Leukemia (2001) 15, 1510–1516  2001 Nature Publishing Group All rights reserved 0887-6924/01 $15.00 www.nature.com/leu combined with and mitoxantrone as front-line therapy in chronic lymphocytic leukemia T Robak1, JZ Błon´ski1, M Kasznicki1,JGo´ra-Tybor1, J Dwilewicz-Trojaczek2, P Boguradzki2, L Konopka3, B Ceglarek3, J Sułek4, K Kuliczkowski5, D.Wołowiec5, B Stella-Hołowiecka6, AB Skotnicki7, W Nowak7, B Moskwa-Sroka8, A Dmoszyn´ska9 and M Calbecka10

1Department of Hematology, Medical University , Ło´dz´; 2Department of Hematology, Medical Academy, Warsaw; 3Institute of Hematology and Blood Transfusion, Warsaw; 4Department of Internal Medicine, Military Medical Academy, Warsaw; 5Department of Hematology, Medical University, Wrocław; 6Department of Internal Medicine, Medical Academy, Katowice; 7Department of Hematology, Jagiellonian University, Krako´w; 8Department of Hematology, City Hospital, Gorzo´w Wielkopolski; 9Department of Hematology, University Medical School, Lublin; and 10Department of Hematology, District Hospital, Torun´, Poland

