, 1995
Mantel N: Evaluation of survival data and two new rank order statistics arising in its consideration. Cancer Chemother Rep 50:163-170, 1966
Kaplan EL, Meier P: Non-parametric estimation from incomplete observation. J Am Stat Assoc 53:457-481, 1958
Peto R, Pike MC, Armitage P, et al: Design and analysis of randomized clinical trials requiring prolonged observation of each patient: II. Analysis and examples. Br J Cancer 35:1-39, 1977
Cox DR, Snell EJ: Analysis of Binary Data (ed 2). London, United Kingdom, Chapman and Hall, 1989
Rai KR: A critical analysis of staging in CLL, in Gale RP, Rai KR (eds): Chronic Lymphocytic Leukemia: Recent Progress and Future Directions. New York, NY, Alan R. Liss, 1987, pp 253
Keating MJ, Kantarjian H, Talpaz M, et al: Fludarabine: A new agent with major activity against chronic lymphocytic leukemia. Blood 74:19-25, 1989
O’Brien S, Kantarjian H, Beran M, et al: Results of fludarabine and prednisone therapy in 264 patients with chronic lymphocytic leukemia with multivariate analysis-derived prognostic model for response to treatment. Blood 82:1695-1700, 1993
Fenchel K, Bergmann L, Wijermans P, et al: Clinical experience with fludarabine and its immunosuppressive effects in pretreated chronic lymphocytic leukemias and low-grade lymphomas. Leuk Lymphoma 18:485-492, 1995
Keating MJ, Flinn I, Jain V, et al: Therapeutic role of alemtuzumab (Campath-1H) in patients who have failed fludarabine: Results of a large international study. Blood 99:3554-3561, 2002
Rai KR, Freter CE, Mercier RJ, et al: Alemtuzumab in previously treated chronic lymphocytic leukemia patients who also had received fludarabine. J Clin Oncol 20:3891-3897, 2002
Bosch F, Ferrer A, Lopez-Guillermo A, et al: Fludarabine, cyclophosphamide and mitoxantrone in the treatment of resistant or relapsed chronic lymphocytic leukaemia. Br J Haematol 119:976-984, 2002(William Wierda, Susan O’B)
ABSTRACT
PURPOSE: The efficacy, toxicity, and tolerability of chemoimmunotherapy with the combination of fludarabine, cyclophosphamide, and rituximab (FCR) were evaluated in previously treated patients with chronic lymphocytic leukemia (CLL). The purpose of this study was to improve the complete remission (CR) rate for previously treated patients and evaluate the quality of bone marrow response.
PATIENTS AND METHODS: One hundred seventy-seven previously treated patients with CLL were evaluated. Treatment consisted of rituximab 375 mg/m2 day 1 of course 1 and 500 mg/m2 day 1 of courses 2 to 6; fludarabine 25 mg/m2/d days 2 to 4 of course 1 and days 1 to 3 of courses 2 to 6; and cyclophosphamide 250 mg/m2/d days 2 to 4 of course 1 and days 1 to 3 of courses 2 to 6. Courses were repeated every 4 weeks.
RESULTS: CR was achieved in 25% of 177 patients, and nodular partial remission and partial remission were achieved in 16% and 32% of patients, respectively; the overall response rate was 73%. Twelve (32%) of 37 complete responders tested achieved molecular remission in bone marrow. Univariate and multivariate analyses were used to identify pretreatment patient characteristics associated with CR and overall remission, longer time to progression, and overall survival.
CONCLUSION: The FCR regimen was an active and well-tolerated treatment for previously treated patients with CLL. Myelosuppression was the most common toxicity. FCR induced the highest CR rate reported in a clinical trial of previously treated patients with CLL. Furthermore, molecular remissions were achieved in a third of patients achieving CR.
INTRODUCTION
Fludarabine is the most active single agent for previously treated patients with chronic lymphocytic leukemia (CLL). The combination of fludarabine, cyclophosphamide, and rituximab (FCR) was developed based on in vitro and in vivo data indicating complementary activity with combination. In vitro, fludarabine inhibits repair of cyclophosphamide-induced DNA interstrand cross links.1 In clinical trials, fludarabine combined with cyclophosphamide (FC) increased the remission rate and duration when compared with single-agent fludarabine in previously treated patients with CLL.2,3 Fludarabine can down-modulate the complement-resistance proteins CD46, CD55, and CD59 on leukemia cells, thereby potentially making cells more vulnerable to rituximab-induced complement-mediated lysis.4 In addition, in vitro, rituximab down-modulates interleukin-10 and bcl-2, making CLL B cells potentially more susceptible to fludarabine-induced apoptosis.5 In this study, previously treated patients with CLL received FCR as second-line or salvage treatment.
