Original Manuscript | Published:


Value of intravenous 6-mercaptopurine during continuation treatment in childhood acute lymphoblastic leukemia and non-Hodgkin's lymphoma: final results of a randomized phase III trial (58881) of the EORTC CLG


Between November 1990 and November 1996, EORTC Children Leukemia Group conducted a randomized trial in de novo acute lymphoblastic leukemia and lymphoblastic non-Hodgkin's lymphoma patients using a Berlin–Frankfurt–Munster protocol to evaluate the monthly addition of intravenous 6-mercaptopurine (i.v. 6-MP) (1 g/m2) to conventional continuation therapy comprising per oral MTX weekly and 6-MP daily. Only during the first 18 months of the randomization period, 6-MP p.o. was interrupted for 1 week after each i.v. 6-MP. A total of 877 patients was randomized to either no i.v. 6-MP (Arm A) or additional i.v. 6-MP (Arm B). A total of 217 relapses (91 in Group A vs 128 in Group B) and 13 deaths in CR (5 vs 8) were reported; a total of 134 patients (55 vs 79) died. The median follow-up was 7.6 years. At 8 years, the disease-free survival rate was lower (P=0.005) in Arm B (69.1% (s.e.=2.2%)) than in Arm A (77.9% (s.e.=2.0%)), and the hazard ratio was 1.45 (95% CI 1.12–1.89). In conclusion, as delivered in this study, i.v. 6-MP was detrimental to event-free survival.


Over the last 30 years, the results of the treatment of children with acute lymphoblastic leukemia (ALL) and lymphoblastic non-Hodgkin's lymphoma (NHL) have greatly improved. After the implementation of the total therapy concept,1 survival rates reached approximately 50%. Further modifications of this total therapy regimen have led to current survival rates of approximately 80%.2 Most of these modifications were introduced in the initial phases of the treatment (induction, consolidation and early or late intensification), whereas continuation therapy remained almost unchanged. The latter combines daily per oral 6-mercaptopurine (6-MP) and weekly, usually per oral, methotrexate (MTX).3 These two drugs inhibit the de novo synthesis of purines and have a synergistic effect. After oral administration, presystemic metabolism of 6-MP causes an important first-pass effect and results in reduced bioavailability and variable plasma concentrations.4 The inactive prodrug 6-MP is further metabolized intracellularly in cytotoxic and noncytotoxic products (reviewed by Estlin et al).5

In a preliminary study by Camitta et al,6 it appeared that intravenous 6-MP during postremission therapy of children with ALL was of marked benefit when compared to historical controls. Several reasons might explain a beneficial effect of the intravenous route:6, 7 by short-cutting the first-pass effect and getting around the high variability of intestinal absorption, parenteral administration should lead to higher tissue levels and greater efficacy.8 As a result of higher plasma concentration, intravenous (i.v.) administration of high doses of 6-MP enhances its penetration into the cerebrospinal fluid and might reduce the incidence of CNS relapses.9, 10, 11 Moreover, parenteral administration guarantees better compliance.

For these reasons, the EORTC Children Leukemia Group (CLG) decided to address the value of i.v. 6-MP given in addition to conventional continuation therapy in a prospective randomized phase III (58881) trial for patients with ALL or lymphoblastic NHL. In the same trial, all or part of the patients had been previously randomized for two other questions: at the start of induction, the first randomization addressed the value of Erwiniase as compared to Escherichia coli asparaginase.12 The second question addressed the value of moderately high-dose cytarabine when combined with high-dose methotrexate during postinduction-consolidation therapy of increased-risk patients.13 The scope of this report is restricted to the evaluation of i.v. 6-MP during continuation therapy.

Patients and methods


The trial was started in July 1989 for patients under 18 years of age with previously untreated ALL or lymphoblastic NHL. Patients with mature B-cell ALL (Burkitt-like) or with non-lymphoblastic NHL were excluded. The ALL patients were stratified in a low, an increased and a very high-risk group as previously defined.12, 13, 14 NHL patients were stratified according to Murphy's clinical stage:15 in the low-risk group if stage I or II and in the increased-risk group if stage III or IV. NHL patients who failed induction therapy were switched to the very high-risk group.12, 13 Patients who relapsed were treated according to an adapted Henze's protocol.16 For registration of a child in the study, an informed consent by his parents or his legal guardian was required. The 58881 protocol has been approved by the EORTC Protocol Review Committee and by the respective ethical committees in each participating country.

