* Supported by the Bundesminister für Forschung und Technologie, FRG.
1 Universitäts-Kinderklinik Münster, FRG
2 Universitäts-KinderklinikHannover , FRG
3 Universitäts-Kinderklinik Düsseldorf, FRG.

A. Introduction

The number of children with acute myelogenous leukemia (AML) who achieve remission and the number of long-term survivors have increased in the last 10 years owing to intensified chemotherapy and better supportive care. This report reviews nine pediatric studies, particularly the German AML studies BFM-78 and BFM-83. A total of 294 children with AML under 17 years of age entered the AML studies BFM-78 (n=151) and BFM-83 (n = 143) between December 1978 and January 1986. The second study is still open for patient entry. The treatment in the first study consisted of a seven-drug regimen over a period of 8-10 weeks, together with prophylactic cranial irradiation, and was followed by maintenance therapy with 6-thioguanine and cytosine arabinoside (Ara-C) for 2 years and additional Adriamycin during the Ist year [1]. In the BFM-83 study an 8-day induction with Ara-C, daunorubicin, and YP16 precedes the BFM-78 protocol. The initial patient data of the two studies are in general comparable age: median 9.11 and 9.3 years; sex: boys 54% and 52% ; WBC: median ( x 10³/µl) 24.0 and 28.5; initial CNS involvement: 9% and 7%, respectively. Extramedullary organ involvement (excluding liver and spleen enlargement) was seen more often in the BFM-83 study (32%); it accounted for only 18% of patients in the BFM-78 study. But the involvement of bone, orbits, and kidney (7% in the BFM-83 study) was not evaluated in the BFM- 78 study. The distribution of the F AB subtypes [2] shows a higher proportion of the F AB M5 type (28%) in the BFM-83 study (only 21% in BFM- 78). In both studies the myeloblastic subtypes Ml and M2 account for 20%-24% of patients, whereas the M3 and the M6 subtypes were rarely seen (2%-4%). The overall results are presented in Table 1. In the BFM- 78 study, 54 relapses (8 with CNS involvement) occurred after a median follow-up time of 5.3 years (range 3.37.0 years). The life table estimations for an event-free survival (EFS, total group) and an event-free interval (EFI, remission group) after 7.0 years are 38% (SO 4%) and 47% (SD 5%), respectively (Fig.1). In the BFM-83 study, 25 relapses occurred (4 with CNS involvement) after a median follow-up of 1.8 years (range 0.2-3.0 years). The life table estimations are EFS 48% (SD 5% ) and EFI62% (SD 6%) (Fig.l). Risk factor analysis shows that hyperleukocytosis (WBC > lOO x 10³/µl) is the main risk factor for early hemorrhage and/or leukostasis (p < 0.001, X² test), for nonresponse (p<0.05, X² test), and also for relapse (p= 0.08, log rank test). In addition, in the monocytic subtypes M4 and M5, extramedullary organ involvement was a risk factor for early hemorrhage and/or leukostasis (p<0.001) and also for relapse (p=0.07, log rank test). The Ml subtype has the best prognosis: EFS 55% (SD 7%) and EFI66% (SD 7%) after 7 years.

Table 1. Results of the AML studies BFM-78 and BFM-83, January 1986

Fig.1. Probability of event-free interval in AML studies BFM-78, and BFM-83. /, patients in CCR
(all patients of BFM- 78 study, last patient entered the BFM-83 study group).
CC R, continuous complete remission

B. Discussion

In most pediatric trials starting before 1976, the median duration of complete remission was short less than 12 months; after 3 or 4 years, life table estimation for EFI was about 30% and for survival 20% in the best studies [3]. Eight recent pediatric chemotherapy protocols with high remission rates and good results are presented in Table 2, together with one bone marrow transplantation (BMT) trial. Even though the induction/consolidation regimens with two to seven drugs differ considerably, they all include one of the anthracyclines and Ara-C. Vincristine and prednisone were also administered in the VAPA [4] and BFM [1] studies. The first St. Jude's study [5] combined Ara-C with 6-azauridine. In consolidation of the BFM studies, cyclophosphamide was given at least twice. In most studies, remission was induced by relatively short and intensive therapy with a seven-plus-three regimen (Ara-C plus daunorubicin), with or without thioguanine, which induced a complete myelosuppression and was followed by a therapy pause of approximately 3 weeks. In contrast, the BFM78 study used a prolonged induction/consolidation regimen for 8 weeks, which also caused severe bone marrow hypoplasia, but in most cases the necessary therapy pauses were short. A new strategy in intensive post-remission therapy called intensification was initiated with the V APA-10 protocol [4] and is now part of most of the new studies presented in Table 2. Lie et al. [7] reported excellent results with high-dose Ara-C as postremission therapy in a small group of children. The results of BMT, which is another way of intensification in remission, are very encouraging, especially in young patients [8]. In conclusion, new therapy strategies including intensive induction regimens together with consolidation and intensification or intensive maintenance with noncross-resistant drugs will improve the treatment results in childhood AML and increase the proportion of patients in long-term remission to 50%. The low incidence of CNS relapses in the BFM studies indicates that prophylactic CNS treatment early in remission can prevent CNS disease, and the increasing number of long-term survivors emphasizes the need for effective prevention of CNS relapse in pediatric patients. It still remains to be seen whether prophylactic cranial irradiation together with intrathecal methotrexate or Ara-C is necessary or whether systemic treatment with Ara-C infusion or especially HD-Ara-C would produce an effective liquor level. Although some results favor BMT, this therapy is currently limited to patients with HLA-compatible donors, and the long-term effects are unknown. Prospective comparisons ofBMT with chemotherapy intensification or maintenance are necessary.