PATIENTS AND METHODS
Patient Group
November 1999 through December 2001, 177 previously treated patients with CLL were enrolled and treated on protocol. All patients provided informed consent according to institutional guidelines. Patients had received at least one prior treatment and had Rai stage III to IV disease or stage 0 to II disease with a National Cancer Institute Working Group (NCIWG) indication for treatment.6 Patients were required to have adequate performance status (Zubrod performance status 3) and kidney (serum creatinine < 2 mg/dL) and liver (total bilirubin < 2 mg/dL) function. Pretreatment evaluation consisted of a history and physical examination; laboratory studies including CBC and differential, serum electrolytes, renal and liver function tests, and serum beta2-microglobulin (2M). Patients underwent bone marrow aspiration with immunophenotyping and bone marrow biopsy. The majority of patients had standard metaphase karyotype analysis and/or molecular study of clonal immunoglobulin heavy-chain variable gene (IgVH). Neither fluorescence in situ hybridization for prevalent chromosome abnormalities nor sequencing of the IgVH to assess mutational status was performed. Other laboratory and radiographic studies were performed as clinically indicated.
Treatment and Monitoring
Fludarabine 25 mg/m2 and cyclophosphamide 250 mg/m2 were administered on days 2 to 4 of course 1 and on days 1 to 3 of courses 2 to 6, and intravenous rituximab 375 mg/m2 was administered on day 1 of course 1, and rituximab 500 mg/m2 was administered on day 1 of courses 2 to 6. Before rituximab, patients received oral acetaminophen 650 mg and diphenhydramine 25 to 50 mg. Before chemotherapy, patients received intravenous ondansetron 24 mg. At the discretion of the treating physician, patients received prophylaxis for Pneumocystis carinii pneumonia with trimethoprim-sulfamethoxazole or equivalent and prophylaxis for herpes virus with valacyclovir 500 mg daily during treatment. Courses were administered every 4 weeks as permitted by recovery of neutrophil and platelet counts. Patients who achieved a stable partial response or who had continued response after their first three courses received a total of six courses. Patients with no response or progressive disease after one to three courses were considered as having failed treatment and were taken off treatment and observed for progression and survival.
Monitoring of patients consisted of CBC with differential every 1 to 2 week for the first course and every 2 to 4 weeks during therapy thereafter. Physical examination and serum chemistries, including kidney and liver function tests, were performed before each treatment course. Dose adjustments were made if pneumonia, sepsis, or other life-threatening infection occurred or if the neutrophil or platelet count did not return to pretreatment level by day 35 of the course. Nonhematologic toxicity of grade 3 or 4 by National Cancer Institutes Common Toxicity Criteria (http://ctep.cancer.gov/reporting/ctc.html) required dose reductions to level –1 or –2, respectively. Dose level –1 was fludarabine 25 mg/m2 daily for 3 days and cyclophosphamide 200 mg/m2 daily for 3 days, and dose level –2 was fludarabine 20 mg/m2 and cyclophosphamide 200 mg/m2 daily for 3 days. There were no dose adjustments for rituximab. Patients were evaluated after their third and sixth course with physical examination, CBC with differential, serum chemistries, and bone marrow biopsy and aspiration with immunophenotyping and, in some cases, IgVH molecular studies.