Definitions and evaluations

When bone marrow involvement was observed, patients with less than 25% of blasts in the bone marrow were considered as having NHL. All cases were studied for specific lymphoblastic cell characteristics, including morphology, immunophenotype and cytogenetics. Morphologic classification of ALL was according to the French–American–British classification.17 Classification of lymphomas was according to the working formulation and the revised European–American Lymphoma Classification.18 Frozen specimens or cell suspensions were evaluated for B- and T-lineage-associated antigens with standard techniques.19 Cytogenetic analysis of the lymphoblastic cells was performed by R or G banding and chromosomes were classified according to the International system for Human Cytogenetic Nomenclature.20

For ALL, complete remission was defined as less than 5% blasts in the bone marrow, recovery of normal hematopoiesis and no evidence of disease at any other site. For NHL patients, CR was defined by the disappearance of all clinical, imaging and cytologic signs of lymphoma.

Central nervous system (CNS) leukemia was diagnosed when neurologic abnormalities related to blastic infiltration of the CNS were observed and/or when blasts were identified on cytocentrifuge examination of CSF in which white blood cell (WBC) count was greater than 5 × 109 cells/l. Infectious, renal, hepatic and neurologic toxicities were evaluated and graded according to the World Health Organization grading system.21


A Berlin–Frankfurt–Munster (BFM)-based protocol with slight modifications was used.22 The therapy regimen was recently described.12, 13 During the course of the study, the randomization period for the first question (Erwinia vs E. coli asparaginase) did not coincide with that of the third question (addition or not of i.v. 6-MP). The registration for asparaginase started 1 year later than the registration for 6-MP: all patients registered during the first year of the EORTC 58881 trial received E. coli asparaginase from Bayer. In a second period, all patients were randomized at diagnosis to receive either E. coli asparaginase from MEDAC or Erwiniase. In a third period, after this randomization was stopped because the inferiority of Erwiniase had become evident, all patients received MEDAC E. coli asparaginase as first-line treatment.12 Increased-risk patients only were randomized to receive or not high-dose cytarabine (HD Ara-C) during interval therapy.13

All patients were eligible for the third randomization if they remained in first complete remission at the start of continuation therapy: they were randomized to receive either conventional treatment with only oral 6-MP and MTX, or to the same treatment with the addition of 4-weekly i.v. 6-MP.

In the conventional arm, continuation therapy consisted of oral 6-MP (50 mg/m2/day) and oral MTX (20 mg/m2/week). In the experimental arm, i.v. 6-MP (1 g/m2 given over 8 h every 4 weeks) was added to the treatment. During the first 18 months of the trial (November 1990–May 1992), daily oral 6-MP (but not MTX) was interrupted for 1 week after each administration of i.v. 6-MP. After the first 18 months of the trial, the treatment protocol was amended and henceforth called for uninterrupted oral 6-MP treatment in all patients. In both arms, doses of per oral drugs were adjusted in order to maintain the number of leucocytes between 2 and 3 × 109/l.

Statistical analysis

Randomization for the 6-MP question was performed centrally (EORTC Data Center, Brussels)23 before starting continuation therapy. Patients were stratified according to center, risk-group and previous randomization arms. The primary end point was the disease-free survival (DFS) from the date of randomization to the date of first relapse or to death in CR. The secondary end points were the time from randomization to isolated CNS relapse, with patients with other types of event (relapse, death in CR) being censored at the time of event, and the duration of survival (the time from randomization to death of any cause). The actuarial curves were computed using the Kaplan–Meier technique, and the standard errors (s.e.) of the estimates were obtained via the Greenwood formula.24 The differences between curves were tested for statistical significance using the two-tailed log-rank test or the log-rank test stratified by a categorical factor.24 To summarize the overall treatment difference, the hazard ratio of having an event per time in arm B vs in arm A was estimated via the Cox proportional hazards model.25 For the overall comparison, the 95% confidence interval (CI) of the hazard ratio was computed, and for two subgroup analyses, the 97.5% CI was used, due to Bonferonni correction for multiplicity. A total of 820 patients was intended to be randomized in order to detect a significant difference in terms of DFS at 5 years (70 vs 80%), corresponding to a hazard ratio of 0.63 (alpha=5%, beta=20%). All analyses were performed according to the intent-to-treat principle.



