A95 3

1 CURRENT APPROACHES TO THE TREATMENT AND MANAGEMENT OF CHRONIC LYMPHOCYTIC LEUKEMIA Emilio M o n t s e r r a t , MD

Apoptosis in Chronic Lymphocytic Leukemia J.L. Binet, F. Mentz, H. Merle-Beral and the French cooperative Group on C.L.L.

Postgraduate School of Hematology "Farreras-Valenti". D e p a r t m e n t of Medicine. U n i v e r s i t y of' B a r c e l o n a . Hospital Clinic. Barcelona, Spain A n u m b e r of f a c t s s h o u l d be c a r a f u H y c o n s i d e r e d w h e n p l a n n i n g t h e r a p y in p a t i e n t s w i t h c h r o n i c l y m p h o c y t i c leukemia (CLL) : 1 ) CLL is u s u a l l y d i a g n o s e d in p e r s o n s w h o s e median a g e is a b o u t 65 y e a r s , w i t h o n l y I0% of t h e p a t i e n t s b e i n g l e s s t h a n 50 y e a r s old; 2) CLL c a n n o t b e c u r e d b y c o n v e n t i o n a l t h e r a p y ; 3) t h e p r o g n o s i s of patients with C L L is variable and some patients may live as l o n g as normal controls; and 4) in patients requiring therapy the achievement of a response is associated with a longer survival. Therefore, for the majority of patients with CLL, the most reasonable goals of therapy are: a) symptoms paIHation, b) correction of cytopenias; and e) switch of the disease to a less advanced stage. In selected patients (e. g., younger persons with poor prognostic features) attempts should be made to achieve a complete response (CR). For the average patient, however, the benefits of achieving a C R should be balanced with the risks of therapy. Treatment is indicated in patients presenting with advanced stage (anemia and/or thromboeytopenia due to bone marrow failure), constitutional "B" symptoms, bulky disease, rapid doubling time, and/or immune hemolytic anemia unresponsive to corticosteroids. Treatment of patients in early stage (Binet A) with chiorambucil has resulted in a delay in the rate of disease progression but no survivai benefit. Measures to treat C L L include: a) radiation therapy; b) chemotherapy (singie agents or combination chemotherapy); e) biotherapy, and d) transplants (aHografts and autografts). The standard treatment has been chlorambueil but remissions with this agent are rare and transient. Purine analogues, particularly fludarabine, give a higher response rate. Whether this will translate into a longer survival, however, remains to be seen. Experience with transplants is accumuiating. Biotherapy is promising but most likely will have a complementary role in the t r e a t m e n t strategy of patients w i t h CLL.

Flndarabine versus Cychiphosphamide, Doxorubicin, Vincristine (CAP) for the Treatment of Advanced Stage Chronic Lymphocytic Leukemia. W. Hiddenmm~ S. Johnson, A. Smith, M. BjOrkholm,H. L6fl'ler, B. Emmerich, P.J. Wyld, J. Binct, for the

French Cooperative Group on eLL The promising results of Fledarabinc single agent therapy in advanced stages of chronic lymphocytic leukemia (CLL) strongly suggest that this agent provides new aspectosin the treatment of this disease. The current study was initiated to compare the efficacy and side effects of single agent Fludarabine treatment with CAP in previously untreated and pretreated patients with advanced stage CLL in way of a prospective randomized clinical trial. Patients with B-CLL at ages older than 18 years were entered into the study if they presented with either (1) previously untreated B-CLL of Binct stages B or C or (2) relapsed B-CLL pretreatod with non- anthracyclinecontainingregimensfor more than six months but less than three years. Patients were randomly assigned to either Fludarabine given at a dose of 25 mg/m2/d over 5 consecutivedays or CAP comprising cyclophosphamide 750 rog/m2/dand doxorobicin 50 mghu2/d on day I and prednisone 40 mg/m2/d on days 1-5. Both regimens were repeated evew 28 days for a total of 6 ~ . Ramiomizallon was stratified for prior therapy, patient age and stage of disease. Response to therapy was assessed according m the criteria defined by the International Workshop on CLL, side effects were evaluated based on the World Health Organization grading system. Results: 208 patients entered the trial of whom 196 cases were fully evaluable. 100 cases were previously untreated, while 96 patients had received prior therapy. Remission rates were significantly higher in Fludambine than CAP treated patients with overall response rates of 60 % vs. 44 %, respectively (p ffi 0.023). Further analysis indicated a higher response to Fhidarabinr in both untreated and pretreated cases which was statistically significant, however, in pretreated cases only (48 % vs. 27 %, p ffi 0.036). Likewise, Fhidarabine appeared superior in both stage B and C disease with the difference being statistically significant in stage C cases only (47 % vs. 23 %, p = 0.018). Patients achieving a CR or PR aRer Hudarabine experienced a significantly longer event free interval than the comparable CAP-treated group and had a tendency towasds a longer survival. Treatment associated side effects consisted predominantly in myelosuppression and grannincytopenia nccoring after 19 % of courses with Fludarabine and after 22 % CAP treatment cycles. CAP treated patients, however, exp~enced a higher frequency and severity of nausea and vomiting (25 % vs. 5 %, p < 0.001) as well as of hair loss and allopecia (65 % vs. 2 %, p < 0.001)

Apoptosis

plays a key role

in Chronic

Lymphocytic Leukemia (C.L.L.) : Spontaneous apoptosis appears in vitro after 48 hours of culture and varies from one patient to another; Corticosteroids, fludarabine, 2 chlorodesoxydesa-minase

and

radiations

induce

apoptosis of B-C.L.L. lymphocytes. We will discuss two aspects of apoptosis in C.L.L. 1/ Comparison of the effect of fludarabine (in which apoptosis is mainly involved), CAP and CHOP in previously untreated stage B and C C.L.L. (first interim results of a randomized clinical trial in 247 patients. 2/ Study of in vitro apoptosis (morphological, functional and pharmacological aspects).

POTENTIAL IMMUNOLOGICAL ACTION OF PURINE NUCLEOSIDE ANALOGS G. Dighiero Institut Pasteur Unitd d'Immunoh6matologie et d'Immunopathologie, France. Purine nucleoside analogs are a new class of drugs with activity against non-dividing lymphocytes; thus, they should play a major role in the treatment of low grade lymphoid malignancies. These drugs have been shown to have strong effect in DNA synthesis on actively dividing cells, through interference with DNA polymerases and ribonucleotide reductase, and for some of them through inhibition of DNA pfimase and DNA ligase. However, the cell cycle kinetics of low grade lymphocytic lymphomas is characterized by the presence of very low growth fractions. Hence, the action of these drugs in slowly progressing lymphoid malignancies cannot be accounted by the same mechanism observed in actively proliferating tumors and needs to be explained through activity agaifist quiescent resting lymphocytes. Recent work has stressed the role of purine analogs in inducing programmed cell death of quiescent lymphocytes, which could be explainde through the induction of accelerated DNA strand breaks. This process leads to consumption of NAD for poly(ADP-ribose) synthesis, which could induce critical depletion of ATP. In addition, as their action extends to normal resting lymphocytes deleterious effects related to their immunosuppressive action are also observed, i.e. prolonged lymphopenia predominating in T cells and especially in CD4 subset and an increased frequency of opportunistic infections. Nevertheless, beneficial effects of this immunosuppressive action have also been reported for 2CDA in the case of autoimmune hemolytic anemia associated to CLL. These last results could offer some promise for these drugs in the field of immunosuppression. To substantiate this possibility, a considerable amount of work needs to be carried out in order to better define the mechanisms of action of these drugs, as well as their potential activity on different immunological effectors. Also, studies in animal models of autoimmune disease should be undertaken.

Ag6

THE MANAGEMENT LYMPHOMAS

OF

LOW-GRADE

NON-HODGKIN

C. Pott-Hoeck and W. Hiddemann, Dept. of Hematolog3'and Oncology, University Hospital Gottingen,Robert-KochStr. 40. 37075 G6Uingen Low-grade non-Hodgkin lyrnphomas (NHL) are a histologic distinctive group of lymphomas with a heterogenous growth pattern and biology. Although a wide array of therapies comprising single agent and combination chemotherapy or combined medality approaches including radiotherapy have been available during the last years, most patients succumb to their disease due to a continous remitting course of the lymphoma. In the early stages I and II and selected cases of stage III disease radiotherapy has a curative potential. Initial chemotherapy for patients with advanced stage III and IV disease induces remission rates of more than 6080 % without any curative potential. The impact of interferone alpha (IFN) on remission duration and survival in addition or after cytoreductive chemotherapy was investigated in different studies. Preliminary results of the German LowGrade NHL Study Group and others reveal a favourable effect of IFN maintenance therapy on prolongation of disease free survival. Treatment at relapse remains an important feature in management of NHL and several new treatment strategies including myeloablative radio-chemotherapy protocols supported by PBSC or ABMT are under investigation. Histologic transformation from low-grade to intermediate or high-grade NHL is a major cause of morbidity and occurs with an increasing frequency over time. Moreover the proof of moleculargenetic markers in clinical remission of NHL promises to be an important tool for monitoring the disease and evolves to be a prognostic factor at least in transformed NHL. With the development of the structurally related purine analogs Fludarabine phosphate, 2-Denxycoformycine and 2-Chlordeoxyadenosinenew perspectives in the therapy of low-grade NHL were provided. Fludarabine has shown to be effective in the treatment of chronic lymphocytic leukemia (CLL) and low-grade NHL, predominantly of the follicular subtype. In consecutive multicenter phase II studies the German Low-Grade NHL Study Group investigated Fludarabine as single agent and in combination with Mitoxantrone and Dexamethasone (FMD) in 76 Patients with relapsed low grade NHL and achieved response rates of 39%. Future perspectives must be directed by an understanding of the biology of lowgrade NHL and the recognition of considerable heterogenity among individuals.

THE FUTURE ROLE OF PURINE ANALOGUES IN ONCOLOGY M.J. Keating, University of Texas M.D. Anderson Cancer Center, Houston, Texas 77030 The purine analogues fludarabine monophosphate (Fludara) and 2-chlorodeoxyadenosine (2CDA) have emerged as major new drugs in the management of hematologic malignancies. The initial activity of these drugs was noted in indolent lymphoid malignancies and 2CDA has emerged as the treatment of choice in hairy cell leukemia (HCL) and Fludara as the most active single agent studied in chronic lymphocytic leukemia (CLL). Numerous effects of these drugs on the activities of e n z y m e s important in DNA synthesis and repair have been identified and a key role in induction of apoptosis has been noted. Fludara and 2CDA both enhance the formation of cytosine arabinoside triphosphate (ara-C} in leukemic cells and have been demonstrated to inhibit repair of DNA damage caused by radiation, cisplatinum, mitoxantrone. This activity presumably extends to other drugs. Fludara has the highest documented complete (CR) and partial (PR) remission rate in both previously treated and untreated patients with CLL. Comparative studies against salvage and frontline regimens have demonstrated superiority in response rate. Combinations of Fludara with anthracyclines, alkylating agents, and mitoxantrone are being actively explored in CLL. The activity of Fludara and 2CDA in low grade lymphoma has led to combinations of these agents with mitoxantrone and corticosteroids. The addition of eorticosteroids appears to enhance the probability of opportunistic infections. The activity of Fludara in enhancing ara-CTP formation in acute leukemia, myelodysplastic syndrome (MDS) has led to exploration of this combination alone, Fludara + ara-C (FA), or together with GCSF (FLAG), and with idarubicin (FLAG/Ida). All three regimens are very active in poor and better prognosis patients with acute myelogenous leukemia (AML) and MDS and are well tolerated. The effect of Fludara in inhibiting repair of D N A damage has led to protocols and solid tumor clinical trials with cisplatinum in ovarian cancer. Trials with cisplatinum or radiation in head and neck cancer are being initiated. Purine analogues as single agents an' the indolent lymphoproliferative d i s e a s e s a r e likely t o b e used for many years. The extended role will be in combinations both in hematologic malignancies and solid tumors. The immunosuppressive effect of the purine analogues is likely to lead to increased use of these drugs in preparative regimens for bone marrow transplantation (BMT) and perhaps use in enhancing cytotoxicity in association with preparative regimens in autologous and allogeneic BMTs.

MECHANISMS OF R E G U L A T E D DRUG SENSITIVITY. E.A.MCCULLOCH Regulated drug sensitivity (RDS) is defined as the influence of growth regulators on the drug sensitivity of cancer cells measured in culture. The blast stem cells of AML provide a suitable model system. Myelopoiefic growth factors, all trans retinoic acid (ATRA) and hydrocortisone (HC) have been found to alter some drug dose response curves, with degree and direction of effect often schedule-dependent. GM-CSF and HC were found to protect AML blasts against cytosine arabinoside (ara-C) killing; G-CSF increased sensitivity. ATRA regularly increased sensitivity only if given after drug. Regulator-mediated change in cell cycle parameters was an potential mechanism for RDS. Measurements using high specific activity tritiated thymidine (3HTdR) to identify stem cells in the S phase provided evidence against this attractive hypothesis. We compared RDS using different drugs. Unlike am-C, anthracyclinesensitivity was changed by ATRA or HC but not by growth factors. Thus, a single mechanism is unlikely to explain RDS. We asked if the proto-oncogene bcl-2, known to influence apoptosis, was part of the RDS mechanism. We found that bcl-2 was down-regulated by retinoic acid. Bcl-2 has an anti-oxidant effect. Exogenous H20 z is metabolized to produce free radicals; we asked if sensitivity to H202 was regulated. We found that ATRA after H202 increased killing, while HC before H202 was protective. We asked whether N-acetylcysteine (NAC) a potent radical scavenger, would alter drug sensitivity. We found that NAC before agent provided partial protection against the lethal effects of both ara-C and HzO2. We concluded that era-C, in addition to direct killing following incorporation into DNA, causes the production of toxic free radicals. ATRA may sensitize and HC protect through mechanisms that influence the outcome of damage by free radicals. The control network that regulates intracellular redox is complex, and includes member of the bcl-2 family and cytochromes; the network provides many targets permitting drug sensitivity to be regulated by more than one mechanism. Ontario Cancer Institute, 500 Sherbourne St. Toronto, M4X 1K9

G-CSF Based Palming Concept for AML Therapy M. Andreeff, U. Consoli, H. Kleine, H. Engel, R. Munker, I. EITounsi, S. Jiang, S. Zhao, A. Goodacre, G. Sanchez-Williams, S. Kornblau, E. Estey. U.T.M.D. Anderson Cancer Center, Houston, TX 77030 Cytokine based "priming" of AML is based on the concept that cells in the ceil cycle provide better targets for chemotherapeutic agents than non-cycling quiescent cells. In vitro and clinical studies have demonstrated that rh cytokines, in particular GCSF, GM-CSF, IL-3, PIXY and SCF can indeed increase the number of cycling cells, and changes in important regulators,r4 of cell proliferation are well documented, including PCNA, KJ-67, Rb, c-myc end other genes controlling transit through G 1 . Changes in the nucleoside transporter, in ARA-CTP formation, MDR1 expression and other cellular functions are also documented. The clinical results with FLAG and FLAG-IDA protocols have resulted in appr. doubling of CR rates for poor prognosis patients, but remission duration has not been affected. While these results support the original concept, cytokines are also known to have anti-apoptotic effects on hematopoieticcells. We are documenting this in normal progenitor cells and in leukemia. In order to elucidate this process, we have studied the effect of G-CSF on bcl-2 expression and found increased and decreased bcl-2 in rive in patients treated with G-CSF. It is suggested that baseline cellular cytokine proliferation states determine the response to exogenous G-CSF The involvement of bcl-2, c-myc and p53 in apoptesis points to diverse mechanisms affected by exogenous G-CSF+ Presently, the effect of G-CSF on minimal residual disease (MRD) is being investigated using the FACS/FISH/BUdR technique, which identifies a few as 1 leukemia in 10,000 normal cells. The level and proliferation of residual leukemic cells were found to predict relapse. Negative regulators of hematopoietic may be useful in controlling MRD in the future.

A97 9

11

G-CSF IN HIGH-RISK ALL THERAPY D. Hoelzer, O.G. Ottmann, E. Gracien, R. Iyer, A. Ganser, R. Reutzel, M. Graf, T. Lipp, F.W. Busch, M. Schwonzen, H. Wandt, G. Heil, P. Koch, A. Heyll, P. Meyer, M. Bentz, S. Peter, H. Diedrich, K. Kolbe. Dept. of Hematology, Frankfurt, Germany In ALL neutropenia occurs not only after chemotherapy but also during induction and is treatment limiting. Therefore a randomised multicentre trial examined whether r-metHuG-CSF (Filgrastim) given concomitantly with chemo-radiotherapy could reduce treatment related neutropenia in adult ALL. 76 consecutive pts were randomly assigned to receive either G-CSF (5 ~g/kg/day sc) (Arm A, n=37) or no growth factor (Arm B, n=39) parallel with wks 5-8 of induction chemotherapy and prophylactic CNS irradiation. G-CSF was continued until neutrophil counts exceeded 3000/~i. The median duration of neutropenia (ANC 2 weeks) chemotherapy delays which occurred in only 24% of patients in Arm A versus 46% in Arm B. These beneficial effects of G-CSF were seen to the same extent in the high and low risk patients. In an adult ALL study with a similar induction regimen where G-CSF was given during weeks 1-8 also the CR rate was improved. In conclusion, concomitant application of G-CSF and induction therapy in ALL is feasible. In all these studies a significant reduction in neutrophil recovery time, a decreased infection rate and better adherence to the treatment schedule were observed. Whether there is a survival benefit has to be determined with longer follow--up.

TRANSPLANTATION O F A L L O G E N E I C FILGRASTIM-MOBILIZED P E R I P H E R A L BLOOD PROGENITOR CELLS (PBPC) N. Schmitz,, P. Dreger, M. Suttorp, T. Haferlaeh, H. 1.6ffler, A. Hunter, and N.H. Russell, Department of Internal Medicine II, University of Kiel, Germany and Nottingham City Hospital Nottingham, UK Transplantation of allogeneic PBPC may have advantages over bone marrow transplantation (BMT) with regard to the kinetics of hematopoietic engraftment, immunologic recovery, and a possibly enhanced graft-vs-leukemia(GVL) effect. The donor could be spared the risks of general anesthesia and the discomfort usually associated with the marrow harvest procedure. However, as the leukapheresis products retrieved from healthy volunteers contain approximately 10 times more Tlymphoeytes as compared to a normal marrow harvest it was anticipated that parents (pts) grafted with aUogeneie PBPC should contract more frequent and/or serious graft-vs-host disease (GVHD). We have now grafted 11 patients (pts) with allogeneic PBPC from HI,A-identical sibling donors, the posttransplant course is currently evaluable in 10 pts. The pts median age was 41 years, 8 pts were male. The patients' diagnoses were AML, CR1 (n=3), relapsed A M L (n=3), A M L CR3, ALL CR2, refractory ALL, and CML in accelerated phase. PBPC were mobilized with filgrastim (5-10 ug/kg) administered for 5-6 days (d) to the HI.A-identical donors. With a median of two l eukaphereses (range 1-3) ax median of 6.9 x l0 t) CD34+ cells, 36.8 x 10" CFU-GM, and 317 x I0" T eells/kg recipient weight were harvested. After myeoablative therapy with busulfan/cyelophosphamide, TBI/cyclophosphamide, or the BAC-regimen the leukapheresis products were transfused without further manipulation. Prophylaxis of GVHD was either eyclosporin A (CyA) or CyA + MTX. Hematopoietic reconstitution was achieved in aU pts with a median of 15 an~ 16 days, respectively, needed to surpass an ANC of 0.5 and 1.0 x 10~/L, respectively. The median time to achieve an unsupported platelet count > 20 x 10~/L was 19 d after grafting. Eight pts are currently alive without evidence of disease up to 655 d after grafting; 2 pts grafted for relapsed AML achieved C R after transplantation but relapsed again and died of leukemia on d +48 and d +70, respectively. A 41-year-eld male grafted from his phenotypical identical 67-year-old father died 359 d after grafting of pneumonia and liver failure with chronic GVHD. We conclude that transplantation of unmanipulated aUogeneic PBPC is feasible, results in long-term engraftment, does not cause detrimental GVHD, and may be a valid alternative to allogeneie bone marrow transplantation.

10

12

PERIPHERAL STEM CELL TRANSPLANTATION IN ACUTE MYELOID LEUKEMIA. 1. REIFFERS. H6pital Haut-Lev&lue, CHRU Bordeaux, 33604 PESSAC

HEMATOPOIETIC GROWTH FACTORS IN LEUKEMIA THERAPY" Th. Btictmer*, W. Hiddemann, B. WOramxm, 1L Rottmann, M. Z~hlsdorf, G. Maschmeyer, W.-D. Ludwig, M.C. Sanerland, J. Frisclx G. Schuh for the AML Cooperative Group

Autologous Peripheral Blood Progenitor Cell (PBPC) transplantation is increasingly used for patients with solid tumors, multiple myeloma and malignant lymphomas, but is nut frequently performed in patients with acute myelo~dleukemia (AML). In 1987, we started a prospective study comparing autolognns unpurged bone marrow transplantation (BMT) and PBPC transplantation in patients in first complete remission (CR). All the patients who were in CR after induction chemotherapy were given a consolidation chemotherapy (high dose Ara-C : 3g/m2 x 8 ; Daanorubicin : 45 mg/nd x 3) before stem cell collection. No growth factors were given. In patients allocated for PBPC transplantation, the coils were collected by means of six leukephereses. Thirty-three patients were transplanted after a conditioning regimen consisting of Busulfan (4 mg/k_gx 4) and Melphalan (140 rag/m2) then the reinfusion of either bone marrow (n = 17) or PBPC (n = 16). The main characteristics of patients in these latter two groups were similar. Alter a median follow-up of 36 months, there was no statistical differences in the actuarial proportion of patients who either had a leukemic relapse or were alive in continuous CR after either type of transplant, ff the platelet recovery did not differ in both groups, the duration of granulucytopenia was significantly shorter after PBPC transplantation than after BMT (lO < 0~005). More recemly, some other patients were transplanted with G-CSF mobilized PBPC and their results will be presented. Within the European Bone Marrow Transplant (EBMT) Group, we have retrospectively reviewed the data of AML patients who underwent either BMT or PBPC transplantation in first CR. A total number of 886 (BMT) and 63 (PBPC) transplantations were reviewed. We found that the 4-year probability of relapse was significantly higher after PBPC transplantation than after either purged BMT anpurged BMT (p < 0.0008). The 4-year probability of surviving without disease was signikicantlyhigher after BMT than after PBPC transplantation (I3 < 0.015). However. by multivariate analysis or match-pair study, we failed to detect any statistically significant difference for the DFS of patients undergoing PBPC transplantation or BMT. The duration of granniucytopenia was significantly shorter after PBPC transplantation. (14 days) than after BMT (29 days) (p < 0.001) but the duration of thrombucytopenia was similar. Thus, further prospective studies are needed to compare prospectively PBPC transplantation and BMT.

Conflicting data have recently been reported from multicenter randomised studies on growth factors in acute leukemias, administered either as a part of supportive care or in order to enhance the antileukemic effect of chemotherapy. Positive supportive effects resulted from G-CSF in ALL (Blood 82 (i):757, 1993; Blood 84 (1):911, 1994) and AML (N Engl I Med 323:871, 1990) while GM-CSF in AML showed either positive (Blood 82 (1):1299, 1993) or no supportive effects (PrtK: ASCO 13:992, 1994). GMCSF priming in AML either improved (Proc ASCO 12:985, 1993) or rather adversely affected pts outcome (Blood 84 (1):909, 1994). We here present an update of our study in 96 pts 49 (16-75) years of age with newly diagnosed AML entering between 7/90 and 1/94. Chemotherapy was TAD followed by HAM for induction, TAD for consolidation and monthly reduced TAD for maintenance. Differently from pts over 60 younger pts received strict double induction with 3 instead of 1 g/ms AraC in HAM. GM-CSF 250 ug/m~/d started 24 h before chemotherapy, continued until neutrophil recovery and was repeated for a total of 5 sequential chemotherapy courses. Among hematologic effects absolute blood blast counts were increased during the initial 3 days to up to 5 fold those in the controls. Under GM-CSF neutrophils recovered earlier after the first TAD course (9 = .007) but not alter the subsequent courses, and platelets showed a progressive delay in recovery from course to course with 22 vs 15 days al~cr consolidation (p = .035). Blasts were cleared from the day 16 bone marrow in 52% GM-CSF and 47% control pts. 81% and 84% went into CR and persistent leukemia ucccured in 2% and 3%. Similarly in both arms disease free survival after 3 years is 45% and in younger pts there is a trend in favour of GM-CSF during the first 2 years. We conclude that GMCSF multiple course priming and longterm administration appears not to protect AML against chemotherapy, delays platelet recovery (stem cell effect) and may delay early relapses. The chemotherapy applied produced 82% CR and 45% CCR in pts of 16-75 years. GM-CSF did not further improve on those favanrable results as of November 1994. *Present address: Department of Medicine/Hematology/Oucology, University of MOnster, Germany

A98

13

15

SM-CSF BASED PRIMING CONCEPTS FOR A.M.L.J.J. Wielenga, University Hospital Rotterdam-Dijkzigt, The Netherlands.

CALGB STUDIES WITH HEMATOPOIETIC GROWTH FACTORS IN PATIENTS WITH A M L C.A. Schiffer', R.M. Stone, B.A. Peterson, J. Moore, for the CALGB. The C A I ~ B has conducted studies evaluating the use of hematopoietic growth factors to shorten the duration of neutropenia, as well as to "prime" leukemia cells to modulate cytokinetic resistance and enhance cytotoxicity with cell cycle specific agents. CALGB 8923 randomized 388 untreated patients > 60 years of age with de novo AML to either placebo or 5/tg/kg GM-CSF beginning the day after completion of standard induction therapy with daunorubicin and ara-c. Complete remission rates (52% vs 54%), incidence of grade 4 and 5 infections and duration of hospitalization were similar in the groups assigned to GM-CSF and placebo respectively. There was no evidence of GM-CSF associated stimulation of leukemia growth. There was a small decrease in the median duration of neutropenia in patients receiving GM-CSF (15 days vs 17 days, p=.02) but this was not felt to be clinically significant. In contrast, CALGB 9022 demonstrated a shortening of the duration of neutropenia by approximately 12 days and a possible shortening of the duration of thrombocytopenia in AML patients < 60 years of age in CR who received mitoxantrone and diaziquone followed by G-CSF for their third course of post remission consolidation [compared to a cohort of similar patients treated earlier with the same drugs alone]. The relapse rate was not increased in patients receiving the G-CSF and this study suggests that growth factors could possibly have a role in attenuating the side effects of post remission therapy. Lastly, CALGB 9021 randomizes patients with AML in first relapse to GM-CSF (3/zg/kg/day by CI for a maximum of 5 days before high dose ara-c) or to no pretreatment; GM-CSF is continued for 24 hours after the ara-c. 151 patients have been treated to date with an overall CR rate of 43%. The results remain coded by treatment arm; accrual should be completed shortly.

Therapeutic use of hemopoietic growth factors (HGFs) includes mitigation of pancytopenia induced by chemotherapy. In vivo stimulation of malignant cells in patients with acute myeloid leukemia (AML) might counteract with the eradication of malignant cells. However, HGFs have been demonstrated to increase in vitro ARA-C cytotoxicity on clonogenic AML cells. Therefore, properly scheduled combinations of HGFs and chemotherapy might result in a more effective treatment than chemotherapy alone. This hypothesis has been tested for GM-CSF in several recently performed clinical trials. In one of these studies, performed by the Dutch and'Belgian HOVON group in cooperation with the Swiss SAKK group, several combination schedules of GM-CSF and remission induction chemotherapy were compared: GM-CSF before and during chemotherapy (group 1A), GM-CSF before, during and following chemotherapy (group 1B), no GM-CSF (group 2A) and GM-CSF following chemotherapy (group 2B). The factorial design of this study enables comparison between groups 1 and 2 for the effect of priming as well as comparison between groups A and B for the effect of GM-CSF on hemopoietic regeneration. The overall Complete Remission (CR) rate was 77 %. A significant difference in CR rate between the groups was not found. The mean day of regeneration of neutrophUs was two day earlier in the B groups as compared to the A groups. The following side effects were observed: fever, skin rash, fluid retention, pericarditis, bone pain and flue like symptoms. In vitro HGF stimulability is different among patients. Therefore, GM-CSF is possibly not the optimal factor for all patients. Furthermore, timing of GM-CSF in relation to the chemotherapy might have been suboptimal Clinical studies can hsrdly answer all such quost!ons. Thereforc the effects ~,; 14GFson in vivo human AML growth is currently under investigation in a SClD mouse model. Clinical studies in the future may be based upon such preclinical data in combination with a mathematical description of AML growth.

"Univ of Maryland Cancer Ctr,22 S. Greene St., Baltimore, MD 21201

14

16

THE USE OF HEMATO POIETIC GROWTH FACTORS (HGF) IN ACUTE

PRECLINiCAL EVALUATION OF GM-CSF/IL3 FUSION PROTEIN (PIXY 321) IN ACUTE MYELOID LEUKEMIA. A. Tafud, T. Valentini, MG. Mascolo, MT. Petrucci, F. Mandelli, L. De Felice, Hematology, Department of Human Biopathology, University "La Sapienza", Rome, ITALY. In vitro pdming with growth factors (GFs) of acute myeloid leukemia (AML) cells has been reported to increase percentage of cycling cells and enhance cytotoxic effects of cycle specific drugs. Proliferative response may be heterogeneous for different GFs, Aim of our study was to investigate in vitro proliferative effects of the novel cytokine GM-CSF/IL-3 fusion protein (PIXY 321) on 20 de novo AML, The same evaluation was performed using other GFs as well as Stem Cell factor (SCF), GM-CSF, IL-3, G-CSF and the GF combinations SCF+PIXY and GMCSF+IL-3. GFs were used in liquid colture for 48-72 hours to determine both proliferative effects on clonogenic cell growth (CFU-LI, and cell cycle changes by flow cytometric DNA/RNA (acridine-orange). AML blasts, after GFs priming, were exposed to different concentrations of Ara-C and cytotexicity was measured as percentage of CFU-L inhibition. Results are expressed as mean values:

MYELOGENOUS LEUKEMIA (AML).

JM Rowe, JW Andersen, JJ

Mazza, JM Bennett, E. Paietta, FA Hayes, PH Wiemik for the Easatern Cooperative Oncology Group (ECOG). Rochester, New York, USA. The use of HGF for AML remains controversial due to conflicting reports. It is important to distinguish between HGF given to shorten the neutropenia or as priming agents prior and during induction therapy. If given following induction, timing could be crucial and may also depend on whether marrow aplasia had been achieved. Furthermore, HGF may have different properties and should not be generically compared without specifying the product; G-CSF, yeastderived GM-CSF and E. Coil-derived GM-CSF. The latter two have different toxicity profiles and the efficacy may not be identical. In a prospective randomized double-blind ECOG study of 124 adult patients (>50-70 years) with AML, yeast-derived GM-CSF (Sargramostim) or placebo was administered 4 days after 1 or 2 courses of standard induction therapy only if marrow aplasia had been demonstrated. The study drug was continued until neutrophil recovery. Under these conditions patients receiving GM-CSF had a CR rate of 60% vs 44% for the placebo group (p= .08); the neutrophil recovery was significantly shortened; the treatmentrelated mortality and infectious toxicity was also reduced with GMCSF (p= .049 and .015 respectively). The median survival for all patients was 10.6 months on GM-CSF vs 4.8 months in the placebo group (p= .048). Thus, while the literature suggests that the use of HGF in AML in general are safe when given during induction therapy, this phase Itl study demonstrated that yeast-derived GMCSF given in induction following demonstration of aplasia significantly shortens the period of neutropenia and likely impacts upon the morbidity, mortality and overall survival of these patients.

GF

I

I

I

GO S CFU-L (fold differences)

RNA-I G1 (Index)

CFU-Lkilling (%at1uM Ara-C)

Control / / / 16.3 52.74 SCF+PIXY 3.44 8 . 9 9 6.72 22.6 77 SCF 1.70 3.48 2.51 19.7 66.18 PIXY 321 1.90 5 . 3 8 3.59 19.8 75.95 GM-CSF 1.55 5.69 2.31 19.3 73.27 IL-3 1.48 5.84 2.92 20.0 72.42 G-CSF 1.31 6.10 2.11 19.0 67.94 GM+IL-3 1.84 6.34 3.53 20.6 73.27 Among the single cytokines PIXY was able to induce the best proliferative effect on clonogenic cell growth comparable to that obtained by GM+IL3. The combination of SCF+PIXY induced an higher proliferative response than other GF conditions and the most pronounced effects in more than 50% of the cases. Cytotoxic effects of Ara-C were also enhanced by SCF+PIXY, although benefit in leukemic cell killing was different from case to case and for the different cytokines tested. In conclusion this in vitro study showed activity of PIXY on AML blast cells pdming and Ara-C sensitization.

A99 19

17 PIXY321 (GM-CSF/IL-3 FUSION PROTEIN) FOLLOWING H I G H - D O S E CHEMOTHERAPY AND AUTOLOGOUS BONE M A R R O W T R A N S P L A N T A T I O N FOR LYMPHOID MALIGNANCY. J. Vose, U n i v e r s i t y of Nebraska Medical Center, Omaha, Nebraska In a phase I/II trial, 50 patients received PIXY321 following ABMT for lymphoid m a l i g n a n c y in doses ranging from 50 - I000 mcg/M z. The cytokine was a d m i n i s t e r e d initially as an IV once daily agent and subsequently as a subcutaneous injection either once or twice daily. Patients were being transplanted either for non-Hodgkin's lymphoma (NHL) or Hodgkin's disease (HD). The PIXY321 was administered for a total of 21 doses or until an absolute neutrophil count of 500 for 3 days and platelet independence was obtained. C o m p a r e d to similar historical control patients treated at the same institutions with the same chemotherapy protocols, the median time to engraftment post t r a n s p l a n t was as follows: PIXY321 GM-CSF Placebo 500 ANC

All = 18 d HD = 22 d NHL = 16 d

19 d

26 d

Plt Ind

All = 23 d HD = 28 d NHL = 20 d

26 d

29 d

Compared to similar historical controls, the PIXY321 treated patients demonstrated similar neutrophil engraftment and an improved platelet independence time. This difference was most m a r k e d in the NHL patients. No clear dose response effect was demonstrated.