The objective of the study was to determine the effectiveness nucleoside analogwith a substituted halogenatom at position and the toxicity of a combined consisting of cla- 2 in its purine ring, which makes it resistant to deamination dribine (2-CdA), mitoxantrone and cyclophosphamide (CMC regimen) in the treatment of previously untreated B cell chronic by . It is mainly used in the treatment lymphocytic leukemia (B-CLL). From August 1998 to December of hairy cell leukemia, but it is also highly active in chronic 2000 2-CdA was administered at a dosage of 0.12 mg/kg for 3 lymphocytic leukemia (CLL), with an overall response rate (CMC3) or 5 (CMC5) consecutive days, mitoxantrone at 10 (OR) and complete remission rate (CR) comparable to that 2 2 mg/m on day 1 and cyclophosphamide at 650 mg/m on day 1 achieved with FAMP.4–7 There are some opinions that 2-CdA to 62 patients with advanced or progressive B-CLL. The cycles were repeated at 4 week intervals or longer if severe myelo- is associated with more frequent and problematic myelosuppr- suppression occurred. Twenty patients received CMC5 and 42 ession. However, direct comparison of both agents in the ran- patients CMC3. Within the analyzed group an overall response domized study of CLL patients has not been published as of (OR) rate (CR+PR) of 64.5% (95% CI: 52.7–76.3%) was reported, this time. including 29.0% CR. There was no difference in the CR rate Despite the fact that new purine analogs induce a higher between the patients treated with CMC5 (30%) and CMC3 (28.6%) (P = 0.9), nor in the OR rate (55.0% and 69.0%, respect- overall response rate and a higher CR rate in patients with 5,8 ively, P = 0.3). Residual disease was identified in seven out of CLL, they do not influence survival time. On the other hand, 18 (38.9%) patients who were in CR, including two treated with both FAMP and 2-CdA administered as single agents can CMC5 and five treated with CMC3 protocols. CMC-induced induce immunophenotypic and molecular remission, which is grade III or IV thrombocytopenia occurred in 12 (19.4%) of especially important in younger patients in order to qualify patients, including four (20%) CMC5-treated and eight (19%) CMC3-treated patients (P = 0.8). Neutropenia grade III or IV was these patients for bone marrow transplantation and offer the 3,9 observed in seven (35%) and 11 (26.2%) patients, respectively possibility of recovery. (P = 0.8). Severe infections, including pneumonia and sepsis, Combined use of purine analogs with other cytotoxic agents occurred more frequently after CMC5 (11 patients, 55.0%) than may increase the CR rate and, possibly, suppress minimal = CMC3 (10 patients, 28.6%) (P 0.03) Fourteen patients died, residual disease and prolongsurvival. Some preclinical studies including six treated with CMC5 and eight treated with CMC3 10–16 (30% and 19%, respectively). Infections were the cause of death and early clinical reports may support such a hypothesis. in nine patients, including four in the CMC5 group and five in Among cytotoxic agents, alkylating drugs and the CMC3 group. In conclusion, our results indicate that the were the primary candidates for use in combination with CMC programme is an active combined regimen in previously purine analogs. Synergistic action of 2-CdA with cyclophos- untreated B-CLL patients; its efficiency seems to be similar to phamide and its derivatives has been shown both in in vitro that observed earlier in B-CLL patients treated with 2-CdA as 10,11 a single agent. However, toxicity, especially after CMC5 admin- and in vivo experiments. These results have been con- istration, is significant. Therefore, we recommend the CMC3 firmed recently by early clinical trials, which have shown that but not the CMC5 programme for further evaluation. Leukemia the combination of 2-CdA with cyclophosphamide and (2001) 15, 1510–1516. is feasible and active in patients with CLL and Keywords: cladribine; cyclophosphamide; mitoxantrone; CLL; other low-grade lymphoid malignancies.13,14 combined therapy; residual disease Mitoxantrone is also a useful drugin the treatment of low- grade lymphomas and can be combined with purine anal- ogs.17–20 However, the advantage of such a combination in Introduction patients with refractory or relapsingdisease over 2-CdA alone has not yet been proven. Thus, the combination of 2-CdA with The newer purine analogs, fludarabine (FAMP), 2Ј-deoxyco- cyclophosphamide and mitoxantrone seems to be the logical formycin (DCF) and cladribine (2-chlorodeoxyadenosine, 2- conclusion of the clinical studies in which 2-CdA was admin- CdA), have been synthesized recently and introduced into the 21 treatment of low-grade lymphoid malignancies.1–3 2-CdA is a istered with each independently. In our previous study, we showed that combined treatment with 2-CdA, cyclophos- phamide and mitoxantrone (CMC programme) in heavily pre- treated CLL and other indolent lymphoma patients is an active Correspondence: T Robak, Department of Hematology, Medical Uni- regimen. In this report, we present our experience with the versity of Ło´dz´, 93-513 Ło´dz´, ul. Pabianicka 62, Poland; Fax: +48 42 6846890 combination of 2-CdA with cyclophosphamide and mitoxan- The preliminary results of this study were presented at the 41st Annual trone in previously untreated patients with B-CLL. Meetingof the American Society of Hematology,San Francisco, CA, December 1–5, 2000. Received 27 February 2001; accepted 10 May 2001 CMC in untreated CLL T Robak et al 1511 Patients and methods In the first part of the study, CMC protocol consisted of 2- CdA administered at a dose 0.12 mg/kg in 2-h i.v. infusion for Patients 5 (CMC5) days, mitoxantrone 10 mg/m2 i.v. on day 1 and cyclophosphamide 650 mg/m2 i.v. on day 1. In the second Between August 1998 and December 2000, 62 previously part of the study, 2-CdA administration was shortened to 3 untreated patients with progressive or symptomatic CLL were days (CMC3) because of excessive hematological toxicity of treated with a combined chemotherapy consistingof 2-CdA, CMC5 protocol. The doses of mitoxantrone and cyclophos- cyclophosphamide and mitoxantrone. The characteristics of phamide were the same as in the CMC5 protocol. 2-CdA the patients are shown in Table 1. All of the patients fulfilled (Biodrybin) was synthesized accordingto the method of Kazi- the National Cancer Institute-Sponsored WorkingGroup cri- mierczuk et al24 and was commercially available from the teria for CLL.22 Pretreatment evaluation included examination Institute of Biotechnology and Antibiotics- Bioton (Warsaw, of medical history, physical examination, complete blood cell Poland). count, differential count of WBC, chemical survey, bone mar- The cycles were repeated over 28 days for three courses. If row examination and serum immunoglobulin level quantit- hematological complications (thrombocytopenia Ͻ50 × 109/l, ation. Cytogenetic analysis was not routinely performed. The granulocytopenia Ͻ0.5 × 109/l) or severe infections developed patients had peripheral lymphocytosis greater than 10 × 109/l the drugs were re-administered at time intervals longer than and more than 30% lymphocytes in normal or hypercellular 28 days, ranging from 2 to 4 months, until the increase of bone marrow. The differences between mean values in both hematological parameters or recovery from infections was groups of patients were not significant. Cell marker studies noted. If a response had been documented, patients were were performed to confirm B cell origin and monoclonal pro- treated until maximal response or prohibitive toxicity. If no liferation. All patients were CD5, CD19, CD20 and CD23 response or progression of the disease was observed after positive and showed monoclonality for light chain immuno- three cycles, the treatment was discontinued. Packed red cells globulin membrane surface receptors. were transfused for symptomatic anemia or prophylactically The clinical stage of disease was determined at the time of if the hemoglobin level was lower than 7.0 g/dl. Platelets were initiation of the treatment accordingto Rai’s classification. 23 administered prophylactically if the platelet count was less The distribution is shown in Table 1. Patients in stage 0, I and than 15.0 × 109/l. Blood products were irradiated. In order to II were eligible if they had evidence of active disease, includ- prevent hyperuricemia, allopurinol (300 mg/daily) was admin- ing progressive lymphocytosis, massive splenomegaly or istered. No patients received antibiotics or antiviral agents bulky lymphadenopathy, recurrent disease-related infections, prophylactically. No patients received hematopoietic growth weight loss Ͼ10% over a 6 month period, temperature of factors prophylactically but G-CSF or GM-CSF were adminis- 38°C related to disease or extreme . Patients with poor tered if the absolute granulocyte count was less than 1.0 × performance status (WHO scale 4), active infection, abnormal 109/l and active infection was present. liver or renal function and Richter’s syndrome were excluded from the study. The study was approved by Local Ethical Com- mittees and the patients had signed the informed consent Response criteria form. Treatment effect was monitored by physical examination, blood count evaluation and bone marrow examination before Treatment modality and after every course of chemotherapy. Guidelines for response were those developed by the NCI-Sponsored Work- The doses and schedule of the treatment agents were based ingGroup. 22,25 Complete response (CR) required the absence on previous studies in cases of CLL relapse, refractory CLL or of symptoms and organomegaly, normal complete blood cell low-grade lymphoma.7,19–21 count (absolute neutrophil count Ͼ1.0 × 109/l, hemoglobin