Between November 1990 and November 1996, a total of 877 patients were randomized: 439 in the no i.v. 6-MP arm, 438 in the i.v. 6-MP arm. The characteristics of both arms are listed in Table 1. Among the 877 patients, 820 were diagnosed with ALL and 57 with NHL: stage I (2 vs 2), stage II (4 vs 1), stage III (19 vs 15) and stage IV (4 vs 10). Patients’ characteristics were generally well balanced in the two arms. In arm B, the male:female ratio was slightly higher (1.39 vs 1.31) as compared with arm A, and the incidence of high WBC (100 × 109/l) was also higher (10.9 vs 8.0%); the infants represented 1.4% in arm B vs 1.8% in arm A, and ALL patients with VHR features represented approximately 6% in both arms. The type of L-asparaginase administered during the induction and intensification phases as well as the type of interval therapy (with or without high-dose cytarabine) were well balanced between the two treatment groups.

Table 1 Patients’ characteristics

Arm B was further divided into two groups: patients registered before 1 January 1992 (randomized during the first 18 months of the protocol), for whom 6-MP per oral was interrupted for 1 week after each administration of i.v. 6-MP, and those registered later who had to receive uninterrupted per oral 6-MP. During the first period, 127 vs 124 patients were randomized in Arms A and B, respectively, the majority of whom (79 vs 76) received non-Medac asparaginase. During the second period, 312 vs 314 were randomized in the two arms, respectively, out of whom 594 (248 vs 246) received Medac asparaginase. Patients’ characteristics were well balanced between Arms A and B in the first as well as the second periods.

Type of event by treatment group

At the time of analysis, the median follow-up was 7.6 years (4–12 years). There were 230 events, including 217 relapses and 13 deaths without relapse. The crude rates of different sites of relapse by treatment arm, in ALL and NHL patients are listed in Table 2. There were more relapses in Arm B than in Arm A: 91 (20.7%) in Arm A vs 126 (28.8%) in Arm B. This difference was mostly due to an increase of isolated bone marrow relapses (46 (10.5%) patients in Arm A vs 60 (13.7%) patients in Arm B) and of combined CNS relapse (17 (3.9%) vs 27 (6.2%)). Isolated CNS relapses (15 (3.4%) vs 17 (3.9%)) were evenly distributed. The 8-year isolated CNS relapse rates were comparable as well: 3.6% (Arm A) vs 4.3% (Arm B), hazard ratio=1.16. A total of 12 patients died in CR: 5 (1.1%) in Arm A and 8 (1.8%) in Arm B.

Table 2 Type of first event in ALL and NHL patients, according to the randomized arm

DFS by treatment group

The DFS rate at 8 years was 69.1% (s.e.=2.2%) for the patients in Arm B and 77.9% (s.e.=2.0%) for those in Arm A (Figure 1) (logrank P=0.005). The estimated hazard ratio was 1.45 (95% CI 1.12–1.89). In ALL patients the estimated hazard ratio was 1.40 (97.5% CI 1.03–1.90) and in NHL patients is was 3.06 (97.5% CI 0.67–13.95). In the latter group, the adjustment for Murphy stage let to a hazard ratio of 2.87 (97.5% CI 0.62–13.38).

Figure 1

DFS from randomization according to the randomized arm. N=number of patients; O=observed number of events (relapses or deaths in first CR). P-value was given by the logrank test.