18

The Specific Role of Idarubicin during Induction Therapy of Childhood AML J. Ritter*, U. Creutzig, M. Zimmermann for the Cooperative Childhood AML-BFM Study Group University Children's Hospital Mtinster, FRG The combination of an anthracycline and cytosine arabinoside (Ara -C) is considered standard therapy for remission induction in acute myeloblastic leukemia (AML). Recently, Idarubicin (4demethoxydaunorubicin, IDA) has been found to be active in AML in adults. Furthermore, the therapeutic index of IDA may be higher than that of daunorubicin (DNR) regarding cardiotoxicity - a major problem, especially in children. Thus, one aim of study AML-BFM93 was to compare IDA with DNR during induction treatment of childhood AML in a randomized trial. From January 1, 1993 to May 31, 1994, 49 children with newly diagnosed AML received 3x t 2 m g l m 2 IDA together with Ara-C and etoposide (ALE) for remission induction treatment. All children received further intensification/consolidation chemotherapy followed by cranial irradiation and maintenance chemotherapy for a total treatment duration of t8 months. Thirty-one of 42 evaluable children had less than 5% blasts in the day 15 bone marrow. Eleven of 49 evaluable children developed a severe infection during bone marrow aplasia which was fatal in two. No severe liver or kidney toxicity (WHO Grade >2) was found in 49 evaluable children. The median time to recovery of the neutrophil count in 23 evaluable children was 26 days. Thirty-six children achieved a complete remission. These preliminary data of the ongoing study show that IDA can be used for remission induction treatment in childhood AML. So far, no significant differences have been found between the IDA and the DNR group concerning toxic deaths, nonresponse and remission rate. However, further follow-up of this study is necessary. *Presentaddress: Universit~ts-Kinderklinik,Albert-Schweitzer-Str.33, 48t29 MOnster,Germany

20

Idarubicin: Intracellular Mechanisms of Action

F. Gieseler, ~ Pfeifer, Th. Briden, M. Clark, K. Wilms Medizinische Poliklinik der Universitdt, Klinikstr.8, 97070 W~rzbur~, Background Anthracyclines induce DNA strand breaks through inhibition oftopo H. The cells react on this genotoxicity by trying to repair the strand breaks while topo II is inhibited and if the cell is sensitive, apoptosis occurs. We examined the time course of anthracycline-induced cell death in CCRF-celIs (human T-lymphoblastic leukemia cells) using daunorubicin and idarubicin. The aim is describe the intracellular pharmacokinetic of anthracyclines and the cellular reaction in order to find "critical points" which are responsible for sensitivity or resistance. Methods The following "checkpoints" have been used: cellular uptake of anthracyclines, nuclear uptake, DNA binding, DNA strand breaks, cell proliferation, viability. To compensate for the different genotoxicity and cytotoxicity, we used 1 I~g/ml ida and 5 ~tg/ml dauno. Results Both anthracyclines appear in the nucleus within minutes. We found, the optimum of DNA binding for ida 30 min after application, the one for dauno 45 min; the highest number of strand breaks for ida and danno within 1 - 3 hrs; proliferation inhibition after 20 - 30 hrs for both drugs. Discussion As in most hematopoetic cells, ida is about five times as potent as dauno in CCRF-CelIs. Differences between both drugs are their time to bind to DNA after arriving in the nucleus and the numbers of induced strand breaks. We found no significant differences in the time to arrive in the nucleus, the time to induce DNA strand breaks, the time to inhibit proliferation or the time to induce cell death. With the methods used, we can differentiate between the "induction phase" of drug induced cell death (end point DNA strand breaks), and the "reaction phase" (end point apoptosis). The induction phase lasts 3-4 hours, whereas the reaction phase takes 2-3 days. Consequences for chemotherapy are discussed.

IDARUBICIN

IN

INDUCTION

THERAPY

OF

AML.

Peter

H.

Wiernik, M.D., Albert Einstein Cancer Center, Bronx, NY USA All trials in which cytarabine + d a u n o r u b i c i n (A+D) were compared w i t h cytarabine + idarubicin (A+I) for induction of AML revealed significant advantages for A+I. Berman et al in a study of patients < age 60 r e p o r t e d a significantly greater CR rate and duration, and survival in patients treated with A+I compared with those t r e a t e d w i t h A+D, and a recent update demonstrates that those differences remain. A multic~nter study of similar design yielded the same significant outcome differences in patients < age 60, and those results still stand in a r e c e n t update. In a Southeast Oncology Group Study (SEG) of similar induction therapy design, the CR rate with A+I was significantly greater than that with A+D. That study incorporated a t e s t of late intensi-fication (LI) with A+I or A+D, depending upon induction therapy randomization. A recent update shows no difference in CR d u r a t i o n or survival based on induction therapy, but CR duration is s i g n i f i c a n t l y longer in patients who received LI compared with those who did not, and patients given A+I during induction and LI had significantly longer CR and survival than other study groups. G I M E M A studied A+I vs. A+D induction in elderly patients. No difference in CR rate or duration, or survival between the 2 groups emerged, but a s i g n i f i c a n t l y greater number of patients achieved CR w i t h 1 A+I course than with 1 A+D course. This was also observed in the Berman and USA m u l t i c e n t e r studies, but not in the SEG study. These studies d e m o n s t r a t e superiority of A+I over A+D as induction t h e r a p y for AML, especially in patients < 60 years. This clinical observation may be related to more favorable p h a r m a c o k i n e t i c s for I than D and less interaction of P - g l y c o p r o t e i n with I than with D.

A IO0

21

23

STATE OF THE ART AND THE CONTRIBUTION OF/DARUBICIN IN ACUTE MYELO/D LEUKEMIA Th. BOclmer*

PHILADELPHIA (PH) CHROMOSOME FREE BLOOD PROGENITOR C E L L S CAN N O W E A S I L Y BE C O L L E C T E D F R O M P A T I E N T S WITH CML NOT PRETREATED WITH ALPHA-INTERFERON (a-IFN). A.M. Carella, Hematology and ABMT Ospedale S. Martino, Genoa (Italy).

An overview of 14 multicenter trials (1) involving 7554 pts shows an overall CR rate of 63% and a rate &ongoing CR after 5 years of 21%. The trials randomly compared different drugs or strategies, and significant differences between alternatives compared were found in 7 trials. Thus, differences in the CR rate were in favor ofACR vs DNR as part of induction treatment (2), IDR vs DNR (3), and DNR full vs half dose (1). On the other hand, differences in CR duration were in favor of AraC infusion vs bolus in induction (4), AraC s.c. vs i.v. in maintenance (4), chemotherapy vs BCG in maintenance (5), maintenance vs no maintenance (6), and AmC high vs standard dose in postremission therapy (7). Improved cure rates of 35 to 40% in younger pts resulted fi'om regimens using high dose AraC in induction(1) or postremission treatment (2,7). /DR compared with DNR in 3 multicenter trials (3,8,9) produced superior CR rates in 2 trials (p = .08 and .03) with no longterm results published so far. Conclusion: Among anthracyelines different drugs and dosages may affect the CR rate with advantages for/DR at doses applied over intermediate dose DNR. For the definite cure ofpts high-dose AraC appears to play a leading role.

The presence of normal hematopoietic progenitor co-existing in the peripheral blood and m a r r o w of CML patients has led to the i m p l e m e n t a t i o n of our biological and clinical trials. The number of Ph-negative cells in the peripheral blood collected d u r i n g early phase of r e c o v e r y after a p l a s i a a p p e a r s to be r e l a t e d to the d u r a t i o n of a-IFN treatment and to the phase of the disease. Blood progenitor cells (BPCs) collected from untreated patients during early chronic phase contained a higher n u m b e r of l e u k e m i a - f r e e cells t h a n g r a f t s c o l l e c t e d f r o m p a t i e n t s who had r e c e i v e d p r i o r a-IFN treatment. Nine patients with CML ineligible for AIIoBMT, who had not received a-IFN therapy, were treated with an i n t e n s i v e chemotherapy regimen including idarubicin, a r a b i n o s y l c y t o s i n e and etoposide. The m e d i a n time from d i a g n o s i s to the mobilization therapy was 3 m o n t h s . Six p a t i e n t s r e c e i v e d H y d r o x y u r e a to r e d u c e W B C to <20xI0~ before mobilization and 3 p a t i e n t s received hydroxyurea for many months before mobilization. In 6 of 9 patients (66%) all metaphases of cells comprising the BPC graft were found to be Ph-negative. The best results in terms of isolat i o n of P h - n e g a t i v e b l o o d p r o g e n i t o r c e l l s w e r e achieved in the patients receiving this approach at diagnosis. The P h - n e g a t i v e cells c o l l e c t e d at diagnosis before a-IFN have been used as autotransplants in 4 patients. At present all patients are alive and their marrow is Ph-negative 3-12 months after reinfusion. All these 4 patients were given a-IFN after Autografting.

References: (1) Curr Opin Hematol

1:172, 1993; (2)Leukemia 5:510, 1991; (3) JCO 10:1103, 1992; (4) Blood 58:1203, 1981; (5) Blood 63:1039, 1984; (6) JCO 12:1583, 1985; (7) N Engl JMed 331:896, 1994; (8) E u r J Cancer 27:750, 1991; (9)Blood 79:313, 1992 *Present address: Department of Medicine/I-Iematology/Oncology, University of M0nster, Germany

22

24

IDARUBICLN, FLUDARABINE, CYTIDINE ARABINOSIDE (ARA-C), AND FILGRASTIM (G--CSF) (IDA-FLAG) LN T H E TREATMENT OF POOR PROGNOSIS AML : A PHASE II TRIAL H.T. Stelnmetz, V. Diehl and P.D. Wickramanayake. F'trst Dept. of Internal Medidne, University Hospital, Cologne, IBonn, FRG.

Antifungal Strategies in Cancer Patients S. Vardivarian, U.T.M.D. Anderson Cancer Center, Houston, TX 77030

A pharmacokinetically-directed sequence of Fludarabine and Ara-C in combination with simultaneously infused G-CSF (FLAG) has proven effective in poor prognosis AML. In order to improve response rate and duration we added Idarubidn and started a phase H trial. The Ida-FLAG regimen consisted of Idarubidn 8mg/m~ di,3,5, Fludarabine 25mg/ms lh-iv dl-5, Ara-C 1000mg/m' ql2h lh-iv dl-5 and Filgrastim 400g.g/m~ eontinoas inlMsion dfl until recovery of ANC > 1000/mcL During a 12 month periode 35 patients (pts.) ware included. 9 pts. (7 men/2 women) with a median age of 48y (22-68) suffered from AML refractory to TAD-HAM induction, g pts. (3m]Sw, median age $9y (30-75)) had an AML in first relapse after TAD-HAM. 18 pts. (12wd6w, median age 56y (45-67)) had a secondary A_ML, or signs or history of myelodysplasia at initial diagnosis (MDS-AML). In January 1995 response to the first course of Ida-FLAG was evaluable in 32 pts. CR waS achieved in 19 pts. (59%) overall, in 1/9 pts. with refractory AML, 7/8 with 1st. relapse of AML, and 11/15 pts. with secondary or MDS-AML. 7 pts. (22%) died within 28 days from sev~e infection. Ti! now 10/19 pts. with CR received a 2 ~, course of Ida-FLAG for consolidation. 3 of them died from sepsis within 28 days. Toxicity was evaluated in 25 pts. surviving the first course of Ida-FLAG. The median (range) of the'duration of G-CSF infusion wa~ 24(I (17-66); of WBC <3000/#I 20d (11..44); of WBC <500/~1 17d (6-31); of ANC <50~//,l 17d (9-31)! of platelets <50000/#1 2,4d (13-68), of days with fever >38.0~ 4d (1-33), af days with iv-antibiotics 16d ($.39), and of days in hospital 29d (19-68). 7 of 10 pts. survived the 2~. course. In 4 of 7 the time till recovery of haematopoiesis was longer than in the first course, in 3/7 it v,~asshorter. In conclusion the regimen has only moderate toxicity in AML patients with poor prognosis. The rate of opportunistic infec*ious did not Iner~tse. Ida-FLAG induces a high rate of complete remisslon In patients with relapsed, secondary, or MDS-AML

Abstract not received

A 101

27

25 ANTIFUNGAL STRATEGIES IN ACUTE LEUKEMIAS Fran~;oise Meunier, EORTC, Brussels, Bdgiura Invasive fungal infections are a major source of morbidity and mortality in patients with acute leukemias, particularly in those undergoing BMT since up to 20% of those patients suffer from lifethreatening mycoses. Any dramatic change in survival of patients with acute leukemias should result from a comprehensive clinical research approach aiming at global management of the underlying disease as well as of opportunistic fungal infection including optimal prevention, empiric management and effective antifuugal therapeutic strategies. Nuraerous challenging fungal pathogens have to be taken into account and there is not a single agent effective to prevent all fungi including C. albicans, C. krusei, Fusarinm, Aspergillus, etc... Good knowledge of local epidemiology is mandatory to define appropriate chemoprophylaxis. Flucenazole is effective against C. albicans but the dose to be recommended (50 to 400 mg daily) is a matter of controversies. Empiric antifungal management is crucial and there are now alternatives to conventional amphotericin B; recommendations for empiric treatment should take into account local cpidemiology, previous chenmprephylaxis and likelihood of candidosis: aspcrgfllosis or other mycoses. It is also difficult to establish when to initiate empiric treatment, it seems unnecessary to initiate it the first day of fever and it is worthwhile to wait for blood cultures results. New therapeutic strategies are currently available and/or under investigation; anaong those, lipid preparations of amphotericin B are safe, usually better tolerated than conventional anaphotericin B but there are still nnmerous questions concerning the optimal dose and duration of treatment. New antifungal compounds are also studied including a cyclodextrin formulation of itraconazole and other classes ofautifungals. Adjuvant antifungal treatment with growth factors or white blood cells transfusion after stimulation of donors are also currently investigated. All these strategies require careful evaluation, i.e. high quality clinical trials involving a large number of patients and quality assurance procedures. Within the fimnework of the EORTC, the EORTC Invasive Fungal Infections Cooperative Group is addressing those important issues of optimal management of fungal infections in cancer patients including those with acute leukemia.

Current Approaches in the Prevention of Invasive Aspergillosis J. Beyer, Dept. of Hematology, Free University Berlin, Germany Aspergillus species are increasingly recognized as major fungal pathogens in patients treated for leukemia or after bone marrow transplantation. As aspergilli conidia are abundant in non-filtered air, invasive pulmonary aspergillosis (IPA) characterized by hyphal invasion and destruction of pulmonary tissue, is the most common manifestation of an aspergillus infection followed by sinus involvement or catheter contamination. Dissemination from these initial ports of entry to other organs can be a secondary event in about 20% of cases or more. The reduction of environmental exposure and eady empiric antifungal treatment with amphotedcin B (araB) in patients in whom IPA is suspected, still remain the standards to reduce the incidence of and the mortality from this condition. However, several novel approaches for the topical and the systemic prophylaxis of IPA are under investigation. Aerosol delivery of amB avoids the toxicities of systemic araB administration and demonstrated prophylactic efficacy of IPA in animal studies as well as in clinical phase 1/11tdals. Yet, these results await confirmation in ongoing randomized studies. At present there are no other approaches for topical prophylaxis available. Systemic prophylaxis of IPA with low-dose intravenous amB desoxycholate was shown to be ineffective, but prophylactic conventional dose araB might prevent recurrence of IPA in patients with previous disease undergoing repetitive periods of neutrepenia. However, this approach is certainly too toxic to be used except in these selected patients with a very high dsk of IPA reactivation. New formulations of araB such as araB liposomes are better tolerated than arab desoxycholate, but so far only limited data are available on their use for IPA prophylaxis. Azole antifungal agents could be promising for the systemic prophylaxis of IPA and are less toxic than amB. However, neither fluconazol nor itraconazol have demonstrated prophylactic efficacy of IPA. Despite in vitro activity against aspergilli, itraconazole has an unpredictable absorption after oral administration and failures with this drug have been linked to low serum levels in one study. Yet, an oral solution of itraconazol as well as newer azols with improved activity against aspergilli and a better pharmacological profile are being investigated in clinical trials.

26

28

ANTIFUNGAL PROPHYLAXIS WITH FLUCONAZOL

FLUCONAZOLE (FCA) V E R S U S A M P H O T E R I C I N B/5-FLUCYTOSINE (ABF) FOR THE T R E A T M E N T OF F E B R I L E N E U T R O P E N I A G. Silling, W. Fegeler, N. Roos, R. S c h o m a k e r , M. Essink, T. B~chner. Dept H e m a t o l o g y / O n c o l o g y , U n i v e r s i t y of MHnster, G e r m a n y Fungal infections are a m a j o r cause of m o r b i d i t y in n e u t r o p e n i c patients. In o r d e r to i m p r o v e the outcome of pts. with hematological malignancies early empiric treatment strategies have been established. Antifungal treatment started on day 4 if fever of more than 3 8 . 5 ~ persisted despite i.v. a n t i b i o t i c s . In case of p u l m o n a r y i n f i l t r a t e s treatment started together with the antibiotic regimen. To c o m p a r e the e f f i c a c y pts. were r a n d o m l y assigned to f l u c o n a z o l e 5.7 m g / k g / d or A m p h o B 0.75 m g / k g / d and 5-FC 150 mg/kg/d. If fever or infiltrates persisted after one week, FCA was r e p l a c e d by ABF. 98 pts e n t e r e d the study, 49 in ~ach arm. N e u t r o p h i l c o u n t s w e r e less than 1000/cmm. U n d e r l y i n g d i s e a s e s in the FCA and ABF g r o u p s w e r e AML, first line treatment in 24 and 23 pts., r e f r a c t o r y or r e l a p s e d in 14 and 12 pts., r e l a p s e d ALL in 4 and 4 pts., MDS in I and 3 pts., CML in 0 and I pts. and l y m p h o m a in 6 and 6 pts.. 41 and 43 pts. suffered from fever of u n k n o w n origin, 8 and 6 pts. had p u l m o n a r y i n f i l t r a t e s w h e n fever occured. Response was a c h i e v e d in 29/49 FCA and 37/49 ABF. This d i f f e r e n c e was not s i g n i f i c a n t . 14 of 19 N o n - r e s p o n d e r to F C A r e s p o n d e d w h e n ABF was given. In this high risk g r o u p 19 pts. died, 6 F C A and 13 ABF. 6 and 9 deaths w e r e d u e to infections, 4 and 5 w e r e fungal infections and I and 4 pts. died from a s p e r g i l l o s i s . We c o n c l u d e that FCA and A B F seem to be e q u a l l y effective in i n f e c t i o n s in n e u t r o p e n i c pts. w i t h hematological malignancies. So the well t o l e r a b l e fluconazole should be a d m i n i s t e r e d early and in case of non-response it should be r e p l a c e d by A m p h o B and 5-FC after one week.

DURING INTENSIVE CHEMOTHERAPY FOR AML RELAPSE

O. Behre (I),Th. B~hner (2),and W. Hiddemann (!)forthe A M L Cooperative Group (I) Dept. of Hematology/Oncelogy,Universityof G0ttingen,Robctt-Koeh-Str.40, 37075 OOttingen;(2)Dept ofHerrmtology/Oncelogy,UniversityofMOnster;Germany

In a multicenter trial, 400 mg flucoanzole daily deea~ased the incidence of death due to deep mycoses in BMT recipients (10 of 177 vs. 1 of 179), most of whom had received allogeneic transplants (Winston et al., N Engl J Med 1992). All the protection afforded seemed to be in deep candidiasis, not aspergillosis. However, fluconazole has not resdted in decreased use of empiric amphotericin B, nor has an effect of overall survival been demonstrated. An increase in C. krusei infeetion occurred at one center but not at others. In most studies, flacouazolr prophylaxis has decrenscd the incidenee of mucoctmmr candidiasis, but this complication can also be diagnosed readily and treated when present. In contrast to BMT recipients, reduction in deep mycoses has not been convincingly demonstrated in patients with acute leukemia (Winston e~ al., Ann Intern Med 1993). Therefore, a randomized trial was initiated at our institutions to evaluate the efficacy of fluconazole prophylaxis in the reduction of the incidence of systemic fungul infections and the requirement of systemic amphotericin B therapy in patients with AML relapse at high risk for invasive mycoses. Patients with relapsed or refractory AML were randomly assigned to either fluconszole 400 rag/day orally administered in addition to antimicrobial prophylaxis with TMPPSMX 160 rag/800 mg PO tid, colistin sulfate capsules 2 mill U PO lid and amphotericin B suspension 240 nag PO qd or to antimicrobial prophylaxis alone.All patients received the sequential HD-AraC/mitoxantrone (S-HAM) + G-CSF salvage chemotherapy. In particular, patients with refractory AML and less than 60 years of age received 3.0 g/m2 AmC while older patients and cases with first relapses after 6 months remission duration were treated uniformly with 1.0 g/m2 AraC. Patients achieving a complete remission and not qualifying for BMT underwent a randomization for postremissinn therapy with interlenlda-2 or no further treatment. The fust interim-analysis of the ongoing trial is scheduled for February 1995.

A 102

29

81

CHROMOSOMALTRANSLOCATIONSPLAYA KEY ROLEIN LEUKEMOGENESIS, Janet D. Rowley, Department of Medicine and of Molecular Genetics and Cell Biology, University of Chicago, Chicago, Illinois.

HEMATOPOIETIC CYTOKINES, BCR/ABL, AND VIABILITY SIGNALS. J.D. Griffin. Dana-Father Cancer Institute, Boston, MA. In vitro, normal hematopoietic cells or factor-dependent cell lines will undergo apoptosis if deprived of growth factors for more than a few hours. In addition to hematopoietic cytokines such as GbI-CSF or IL-3, oncogenes such as BCR/ABL can provide viability signals. The signaling pathways involved in maintaining viability are largely unknown. To understand how normal myeloid cells stay alive, we looked for viability domains in the GM-CSF receptor. Using chimeric receptors, we found that the cytoplasmic domain of the I] chain, but not the ~ chain, was required for viability. A series of deletion mutants of the 13 chain were constructed and expressed in Ba/F3 cells. Two viability domains were identified, a major domain encompassing aa 626-763, and a secondary domain between 517 and 544. When the I~ chain was truncated at aa626, GM-CSF could no longer support viability in serum-free medium. A major tyrosine phosphorylation site was detected at aa 750, and mutation to phenylalanine also resulted in a receptor which was defective in maintaining viability. Both the 626 deletion and the 750Y-F mutants were found to be able to activate JAK2, STATs, and most known signaling pathways. However, tyrosine phosphorylation of Shc was defective. She is a widely expressed signaling molecule which contains an SH2 domain, a binding site for GRB2-SH2 , and a proline rich domain. Shc has been linked to activation of p21 ras. Our results implicate She in a pathway that leads to viability signals, perhaps also involving ras. We compared the viability signals from BCR/ABL with those from the GM-CSFR. Cell lines were constructed which expressed temperature-sensitive mutants of BCR/ABL. At the non-permissive temperature (38~ the cells are wild-type and dependent on IL-3 for proliferation. At the permissive temperature (33~ the BCR/ABL tyrosine kinase is activated and the cells no longer need IL-3 for viability. Shc was found to be a major substrate of BCR/ABL and to form a complex with BCR/ABL and several other signaling molecules. These results indicate that BCR/ABL can provide a viability signal for myeloid cells that will replace that normally supplied by cytokines such as IL-3 or GM-CSF. Activation of Shc and the p2 lras pathway is indirectly implicated in this process.

Chromosome translocations have been most closely associated with human leukemias and lymphemas. Recently, they have been shown to occur in sarcomas as well. In each type of tumor, the translocations are r e l a t i v e l y specifically associated with particular subtypes of these tumors. Cloning the translocation junctions has identified the genes affected by the breakpoints in these tumors. Well over three dozen new genes have been identified in this process. Someof these genes are not normally active in hematopoietic cells. Although the genes that are involved participate in a number of steps in the complex pathway of transmitting growth regulatory signals from the cell surface to the nucleus, most of those identified in the acute leukemias and lymphomasact as transcriptional activators. That is they are DNA binding proteins that d i r e c t l y regulate the level of transcription of the target genes. Someof the genes recently cloned from breakpoint junctions in sarcomas, also act to regulate transcription. I t has been recognized for some time that chromosome abnormalities, especially the translocations, have prognostic significance in the acute leukemias and lymphomas. Cloning of translocation breakpoints provides DNA probes for FISH analysis as well as for Southern blot and RT-PCR analysis of patient samples, some of which have either an apparently normal karyotype or are inadequate for cytegenetic study. These probes w i l l permit much more accurate genotypephenotype correlations than have been possible in the past. The challenge for the future is to understand the mechanisms leading to recurring translocations and to understand the tumor-specificity of the translocations. This increasing insight into the biology of these malignant diseases should lead to more accurate diagnosis, to improved therapy, and possibly to more effective prevention.

80

HEMATOPOIETIC G R O W T H FACTORS AND THEIR RECEPTORS IN A C U T E LEUKEMIA. Ivo P. Touw, Daniel den Hoed Cancer Center and Erasmus University, Rotterdam, The Netherlands. The application of in vitro culture assays has contributed significantly to our understanding of the growth characteristics of human acute leukemia cells. In this presentation, some of the most prominent features of hematopoietic growth factor (HGF) responses of acute leukemia cells, including autocrine mechanisms of growth, will first be summarized. Subsequently, investigations dealing with the function of the receptor for gmnulocyte colony-stimulating factor (G-CSF-R) in two pathological conditions characterized by defective myeloid maturation, A M L and severe congenital neutropenia (a disease that may progress towards AML) will be discussed. Collectively, these studies serve as examples of how the analysis of HGF responses and HGF receptor function provides new insights into the heterogeneous pathobiology and clinical behaviour of acute leukemia and related diseases.

A CLONING STRATEGY TO IDENTIFY GENES THAT FUNCTIONALLY INTERACT WITH BCR-ABL FOR TRANSFORMATION.

Stefan Maidmann 1, Daniel E. H, Afart, Jami McLaughlin3, Andrei Goga2 and Owen N. Wittr 3. Department of Microbiology and Molecular Geneticst, Molecular Biology Institute2, and Howard Hughes Medical Institute, University of California-Los Angeles, Los Angeles, CA 90024-1662, USA. We have developed an expression cloning system to identify genes which functionally interact with Bcr-Abl for transformation. This system is based on a complementation strategy we used recently (1) to define the signaling pathways activated by the Bcr-Abl tyrosine kinase. Point mutations were generated in domains of Bcr-Abl that may play a role in connecting its tyresina kinase signal todownstream effector molecules. Single point mutations in the Sre-homology (SH2) domain, the major tyrosine autophosphotylation site of the kinase domain, and the Grb-2 binding site in the Bcr region impaired the transformation of fibrablasts by Bcr-Abl. Hyperexpression of c-Myc, a gene known to act downstream of a Bcr-Abl signal, was able to only rescue the SH2 mutant for transformation. The ability of c-Myc to differentially complement the Bcr-Abl point mutants signified that at least two different signaling pathways are generated by Bcr-Abl. Based on the observation that c-Myc complements the SILl mutant, we developed an expression cloning strategy to identify unknown downstream genes in the complex signaling network of Bcr-Abk A lymphoid progenitor eDNA expression librmy was constructed in a retroviral vector (2) and introduced into fibroblasts, stably expressing a Bcr-Abl point mutant, eDNA clones, capable of complementing a mutant for tmm~ormation should result in colony formation in a soil agar assay. Thus far, we have obtained numerous colonies in several assays. Additional steps am necessary to isolate a complementing eDNA from a colony in soft agar. These include: Rescue of the complementing vin~ by a helper virus, identification of the complementing eDNA based on a PCR strategy, subelonlng and sequencing. This cloning strategy should be useful to identify numerous genes that functionally interact with Bcr-Abl in distinct signaling pathways. Literature: Afar, D. E. H. Goga, A. McLaughlin, J., Witte, O.N. and Sawyers, C. L. (1994) Science 264, 424-426. Rayner, J. 1L and Gonda, T. J. (1994) MCB 14, 880-887.

A 103 3,3

35

BIOLOGY AND DETECTION OF t(4; 11). F:Griesinger. Department of Internal Medicine,University of Gtttingen, Germany. The elucidation of the genetic basis of various entities of acute leukemias has allowed to gain insight into the physiologic function in lymphohematopoiesis and the transforming potential of genes involved in malignant transformation. Furthermore, specific genetic alterations/chromosomal translocations are associated with distinct biological and prognostic characteristics, and serve as criteria for stratification to risk adapted therapy in ALL. The t(4;11) leukemias are characterized by a unique epidemiology and specific biologic features. t(4; 11) pre-pre-B-ALL has an incidence of 50 to 60% in infant ALL with a preponderance in females, and an ineidenoe of 5-10% in pediatric and adult ALL. Immunophenotypically, this translocation is almost exclusively associated with a pre-t~re-B-ALL phenotype and coexpression of the myeloid differentiation antigens CD15 and CDw65. Induction of differentiation into the myeloid lineage has been described, and it has been hypothesized that a common lymphoid and myeloid progenitor is transformed in t(4; 11) pre-pre-B-ALL. Prognosis in infants is dismal with chemotherapy alone. With intensification of post-remission chemotherapy, prognosis may be more favorable in adults. HRX (ALL-I, Htrx, MLL) on I lq23, a potential DNA-binding protein with homology to Drosophila trithorax, and FEL (AF-4) on 4q21, a gene encoding a serine/proline-rich protein of unknown function, are fused to each other by the 4;11 translocation. A variety of different fusion transcripts have been described, in all of which all or most of the zinc finger domains of HRX are deleted, while potential DNA transactivating domains of HRX are preserved. Whether different fusion transcripts are associated with different biology and age groups is still under investigation. The molecular similarity of t(4;ll) pre-pre-B-ALL to secondary epipodophyllotoxinassociated leukemias with translocations of I 1q23 to a variety of different fusion partners and the identification of topoisomerase II binding sites in the vicinity of 1 lq23 breakpoints makes this leukemia an attractive model to investigate mechanisms of leukemogenesis.

DEXAMETHASONE AND ANTHRACYCLINE, S FOR CHILDHOOD ACUTE LYMPHOBLASTIC LEUKEMIA (ALL). WHERE ARE WE? W.A. Kamps, Beatrix Children's Hospital, Univ Hosp, Postbox 30001, 9700 RB Groningen, The Netherlands

04

36

RISK ADAPTED TREATMENT IN CHILDHOOD ALL: DATA FROM THE BFM GROUP. Riehm, H. t, Schrappe, M. x, Ludwig, W.-D. 2, Harbott, J.a,. Reiter, A. 1. Hannover t, Berlin2, Giet3en3, FRG.

MECHANISMS OF TREATMENT FAILURE IN C H I L D H O O D ACUTE LYMPHOBLASTIC (ALL) LEUKEMIA: CHILDRENS CANCER GROUP (CCG) INITIATIVES: Paul S. Gaynon, Bruce C. Bostrom, Gregory H. Reaman, H a r l a n d N. Sather, Michael E. Trigg, David G. Tubergen, and Fatih M. Uckun for the Childrens Cancer Group, Arcadia, CA, USA

Experience with approx. 3700 pts in the last four multicenter trials for childhood acute lymphoblastic leukemia (ALL) performed over the last 13 years by the BFM study group established and reconfirmed basic knowledge on risk of relapse. A series of randomized and prospective questions, however, allowed new insights into additional biological and therapy-related thctors. In trial ALL-BFM 83, evidence was generated on the importance of delayed intensification even tbr standard risk ALL that was later confirmed in the following trial ALL-BFM 86 and other trials. Pts with standard risk ALl, treated without reinduction therapy had an higher risk of relapse than pts stratified into the medium risk group. From 1981 to 1986, the ir~pact of adequate maintenance therapy with 6-MP sad MTX on event-free survival was evaluated by randomization of 18 vs. 24 months total therapy duration proving the impact of prolonged exposure on disease-free survival. Also, in trial ALL-BFM 83 the prognostic significance of blast cell reduction as a parameter for early response to induction therapy with prednisone and MTX i t. was prospectively evalumed. Thus, a new independent risk factor called the 'prednisone poor response' (PRED-PR) could be generated that identified approx. 10% of all pts with ALL. Prognosis for this group, however, has not yet improved since being discovered in trial ALL-BFM 83, despite various attempts of treatment intensification. Pts in this group are predominantly boys, have an high initial leukemic cell burden, present with T-cell ALL in 40%, and have an increased risk for nonresponse to therapy. The probability of event-free survival (pEFS) in the group is less than 50%. Among these pts with PKED-PK a subgroup can be identified that is at even higher risk of relapse: with coexpression of myeloid markers on their blast population, with T-ALL, or with very high WBC. Transloeation t(9;22), the equivalent BCKABL recombination, or t(4;ll) are together with nonrespoase to primary trcatment additional high risk features that qualify any pt for maximum therapy including allogeneic BMT. Despite difficulties for high risk pts the overaU result of 75-80% pEFS is encouraging.

Prednisone (PDN) has a long and invaluable tradition in ALL therapy and BFM studies have proved the good correlation of in vivo response of ALL cells to PDN and risk of treatment failure. Substitution of dexamethasone (DXM) for PDN became popular after Jones reported excellent CNS prophylaxis using DXM in place of PDN. To improve EFS and abandon cranial irradiation the DCLSG ALL-6 study used DXM in place of PDN resulting in > 20 % improvement of EFS over historical controls. Higher anti-leukemic activity of DXM over PDN is seen in MTT in vitro assays, and lower protein-binding of DXM versus PDN results in better cerebrospinal fluid drug levels, both providing a rationale for substitution. Results from the recent DFCI trial (3 dose levels DXM versus PDN) and the CCG trial (DXM versus PDN in induction and maintenance) may justify more use of DXM and this requires monitoring for steroid side effects like behavioral problems, obesity, avascular bone necrosis, and linear growth retardation. Anthraeyclines added as induction agent can hardly improve the over 95 % complete remission rates that are obtained in patients using vincristine, PDN and asparaginase. Better EFS may result, but old trials like DCLSG ALL-5 only have only shown slight survival benefit with daunorubicin added as fourth agent to induction of standard risk patients. Many groups use anthracyclines for (re)induction without having evaluated on a single drug base the additive value. With cardiotoxicity being the major obstacle to use anthracyclines we should evaluate alternative administration ways and potentially less toxic analogs but also the overall need of anthracyclines especially in standard risk patients.

CCG h a s c o n d u c t e d clinical trials for c h i l d r e n w i t h n e w l y diagnosed ALL since 1968 and enrolled more than 12,000 children. M o r e t h a n 1,000 c h i l d r e n w e r e e n r o l l e d in 1994 alone. Preliminary analyses show a 3-year event free survival of 81% for the current concluding CCG-1800 series and successive statistically significant 25% reductions in relative risk of failure over the last two series of studies. Several themes m a y be derived from this experience. Early response to therapy is a consistent d e t e r m i n a n t of outcome. Effective intensification of systemic t h e r a p y a f t e r induction reduces the likelihood of m a r r o w a n d extramedullary r e l a p s e for p a t i e n t s w i t h e i t h e r f a v o r a b l e or u n f a v o r a b l e presenting features. Effective presymptomatic CNS therapy m a y be p r o v i d e d to all infants a n d most c h i l d r e n w i t h o u t cranial irradiation. However, 18 Gy cramal irradiation may btili be useful for children with WBC > 100,000/tll and those older than 10 years. Recent o b s e r v a t i o n s p o i n t to p r i m a r y disease resistance to c h e m o t h e r a p y and unfavorable host thiopurine p h a r m a c o l o g y as c o m m o n m e c h a n i s m s of t r e a t m e n t failure. R e s i s t a n c e to chemotherapy m a y be linked to resistance to apoptosis. Patients with primary disease resistance m a y be identified t h r o u g h in vitro assays or poor early m a r r o w response a n d a s s i g n e d to m o r e prolonged, more rigorous intensification. We propose to e m p l o y laboratory assays of e n d - i n d u c t i o n leukemic b u r d e n to measure the specific contribution anti-CD19-pokeweed antiviral p r o t e i n conjugate in the context of complex therapy in k n o w n resistant disease. Failure to achieve a d e q u a t e intracellular levels of thiopurine metabolites m a y possibly be redressed by parenteral administration a n d / o r substitution of 6-TG for 6-MP. The reasons for t r e a t m e n t failure in cases of P h i l a d e l p h i a c h r o m o s o m e positive ALL remain to be elucidated.