Table 1 Clinical characteristics of B CLL patients before CMC treatment

Characteristics All patients CMC5 CMC3

Total 62 20 42 Sex Male 38 (61.0) 13 (65) 25 (59.5) Female 24 (39.0) 7 (35) 17 (40.5) Age in years, median ± s.d. (range) 62.5 (33–77) 63.5 (36–77) 61 (33–76) Rai stage before CMC 0 1 (1.6) 1 (5) 0 I 8 (12.9) 3 (15) 5 (11.9) II 18 (29.0) 6 (30) 12 (28.6) III 20 (32.3) 4 (20) 16 (38.1) IV 15 (24.2) 6 (30) 9 (21.4) Median disease duration in months median (range) 8.9 (3.0–156.3) 14.6 (4–26) 8.2 (3.0–156.3) Mean number of WBC × 109/l (range) 84.0 (5.4–376.0) 103.2 (10.2–606.0) 82.6 (5.4–376.0) Mean Hb concentration g/dl (range) 12.0 (4.0–17.0) 12.0 (4.0–14.0) 11.0 (5.1–17.0) Mean number of platelets × 109/l (range) 150 (28–314) 133 (15–231) 174 (28–314)

Leukemia CMC in untreated CLL T Robak et al 1512 concentration Ͼ11.0 g/dl, platelet count Ͼ100 × 109/l and Results bone marrow with less than 30% lymphocytes for at least 2 months. Partial response (PR) was considered in the case of a Sixty-two patients with B-CLL entered the study and all of 50% or greater decrease in the size of lymph nodes, liver and them were analyzed. The median time from the diagnosis to spleen, and peripheral blood findings either identical to those CMC administration was 8.2 months (range 4–26). The of CR or improved over pre-therapy values by at least 50%. characteristics of the patients are presented in Table 1. Twenty The patients who had not achieved CR or PR were classified patients were treated accordingto CMC5 protocol and 42 as non-responders (NR). Clinical relapse was defined accord- received CMC3 chemotherapy. The total number of CMC ingto Robertson et al9 as an increase in the absolute lympho- courses administered to the entire group of patients amounted cyte count above 10 × 109/l, more than 50% increase in the to 237 (84 CMC5 and 153 CMC3 courses). The median num- sum of the size of at least two lymph nodes, appearance of ber of CMC3 courses was 3.8 (range 1–6), and the median new lymph nodes, more than a 50% increase in the liver or time needed to deliver the total number of cycles was 4.6 spleen below the costal margin, the new appearance of pal- months (range 1–8.7). The patients treated with CMC5 pable hepatosplenomegaly or development of an aggressive received from 1 to 6 (median 2.5) courses, and the median lymphoma. time needed to deliver the total number of cycles was 5.7 months (range 1–10.2). In the CMC3 group 8 (19%) patients, and in the CMC5 group 9 (45%) patients, could not reach the Immunophenotype analysis total of three cycles because of early death or severe myelo- toxicity. The criteria for CR were fulfilled in 18 patients (29.0%) and Immunophenotypingwas performed on peripheral blood and PR in 22 (35.5%), giving an overall response rate in 64.5% bone marrow by flow cytometry usinga simultaneous dual- (95% CI 52.7–76.3%) of total patients (Table 2). There were color stainingtechnique before treatment and after obtaining no differences in the frequency of OR between the CMC5- 26 CR, accordingto the method described by Brugiatelli et al. A and the CMC3-treated groups (55% and 69%, respectively, P combination of phycoerythrin (PE)-conjugated and fluorescent = 0.3). The CR rate was also similar in both groups (30% and isothiocyanate (FITC)-conjugated monoclonal antibodies was 28.6%, respectively) (P = 0.9). In the group treated with CMC5 utilized. Residual disease was determined by co-expression of the patients attainingCR received a median of three courses CD5/CD19 and CD5/CD20 on B lymphocytes in conjunction (range 2–6) and in the group treated with CMC3 a median of with monoclonality of surface light-chain expression on CD5- three courses (range 2–6) was also administered. CRs were positive B cells. The presence of more than 10% of the total observed more frequently in the early stages of the disease lymphocytic population co-expressingCD5/CD19 and (Rai 0, I and II) – 11/29 (37.9%) than in the more advanced CD5/CD20 with monotypic light-chain expression (a ␭:␬ or = ␬ ␭ stages (Rai III and IV) – 7/33 (21.2%) (P 0.8). : ratio exceeding3:1) was considered positive for residual Surface immunophenotypingby flow cytometry usingdual- disease, accordingto the criteria developed by Robertson et color stainingon peripheral blood and/or bone marrow was 9 al. performed in patients who achieved CR. Residual disease was identified in seven out of 18 (38.9%) patients who achieved remission. Toxicity monitoring