Multivariate analysis based on the Cox model indicated that the difference remained highly significant (P=0.006) between the two groups (hazard ratio=1.44, 95% CI 1.11–1.88) after adjustment for the following factors: initial WBC count (P<0.0001), type of asparaginase (P=0.0001), age (P=0.0003) and sex (P=0.01) (Table 3). Period of recruitment was of prognostic importance only in univariate analysis (P=0.03). In multivariate analysis, it was not (P=0.78) and did not influence the magnitude of the treatment difference (P=0.64). Moreover, in multivariate setting, after adjustment for the same prognostic factors, no significant interaction could be determined between the type of asparaginase (Medac vs non-Medac) and the randomized group (P=0.22). In patients who had received Medac E. coli asparaginase during induction and intensification, the estimated hazard ratio of the comparison Arm B vs Arm A was 1.29 (97.5% CI 0.87–1.91) in multivariate setting, and in those who had received one of the other, less potent asparaginases (E. coli asparaginase from Bayer or Erwiniase) it was 1.73 (97.5% CI 1.09–2.75). Using a Cox model, a prognostic score was defined based on the initial WBC, the type of asparaginase, age and sex. Treatment remained significant once adjusted for this score, but the interaction between this score and the treatment arm was not significant (P=0.38), indicating that the treatment difference was quite uniform among the risk groups defined by the above-mentioned factors. Previous randomized treatment during interval therapy, high-dose or no high-dose Ara-C, had absolutely no influence on the treatment difference.

Table 3 Results of the Cox proportional hazards model regarding the DFS and survival

Duration of survival by treatment group

A total of 134 patients died: 55 in Arm A and 79 in Arm B. The 8-year survival rate was lower for the patients randomized in Arm B than for those randomized in Arm A: 81.3% (s.e.=2.0%) vs 87.2% (s.e.=1.7%) (Figure 2) (logrank P=0.03). The estimated hazard ratio was 1.46 (95% CI 1.04–2.07). In ALL patients, the 8-year survival rates were 81.5% (s.e.=2.0%) vs 86.5% (s.e.=1.7%), and the hazard ratio was 1.36 (97.5% CI 0.91–2.0.4). In NHL patients the 8-year survival rates were 77.7% (s.e.=8.1%) vs 96.6% (s.e.=1.8%), and the hazard ratio was 6.78 (97.5% CI 0.60–95.8). After adjustment for Murphy's stage, it was 5.61 (97.5% CI 1.88–5.61).

Figure 2

Duration of survival from randomization according to the randomized arm. N=number of patients; O=observed number of deaths. P-value was given by the logrank test.

Multivariate analysis based on the Cox model led to the same conclusions as for DFS (Table 3). Period of recruitment was not of prognostic importance whether in univariate or multivariate analysis (P=0.88). No interaction between recruitment period and treatment difference was observed (P=0.82). The results in Arm B were consistently worse than in Arm A independently (P=0.55) of the type of asparaginase received. After adjustment for the prognostic score defined above, the difference remained significant. The interaction between this score and treatment arm was far from being significant (P=0.79), indicating the independence of the deleterious effect of i.v. 6-MP from this score.

Toxicity and protocol compliance

Generally, toxic side effects were observed in both arms with comparable frequency. Moderate and major infectious complications were evenly distributed between the two groups: 38.7% in Arm A vs 36.4% in Arm B. The incidence of liver biologic toxicities was higher in Arm B (22.0%) than in Arm A (16.6%).

The average daily dose/m2 of oral 6-MP actually given throughout continuation therapy was lower (P<0.0001) in Arm B as compared with Arm A: the median (mean) was 55.7 (57.0) mg/m2 in Arm A vs 47.9 (48.3) mg/m2 in Arm B. Patients randomized in the first period of the trial received a median (mean) daily dose of oral 6-MP of 53.2 (53.6) vs 48.1 (45.6) mg/m2 in Arms A and B, respectively. In the second period, this was 56.0 (57.9) vs 47.9 (49.2) mg/m2 in Arms A and B, respectively. In Arm B, the median (mean) 4-weekly dose of i.v. 6-MP was 0.83 (0.73) g/m2; it was 0.90 (0.79) and 0.81 (0.70) g/m2 in the first and second periods respectively.


In contrast to what was expected from previous findings,6 randomization to i.v. 6-MP during continuation therapy has led to significantly worse outcome. At 8 years, the DFS and survival rates were approximately 9% (69.1 vs 77.9%) and 6% (81.3 vs 87.2%) lower, respectively, in Arm B.

Other studies have evaluated the efficacy of intermittent i.v. 6-MP as compared to protracted daily per oral 6-MP during post-remission therapy either for all ALL patients26 or for some risk-groups of patients.27, 28, 29 None of these studies showed any benefit but no detrimental effect either from the use of i.v. 6-MP. These studies differed from ours in at least two respects: they evaluated i.v. 6-MP given instead of and not in combination with per oral 6-MP. Moreover in two of these studies,27, 29 i.v. 6-MP was started early during postremission therapy, concomitantly with and before further intensive multiagent chemotherapy or continuation therapy. Any possible detrimental effect of i.v. 6-MP could thus have been erased by the efficacy of the concomitant and subsequent treatment courses. By contrast, in our study the addition of i.v. 6-MP jeopardized the efficiency of the sole two drugs or at least of one of the two used in order to eradicate any minimal residual disease still present.