A 104

37

89

RISK-ADAPTED TREATMENT FOR ALL - ST. J U D E FINDINGS Ching-Hon Pui, Gaston K. Rivera, John T. Sandlund, Raul C. Ribeiro, James M. Boyett, William E. Evans, William M. Crist. St. Jude Children's Research Hospital and The University of Tennessee, Memphis, Tennessee, USA. Current approaches to therapy for acute lymphoblastic leukemia (ALL) are based on stringent risk assessment at diagnosis, so that only those patients at high risk of relapse are treated aggressively, with less toxic regimens reserved for lower-risk patients. In the St. Jude Total Therapy Study XIII for childhood ALL (1991 to 1994), cases were classified as "higher risk" if they had one or more of the following features at diagnosis: age <1 or >10 years, leukocyte count >25 X 109/L, DNA index <1.16 or >1.60, CNS-3 status (>5 leukoeytes/#L of cerebrospinal fluid with leukemic blasts apparent in cytocentrifuged preparations or the presence of cranial nerve palsy), and lymphoblasts containing the Philadelphia chromosome or the t(1;19) in association with a pre-B phenotype. With intensive early treatment, addition of a reinduction/consolidation phase and rotational use of pairs of effective antileukemic drugs during the continuation phase, 137 of 154 high-risk patients continue to be free of adverse events. It has become apparent in recent analyses that cases with DNA indices from >1.16 to <1.60 have an excellent treatment outcome with antimetabolite- based therapy, irrespective of age and WBC, and that early treatment responses to chemotherapy, certain molecular abnormalities and CNS status have particular prognostic significance (data to be presented). Thus, in our current Study XIIIB, we have defined lower-risk ALL as a disease with a DNA index >1.16 and <1.60 or age 1 to 9 years with WBC <50 X 109/L, provided that the patients lack CNS-3 status, testicular involvement, T-cell phenotype, eytogenetic or molecular evidence for the t(9;22), t(4;ll) or t(1;19) with a pre-B phenotype or more than 5% bone marrow blast cells on day 15 of remission induction. With this definition, 50% to 55% of cases would be classified as having lower-risk ALL and from experience in Study XI, we project a long-term event-free survival of 90% or more for this subgroup.

RISK ADAPTED TREATMENT (GM-ALL RESULTS) D. Hoelzer, R. Arnold, Th. B~chner, M. Freund, W. Gassmann, N. G~kbuget, W. Hiddemann, P. Koch, H. L~ffler, W.-D. Ludwig, G. Maschmeyer, E. Thiel, B. V~Ikers, D. Messerer for the German Adult ALL Study Group, Dept. of Hematology, Frankfurt, Germany Within B-lineage ALL significant progress in LFS rates has been made for B-ALL and now, apparently, also for pre-pre-B-ALL. In B-ALL, which formerly had a very poor survival rate of <10%, recent use of adapted childhood regimens with short intensive cycles including high doses of cyclophosphamide or ifosphamide, methotrexate and cytosine arabinoside (HD-AraC), has increased the CR rate to 77% and LFS to 58% in the German adult ALL studies. The outcome for pre-pre-B-ALL which was also associated with a poor prognosis, is apparently improved. With intensified protocols (GM-ALL 04/89) LFS rates are 60%, including those patients with a t(4;ll) translocation. The most frequent B-lineage subtype, common ALL (57%), has however not improved and the LFS in most larger adult ALL series is still 30%. This may be partly due to the poor prognostic Ph/BCR-ABL positive ALL which constitutes 40% of this subtype. But also in Ph/BCR-ABL negative common ALL the outcome remains unsatisfactory with a continuous relapse rate up to 5 or even 7 years. Thus it seems that common ALL patients have not profited from intensification with HD-AraC combinations tested in several trials. The B-lineage associated Ph/BCR-ABL positive ALL, constituting 20% - 25% of the total adult ALL population, still has a poor LFS between 0 and 15%. Intensification of induction increases the CR rates to 70% - 80%. PCR analysis in those patients reveals that the majority are still BCR-ABL positive. Intensification of chemotherapy, BMT, allogeneic, autologous with purging or MUD, are recommended, and substitution of conventional ineffective maintenance by IL-2 and/or s-interferon should be explored.

88

40

Risk Adapted Therapy for ALL: The Pediatric Oncology Group (POG) Experience. B Camitta, V Land, S Lauer, D Mahoney, J Pullen, J Shuster. POG, Chicago, IL, USA. Prognosis in childhood ALL depends upon a complex interaction of patient and leukemia cell characteristics with treatment. Although age and WBC are important predictors of outcome in most studies, they have been used in multiple ways to define risk groups. This hinders comparison of results between centers. At a National Cancer Institute (NCI) sponsored meeting of major USA Centers the following consensus risk group definitions were adopted: Lower Risk (age 1.00-9.99 and WBC <50,000), Higher Risk (age >10.00 or WBC >50,000). Other data to be collected for prognostic analyses include: DNA index, cytogenetics, immunophenotype, CNS status at diagnosis, and early response to therapy (day 8 or day 15) POG has previously shown that leukemia cell DNA content (DNA Index, DI) is an important additional prognostic factor in B-precursor ALL. Using the same patients (ALinC14; POG 8602) we have shown that patients with trisomies of both chromosomes 4 and Event-Free Survival on P O G 8602 t0 (43/103 +) have a highly favorable vs DI/Cytogenetics 4 yr EFS. Although both DI and tdsornies 4/10 are significant predicD i c s ~ >1 16 <1.16 Genet " tors of outcome, (DI > I. 16 vs < 1.16 adjusted for 43/103, p=.0059; 43/103 + 43/103 + vs - adjusted for DI, p= 90% 87% .0077) when informative cytogenetics is available trisomies of 4/10 is more 43/103 important (see Table). Preliminary 78% 60% evaluation ofALinC 15 (POG 9005/ 9006) has confirmed these data. In ALinCl6 the consensus risk groups, Normal or 91% 76% DI and cytogenetics are used to idenUnknown tify patients who receive different intensities of treatment. It is our expectation that adapting intensity of therapy to risk of relapse can maximize cure while minimizing short and long-term toxicities.

TREATMENT OF BIOLOGICALLY DEFINED SUBSETS OF ACUTE LYMPHOBLASTIC LEUKEMIA. R.A. Larson for the Cancer and Leukemia Group B (CALGB). Chicago, IL USA In a phase II multicenter clinical trial, we treated 197 newly diagnosed patients (pts) with ALL (ages, 16"-80 years (yrs); median, 32) using cyclophosphamide, daunorubicin, vincristine, prednisone, and L-asparaginase: 167 pts (85%) achieved a complete remission (CR), 13 (7%) had refractory disease, and 17 (9%) died during induction. A higher CR rate was observed in younger pts (94% for those < 30 yrs old, 85% for those 30-59 yrs old, and 39% for those > 6 0 yrs old, p<0.001 ), and in those who had a mediastinal mass (100%) or blasts with a T-cell immunophenotype. Eighty percent of B-lineage and 97% of T-cell ALL pts achieved a CR (p =0.01 ). The co-expression of myeleid antigens did not affect the response rate or duration. Seventy percent of those with cytogenetic or molecular evidence of the Philadelphia (Ph) chromosome and 84% of those without such evidence achieved a CR (p=0.11). Pts in remission received multi-agent consolidation treatment, central nervous system prophylaxis, late intensification, and maintenance chemotherapy for a total of 24 months (mos). After a median follow-up time of 43 mos, the median survival for all 197 pts is 36 mos; the median remission duration for the 167 CR pts is 29 mos. Favorable pretreatment characteristics relative to remission duration or survival are younger age, the presence of a mediastinal mass or lymphadenopathy, WBC <30,O00//tl, L1 morphology, T or TMy immunophenotype, and the absence of the Ph chromosome. The estimates of the proportion surviving at 3 yrs are 69% for pts < 3 0 yrs old, 39% for those 30-59 yrs old, 89% for those who had a mediastinal mass, 59% with WBC < 30,O00/pl, 63% with L1 morphology, 69% for T or TMy antigen expression, and 62% for those who lack the Ph chromosome. Fifteen pts (8%) had no unfavorable prognostic factors and have an estimated probability of survival at 5 yrs of 100% (95% confidence interval, 77-100%). This intensive chemotherapy regimen produces a high remission rate and a high proportion of durable remissions in adults with ALL.

A 105

41

43

ALL IN ELDERLY: THE GIMEMA TRIALS F. Mandelli, L. Annino, A. Ferrari for the GIMEMA Group (Italy).

Bone Marrow Transplantation vs Chemotherapy in ALL. Joint Trial MRC UKALL-12 with Eastern Cooperative Oncology Group (ECOG) E2993.

In the past GIMEMA ALL 0183 trial no specific arm was designed for pts. > 60 yrs.; 12 pts. were enrolled, of them 8 achieved CR,2 died in induction (I.D.) and 2 were refractory. At l0 yrs follow-up, out of 8 CRs 1 pt was alive in 1st CR. The subsequent GIMEMA ALL 0288 trial (Mandelli et al.,1992) envisages for pts>60 yrs a specific ann consisting in a non-randomized 4-drug (PDN,VCR,DNR, ASP) induction, followed in the post-CR phase by intensification + 2 yrs conventional maintermace. This study, activated in January 1988 is still open .As of September 1994, 80 pts were enrolled-32 were males,median age was 64.9 yrs (range 60.1-78.2)-;median WBC count was 11.5x109/1 (range 0.4-290).According to FAB classification,25 pts were classified as L1, 51 as L2,2 as L3 , 2 were not evalaable. As to immunophenotype,56 exhibited B-lineage, 15 T-lineage, 1 mixed lenkemia,1 AUL,7 were not evaluable.Cytogenetics was not performed routinarly: only 28 eases were studied. No metaphases were found in 10; of the 18 in whom karyotype analysis was available, 6 were normal, 5 showed aspecific abnormality, 7 resulted Phi+. 77 out of 80 pts were evaluable for induction response: 36 (47%) achieved CR,32 w'..ere I.D., 9 were refractory . As the 7 Phi+ pts, 4 achieved CR, 2 were refractory and 1 died during induction.The main I.D. cause was infection,4 pts only experienced fatal cardiac toxicity. Among the 36 CRs, 10 (28%) went off-study early because of chemotherapy related toxicity (8) or protocol violation (2);3 died in CR, 4 relapsed, 8 were on intensification therapy,finally 11 were given maintenance therapy. At time of this analysis, at 2 yrs,19% of pts are projected to be survivors and 13% are projected to be EFSs. In univariate as well as in multivariate analysis age>65 yrs,PDN-pretreatment response, WBC count >50x109/1 and immunopbenotype were tested, in univariate attalysis age revealed to be the only prognostic factor influencing significantly the achievement of CR (59% vs. 34%;p=0.02). in multivariate analysis age confirms to be the only indipendent prognostic factor on EFS.

A.I-[ Goldstone, H.M. Lazarus, I.M. Franklin, S. Richards, J.W. Anderson, J.M. Rowe

42

44

Results of different post remission treaanents in adult acute lymphoblastic leukemia (ALL) : chemotherapy, allogenic (allo) or autologous (auto) bone marrow transplantation : basis of a more accurate therapeutic stratification. D. Fibre, C. Sebban, JP Vcmant, F. Rigal Huguet, J. Rciffers, E. Lcpage, L. Sutton, X. Thomas. HOpital Edouard Herriot 69437 LYON CEDEX 03 France

APPROACHES FOR DETECTION OF MRD

In the prospective study of adult ALL incleding 572 patients, complete remission (CR) was achieved after a first chemotherapy course comparing two anthracyclin containing regimens, :i: a salvage therapy course, if CR was not achieved at day 28. Patients in CR were treated according to age and results of HLA typing. Allo BMT was scheduled for patients 15-40 with an identical sibling donor. Patients over 50 received chemotherapy with consolidation (C) and maintenance (M). Others were randomized after C between (auto) BMT and M. CR rate (76 %) did not differ between the two induction regimens. Death during induction increased with age and P h l + ALL got a lower CR rate. Allo BMT group was compared to a control group of patients 15-40 with at least one sibling but without HLA matched donor. Outcome of allo BMT was not significantly superior to those of control group in term of disease rice survival (DFS) antJ survival. If all patients were. stratified according to high risk criteria of relapse, in the high risk sa'atified group allo BMT had superior DFS and survival than control but, in standard risk, allo BMT had the same outcome than in control. Results of auto BMT were not significantly better than chemotherapy but the number of late relaspos (9 versus 24) was inferior and the duration of therapy was shorter. Results have also been analyzed according to the T or B lineage of ALL cells, in patients up to 40. Chemotherapy was efficient in T ALL and the difference with B lineage ALL was highfly significant with 3 years D.F.S. of 59 versus 20 % (p = 0.002). For auto or allo BMT, B or T lineage are not a prognostic factor. So standard risk ALL or T ALL will be scheduled to ~.eive chemotherapy for post remission treatment with allo BMT only proposed in CR2. In high risk ALL and in P h l + ALL intensive myelo ablative therapy (aLlo or auto) is proposed in first CR to improve outcome.

We have initiated a randomized study of BMT in adults aged 15-55 years with ALL. Those with a matched sibling will have allogeneic transplantation and this will be compared with a randomisation between either myeloablative chemotherapy with VP16, TBI and autolognus bone marrow transplantation or intensive consolidation and maintenance chemotherapy. Patients with the Philadelphia chromosome are eligible to receive an allogeneneic bone marrow transplant from a MUD donor and to receive alpha-interferon during maintenance. By October 1994 171 patients had been registered in the MRC arm and 56 in the ECOG arm (total 227). Of known characteristics at diagnosis 129 were Ph negative and 24 Ph positive and male:female is 125:78. Remission rates are available for the first 197 patients. O f these 179/197 (90 %) remitted and 99 of these were in remission by day 21. Of the 32 known not to be in remission by day 21, 17 remitted by day 28. In the MRC group the number of deaths in induction remains lower than the previous adult trial UKALL-10 despite the increased intensity of the protocol. The causes of infection in induction in UKALL-12 have been mainly bacterial but there have been significant numbers due to pneumocystis and fungi. Out of 11 initial MRC patients with pneumocystis, 4 have died. Cerebral vein thrombosis has occurred in 2 trial patients and several pilot patients treated on the same protocol before the randomised study opened. Detailed coagulation monitoring is taking place during the periods that the patient is receiving L-asparaginase.

J.F.San Miguel. Servicio de Hematologia.HosrYitalUniversita6o, Salamanca. Patients with acute leukemia is complete

remission are currently

indiscriminately subjected to consolidation treatments including bone marrow transplants, in order to eliminate residual leukemia cells below the detection level of conventional morphology (<5%) -minimal residual desease (MRD). Therefore more sensitive techniques are needed to evaluate the effectiveness of treatment that will permit a more precise evaluation of the tumor mass and predict impeding relapses prior to clinical manifestation.The efficacy of the different techniques available for the detection of MRD are based on : 1) Specificity (to discdminata malignant from normal cells, without false negative and positive results) ; 2) Sensitivity (detection limit) and 3) Reprodectibility and applicabitlty (easy standardization and speed in collecting results for their clinical application). Techniques for MRD can be classified depending on the cellular structure i0en~ed: a) overall charectedstics of the neoptashc cells, such as the morphology and the "in v'~'o" cotony growth, b) cytogenetic characteristics, evaluated either by conventional chromosomal

analysis or fluorescence in s~tu hybridization

(FISH), c) the antigenic immunophenotype of blast cells assessed by muifiparametrie flow cytometry, d) DNA content-aneuploidy- detected by flow cytometry, and e) DNA structure, analyzed either by Southern blot or petymerase chain reaction (PCR).At present the two most sensitive techniques are : i) The use of double and tnple antigenic markers analyzed at flow cytometry which permits the detection of "uncommon" or "aberrant" phenotypes at a level ranging from 10-3 to 10-5 and ii) the PCR technique based on the detection of Junctional regions of rearranged Igf'l'CR genes (at DNA level) and fusion regions of chromosome aberrations (at mRNNcDNA level), with a detection limit of 10-5 or 10.6 .

A106 45

47

MOLECULAR BIOLOGICAL DETECTION OF MINIMAL RESIDUAL DISEASE IN ACUTE LYMPHOBLASTIC LEUKEMIA

MOLECULAR BIOLOGICAL D ~ O N DISEASE IN AML

Claus R. Bartram, Taku Seriu, Thomas E. Hansen-Hagge, Johannes W.G. Janssen, Wolf-Dieter Ludwig 1.

A.Biondi*, D. Diverio~ Luciano*, and F. Lo Coco~ *Cliniea Pediatrica Universit~ Milano, H.S. Gerardo, Monza; o Ematologia, Dipt. Biopatologia Universi~ La Sapienza, Roma.

Section of Molecular Biology, Department of Pediatrics II, University of UIm, 1Department of Medical Oncology and Applied Molecular Biotogy, Robert-RSssle Cancer Center, Free University of Berlin, FRG. During the last 5 years numerous PCR studies have determined the remission status of ALL patients. These mostly retrospective analyses indicate that 1) molecular markers allow the monitoring of almost every patient, 2) in particular clonospecific lg and TCR probes can be generated for 90% of cases, 3) oncogene recombinations constitute a complementary, steadily expanding marker category, 4) (semi}quantitative serial PCR analyses appear mandatory for clinically relevant predictions, 5) all methods have limitations so that the concurrent application of independent strateqie.~ is recommended, and 6) the MRD status depends very much on both the biologically/genetically defined ALL subtype as well as therapeutic modalities. It became also clear, however, that data from large prospective PCR studies have to be awaited before individualized protocols can seriously be considered for the treatment of ALL patients. Along this line we will discuss recent data from our laboratory particularly in the context of the European ALL-BFM Study Group.

OF MINIMAL RESIDUAL

The role of molecular detection of minimal residual disease (MILD) in different form of acute myeloid leukemia (AML) is currently being explored. PCR methods that can amplify a unique nucleotide sequence from 1 in 105 ceils can detect residual disease with a sensitivity that far exceeds standard light microscopy and routine cytogenetics. In acute promyeloeytic leukemia (APL) appropriate oligoprimers complementary to PML and RAR-[x sequences nearby the DNA breakpoints have been successfully used in PCR experiments to amplify the PML/RAR-a hybrid gene and sensitively detect MRD. Retrospective analysis of APL patients in different series, have indicated the prognostic relevance of PCR evaluation of the PML/RAR-a rearrangement, as a predictor of impeding relapse in APE Moreover in APL patients during long term remission (4 to 5 years) RT PCR analysis revealed the absence of PML-RAR-a fusion transcripts. Data of RT-PCR monitoring analysis of APL patients in a large longitudinal prospective Italian study will be presented. By comparison the recent findings on the persistence of the t(8;21) transloeation in patients with AML in long-term remission will be discussed for the different implications of the detection of transloeationcarrying cells in acute myeloid leukemia.

46

48

IMMUNOPHENOTYPIC DETECTION OF MINIMAL RESIDUAL DISEASE (MRD) IN ACUTE LYMPHOBLASTICLEUKEMIA (ALL) J.Ciudad, J.F.San Miguel1, M.C.Lopez-Bergasl, B.Valverde1,B. Vidriales1, A.Lopez, M.Gonzalez1, R.Garcia-Sanz1, A~Orfao.Unitof Cytometry, University of Salamanca and Dept. of Haematology, University Hospital of Salamanca, Spain. The aim of the present study was to analyze: 1) the incidence of "leukaemJc" phenotypes within patients with ALL, 2) to determine the frequencyand causes of phenotypic changes both in non-remission(NR) and in relapse (R) samples, and 8) the clinical value of the immunophenotypicdetection of MRD in ALL. For that purpose a total number of 72 ALL cases with a mean follow-upof 85+__28 months (range: 12 to 100 months) were included in the study. A total of 323 complete remission (CR) bone marrow (HM) samples were analyzed. All patients were uniformelytreated accordingto the protocolof the SpanishPETHEMA Cooperative group. Sixteen cases were T-lienage ALL and the remaining 56 patients corresponded to B-lineage ALL. All 16 T-lineage ALL cases displayed at diagnosis, phenotypes which are extremely rare or absent in normal BM and that were used to investigate bIRD. Regarding B-lineage ALL cases, 72% showed aberrant phenotypes. The overall incidence of leukemic phenotypes was: lineage "infidelity" 55%, asynchronous antigen expression 15%, antigen overexpreesion 17%, tissue restricted phenotypes 15% and DNA aneuploidy 15% of the cases. Interestingly, 75% of these patieii~s displayed more than one aberrant pbenotype,Follow-up studies have shown that upon consideringindividually each immtmologicalmarker analyzed, phenotypicchanges are relatively rare ~ 4%, R: 18%) although they might be detected in a relatively high proportion of patients (NR: 25%, 1~ 36%). However the aberrant phenotypes are usually retained both in the NR and the R blast cells (93% of cases). For those cases which display phenotypic changes, they are frequently related to the existence of different blast cell subsets already detectable at diagnosis (NR: 38%, R: 44%), and to technical artifacts(N'R: 25%, R: 17%).Regarding clinical outcome, relapses (n=25) were predicted in 84% of the cases upon analyzing CR BM samples, 1 to 20 months earlier than by use of conventional cytomorphotogy.With respect to those patients who are still in CR and that displayed a leukemia associated phenotype at diagnosis, two groups could be differentiated fromthe investigationof MRD: Dpatients in whomeither no leukemic cells could be detected in the last study (< (].(]1to <(].(](]1%)(5(]%of casas) or it has been decreasing progressively from diagnosis (1(]% of patients) and 2) those patients who displayed an increase in the percentage of detected "blasts" cells without relapse (40% of the cases). Nevertheless, within this last group there is a high proportion of cases 25%/40%, in which the increase in the percentage of cells displaying "leukemic"-associated phenotypes was recently detected.

DETECTION OF MINIMAL RESIDUAL DISEASE IN ACUTE MYELOID LEUKEMIA BY CELL SURFACE IMMUNOPHENOTYPING B. W6rmann I D. Grove I S. K6nemann 1, F. Griesinger 1, S. Toepker2, G. Innig2 Th. BQcllner2, W. Hiddemann 1, LW.M.M Terstappen ~ 1Department of Internal Medicine, University o f G~ttingen, Germany 2Department of Intemal Medicine, University of MUnster, Germany 3Becton Dickinson Immunocytometry Systems, San Jos~, CA, USA Multiparameter immunophenotyping has become a standard procedure in the diagnostic workup of newly diagnosed AML. Leukemic blasts are frequently characterized by a cetl surface antigen presentation, which allows discdmination from normal myeloid progenitor cells. This can be used for monitoring of persistent leukemic cells in hematological CR. The most frequent, informative leukemia-associated antigen pattems are (in descending frequency): expression of lymphoid-lineage associated antigens; asynchronous expression overexpresston or absence of myeloid-lineage associated antigens, expansion of physiological y infrequent cell populations. In 1989 we started a prospective study in patients with newly diagnosed acute myeloid leukemia with the aim of assessing the clinical significance of persistent leukemic cells in hematological CR. In the recent update follow-up data are available in 96 patients. All patients were treated according to the protocols of the German AML cooperative group. The median age of patients was 53 years, 50 patients were female, 46 were male. Persistent cells with the leukemia-associated immunophenotype were detected in 60 of 96 (63 %) patients at achievement of hematologcial CR. 26 of 57 (46 %) patients had persistent leukemic cells 4 to 6 months after achievement of hematological CR. Analysis of subgroups of patients revealed significant differences: 9 of the patients with persistent leukemic cells at achievement of CR had a clearance of leukemic cells through consolidation therapy. 7 of these 9 patients are in CCR with a median observation time of 22 months. 3 patients with no residual leukemic cells at achievement of CR had reappearance of the leukemia-associated immunophenotype at further follow-up. All relapsed within a period of 7 months. The best prognosis was found in 8 patients with an immunophenotypic blast clearance after the very first course of chemotherapy and no reappearance at any further timepoint of evaluation. 7 of the 8 patients are in CR with a median observation time of 48 months. In conclusion, the follow-up data of patients with persistent leukemic cells as detected by multiparameter flow cytometry confirms the unfavorable prognosis of these patients. These data may form a basis for interventional clinical trials with early intensification therapy in patients with minimal residual disease.

A 107

49

51

Timed Sequential Induction Therapy Improves Post-Remission Outcome in Acute Myeloid Leukemia (AML): Woods WG, Kobrinsky N, Buckley J, Lee JW, Sanders J, Neudorf S, Gold S, Barnard D, DeSwarte J, Dusenbery K, Kalousek D, Arthur D, and Lange BJ for the Childrens Cancer Group (CCG)

A randomized comparison of purged autologous bone marrow transplantation (ABMT) and consolidation chemotherapy (CC) for acute myelogenous leukemia in children: Pediatric Oncology Group(POG) study 8821. Ravindranath Y, Krischer J, Yeager AM, Steuber CP, Graham-Pole J, camitta B, and Weinstein HJ for the POG, Chicago, IL. To address the role of ABMT in firstremission of childhood AML, we conducted a randomized comparison of ABMT vs CC. Patients (pts) <18 years of age with untreated AML received remission induction with one cycle of daunomycin (D), cytarabine (A), and Thioguanine (T), followed by high dose cytarabine (HdA). Allogeneic BMT (allo BMT) was offered to pts with matched sib donors. Pts ineligible for allo BMT were randomized after HdA to receive either CC (sequential cycles of D/HdA, DAT, VPI6/AZ, HdA, DAT and VPI6/AZ) or ABMT (busulfan 16 mg/kg and cytoxan (CY) 200 mg/kg followed by infusion of 4hydroperoxy-CY purged marrow). Five hundred and fifty-five of 648 evaluable pts (86%) attained complete remission (MI or M2A marrow), of the 555 pts who attained CR, 87 underwent allo BMT, 63 were ineligible for randomization, (Down syndrome, secondary AML, etc)) and 21 were taken off study for other reasons. Of the remaining 384 pts, 226 (59%) were randomized to ABMT (113) or CC(I13). For the randomized pts, disease-free survival (DFS) at 3 years for CC or ABMT was not statistically different when analyzed by intent to treat (CC 34% vs ABMT 36%) or as treated (CC 37% vs ABMT 48%). The 3 yr DFS was 56% for pts given allo BMT. The 3 yr event-free survival for all 648 pts was 37%. New strategies during induction or in the postremission phase of therapy are clearly needed to further improve upon these results.

Backgrounck.: Recruitment and synchronization of ceils by kinetically sequenced timing of phase specific chemotherapy has been proposed as an effective way to eradicate leukemic myeloblasts. To determine the impact of timed sequencing on remission rate and long term outcome, CCG compared sequentially timed therapy to conventionally timed induction therapy in children with AML. Methods: At diagnosis, 591 patients were randomly assigned to receive one of two schedules of a four clay cycle of dexamethasone, cytosine arabinoside, 6-thioguanine, etoposide, and rubidomycin (DCTER). The timed sequential schedule mandated DCTER on days 0 to 3 and 10 to 13. The conventional schedule prescribed the second cycle of DCTER on day 14 or later depending on bone marrow status. Upon marrow recoveq! patients received a second course of sequentially timed or conventionally timed DCTER. Patients in remission with a matched related donor were allocated to allogeneie marrow transplant; those without a donor were randomly assigned to 4-HC purged autologous transplant or intensive chemotherapy (Coded regimens X, Y and Z). Results'. Induction outcomes are tabulated below: Timing Fail Death Remission I Sequential (n=296) 10 % 14 % 76 %* (P=0.24)* Convent ona (n=295) 21% 5% 74 %* Actuarial event-free survival at 3 years is 43__.7%in the sequential timing regimen and 27_+6 in the conventional timing group (P=0.0008). Actuarial post remission disease-free survival at three years is listed below accordin9 to coded regimen X Y or Z.

Timing

X

Y

Z

Combined

Sequential 49_+16 79+14 59-2:18 60-+9* (P=O,O002)* Conventional 30-+17 52+17 39-+16 37_+9* Conclusions: Sequential timing of induction therapy improves postremission outcome of allogeneic transplant, autotogous transplant and chemotherapy.

I

I

50

52

INTENSIFIED EARLY THERAPY FOR CHILDHOOD ACUTE MYELOID LEUKEMIA (AML): PILOT STUDIES OF THE AIEOP COOPERATIVE GROUP.

P R O G N O S T I C F A C T O R S AND T R E A T M E N T IN P E D I A T R I C A C U T E M Y E L O B L A S T I C L E U K E M I A (AML). F R E N C H TRIALS. G Sehaison, G Leverger, G Michel, T Leblanc, D Sommelet, I Thuret, J Landman-Parker, E L e Gall, Y Perel, E Gluckman, J P Vannier, B

S. Amadori, A.M_ Testi, M.L Moteti, F. Giona, M. RoUa, A. Pessioa, IL RondeUi, Iv[. De Rosa, M.T. Di Tnilio, F. MandeUi. Improvement in AML outcome over current treatment may depend on more eff~tive induction therapy to attain higher complete remission (CR) rates and ameliorate CR quality. In 1992 the AIEOP cooperative group initiated a pilot study (LAM 921>) to explore the feasibility of administering two courses of an intensive induction regimen combining Idarubicin, Etoposide and Cytarabine (~[CE; 3+5+10 schedule), followed by a single course of high dose Cytarabine and Idarubicin as postremission therapy for newly diagnosed children with AML. Patients in CR with HLA-compatible siblings were offered bone marrow transplantation (BMT). From April 1992 to March 1993, 20 patients aged < 17 years entered the study from 7 AIEOP centers: 12 males, 8 females, median age 9 years (range 1 month to 17 years); FAB M0 I, MI 5, M2 2, M3 5, M4 2, M5 5; de novo AML 19, AML secondary to Ewing sarcoma 1 CNS+ 1. Of the 19 evahiable patients 1 (5%) died during induction, 3 (16%) were resistant and 15 (79%) achieved CIL Of the 15 responders, one suffered from an early relapse and another was withdrawn from the study because of a severe pulmonary fangal infection. The remaining 13 complete responders completed the therapeutic program (allogeneic BMT 4; pest-remission chemotherapy 9): 1 died in CR of disseminated fungal infection occurring during neutropenia induced by post-remission chemotherapy;, 4 had recurrent disease at 3-10 months (1BMT; 3 chemotherapy); 8 remain in continuous CR with a median follow up of 18 months. In April 1993, a new pilot study (LAM 93P) was activated with the following changes: 1) the dose of Idarubicin during induction was decreased to reduce gastrointestinal toxicity; 2) a third course of ICE was added for patients reqniring two courses to enter CR; 3) autologous BMT was considered as post-remission therapy for all complete responders lacking compatible sibling donors; 4) FAB M3 was excluded. As of May 1994, 34 children were enrolled from 16 AIEOP centers: 16 males, 18 females; median age 8 years (range 6 monts 15 years); FAB M0 1, MI 6, M2 9, M4 9, M5 g, M6 1; SNC+2, CR was achieved in 25/32 evaluable patients (78%), 2 (6%) were resistant, 5 (16%) died during induction (4 from infection and 1 from hemorrage). Of the 25 responders, 5 are too early, 2 suffered from an early relapse, I was withdrawn for toxicity. The remaining 17 patients were submitted to BMT (8 allogeneic, 9 autologous), and are presently in CCR with a median follow up of 6 months. The study is still ongoing; updated results will be presented.

Nelken, A Baruchel for the French Society of Pediatric Hematology and Immunology (SHIP). Supported by a Grant from ARC. Protocol LAME 89-91 was designed to assess the comparative value of BMT versus aggressive post remission intensification in children with AML. Secondary AML and M0 subtype were excluded. Induction therapy was a combination ofMitoxantrone (Mit) 12 mg/m 2 x 5 d and Cytarabine (Ara-C) continuous infusion 200 mg/m2 x 7 d. CR patients were given a first 4 d intensification with VP 16 100 mg/m2/d, Daunorubicine 40 mg/m2/d and AraC 100 mg/m2/d followed b y a second intensification with high dose Ara-C + Asparaginase + Amsacrine. Triple IT, CNS prophylaxis was restricted to pts with M4-M5 or WBC > 50.000/ram 3. Skull irradiation was mandatory in pts with initial CNS involvement. I t s were then randomized with or without 18 m maintenance program. All pts with an HLA matched sibling underwent BMT after the 1n intensification. Conditioning regimen includes TBI in 10 pts and Busulfan-Cytoxan in 23 pts. Median time between CR and BMT was 80 _+ 13 d. From 1-89 to 11-93, 171 newly diagnosed pts were registered from 17 institutions, 149 (87%) achieved CR. The median follow up from CR is 25 m. The 3 y DFS and EFS for all pts are 57 + 9% and 49 4- 9%. BMT CHEMOTHERAPY Number 33 116 3 y Actuarial Risk of Relapse 22+17% 43_+11% p = 0.025 Treatment related mortality 3% 6.4% NS 3 yDFS 75+17% 51+10% p =0.018 Age less 1 y and M5 are the 2 adverse prognostic factors for the chemotherapy group. In multivariate analysis WBC, initial CNS involvement, bulky disease are without prognostic value. We conclude 1) Mit + Ara-C is an appropriate induction regimen. 2) A very early BMT is a non toxic procedure. 3) BMT is the most effective treatment modality. 4) 18 m maintenance program don't increase C R duration in pts receiving two aggressive intensifications.