Response Criteria
Patients were evaluated for response using the NCIWG criteria.7 Molecular monitoring for residual disease was performed using a polymerase chain reaction (PCR)-based ligase assay for patient-specific clonal IgVH (Manshouri et al, submitted for publication). Briefly, genomic DNA was isolated from leukemia cells, and the IgVH CDR3 region was amplified by PCR using FAM-labeled consensus sequence primers. The sequences of the FR3 and J primers are 5'-ACACGGCCGTGTATTACTGT-3' and 5'-GTGACCAGGGTNCCTTGGCCCCAG-3', respectively. The resulting unique PCR product was used to generate patient-specific CDR3 segments by digestion with HaeIII restriction enzyme. The digested fragments, including the FAM-labeled fragment, were isolated by gel electrophoresis and subsequently used for a ligase reaction. The RAS oncogene was amplified with each IgVH reaction using fluorescent primers as previously described.8 The amplification products of the RAS oncogene were used to verify the integrity of the DNA and to normalize the levels of CDR3 amplification products in each sample. The ligase reaction was performed using standard reaction. The products were quantified using the ABI Prism 310 or 3100 genetic analyzers (Applied Biosystems, Foster City, CA). Peak areas were measured using GeneScan analysis software (Applied Biosystems). The ratio of IgVH to RAS (PCR ratio) was calculated and used to quantify the levels of residual disease.9,10 A ratio of 0.001 to 0.10 was considered to be low positive ratios, and higher ratios were considered to be high positive ratios. This assay was able to detect one leukemia cell in 105 mononuclear cells, as determined by cell-mixing studies.
Statistical Considerations
The goal of this study was to improve the complete remission (CR) rate, and the statistical design evaluated both response and morbidity according to the method of Bryant and Day.11 On the basis of a historical CR rate of 12% for FC,2 the desirable CR rate was 25%, with an acceptable toxicity rate being 20%; a 40% toxicity rate was unacceptable.
Associations between categoric patient characteristics and CR or overall response (OR) were evaluated by 2 analysis.12 The Wilcoxon two-sample test was used to evaluate associations between continuous patient characteristics and CR or OR. For multivariable analysis, logistic regression analysis was the statistical methodology used for examining the relationship of a binary dependent variable (CR v nodular partial remission [NPR], partial remission [PR], and no response, or OR [CR + NPR + PR] v no response) with one or more independent variables (for example, age, sex, and so on). The independent variables could be continuous or categoric. There were more than 10 potentially important independent variables in this study; therefore, a stepwise (backward) procedure was carried out once a full model was fitted.
The distribution of time to progression and survival were estimated by the Kaplan-Meier method.13 The Kaplan-Meier curves for the categoric variables were plotted for disease-free and overall survival. The P values for testing the differences between subgroups and levels for each variable were calculated by the log-rank test.14 Time intervals were measured from the first day of treatment until progression, relapse, or death, and deaths from all causes were included.
The Cox proportional hazards model was used to perform survival regression analysis and, thereby, examine the relationship of survival time with independent variables.15 Additionally, the final Cox model for time to disease progression was fitted.
RESULTS
Patient Characteristics
Patient characteristics are listed in Table 1. Five patients (3%) had low-risk disease, 84 (47%) had intermediate-risk disease, and 88 (50%) had high-risk disease using the modified Rai staging criteria.16 Immunostaining and flow cytometry analysis of bone marrow lymphocytes demonstrated that 107 (68%) of 158 patients had 30% of lymphocytes expressing CD38; and 32% of patients had more than 30% CD38-positive lymphocytes. Metaphase karyotype analysis on bone marrow aspirate was obtained in 129 patients. Three patients (2%) had 13q deletion as the sole abnormality, 84 patients (65%) had diploid karyotype, and 42 patients (33%) had complex abnormalities, 11q deletion, 17p deletion, or trisomy 12 as sole abnormalities or as part of a complex abnormal karyotype.
The median number of prior treatments was two (range, one to 10 treatments). Thirty-two patients (18%) had received only alkylating agent therapies; 145 patients (82%) had previously received fludarabine alone or in combination, 108 patients (61%) had fludarabine-sensitive disease, and 37 patients (21%) were refractory to fludarabine. Fludarabine refractoriness was defined as failure to achieve at least a PR with the last fludarabine-based treatment or progression within 6 months of treatment. Twenty-two patients (12%) had previously received rituximab, either alone or in combination; and seven patients received rituximab as their only prior treatment. Thirty-four patients (19%) were previously treated with FC; 30 of these patients had fludarabine-sensitive disease, and four patients were refractory to fludarabine.
Response
CR was achieved in 45 patients (25%), NPR was achieved in 28 patients (16%), and PR was achieved in 57 patients (32%), for an OR rate of 73%. Responses by prior treatment are listed in Table 2. CR and OR rates were comparable for patients previously treated with an alkylating agent versus patients previously treated with fludarabine with or without cyclophosphamide.