Hematological toxicity was evaluated according to the criteria Toxicity developed by the NCI-Sponsored WorkingGroup. 22 Drug- induced anemia, thrombocytopenia and neutropenia were Myelosuppression was the major toxicity of the CMC therapy diagnosed if after the treatment course a further decrease of (Table 3). Grades III and IV neutropenia were observed in 18 hemoglobin level and/or platelets and neutrophils numbers (29.0%) out of 62 patients and after 30 (12.7%) out of 237 were observed. Other side-effects were assessed and moni- CMC courses. There was no significant difference between the tored accordingto WHO criteria. 27 frequency of severe neutropenia in the 7/20 (35%) patients Only major and moderate infections were recorded, includ- treated with CMC5 or 11/42 (26.2%) patients treated with = inglife-threateningepisodes such as pneumonia and dissemi- CMC3 regimens (P 0.8). Hematopoietic growth factors (G- nated infections that required oral or parenteral antibiotic CSF or GM-CSF) were administered after four and 13 therapy, antifungal and antiviral therapy and/or hospitaliz- courses, respectively. ation. Fever of unknown origin (FUO) requiring parenteral The frequency of thrombocytopenia was similar (20% and = antibiotic therapy was also recorded as an infectious event. 19%) in both groups (P 0.8). Grade III or IV anemia was the Infections were reported as CMC related if they developed on rarest hematological complication and was observed in only therapy or within 4 months of the completion of the CMC four patients. Autoimmune hemolytic anemia (AIHA) treatment. developed in two patients treated with CMC3. These patients had no clinical or laboratory symptoms of this complication before CMC treatment. Severe infections and fever of unknown origin (FUO) requir- Statistical analysis ingparenteral antibiotic therapy occurred more often in patients treated with the CMC5 regimen – 11 patients (55%) Statistical analysis of the differences in the percentages of than CMC3 – 10 patients (23.8%) (P = 0.03). The number of responses in patients treated with CMC3 and CMC5 were severe infections and FUO was also more frequent after CMC5 evaluated by the ␹2 test. Ninety-five percent confidence inter- courses – 15 (17.9%) than CMC3 courses – 12 (7.8%) (P = vals for response probability were calculated usingthe 0.03). Pneumonia and upper respiratory infections (sinusitis method described by Duffy and Santner.28 and bronchitis) occurred in 23 patients, including12 treated

Leukemia CMC in untreated CLL T Robak et al 1513 Table 2 Results of the treatment of B CLL patients with CMC accordingto the treatment protocol

Rai All patients (%) CMC5 (%) CMC3 (%) stage n = 62 n = 20 n = 42

n CR PR NR n CR PR NR n CR PR NR

0 110011000000 (100.0) (100.0) I + II 28 10 11 7 9 3 2 4 19 7 9 3 (35.7) (39.3) (26.0) (36.7) (22.2) (44.4) (36.8) (47.4) (15.8) III + IV 33 7 11 15 10 2 3 5 23 5 8 10 (21.2) (33.3) (45.5) (20.0) (30.0) (50.0) (21.7) (34.8) (43.5) P valuea 0.8 0.4 0.4 Total 62 18 22 22 20 6 5 9 42 12 17 13 (29.0) (35.5) (35.5) (30.0) (25.0) (45.0) (28.6) (40.4) (31.0) aP values for responses in different Rai stages.

Table 3 Severe hematological toxicity of the CMC programme

Protocol Neutropenia Thrombocytopenia Anemia grade III and IV grade III and IV grade III and IV

n (%) n1 (%) n (%) n1 (%) n (%) n1 (%)

CMC5 7 (35.0) 9 (10.7) 4 (20.0) 6 (7.1) 2 (10.0) 4 (4.8) n = 20 n1 = 84 CMC3 11 (26.2) 21 (13.7) 8 (19.0) 14 (9.2) 2 (4.8) 2 (1.3) n = 42 n1 = 153 P value for 0.8 0.9 0.7 0.8 0.8 0.2 CMC5 vs CMC3 Total 18 (29.0) 30 (10.3) 12 (19.4) 20 (8.4) 4 (6.5) 6 (2.5) n = 62 n1 = 237 n, number of patients; n1, number of courses. with CMC5 and 11 treated with CMC3. Herpes zoster reacti- CMC3. All six patients in the CMC5 group died before com- vation and herpes simplex infections were observed in 10 pletingthree courses of treatment. In CMC3 groupsix out of patients (two after CMC5 and eight after CMC3). Tuberculosis eight patients died before completing three courses of treat- was diagnosed in one patient after two courses of CMC5. ment. The causes of death are shown in Table 5. Compli- Other opportunistic infections were not seen. cations due to infections were the cause of death in nine In four patients, grade III according to the WHO patients, includingfour treated with CMC5 and five after classification was observed and in one DIC syndrome CMC3 therapy. developed. Other non-hematological side-effects including alopecia, abnormal aminotransferases, transient increased LDH and creatinine levels were observed only very rarely Discussion (Table 4). Secondary cancers were observed in two patients includingone case of pulmonary cancer (adenocarcinoma) In spite of intensive research and important progress in the after CMC3 in a patient without the history of tobacco last years, CLL still remains incurable. The results of the ran- exposure and one case of malignant histiocytosis after CMC5. domized studies published so far indicate that both FAMP and Richter syndrome has not been observed to date 2-CdA, applied as first-line therapy, significantly increase the The intervals between CMC courses due to myelosuppres- rate of complete remissions when compared to , sion and/or infections were prolonged from 4 to 8 weeks in which is still regarded as the gold standard in CLL treatment.5,8 23 (37%) patients and after 38 (16%) courses, including10 However, neither study proved that the higher remission rate patients (18 courses) treated with CMC5 and 13 patients (20 is translated into longer survival. That is why further research courses) treated with CMC3. In four patients (two treated with is justified, with the objective to increase the remission rate CMC5 and two with CMC3), therapy was stopped before the and quality, and in this way, to influence survival time. In the completion of three courses of treatment because of trials finished to date, alkylatingdrugsand anthracyclines with prolonged severe myelosuppression. mitoxantrone were most often combined with either FAMP or Altogether, 14 (22.6%) patients died during the study, with 2-CdA.10–18,29,30 includingsix (30%) treated with CMC5 and eight(19.0%) with Our study is the first, to our knowledge, in which previously