A 108 53

55

C o m p a r i s o n of B i o l o g i c a l Data and P r o g n o s i s in C h i l d h o o d and A d u l t A M L U. Creutzig*, T. Btichner, J. Ritter, G. Schellong, M. Zimmermann, A. Heinecke, W. Gassmann, W. Hiddemann

DYSPLASTIC FEATURES IN DE NOVO AML. CLINICAL IMPLICATIONS

Differences in prognosis between adults and children with AML are mainly attributed to the factor age and different biological features. In order to assess these findings we investigated the initial clinical, morphological and cytogenetic data and outcome in comparable groups of 307 children of study AML-BFM-87 (age 0- 17), and subsets of 704 adults of study AMLCG-86 (age 16 - 59). Children were treated with Ara-C, daunorubicin and etoposide (ADE) induction, a 6-week/7-drug consolidation, 2 blocks of high-dose Ara-CNP-16 + cranial irradiation and continuous maintenance up to a total duration of 1.5 yrs. Adults received a similar induction (thioguanine, Are-C, daunorubicin = TAD) followed by a second induction TAD or high-dose Ara-Clmitoxantrone (HAM), a third TAD course, and 3 years of monthly maintenance. Comparison of initial hematological data showed higher leukocyte counts (median 27.000/mm* vs. 18.8001mm=) and a lower Hb value (8.2 gldl vs. 9.1 g/dl; p< .0001) in children. The FAB subtypes M4Eo, MS, M0, M7 were more common in children than in adults. The chromosomal aberrations t(8;21) and 11q were more frequent in children (21% vs. 16% and 20% vs. 4.5% of pats.), while the incidence of inv16 and -7 was similar in both groups. Overall survival, event-free survival and relapse-free interval at 5 years were .47 (.03), .40 (.03); .52 (.05) in children and .31 (.02); .23 (.02), .35 (.03) in adults. Although less impressive in adults, the known good risk groups in children, e.g. M2 with Auer rods, M3, M4Eo were predictive for a better survival. In comparison, M4 without eosinophils had a poor outcome, especially in adults [EFS: .19 (.06)]. In all age groups the FAB subtypes M0/M5/M6/M7 proved to be the most unfavorable. Corresponding to the morphological data the aberrations t(8;21), t(15;17), inv16 were favorable in adults, and especially in children. We conclude, that patients of all age groups with M4Eo have a relatively good prognosis. Overall, in children the response rate is slightly higher with fewer relapses. Since the difference in outcome cannot be explained sufficiently by biological data alone, the impact of different therapy strategies or different therapy tolerances need further investigations. *Present address: UniversityChildren's Hospital, DepLof Hematology/Oncology Miinster, Germany

T. Hafedach, H. L6ffler, W. Gassmann, C. Fonatsch, B. Schlegelberger, W.-D. Ludwig, E. Thiel, M. C. Sauedand, A. Heinecke and Th. B0chner for the AML Cooperative Group, Kiel, Germany Cases of de novo AML are classified according to the FAB critena into the categories of AML M0 to M7. In addition, some groups tested the prognostic value of dysplastic features. We have started a complete morphological analysis of 387 new cases of AML recruited for the AMLCG-92 study. This analysis included the above mentioned cdteda as well as the following cytomorphological details: Cellularity and the percentage of myeloblasts type I, 11, and III, monocytes and monoblasts (for M4 and M5) and erythroblasts (M6) were examined. Pappenheim and myeloperoxidase (POX) stain were analyzed for the absence or presence of auer rods and for POX deficiency in polymorphonuclear neutrophils (PMN). Non-specific esterase was used for the definition of subtypes M4 and M5. The number of eosinophils and abnormal eosinophils (M4Eo) and basophils, as well as the percentage of erythroid precursors were registered. The dysplasia in granulopoiesis (Dy~G), in e~th~opoiesis (DysE) and in megakaryopoiesis (DysM) was analyzed according to standard criteria of the FAB group and the AMLCG. The percentage of dysplastic PMN, or erythroid precursors, or megakaryocytes were registered. Cases without any dysplasia were seperated from cases with dysplasia in one cell line, a combination of two cell lines, and trilineage dysplasia (TLD). In a multivanant analysis the prognostic impact of all these cytomorphological details for remission rate and disease-free survival in de novo AML was investigated in order to design optional subtype-specific therapeutical strategies for each patient. Results in an ongoing study will be presented.

54

56

GENETIC INSTABILITY AND KARYOTYPE EVOLUTION IN SECONDARY LEUKEMIAS D. Haase, C. Fonatsch*, M. Fenring-Bnske, H. Gudat*, C. Schoch*, B. WOrmann, W. Hiddamann: University of Gottingen, Dept. of HematulJOncul., Robert-Koch-Str. 40, 37075 GOttingen, Germany, * Medical University of Lfibcck, Tumorcytogeneties Workin~ Group, Ratzeburser Allce 160, 23538 Lfibeck Secondary leukemins occur after a preceding myelodysplastic prephase or following a mutagen exposition. In secondary AML clonal chromosomal abnormalities can be observed in a higher frequency as compared to AML de novo. In an analysis of 103 patients with secondary AML, performed in Lfibcck, 72% displayed chromosomal abnormalities as compared to 59% with de novo AML 3q-rearrangements (t 1%), -5/5q(25%), -7/structural 7-abanrmalities (10%), 11q- and 12p-abanrmalities (11% each), monosomy 18 (12%) and trisomy 21 (13%) were the most frequent findings in this AML group, clearly discriminating secondary from de novo AML Beneath these clonal abnormalities the phenomenon of karyotype evolution, which is closely related to the complexity of the rearranged genome, is a frequent and characteristic finding in secondary leukemias and lvlDS and indicates rapid progression and therapy refractoriness. There is increasing evidence that in neoplnsia the accumulation of genetic defects takes place in a nonrandom sequence. In an analysis of AML patients (AML cooperative group) with complex chromosome abnormalities we found 5q- or -5 as initial abnormality in 64%, and -7Hq- and +8 in 12% each. To determine the localization of initial cytogenetic events and karyotype evolution in AML and MDS within the hematopoietic hiemrehy, we performed cytogenetic analyses of highly purified (FACS) CD34+ subpopulations characterized by the expression of CD34 and CD 38. We recently demonstrated primary inv(16) as well as additional trisomy 8 and monnsomy 12 in the CD34+/38- progenitor cells ia a case of de nova AML M4 Eo. In a case of secondary MDS, transforming to acute myeloid leukemia we were able to demonstrate karyotype evolution with 5q- as primmy abnormality, initiating 4 depending cell clones in the CD34+, sorted subpopulation. Our results show that disease initiation as well as disease progression might be driven from the stem cell compartmem in de novo and secondary AML. Furore research should focus on the pathogenesis of karyotype evolution as a prerequisite for the accumulation of genetic defects. Several lines of evidence imply that the loss of genetic material from 5q and possibly 7q might initiate the multistep process of karyotypo evolution. Recently, DNA-repair - and cell cycle controlling genes have been localized on 5q and 7q. Functional and molecular studies of DNA-repair and cell cycle regulation are needed to clarify the molecular basis of genetic instability in MDS and AML manifesting as karyotypo evolution leading to complex abnormalities.

INCIDENCE OF SECONDARY LEUKEMIAS AFTER THERAPY OF CHILDHOOD HODGKIN'S DISEASE (HD) WITHOUT NITROGEN-MUSTARD G. Schellong, M.Riepenhausen, H. Gadner 1, G. Mann 1, U. Creutzig, J. Ritter, for the German-Austrian Pediatric HD-Group - Univ.-Children's Hospital MOnster (FRGI and St. Anna Children's Hospital Wien 1 (Austria) There is good evidence that the risk for secondary leukemiaslpreleukemias (SL) after Hodgkin's Disease (HD) depends to a large extent on the cumulative dosis of alkylating agents (AA), especially nitrogen-mustard (NM). The incidence of SL after MOPP chemotherapy (CHT} is higher than after ABVD. - The occurrence of SL in 4 consecutive German-Austrian multicenter trials using combined modalitiy treatment without NM is analysed. 667 pats under age 16 were enrolled in the trials HD-78, HD-82, HD-85 and HD-87 between June 78 and Sept. 90. Primary CHT consisted of 2 cycles of OPPA (VCR, PRED, PC, ADR) or OPA (without PC) in stages IA/B and IIA (group I ), and of 2 0 P P A or OPA + 2, 4 or 6 cycles of COPP or COMP (C=cyclo, M =MTX) in advanced stages (group 2). CHT was followed by radiotherapy (RT), in HD-78 to the extended fields, in the later trials to the involved fields, with decreasing doses (40-20 Gy). As salvage-therapy IEP fiFO, VP16, PRED) combined with ABVD, COPP and/or CEP (CCNU, VP16, Prednimustin) was used (group 3). Follow-up ranges from 4 to 16 y (median 9 y) in pats alive. - Results: Until August 1994 4 ANNL and 1 MDS were observed, 3 in 1st remission (50, 63 and 122 m after start of treatment) and 2 after relapse and salvage therapy (after 38 and 40 m). NHL was not seen. The table presents the number of pats (N), survival rate (S) at 12 y and the cumulative risks for SL after 5, 10 and 15 y in the total group and in group 1-3: Cumulative Risk (%) N S SL 5y 10y 15y total group (1) 2x OPPA or OPA (2) plus COPP or COMP (3) plus salvage-therapy

667 296 296 75

94 % 100% 94 % 77 %

5 1 2 2

0,5 0,4 0 3,3

0,7 0,4 0,4 3,3

1,2 0,4 1,4

The differences of group 1 vs 3 (P=0.02) and 2 vs 3 (P=0.02) are significant. - Conclusion: The presented risks of SL are smaller than in other series (e.g. Meadows et ah Total group 4% at 15 y, after salvage 8%), probably due to absence of NM in our CHT, the lower cumulative dose of AA and the smaller number of pats needing salvage. Pats remaining in CCR have a lower risk than those who received also salvage. Differences in the risk for SL reflect the differences in the overall treatment strategies.

A 109 57

59

INTENSIVE THERAPY OF MYELODYSPLASTIC SYNDROMES (MDS) AND SECONDARY LEUKEMIAS (SL) : THE FRENCH EXPERIENCE E.Wattel, E.Sotary, D.CaUlot,, N.Ifrah, A.Bdon, N Milpied, M.Janvier, A.Guemi, H.Rechant, C.Cordonnier, F. Dreyfus, ANeil, AM.Stoppa, N.Gratecoe, A.Sadoun, H.Tilly, P.Brice, B.Lioure, B.Desablens, B.Pignon, JP.Abgrall, M.Leporrier, B.Dupriez, D.Guyotat, A.Merlat, P.Fenaux, Groupe Fran~ais des My61odysplasies (GFM) and Groupe GOELAM Low dose chemotherapy, in MDS, gives overall limited results in =high risk* MDS. Intensive chemotherapy, in MDS and secondary leukemias (ie AML following de nero or therapy related MDS), gives lower CR rate than in de novo AML, possibly due in part to higher incidence of mdr gene expression (Leukemia 1994, 8, 998) Furthermore CR duration is generally short in MDS and SL following this treatment and consolidation chemotherapy (Br.J.Haematol, 1991,77,497). We therefore designed a tdal of intensive chemotherapy : Mitoxantrone 12mg/m2/d d2-5 + AraC 1g/m2/12h dl.5, with or without (randomized) quinine 30mg/kg/day, an agent capable of reverting the mdr phenotype. Pts <55 years achieving CR were scheduled to be autografted (ARMT) after bone marrow (and more recently peripheral blood : 3 cases) stem cell collection. The conditioning regimen was: Ex 50mg/Kg/d and Bu 4mg/K/d during 4d. CR criteria were stringent: normalization of cytopenias and of karyotype, in addition to disappeaffance of blasts and MDS features. Pts >55 or who could not be aulografted received consolidation CT. From Oct 1992 to Oct 1994, 84 pts were included : median age was 54 (range: 18-68, 34 pts>55) ; 49 pts had progressed to AML and 35 were still in MDS phase (8 RAEB, 26 RAEBt, 1CMML). 72 pts are currently evatuable for response. 29 pts (40%) achieved CR, 8 pts (11%) achieved PR, 15 (21%) had eady death and 20 (28%) had resistant disease. Analysis of pretreatment prognostic factors of CR is currently underway. Of the t6 pts aged < 55 years who achieved CR, 10 were actually autografted. Reasons for not doing ABMT were : early relapse (3 pts), poor clinical condition (1 pt), allogenle BMT (lpt), patient refusal (lpt). 1 pt where stem cells were collected is awaiting ABMT. Hematological n-=constitution occurred in all pts, after delays similar as for AML. 16 of the 29 pts who achieved CR have relapsed after 2 to 18 months and 13 are still in CR after 1 to t9 months. 6 of the autografted pts have relapsed, after 4 to 18 months, 1 died after the procedure and 3 are still in CR after 8 to 9 months. Median overall survival from the onset of treatment was 8 months. Median CR duration was 10.5 months. The CR rate, CR duration and survival from treatment were 42 % and 38 %, 11 months and 10 months, 8 months and 7.5 months, respectively in pts treated with and without quinine (difference NS). The effect of quinine in pts expressing mdr at the onset of treatment ls currently examined: preliminary results in 14 pts expressing MDR1 at the onset of treatment show a CR rate of 28% (as compared to 14% is our previous experience without quinine in mdr positive MDS) (Leukemia 1994, 8, 998). In those 14 pts, 4 of the 9 cases (44%) treated with quinine and 0/5 (0%) in pts treated without quinine achieved CR (p=0.07). In conclusion. (1) we confirm that relatively small CR rates are obtained with intensive chemotherapy in MDS; (2) preliminary findings suggest a potential benefit of quinine in mdr positive cases; (3) ABMT is feasible in MDS after CR achievement but it remains to know if it will prolong CR by compadecn to consolidation chemotherapy.

Mediators in Sepsis and Septic Shock. J. Kienast and H.Ostermann, Department of Internal Medicine, Haematology / Oncology Section, University of Miinster, Germany

58

60

Intensive Therapy of High Risk Myelodysplastic Syndromes with Sequential Intermediate Dose Cytosine Arabinoside and Mitoxantrone with or without GM-CSF W. I-Iiddemann, Th. B0.chuer, B. W6rmann, P. Koch, C. Aul, L. Balleisen, 3r. Bennett. Dept. of Hematology/Oncology, University of G6ttingen, Germany

INTERVENTIONAL ANTIMICROBIAL THERAPY tN FEBRILE NEUTROPENIC PATIENTS H. Link, G. Maschmeyer, P. Meyer, W. Hiddemann, W. Stille, M. Helmerking and D. Adam, for the study group of the Paul Ehrlich Society for Chemotherapy (PEG)

Myelodysplastic syndromes (MDS) represent a heterogenous group of disorders with different biology and prognosis. Based on the percentage of blast ceils and the degree of hematopoietic insuffidency a high risk group of patients can be identified with an expecteA survival time of less than 12 months only. The outcome for these patients remains dismal in spite of therapeutic intervention with intensive or prolonged low dose cytostatic therapy. The discovery and characterization of hematopoietic growth factors provides a new approach both by the ability to possibly increase the sensitivity of blast cells to subsequent cytostatic therapy as well as by the stimulatory effect on hematopoietic precursor cells shortening the period of treatment associated cytopenia. Both aspects are addressed by the current study. Patients with high risk MDS ~ , RAEBt, CML) are randomized in a double blind fashion to receive GM-CSF or placebo by 1 x daily s.c. injection starting 48 hours prior to cytostatic therapy with S-HAM. At the present time 28 patients have entered the study including RAEB 6, RAEBt 15, CMML 3 and unclassified 4. From the 20 currently fully evaluahie patients 10 (42 %) achieved a complete remission, 6 (25 %) had persisting MDS. Time to recovery from critical cytopenia was 2g days (range 19 - 55). Unmaintained remission duration at 3 years is estimated to be 35 %. These data indicate that intensive chemotherapy may be applied to this high risk group of patients and that the preliminary data on remission duration look promissing.

Proinflammatory cytokines, namely interleukins(IL)-l, -6, -8, and tumor necrosis factor-or (TNF-ot) are pivotal mediators of septic shock and multiple organ failure in non-leucocytopenic patients. Circulating neutrophils and mononuclear cells are thought to be a major source for their release in response to infectious stimuli. We have prospectively measured plasma levels of pro- and antiinflammatory cytokines in leucocytopenic patients with febrile infections. Cytokine concentrations were related to the severity of the inflammatory reactions and outcome. Two prospective studies were consecutively conducted in a total of 52 patients with AML who developed fever in the presence of infection during chemotherapy-induced leucocytopenia. A significant increase in TNF-a, IL-6 and IL-8 was observed at the onset of fever, with peak levels being related to the severity of the febrile episode (sepsis < severe sepsis < septic shock). Simultaneously, an increase in endogenous antiinflammatory mediators (IL-10, soluble TNF-receptor, and II-1 receptor antagonist) was observed. Markers of neutrophil or monocyte activation, such as elastase or neopterin, were not significantly elevated. We conclude that proinflammatory cytokines originating from sessile mononuclear cells or endothelium play a key role in mediating the systemic inflammatory response to infection also in leucocytopenic patients. The release of antiinflammatory mediators parallels rather than follows the proinflammatory response. Data from the first study suggested that a cytokine score based on peak values for TNF-c~ (cut off: 60 pg/ml), [L-6 (400 pg/ml), and IL-8 (4500 pg/ml) allows the prediction of lethal outcome. This hypothesis was validated in the subsequent independent study. Mortality rates in patients with _< l, 2 or 3 of the above cytokines exceeding threshold levels were 8%, 50% and 86% respectively (p < 0.005).

In this prospective multicentre trial, treatment strategies for 1573 patients with neutropenia < 1000/pl and fever ;~38.5~ after cytotoxic chemotherapy were compared. Patients with unexplained fever were randomised to a three-phase sequential study for different established drug regimens. If an infection could be defined microbiologically or clinically, treatment modifications were determined. In phase I, treatment for all patients consisted of acylaminopenicillin (PEN) plus aminoglycoside (AMG); or third generation cephalosporin (CEPH) plus AMG; or PEN plus CEPH. In 800 patients with unexplained fever the response rates were: PEN/AMG (n =258): 74,4%, CEPH/AMG (n =252): 73.4%; PEN/CEPH (n =290): 70.0%, Total response rate was 72.5%. In Phase II, patients not responding after three days received PENICEPHIVancomycin ( n = 7 0 ) or PEN/CEPH/AMG (n = 74). The respectively response rates were 52.9% and 55,4%, total 54,2%. If fever did not resolve, the patients received either PEN/CEPH (n =40) or Imipenem/Cilastatin ( n = 5 9 ) both in combination with Amphotericin-B/5-Flucytosin/Rifampin. The response rates were 62.5% and 79.7%, respectively (p=0.07), total 72.7%. No significant differences between the treatment modalities compared were found. Analysing all three phases together, 91.3% of patients with unexplained fever were cured. The response rate was also analysed by patients with gram-positive bacteraemia (n = 183), response rate = 82.5%; gram-negative organisms (n = 145) 78.6%; fungaemia (n =51 ) 43.1% (p<0.O01); lung infiltrates (n=269) 61.3% (p<0,001); clinically documented infections (n = 198) 84.4%; and in clinically and microbiologicaily documented infections (n = 84) 82.1%. tf infections were diagnosed after at least five febrile days, more lung infiltrates and fungal infections occurred (p
A110 61

63 FUNGEMIA IN CANCER PATIENTS IN EUROPE Fran~oise Meunier, EORTC, Brussels, Belgium

Fungemia is an increasing complication during management of cancer patients. For too long, this infection has been considered as a benign, transient phenomenon, but there is a need to increase the awareness of physicians and nurses about the morbidity and mortality related to fungemia in cancer patients. The EORTC Invasive Fungal Infections Cooperative group initiated for a period of two years (1993 and 1994) a survey on fimgemia observed in cancer patients, including those with leukemia, in Europe. As of October 1994, more than 225 episodes of fungemia (defined as at least one positive blood culture for yeasts) have been registered at the EORTC Data Center by 29 institutions. The study will be terminated at the end of 1994. Preliminary data provide the characteristics of those patients as well as their outcome and death rate within 30 days after fnngemia. Half of the patients reported in this survey have acute leukemia as underlying disease. Overall, mere than 50% of patients had fungemia caused by non-albicans Candida and half of the patients had neutropenia at the time of fungemia. Fungemia appears mainly as a nosocomial infection since most patients were hospitalized for a prolonged period prior fungemia (medim: duration of hospitalization prior fimgemia 22 days, range 1-46). Breakthrough fungemia (i.e. occurring while the patients were receiving chemoprophylaxis or empiric antifungal treatment) was not uncommon, about 40% of patients with fengemia did not receive prior antifungal prophylaxis wbile more than 30% received polyenes as prophylaxis. Clinical signs of fungemia were often lacking. The overall mortality (within one month after fungemia) was 45% stressing the fact that fungemia is a severe opportunistic complication in cancer patients including in those with acute leukemia.

NEW APPROACHES TO THE MANAGEMENT OF PATIENTS ALLOIMMUNIZED TO PLATELET TRANSFUSION. C.A. Schiffer, Univ. of Maryland Cancer Center, Baltimore, Maryland, USA. Approximately 30% of patients with leukemia receiving multiple platelet transfusions become alloimmunized and refractory to transfusions from random donors. Perhaps of greatest biologic interest is that the majority of patients do not make antibody against histocompatibility antigens despite these multiple exposures, and behave as if they were immune "tolerant". The mechanism for this tolerance is poorly understood. Detection of anti-HLA antibody using either lymphocytotoxicity or antiplatelet antibody assays can distinguish alloimmunization from other causes of platelet destruction and indicate the need for single donor histocompatible platelets. A variety of donor selection strategies by HLA typing (e.g. use of platelets mismatched for antigens poorly expressed on platelets such as HLA B44) or, more recently, by platelet cross matching, can be helpful in such patients. We have noted that platelet cross matching can sometimes identify compatible donors who might not have been identified by HLA typing, suggesting that antibody against other antigen systems may contribute to refractoriness. Clinical trials are in progress evaluating the benefit of leukocyte depletion by filtration or UV irradiation of the platelet product as a means of preventing alloimmunization. Preliminary data from small trials are promising but not conclusive, and such approaches, which substantially increase the cost of transfusion, should be considered investigational. A large, multi-institution randomized trial (TRAP study) will complete accrual shortly in the US and should provide definitive information about the relative efficacy of leukocyte depletion and UV irradiation.

62

64

CMV-Infection: Role of Gancielovir and Hyper-Immunoglobulines (Canadian BMT Study Results) H.A. Messner, University of Toronto, The Princess Margaret Hospital, Toronto, Canada

PROGNOSTIC FACTORS IN AML (AMLCG RESULTS) Th. Biiehner*, W. Hiddemann, H. I_,rffler, L Goasguen, J. Bennett, W. Gassmann, T. Haferlach, C. Fonatsch, R. Becher, D. Hossfeld, M.C. Sauerland, A. Heinecke

Abstract not received

From recent reports intensive induction (Blood 84 (1):914, 1994) or postremission therapy (Leukemia 5:510, 1991, N Engl J Meal 331:896, 1994) using high-dose AraC substantially improve the cure rate in AML. Compatibly to bone marrow transplantation more than 50% cures can be achieved in special subgroups like favourable karyotype (Blood 94 (1):431, 1994). Since favourable karyotype accounts for less than 20% pts additional prognostic factors were tested in a study by AMLCG where pts adapted to age received double induction - consolidation maintenance chemotherapy. In the 1044 pts of all ages the CR rate is 62% and CR's after 5 years are 33% as from intention to treat. By univariate analysis factors predecting superior disease free survival were lower LDH (p = .0001) and WBC (.0003), favourabte karyotype (.0004), younger age (.0007), M3 morphology (.005), and absence of dysmegakaryopoiesis or dyserythropoiesis (.046). A cure rate of more than 50% was predicted by favourable karyotype and/or M3 morphology. 50% cures were also achieved in pts characterised by age 26-60 and WBC < 25000/cram and LDH < 1000 U/1 which applies for 45% of pts under 60. We conclude that by commonly available disease characteristics - in particular LDH - larger groups of pts than by cytogeneties with favourable "transplant like" outcome can be identified. *Present address: Department University of Mtinster, Germany

of Medicine/Hematology/Oncology,

Alll 65

67

P r o g n o s t i c F a c t o r s in A M L - T h e L o n e S t a r E x p e r i e n c e

INTENSIFIED INDUCTION IN ACUTE MYELOID LEUKEMIA (AML) J.F. Bishop, J.P. Matthews, G.A. Young, J. Szer, A. Gillett, D. Joshua, K. Bradstock, A. Enno, M. Wolf, R. Fox, R. Cobcroft, R. Herman, R. Lowenthal, F. Page and S. Juneja for the Australian Leukemia Study Group (ALSG) Melbourne, Australia.

Estey E, M.D. Thall P, Ph.D.

From the Department of Hematology, University of Texas M.D. Anderson Cancer Center, Houston, Texas, USA At M.D. Anderson prognostic factor analysis is becoming increasingly sophisticated. As examples we are now using "smoothed martingale residual plots" (SMRP) to select optimum "eutpoints" for continuous covariates, considering interactions between prognostic covariates, and using models that are not dependent on the assumption of proportional hazards. Traditionally when one considers the relation between clinical outcome and a continuous variable such as % of cells expressing CD7 the "minimal p-value" method is used, in which for all possible cutpoints between distinct CD7 values in the data, CD7 is dichotomized with survival compared in. patients with values > and < than the cutpoints, and the CD7 outpoint with the minimal p-value chosen. Of course the p-value then has to be adjusted upward to account for the many tests involved in this process. However, the use of SMRP constructed after first fitting the Cox model with CD,7 as predictor can permit the search for the optimum cutpoint to be limited to a small range, therefore limiting the necessary upward p-value adjustment. A recent model using a CD7 cutpoint of 10.5% found an interaction between cytogenetics and CD7 such that whereas both were independent predictors of survival in AML CD7 was predictive only in patients with unfavorable cytogenetics. Finally an analysis of the effect of treatment on survival in MDS indicated that the (negative) effect of treatment became more pronounced with time. To accommodate this an accelerated failure rather than a propostimal hazards model can be used. The application of these techniques to development of prognostic models in AML will be illustrated.

Optimal initial treatment has made a major impact on a number of curable cancers. To study methods to improve outcome in AML the ALSG has intensified induction therapy in two consecutive randomized trials. In the first trial, patients were randomized to receive standard daunorubicin 50mg/m2 in daily x 3 and cytarabine 100mg/m2 as a continuous infusion for 7 days (7-3) or 7-3 as above with additional etoposide 75mg/m2 daily x 7 (7-3 vs 7-3-7, Blood 75: 27-32, 1990). Of 264 eligible patients the remission duration was significantly longer with 7-3-7. In a subset analysis patients aged <55 years had significantly superior survival on the etoposide ann. In a dose intensity multivariate analysis, we have shown that the initial dose of cytarabine and dannorubicin significantly influenced the duration of remission in AML (Leukemia and Lymphoma 15: 79-84, 1994). To test if eytarabine dose escalation in induction could influence outcome, the ALSG randomized 301 eligible de novo patients to either standard dose c3~rabine in the 7-3-7 regimen or to high dose cytarabine at 3g/m: twice daily on days 1, 3, 5 and 7 with daunorubicin and etoposide as above (HIDAC-3-7). HIDAC-3-7 significantly prolonged remission duration with a median for HIDAC-3-7 of 70 months vs 12 months for 7-3-7 (P=0.0004). The percentage of patients relapse free at 5 years was 51% on HIDAC-3-7 and 23% on 7-3-7. The rate of relapse for HIDAC-3-7 patients was 0.52 relative to 7-3-7 patients and the rate of death was 0.74 relative to 7-3-7 patients. HIDAC-3-7 was associated with significantly more toxicity in induction with more leucopoenia, thrombocytopenia nausea and vomiting and eye toxicity (all p<0.001). Although both arms received the same post-induction therapy (5-2-5), patients who received HIDAC-3-7 in induction had significantly more myelosuppression. We conclude that intensifying induction therapy with etoposide and with high dose cytarabine can significantly prolong relapse free survival with increased but acceptable toxicity.

66

68

Intensified Induction in Acute Myeloid Leukemia - AMLCG Results W. I-Iiddemann and Th. B0chner for the AML Cooperative Group G6ttingen, Germany

ALL TRANS RETINOIC ACID (ATRA), A TARGETTING DIFFERENTIATION THERAPY FOR ACUTE PROMYELOCYTIC LEUKEMIA : CLINICAL RESULTS Laurent DEGOS. Institut d'Htmatulogie, Hopital Saint Louis, Paris 75010, FRANCE

Increasing evidence suggests that intensification of chemotherapy by high dose Ara-C during induction or consolidation may improve the Iongterm outcome in adult patients with acute myeloid leukemia. The German AMLCG investigated the impact of intensified induction therapy by applying double induction treatment to patients younger 60 years of age and further analyzed in way of a prospective randomized comparison whether the introduction of high dose Ara-C in combination with mitoxantrone as second induction cycle was superior to conventional dose Ara-C combined with 6-thioguanine and daunorubicin (TAD). All responders uniformly received TAD consolidation and three years monthly maintenance. Between 1986 and 1992 692 patients 16 - 59 years of age entered this study and were eligible for randomization. Complete remission rates were comparable with 66 % CR after TAD/TAD and 73 % CR after TAD/HAM. However, in patients with a significant residual blast cell infiltration after the first induction course TAD/HAM proved superior with 45 % of patients achieving remission as compared to 25 % of cases randomized to TAD/TAD. With no exclusions after treatment start the event free survival of all patients and disease free survival showed a clear tendency in favor of TAD/HAM with a probability of remaining relapse free at 5 years of 40 % as compared to 30 % in the TAD/TAD arm. These data indicate that double induction treatment is feasable and associated with a remarkably high rate of Inngterm remissions. TAD/HAM appears favorable to TAD/TAD especially in terms of longterm prognosis.

Differentiation of leukemic cells from patients with acute promyelocytic leukemia (APL) is also ascertained by PML products molecules on the outer shell of a specific nuclear body (0.3 - l~tm diameter) encoded by PML gene which have a speckled nuclear pattern in normal ceils. PML molecules in APL patients have a micropunctuated pattern and are delocalized in the cytoplasm. ATRA therapy rapidly mobilizes the molecules and induces a normal pattern. Clinical results : ATRA treatment was proposed at doses of 45 mg/m2/day. Lower doses are able to induce CR but do not prevent the adverse effects (hyperleucocytosis and retinoic acid syndrome). Survival times are dramatically increased by using ATRA before chemotherapy. Event free survival was estimated at 82% versus 50% at 12 months in ATRA plus chemotherapy cohort compared to chemotherapy alone (p < 0.005) in a multicentric randomized trial. The last analysis showed an EFS, relapse rate and survival at two years of 68%, 25% and 81% in the ATRA group as compared to 23%, 56% and 52% (p=10 -4, p=10 -4, p = .009 respectively) in the chemotherapy group. Rapid disappearance of bleeding diathesis is a major advantage of ATRA treatment. The major disorder is a primary fibrinolysis, part of an extended proteolysis, which is removed by ATRA treatment. Mild DIC and procoagulant tendency are not modified by ATRA treatment, and are responsible for thrombotic events. Hyperleucocytosis (and retinoic acid syndrome) are efficiently avoided by the addition of chemotherapy when WBC increase during ATRA treatment. In the multicentric european trial APL 93 in which 175 patients are actually included, the low rate of early mortality (5,4%) leads to a high rate of CR (94,5%). The absence of primary resistance is confirmed. Secondary (second tratment) resistance is due to an induced catabolytie process which is reversible after 6 months out of ATRA treatment. In conclusion, ATRA in APL is the first model of differentiation therapy in malignancies, the first specific treatment for a genetic defect, the first evidence for a reversible nuclear structure disruption and greatly improves the survival of patients.

Al12 69

71

INTENSIFIED POSTREMISSION CHEMOTHERAPY FOR ADULTS WITH AML. THERAPEUTIC RESULTS AND PROGNOSTIC FACTORS, RJ Mayer, CD Bloomfield, DT Berg, RB Davis, DC Arthur, D Lawrence, CA Schiffer. Cancer and Leukemia Group B (CALGB), Lebanon NH and Dana-Farber Cancer Institute, Boston, MA, USA While complete remission (CR) may be achieved in 65% of adults with de novo AML such CR's lack durability when conventional postremission therapy is administered. Uncontrolled trials have suggested that intensive postremission therapy may prolong these CR's. To assess this concept, the CALGB administered induction therapy (ara-C and daunorubicin) to 1088 adults (median age:52 years). Of the 693 (64%) patients (pts.) who achieved CR, 59% were randomly assigned to receive 4 courses of single agent ara-C in one of the 3 dose schedules: 3 g m / m 2 in 3 hr infusion q 12 hr (twice daily) on days 1,3, and 5 (HiDAc)(187 pts.); 400 m g / m 2 / d x 5 days (CI) (206 pts); or 100 m g / m 2 / d x 5 days CI (203 pts.). All pts. then received 4 courses of monthly maintenance treatment. After a median follow-up of 63.5 months, the disease-free survival rates in the three treatment groups were significantly different (p=0.001). The probability of remaining in continuous (CCR) after 4 years for pts. 60 years of age or younger was 43% for HiDAc, 31% for 400 mg, and 23% for 100mg (p=0.0007). In contrast for pts. older than age 60 years, the probability of CCR after 4 years was 16 percent or less in each of the 3 ara-C groups. Adequate pretreatment cytogenetic data were available from 615 pts., of whom 388 (63%) achieved a CR and 347 were randomized to one of the ara-C dose arms. Pts. were divided into favorable (t[8;21] or inv[16]) (62 pts.), intermediate (t[15;17] or normal) (32 pts.), or unfavorable (other karyotype) (160 pts.) cytogenetic subsets and had probabilities of CCR after 4 years of 51%, 32%, and 16% respectively. Higher doses of postremission ara-C did not influence CCR in the unfavorable group, improved outcome in the intermediate group, and strikingly prolonged CR in the favorable subset. To identify prognostic factors for CCR, 10 risk factors were subjected to evaluation. Multivariate analysis showed cytogenetics and ara-C dose to be most predictive for cure. These data support the concept of a dose-response effect for araC in pts. age 60 years or younger with the probability of benefit strongly related to pretreatment karyotype.

POSTREMISSION THERAPY: THE ROLE OF BONE MARROW TRANSPLANTATION IN ACUTE MYELOGENOUS LEUKEMIA

(AML)

T. de WiRe, F. Mandelli, R. Zittoun, R. Willemze, P. Muus, M.C. Petti, S. Suciu, G. Solbu, M. Dardenne, M.L. Vegna, S. Keating, M. Peetermans. EORTC Data Center, Bruxelles, Belgium. The AML 8A protocol of the EORTC and GIMEMA Cooperative Groups studied the value of allogeneic (allo-BMT) and autologous bone marrow transplantation (ABMT) in adult AML under 45 years old (or 60 according to policies of centers) when performed during first complete remission (CR). In this protocol 990 patients with AML were registered since November 1986. Following 1 or 2 courses of remission-induction treatment, 66% of patients achieved a CR. Of the 623 evaluable patients achieving a CR, 576 received an intensive consolidation course. Then 168 patients who had a HLA-identical sibling were assigned to alIo-BMT, while 254 were randomized for an ABMT using an unpurged bone marrow, or for a second intensive consolidation course. At a median follow-up of 3 years DFS of the intensive chemotherapy arm was 30% at 4 yrs, the DFS in the alIoBMT and ABMT arms were 55% and 48% respectively. The two BMT arms gave significantly better results than the intensive chemotherapy arm (p= 0.03). The main reason for failure is relapse in both the ABMT and the chemotherapy (CT) arm, while treatment-related mortality is higher in the alloBMT arm. A preliminary, retrospective analysis of the role of alIoBMT showed that patients in whom no HLA-typing was performed had a poor prognosis. Furthermore only 68% of the patients in CR1 with a donor received an alloBMT. Patients with a donor and who received no alloBMT in CR1 had a poor prognosis. Only 60% of the CR patients without a donor were randomized between chemotherapy and autoBMT. Moreover only 75% of the patients randomized for autoBMT received an autograft. An intention-to-treat analysis resolved this last selection bias partly. The rate of patients completing the full treatment protocol is still relatively low. In the AML 8A study only 33.5% of registered patients completed the full treatment protocol. Results of BMT are difficult to interpret because of these selection biases. Most selection biases, such as delay of BMT after achieving CR, are similar for autoBMT and alIoBMT. Genetic selection bias is not operational in autoBMT, but exclusion of patients due to insufficient number of CFU-GM in the harvest forms a specific selection bias in autoBMT. Careful anatysys of these selection biases are needed. Prospective randomized trials are extremely useful for this approach.