As to why the addition of i.v. 6-MP was detrimental, two possible reasons not necessarily mutually exclusive, may be considered. The patients assigned to the i.v. 6-MP arm received a moderately lower cumulative per oral dose. This difference resulted, during the first 18 months of the trial, from the mandatory interruption of per oral 6-MP during every fourth week of the treatment and, throughout the whole of the study, from more downward adjustments of the per oral 6-MP dosing. However, for each period of the study, the deleterious effect of i.v. 6-MP, adjusted for the type of asparaginase previously administered and for other confounding factors remained approximately the same: the increase of the event (relapse or death in CR) rate per time unit in Arm B vs Arm A was 55% in the first period and 39% in the second. Overall, the excess of the event rate was 45% (hazard ratio=1.45). It has already been suggested that the total dose of 6-MP administered is positively correlated with the DFS.30 The inferiority of Arm B could also result from an inhibition by the i.v. 6-MP of the cytotoxic activity of per oral 6-MP. This activity is mainly related to the intracellular accumulation of 6-thioguanine nucleotides, which are incorporated into DNA and lead to its damage.31 Alternatively, 6-MP is catabolized into inactive thiouric acid by xanthine oxidase and methylmercaptopurine by thiopurine methyltransferase. Unfortunately, we have no pharmacological data that would have allowed us to verify the hypothesis that i.v. 6-MP could have modified the ratio of activating vs inactivating enzymatic activities and thereby have led to a lower steady-state intracellular concentration of thioguanine nucleotides.

Alternatively, Schmiegelow et al32 reported that if the doses of 6-MP given during continuation chemotherapy were tailored in order to sustain high 6-TG levels, the incidence of relapse increased after cessation of therapy in girls with ALL. These authors speculated that above a hypothetical level of 6-TG, residual leukemic cells escaped apoptosis by being maintained in the G0 phase of the cell cycle and allowed to proliferate again after chemotherapy had been stopped. In this regard, it is noteworthy that in our study as well, the divergence of the two DFS curves approximately started from completion of the treatment (Figure 1).

In conclusion, whatever the mechanism involved, addition of 6-MP to continuation therapy was detrimental. Our results also show that, notwithstanding the intensity of the initial induction, consolidation and intensification courses of the BFM protocol, small changes of the continuation therapy may lead to important changes in outcome, which underlines the persisting importance of this treatment phase.


  1. 1

    Pinkel D . Five-year follow-up of ‘total therapy’ of childhood lymphocytic leukemia. JAMA 1971; 216: 648–652.

  2. 2

    Pui CH, Evans WE . Acute lymphoblastic leukemia. N Engl J Med 1998; 339: 605–615.

  3. 3

    Riehm H, Feickert H, Lampert F . Cancer in children. Clinical Management. Berlin: Springer, 1986.

  4. 4

    Zimm S, Collins JM, Riccardi R, O’Neill D, Narang PK, Chabner B et al. Variable bioavailability of oral mercaptopurine. Is maintenance chemotherapy in acute lymphoblastic leukemia being optimally delivered? N Engl J Med 1983; 308: 1005–1009.

  5. 5

    Estlin EJ, Lowis SP, Hall AG . Optimizing antimetabolite-based chemotherapy for the treatment of childhood acute lymphoblastic leukaemia. Br J Haematol 2000; 110: 29–40.

  6. 6

    Camitta B, Leventhal B, Lauer S, Shuster JJ, Adair S, Casper J et al. Intermediate-dose intravenous methotrexate and mercaptopurine therapy for non-T, non-B acute lymphocytic leukemia of childhood: a Pediatric Oncology Group study. J Clin Oncol 1989; 7: 1539–1544.

  7. 7

    Camitta B, Mahoney D, Leventhal B, Lauer SJ, Shuster JJ, Adair S et al. Intensive intravenous methotrexate and mercaptopurine treatment of higher-risk non-T, non-B acute lymphocytic leukemia: A Pediatric Oncology Group study. J Clin Oncol 1994; 12: 1383–1389.