Leukemia CMC in untreated CLL T Robak et al 1514 Table 4 CMC-induced grade III and IV non-hematological side-effects

Side-effect Total CMC5 CMC3

n = 62 n1 = 237 n = 20 n1 = 84 n = 42 n1 = 153 n (%) n1 (%) n (%) n1 (%) n (%) n1 (%)

Infections and FUO 23 (37.1) 37 (15.6) 11 (55.0) 15 (17.9) 10 (23.8) 12 (7.8) Vomiting grade III or IV 4 (6.5) 7 (3.0) 2 (10.0) 4 (4.8) 2 (4.8) 3 (2.0) Alopecia 1 (1.6) 1 — 11 Aminotransferases increased 1 (1.6) 1 — 11 Creatinine increased 1 (1.6) 1 — 11 IHD 1 (1.6) 1 — 11 DIC 1 (1.6) 1 — 11 Secondary neoplasms 2 (3.2) 1 1 1 1 1

n, number of patients; n1, number of courses; FUO, fever of unknown origin; DIC, disseminated intravascular coagulation; IHD, ischemic heart disease.

Table 5 Cause of death in B CLL patients treated with 2-CdA patients with CLL or low-grade non-Hodgkin’s lymphoma as first-line treatment. Such combinations were, however, Cause of death Total CMC5 CMC3 administered to earlier pretreated patients.21,36,37 These were = = = n 14 n 6n8 the phase I/II studies that did not prove that the combination of purine analogwith cyclophosphamide and mitoxantrone Pneumonia 5 (35.7%) 2 (33.3%) 3 (37.5%) seemed more efficient that the monotherapy with FAMP or Septic shock 4 (28.6%) 2 (33.3%) 2 (25.0%) Hemorrhage 1 (7.1%) — 1 (12.5%) 2-CdA. Second neoplasm 2 (14.3%) 1 (16.7%) 1 (12.5%) An additional aim of our study was the evaluation of the Not connected with CLL 2 (14.3%) 1 (16.7%) 1 (12.5%) treatment-related toxicity of the CMC programme. We have shown that the most important toxic effect of this combination was myelosuppression. Grade III/IV neutropenia was observed in 29% of the patients and severe thrombocytopenia in 19.4% untreated CLL patients received two other drugs in combi- of the patients. It should be remembered, though, that myelo- nation with 2-CdA: cyclophosphamide and mitoxantrone toxicity is the limitingfactor for the administration of any drug (CMC programme). In this study, the daily doses of 2-CdA of the CMC programme when they are applied in monother- (0.12 mg/kg) were identical in both groups, but the duration apy or combined with other cytotoxic agents. Important myel- of the treatment was different: 3 days (CMC3) or 5 days otoxicity of the CMC programme may indicate the necessity (CMC5). The doses of cyclophosophamide and mitoxantrone of reducingthe dose of 2-CdA in this combination, and this were the same in both programmes. Our study indicates that supports the CMC3 rather than the CMC5 programme. The CMC combination is active in CLL. The CR rate was 29% and justification of such an approach has also been found in other the OR rate was 64.5%. It should be emphasised, that in spite studies, where the dose of purine analogin combination was of the large proportion of patients who have responded to lower than in monotherapy.16–18,29,38 treatment, the response rate does not seem greater than in our The infections were frequent in the patients treated with the earlier studies with 2-CdA in monotherapy or combined with prednisone.7,8 It is of interest that the shortened administration CMC programme and they were most commonly the cause of time from 5 (CMC5) to 3 (CMC3) days has not resulted in a death. These results are consistent with the observations from other studies, which also indicate that infections are frequent lower CR rate (30% and 28.6%, respectively) or overall 39,40 response rate (55% and 69%, respectively). Similarly, we in patients treated with purine analogs. It should be observed earlier the lack of influence of a shorter adminis- emphasized that the frequency of infections was significantly tration time of 2-CdA within a CMC programme on the out- lower in patients treated with CMC3 (23.8%) than in patients = 41 come in pretreated indolent lymphoma.21 Previous studies treated with CMC5 (55%) (P 0.02). Betticher et al found revealed that the response rate was not lower when 2-CdA or that the reduction in the dose of 2-CdA from 0.7 mg/kg per FAMP were combined with either cyclophosphamide,14,31–33 cycle to 0.5 mg/kg per cycle in pretreated patients with malig- mitoxantrone17,34 or .35 nant lymphomas did not decrease the activity of this Minimal residual disease (MRD) was evaluated by two- compound but greatly reduced the incidence of infections. color immunophenotypingin the patients who fulfilled mor- Further reduction of myelosuppression and infections may phological criteria for CR. MRD was identified in seven out be most likely achieved with the prophylactic use of growth of 18 patients (38.9%). A similar evaluation was performed in factors followingthe chemotherapy cycle. 42,43 In our study, the CLL patients treated with 2-CdA in monotherapy and MRD G-CSF was applied only as a treatment for severe infections was diagnosed in five out of 17 (29.4%) patients with CR.3 to the patients with neutropenia Ͻ1.0 × 109/l. O’Brien et al42 These data indicate that the combined therapy with CMC does showed, however, that prophylactic administration of G-CSF not appear to reduce the incidence of MRD more than 2-CdA to patients treated with FAMP may reduce the incidence of in monotherapy. pneumonia, but does not interfere with other infections. It To the best of our knowledge, our study is the first in which should also be remembered that prophylactic administration a purine analogwas combined with two other cytostatics in of hemopoietic growth factors may greatly increase the cost