70

72

INTENSIFIED POSTREMISSION CHEMOTHERAPY IN A D U L T S WITH AML (THE FINNISH LEUKEMIA GROUP RESULTS). Elonen E, Nmqvtst A, H~nninen A, Jansson S-E, Jirvenlie G, Koistinen P, Koivunen E, Lahlinen R, Lehtinen M, Nousiainen T, Pelliniemi T-T, RajamSkl A, Remes K, Timonen T, Vilpo J, Volin L, and Ruutu T.

The Value of Bone Marrow Transplantation in AML in CR1, in Intensively

248 consecutive patients under 66 years of age with de novo acute myeloid leukaemia (FAB M1 50, M2 82, M3 50, M4 56, M5 23, M6 4, unknown 3) admltted to the participating six centers were included In the study. Median age was 46 yrs (range 16 to 65). The patients In remlssion (CR) after two cycles (C) were randomized to receive either 2 (short ann n=73) or 6 (long arm n--66) further intensive cycles. C1 and C2: Daunorubicin (DR) 50 mg/m 2 d 1,3 and 5, AreC 150 mg/m 2 cont.inf, and thioguanine 75 mg/m 2 x 2 d 1 to 9. C3: Amsacrine 115 mg]mZ d I t o 5 , AraC 3g/mZ x 2 d 1 and2. C4: AraC 2 g/m2 x 2 d I to 5, DR 30 mg/nl~ d 6 to 8. C5 and C6: Aclarubicin 25 mg/n~ d 1 to 7, etoposide 50 mg/m 2 x 2 d 1 to 5, vlncristine 1 mg/n~ d 1 and 5, prednisone 50 mg/m 2 d 1 to 7. C7 DR 30 mg/m 2 d 1 to 3, AraC 500 mg/m 2 cont.inf, d I to 3 and 10 to 12. C8: Amsacrine 115 mg/n~ d 1 to 5, AraC 2 g/m:' x 2 d 1 and 2. Completely refractory patients ( ~ ) were b'eatsd with (::3 instead of C2 after C1. 192 of ti.m 248 patients (77%) achieved CR with 1 or 2 cycles, 69% with C1, 25% with C2, and 6% with C3. 15% of the patients were refractory and 8% died during induction. 61 patients have been transplanted, 41 in first CR, all of them were censored at the time of transplantation. The median follow up time of the living patients is 57 mo (from 25 to 103 mo). The median relapse free survival (RFS) is 17 mo, and 36% of the patients live relapse free for 5 years. The median survival lime is 25 mo, and 35% of the patients are alive for 5 years. There was no significant difference in survival or RFS between the randomized arms. The median RFS are 21 and 17 mo, and median survival times are 43 and 39 mo In the short and long arm, respectively. Because of uneven mortality of patients during C3 and C4 (same treatment in both arms) the patients in CR after C4 (short arm n=61, long arm n=53) were analyzed separately. RFS of the patients under 46 years of age Is longer (p=0.016) in the long arm than in the short arm but no difference in survival limes is seen. 46 patients in the long arm received more than 4 cycles of chemotherapy; two elderly p a tients died of treatment related toxicity. It is concluded that 8 cydes of intensive chemotherapy is not better than 4 cycles in the whole study group but younger patients may benefit from the longer chemotherapy.

Treated Patients: The MRC AML 10 Trial. A K BurneR, A H Goldstone, R Stevens, I Hann, J Rees, K Wheatley A major aim of this trial was to evaluate the role of BMT (allo or auto) as extra consolidation of patients aged <55yrs who had already received intensive chemotherapy. Patients, including those with preleakaemic phase, were first randomised to receive either DAT (3+10) or ADE (3+5+ 10) followed by DAT (8+3) or ADE (8+3+5) as course 2, MACE as course 3 and MidAc as course 4. Those with a matched sibling donor should receive allogeneic BMT, the remainder were harvested and randoraised to receive autologous BMT, or no further treatment (STOP) with the provision that those who relapsed could receive an auto BMT in CR2. The trial opened in mid 1988 and closed in October 1994. Nineteen hundred and ten patients were entered. The overall remission rate was 81% (ranging from 91% in children to 75% in adults aged >45 years). There were no significant differences between DAT or ADE arms. The overall

survival from entry is 41% at 5 years. The strongest predictors of survival were cytogenetics (t15:17; t8:21; inv 16 are favourable while complex karyotype or abnormalities of 5 or 7 were unfavourable) or the percentage of blasts in marrow at the end of course 1. These factors dominate survival irrespective of subsequent treatment modality. 365 Patients were randomised between STOP (n= 182) and autograft (n= 183, 142 completed). 318 had a matched sibling of whom 213 have received Allo BMT. 40% of patients in CR who were eligible, were randomised. In the comparison of AUTO vs STOP there was no significant difference in survival or event-free survival (55% vs 48% and 52% vs 37%) at 5 years, but autograft significantly reduced relapse risk 39% vs 61% (p=0.008). All analysis are based on intention to treat. If the allografts are analysed on an intention to treat basis ie donor available vs no donor, there is at present no significant survival difference in good or poor risk patients, but in standard risk patients the donor group has a significantly superior survival.

Al13 78

75

POSTREM1SSION THERAPY FOR ADULTS WITH ACUTE MYELOGENOUS LEUKEMIA: THE ROLE OF ALLOGENEIC BONE M A R R O W TRANSPLANTATION (IBMTR RESULTS). MM Horowitz, T Bt~chner, RP Gale, MJ Zhang~ International Bone Marrow Transplant Registry, Medical College of Wiscousin, Milwaukee, WI, USA.

FIFTY-SIX CASES OF ACUTE MYELOID LEUKEMIA WITH INVERSION INV(16)(P13Q22): DO ADDITIONAL CHROMOSOMAL ABERRATIONS INFLUENCE PROGNOSIS?

There is controversy whether adults with acute myelogenous leukemia (AML) in first remission are best treated with further chemotherapy or an HLA-identical sibling bone marrow transplant. We addressed this issue using data for 1097 adults 16-50 years old with AML in first remission. Results of transplants from HLA-identical siblings reported to the International Bone Marrow Transplant Registry (IBMTR; N=901) were compared with results of chemotherapy in persons treated by the German AML Cooperative Group (GAMLCG; N=196). Preliminary analyses were done to determine variables (I) differing between the cohorts; and (2) associated with treatment outcome in one or both cohorts. Leukocyte count at diagnosis and age were identified as important potentially confounding variables. These variables were adjusted for in subsequent comparisons of treatment-related mortality (TRM), relapse and treatment-failure (leukemiafree survival [LFS]) between the cohorts. Additional adjustment was made for potential bias introduced by the waiting time for transplant, i.e. only patients surviving in remission tong enough to receive a transplant were included in the transplant cohort. Median (range) time from first remission to transplant was 3.3 (<1 to 17.1) mos. Adjusted five-year probability of TRM was significantly greater and five-year probability of relapse significantly less with transplants than chemotherapy. Five-year probability of LFS was greater for transplants than chemotherapy (46% [42-49%] versus 34% [26-42%]; P<0.01). These data indicate that bone marrow transplants from HLA-idenficat siblings result in greater TRM, fewer relapses and higher LFS than postremission chemotherapy in adults 16-50 years old with AML in first remission.

C. Schoch 1, Th. B0chner2, M. Freund 3, H. Link4, H. Bartels5, T. Haferlach 6, H. Lbffler6, C. Sauerland 7, Ch. Fonatsch 1 lAG Tumoroytogenetik, Medizinische Universit~t zu L0beck, Ratzeburger Allee 160, 23538 L0beck, 2Medizinische Universit~itsklinik A, M0nster,3Abt. H~imatologie/Onkologie der Univarsit~t Rostock, 4Abt. H~matologie/Onkologie, Medizinische Hochschule Hannover, 5St~dtisches Krankenhaus SOd, L0beck, 62. Medizinische Klinik, Universit~t Kiel, 71nstitut for Medizinische Informatik und Biomathematik, MOnster The pericentric inversion inv(16)(p13q22) represents one of the most common structural rearrangements in acute myeloid leukemia. It occurs in 2% to 7% of patients with acute myeloid leukemia. Nearly all cases show a morphology consistent with AML FAB group M4 with eosinophilia. Patients, who show an inv(16) in their leukemic blasts have a high remission rate and probably the longest survival compared to other AML entities. Usually inv(16) is the sole cytogenetic abnormality, but some patients show chromosome aberrations in addition to inv(16). The influence of additional chromosome aberrations on the clinical outcome of patients with inv(16) is unclear. We analyzed 56 cases of acute myeloid leukemia carrying a pericentric inversion 16, 53 occurred de novo, 3 were observed secondarily after treatment with cytotoxic drugs for a primary tumor, Clinical follow-up data were available for 51 patients. Seven patients (14%) died within one month after diagnosis of AML due to infection or bleeding, 20 (39%) are alive 3 to 109 months after diagnosis (median folllow-up 36 months). Additional chromosome aberrations were observed in 15 patients (26%). Seven of them showed an additional chromosome 8, in 2 patients a gain of chromosome 22 was observed. Different structural rearrangements in addition to inv(16) were found in 5 patients. In all three patients suffering from secondary AML additional chromosome anomalies were observed. Secondary clonal chromosome abnormalities in patients with a pericentric inversion 16 had no negative influence on prognosis.

74

76

AUTOLOGOUS BONE MARROW TRANSPLANTATION VERSUS C H E M O T H E R A P Y IN A D U L T A C U T E M Y E L O B L A S T I C L E U K E M I A (AML) PATIENTS (PTS) - A R G E N T I N E A N STUDY R E S U L T S ~ S. P A V L O V S K Y and I. FERNANDEZ,r from A R G E N T I N E GROUP OF T R E A T M E N T OF A C U T E L E U K E M I A (GATLA) and ARGENTINE G R O U P OF BONE M A R R O W T R A N S P L A N T A T I O N (GATM0). Buenos Aires, Argentina. PURPOSE: To compare the impact of autologous progenitor cells transplantation (APCT) versus s t a n d a r d dose i n t e n s i f i c a t i o n (SDI) in A M L pts in first complete remission (CR), in event free survival (EFS) a n d overall survival (OS). PATIENTS A N D METHODS: A total of 43 pts registered at G A T M O w h o r e c e i v e d an A P C T b e t w e e n 6/91 and 5/94 w e r e c o m p a r e d w i t h 52 pts who r e c e i v e d a SDI r e g i s t e r e d at G A T L A f r o m 3/86 to 3/94. All the pts w i t h SDI h a d m o r e than 5 m o n t h s in CR to be eligible. Pts of b o t h groups w e r e m a t c h e d b y age. M e d i a n age of b o t h groups was 37 years, range 3-65 for A P C T and 3-64 for SDI. Seven of b o t h groups w e r e u n d e r 15 years old. T w e n t y one w i t h A P C T and 25 with SDI w e r e males. The ablative regimen was the classical B u s u l f a n 1 6 / m g / k g in 4 days plus C y c l o p h o s p h a m i d e 120 m g / k g in 2 days. Pts who need 2 induction treatment to a c h i e v e d CR r e c e i v e d in a d d i t i o n VPI6 2400 m g / m 2 in 36 hrs CI. RESULTS: At p r e s e n t 18 out of 43 pts w i t h A P C T h a d an event (4 treatment r e l a t e d deaths [TRD] a n d 13 relapses). No r e l a p s e was o b s e r v e d after 20 months. Of the 52 pts w i t h SDI 34 had an event (3 TRD and 31 relapses) w i t h relapses until 43 months. At 36 m o n t h s remains e v e n t - f r e e and alive 47 % and 50 % v e r s u s 32 % and 35 % of those treated w i t h A P C T and SDI r e s p e c t i v e l y (p=0.130 and. p=0.356). CONCLUSIONS: Pts who r e c e i v e d an A P C T h a d b e t t e r EFS and 0S than SDI. This is m a i n l y due to lack of relapse a f t e r 20 m o n t h s in the group w i t h APCT. The small n u m b e r of pts and short follow-up p r o b a b l y p r e c l u d e to have statistical differences.

CHROMOSOME ABNORMALITIES ANDKARYOTYPE EVOLUTIONINTHE STEM CELL COMPARTMENTS OFAMLANDMDS

D. Haase, M. Feuring-Bnske, S. KOnemarm, W. Verbcok, C. Tmff, C. Fonatsch, W. Hiddemann, and B. WOrmann University of G6~ngea, Dept. of Heraatol./Oncol.;'Robert-Koch-Str. 40, 37075 GOttingen, Germany The phenotype of the (pre)leakemic stem coil in AML and MDS has not been identified as yet. We have investigated whether cells with a stem coil-like immunophenotype (CD34+/CD38-) are involved in the leukemogenic process. In 12 patients with AML and in one patient with MDS the identification of (pro)leukemic cells was performed by classical cytogenetics. Highly purified (96 to 98% purity) cellular subpopulations were generated by fluorescence activated ceil sorting according to the expression of CD34 and CD38 (with increasing maturity: CD34+/38-, CD34+t38+, CD34-/38+, CD34-/38-). Following a short time incubation with a cytokine cocktail (EPO, G-CSF, GM-CSF, IL-3, SCF) the sorted subpopdations were processed for chromosome analysis. In 9/12 AML patients clonal karyoty~ abnormalities were observed in the unsorted material ( 2q+ [M1], -7 [M2],+g [1VI2],6p- [M4], +4 [M4], inv(16) [3 x M4Eo]. inv(16),+g,-12 [M4Eo]). In 719 cases the chromosomal changes found in unsorted specimens were also detected in the immature stem coil-like population (CD34+/CD38-). In 919 cases the kaiyotype abnormalities were diagnosed in the popniadon of committed progenitors (CD34+/CD38+). Additional, secondary abnormalities (+8, -12) were also demonstrable in both subpopulations in a case of M4Eo with inv(16). In the patient with a secondary MDS RA cytogenetic analysis revealed a mosaic of normal and abnormal cons with 5 different dependent cell clones (I: 5q-; II: 5q,17p-;III: 5q-,17p-,der(17)(p?); IV.' 5q-,17p-,derl7,-20;V: 5q-,-7,17p-,derlT) in the unsorted bone marrow specimen. Two subpopulations, CD34+t38- and CD34+/38+ were sorted together (CD34+138+-) because of the rarity of available cells. All abnormal cog dunes found in the unsorted material were also demonstrable in the CD34+, sorted cell population. The relevanco of our findings for pathogenesis, autologons stem cell transplantation and targeted gene therapy will be discussed.

Al14 79

77

KARYOTYPIC A B N O R M A L I T I E S IN S E C O N D A R Y L E U K E M I A OF CHILDREN* J.HARSO'i-r 1, I.REINISCH-BECKER 1, J.RI'FrERBACH 1, U.CREUTZIG 2, W _D.LuDWIG 3, P.GUTJAHR 4, A.8ORKHARDT 1, F.LAMPERT 1

1)CHILOREN'SUNIV. HOSPITAL,FEULGENSTRASSE12, 35385 GIESSEN,GERMANYj.2)CHILDREN'SUNtV* HOSPITAL, D-48129 MONSTER, GERMANY;3)ROBERT'ROSSLE-KLINIK,DEPT. UNCOLOGy/MOLECULAR BIOLOGY,13122BERLIN,GERMANY;4)CHILDREN*SUNIV.HOSPITAL,55101 MAINZ,GERMANY;

By the use of prolonged intensive chemotherapy combined with radiation a high cure rate in childhood malignancies has been achieved. In consequence to this improvement, however, the incidence of secondary and therapy related leukemias has increased. Within a period of ten years the bone marrows of 27 children with secondary leukemias were cytogenetically analyzed. Eighteen children had ANLL, 7 ALL, and 2 MDS. Two of the patients (1 AUL; 1 ANLL) had received a kidney transplantation, whereas the others developed their secondary leukemia during a time period between 1 month and 20 years after treatment of different solid tumors, lymphomas, or leukemias. The majority of the children (21/27) showed an aberrant leukemia karyotype, most frequently involving breakpoint 11q23. In the MDS patients only the MDS-specific changes (-7 and 7q-) could be detected. Five of 7 children with ALL showed one of the subtype specific abnormalities, whereas only one of these abnormalities was found in ANLL. In 4 patients of the latter group a chromosome 11-involvement was found, whereas chromosome 5 which is described to be typical for adults was seen only twice. The cytogenetic results will be compared with outcome, clinical data, and given drugs. ..................................................

*) Supportedby the "Parents' LeukemiaResearch Fund GieBen" and the "ForschungshilfeStationPeiper"

S Tosi 1,5, j Harbott2, M Fitchett3, F Ross4, A Biondi5, L Kearney 1 1. MRC Molecular Haematology Unit, Institute of Molecular Medicine, John Radcliffe Hospital, Oxford, UK 2. Oncocytogenetic Laboratory, Children's University Hospital, Giessen, Germany 3. Department of Medical Genetics, Churchill Hospital, Oxford, UK 4. Wessex Regional Genetics Laboratory, Salisbury, UK. 5. Clinica Pediatrica dell'Universita' di Milano, Ospedale S. Gerardo, Monza, Italy Abnormalities of the long arm of chromosome 7 are commonly observed in acute myeloid leukemia (AML) and myelodysplastic syndrome (MDS). These often involve deletions of 7q. We applied fluorescent in situ hybridization (FISH) using a chromosome 7 paint and a 7q tetomere specific probe for a better characterization of 7q abnormalities in a series of patientc w{th ncute leukemia. Of the six patients studied, five had structural abnormalities of 7q involving band q22" four of these were deletions and one had additional material on 7q. Only one patient had a breakpoint of 7q31. Four patients had a complex karyotype: only two had chromosome 7 aberrations as the sole abnormality. The use of FISH revealed the presence of an interstitial deletion in only two patients analyzed so far. A further two patients with reported 7q deletions were shown to have unbalanced translocations, with additional material present on 7q. The remaining two patients with other abnormalities of 7q were also identified as having unbalanced translocations. We are continuing these studies in conjunction with FISH using locus specific probes from 7q to define the critical deleted region involved in abnormalities of 7q in acute leukaemia. This work was supported in part by the Associazione Italiana per la Ricerca sul Cancro (AIRC)

80

78 Akute Leukemia VI Prognostic strategies F e b r u a r y 25 - M a r c h i, 1995, M Q n s t e r DELETION OF CHROMOSOME 6p AS ANOMALY

Molecular Cytogenetic Analysis of 7q Abnormalities in Acute Leukemia

IN

SECONDARY

ACUTE

factors

PRIMARY

NONLYMPHOCYTIC

and

treatment

KARYOTYPIC LEUKEMIA

Simone M e t z k e (w M. H e r o l d ($), G. A n g e r ($), and U. Claussen(w (w I n s t i t u t e of H u m a n G e n e t i c s and A n t h r o p o l o g y , D - 0 7 7 4 0 Jena ($) D e p a r t m e n t of Internal M e d i c i n e , K l i n i k u m Erfurt, D - 9 9 0 1 2 Erfurt Structural rearrangements on the short arm of chromosome 6 with involvement of p21-pter have been reported repeatedly in therapyrelated acute nonlymphocytic leukemia (ANLL) or myelodysplastic syndrom (MDS). Most often the 6p rearrangements have been observed in complex karyotypes including aberrations of chromosomes 5 and 7. Only occasionally 6p rearrangements were described as single anomaly. Here we report on a 43-year-old male, who was treated for Hodgkin's desease(HD), nodular sclerosis, stage IIh with combination radio-chemotherapy. He was treated with six cycles of cyclophosphamide, vinblastine, procarbazine prednisolone, doxorubicin, vincristine, bleomycin and involved field radiation (25 Gy). At first the patient achieved only a partial remission however, after six cycles of CLMP, including CCNU, leukeran, methotrexat and prednisolone complete remission was achieved. He relapsed one year later and then was treated with six cycles of CLMP and involved field radiation (20 Gy). He achieved complete remission but developed a secondary ANLL-M2 30 month after therapy for HD. Cytogenetic analysis of the bone marrow at time of diagnosis of ANLL showed a deletion 6(p21-pter)in 15 out of 22 cells, which occured as a primary and single cytogenetic event. The observed deletion on the short a r m of chromosome 6 was terminal, suggesting that probably only one break happened. The breakpoint p21 is located within the region involved in 6p rearrangements which are reported in therapy-related hematological malignancies. This suggests that the involvement of 6p in complex 6p rearrangements or 6p translocations is the specific karyotypic anomaly. Therefore we conclude that a gene is located in 6p21-6pter that is responsible for the development of therapy-related hematological malignancies.

Unusual karyotype 46,XX, inv (8) (p21;q24), del (20) ( q l l ; q l 3 ) in children's acute leukemia M6

S. Donskaya 1, S. Andreeva2, A. Galanta2, L. Sclyarenko 3, N. Derbenyova I We present a report about an unusual case of acute leukemia M6 in a 6years old girl. The initial clinical data consisted of hepatosplenomegaly and mild cutaneous hemorrhagic symptoms. Blood analysis before treatment did not show anemia (Hb level 10, 6-11, 9 g/d0, the range of thrombocytes was 34-72x10/1. The conclusion after cytomorphologic and cytochemical bone marrow investigation was acute mydoid leukemia M6. Cytogenetic analysis of the bone marrow indicated: 46,XX, inv (8) (p21;q24), del (20) (ql I;q13), 8 % of the metaphases were near tetraploid. The disease was resistent to chemotherapy according to a modified BFMprotocol (combination of Adriamycin, Vincdstin, Prednison and AraC in medium doses), palliative treatment with low dose AraC and mercaptopurin gave moderate clinical improvement. Peculiarity of this case in our opinion is the combination in the same clone of an abnormality registered in the M6-variant of AML [inv (8) (p21;q24)] together dd (20) (ql 1;q13) which was described in Polycytemia vera. Thus, oth abnormalies could be connected with pathologic erythroblnsts and may cause initial resistencr to chemotherapy. ] Kicv Institute of Medical Training 2 Kiev Institute of Hematology and Blood Transfusion 3 ILE, Kavetsky Institute of Experimental Pathology, Oncology and Radiobiology Academy of Sciences, Ukraine

Al15

81

83

Acute myeloblastie Leukemia M2 with the 8;21 translocation associated with grauulocytic sarcoma E. Dubrovina, O. Kryzhanovsky, A. Maschan, E. Samochatova, L. Baydun, O. Bartseva Institute of Children Hematology of Russia, Moscow

ADDITIONAL CHROMOSOME ABNORMALITIES IN PHILADELPHIA CHROMOSOME POSITIVE ADULT ACUTE LYMPHOBLASTIC LEUKEMIA (ALL): MORPHOLOGIC, IMMUNOLOGIC, MOLECULAR GENETIC AND PROGNOSTIC CORRELATIONS Rieder, H. ~, Ludwig, W.-D. 2, Gassmann, W. ~, Thiel, E. 4, L6ffler, H. 2, Bartram, C.R. s Maurer, J. 4 Janssen, J.W.G. s, G6kbuget, N. e, Hoelzer, D. s., and Fonatsch, C. 1

Not more then several tens of patients with AML, M2 variant, with the 8;21 translocation [t(8;21)] who developed granulocytic sarcomas or chloroma, are described. There are only 18 such cases in children. Review of previously documented cases suggests that this association is more common than generally recognized. Patient Y..K., a 13 year girl, presented with 2 compact painless subcutaneous tumors of 3 cm on the scalp in June 1993. AML FAB M2 was diagnosed. A tumor in the spinal chanel (L1-L2) was also revealed by MRI. Cytogenetic analysis showed two clones: clone 1 46 XX (20 %) and clone 2 46 XX t(8;21) (q22;q22) and t(1;17) (q21;p13) - 80 %. A complete remission was achieved after treatment according to BFMAML-87. 3 months later bone marrow relapse occurred. A CT scan disclosed tumor masses in uterus and right ovary. After intensive treatment with courses with HD AraC, VP-16, mitoxantrone, amsacrin a second remission was achieved and autoBMT was performed. A maintenance treatment with 6MP together with immunotherapy with IL-2 is continuing up to now. The case described reflects the supposition that patients with AML M2 t(8;21) should be observed closely for signs of granulocytio sarcoma. If it is present they have to be treated as patients with high risk of relapse.

AG Tumorcytogenetik, Institut for Humangenetik, Medizinische Universittit, Lebeck; 2 Robert-R6ssle Klinik, MDC, Bedin; 3 Innere Medizin II, St~idt. Klinik, Kiel; 4 Innere Medizin m.S. H~imatologie/Onkologie, Universitatsklinik Benjamin Franklin, Bedin; s A b t P~diatrie I1, Sektion Molekularbiologie, UniversitSt UIm; Abt. Hamatologie/Onkologie, Zentrum Innere Medizin, Universit~it Frankfurt; Germany. * for the participants of the German Multicenter Adult Acute Lymphoblastic Leukemia (GMALL) Therapy Study In 64 adult patients with ALL a Philadelphia (Ph) translocation was identified. Sixty-two (97%) showed a standard Ph-translocaUon and two cases (3%) had variant Ph-translocations, one simple vadant and one complex variant. In 46 cases (72%) additional chromosome abnormalities were present. Among these, chromosome abnormalities leading to loss or rearrangement of the short arm of chromosome 9 (n=13; 20%), a hyperploid karyotype >50 (n=11; 17%), and monosomy 7 (n=9; 14%) were observed most frequently. Immunophenotyping was done in 61 cases and disclosed common-ALL in 51, pre-pre-B- in 3 and pre-B-ALL in 7 patients. In two cases with common-ALL the presence of a minor compartment with POX-positive blasts indicated a biphenotypic leukemia, one of these had monosomy 7 in addition. By RT-PCR the chimeric bcr-abl gene was not detected in two out of 41 patients investigated (5%), one case with standard Ph-translocation and one with a complex variant Ph-translocation. In the latter, the bcr-abl rearrangement was verified by fluorescent in situ-hybridization. In a preliminary study survival of patients with hyperploid karyotypes >50 seemed to be superior to that of patients with monosomy 7 or 9p-abnormalities. This work was supported in part by the Deutsche Krebshilfe

82

84

CYTOGENETIC SUBGROUPS IN CHILDHCOD ACUTE MYELOBLASTIC LEUKEMIA (AML) IN A NATIONAL CENTER OF PEDIATRIC ONCOLOGY DEPARTMENT IN ISRAEL B. Stark*,** M. Jeison**, R. Gobuzov**, I. Kadyshevitch**, I.J. Cohen*, I. Yaniv*, Y.Goshen*, H. Tamary*, S. Ash*, J. Stein*, S. Fischer*, R. Zaizov*,** *Pediat Oncology Depart. **Ca. Cytogenetic Laboratory, children's Medical Center of Israel, Beilinsom Medical Center, Sackler Faculty of Medicine, Tel Aviv University, Israel. From July 1988 to Oct. 1994, 39 out of 44 newly diagnosed children (age 1 mo to 17 yrs) with AML, were analyzed cytogeneticaly. Clonal aberrations ware detected in 32 patients (80%). In 4 patients with Therapy-related leukemia, t(9:ll) (in 2 pts) and add(llpl5) occured following topoiscm~rase II inhibitors, ,while inv(3q) ,-7 appeared following alkylating agents. Among the 28 De-Novo AML patients, t(8:21) was the most prominent aberration, found in 8 pts (28%). It was followed in frequency by the known aberrations; involving 3q in 4 pts [two with t(3:5) one with t(3:21)], involving iiq23 in 3 pts [t(9:ll), t(l:ll), add(llq23)] and t(15:17) in 2 pts. Other specific chromosomal translocations, such as t(6:9) or its variant, was found in 2 pts (one of them with preceeding MDS), del(9q) in 2 pts, and t(8:19) variant of t(8:16), t(10:ll:12), +8, -7 in one patient each. Comparing our results to those reported in children from other centers, indicate excess of t(8:21) while no inv(16) could be detected. Morover, rare specific translocations could be found in this small series. This finding may suggest a geographic/ethnic variation in childhood AML subtypes, and should be confirmed in a more extended population. Correlations with clinical and blast cell characteristics and induction success will be presented.

Do Cdntradictory Cytogenetic Results at Initial Diagnosis and in Relaps Indicate SeCondary Leukemia? IIse Chudoba*, Simone Metzke*, Klaus Blumenstengela,Johannes HermannA Kerstin BtumenstengelQ, Hans-J6rg Fdcken, Ke~tin Junker*, Cordula Bleck*, Klaus H6ffken~ Uwe Claussen* *lnstitut fer Humangenetik und Anthrepologie, aKlinik fur Innere Medizin und ~Klinik fLir Kinder- und Jugendmedizin tier Fdeddch Schiller Universit~t, Jena In acute leukemias as well as myeledysplastic syndromes the diagnosis of cytogeneticafly completely different results at different acute phases of the disease is a recurrent phenomenon. The questions adsing from these findings are whether cytegenetic investigations are not sensitiv enough to detect the malignant cell clone in certain stages of the disease or whether a relaps with different chromosome findings might be another disease, comparable to leukemias developing after non leukemic neoplasias as therapy-related secondary leukemias. AIItogether we present six cases with unexplained differences. Two cases revealed karyotypic evolution, two further cases had aberrant karyotypes at diagnosis and normal cytogenetic results in hematological relaps, and two cases showed completely different abnormal karyotypes in relaps compared to the initial diagnosis. Numerical aberrations diagnosed initially and disappeadng in relaps were analyzed by means of Fluorescence-ln-Situ-Hybridisation (FISH). This method provides a higher sensitivity and the cell population examined was enlarged by cells with no mitotic activity. The same procedure was applied to cases with karyotypic evolution. The stored cell suspensions of bone marrow preparations from earlier taken samples with normal karyotypes were reexamined with FISH after diagnosis of aberrant cell lines in relaps. In two cases a change of the immunophenotype during the course of the disease has bee observed, which suggests therapy-related secondary leukemia. Still there are two other cases where at initial diagnosis and at time of relaps different abnormal karyotypes were found. This indicates that different cell clones were involved in these malignancies. In these cases, however, the cytological and immunophenotypic data remained the same at different stages of the disease. The other two cases, one with a trisomy 21 as the sole abnormality and another one with a trisomy 21 as a part of a highly Complex karyotype at the initial diagnosis, were examined with FISH. In both cases 200 interphases were analyzed, and the initially diagnosed cell line could not be detected in later bone marrow samples. In these cases an involvement of different cell lines in the malignant disorder at different times of the disease seems to be most likely.

Al16 85

87

Cytogenetic Findings of Acute Leukemias in U k r a i n e S. Andreeva 1, S. Donskaya 2, A. C,-atantha 1, O. Ryzak 2, A. Nadgornaya 3

DETECTION OF A N EXTRAMEDULLARY RELAPSE OF ACUTE MYELOID LEUKEMIA (AML) BY FLUORESCENCE IN SITU HYBRIDIZATION

From April 1992 to August 1994 bone marrow or peripheral blood samples o f 25 children with acute ieukemias were analysed. 21 eases (84,0 %) were determined as acute lymphoblastic leukemia (ALL), 4 (16,0 %) - as acute myeloid leukemias (AMI.,). Clonal abnormalities were found in 16 (76,2 %) children with ALL and in 4 children with AML. Besides qualitative and quantitative abnormalities which were characteristic for different variants o f disease we noted an unusually high frequency o f near tetraploid metaphases. 6 children (28,6 %) with ALL had 20,9 % near tetraploid metaphases ranging from 10,4 % to 50,0 %. 3 cases with AMI., had 23,0 % tetraptoid metaphases ranging from 8,0 % to 41,6 %. Similar findings were found in 6 adults with acute leukemia. 3 of'them were determined as AML and 3 as ALL. One patient with M3-A.ML had karyotype 46,XX, t(15;17)(q22;qll) and t l , 0 % - near tetraploid metaphases. All cases with ALL had 34,7 % near tetraploid metaphases ranging from 6,0 % to 84,2 %. Clinical significance o f this polyploidy is under investigation, however w e could assume that its high frequency may be caused by the special ecological situation in Ukraine. 1 Kiev Research Institute o f Hematology and Blood Transfusion 2 Kiev Institute o f Medical Training 3 R.E. Kavetsky Institute o f Experimental Pathology, t h e o l o g y and Radiobiology, Academy o f Sciences o f Ukraine

K. Pompe, J. Krauter, H. Heimpel and G. Hell Department of Internal Medicine Ill, University of UIm, 89081 UIm, Germany Fluorescence in situ hybridization (FISH) allows the detection of several numerical and/or structural chromosomal abnormalities using DNA sequences specific for either the centromeric region of certain chromosomes or for structural aberrations. In 1992 a 26 year old male patient was admitted with the first diagnosis of a de-novo AML, FAB M 1. A trisomv 8 could be demonstarted by FISH analysis in 70% of the peripheral blood mononuclear cells. For the detection of the chromosomal aberration a FITC-labelled alpha-satellite DNA probe (Oncor, Braunschweig), specific for the centromere region of chromosome 8, was used. A complete remission (CR) was achieved by a standard induction therapy with daunorubicin, cytosin-arabinoside and etoposide (DAV). Postremission therapy included one course of high-dose ara-C (3 g/m 2, 12 doses) and daunorubicin (30mg/m 2, 3 doses). Sixteen months after achievement of the CR t w o subcutaneous nodules of 0.5 cm in diameter were detected in the skin of the patient's left arm. A t the same time the patient presented with normal peripheral blood counts and no evidence for a bone marrow relapse was found by morphological analysis. The nodules were removed and slide imprints of the excised material were performed. FISH analysis of these cells revealed a trisomy 8 in 70% of the cells, thus proving an extramedullary relapse of the AML. In addition, FISH analysis of the bone marrow mononuclear cells revealed a trisomy 8 in 2-3% of the cells. These data confirm, that FISH analysis of AML blasts is a rapid and sensitive tool for the detection of chromosomal abnormalities, which can provide decisive diagnostic information with respect of an extramedullary manifestation and/or bone marrow involvement of the disease.

86

88

Detection of the Most Frequent Numerical and Structural Chromosome Aberrations in Acute Myeloid Leukemia by Fluorescence In-Situ Hybridization Using Yeast-Artificial-Chromosomes.