References

1. Creutzig U, Ritter J, Riehm H-J, et al. (1985) Improved treatment results in childhood acute myelogenous leukemia. a report of the German cooperative study AML-BFM 78. Blood 65.298-304
2. Bennett JM, Catovsky D, Daniel MT, et al. (1976) Proposals for the classification of the acute leukaemias. Br J Haematol 33.451-458
3. Ritter J, Creutzig U, Riehm H, et al. (1984) Acute myelogenous leukemia: current status of therapy in children. In. Thiel E, Thierfelder S (eds) Recent results in cancer research, vol 93. Springer, Berlin Heidelberg New York, pp 204-215
4. Weinstein H, Grier H, Gelber R, et al. (1987) Post remission induction intensive sequential chemotherapy for children with AML treatment results and prognostic factors. In. Biichner T, Schellong G, Hiddemann Wet al. (eds) Haematology and blood transfusion, vo130. Acute Jeukemias. Springer, Berlin Heidelberg New York, pp 88-92
5. Dahl GV, Kalwinsky DK, Mirro J, et al. (1987) A comparison of cy to kinetically based versus intcnsive chemotherapy for childhood acute myelogenous leukemia. In. Biichner T, Schellong G, Hiddemann Wet al. (eds) Haematology and blood transfusion, vol 30. Acute leukemias. Springer, Berlin Heidelberg New York, pp 83-87
6. Marcus RE, Catovsky D, Prentice HG, et al. (1987) Intensive induction and consolidation chemotherapy for adults and children with acute myeloid leukaemia Joint AML trial 1982-1985. In: Biichner T, Schcl1ong G, Hiddemann Wet al. (eds) Haematology and blood transfusion, vol 30. Acute leukemia. Springcr, Berlin Heidelberg New York, pp 346-351
7. Lie SO, Slordahl SH (1987) High-dose cYtosine-arabinoside and retinol in the treatment of acute myelogenous leukemia in childhood. In. Biichncr T, Schel1ong G, Hiddemann Wet al. (eds) Haematology and blood transfusion, vol 30. Acute leukemias. Springer, Berlin Heidelberg New York, pp 399-402
8. Appelbaum FR, Thomas ED (1985) The role of marrow transplantation in the treatmcnt of Icukemia. In: Bloomfield CD (ed) Chronic and acute leukemias in adults. Nijhoff, Boston, pp 229-262

Additional participating members of the BFM-AML-Study Group
M. Neidhardt (Augsburg); G. Henze (Berlin); H.-J. Spaar (Bremen); M. Jacobi (Celle); w. Andler (Datteln); J .-D. Beck (Erlangen); B. Stollmann (Essen); B. Kornhuber (Frankfurt); A. Jobke (Freiburg); G. Prindull (Göttingen); F. Lampert (Gießen); W. Brandeis (Heidelberg); N. Graf (Hombilrg/Saar); H. Kabisch (Hamburg); G. Nessler (Karlsruhe); H. Wehinger (Kassel); M. Rister (Kiel); F. Berthold (K61n-Univ.); W. Sternschillte (K61n); 0. Sailer (Mannheim); C. Eschenbach (Marbilrg); P. Giltjahr (Mainz); K.-D. Tympner (München-Harlaching); Ch. Bender-G6tze (München-Univ.); St. Müller-Weihrich (M ünchen-Schwabing); R. J. Haas (München v. Hailnersches Spital); A. Reiter (Nürnberg); W. Ertelt (Stilttgart); D. Niethammer (Tübingen); G. Gaedicke (Ulm); Th. Luthardt (Worms)