Leukemia CMC in untreated CLL T Robak et al 1515 of treatment. As an alternative, some authors are in favor of prednisone as first-line therapy in chronic lymphocytic leukemia: prophylaxis with antibacterial and antiviral agents in patients report of a prospective, randomized, multicenter trial. Blood 2000; treated with purine analogs.44,45 This approach is intended to 96: 2723–2729. 9 Robertson LE, Huh YO, Butler JJ, Pugh WC, Hirsch-Ginsberg C, reduce the incidence of opportunistic infections. We have not, Stass S, Kantarjian H, KeatingMJ. Response assessment in chronic however, observed in our patients infections caused by Pneu- lymphocytic leukemia after fludarabine plus prednisone: clinical, mocystis carinii or other opportunistic infections. Such an pathologic, immunophenotypic and molecular analysis. Blood approach should also not be recommended because of the 1992; 80: 29–36. broad spectrum of potential infections that may necessitate the 10 Van Den Neste E, Bontemps F, Delacauw A, Cardoen S, Louviaux use of numerous potentially toxic drugs.46 I, Scheiff JM, Gillis E, Leveugle P, Deneys V, Ferrant A, Van den Berghe G. Potentiation of antitumor effects of cyclophosphamide In conclusion, combined treatment with 2-CdA, mitoxan- derivatives in B-chronic lymphocytic leukemia cells by 2-chloro- trone and cyclophosphamide (CMC programme) shows sig- 2Јdeoxyadenosine. Leukemia 1999; 13: 918–925. nificant therapeutic activity in previously untreated patients 11 Go´ra-Tybor J, Robak T. Synergistic action of 2-chlorodeoxyadeno- with CLL. However, the toxicity of the CMC programme is sine and cyclophosphamide on murine leukemia L1210 and P388. significant although it is restricted mainly to myelosuppression Acta Haematol Pol 1993; 24: 177–182. and infections. The reduction of total dose of 2-CdA per cycle 12 Hoffman M, Xu JC, Lesser M, Rai K. Cytotoxicity of 2-chlorodeoxy- in the CMC programme does not seem to decrease its antileu- adenosine (cladribine 2-CdA) in combination with other chemo- therapy drugs against two lymphoma cell lines. Leuk Lymphoma kemic activity but significantly reduces the risk of infections. 1999; 33: 141–145. At present, it is not clear whether the CMC programme is more 13 Van Den Neste E, Louviaux I, Michaux JL, Dellannoy A, Michaux efficient in comparison to 2-CdA in monotherapy or the com- L, Sonet A, Bosly A, Doyen C, Mineur P, Andre M, Straetmans N, bination of 2-CdA with cyclophosphamide in previously Coche E, Venet C, Duprez T, Ferrant A. Phase I/II study of 2- untreated CLL patients. Nevertheless, in our opinion, the chloro-2Ј-deoxyadenosine with cyclophosphamide in patients CMC3, but not the CMC5 programme, deserves further with pretreated B cell chronic lymphocytic leukemia and indolent non-Hodgkin’s lymphoma. Leukemia 2000; 14: 1136–1142. evaluation. 14 Laurencet FM, Zulian GB, Guetty-Alberto M, Iten PA, Betticher DC, Alberto P. Cladribine with cyclophosphamide and prednisone in the management of low-grade lymphoproliferative malig- Acknowledgements nancies. Br J Cancer 1999; 79: 1215–1219. 15 Kano Y, Akutsu M, Tsunoda S, Suzuki K, Ichikawa A, Furukawa This work was supported in part by grant No. 4P05B06019, Y, Bai L, Kon K. In vitro cytotoxicity effects of fludarabine (2-F- ara-A) in combination with commonly used antileukemic agents from KBN, Warsaw, Poland. We thank Mr Shane Gollop for by isohologram analysis. Leukemia 2000; 14: 379–388. the correction of the English version of the manuscript. 16 Robertson LE, O’Brien S, Kantarjian H, Koller C, Beran M, And- reeff M, Lerner S KeatingMJ. plus in pre- viously treated chronic lymphocytic leukemia. Leukemia 1995; 9: References 943–945. 17 McLaughlin P, Hagemeister FB, Romagnera JE, Sarris AH, Pate O, 1 Tallman MS, Hakimian D. Purine nucleoside analogs: emerging Younes A, Swan F, KeatingM, Cabanillas F. Fludarabine, mitoxan- roles in indolent lymphoproliferative disorders. Blood 1995; 86: trone and dexamethasone: an effective new regimen for indolent 2463–2474. lymphoma. J Clin Oncol 1996; 14: 1262–1268. 