  8. 8

    Jacqz-Aigrain E, Nafa S, Medard Y, Bessa E, Lescoeur B, Vilmer E . Pharmacokinetics and distribution of 6-mercaptopurine administered intravenously in children with lymphoblastic leukaemia. Eur J Clin Pharmacol 1997; 53: 71–74.

  9. 9

    Covell DG, Narang PK, Poplack DG . Kinetic model for disposition of 6-mercaptopurine in monkey plasma and cerebrospinal fluid. Am J Physiol 1985; 248: R147–R156.

  10. 10

    Zimm S, Ettinger L, Holcenberg J, Kasen B, Vietti T, Belasco J et al. Phase I and clinical pharmacological study of mercaptopurine administered as a prolonged intravenous infusion. Cancer Res 1985; 45: 1869–1873.

  11. 11

    Adamson PC, Zimm S, Ragab AH, Steinberg SM, Balis F, Kamen BA et al. A phase II trial of continuous-infusion 6-mercaptopurine for childhood solid tumors. Cancer Chemother Pharmacol 1990; 26: 343–344.

  12. 12

    Duval M, Suciu S, Lutz P, Benoit Y, Robert A, Thyss A et al. Comparison of Escherichia coli-asparaginase with Erwinia-asparaginase in the treatment of childhood lymphoid malignancies: results of a randomized European Organisation for Research and Treatment of Cancer-Children's Leukemia Group phase 3 trial. Blood 2002; 99: 2734–2739.

  13. 13

    Millot F, Suciu S, Philippe N, Benoit Y, Mazingue F, Uyttebroeck A et al. Value of high-dose cytarabine during interval therapy of a Berlin–Frankfurt–Munster-based protocol in increased-risk children with acute lymphoblastic leukemia and lymphoblastic lymphoma: results of the European Organization for Research and Treatment of Cancer 58881 randomized phase III trial. J Clin Oncol 2001; 19: 1935–1942.

  14. 14

    Langermann H, Henze G, Wulf M, Rhiem H . Abschatzung der tumorzellmasse bei der akuten lymphoblastischen leukamie im kindesalter: prognotische bedeutung- und praktische anwendung. Klin Ped 1982; 194: 209–213.

  15. 15

    Murphy S . Classification, staging and end results of treatment in childhood non-Hodgkin's lymphoma: dissimilarities from lymphomas in adults. Semin Oncol 1980; 7: 332–339.

  16. 16

    Henze G, Fengler R, Hartmann R, Kornhuber B, Janka-Schaub G, Niethammer D et al. Six-year experience with a comprehensive approach to the treatment of recurrent childhood acute lymphoblastic leukemia (ALL-REZ BFM 85). A relapse study of the BFM group. Blood 1991; 78: 1166–1172.

  17. 17

    Bennett JM, Catovsky D, Daniel MT, Flandrin G, Galton DA, Gralnick HR et al. Proposals for the classification of the acute leukaemias. French–American–British (FAB) co-operative group. Br J Haematol 1976; 33: 451–458.

  18. 18

    Harris NL, Jaffe ES, Stein H, Banks PM, Chan JK, Cleary ML et al. A revised European–American classification of lymphoid neoplasms: a proposal from the International Lymphoma Study Group. Blood 1994; 84: 1361–1392.

  19. 19

    Linder J, Ye Y, Armitage JO, Weisenburger DD . Monoclonal antibodies marking B-cell non-Hodgkin's lymphoma in paraffin- embedded tissue. Mod Pathol 1988; 1: 29–34.

  20. 20

    Group FMCS. Morphologic, immunologic, and cytogenetic (MIC) working classification of acute lymphoblastic leukemias. Report of the workshop held in Leuven, Belgium, April 22–23, 1985. First MIC Cooperative Study Group. Cancer Genet Cytogenet 1986; 23: 189–197.

  21. 21

    Miller AB, Hoogstraten B, Staquet M, Winkler A . Reporting results of cancer treatment. Cancer 1981; 47: 207–214.