Detection of minimal tumor cell populations by FICTION

K. Fischer, S. FrOhling, C. Scholl, R. Sehl[nk, M. SchrOder, M. Bentz, P. Lichter, H. D0hner. Specific chromosome aberrations are one of the most important prognostic factors in acute myeloid leukemia [AML] and are now increasingly being used for selecting postremission therapy. A prerequisite for such studies are rapid and reliable methods for cytogenetic analysis. Fluorescence in situ hybridization (FISH) has greatly improved our ability to detect specific chromosome aberrations, since interphase cells can be analyzed. The objective of our study is to design DNA probe sets, mosdy consisting of yeast-artificial-chromosomes [YACs], that allow the detection of all AML-relevant numerical and structural chromosome aberrations. At present, the following aberrations are being identified by interphase cytogenetics lDNA-clone; chromosomal Focalization]: -5/5q- [yPR411 recognizing the FMS gene; 5q33]; -7/7q- [HSC7E161; 7q22]; § [935_A_12; 8q26}; 9<]-[253_F_8; 9q13-q21}; translocations involving band 11q23 [13HH4 spanning the MLL gone; 1lq23]; +l 1 [D11Z1; #1 [ centromere]; +13 [phage pool recognizing the RB-I tumor suppressor gene; 13q14]; 17p- [cosmid pool recognizing tbe p53 tumor suppressor gene; 17p13]; 20q- [D20S99; 20q]; +21 [2D3; 2Iq21]; +22 [DI07Fg; 22q11]; and-Y [DYZ3; Y-centromerel. Probes for the detection of other AML-sl~ecific translocalion~, such as the 1(8;21). are under development. So far, 32 pts with adult AML entered on a multicenter treatment trial [AML HD93] have been investigated using this probe set. Single aberration on Gbanding, n=4: only one (5q-) of d~eseaberrations was ideutilied by FISH; two pts. had an inv(16), and another had an ins(2;3) that could not be detected by the probe set. Complex aberrations on G-banding, n=10: in all pts., the aberrations were fully or partially detected; in 3/10 cases, where the karyotype could I)ot be completely evaluated due to poor chromosome morphology, the clone was further described by F[SIL Normal karyotype on G-banding, n=10: No aberrations were found by FISH either. No metuphases on G-banding, n=8:3/8 pts exhibited clonal aberrations by FISH; one pt. had three 935_A 12 alguals ["t~isomy8"7, a second pt. had Ihree 935_A 12 and five 253 F_8 signals [complex karyotype], and a third pt. had a del(20q). In conclusion, FISH using the current DNA probe set allowed the detection of ciot~alaberrations in almost 80% of pts. with an abnormal karyotype on baz~dinganalysis and provided a tool for the identification of clonal aberrations in pts. in whom banding analysis was not successful. Beyond its application at diagnosis, the probe set will provide ,.he basis for monitoring the remission status and for detecting residual disease in autologous peripheral blood stem cell and autologous bone marrow gr~'ts by FISH. Medizinische Kllnik and Poliklinik V, University of Heidelberg; and Deutsches Krebsforschungszentrum, Heidelherg, Germany.

Department of Human Genetics, University of Kiel K. Weber-Matthiesen, J. Deerberg, B. Schlegelberger Sensitive techniques for the detection of minimal tumor cell populations are gaining in importance. In the field of hematological oncology, for example, they are required in monitoring residual disease after therapy, detection of minimal bone marrow infiltration by malignant lymphoma cells and exclusion of tumor cell contamination after leukapheresis. PCR is the most sensitive technique, however, it is only applicable if the tumor cells have defined structural chromosome aberrations. Interphase cytogenetics (fluorescence-in-situ-hybridization = FISH) may be considered for tracing tumor cells with numerical chromosome aberrations. However, small tumor cell populations ( < 1%) cannot be detected by FISH, because artificial hybridization signals found in up to 1% of normal cells may falsely pretend aneuploidy. The solution to this problem could be the new FICTION technique, a combination of fluorescence immunophenotyping and interphase cytogenetic analysis. By means of FICTION, cells are searched selectively on the basis of tumor cell associated immunophenotype, The selected cells are studied simultaneously regarding the hybridization signals. Thus, numerical chromosome aberrations are evaluated only within a few cells with a certain tumor cell associated immunophenotvpe. For example, residual c-ALL blasts in peripheral blood after therapy can be monitored by detecting numerical chromosome aberrations in CD 1O-positive cells. Within the selected cell population the percentage of tumor cells may be relatively high. This selective evaluation is the equivalent of physical tumor cell enrichment by means of magnetic beads or FACS. This FICTION approach has the potential to overcome the detection limit of FISH. FICTION-studies can be done on cytocentrifuge slides, smears and touch preparation. The results are available within 24 hours. In cases, where PCR is not applicable, FICTION appears to be a promising alternative to detect obscure tumor cell populations. Here we demonstrate the powerful potential of FICTION to detect obscure tumor cell populations.

All7 89

91

HE)MOZYGOUS DELETION OF THE p16 GENE IN CHILDHOOD ACUTE LYMPHOBLASTIC LEUKEMIA (ALL)

TEL is fused to a previously unknown gene, STL, in a t(6;12)(q23;p13) translocation in a patient with acute lymphoblastic leukemia

Taku Seriu, Seisho Takeuchi, Carl W. Miller, Andreas ToNer, Johannes W. G. Janssen, Alfred Reiter, Wolf-Dieter Ludwig, Martin Zimmermann, J0rg Schwaller, Elizabeth Lee, Isao Miyoshi, H. PhilUp Koeffler, and Claus R. Bartram.

p16/ MTS1 (multiple tumor suppressor)/ CDK41 (cyclin-dependent kinase 4 inhibitor) is a putative tumor suppressor gene involved in carcinogenesis in a wide variety of tissues. To study the involvement of the p16 gene in leukemogenesis, we performed Southern blot analyses in childhood ALL. DNA samples from 112 primary ALL (22 T-ALL, 90 precursor-B ALL) revealed homozygous deletions in 30/112 cases (27%), and showed a significant association with the T-lineage [17122 T-ALL (77%) versus 13190 precursor-B ALL (14%); P<0.001]. Additionally, hemizygous deletions were observed in 2 T-ALL and 9 precursor-B ALL. None of 19 cases with p16 deletions showed evidence of germline alteratiorl~s. This study demonstrates that homozygous loss of the p16 gene represents the most frequent genetic alteration in childhood ALL, and suggests that inactivation of p16 is a critical event in leukemogenesis.

Yoshimasa Suto, Yuko Sato, Steven D. Smith, Janet D. Rowley, Stefan K. Bohlander The University of Chicago, Chicago, IL, USA. Reciprocal transiocations affecting band 12p13 are found as recurring chromosomal changes in myeloid and lymphoid malignancies. These translocations occur between band 12p13 and a number of other chromosomal bands (e.g.: 2p11, 2q11, 5(:133, 7q32, 9p11, 16q13, t7p11, 20qll and 22q12). The t(5;12)(q33;p13) from a patient with chronic myelomonocytio leukemia resulted in the fusion of the platelel-derived growth factor receptor-I] gene from 5q33 and TEL from 12p13. TEL (Translocation Ets Leukemia) is a new member of the ETS family of transcription factors which seems to be involved in most of the balanced 12p13 translocations. A cell line derived from the leukemic cells of a patient with acute lymphoblastic leukemia (ALL) thal contained a t(6;12)(q23;p13) translocation was used to clone the fusion partner of TEL from 6q23. Two different types of fusion cDNAs were characterized. The first type had 5' TEL sequences fused at cedon 55 of TEL to a previously unknown gene on 6q23. The sequences from this novel gene (designated STL: Six-Twelve Leukemia) were apparently derived from the 3' untranslated region, because there were multiple stop codens in all three reading frames. The resulting fusion gene product would be a truncated TEL lacking both the HLH domain and the ETSDNA bindingdomain.The second type has TEL sequences from codon 337 at the 3' end and a short stretch of novel sequence with an open reading frame at the 5' end. The unrearranged TEL allele is deleted in this cell line, and thus there is no normal TEL transcript. These observations in conjunction with the fact that TEL is frequently hemizygously deleted and that it is more closely related to the negative transciptional regulator yan/pok from Drosophila than to other human ETS family members suggests that TEL may function as a negative regulator of cell growth.

90

92

DETECTION OF p 16 GENE DELETIONS IN HEMATOLOGIC DISEASES BY INTERPHASE FLUORESCENCE IN SITU HYBRIDIZATION. M.H. Drevling. O.1. Olooade and S.K. I~Qhlander. The University of Chicago, Chicago, IL Hemizygous and homozygous deletions of chromosomal band 9p21 have been detected in vadous tumor types as well as more than 20% of acute lymphoblastic leukemia. Recently, the CDKN2 gene (p 16, MT$ I) has been proposed as a candidate tumor suppressor gene because it is frequently deleted in cell lines derived from various tumor types. However, its role in pdmary neoplasias remains controversial. We analyzed 1 S hemat~ological neoplasias (9 ALL, Z AML, 4 NHL) with cytogenetically detected 9p rearrangement by fluorescence in situ hybridization (FISH) and molecular analysis. FISH probes were generated by a sequence-independent amplification technique (SIA) from yeast artificial chromosome (YAC) clones and a cosmid contig (Z50 kb) of the CDKN2 region. Using the cosmid contig, FISH analysis detected homozygous deletion of the CDKN2 region in 5 cases (3 ALL, Z AML) and hemizygous deletions in 6 cases (6 ALL). While interphase FISH analysis detected all homozygous deletions detected by molecular analysis, southern blot analysis did not detect the hemizygous deletions in sub-populations of cells. Therefore, interphase FISH analysis will be a sensitive tool to define the clinical significance of 9p deletions in hematological neoplasias.

DIVERSITY OF THE TCR6 GENE REARRANGEMENTS INDICATES SUBCLONE FROMATION IN ACUTE B CELL PRECURSOR LEUKEMIAS

Section of Molecular Biology, Department of Pediatrics II, University of UIm, UIm, FRG; Division of Hematology/Oncology, Cedars-Sinai Medical Center, UCLA School of Medicine, Los Angels, CA, USA; Department of Pediatrics IV, Hannover Medical School, Hannover, FRG; Department of Medical Oncology and Applied Molecular Biology, Robert-ROssle-Klinik, Free University of Berlin, Berlin, FRG; Central Hematology Laboratory, University of Bern, Bern, Switzerland; and Department of Internal Medicine, Kochi Medical School, Kochi, Japan.

E.R. Panzer-Grtimayer, D. Ghali, S. Panzer*, S. Fischer, A. Argyriou-Tirita, OA. Haas, H. Kovar, H. Gadner. Acute B cell precursor leukemias (BCP-ALL) have been demonstrated to be oligoclonal at the IgH gene level in up to 40% by Southern blot hybridization. In contrast, oligoclonality as deduced from diversity of the TCR6 gene rearrangements of the immature types has not been reported, so far. We detected in four of 20 childhood BCP-ALLs oligoclonality characterized by the coexistence of different junctional regions variying in size by three to 25 nucleotides of identical V62-D63 rearrangements. No variation was found in the lgH gene status. Two of the leukemias were also homogenous in the TCR,/gene. Two leukemias displayed the variants in an unequal proportion, in the other two remaining leukemias, for which similar quantities of the coexisting rearrangements were detected, single-cell nuclei PCR revealed two seperate leukemic populations. We therefore define these populations as subclones. The variants arose independently from each other as deduced from their individual sequences. We propose that in a BCP leukemic cell population TCR6 gene diversity arises after rearrangements of the IgH genes resulting in apparent clonality at the IgH gene level. However, cells are oligoclonal, if the TCR6 gene rearrangements are considered. Children'sCancer Research Institue, St. Anna Kinderspital, *Dept. of Blood group Serology,Univ. Vienna, Vienna,Austria,

All8

93

95

MOLECULAR

ANALYSIS

BREAKPOINT

AND

NOVEL

ACUTE

OF THE

FUSION

CHROMOSOMAL

TRANSCRIPTS

LYMPHOBLASTIC

SEM

IN T H E

CELL LINE

W I T H C H R O M O S O M A L T R A N S L O C A T I O N t(4; 11 ) R. Marschalek, J. Greil$, K. L6chner, I. Nilson, G. Siegler, I. Zweckbronner, J. D. Beckr and G.H. Fey Department of Genetics, University of Erlangen-Nuernberg, Staudtstr. 5, 91058 Erlangen, Germany, SDepartment of Paediatrics, University of Erlangen-Nuernberg, Loschgestr. 15, 91054 Erlangen, Germany. The chromosomal breakpoint and break-region transcripts of the novel pre B leukemia-derived SEM cell line carrying a t(4;11) translocation were analyzed. The chromosomal breakpoint from derivative chromosome der4 was cloned and sequenced. The crossover site was localized in intron 7 of the ALL-1 gene on chromosome 1 lq23 and in a large intron of the AF-4 (FEL) gene on chromosome 4. RNA transcripts from both wild-type genes and both hybrid genes were detected by reverse transcriptase polymerase chain reaction (RT-PCR) assays. In addition, alternatively spliced mRNA species derived from the der4 chromosome were found. They were generated by using the exon 5' of the breakpoint on der4 as a common splice donor site and the 5' boundaries of exerts 8 or 9 of the ALL-1 gene as alternative splice accepter sites. The hypothesis is proposed, that selective pressure operates to maintain the presence of both derivative chromosomes as important elements in the leukemogenic process.

94 MOLECULAR GENETIC DETECTION OF CHROMOSOMAL ABNORMALITIES AT 11Q23 IN PATIENT,S WITH DE NOVO AND SECONDARY ACUTE LEUKEMIA-A.Borkhardt, R.Repp, S.Schlieben, J.Ha~rbott, S.Brettreich, M.Mitteis, F.Lampert, Children's University Hospital Giessen, 35392 Giessen,Feulgenstr.12, Germany We recently reported that it is possible to detect four different 1 lq23 abnormalities by multiplex-PCR (Repp et al., Leukemia, in press). Meanwhile, we expanded the scope of our method to four additional chromosomal translocations involving region 1 lq23. Using an upstream primer located in exon 5 of the MLL gene at 1iq23 and eight downstream primers specific for AFlp, AF4, AF6, AF9, AF10, AF17, ENL, and AFX, respectively, we were able to identify the molecular equivalents of t(1;1 1), t(4;11), t(6;11), (t9;11), t(10;1 I), t(11;17), t(11;19) and t(X;11). Permanent leukemic cell lines expressing the corresponding fusion mRNA's have not been established for all of these abnormalities in question. We, therefore, designed a protocol in order to synthesize chimeric RNA molecules for use as positive controls in our RT-PCR assays. The method is based on recombinant PCR technology in combination with the socalled "solid-phase in vitro transcription". Using the described multiprimer PCR we retrospectively investigated 50 patients with de novo and secondary leukemias as well as a panel of permanent leukemic cell lines. Various rearrangements of the MLL gene were detected in 24 of these patients or cell lines, respectively. Moreover, sequence analysis revealed variants in the exact chromosomal breakpoints. Our approach enables the assessment of particular clinical and prognostic parameters associated with different MLL rearrangements. supported by "Forschungshilfe Station Peiper"

SEQUENCE ANALYSIS OF THE GAP-RELATED DOMAIN OF THE NF1 GENE AND ALL THREE RAS PROTOONCOGENES.IN PATIENTS WITH SECONDARY ACUTE LEUKEMIAS S.Brettreich, A. Borkhardt, J. Hammermann, A-.K. Borkhardt, R. Repp, F.Lampert, Children's University Hospital Giessen, 35392 Giessen,Feulgenstr.12, Germany We investigated mutations in the GTPase ac:ti~ating domain of the Neurofibromatosis 1 gene (NF1) as well as in the N-ras, Kras and H-ras protooncogene in 11 patients with secondary acute leukemia. Eight patients suffered from secondary acute myelogenous leukemia after previous chemo-and radiotherapy for acute lymphoblastic leukemia (4), non-Hodgkin lymphoma (1), Ewing Sarkoma (1), hepatoblastoma (1) or Wilms-Tumor (1), respectively. Three patients had a secondary acute lymphoblastic leukemia after previous therapy for a neuroblastoma, astrocytoma, or rhabodomyosarkoma, respectively. After isolation of genomic DNA and selective PCR amplification for codon 12, 13 and codon 60, 61 of all ras genes as well as the so-called FLR exon of the NF1 gene, all PCR products were directly sequenced in both directions. Sequencing was performed by the didesoxy-chain termination method using an "Prism T7 terminator sequencing kit" from Applied Biosystems (ABI, Foster City, California, USA). We found one point mutation in K-ras codon 13 (GGC to GAC) leading to an aminoacid exchange from Gly to Asp. The entire FLR exon of NF1 as well as all other ras fragments were found to be of wild type sequence. In conclusion, we assume that the NF1 gene does not seem to be involved in development of secondary leukemias. In contrast, the ras protooncogenes might be in an acivated state in a small subset of secondary leukemias. * s u p p o r t e d b y " F o r s c h u n g s h i l f e Station Peiper"

96 MOLECULAR GENETIC SCREENING OF CHILDREN WITH ACUTE LEUKEMIA BY RT-PCR* S. SCHLIEBEN, [. REINISCH-BECKER, A. BORKHARDT, F. LAMPERT, J. HARBOTF " ONCOGENETICLABORATORYjCHILDREN'SUNIVERSITYH~PITAL, FELILGENSTR.12, 35392 GIlES*SEN

A variety of oncogenes are activated by specific chromosomal translocations and induce leukemogenesis. Some of them are of clinical relevance because of their prognostic meaning. The t(9;22), t(4;11), and t(9;11) are characterized by a poor prognosis whereas t(8;21) or inv(16) seem to be more favorable. An increasing number of these abnormalities can be detected by PCR. The genes rearranged in these translocations are BCR and ABL (t(9;22)), MLL and AF4 (t(4;11)), MLL and LTG9 (t(9;11)), AML1 and ETO (t(8;21)) and CBF6 and MYH11 (inv(16)). We used polymerase chain reaction (PCR) to detect BCR/ABL-, MLIJAF4-, MLL/LTG9, AML1/ETO-, and CBFS/MYH11-rearrangements9 In a prospective study 1065 bone marrow samples of children with acute leukemia were analyzed9 The incidence of BCR/ABL-rearrangement was 3.1% in all de novo c-ALL and pre-B-ALL (17/544), whereas in relapse patients the BCR/ABL-transcript was found in 8.0 % (10/115). Analyzing 18 patients with pre-pre-B-ALL or infants we detected 5/14 infants and 1/4 older patients with a MLL/AF4 rearrangement. The fusion transcript AML/ETO was detected in 4/10 patients with ANLL-M2, -M1. All cytogenetically t(9;11) positive patients with ANLL-M5 showed the corresponding MLL/LTG9 rearrangement, respectively inv(16) positive patients with ANLL-M4eo showed the CBFB/MYH11-rearrangement. The screening of ANLL-M5 and ANLL-M4eo patients for the corresponding rearrangements was therefore started. 9supporteoby the DeutscheKrebstliHe- MildredScheelStiftung,Parents-lnibativeGie~enandtheFo~schungshilfeSta6onPeiper.

Al19 97

99

THE WILMS' TUMORGENE(WT-I) lS FREQUENTLYEXPRESSEDIN ACUTE MYELOBLASTICLEUKEMIASANDMAYPROVIDESENSITIVEDETECTIONOF RESIDUALBLASTCELLSBY PCR.'~ L.BergmamL J.Brieger, E.Weidmann, U.Manrer, D.HoeLzer and. P.S.Mitrou Medical Clinic III, Hematology/Oncology, J.W.Goethe University, Frankfurt/M, FRG.

DETECTION OF MINIMAL RESIDUAL DISEASE (MRD) ON BONE MARROW SMEARS BY REVERSE TRANSCRIPTASE POLYMERASE CHAIN REACTION (RT-PCR) Schoch, R.+ Jelnsch S.,* Haferlach. T+. L6ttler, H.+, Mtiller-Ruchhohz. W.,* " Gassmann. W.+: SecondDep. of Internal Medicine (+) an(t Institute of Immunology(*). Christian-Albrechls-Universityof Kiel. Germany The polymerase chain reaction (PCR) is now oRen used for the identification of the three most common chromosomal aberrations t(8;21 ), t(l 5; 17) and inn(16) in adult AML. In addition, the high sensitivity of this technique has provided the possibility to detect residual leukemia at previously unobtained levels. For the application of primary diagnosis one bone marrow aspirate smear is sufficient for the RT-PCR assay. This source has the advantage, that no special handling and transport of the specimen is necessary. The aim of the present study was to detect minimal residual disease on bone marrow smears and to compare the results with the detection of the fusion transcript in bone marrow aspirates. From twenty patients with an AML in complete clinical and haematological remission (8 pts with t(15;17), 5 pts. with t(8;21) and 7 pts. with inv(16)) a bone marrow smear and three m~ of the bone manow aspii'ate were invest;gated. Farm the glass slide smear RNA was dissolved directly on the glass slides in a buffer containing guanidinium thiocyanate and Phenol (RNAzoI| The bone marrow aspirate was isolated on FicolI-Hypaque and RNA was extracted from the mononuclear cells. For both test groups PCR was performed with one primerset for a normal gene (RARe, AML or J3-Actin) and one for the fusion gene with a nested system. The normal gene could be detected from the glass slide smear and the bone marrow aspirate in every case. The specific translocation as a marker for minimal residual disease was observed in 9 out of 20 cases from the bone marrow smear. In the control with the bone marrow aspirate in only 6 out of 20 cases the specific translocation could be detected. These results demonstrate, that bone marrow smears could be a suitable source for a RTPCR- based detection of minimal residual disease and seem to be superior to the use of bone marrow aspirates. Further studies should estimate this possibility.

The Wilms' tumor, gene (wt-l) is a tumor suppressor gene generally expressed in developing fetal kidneys, gonads and in Wilms' tumors. Recently some cases of AML have been described to express wt-1 mRNA. This study was designed to determine the frequency ofwt-1 gene expression in acute leukemias and its applicability as genetic marker for the detection of blast cells. Leukemic cells from 83 patients with AML and four leukemia derived cells fines were examined for the expression of wt-1 mRNA. As normal controls peripheral blood or bone marrow derived MNC of healthy donors were used. Cells were isolated from bone r~arrow or peripheral blood. Total RNA was extracted and wt-1 transcription was studied via RT-PCR. In dilution experiments the detection limit was assessed as 1 out of 10000 cells. Wilms' tumor mRNA was detectable in 67/83 cases of AML (81%). Only one of the examined cell lines was wt-1 negative. None of the 13 controls expressed wt-l. The expression of wt-1 nkRNA was not associated with response, phenotype or FAB-classification. After chemotherapy, 18/23 patients in CR lost wt-1 expression completely. In 6 pts. wt-1 mRNA reappeared and was associated with a following relapse in5/6 pts. All patients with persisting wt-1 mRNA were non-responders or developed early relapses. Studies for the evaluation of the relevance of wt-1 as a prognostic factor and as an early marker of relapse are in progress. So far, the data suggest the persistence or occurrence of wt-1 in CR to be associated with an early relapse. We assume that analysis ofwt-1 gene-expression via PeR may be a sensitive technique for the detection of residual blast cells in AML. *supportedby DeutscheKrebshiffegrant W22/92

98

100

CBFg/MYH11 - REARRANGEMENT IN M4Eo WITH INVERSION 16 A NOVEL MARKER FOR RT-PCR-MEDIATED MINIMAL RESIDUAL DISEASE DIAGNOSIS. E.Koller1, M.PfeilstOcker1, B.Pavlova 1, H.M0filberger ~, H.Nowotny~, A.Zoubek2, H.Kasparu3, R.Heinz ~, E.Pittermann~, HKarlic1. 13rd Medical Department and Ludwig Boltzmann-lnstitute for Leukemia Research and Hematology, Hanuschhospital, Vienna; 2St.Anna Kinderspital, Vienna;3KH Elisabethinen, Linz, Austria.

Evaluation of Chimerism after Allogeneic Bone Marrow Transplantation (BMT) using Amplification by Polymerase Chain Reaction (PCR) of Hypervariable Regions of the Human Genome

The inversion of chromosome 16(p13q22) is a characteristic cytogenetic marker for acute myelogenous leukemia sutype M4Eo. Recently the fusion gene product of this pericentric translocation was characterized (Pu Liu et el., Science 1993). The CBF6 gene is translocated from 16q in the region of MYH11, a smooth muscle gene on the short arm of chromosome 16. This may result in an altered regulation of transcription and subsequent leukemic transformation. We wanted to evaluate the value of this rearrangement as a potential new target for MRD-diagnosis which is routinely carried out by RT-PCR. BM and PB-samples of 6 patients with M4Eo were analysed at diagnosis, in clinical remission and relapse. The detection of inv(16) was always correlated with a CBF6/MYH11 rearrangement. 5 patients showed a type A transcript, one patient a type B fusion. PCR- negativity in remission was observed in one patient who is relapse-free for more than 20 months. Samples of 3 other patients investigated in clinical and cytogenetic remission remained PCR-posilive, 2 patients relapsed so far. we conclude that the CBF6/MYH11 rearrangement is highly sensitive, residual leukemia can be detected during remission. The value for MRD-evaluation needs confirmation on a larger study of prospectively analysed patients.

Demidova IA, Surin VL, SavehenkoVG, LubimovaLS, Meadeleeva LP We used PCR as novel and sensitive technique to evaluate posttransplant chimedsm. Polymorphosis in tandemly repetitive core sequences of regions with variable number of tandem repeats (VNTRs) served as genetic markers of donor and recipient cells. For our study we selected the following VNTR regions: A poliprotein B3' hypervariable region, DX $52 locus of X chromosomeand VNTR in intron 40 in von WiUebrand factor gene. 8 patients (pts) underwent BMT for acute leukemias and their donors were tested using these systems. 4 pairs doner/reeipient were available for study. Highmolecular-weightgenomic DNA was extracted from whole blood sample and bone marrow according to the methods previously described. In one ease DNA was extracted from the patient's granulocytes isolated by the Ficoll density eentrifugation (this patient was treated by donor lymphocyte transfusions for posttransplant relapse). Recipient and donor pretransplant DNA and recipient posttransplant DNA at varying intervals after BMT were amplified. The resultant fragments were analyzed after gel electrophoresis. Sensitivity of the method was determined by mixing various proportions of recipient and donor DNA; the limit of detection of the minor componentin a mixture was 1%. Evaluation of 4 available cases indicated complete ehimerism (3) and mixed ehimerism (1). In all cases of complete chimerism pts demonstrated complete remission of acute leukemia. In the ease of mixed chimera we detected both donor and recipient DNA since day +28 after BMT. On day +90 the patients relapsed and was treated with donor lymphocyte transfusion. Mixed chimerism was detected during all courses of treaUnent. The patient died of relapse at the day +120 post BMT. PCR amplification of VNTR for evaluation of ehimerismwas comparedwith the method of red blood cell phenotyping and demonstratedthe advantages over the latter such as high sensitivity, use of small amounts of blood or bone marrow, ability of early detection of posttransplant engratWaeat. That's why it seems to be the most useful method for long-term observation of chimefism dynamics in pts after BMT. National Research Center for Hematology, Novozykovsky 4a, 125167 Moscow, Russia

A120 101

103

CYTOKINE PRODUCTION IN ADHERENT LAYERS OF LEUKEMIC LONG TERM BONE MARROW CULTURES AND IN PASSAGED

Insulin-like growth factors (IGF)-I,-II and insulin-like growth factor binding proteins -2 and -3 serum levels in children with leukemia. P.Vorwerk~, K.Mohnike1, U.Kluba ~, V.Aumann~, W.F.Blum2, U.Mittler~

NORMAL

BONE

MARROW

FIBROBLASTS

DETECTED

BY

INTRACELLULAR IMMUNOFLUORESCENCE M.Ya. Lisovsky, G.Yu. Miterev, V.G. Savchenko Deficiency in stromal miercenvironment can possibly contribute to inhibition of normal bemopoiesis ha acute (AML) and chronic (CML) myeloid leukemias. Recently we showed that the formation and hemopoietic supportive ability of adherent stromalcell layers (ASCL) in long term bone marrow cultures (LTBMC) of patients with CML and AML was impaired. The aims of this investigationwere: a) to evaluate the production of some stimulator,/ and inhibitory eytokinesby ASCLs of LTBMCs of patients with AML and CML: b) to evaluate the production of r and some proteins of eytoskeleton and of extracellular matrix by normal passaged bone marrow fibroblasts (PBMF), which are known to haveno ability to support long-term haemopoiesis in vitro. Cultures were established from bone marrow aspirates obtained from 4 normal donors of bone marrow transplant, 3 patients with chronicphase Ph-positive CML and 3 patients with AML at the time of presentation. Detection of G-, GM--CSF,SCF, IL-I, IL-3, IL-6 and TGF-~3 in trypsinized acetone-fixed ASCL cells of LTBMCs and in PBMFs of third passage was porformcdby indirect semiquantitafive immanoflnorescenceanalysis using moaocloeal antibodies. Compared to ASCLs of normal LTBMCs, stroma of CML LTBMCs from all three patients was deficient in IL-I production. In ASCLs ofAML LTBMCs produetiun of the majority of cytokines was variably decreased. Normal PBMF were overtly deficient in G-CSF production and synthesized increased amounts of TGF-13. Flow cytomeUy confirmed the lack of or negligible production of G-CSF by PBMF. There was no difference between PBMF and stromal cells of normal LTBMCs in expression of a-smooth muscle aetin, collagen 1V and laminin. Opposite to it,, PBMF, but not normal stromal cells, were strongly positive for CD68. The results demonstrate variable alterations of eytokine production in stromal cells of CML and AML LTBMCs aad in normal PBMF. These alterations may be responsible for the lack of haemopoieticsupportive ability of studied stromal cell populations and hence for inhibitionof normal haemopoiesis in CML and AML. Present address: Dept. of Hematological Oneology and BMT, Hematological ScientificCenter, Novozykovsky pr. 4A, 125167 Moscow, Russia

The insulin-like growth factor (IGF) signaling pathway is of major importance for the proliferation of different tumors (e.g. breast cancer and Wilms' tumor). The bioavailability of both ligands IGF-I and IGF-II is regulated by specific IGF binding proteins (IGFBPs). IGFBP-2 is the predominant binding protein during fetal life, where it is expressed in most tissues. In contrast, postnatally it is mainly released by specific cell types (hepatocytes, astroglia, kidney, prostate) and a range of tumor cell lines. Furthermore PHA stimulated normal lymphoblasts and malignant lymphoblasts express IGFBP-2. In order to investigate the IGF regulatory pathway we determined serum levels of IGF-I, IGF-II IGFBP-2 and IGFBP-3 by ILIA in 28 leukemic children at diagnosis. Whereas serum levels of IGF-I (mean/range:-2,7/-0,5 to -6,7 SDS), IGF-II (-3,6/-1,3 to -8,7 SDS) and IGFBP-3 ( -2,0/+0,85 to -7,1 SDS) were significantly decreased, IGFBP-2 levels (+4,0/-0,45 to +7,4 SDS) were found to be elevated and inversely correlated to IGF-I (r=-0,51, p=0,013). After the hematological remission all 4 parameters had normalized in the 16 reinvestigated children. Similar findings have been observed in one boy with a relapse including CNS leukemia. This study demonstrates that decreased serum levels of IGF-I, IGF-II and IGFBP-3 are present in spite of the proliferation of malignant lymphoblasts (at diagnosis vs. treatment). It further indicates that serum levels of IGFBP-2 may be directly related to the proliferation of lymphoblasts. aotto-von-Guericke Universit~it Magdeburg, Klinik fur P~idiatrische H~imatotogie und Onkologie D-39112 Magdeburg, Halberst~idter Str. 13 ~Universit~itskinderklinikTiibingen, Abtl. Allg. P~tdiatrie und Onkologie

102

104

INTt~RLEUKIN (HI-l, IL-2, IL-4, IL-6) AND TNF cc PRODUCTION Ilq CHILDREN WITH ACUTE LYMPHOBLASTIC LEUKEMIA (ALL)

INFLUENCE OF TGF-BETA AND IFN-ALPHA ON APOPTOSIS OF B-LINEAGE ALL

A. Chybicka, J. Salwa, J. Bogustawska-Jaworska, Cz. Radzikowski, W. Jaworski. Department of Children Hematology and Ontology, Department of Children Surgery, Wroclaw University of Medicine, Department of Tumor Immunology, Institute of Immunology and Experimental Therapy, Polish Academy of Science, Wroclaw, Poland. This study was undertaken to establish the role of cytokines (IL-l, IL-2, IL-4, IL-6 and TNF in the pathogenesis and the clinical course of childhood ALL:Total number of 150 children with ALL, 88 boys and 62 girls, aged from 0.5 to 15 years were included to the study. TNF production was studied in supernatants deriving from 24 hours PBSC mononuclear culture ace. the method based on growth inhibition of sensitive to TNF a 929 mice fibroblasts, IL-I production ace. method based on inhibition of autologous rosette formation by thymocytes of CBA mouse and IL-2, IL-4, IL-6 production ace. to conventional ELISA Genzyme-test. Thirty seven healthy children served as the control group. The TNF and IL-1 production decreased after starting chemotherapy of ALL, while relative increase of IL-6 was observed. It was found that in children with ALL during the whole period to therapy the IL-1 production, was significantly lower than that observed in the control group of healthy children (p 0.005). No IL-2 and IL-4 production was observed in children with ALL during the 2,5 years period of observation. After cessation of the therapy IL-1 and TNF production grew up, while IL-6 production remain on the same level. During 10 years period after the end of therapy IL-1 median values didn't reached values of the control group. EFS at 54 ruth in ALL children with ]I,-1 production < 10p. before therapy was higher than that in children with IL-I < 10p (EFS 90% v 70%). EFS at 60 ruth in ALL children with not detectable TNF production was longer than that observed in children with detectable TNF production (80% vs 71%).The IL-1 and TNF production could be a useful prognostic maker.

C. Buske, M. Feuring-Buske, K. SchreibeL W. Hiddemann, B. W6rmann Department of Hematology and Oncology, University of G6ttingen, Robert-Koch-Str. 40, 37075 G6ttingen, Germany Prognosis of patients with B-lineage ALL has no longer significantly improved during the past five years. New approaches may be based on identification of growth-inhibitory signals for malignant pre-B cells. Mononuclear calls of 11 patients with the diagnosis of cALL and 2 patients with pre-pre-B ALL were cultured with routine fibroblasts as feeder cells, alone or plus TGF-beta (10 ng/ml) or IFN-alpha (1000 U/ml). Cell viability was tested by trypan-blue exclusion. Apoptosis and percentage of calls in S-phase were assessed by propidium-iodidestaining. In 9 of 12 samples TGF-beta induced apoptosis with a maximum increase of 20 % compared to the controls. Also in 9 of 12 samples TGF-beta reduced cell viability with a maximum of 85 % of the call count compared to the controls. Seven of 11 samples showed a reduction of cells in S-phase up to 50 % of the control when incubated with TGF-beta. IFN-alpha induced apoptosis in 5 of 9 samples with a maximum increase of 15 % and reduced celt viability in 4 of 9 samples. Only in 3 of 9 patients a loss of cells in S-phase was observed. All experiments were characterized by a great interindividual variability of results. The data identify TGFbeta and IFN-alpha as effective inhibitory cytokines for B-lineage ALL. The identification of negative regulators in B-lineage ALL may have direct implications for new therapeutic strategies.