2 Juliusson G, Liliemark J. Long-term survival following cladribine 18 Saven A, Lee T, Kosty M, Piro L. Cladribine and mitoxantrone dose (2-chlorodeoxyadenosine) therapy in previously treated patients escalation in indolent non-Hodgkin’s lymphoma. J Clin Oncol with chronic lymphocytic leukemia. Ann Oncol 1996; 7: 373– 1996; 14: 2139–2144. 379. 19 Robak T, Go´ra-Tybor J, Urban˜ska-Rys´ H, Krykowski E. Combi- 3 Robak T, Błon´ski JZ, Urban´ska-Rys´ H, Błasin´ska-Morawiec M, nation regimen of 2-chlorodeoxyadenosine (cladribine), mitoxan- Skotnicki AB. 2-Chlorodeoxyadenosine (Cladribine) in the treat- trone and dexamethasone (CMD) in the treatment of refractory and ment of patients with chronic lymphocytic leukemia 55 years old recurrent low grade non-Hodgkin’s lymphoma. Leuk Lymphoma and younger. Leukemia 1999; 13: 518–523. 1999; 32: 359–363. 4 KeatingMJ, O’Brien S, Lerner C, Koller C, Beran M, Robertson LE, 20 Robak T, Go´ra-Tybor J, Chojnowski K. Cladribine as monotherapy Freireich EJ, Estey E, Kantarjian H. Long-term follow-up of patients or combined with dexamethasone and or mitoxantrone with chronic lymphocytic leukemia (CLL) receivingfludarabine in previously treated patients with low-grade lymphoid malig- regimens as initial therapy. Blood 1998; 92: 1165–1171. nancies. Haematologica 2000; 85: 215–216. 5 Rai KR, Peterson BL, Appelbaum FR, Kolitz J, Cheson BD, Schiffer 21 Robak T, Go´ra-Tybor J, Lech-Maran´da E, Błon´ski JZ, Kasznicki M. CA. Fludarabine compared with chlorambucil as primary therapy Cladribine in combination with mitoxantrone and cyclophos- for chronic lymphocytic leukemia. N Engl J Med 2000; 343: phamide (CMC) in the treatment of heavily pretreated patients 1750–1757. with advanced indolent lymphoid malignancies. Eur J Haematol 6 Delannoy A, Martiat P, Gala JL, Deneys V, Ferrant A, Bosly A, 2000; 66: 188–194. Scheiff JM Michaux JL. 2-chlorodeoxyadenosine (CdA) for patients 22 Cheson BD, Bennett JM, Grever M, Kay N, KeatingMJ, O’Brien with previously untreated chronic lymphocytic leukemia (CLL). S, Rai KR. National Cancer Institute-Sponsored WorkingGroup Leukemia 1995; 9: 1130–1135. guidelines for chronic lymphocytic leukemia: revised guidelines 7 Robak T, Błon´ski JZ, Kasznicki M, Konopka L, Ceglarek B, for diagnosis and treatment. Blood 1996; 87: 4990–4997. Dmoszyn˜ska A, Soroka-Wojtaszko M, Skotnicki AB, Nowak W, 23 Rai KR, Sawitsky A, Cronkite EP, Chanana AD, Levy RN Pas- Dwilewicz-Trojaczek J, Tomaszewska A, Hellmann A, Lewandow- ternack BS. Clinical staging of chronic lymphocytic leukemia. ski K, Kuliczkowski K, Potoczek S, Zdziarska B, Hansz J, Kroll R, Blood 1975; 46: 219–234. Komarnicki M, Hołowiecki J, Grieb P. Cladribine with or without 24 Kazimierczuk Z, Cottam HB, Ravanliar GR, Robins RK. Synthesis prednisone in the treatment of previously treated and untreated B- of 2Ј-deoxytubercidin, 2Ј-deoxyadenosin and related 2Ј-dexynu- cell chronic lymphocytic leukaemia – updated results of the multi- cleosides via a novel direct stereospecific serum salt glycosylation centre study of 378 patients. Br J Haematol 2000; 108: 357–368. procedure. J Am Chem Soc 1984; 106: 6379–6382. 8 Robak T, Błon´ski JZ, Kasznicki M, Błasin´ska-Morawiec M, Kry- 25 Cheson BD, Bennet MJ, Rai KR, Grever MR, Kay NE, Schiffer CA, kowski E, Dmoszyn´ska A, Mrugała-S´piewak H, Skotnicki AB, Oken MM, KeatingMJ, Boldt DH, Kempin SJ, Foon KA. Guidelines Nowak W, Konopka L, Ceglarek B, Maj S, Dwilewicz-Trojaczek for clinical protocols for chronic lymphocytic leukemia: rec- J, Hellmann A, Urasin´ski I, Zdziarska B, Kotlarek-Haus S, Potoczek ommendations of the National Cancer Institute-Sponsored Work- S, Grieb P. Cladribine with prednisone versus chlorambucil with ingGroup. Am J Hematol 1988; 29: 152–163.