Univariate Analysis and Multivariate Analysis
Pretreatment characteristics were evaluated for correlations with response (CR and OR) in univariate analysis of categoric variables and continuous variables using 2 (Table 3) and Wilcoxon two-sample (Table 4) analysis, respectively. CR was significantly (P < .05) associated with younger age (mean, 55 years), earlier Rai stage (P = .01), fewer number of prior treatments (P < .001), remission with last fludarabine-based regimen (P = .002), lower 2M ( 4 mg/L; P < .002), higher hemoglobin (mean, 13 g/dL), higher platelet count (mean, 150 k/μL), and higher albumin (mean, 4.1 g/dL). A greater likelihood for CR was associated with diploid cytogenetics or 13q deletion as the sole karyotype abnormality (P = .083) versus abnormal cytogenetics. No statistically significant association was noted between CR and sex, pretreatment WBC count, absolute lymphocyte count, creatinine, percentage of lymphocytes or prolymphocytes in bone marrow, or presence of less than 20% to 30% CD38-positive lymphocytes in bone marrow. OR was significantly (P < .05) associated with earlier Rai stage (P < .001), remission with last fludarabine-based treatment (P = .039), lower 2M ( 4 mg/L; P < .02), higher hemoglobin (mean, 12.4 g/dL) and platelet count (mean, 137 k/μL), diploid cytogenetics or 13q deletion as the sole karyotype abnormality (P = .001), and presence of less than 30% CD38-positive lymphocytes in bone marrow (P < .001). No statistically significant association was noted between OR and age, sex, number of prior treatments, albumin, WBC count, absolute lymphocyte count, creatinine, or percentage of lymphocytes or prolymphocytes present in bone marrow.
Logistical regression multivariate analysis was performed to identify characteristics (both categoric and continuous) associated with CR and OR (Table 5). The following pretreatment variables were analyzed in the model: age, sex, Rai stage, refractoriness to fludarabine, number of prior treatments, WBC count, absolute lymphocyte count, hemoglobin, platelet count, creatinine, 2M, albumin, percentage of CD38-positive lymphocytes in bone marrow, and abnormal karyotype. Characteristics that were not significant were dropped from the model in a stepwise fashion; the final model contained only statistically significant characteristics. In the final model, CR was associated with higher pretreatment platelet count, fewer number of prior treatments, and lower pretreatment 2M. In the final model, OR was associated with higher pretreatment hemoglobin, and treatment failure (nonresponders) was associated with fludarabine refractoriness and higher pretreatment creatinine. There was no correlation between advanced age and higher creatinine (data not shown).
Quality of Remission
Molecular evidence of residual disease was assessed by performing PCR to detect the IgVH gene of the malignant clone in the bone marrow of 37 patients achieving CR; in 12 patients (32%), the clonal population was not detected (Table 6). PCR was performed in 16 patients with NPR, and two patients (13%) were negative. Of the 22 patients with PRs who were tested, two patients (9%) were negative for the presence of the malignant population. Two-color immunostaining and flow cytometry analysis for CD5-positive and CD19-positive lymphocytes was also performed on bone marrow in some patients (Table 6).
Disease Progression and Survival
The current median follow-up time for all patients is 28 months (range, 1 to 50 months), and for surviving patients, it is 35 months (range, 14 to 50 months). The Kaplan-Meier estimates of time to progression for responding patients and overall survival for all patients are shown in Figure 1. The median time to progression was 28 months (range, 3 to 49 months). Moreover, the median time to progression for patients achieving CR, NPR, and PR was 39, 33, and 15 months, respectively (Fig 2). Time to progression by bone marrow PCR status at end of treatment is shown in Figure 3. There is not yet a statistical difference in time to progression by bone marrow PCR status in patients with available data; however, follow-up continues.
The multivariate Cox proportional hazards models was used to explore correlations between pretreatment characteristics and time to progression for the 129 responders and survival for the entire 177 patients (Table 7). The following variables were analyzed in the model: age, Rai stage, response to FCR, refractoriness to fludarabine, number of prior treatments, WBC count, absolute lymphocyte count, hemoglobin, platelet count, creatinine, 2M, albumin, percentage of CD38-positive cells in bone marrow, percentage of lymphocytes and CD5-positive and CD19-positive cells in bone marrow, and abnormal karyotype. In the Cox model for time to progression, there were three significant independent variables; these variables were 2M, response to FCR, and cytogenetics (Table 7). Time to treatment failure was significantly shorter for patients with increased 2M (hazard ratio, 1.1). The time to progression was significantly shorter for patients not achieving a remission versus responders (CR, NPR, or PR; hazard ratio, 7.8). Patients with diploid karyotype or 13q deletion as a sole abnormality had longer time to progression than patients with karyotype abnormalities.