A121 105

107

Abstract: Acute Leukemias VI: Prognostic factors and treatment strategies CYTOKINE'SERUM LEVELS DURING THE COURSE OF THE DISEASE IN ACUTE MYELOID LEUKEMIA M. Stauch, D.Fritsche and K. H6flken. Dep. of Internal Medicine II, Div. Oncelogy-Hematology,FSU Jena, 07740 Jena.

G M - C S F Dose R e s p o n s e Progenitors for Ara-C

Cytokines are a group of hormone-like regnlatory pnlypeptides. Several investigators have shown that cytokines which are essential for in vitro proliferation and differentiation of normal hemat~poietie cells, may also be involvedin the process of malignanttransformation of bematopoistic cells. The role of the in vivo production of eytokincs in the transformational process of acute myeloid leukemia (AML) remains uncertain. We studied the serunalevels of IL-1, IL-113,IL-2, soluble IL2 receptor (IL-2R), IL-6, tumor necrosis factor-alfa fl'~NF~), granulocyte-maerephage coleny-stimulating factor (GM-CSF), grunulocyte colony-stimulating factor (G-CSF) and erythrepoietin (EPO), using enzyme-linked immunosorbentassays, in 45 patients with de novo AML, during hematological reconstitartion after induction and consolidation chemotherapy, during maintenance treatment, 12 weeks after treatment, and in 40 normal persons. We found a significant increase of the serum levels of TNF-et, IL-6, G-CSF, GM-CSF, EPO and IL-2R in patients with de novo AML compared with the control group, in these patients with AML, a clear correlation was found betwe~a the elevated serum concentrations of this cytokines and the Creactive protein (CRP). The elevated serum level of EPO correlated with decreased haemoglobin and haematocrit. The increased serum concentration of the soluble IL-2R did not correlate with the absolute number of blast cells but correlated with the n'amber of PMN of the peripheral blood. After induction chemotherapy,the serum levels of IL6, G-CSF, EPO and IL-2R remained increased. IL-6 and G-CSF correlated in fltis phase of the disease with the serum CRP. In addition, IL-6 showeda positive correlation to the absolute number of blast cells and a negative correlation to the PMN of the peripheral blond.There were normal cytokine serum levels 12 weeks after treamaent. Our data show that the incerese of the endogenous cytokines studied in patients with de novo AML was not only due to the underlying disease. However, EPO, IL-6 and G-CSF seems to be involved in the hematological reconstitutionaider chemotherapy.

Curves

in

Priming

of

AML

M. Ziihlsdorf V. Skobin, and Th. Biichner Div. of Haematol. & Oncol., Dept. of Int. Med., University of MUnster A.-Schweitzer.-Str. 33, D-48 129 Mtinster Myeloid leukemic progenitors can be primed for ara-C cytotoxicity by haemopoietic growth factors such as GM-CSF. This effect is highly variable for different AML samples. Previous studies have used highly saturating concentrations of GM-CSF for priming. In this study, we sought to establish dose response curves for GM-CSF for AML samples. Up to date, 4 samples are evaluable. AML blasts were isolated by FicollHypaque and adherence depletion. Pretreatment with 0, 1, 10 and 100 U/ml of GM-CSF was done for 48h in IMDM, 10% FBS. Ara-C was added at 0 to 100BM for 12h. Cells were washed 3 times and seeded into a serum-free colony assay based on methylcellulose. Leukemic colonies were scored on day 14. For each GM-CSF dose, a D50 for ara-C was calculated using the median effect principle. These D50 values are summarized for the 4 samples in the table. GM-CSF (U/ml) #1 #2 #3 #4 0 8.25 7.34 39.7 11.,3 1 4.83 3.36 7.32 8.95 10 4.19 ne 1.67 6.35 100 1.05 2,94 0.14 3.84 A dose response relation was found for each sample. However, the extent of priming is highly variable for different samples. A half maximal effect could be reached between 1 and 10 U/ml of GM-CSF.

106

108

EXPRESSION OF FLT3 AND FLT3-LIGAND IN A PANEL OF HUMAN LEUKEMIA-LYMPHOMA CELL LINES. G. MeierhofP. W. Dirks t, Z.B, Hu1, U. DehmeP, Q, Rosne~, H.J. Grassa & H.Q, Drexler t. t DSM-German Collection of Microorganisms & Cell Cultures, Human & Animal Cell Cultures, Braunschweig, G~rmany; 2 INSERM U.119, Marseille, France; 3 Immanex Res. & Develop. Corp., Seattle, WA, USA. The FLT3/FLK2 (human/mouse) rec~tor tyrosine kinase is closely related to three receptors, e-KIT, c-FMS, and PDGF-R which function with their respective ligands, stem cell factor, macrophage colony-stimulating factor and plateletderived growth factor, to control cellular growth/differentiation. Another acronym for FLT3 is STK-1. The ligand for FLT3 (FL) was cloned and found to encode a transmembrane and a soluble protein. Here, we examined the expression of FLT3 and FL in a large panel of human lenkemia-lymphoma cell lines at the mRNA level using Northern blotting. Total RNA was prepared using the guanidinium isothioeyanate/CsC1 method. RNAs were separated by agarose gel electrophoresis and specific bands we,re visualized by hybridization with :ep_ labeled cDNAs and autorediography. The cell lines were assigned to different categories based on their origin and phenotypic expression of cellular characteristics. The results for FLT3 mRNA expression were as follows for 95 cell lines examined: 10113 pre B-cell were positive; 1/6 B-cell; 0/10 myeloma; 0/11 T-cell; 0/2 NK cell; 15/20 myeloid; 8/15 monocytic; 1/10 megakaryocyticerythroid; and 0/8 Hodgkin. Clearly, FLT3 is expressed mainly by pre B-cell, myeloid and monocytic leukemia cell lines. A previously suggested association of FLT3 expression with CD34 positivity could not be confirmed in the panel of cell lines investigated. Northern blots showed widespread expression of the FL mRNA transcripts: 7/13 pre B-cell; 5/6 B-cell; 7/10 myeloma; 10/11 T-cell; 2/2 NK cell; 14120 myeloid; 7/15 monoeytie; 9/10 megakaryoeytie-erythroidand 3/8 Hodgkin. The intensity of the signal was in most samples rather weak. 22195 ceil lines expressed both receptor and ligand mRNA suggesting a possible autocrineparacrine stimulation. BHK-21 cells were transfected with FL; supernatant from an FL producing done was harvested. Out of 6 FLT3+ cell lines (2 lymphoid, 3 myeloid, 1 monocytic) only one cell line showed a proliferative response. In summary, we have identified the expression of FLT3 and its ligand FL in a panel of leukemia-lymphoma cell lines (covering all major hematopoietic cell lineages and various stages of differentiation). Further functional studies are underway to identify and modulate the effects of this new cytokine on hematopoietic cells.

Different Potential of G-CSF, GM-CSF and PIXY in Priming of Myeloid Leukemic Progenitors for Ara-C

M. Ziihlsdo~ M. Brand, V. Skobin, and Th. Biichner Div. of Haematol. & Oncol., Dept. of Int. Med., University of MUnster A.-Schweitzer.-Str. 33, D-48 129 Mtinster Cell kinetic or metabolic resistance of myeloid leukemic cells to cycle-specific cytostatics, such as cytosine arabinoside (ara-C) may be overcome by priming with haemopoietic growth factors. The extent of priming with GMCSF and IL-3 have been highly variable for myeloid leukemic progenitors in previous stvdies. We tested 7 differem AML samples for their ara-C sensitivity and the priming effects of G-CSF, GM-CSF, GM-CSF plus IL-3 as well as pIXY321. Cells were isolated on Ficoll-Hypaque, adherence depleted and washed. All factors were used at 100 U/ml for 48h, controls were incubated in medium only (IMDM, 10% FBS). Ara-C was added for 12h at concentrations from 0 to 1001.tM. Cells were washed 3 times and seeded into a colony assay based on methylcellulose, with serum-free conditions using solubilized cholesterol and linoleic acid, BSA, transferrin, insulin, 13-mercaptoethanol and a mixture of 7 haemopoietic growth factors. After 14d, colonies were scored. From the dose response curves for ara-C, a D50 was calculated using the median effect principle. The D50 values were compared for different pretreatments, showing shifts in ara-C ~ensitivity due to priming. Pretreatment control G-CSF GM-CSF .GM-CSF+IL-3 vIXY321 samples n 7 7 6 5 5 median DSO 0.178 0.050 0.120 0.046 0.017 x-fold change 3.58 1.48 3.91 10.7 D50 values for ara-C covered a range of 1 log, in previous studies over 2 logs with a larger sample size. Priming with various factors decreased the D50 values for the majority of samples with median shifts as given in the table. The largest shifts were obtained with pIXY321 or the aequimolar mixture of GM-CSF and IL-3 at a factor of 1:1500. Samples unresponsive to priming with one growth factor responded to another factor. Therefore, priming with a combinatic, n of factors, triggering different signal transduction pathways, might be more promising.

A 122 109

111

An alternative transcript of the human GM-CSFR alpha lacking the signal sequence is found in AML-blasts and normal controls. HM Waaner, RE Gale*, DC Linch* and R Andreesen Klinik und Poliklinik for Innere Med. I der Universit&t Regensburg, Abt. H&matologie/Onkologie, *Dept. of H~aematology, University College London Medical School Blast cells from 70% of patients with acute myeloid leukaemia (AML) show partial or total autonomous growth characteristics. One of the autocrine growth factors involved is GranulocyteMacrophage Colony-Stimulating Factor (GM-CSF). The receptor for GM-CSF (GM-CSFR) is composed of at least 2 chains. The alpha chain binds GM-CSF specifically but with low affinity. The beta chain confers high affinity binding when associated with the alpha chain. So far at least 6 alternative transcripts of the alpha chain have been described which differ from each other in the 3' end of the coding sequence. Using RT-PCR of RNA from normal bone marrow mononuclear cells and blast cells from patients with AML we have found an mRNA transcript with a deletion of 102 nucleotides (nt) between nt 124 and 225 of the published sequence. This deletion removes the original initiation codon and the signal peptide sequence. Translation would still be possible starting from ATG (nt 282-284) creating an alpha chain lacking the signal peptide and the first 22 amino acids of the extracellular protein sequence. In RNAse protection assays transcripts carrying the 102 nt deletion were expressed at levels of about 20% of the total alpha chain message in normal bone marrow cells. In blasts from AML patients expression varied between 5 and 60%. Full length GM-CSFR alpha RT-PCR products from a patient with AML showed that transcripts carrying the 102 nt deletion also lack the sequence coding for the transmembrane domain. Translation of this mRNA transcript would therefore produce a signal peptide deficient soluble GM-CSFR alpha isoform that might be confined to the cytosol. It is possibe that such a cytosolic receptor chain could be involved in intracellutar autocrine loops.

HUMAN ALLOGENEIC NK-.CELLS ACTIVATED BY IL~2 AND II.,-12

EFFECIENTLY LYSE FRESH ACUTE LEUKEMIA ~BLASTS. B. Glass, M. Zeis, L. Uharek, T. Gaska, J. Steinmann, W. Gassmann, H. L6ffier and W. Mueller-Ruchholtz. Institute for Immunology and Department of Internal Medicine II, University o f Kiel. We have recently shown that the transfer of allogeneic natural killer cells in the context of cellular immunotherapy is able to cure mice from leukemia after allogoneic BMT. In humans, however, thg high resistance of leukemia blasts to NK-cell-mediated lysis may be an obstacle to this type o f therapy. In this study we investigated the role of IL-12 (known as natural killer stimulatory factor, NKSF) and IL-2 in the generation of potent activated allogeneic NK cells against fresh acute myeloid (n=10) and lymphatic leukemia (n=4) blasts. Method: PBMC from healthy donors were separated for removal o f T cells by using anti CD3 MoAb coupled to immunomagnetic beads.This enriched NK-cell fraction (about 65% CD56+) was treated with various concentrations of IL-2 (100 U/ml to 1000 Ulml) and IL-12 (1 U/ml to 100U/ml) for 24 h at 37~ Cytotoxicity was measured against acute leukemias in a conventional 4h 51Cr.Releas e Assay. Results: Pretreatment of NK-cells with optimal doses o f IL-2 (I000 U/ml) resulted in a more than two-fold enhancement of cytotoxicity against leukemia blasts whereas the response to optimal doses of IL-12 (100U/ml) was slightly lower. The combination with IL-2 (500 U/ml) and I1-12 (100U/ml) most effectively activated the cytolytic capacity o f NKcells tested against myeloid and lymphatic leukemia blasts. NK-cells stimulated by the combination o f these eytokines augmented eytotoxicity for more than twofold compared to optimal IL-2 pretreatment alone. Conclusions: Our results indicate that such IL-2 and 11.,-12 stimulated allogeneie NK-cells may have interesting therapeutic implications in the context of cellular immunotherapy against leukemia in humans. Experimental studies evaluating this phenomenon in a murine leukemia model are now underway.

110

112

MARK EFFECT OF THE COMBINATION OF A NOVEL POTENT VITAMIN D3 ANALOG (KH 1060) WITH 9-CIS-RETINOIC ACID ON PROLIFERATION OF HL-60 CELLS. E. Elstner*, T. Umiel**L. Binderup***, I. Grillier*, J. Said****, and H. P. Koeffler*. Division of Hematology-Oncology*, Depts. of Medicine, Pediatrics** and Pathology****, UCLA, School of Medicine, Cedass-Sinai Medical Center; Los Angeles, CA, USA; Dept. of Biology, Leo Pharmaceutical Products, Ballemp, Denmank***. AIl-traas RA induces terminal differentiation of promyelocytes to granulocytes in acute promyelocytic leukemia (APL) and is the first highly effective differentiation-inducing agent for remission induction in patients with APL, but remissions are short-lived beeanse the treatment fails to eliminate the malignant clone and because clinical resistance develops rapidly. To prevent or overcome resistance and to eliminate the malignant clone, in analogy with chemotherapy, combination of new potent differentiation-inducing drugs working through different receptors and signal pathway may be useful. The active form of vitamin D3 [ 1,25 dihydroxyvitamin 1)3, (VD3)] is an inhibitor of proliferation ~ effector of differentiation of myeloid leukemic cells to monoeytes. The 20-epi-22oxa-25a,26a,27atri-homo-lct,25(OH)2D 3 (KH 1060) belongs to the family of 20-epi-VD3 analogs and is considerably more potent in vitro than VD3 as a regulator of growth and inducer of differentiation of the HL-60 cell line. The aim of this study was to evaluate the affect of the combination of KH 1060 with 9-cis-RA on proliferation of the human promyelocytic leukemia cell line HL-60. The 9-eis-RA ( 10-8 M) produced a 30 % inhibition of elonal proliferation of ilL-60, and KH 1060 at 10-9 M inhibited 50 % eloualTfnoliferation of the lenkemie cells. No colonies were detectable after incubation with 10- M KH 1060. The combination of 9-cis-RA (10-8 M) with KH 1060 (10-9 M)produced a 95 % inhibition of clonal growth. When the HL-60 cells were cultured in liquid medium with either 10-7 M of KH 1060 or 9-cis-RA for three days, washed and plated in soft agar, KH 1060 and 9-eisRA inhibited 71% and 25 % of the HL-60 clonogenic cells, respectively. In contrast, when the cells were eultured in liquid eulture with 10-7 M KH 1060 and 9-cis-RA together for three days, washed and plated in soft agar, no colonies were detectable. In order to gain insight into the remarkable antileukemie effect of the combination of these analogs, apoptosis ('cell suicide") and expression of bel-2 ("snicide-bloeking" gene) were examined. After 60 hours of culture of HL-60 cells with the combination of KH 1060 (10-7 M) and 9--eis-RA (10-7 M), apoptosis was induced in 72 % of cells, as detected by measurement of morphological changes. No apoptosis was detectable after cultivation with KH 1060 or 9-cis-RA alone. Also after 60 hours, bcl-2 protein, as analyzed by immunohistochemistry, became nearly undetectable when cells were cultured with both KH 1060 and 9-cis-RA (10-7 M). When cultured over the same period with either KH 1060 or 9-cisRA (10-7 M), 26 % and 36 % of cells expressed bcl-2; and 100 % of wild-type HL4i0 expressed bcl-2. These effects were observed before terminallyMifferentiated ceils wore morphologically detected. In sunaum% our data demonstrate that the combination of KH 1060 and 9--cis-RAsynergistically inhibits clonal growth of HL-60 cells as welt as induces apoptosis in HL-60 cells with a strong decrease of bcl-2.

IN VITRO GENERATION OF ALLOGENEIC MOUSE T CELL-LINES FOR LEUKEMIA THERAPY. L. Uharek, J. Steinmann, T. Sebens, B. Glass, M. Zeis, W. Gassmann, W. Mtiller-Ruchholtz. Institute of Immunology and Department of Internal Medicine II, University of Kiel We tried to establish allogoneic T cell-lines that are capable of mediating a cytotoxie effect against leukemic cells without attacking normal graft-recipient cells. This approach implies that the cytotoxic T lymphocytes (CTL) must be specific for tumor-associated antigens in the context of allogeneic major histocompatibility complex molecules. To provide evidence for the feasibility of this approach in the murine system, the following experiments were carried out: C3H (H-2k), C57 (H-2 b) and DBA (H-2 d) mononuelear spleen cells were primed against irradiated A20, WEHI-3 and PU cells, i.e. cells of three d hematopoietie tumors o f Balb/c ( H - 2 ) origin. All three tumors are proved to be nonimmunogenic in Balb/c mice in vivo and in vitro. After nineteen days the cytotoxic activities o f the generated CTL-lines were tested against the three leukemias in a SICr-release-assay. In all nine combinations the highest specific lysis was achieved against the stimulating tumor. As the three tumors are of the same origin, their different susceptibility against the different CTL-linas appears to be caused by tumor-associated features. In a second set of experiments with the same nine CTL lines an attempt was made to reduce the cytotoxie effect against the tumor cells by cold target inhibition. As expected the activity against the 51Cr-labelled leukemia cells could almost be totally inhibited by adding a sufficient amount of unlabelled native tumor cells. On the other hand it was impossible to inhibit the cytotoxic activity to the same amount by adding unlabelled Concanavalin A-induced Balb/c blasts. In a facs analysis we were able to prove that the different susceptibilities and inhibitive potentials of tumor cells and nonmalignant blasts cannot be caused by differences in the MHC expression o f the target ceils, as all three leukemias as the ConA blasts showed approximately the same expression of MHC (class I and II) molecules on their cell surfaces. These findings indicate that allogeneic CTL are capable o f discriminating malignant from nonmalignant cells, which implies that a T cell mediated GvL effect is not necessarily linked with GvH reactions. First experimental studies were initiated in order to investigate the antileukemic efficiency o f such allogeneie CTL in an immunotherapeutic setting.

A123 113

115

POTENT ANTILEUKEMICEFFECTS OF ALLOGENEICNK-CELLS GIVEN AFTER ALLONEEICBONEMARROWTRANSPLANTATIONIN MICE.M. Zeis, L. Uharek, B. Glass, T. Gaska, J. Steinmann, W. Gassmann, H. Lrffler and W. MuellerRuchholtz. Institute for Immunolgy and Departmetit of Internal Medicine II, University of Kiel. We previously demonstrated GvL effects independent from GvHD and correlated them to the donor's NK activity in different murine models. Aim of the present study was to determine the therapeutic potential of separated and ex vivo activated allogeneic NK cells given shortly after BMT. Method: lxl05 A20 leukemia cells (H-2 d, B-tymphoblastic neoplasm, nonimmunogeneic) were i.v. injected into Balb/e (H-2d) mice. After two days, the animals were treated with 7.5 Gy of TBI and received 2x107 marrow cells of syngeneic (Balb/c) or semiaUogneic (Fl(Balb/e x C57)) (H-2dxb) origin, Additional cellular immunotherapy consisted of injection of lxl07 F1 effector cells immediately after BMT. Effeetor cells were either untreated F1 cells or FI NK cells that had been incubated with 200 U/ml fi)r 24 h. NK cells were separated by immunomagnetic elimination of T cells, B cells and monocytes with antiCD3 and anti-MHC class II MoAbs coupled to dynabeads. Leukocyte counts and the percentage of NKI.I+, CD3+, CD4+, CD8+ peripheral cells were determined post-transplant. Post-mortem analysis of leukemia infiltration in spleen and liver was performed. Results: The immunomagnetie separation of NK cells and their activation with IL-2 increased the antileukemic activity in vitro more than 6-fold. The following relapse rates in vivo were observed: no treatment 100% (n=12, MST 26 days); syngeneic BM 82% (n=12, MST 34 days); F1 BM 70% (n=12, MST 43 days); semiallogeneie BM plus spleen cell injection 54% (n=13, MST 50 days9, F1 BM plus injection of IL-2 activated NK cells 10% (n=12, MST >120 days). Conclusion: (I) Immunomagnetic separation techniques can be used to obtain highly enriched NK cell fractions with preserved antileukemic activity, (II) Treatment with allogeneic IL-2 activated NK cells given post-transplant prevents leukemia relapse independent from GvHD.

Monitoring leukemic cells during the course of disease of an acute leukemia by Immunphenotyplng and genotyplng, K. Pachmann, H, Zabnienska, U. Goehly, J. U. Walther and B. Emmerich Medizinische Klinik und Kinderpoliklinik des Klinikums Innenstadt der Universltgd M0nchen Immunphenotyping, cytogenetics, restriction fragment length analysis, polymerase chain amplification and in situ amplification were used to detect and monitor the malignant clone in an acute lymphatic leukemia during the course of disease. The percentage of cALLA positive cells correlated well with the percentage of blasts detected morphologically. The concomittant decrease in cALLA positive, CD19 positive and HLADR positive cells in response to therapy indicated that these three markers were expressed on the malignant clone. The bcr/abl translocation was detected by cytogenetic analysis and it could be shown by Southern blot analysis that the abl gene was rearranged. The hybrid RNA as a result of the translocatlen could be- amplified by PCR and detected in individual cells of the leukemia directly on the slides using reverse transcription and subsequent incorporation of radioactive nucleetldes during amplification. Relocalization of the labeled cells and comparison with morphology, however, revealed that not all and not only cells diagnosed as blasts were labeled. The negativity of some blasts may be due to the fact that these cells do not transcribe the RNA although they do carry the translocatlon. On the other hand, cells with lymphoid appearance were labeled, too. This may indicate that cells of the leukemic clone still have limited maturational capacity. Lymphoid infiltration may have been more mature cells of the leukemic clone. Monitoring of the proportion of cells carrying the translocation at different times revealed labeled cells even at instances, when no leukemic cells were detectable either morphologically or by immunphenotyping. Thus this method can contribute to detect minimal amounts of leukemic cells.

114

116

CHARACTERIZATION OF NORMAL AND LEUKEMIC LYMPHOHEMATOPOIETIC PRECURSOR CELLS (LHPC) BY THREE-COLORFLOW-CYTOMETRY AND DETECTION OF CLONE-SPECIFIC LEUKEMIAASSOCIATED T-CELL-RECEPTOR 5 REARRANGEMENTS. F.Grlesinger, N. Kaib, S. KOnemann, B. Maecker, M. Feuring-Buske, I~1.Pirb-Noack, M. Falk, J. Ritter, Th. BOchner, L.W.M.M. Terstappen, W.Hiddemann, B. WOrmann. Dept.'s Internal Medicine,University of GOttingen and MOnster, Dept. Pediatrics, University of Mnnster, Becton Dickinson, San Jos~, CA, G6ttingen, Germany. Acute lymphoblastic leukemias (ALL) represent clonal expansions of malignant transformed LHPC. The molecular genetic basis of ALL is unravelling and progress has been made in deducing distinct differentiation stages of leukemic cells based on expression patterns of differentiation antigens (DAg). However, although normal LHPC compartments haw~ been defined functionally and immunologically, very few direct data exist on the leukemogenic "stem" cell in ALL. Its identification would lay the basis for targeted gene therapy as well as for exclusion strategies in autologous stem cell retransfusion.Therefore, the composite immunophenotype of diagnostic bone marrow mononuclear cells including the most immature progenitor cells was analyzed by three-color-flowcytometry in 218 consecutive ALL patients. Leukemia-associated asynchronous and or aberrant composite expression patterns of DAg were detected in >90% of ALL. These leukemia-specific immunophenotypes have been proved to be potentially useful for the prediction of relapse in 21 ALL (4 relapses, 17 CCR) and an update of follow-up marrows of additional patients will be presented. In an interim analysis of 32 c-ALL and 6 pre-pre-B-ALL, the median percentage of the most immature (CD34++CD38-DR-) LHPC was 0.03%, while CD34+CD38++DR+ or CD34-CD38+DR+ LHPC represented the predominant leukemic populations. Seven c-ALL carded a clone-specific leukemiaassociated genetic marker, i.e. a TCR 8 rearrangement. Highly purified CD34++CD38- and CD34+CD38++ LHP and CD3+ ceils (negative control) were isolated by FACS sorting. V52-D~53rearrangements were amplified by per from a minimum of 200 cells and hybridized to junctional region specific olignucleotides. Preliminary data suggest, that the clone-specific V62-Dc53 rearrangement can be detected in both the CD34++ CD38- as well as in the CD34+CD38++ compartments, while normal T-lymphocytes were negative. The results of additional sorted c-ALL will be presented at the meeting. If these data hold true, they wilt have an impact on autologous stem cell isolation and underscore the importance of designing strategies to discriminate leukemic versus normal cells within the "stem" cell compartment.

Flowcytometric detection of aneuploidies in Acute Lymphoblastic Leukemia (ALL) before and after chemotherapy U, Oelschl~igel 1, R. Nowak 1, R. Hofmann 1, H. Zengler 1, R. Siegert2 and G. Ehninger 1 We performed flowcytometric determination of DNA-content in immunophenotyped ALL cells. With this method one can increase the certainty of detecting aneuploidies compared with the common used one-parameter DNA-analysis. W e ' v e found aneuploidies in 30% (7 of 23) of ALL patients. The proposed method can be used for detecting residual aneuploid leukemia cells after chemotherapy. Therefore we carried out experiments in which w e ' v e mixed diploid bone marrow cells with the leukaemic aneuploid cells for evaluating the sensitivity of our method. The results show that the sensitivity of the described flowcytometrie method is much higher than those of the microscopical evaluation to look for residual leukemic cells. Important factors which influence the aneuploidy detection are the DNA-index of the aneuploid population and the coefficient of variation of the GO/Gl-peak of the diploid cells. It should be further investigated if this method can be used in monitoring patients with aneuploid leukemia cells. 1Medical Clinic I of Technical University Dresden;Dept.of Haematology Ontology; 2Clihic of Pediatrics; Fetscherstr. 74; 01307 Dresden

A124 117

119

Comparison of growth patterns of acute myeloid leukemia cells in vivo and in vitro by flow cytometry

Correlation of Deoxycytidine Deaminase, Thymidine Kinase and Polymerase cz Activity and Deoxycytidine Deaminase Gene Expression in Vivo with Clinical Response to TAD-9 + GM-CSF Induction Therapy in Acute Myeloid Leukemia.

K. Nowak, U. Oelschlagel, R. Hofmann, M. Palitzsch, H. Zengler and G. Ehninger

G. Jahns-$trenbel1.,M. Unterhaltl,E. Sehleyerl,C. Reute~,B. W6mmnnl,T.Biiehner3and W. I-liddemannj. l Dept. of Hematoologyand (9neology,,Universityof G0ttingen, RobertKoeh-Str.40, I)-37075 G0ttingen; "Dept.Mierohiologyand Cancer Center, Universityof Virginia; 3 Dept. of Hematologyand Oneology,Universityof Mimster.

Cells from patients with acute myeloid leukemia were analysed by flow cytometric detection of PI-staining for evaluation of proliferative activity. If the percentage of blasts was lower than 60%, additional immunophenotyping beneath cell cycle phase analysis was performed. The distribution in cell cycle was measured immediately after bone marrow aspiration, after 72h cultivation without and with Interleukin 3 (m3). The growth patterns result from alterations in S-phases of more than 50% between the different samples. In this way, of 9 possible growth patterns only 6 were detected. In most cases (18/32) blasts were stimulated with IL-3 as expected. But in two leukemias an increase was observed in control culture compared to in vivo S-phase. TFJs may be induced by autocrine loops. IL-3 sensitive cases showed a trend towards better clinical outcome compared to the group without increase of S-phase after IL-3 culture. The 2 cases with a decrease of S-phase in control culture compared with in vivo and no IL-3 induced increase have favourable clinical outcome. Both are relapse free for 1160 and 387 days respectively. Most cases of FAB M4/M5 (6 of 8) were not sensitive to IL-3, when evaluated with this method. Medical Clinic I of Technical University Dresden; Fetscherstr. 74; 01307 Dresden

Although signiticaut advances have been made in the treatment of acute myeloid leukemia with ma42 containing regimes, a considerable numberof patients does not reach an adaequate blast cell redaction after induction therapy. Therefore it is an impamut task to find sensitiveand reliable pmgnestieparametersfor clinical response in orderto screenpatientsbeforethe start of therupyand to optimize inductiontherapy. Bone marrowcells from 32 patients with newly diagosed AML were analysed for the activities of the followingenzymes,relevantfor the metabolismof ara-C: Deoxyeytidine kinase (DCK),Thymidinekinasr (TK),DeoxycytidineDeaminase(DCD),Polymerase (Poly ct) and overall Polymerase. Median values and range were for DCK 27.2 prnottmin/mgprotein (3.8 - 196), for TK 3.1 pmol/min/mgprotein (0.0 - 21.6), for DCD 1.64 nmol/min/mgprotein (0.0 - 8.6), for overall Polymersse55.9 pmol/min/ragprotein (17.3 - 13 l) and for Polytz 1.22 pmol/min/mg protein (0.03 - 8.8). The data oftbe enzymeactivities were related to clinical ~'ponse of TAD-9 +GM-CSF induction therapyin 29 patients, taking the blast cell reductionat day 10 or day 16 as a parameter of early response. Neither DCK- nor overall Polymerase-activity were predictive for response.Differencesin TK- and Poly et activity between the responder groupand the nonrespandergroup were statistically significant(p= 0.0064 for TK and lr=0.0062 for Polymerase o~). The DCD activity turned out to be the most sensitive parameter for clinical responseto TAD-9 +GM-CSFinduction therapy. The difference betweenthe respandergroup(median:0.333 nmol/min/mgprotein, range: 0.0 - 4.1) and the nonrespondergroup (median: 4.8 umol/min/mgprotein, range: 0.11 - 8.45) was statistically significant (it=0.0012). In order to gain an insight into the molecular mechanism underlying these varying enzyme activities transcription of the DCD gene yeas investigated in leukemic blasts of 15 AML patients. RT-PCR of ~-actin served as control for integrity of the RNA and as reference for quantification of DCD gene expression. Spceifitywas checked by direct DNA-sequeneing.It could be demonstrated that high DCD activity was associatedwith high transcriptionrate of the DCD gene and vice versa. It can be eonelndedfromthis studythat high activities of the S-phasedependent enzymes TK and Polyexand low activity of DCD are associatedwith goodclinical outcome. The resultsdemonstrate,that these factorsare valuable parametersin orderto stratifypatients in favorableand unfavourableprognosisgroupsfor inductiontherapy, whichis the basis for risk-adaptedtherapyprotocols.

118

120

Z.Kuliczkowski, S.Kotlarek-Haus, K.S~dek

CD7+CD4-CDS- BLAST CELLS IN ACUTE LEUKEMIAS DF Gluzman, IV Ahramenko, NI Belous, LM Sldyarenko, LY Poludnenko, SN Gaidukova, VD Drozdova, SB Donskaya, EA Lyvshits Institute of Experimental Pathology, Oncolog~, and Kadiobiology, Kiev, Ukraine

The influence of short term l i q u i d leukemic blasts on their phenotype. Department of Haematology, Wroclaw, Poland

culture of

Medical University of

The i m m u n o p h e n o t y p i n g of acute leukemia blasts is not always s a t i s f a c t o r y in precise d e t e r m i n a t i o n of the type of p r o l i f e r a t i n g cells. An attempt was made to test if a short 3 days' culture of leukemic cells in RPMI 1640 + 10% FCS m e d i u m will disclose their additional features. Blasts were classified acc. to FAB and p h e n o t y p e d w i t h the following monoclonal antibodies (DAKO): HLA-DR, CDI0, TdT, CD3, CDI9, CD33, CDI3 and CDI5 prior to and after the culture. Studies of 16 p a t i e n t s (13 with AML and 3 w i t h ALL) showed an increase or inversely a decrease of percentage of cell type w h i c h seemed to predominate before culture. Three e x a m i n e d cases had cells suspected as b i - p h e n o t y p i c on morphological and cytochemical grounds and it was confirmed after 3 days' culture (in two AML lymphoblasts appeared and in one ALL myeloblasts). Blasts of one case of AML Mo revealed lymphoblastic features after the culture, and in one case w i t h bi-phenotypic cells (myeloblasts and lymphoblasts) only lymphoblasts proliferated. In c o n c l u s i o n it seems that short time 3 days' culture of leukemic blasts may contribute to more precise diagnosis of acute leukemia.

Immunocytochemical methods (PAP and APAAP) and monoelonai antibodies (MoAbs) were applied against HLA-DR CD4, CD5, CD7, CD8, CD13, CDI4, CD15, CD20, CD33, CD34, CD45RA, CD45RO, to GMCSF and SCGF receptors, proteins of myeloid cells p8 and p14. In 14 of 89 examined children and 112 adult patients with different forms of acute leukemias (AL) blast cells with the phenotype CDT+CD4-CD8- were revealed. Two cases were diagnosed as pre-T ALL, 6 cases as AL of myeloid origin (AML M0, M1, M4 and MSa by FAB classification). In bipbenotypir AL lymphoid (CD10), myeloid (CD33) and erythroid (HAE9) antigens or markers of lymphoid cells (CD10, CD37) and monocytes (CD14) were eoexpressed on CD7+CD4-CD8- blast cells. In 4 patients CD7+CD4-CD8- blast cells had minimal signs of differentiation. The last two subvariants can be regarded as AL from transformed hemopoietic stem cells (Kurtzberg et al., 1989). For singling out CD7+CD4-CD8- AL as independent subvariant of disease it is necessary to use immunophenotyping with obligatory use of broad spectrum MoAbs, TdT, rearrangement of the immunoglobulin heavy chain and TCR genes.