Leukemia CMC in untreated CLL T Robak et al 1516 26 Brugiatelli M, Claisse JF, Lenormand B, Marabiro F, Callea V, Mal- 36 Bocchia M, Bigazzi C, Marconcini S, Forconi F, Marotta G, Algeri loum K, Chevret S, Binet JL, Dighiero G, Travade P. Long term R, Lauria F, Favorable impact of low-dose fludarabine plus epirub- clinical outcome of B-cell chronic lymphocytic leukaemia patients icin and cyclophosphamide regimen (FLEC) as treatment for low- in clinical remission phase evaluated at phenotypic level. BrJ grade non-Hodgkin’s lymphomas. Hematologica 1999; 84:716– Haematol 1997; 97: 113–118. 720. 27 Miller AB, Hoogstraten B, Staquet M, Winkler A. Reporting results 37 Perales M, Bosch F, Esteve J, Fludarabine, cyclophosphamide and of cancer treatment. Cancer 1981; 47: 207–214. mitoxantrone (FCM) in resistant or relapsed chronic lymphocytic 28 Duffy DR, Santner TJ. Confidence intervals for binomial parameter leukemia and . Ann Oncol 1999; 10 (Suppl. based on multistage tests. Biometrics 1987; 43: 81–93. 3): 128 (Abstr. 462). 29 Rummel MJ, Kafer G, Pfreundschuh M, Jager E, Reinhardt U, 38 Weiss M, Spiess T, Berman E, Kempin S. Concomitant adminis- Mitrou PS, Hoelzer D, Bergmann L. Fludarabine and tration of chlorambucil limits dose intensity of fludarabine in pre- in the treatment of chronic lymphocytic leukaemia: a German viously treated patients with chronic lymphocytic leukemia. Leu- multicenter phase II study. Ann Oncol 1999; 10: 183–188. kemia 1994; 8: 1290–1293. 30 Zinzani PL, Magagnoli M, Moretti L, De Renzo A, Battista R, Zac- 39 Cheson BD. Infectious and immunosuppressive complications of caria A, Guardigni L, Mazza P, Marra R, Ronconi F, Lauta VM, purine analogtherapy. J Clin Oncol 1995; 13: 2431–2448. Bendandi M, Gherlinzoni F, Gentilini P, Ciccone F, Cellini C, Ste- 40 Anaissie EJ, Kontoyiannis DP, O’Brien S, Kantarjian H, Robertson foni V, Ricciuti F, Gobbi M, Tura S. Randomized trial of fludarab- L, Lerner S, KeatingMG. Infections in patients with chronic lym- ine versus fludarabine and idarubicin as front line treatment in phocytic leukemia treated with fludarabine. Ann Intern Med 1998; patients with indolent or mantle-cell lymphoma. J Clin Oncol 129: 559–566. 2000; 18: 773–779. 41 Betticher DC, von Rohr A, Ratschiller D, Schmitz SF, Egger T, Son- deregger T, Herrmann R, Kroner T, Zulian GB, Cavalli F, Fey MF, 31 O’Brien S, Kantarjian H, Beran M, Freireich E, Kornblau S, Koller Cerny T. Fewer infections, but maintained antitumor activity with C, Lerner S, Gilbreath J, KeatingM. Fludarabine (FAMP) and cyclo- lower-dose versus standard-dose cladribine in pretreated low- phosphamide (CTX) therapy in chronic lymphocytic leukemia grade non-Hodgkin’s lymphoma. J Clin Oncol 1998; 16: 850–858. (CLL). Blood 1996; 88 (Suppl. 1): 480a (Abstr. 1910.) 42 O’Brien S, Kantarjian H, Beran M, Koller C, Talpaz M, Lerner S, 32 Flinn IW, Byrd JC, Morrison C, Janison J, Diehl LF, Murphy T, KeatingMJ. Fludarabine and granulocytecolony-stimulatingfactor Piantadosi S, Seifter C, Ambinder RF, Vogelsang G, Grever MR. (G-CSF) in patients with chronic lymphocytic leukemia. Leukemia Fludarabine and cyclophosphamide with filgrastim support in 1997; 11: 1631–1635. patients with previously untreated indolent lymphoid malig- 43 Flinn IW, Byrd JC, Morrison C, Jamison J, Diehl LF, Murphy T, nancies. Blood 2000; 96: 71–75. Piantadosi S, Seifter E, Ambinder RF, Vogelsang G, Grever MR. 33 Cazin B, Binet JL, Divine M, Lepretre S, Lederlin P, Travade P, Fludarabine and cyclophosphamide with filgrastim support in Lemaire G, Guibon O. Oral fludarabine and cyclophosphamide patients with previously untreated indolent lymphoid malig- in previously untreated chronic lymphocytic leukemia. Blood nancies. Blood 2000; 96: 71–75. 2000; 96 (Suppl. 1): 515a (Abstr. 2218). 44 Sudhoff T, ArningM, Schneider W. Prophylactic strategiesto meet 34 O’Brien S, Kantarjian H, Beran M, Robertson LE, Freilich E, infectious complications in fludarabine treated CLL. Leukemia Kornblau S, Koller C, Estey E, Lerner S, KeatingM. Fludarabine 1997; 11: S38–41. and mitoxantrone therapy in chronic lymphocytic leukemia. 45 Byrd JC, Hargis JB, Kester KE, Hospenthal DR, Knutson SW, Diehl Blood 1996; 88 (Suppl. 1): 588a (Abstr. 2341). LF. Opportunistic pulmonary infections with fludarabine in pre- 35 Rummel MJ, Stilgenbauer S, Gamm H, Karakas T, Mitrou PS, viously treated patients with low-grade lymphoid malignancies: a Doelmer H, Hoelzer D, Bergmann L. Fludarabine versus fludarab- role for Pneumocystis carinii pneumonia prophylaxis. Am J Hema- ine plus epirubicin in the treatment of chronic lymphocytic leuke- tol 1995; 49: 135–142. mia – preliminary results of a randomized phase III multicenter 46 Kalil N, Cheson BD. Management of chronic lymphocytic leu- study. Blood 2000; 96 (Suppl. 1): 2946 (Abstr. 5013). kaemia. Drugs Aging 2000; 16: 9–27.

Leukemia