The Kaplan-Meier estimate of overall survival is shown in Figure 1. The estimated median survival for the 177 patients treated on this study is 42 months (range, 1 to 50 months). Survival by response to FCR is shown in Figure 4. The estimated median survival times for all responders, complete responders, and nodular partial responders have not been reached but are estimated to be 45 and 30 months, respectively. The estimated median survival time for the partial responders is 39 months; for nonresponders, it is 11 months (Fig 4). In the Cox model for survival, there were three significant independent variables; these were 2M, response to FCR, and cytogenetics (Table 7). The hazard function for risk for death increased with higher levels of 2M, failure to achieve a CR, NPR, or PR, and unfavorable karyotype (Table 7).
Toxicity
All six intended courses of treatment were administered to 81 (46%) of 177 patients, and 68 of these patients received full-dose treatment. Sixty-six patients (37%) received one to three courses, 16 patients (9%) received four courses, and 14 patients (8%) received five courses of FCR. Myelosuppression was the most common reason for discontinuing treatment before completing six courses and occurred in 46 patients (26%). Sixteen patients (9%) stopped treatment because of disease resistance, and 11 patients (6%) stopped because of infection. Other reasons for early treatment cessation included death (n = 8), physician preference (n = 4), stem-cell transplantation (n = 2), and Richter’s transformation (n = 2). Dose reduction occurred in 25 patients, and four patients had two dose reductions. The primary reason for dose reduction was myelosuppression.
Rituximab-associated toxicities were associated with the first infusion. One hundred eleven patients (63%) had adverse events associated with the first infusion of rituximab, and all events were grade 1 or 2, with the exception of one episode of grade 3 shortness of breath, and all were self-limited. There were no grade 4 toxicities associated with any rituximab infusion. Of the grade 1 to 2 toxicities, fever occurred in 65 patients (37%), chills occurred in 89 patients (50%), hypotension occurred in 23 patients (13%), nausea and vomiting occurred in 22 patients (12%), and shortness of breath occurred in five patients (2%). Rash, fatigue, headache, and arthralgias were all grade 1 and occurred in less than 5% of patients.
The most common toxicity with FCR was neutropenia, with grade 3 and 4 neutropenia noted in 21% and 41% of 529 assessable courses. Fifteen percent and 66% of patients experienced one or more episode of grade 3 or 4 neutropenia, respectively, as their most severe grade of toxicity during treatment. Thrombocytopenia occurred in 17% of 539 assessable courses (grade 3 in 10% and grade 4 in 7%). Sixteen percent and 18% of patients experienced one or more episode of grade 3 or 4 thrombocytopenia, respectively, as their most severe grade of toxicity during treatment. Anemia (grade 3 or 4) was observed in 24% of patients during this study.
Major infections, defined as sepsis, pneumonia, or infection requiring hospitalization, occurred in 34 (5%) of 745 assessable courses or 16% of treated patients. Fever of unknown origin occurred in 78 (10%) of 745 assessable courses or 31% of treated patients, and minor infections occurred in 59 (8%) of 745 assessable courses or 18% or treated patients. Minor infections were defined as upper respiratory tract infection, bronchitis, cellulitis, and so on. One patient was removed from study because of cytomegalovirus pneumonitis. Herpes simplex and zoster were associated with four (1%) and six (1%) of 745 assessable courses, respectively. The frequency of major infections, herpes reactivation, and minor infections was similar for fludarabine-refractory patients compared with fludarabine-sensitive patients.
We defined late neutropenia as grade 3 or 4 neutropenia that occurred more than 30 days after the last course of FCR in patients who had recovered their absolute neutrophil count to greater than 1,000/μL. Late neutropenia occurred in five of 45 CR patients. It was transient in all cases, and bone marrow biopsy at the time of neutropenia showed no disease in any of these patients. The grade 4 events were isolated events; one occurred on day 82 and the other occurred on day 321 after the last course of FCR. One of the grade 3 events lasted 3 weeks; the others were isolated events. There were five other late neutropenic events in CR patients; however, at the time of the event, all of these patients had lymphoid aggregates or overt disease relapse in bone marrow.