  22. 22

    Reiter A, Schrappe M, Ludwig WD, Hiddemann W, Sauter S, Henze G et al. Chemotherapy in 998 unselected childhood acute lymphoblastic leukemia patients. Results and conclusions of the multicenter trial ALL-BFM 86. Blood 1994; 84: 3122–3133.

  23. 23

    Vilmer E, Suciu S, Ferster A, Bertrand Y, Cave H, Thyss A et al. Long-term results of three randomized trials (58831, 58832, 58881) in childhood acute lymphoblastic leukemia: a CLCG-EORTC report. Children Leukemia Cooperative Group. Leukemia 2000; 14: 2257–2266.

  24. 24

    Buyse M, Staquet M, Sylvester R . Cancer Clinical Trials: Methods and Practice. Oxford, England: Oxford Medical Publication, 1984.

  25. 25

    Cox D, Oakes D . Analysis of Survival Data. New York: Chapman & Hall, 1984.

  26. 26

    Silverman LB, Declerck L, Gelber RD, Dalton VK, Asselin BL, Barr RD et al. Results of Dana-Farber Cancer Institute Consortium protocols for children with newly diagnosed acute lymphoblastic leukemia (1981–1995). Leukemia 2000; 14: 2247–2256.

  27. 27

    Bostrom B, Sensel M, Sather H, Gaynon P, La M, Johnston K et al. Dexamethasone versus prednisone and daily oral versus weekly intravenous mercaptopurine for patients with standard-risk acute lymphoblastic leukemia: a report from the Children's Cancer Group. Blood 2003; 101: 3809–3817.

  28. 28

    Mahoney Jr DH, Shuster JJ, Nitschke R, Lauer S, Steuber CP, Camitta B . Intensification with intermediate-dose intravenous methotrexate is effective therapy for children with lower-risk B-precursor acute lymphoblastic leukemia: a Pediatric Oncology Group study. J Clin Oncol 2000; 18: 1285–1294.

  29. 29

    Kamps WA, Bokkerink JP, Hakvoort-Cammel FG, Veerman AJ, Weening RS, van Wering ER et al. BFM-oriented treatment for children with acute lymphoblastic leukemia without cranial irradiation and treatment reduction for standard risk patients: results of DCLSG protocol ALL-8 (1991–1996). Leukemia 2002; 16: 1099–1111.

  30. 30

    Relling MV, Hancock ML, Boyett JM, Pui CH, Evans WE . Prognostic importance of 6-mercaptopurine dose intensity in acute lymphoblastic leukemia. Blood 1999; 93: 2817–2823.

  31. 31

    Elion GB . The purine path to chemotherapy. Science 1989; 244: 41–47.

  32. 32

    Schmiegelow K, Bjork OG, Glomstein A, Gustafsson G, Keiding N, Kristinsson J et al. Intensification of mercaptopurine/methotrexate maintenance chemotherapy may increase the risk of relapse for some children with acute lymphoblastic leukemia. J Clin Oncol 2003; 21: 1332–1339.

Download references


This work was supported in part by grants from Télévie 2001 (Grant No 7.4561.01) and 5U10CA11488-18S2 and 2U10CA11488-19 through 5U10-CA11488-34 from the National Cancer Institute. Purinethol (6-mercaptopurine) for intravenous use, a product of Wellcome, was provided free of charge by the National Cancer Institute. The contents of this work are solely the responsibility of the authors and do not represent the official view of the National Cancer Institute or of Wellcome. Additional members of the EORTC CLG participated to this study: B Nelken (Lille), A Babin-Boilletot (Strasbourg), C Béhar (Reims), N Francotte (Montegnée), B Lescoeur (Paris), G Margueritte (Montpellier), F Millot (Poitiers), X Rialland (Angers) and F Méchinaud (Nantes). The members of different subcommittees: Dr M Malet (Caen), A Falkenrodt (Strasbourg), M Fournier (Lille), Dr N Dastugue (Toulouse), Drs A-M Manel and M-P Pages (Lyon). A warm thank to the previous data managers: I VandeVelde and G Solbu. Thanks to E Jacqz-Aigrain and G Milano for their careful review of an earlier version of this paper and their input in the Discussion.

Author information

Correspondence to J Otten.

Rights and permissions

Reprints and Permissions

About this article


  • 6-MP
  • childhood
  • ALL
  • NHL
  • randomized trial

Further reading