To date, one patient developed acute myelogenous leukemia (French-American-British classification M5) 2 months after his fourth course of FCR. This patient had previously received single-agent fludarabine, and there were no cytogenetic abnormalities associated with his acute myelogenous leukemia. Four patients have developed myelodysplastic syndrome (1.5 years, 1 year, 9 months, and 3 months after five or six courses of FCR). Three of these patients had cytogenetic abnormalities (–5,–7 in two patients and –7 in one patient). Two of these patients were previously treated with single-agent purine analog, and two were previously treated with purine analog combined with alkylating agent.
DISCUSSION
The ultimate goal for treatment of CLL is cure; however, this has not yet been demonstrated with standard-dose chemotherapy. Historically, responders to treatment live longer and fare better, particularly complete responders. Recently, more sensitive assays, such as flow cytometry and PCR, have been developed to detect residual disease. The results of these assays have not yet been incorporated into formal response criteria. However, early data suggest that negative results with these more sensitive tests may predict longer remission duration and improved survival. The objective in developing new treatments for CLL is to achieve the highest CR rate and the highest quality CR as assessed by sensitive tests such as molecular and flow cytometry assay.
Typically, the CR and OR rates achieved with fludarabine as a salvage treatment were 5% to 10% and 40% to 60%, respectively. The median time to progression was 12 to 13 months, and median survival was 20 to 25 months.17-19 The combination of FC produced CR and OR rates of 10% to 15% and 60% to 70%, respectively, with a median response duration of 20 to 25 months in previously treated patients with CLL.2,3 This report of 177 previously treated patients with CLL receiving FCR is the largest series of such patients treated with a single regimen reported in the literature. The patient characteristics were similar to the characteristics of patients treated on historic single-agent and combination trials. Both CR and OR rates were higher than with single-agent fludarabine or FC. Furthermore, the median time to progression for responding patients was longer (28 months). More importantly, the estimated median overall survival for all patients (42 months) was longer than in historical trials. There are potential limitations regarding the comparability to previously reported single-center phase II studies. These limitations relate, in part, to patient characteristics, treatment, and post-trial therapy. We performed extensive evaluation to identify pretreatment characteristics correlated with response and survival and have reported these. This analysis is intended to provide information that may be applied to patients seen in practice.
Patients who were previously treated with FC were included in this trial. Thirty-four such patients were treated; 30 were fludarabine sensitive, and four were fludarabine refractory. The CR and OR rates for these patients were similar to those for the entire group and to those for patients previously treated with only an alkylating agent (Table 2). This suggests that prior treatment with FC does not have a detrimental impact on response to FCR. Patients who are refractory to fludarabine typically have low response rates with fludarabine-based regimens. Consistent with this, the CR and OR rates in the fludarabine-refractory patients were 6% and 58%, respectively. Treatment of fludarabine-refractory patients with the monoclonal antibody alemtuzumab produced CR and OR rates of 2% and 33%, respectively.20,21 The results with the FCR regimen seem comparable.
The NCIWG criteria for evaluating remission incorporate blood count with differential, morphologic evaluation of the bone marrow, and physical examination for adenopathy and organomegaly, without radiographic evaluation. Radiographic examination by computed tomography of chest and/or abdomen was not routinely performed and not required in this study to confirm response. Although responses tend to be uniform across nodal sites, it is conceivable but unlikely that there was residual intrathoracic or intra-abdominal lymphadenopathy or splenomegaly in a fraction of the responders. Flow cytometry and molecular techniques to detect malignant clonal cells are now being applied to patients as measures of the quality of remission. The importance of molecular remission was recently reported in a clinical trial of fludarabine combined with cyclophosphamide and mitoxantrone in previously treated patients with CLL.22 With fludarabine, cyclophosphamide, and mitoxantrone, 50% of 60 treated patients achieved CR, and 10 of 25 complete responders who were tested were negative for residual disease by PCR in bone marrow. This correlated with longer disease-free and overall survival. With FCR, 32% of 37 complete responders were negative by PCR for residual disease in bone marrow, and this was correlated with longer progression-free survival. Further follow-up is needed to fully assess a potential survival benefit in these patients.
Identifying prognostic features for response, duration of response, and survival is important because it aids in counseling patients before treatment and in stratifying patients in clinical trials. In addition, insight into the biology of the disease and the mechanism of action of therapeutic agents may be gained. Age was correlated with the CR rate but not the OR rate. Also, advanced age correlated with fewer number of courses of FCR received. Patients of advanced age may tolerate less chemotherapy, possibly because of attrition of normal bone marrow elements. Alternatively, older patients may have disease that is intrinsically resistant to treatment. Both hemoglobin and platelet count correlated with CR and OR rates; lower hemoglobin and platelet count predicted for lower CR and OR rates. Lower hemoglobin and platelet count may reflect more extensive bone marrow infiltration and/or resistance to treatment in cells residing in bone marrow. The use of growth factors during treatment may benefit such patients and merits evaluation.
The log-rank test demonstrated that lower 2M, higher albumin, early Rai stage, and response (CR and OR) to FCR correlated with longer survival. For both longer time to progression and overall survival, multivariate Cox model analysis demonstrated a correlation with lower serum 2M, diploid cytogenetics or 13q deletion as a sole chromosomal abnormality, and response (CR, NPR, or PR) to FCR. Lower 2M was also associated with likelihood for CR and OR, and 4.0 mg/L (normal range, 0.6 to 2.0 mg/dL) was a significant cut point. The biologic basis of this correlation is unclear, but the association has repeatedly been demonstrated in CLL as well as other B-cell malignancies such as multiple myeloma and lymphoma. The importance of 2M as a prognostic factor for response, time to progression, and survival was intriguing. The source of the soluble 2M is most likely the leukemia cells; however, this has not been definitively demonstrated. 2M is expressed on the surface of nucleated cells as a heterodimer with HLA class I. It is a well-established prognostic factor in CLL and other B-cell neoplasms. 2M is cleared from circulation by the kidneys and may be elevated in patients with renal insufficiency. Patients enrolled onto this study were required to have a creatinine of 2.0 mg/dL. Therefore, creatinine was not likely the major factor influencing 2M levels. Consistent with other studies, abnormal cytogenetics predicted for failure to achieve CR and OR and were associated with shorter time to progression and shorter survival.
The toxicities for FCR were typical and expected for such a group of previously treated patients.2,3 Infusion-related toxicities with rituximab were similar in frequency and severity to the toxicities seen in other studies. FC has been evaluated in previously treated patients with CLL.2,3 Myelosuppression was also the most frequent toxicity in these studies. In one previous study of 101 previously treated patients, one or more cycle of treatment was associated with grade 3 and 4 neutropenia in 75% and 48% of patients, respectively. Of note, these patients received higher doses of fludarabine (30 mg/m2 daily for 3 days) and cyclophosphamide (300 mg/m2 daily for 3 days).2 In this previous trial, infections were more frequent, with sepsis or pneumonia in 25% of patients and fever of unknown origin in 25% of patients. This may be related to differences in use of prophylactic antibiotics. The incidence of grade 3 or 4 thrombocytopenia in patients treated with FC was 19%, which is similar to the incidence noted in this study with FCR. Thus, the incidence of neutropenia, thrombocytopenia, and infection was similar for patients who received FCR compared with previously treated patients in a previous trial who received FC, indicating that the addition of rituximab did not significantly increase toxicity. Late neutropenia was transient and rare in the absence of relapsed or residual disease in the bone marrow.
FCR is a well-tolerated regimen with significant activity, even for patients who have received a substantial amount of prior therapy. Work continues in developing new combination regimens with complementary agents, with the overall objective to improve CR rate and quality of CR.
Authors’ Disclosures of Potential Conflicts of Interest
The following authors or their immediate family members have indicated a financial interest. No conflict exists for drugs or devices used in a study if they are not being evaluated as part of the investigation. Performed contract work within the last 2 years: William Wierda, Berlex Laboratories Inc, Xycte Therapies Inc, Ligand Pharmaceuticals.
NOTES
Authors’ disclosures of potential conflicts of interest are found at the end of this article.
REFERENCES
Yamauchi T, Nowak BJ, Keating MJ, et al: DNA repair initiated in chronic lymphocytic leukemia lymphocytes by 4-hydroperoxycyclophosphamide is inhibited by fludarabine and clofarabine. Clin Cancer Res 7:3580-3589, 2001
O’Brien SM, Kantarjian HM, Cortes J, et al: Results of the fludarabine and cyclophosphamide combination regimen in chronic lymphocytic leukemia. J Clin Oncol 19:1414-1420, 2001
Hallek M, Schmitt B, Wilhelm M, et al: Fludarabine plus cyclophosphamide is an efficient treatment for advanced chronic lymphocytic leukaemia (CLL): Results of a phase II study of the German CLL Study Group. Br J Haematol 114:342-348, 2001
Di Gaetano N, Xiao Y, Erba E, et al: Synergism between fludarabine and rituximab revealed in a follicular lymphoma cell line resistant to the cytotoxic activity of either drug alone. Br J Haematol 114:800-809, 2001
Alas S, Emmanouilides C, Bonavida B: Inhibition of interleukin 10 by rituximab results in down-regulation of bcl-2 and sensitization of B-cell non-Hodgkin’s lymphoma to apoptosis. Clin Cancer Res 7:709-723, 2001
Cheson BD, Bennett JM, Grever M, et al: National Cancer Institute-sponsored Working Group guidelines for chronic lymphocytic leukemia: Revised guidelines for diagnosis and treatment. Blood 87:4990-4997, 1996
Cheson BD, Bennett JM, Rai KR, et al: Guidelines for clinical protocols for chronic lymphocytic leukemia: Recommendations of the National Cancer Institute-sponsored Working Group. Am J Hematol 29:152-163, 1988
Slavickova A, Ullmannova V, Klener P: Optimized multiplex IgH/ras PCR: A tool for quantitative monitoring of B-lymphoproliferative disorders. Biotechniques 28:716-718, 720-721, 2000
Albitar M, Wu WI, Feltz E, et al: Simplified reverse dot blot analyses for detecting of ras oncogene mutations. Mol Diagn 2:169-176, 1997
Schultze JL, Donovan JW, Gribben JG: Minimal residual disease detection after myeloablative chemotherapy in chronic lymphatic leukemia. J Mol Med 77:259-265, 1999
Bryant J, Day R: Incorporating toxicity considerations into the design of two-stage phase II clinical trials. Biometrics 51:1372-1383, 1995
Mantel N: Evaluation of survival data and two new rank order statistics arising in its consideration. Cancer Chemother Rep 50:163-170, 1966
Kaplan EL, Meier P: Non-parametric estimation from incomplete observation. J Am Stat Assoc 53:457-481, 1958
Peto R, Pike MC, Armitage P, et al: Design and analysis of randomized clinical trials requiring prolonged observation of each patient: II. Analysis and examples. Br J Cancer 35:1-39, 1977
Cox DR, Snell EJ: Analysis of Binary Data (ed 2). London, United Kingdom, Chapman and Hall, 1989
Rai KR: A critical analysis of staging in CLL, in Gale RP, Rai KR (eds): Chronic Lymphocytic Leukemia: Recent Progress and Future Directions. New York, NY, Alan R. Liss, 1987, pp 253
Keating MJ, Kantarjian H, Talpaz M, et al: Fludarabine: A new agent with major activity against chronic lymphocytic leukemia. Blood 74:19-25, 1989
O’Brien S, Kantarjian H, Beran M, et al: Results of fludarabine and prednisone therapy in 264 patients with chronic lymphocytic leukemia with multivariate analysis-derived prognostic model for response to treatment. Blood 82:1695-1700, 1993
Fenchel K, Bergmann L, Wijermans P, et al: Clinical experience with fludarabine and its immunosuppressive effects in pretreated chronic lymphocytic leukemias and low-grade lymphomas. Leuk Lymphoma 18:485-492, 1995
Keating MJ, Flinn I, Jain V, et al: Therapeutic role of alemtuzumab (Campath-1H) in patients who have failed fludarabine: Results of a large international study. Blood 99:3554-3561, 2002
Rai KR, Freter CE, Mercier RJ, et al: Alemtuzumab in previously treated chronic lymphocytic leukemia patients who also had received fludarabine. J Clin Oncol 20:3891-3897, 2002
Bosch F, Ferrer A, Lopez-Guillermo A, et al: Fludarabine, cyclophosphamide and mitoxantrone in the treatment of resistant or relapsed chronic lymphocytic leukaemia. Br J Haematol 119:976-984, 2002(William Wierda, Susan O’B)