A125 121

123

ADHESION MOLECULE RECEPTOR PROFILE IN ACUTE LEUKEMIA

LYMPHOID ANTIGEN EXPRESSION IN ACUTE MYELOID LEUKEMIA G. Specchia, G. Palumbo, A. Mestice, D. Mininni, P. Carluccio, A. Quarta, G. Debellis,V. Liso Hematology - Universityof Bari - Italy Leukemic blasts of some patients with AML display surface lymphoid-associated antigens (AML Ly+). The incidence of these lymphoid antigens in AML varies considerably among different studies (from 13 to 60%). Frequently, the presence of asynchronous and mixed-lineageantigen expression is associated with poor outcome in At_ Immunophenotypes for 191 patientswith AMLwere analyzed using awide panel of Monoclonal Antibodies (MoAt)s) and surface marker analysis was performed by immunofluorescence (FACScan) (positive cut-off level 20%) and immunocytochemical methods (APAAP). All cases expres.,~l one or two myeloid-essociated antigens (CD13, CD33, CD14, CD15). 80 out r 191 cases (41.9%) expressed at least one 'lymphoid' associated antigen (CD7 22.9%, CIM 22.3%, CD2 13.9%, CD3 5.6%, CD19 3.5% and CD10 3.1%). The incidence of Ag Ly expression was 83.3% in MO cases (10/12), 63.6% in M1 (7/11), 28.9% in M2 (22176), 34.5% in M3 (10/29), 51.9% in M4 (14/27), 48.0% in M5 (12/25), 50.0% in M6 (5/10) and 0% in M7. No significant differences were found between Ly+ and Ly- AML patients as regards presenting clinical features and complete remission rate. CD7, CD4 and CD2 antigen expression was found 1~obe correlated with FAB MO/M1, M4/M5 and M3 morphology respectively. We confirm the relevant frequency of 9lymphoid" markers, particularly T (45.2%), less often B (5.2%), on the surface of AML blasts. Further studies are needed to assess the clinical significance of lymphoid-associetedantigen expression in adutt AML.

M.A. Reuss-Borst: G. Klein, H.D. Waller and C.A. Miiller, Medical University Clinic II, Section of lmmunohematology and Transplantation lmmlmology, D-72076 Tiibingen Since variant expression patterns of adhesion molecules have been suggested to alter the adhesive qualities of leukemic blasts, a large number of leukemic samples were investigated for the expression of 131-, f12-, l~3integrins, CD44, the three selectins and several members of the immunoglobulin family. In summary, adhesion molecules such as CD44 (167/169), LFA-3 (158/169), the I$1-integrins VLA-4 (120/123) and VLA-5 (45/51) and the g2-integrin LFA-1 (149/157) were found on >70% of blasts in most lenkemias. Other adhesion molecules seemed to be restricted to specific differentiation stages and lineage. The g2-integrins Mac-1 (CD1 lb) and gp 150,95 (CD1 lc) were preferentially expressed on AMLs of the differentiated subtypes M4 and M5, and NCAM (CD56) was only found on a subset of acute myeloid leukemias. Unexpectedly, some adhesion molecules such as the fil-integrins VLA-I (1/51), VLA-2 (18/123), VLA-3 (5/43), VLA-6 (15/29) and the E-selectin (2/47) were expressed on >70% blasts on a subset of leukemias not characterized by a common phenotype; these molecules were not found on CD34+ hematopoietic precursors of the bone marrow. In addition, the analysis of simultaneously obtained blood and bone marrow samples generally revealed a higher percentage of positive blasts in the blood than bone marrow, possibly due to the up-regulation of these antigens. The functional significance of antigen expression in these cases remains to be elucidated; however, our observations suggest that in leukemia these antigens are displayed in a non-adherent phenotype that cannot be converted to an adherent, functionally active conformational state due to defective, yet unknown regulation mechanisms.

122

124

CD54 EXPRESSION AND ITS ROLE IN HOMOTYPIC AGGREGATION OF THE BLASTS OF ACUTE MYELOBLASTIC (AML) AND ACUTE LYMPHOBLASTC LEUKEMIAS (ALL)

Expression of Adhesion Molecules on Blood Derived CD34+ Cells immediately after Thawing and after Ex Vivo Incubation with Cytokines Kcenigsmann M, Notter M, Miicke C, Thiel E, Berdel WE, Division of Hematology and Oncology, Benjamin Franklin Hospital of the Free University of Berlin, Germany

J. Krautsr, K. Pompe, D. Bunjes, H. Heimpel and G. Hell Departmont of Internal Medicine III, University of Ulrn, 89081 UIm

CD54 (ICAM-1) expression was studied on blasts of 13 cases of de-novo AML and 10 cases of ALL prior to and after stimulation by Interferon-gamma (IFN-r). Furthermore, its functional activity was studied by analysis of the influence of CD54 on homologous interaction of the blasts. Prior to culture only a minority of the AML and ALL blasts displayed CD54 positivity. Timed incubation of the AML blasts in suspensions cultures in RPMI-1640/FCS (final concentration 10%) or in serum-free medium for up to 24h at 37"C in 5% CO2 in a fully humidified atmosphere increased the percentage of CD54 positive cells in 11/13 AML-cases. Incubation of the AML blasts in the presence of IFN-T (500 IU/ml) further enhanced CD54 positivity in 6/11 cases, whereas in 2/13 cases no CD54 upregulation could be detected at all. Incubation of the ALL blasts in RPMI/FCS or serum-free medium alone led te an increase of CD54 expression in 2/10 cases, while the addition of IFN-T to the media led to an increase in CD54 expression in 6/10 ALL cases. In AML ICAM-1 expression was paralleled by homologous aggregation of the blasts in that (1) in all but the two CD54 negative cases autologous cluster formation could be detected. (2) IFN-T enhanced cluster formation in 5/6 cases in which it had enhanced CD54 upregulation. (3) Incubation of the blasts in the presence of an anti-CD54 MoAb (clone 84H10) reduced the "spontaneous" and IFN-T induced cluster formation in a majority of the cases. In ALL only 2/6 cases, in which CD54 upregulation could be induced, displayed homologous cell aggregation, which could be slightly enhanced by IFN-Tstimulation and inhibited by the anti-CD54 MoAb. Taken together, CD54 expression on AML and ALL blasts is heterogenous with respect to (1) its "constitutional" expression and (2) its "spontaneous" and IFN-T induced upregulation. While on AML blasts CD54 seems to be functionally active once expressed on the surface membrane, in at least some ALL cases its function seems to depend on additional activation.

In order to understand the role of specific adhesion for stem cell engraftment after autolognus transplantation we investigated the expression of adhesion molecules on blood derived CD34+ cells from cancer patients immediately after thawing and after ex vivo incubation with or without cytokines. Peripheral blood progenitor cells (PBPC) were collected by lenkapharesis (COBE Spectra) after conventional chemotherapy and G-CSF given during the recovery phase. The buffT coat cells were stored in liquid nitrogen. Aliquots were used for our study with the patients informed consent. After thawing, CD34+ cells were enriched by immunomagnetic mierobead separation (Miltenyi Biotech, Cologne, Germany). CD34 and adhesion molecules were detected simultaneously by dual-cxflorflow-cytometry. Immediately after CD34 cell enrichment we saw bright signal intensities for the integrin subunits CD29, CD18, CD49d, CD41, for CD44, and for the selectin CD62L. Dim intensity was found for CD49e and CD61. We did not detect CD49a, 49b the beta 4 integrin subunit, CD51, CD62P, and the CD44 splice variants V6, VT, V7-8. Some molecules were studied after 14 hours of incubation in human AB serum. Without cytokine addition, an increase in signal intensity was seen for CD49e, CD29, and CD62L. Also, the fraction of the CD34+ ceils expression these three molecules increased. The addition of a cytokine cocktail containing recombinant human GCSF (Rh6ne-Poulenc Rorer Inc., Paris, France), SCF, 1I,-3 and IL-11, each at 100 ng/ml, did not clearly change the signal intensities for these adhesion molecules. Before cryopreservation, CD29, CD18, and CD26L were brightly expressed on most CD34 cells. We conclude that 1. Immediately after thawing, the expression of some adhesion molecules like CD29, CD49e, and CD62L on CD34+ cells is decreased while it is maintained for others like CD18 and also the CD34 molecule itself; 2. Ex vivo incubation for 14 hours leads to re-expression or increased expression of CD29, CD49e and CD62L. 3. High concentrations of G-CSF, SCF, IL-3, and IL-11 do not contribute to inducing rapid re-explessien of these adhesion molecules. Supported by Dr. Mildred Schcel Foundation, grant No W27/92/Kol and by Rh6neoPoulencRorer Inc. Paris, France.

A126 125

127

C~(TOMORPHOLOGY AND CLINICAL OUT-COME IN 5 4 P A T I E N T S WITH AML M4EO AND INVERSION 16

Studies on p l a s m a t i c c o a g u l a t i o n in Children with ALL u n d e r g o i n g t h e r a p y using t h e C O A L L - 0 5 - 9 2 p r o t o c o l l

T. Hafedach, H. LOftier, J.M. Bennett, J. W. Andersen, W. Gassmann, W.-D. Ludwig, E. Thiel, E. Paietta, P. A. Cassileth, C. Fonatsch, C. Schoch, B. Schtegelberger, R. Becher, M. C. Sauedand, A. Heinecke and Th. B~chner for the AML Cooperative Group and the ECOG; Kiel, Germany and Rochester, N.Y., USA

S. Eckhof-Donovan, R, Kirschke, I. Michelmann, D. S c h w a m b o r n , U. G 0 b e l

The acute myelomonocytic leukemia with abnormal eosinophils (FAB AML M4Eo) is considered a distinct subtype of AML with characteristic clinical, morphological, cytogenetic and prognostic factors. We analyzed specific morphological features and treatment response in 54 patients with the cytogenetic detection of inversion 16 and the morphological appearance of M4Eo who underwent aggressive chemotherapy in two cooperative study groups (AMLCG: n=34; ECOG: n=20). Treatment schedules are comparable and included thioguanine, ARA-C, mitoxantrone, daunorubicin or idarubicin. Median age was 39 years (range 20-66), 29 female and 25 male were studied. The percentage of blasts (type I,I1, and III combined including monoblasts) ranged between 45 and 90% (mean 66%); the percentage of monoblasts and monocytes ranged between 12 and 41% (mean 24%). Erythropoiesis was decreased to 3-25% (mean 8%). The number of specific abnormal eosinophils with basophilic granules that led to the morphological diagnosis of AML M4Eo ranged between 1 and 31% (mean 10%). In 28 cases up to 10% of abnormal eosinophils were observed in the bone marrow, in another 18 cases eosinophils differed between 11 and 20%, and in 8 cases more than 20% of abnormal eosinophils were seen. Dysgranulopeiesis (DysG) was observed in 8 pts. (15%); one pt. had DysG in combination with dysmegakaryopoiesis (DysM). 14 pts. had only DysM. Only one pt. had DysM combined with dyserythropoiesis. One pt. had tdlineage dysplasia, tn the majority of cases (28/54 pts. = 52%) no dysplasia was detected. Response to chemotherapy was very high in pts. with M4Eo and inv16:50/54 of our pts. (93%) achieved a complete remission, one pt. was considerd as eady death and three pts. as no remission. Prognostic factors and survival data will be presented in detail.

Backaroun0; A survey c o n d u c t e d in 1993 in p a t i e n t s u n d e r g o i n g t r e a t m e n t using t h e C O A L L - p r o t o c o l l r e v e a l e d "that not only the use of venous c a t h e t e r s , b u t also the induction p h a s e as well as the administration Of asparginase r e p r e s e n t e d risk f a c t b r s for thrombosis. In o r d e r to b e t t e r u n d e r s t a n d plasmatic c o a g u l a t i o n a c t i v i t y during these phases of t r e a t m e n t , d e t a i l e d c o a g u l a t i o n analysis was p e r f o r m e d . Patients: ] ] girls a n d 10 b o y s suffering from ALL, a g e at t h e time of diagnosis ranges from 1,4 t o 13,1 years, l e u c o c y t e count a t diagnosis b e t w e e n L 100 to 1,258,000/~d; ] ] p a t i e n t s w e r e t r e a t e d a c c o r d i n g t h e high-risk (HR-) a n d 10 p a t i e n t s a c c o r d i n g to the low-risk (LR-) protocotl COALL 05-92. M e t h o d s ; Cectgulc#ion ctcfivc/tion pc/rameters: thrombinantithrombin-III complex (1AT), modified antithrombin I11 (ATM), p r o t h r o m b i n f r a g m e n t (F ]+2); anticoogulcttory porame/ers: antithrombin III (ATIII), protein C (PC); parameters o f fibrinolyfic capacity: plasminogen a c t i v a t o r inhibitor (PAl). plasminogen (Piing), ontiplasmin (AP). Results: Evaluation of the a v a i l a b l e d a t a shows c o a g u l a t o r y a c t i v i t y is increased in b o t h the HR a n d LR groups ( m e d i a n values for TAT t, ATM t, F1+2 t). While t h e median values for All11 remained within t h e r e f e r e n c e boundries, PC values w e r e low. particularly following ASP administration. "the median values for PAl w e r within t h e normal r a n g e in b o t h groups in at~ phases o f t r e a t m e n t , a n d all o t h e r m e a s u r e d fibrinolytic p a r a m e t e r s w e r e also unremarkable, Conclus.ien: Not only a t t h e time of diagnosis, b u t also during t h e r a p y , children with ALL h a v e b e e n shown t o h o v e increased coagulation activity which does not cause serious d e r a n g e m e n t of fibrinolytic c a p a c i t y . Protein C w o u l d a p p e a r t o b e a more sensitive p a r a m e t e r t h a n AT III.

This work is supported b E the Elterninifiative Kinderkrebsklinik e. V.

126

128

PROGNOSTIC RELEVANCE OF P R O T E O L Y T I C A L L Y INACTIVATED ALPHA1-ANTITRYPSIN EXCRETED IN THE URINE OF PATIENTS WITH ACUTE MYELOID LEUKEMIA. R. Dengler, R. Busch, G. Eger* and B. Emmerich. Med. Klinik Innenstadt, H~.matologie und Onkologie, D-80336 MOnchen; and *Med. Klinik Ill, Universit~.t Erlangen-N0rnberg, 91054 Erlangen, Germany. During remission induction chemotherapy, a fragment of alphal-antitrypsin (ell-AT) can be observed in the urine of patients with acute myeloid leukemia (AML). By using immunoblotting, AML as well as other patients with haemato-oncotogical diseases were screened for the excretion of this inhibitor fragment. In 74 % of AML the patients (n = 27), the protein was detected. Mean concentration of peak excretion was found to be 6.7 ug/mg creatinine (cr). It could also be detected in one out of 10 ALL and 3 out of 15 NHL patients, but not in patients with solid tumors, infections, kidney diseases or healthy donors. Among the AML patients, those who responded completely to induction chemotherapy (< 5% residual marrow blasts) exhibited significantly higher peak concentrations of this protein than the nonresponders (9.45 vs. 1.50 ug/mg cr, p < 0.05, Log rank test). The probability of median time to reach remission was significantly shorter in patients excreting the inhibitor fragment than in the nonexcreters (40 vs. 100 days, p < 0.02, Log rank test). In addition, Kaplan-Meier analysis revealed that overall survival differed significantly when using a best cut-off strategy which showed that patients with > 3 uglmg cr had a median survival of 28 months compared with 10 months in the group with < 3 ug/mg cr (p < 0.05). The data presented here are the result of a pilot study with a limited number of patients. Clearly, future studies with larger patient numbers are required to test whether or not the excretion of proteolyzed ~ I - A T could indeed be a marker for response and prognosis in AML patients.

Dependence of asparaginase - induced coagulopathy on serum asparaginase activities : a randomized trial. Nowak-G6ttl U., Boos J., Werber G., Ahlke K, J~gens H. Paediatric Haematology/Oncology, University Hospital Miinster, Germany Hypercoagulability and thrombosis are well documented side effects of leukemia treatment with l-asparaginase (1 - asp). To determine haemostatic alterations of different K coli asp preparations a randomized prospective study was carried out in 20 leukemic children (ALL- BFM -90): 10 patients received 1 -asp Crasniting/Bayer, Germany (group A) and 10 patients medacR (group B). Blood samples for coagulation studies were obtained prior and whilst on medication. L- asp activities were higher in group B. Ash depletion did not differ in both groups. Alterations of haemostatic parameters observed (Table 1: Median and ranges) were more pronounced in group B: Values of fibdnogen, antithrombin HI, plasminogen and ct2-anfiplasmin were significantly (* p< 0.001, ** p< 0.005) decreased, D - Dimer formations were found to be clearly higher than in group A. Changes of VWf:ag and plasminogen activator inhibitor 1, the latter only enhanced at one time point in group B, did not differ in both groups. In group B a 13 year old boy developed thrombosis and a 7 year old girl acquired intermediate insulin dependent hyperglycemia. Day Act

IU/1 Rb m~dl AT III %

12

21 - 24

A

B

<20

< 20

110 57-213 113 75-135 Plasm 96 % 70-130 ~2AP 100 % 70-108 D-Dimer 28 ug/t t 8-400

110 48-110 109 84-169 92 48-126 106 80-107 42 16-115

A

79 * 8-192 89 ** 40-164 70 ** 66-99 67 * 41-89 79* 69-109 19 ** 8-42

27 - 30 B

A

NV B

278 61* 291 84-432 20-156 85-864 41 77 * 37 25-61 47-133 26-54 56 78 * 57 34-73 4 8 - 9 2 41-67 35 65 ** 43 32-63 4 5 - 9 8 35-58 43 82* 48 37-63 61-91 39-81 18 ** 130 110 33-350 6-300 10-1450

< 20 200-400 80-138 75-140 80-120 4-25

A127 129

131

SIGNIFICANT DIFFERENCES IN THE PHARMACOKINETICS OF TWO L-ASPARAGINASEPREPARATIONS FROM ESCHERICHIA COLI

ASPARAGINASE ACTIVTIES IN VITRO ARE HIGHLY SENSITIVE TO

J.Boosl, G.WerberL2,E.Ahlke1, U.Nowak-Gfttl t, E.,Verspohl2, H.Jiirg~ast

Gisela Werber1'2, E.Ahlke1' U.Nowak.G6ttl1,H.J~rgensl,, E.J.Verspohl2, j.Boos 1 (1) Department of Pediatric Hematology/Oncology, (2) Institute of Pharmaceutical Chemistry, Univ. of MLinster, 48149 Mtinster, FRG Different asparaginase (Ase) preparations from various biological sources (Eschedchia coil and Erwinia carotovora) are currently in use as potent antileukemic enzymes in the treatment of acute lymphoblastic leukemia in children. According to ALL/NHL-BMF-90 protocol therapy is initiated with E.coli-Ase (Asparaginase| MEDAC, MEDAC GmbH, Crasnitin | BAYER AG) and in case of allergic reaction substituted by Erwinia-Ase (Erwinase | PORTON). There are, however, important questions on comparability regarding these different preparations, Therefore, we compared the enzyme activities in vitro of E.coli-Ase (obtained from MEDAC and BAYER) and Erwinia-Ase (obtained from PORTON) at different pH-conditions and various (eight) buffer conditions (pH 7.0 - 8.5, each expedment 5-times). Physiological pH of PBS-buffer resulted in only 50 to 65 % activity regarding the manufactores declarations in Erwinia-Ase and in 80 to 95 % in E.coli-Ase. The expected activities were detectable at pH 8 and pH 8.5. In saline solutions (0.9 %) the activity decreased to 20 % and 60 %, respectively. Buffers containing bodc acid (pH 8.5) or Tds (pH 7.3) and additional BSA resulted in a more reliable determination of enzyme activity according to the declarations on the vials (100 + 10 %). In most experiments E.coli Ase showed higher results than Erwinia-Aae. Conclusion: The activity of asparaginase preparations depends on the conditions of the dissolution and Erwinia-Ase is more sensitive compared to E.coli-Ase.

(1) Department of Pediatric Hematology/Ontology, (2) Institute of Pharmaceutical Chemistry, Univ. of Miinster, 48129 Mtinster, FRG Asparaginase (ASE) preparations from different biological sources (Eseherickla eoli and Erwinia caratovora) and companies are in clinical use. Current treatment protocols prescribe the dose and schedule but not the preparation to be used. As a rule, Escherichia cell preparations are given (Asparaginase medac| Medao-GmbH or Crasnitin| Bayer AG) until allergic reactions occur. We initiated a drug monitoring measuring asparaginase activity as well as serum amino acid levels in children on the ALIJNHL-BFM 90 treatment protocols. Samples were taken immediately prior to the application and additionally on the occasion of routin+e diagnostic punctures. In protocol I, the dose administered was 10,000 U / m 2 every three days (day 12-33). Asparaginase activity was significantly higher in children treated with Asparaginase medac (162 samples/38 patients: median 441 U/l) than in children on Crasnitin (49 samples/10 patients: median 79 U/L). The serum asparagine levels were in the order of the detection limit (0.1 p_M) in both groups. While glutamine was similar (365/368 lttM) glutamic acid was elevated in the Medac group (77 vs 146 lttM median). The pharmacokinetic calculation of pooled data resulted in a T 89 of about 24 hours for Asparaginase medac and about 15 h for Crasnitin. Discussion: There are significant differences between asparaginase preparations even when likewise obtained, from Escherichia coil The relatively high activity observed with Asparaginase Medac is not necessary to reach complete asparagine depletion. In the discussion about side effects o f Asparaginase, the preparation should, therefore, be specified in every study and the dosage.,; should be defined taking pharmacokiaetic differences into account.

DIFFERENT BUFFERCONDITIONS

Supported by BMFT

130

132

I:~ARMACOKINETIC ] PHARMACODYNAMIC MONITORING OF L-ASPARAGINASE IN CHILDREN

Influence of L-asparaginase E.coli and Erwinia on antitromhin III and coagulation factors during induction therapy of ALL in children.

G.Werber1"2, J.Boos t , E.Ahlkel, U.Nowak-Gtl/l l, E.Verspohl2, H.JiirgensI (1) Department of Pediatric Hematology/Oncology, (2) Institute of Pharmaceutical Chemistry, Univ. of MOnster, 48129 M/inster, FRG Asparaginase (ASE) preparations from different biological sources and companies are in clinical use. In the event of adverse reactions substitutions are common. The half-life of Erwinia asparaginase is s~gniflcantly shorter compared to asparaginase obtained from Escherichia coil Questions remain regarding the comparability of pharmaeodynamic effects. Therefore, we performed an asparaginase monitoring focussing on the comparability of asparaginase activity as well as asparagine serum levels. Therapy according to the protocol ALL/NHI~BFM 90 was initiated with E. coti-ASE (Asparaginase Medac~ / Crasaitin| Bayer) and in case of allergic reactions continued with Erwinia asparaginase (Erwinase~ Speywood ). All children underwent monitoring for asparaginase activity and serum asparagine levels; samples were taken immediately prior to the applications. The dose administered was 10,000 U/m2 every 3 or 4 days according to the protocol II. 50 protocol II patients, i.e. 26 children on Asparaginase Medac, 8 on Crasnitin and 16 on Erwinase were compared: The asparaginase activity on day 3 (drop level) was higher in children receiving E.coli asparaginase (Asparaginase Medac 528 E/t; Crasuitin 49 E/1 median) than in those on Erwinase (< limit of detection). The corresponding asparagine levels remained above the limit of detection (100 riM) in only 6 of 26 children on asparaginase Medac (all < 0.5 raM). On Crasnitin, 2 of 8 children were completely depleted, 4 were nearly depleted ( >0.1, <0.5 mM,) and 2 had normal asparagine levels. In the group switched to Erwinase only 2 of 16 children showed complete asparagine depletion was complete and in 2 others asparagine near the detection limit. In 12 children the levels covered the whole range up to normal concentrations. Supported by BMYT

A.Krauzel, M.Jelefiska2,M.Palester-Chlebowczyk2,M.Ochocka 1 1 Warsaw University Medical School, Department of Pediatic Hematology -Oncology ,Dziatdowska 1 ,Warsaw Poland 2. Warsaw University Medical School ,Department of Vascular Surgery and Transplantology , Banacha 1 ,Warsaw Poland Activity of AT III and factors I ,II ,V, VII ,X were investigated in a prospective study in two groups of children.The first group of 30 children received L-asparaginase E.coli ~the second group of 14 children received Erwinia..Both groups were treated according to protocol ALL BFM 86 .Coagulation factors were measured before the therapy ,during the first, the second, the third week of therapy and after finishing treatment with L-asparaginase.Results were analized statistically .Our studies showed a significant decrease in the activity of antitrombin III and fibrinogen during L-asparaginase therapy in the first group of patients .At variance in patients receiving Erwinia there was no such significant decrease in the activity of AT III and fibrynogen.

A 128

133

135

Infusion rate does not influence ifosfamide side-chain metabolism H.SiliesI~, J.Boos 1, G. Blaschke2, H.Jiirg~ns1

Topoisomerase 11 alpha and beta activity and cellular sensitivity to topo I1 inhibitors. Diffferences between lymphopoietic and granulopoietic leukemia cells

(1) Department of Pediatric Hematology/Oneology, (2) Institute of Pharmaceutical Chemistry, Univ. of MOnster, 48129 Mtinster, FRG Ifosfamide (IFO) is a widely used anticancer drug which requires metabolic activation by hydroxylation of the ring system to exert cytotoxic activity. A second metabolic pathway produces the eytostatically inactive metabolites 2-dechloroethylifosfamide (2-D-IFO) and 3-dechloro ethylifo sfamide (3-D-IFO) releasing chloroacetaldehyde. This pathway of side-chain metabolism has been discussed as a source o f CNS-toxir and in connection with renal toxicity. As regards neurotoxicity the incidence seems to be strongly schedule dependent. We compared the urinary excretion of IFO, 2-D-IFO end 3-D-IFO on short-term and continuous ffosfamide infusion (3g/m2). In addition, we compared schedules with dosages ranged between 400 nag to 3 g/m2. Urine was sampled up to 72 hours and investigated by gaschromatography. 464-13% of the applied dose could be recovered in the urine (n=26 patients). Main metabolite was IFO (22=1=8%) followed by 3-D-IFO (14• 2-D-IFO (9• and total acrolein (0.8a:1%). Neither the total amount formed nor the excretion kinetics of sidechain metabolites showed relevant schedule dependency. Even with 1hour infusion there was a lag of 3-6 hours until dechloroethylation became relevant. The excretion pattern of side-chain metabolites as well as unmetabolized IFO was nearly superimposable. Therefore, differences in toxicity and efficacy cannot be explained by an influence of the application time on the metabolic profile o f ifosfamide.

F Gieseter, W. Pfeifer, T. Brieden, M.Clark. Medizinische 9 Poliklinik, Wi~rzbur~ . We have performed ex vivo experiments with cells from

patients wit AML, CML, ALL and CLL and determined the activity oftopo lI alpha and topo II beta and the sensitivity of the cells to a number of tope II inhibiting drugs. We found no correlation between the sensitivity of the cells and the activity of either topo II alpha or beta, bat a good correlation between the ratio of topo II alpha to beta and the sensitivity. Basically, if the ratio (c~ : 13) is higher than 2, the cells are sensitive, if the ratio is lower than 2, the cells are resistant. This is true as well for lymphopoietic, as for granulopoietie leukemia cells. Comparing cellular sensitivity to two anthraeyclines, daunorubicin and idarnbicin, we found significantly differences in the lymphopoietic and the granulopoietic line. Whereas resistant granulopoietic cells (AML and CML) still are relatively sensitve to idarubicin, resistant lymphopoietie cells (ALL and CLL) exhibit a almost linear correlation between the LD50 ida and the LD50 dauno. We are currently investigating molecular resistance mechanisms, comparing both hematopoietic lines. Evidently, in lymphopoietie cells, the cellular reaction on the DNA damage due to topo I1 inhibition is disturbed, which results in a wide resistance - spectrum to DNA damaging agents. Consequences for therapy are discussed, although it seems rational to swich from dauno to ida in the treatment of resistant granulopoietic cells, this is not the case in resistant lymphopoietic cells.

134

136

BUSULFAN PHARMACOLOGY IN BONE MARROW TRANSPLANT PATIENTS. Voges K, Kolb H-J, Mittermiiller J, Jehn U, Wilmanns W Heinemann V. Dept. Hematology, Klinikum Grosshadern, University of Munich, FRG.

Q u a n t i f i c a t i o n o f I n t r a c e l l u l a r A r a - C Metabofites J, Pfdrtner, J. Braess, C.C. Kaufmann, M. Unterhalt, B. Ramsauer, W. Hiddemann and E. Schleyer Department of Haematology/Oncology, University of Gtttingen, 37075 G/Sttingen, FRG

Thirtyone patients, 13 AML, 16 CML, 2 RAEBT (median age 47 yrs, range 26-58 yrs) received high-dose busulfan as conditioning regimen for bone marrow transplantation (22 allogen and 2 autologous). Busulfan was given at a dose of 1 mg/kg (2mg tablets) every 6 hours until a total dose of 16 mg/kg was achieved. Peak plasma concentrations of busulfan were measured 60-90 rain after oral intake. A limited sampling of plasma at 0, 60, 180 and 360 min allowed a reliable determination of busulfan AUCs. Busulfan plasma concentrations showed a considerable variability between different patients (20-fold), but also within the respective treatment courses. To allow a targeting of busulfan plasma concentrations by appropriate dose adjustment, it appeared necessary to reduce the intrapatient variability of busulfan plasma concentrations. We therefore evaluated the impact of food uptake on gastrointestinal absorption and pharmacokinefics of busulfan. Thirteen orally fed (OF) pts were compared to 18 parenterally fed (PF) pts. In fact, interpatient and intrapatient variability could be significantly (p=0.016) decreased by parenteral feeding. However, trough values of busulfan were significantly lower in PF pts compared to OF pts (350 vs 500 ng/ml p<0.001). We further performed a longitudinal analysis of busulfan trough concentrations during the whole treatment period. The pharmacological evaluation of 30 pts demonstrated that busulfan plasma concentrations significantly (13=0.002) decreased from dose 4 to dose 15 (413 vs 281 ng/ml). All pts received an anticonvulsaat prophylaxis with phenytoin during busulfan treatment. It is, therefore, conceivable that phenytoin, a known enzyme inducer, enhanced hepatic metabolism of busulfan and thus increased busulfan clearance from plasma during the course of busulfan conditioning treatment.

When investigating the antineoplastic mechanisms of Ara-C in the past the focus has been primarily on its triphosphate (Ara-CTP). This approach is not sufficient to explain the efficacy of high dose Ara-C regimens featuring saturation of deoxycytidine kinase and totally ignores the postulated influence of Ara-CDP-Cholin on Ara-C cytotoxicity. Therefore an HPLC assay has been developed that is able to separate in a single run all major intracellular metabolites of 5-3H-Cytosine-beta-Darabinoside in an in vitro model. Using this assay the following substances can now be separated and quantified: Ara-C, Ara-CMP, Ara-CDP, Ara-CTP, Ara-U, AraUMP and Ara-CDP-Cholin. The metabolites are separated by ion-pairing involving a 250/8/4 5/z Cis reversed phase column, a C18 precolumn and two mobile phases with the first eluent consisting of 0.0025 mol PIC A, 0,2 % CH3CN at a pH of 3,0 and the latter involving 0. l tool NaI-I2PO4, 0.005 tool PIC A, 0.5 % CH3CN at a pH of 2,6. Cells were incubated with 3H-Ara-C then washed with normal saline and finally lysed using the weaker eluent. Following injection of 100 /zl of the cell extract into a parking loop the first (and weaker) eluent will transport the sample onto the solid phase thereby separating all the aforementionned substances except Ara-CTP. The latter will subsequently be eluated by the stronger second mobile phase thereby reducing the elution time for a single run to 45 minutes. Several eell-lines have been analysed so far including HL60, K562, Raji, Nalm, Blin and U937. Ratios of (Ara-CMP + Ara-CDP)/Ara-CTP and Ara-CDPCholin/Ara-CTP were calculated in order to demonstrate the significance of these metabolites in relation to Ara-CTP. The ratio of the mono- and diphosphate to AraCTP ranges from 0.08 (Nalm-16) to 0.43 (Blin) which means that at least in some cases a substantial amount of not yet completely phosphorylated Ara-C metabolites can be found. In the ease of Ara-CDP-Cholin/Ara-CTP results ranging from 0.06 (Nalm-16) to 0.65 (U937) were calculated indicating that due to its abundance in several eell-lines this metabolite is attractive for further investigation. Only minimal amounts of phosphorylated metabolites of Ara-U could be found during these analyses in all eell-lines. In the future individual profiles of Ara-C metabolites in leukemic blasts and in normal bone marrow during in vitro dose escalation will be established.

A 129

137

139

THE ROLE OF dCTP FOR FEED-BACK REGULATION OF DEOXYCYTIDINE KINASE AND PHOSPHORYLATION OF CYTOSINEARABII'~OSIDEIN WHOLE LEUKEMIA CELLS. Schulz L 2, lssels R. D.1, 2, Debus A. 1, Wilma0ns W. 1,2, Heinemarm v.l.lUniversity Hospital Grosshadern, University of Munich, FRG. 2GSFInstimt for Clinical Hematology, Munich, FRG.

ADULT ACUTE M Y E L O B L A S T I C L E U K E M I A (AML): IN VITRO DRUG R E S I S T A N C E PREDICTS PROGNOSIS. AJP Veerman, E Klumper, GJ Ossenkoppele, MMA Rottier, G Westra, R Pieters. Dept. Pediatrics and Dept. Hematology, Free University Hospital, Amsterdam, the Netherlands INTRODUCTION. In vitro drug resistance assays show a strong correlation with prognosis in various mallgnant disorders. In two studies, one retrospective and one prospective, we could show that in vitro drug resistance is the strongest prognostic factor in childhood ALL. We now determined in vitro drug resistance on 52 samples o f ' n e w l y diagnosed adult patients with AML. The short term (4 days) MTT assay was adapted for use on fresh and cryopreserved samples of AML. All drugs used in the current Dutch protocol were tested: cytosine-arabinoside (ARA-C), daunorubicin (DNR), mitoxantrone (MITOX), m-amsacrine (AMSA) and etoposide (VPI6). RESULTS. Non-responders after one course of chemotherapy (ARA-C and DNR) were 2.4-fold more resistant to DNR (p=0.02) and 1.7-fold more resistant to ARA-C (p=0.07) than complete responders. The probability of CCR for patients in vitro sensitive to ARA-C was 61% at 34 months, whereas all patients in vitro resistant to ARA-C relapsed within 18 months (p = 0.02). This difference was independent of age, sex, white blood cell count, FAB classification, karyotype and CD34 expression. However, patients over 60 years were relatively resistant to ARA-C and AMSA (2.5 fold) and CD34 positive samples were 2 fold more resistant to DNR and MITOX than CD34 negative samples. CONCLUSION. The CR rate was related to resistance to both ARA-C and DNR, the risk of relapse was strongly related to ARA-C resistance. The other drugs tested did not correlate with prognosis in this limited study. The risk groups older than 60 years and CD34 positive AML showed relative drug resistance.

Denxycytidine (dCyd) kinase (dCK) is the rate limiting enzyme in the activation of a variety of,antimetabolites such ,as 1-1~-D-arabino-furanosylcytosine (ara-C), 2'-ehlorodeoxyadenosine or 2',T-diflnorodeoxyeytidine. In vitro experiments with purified dCK led to the conclusion that dCK activity is tightly regulated by. the concentration of dCTP. This has, however, not been conclusively demonstrated in whole cells. In fact, HU, a potent inhibitor of ribonueleotide reductase (RR) depleted dCTP in CEM and Raji ceils, but enhanced ara-CTP formation only in the latter call line. In this study we analysed dCK activity as a function of dCTP concentration in whole Raji cells and used ara-CTP formation as a measure of enzyme activity. Depletion of dCTP was achieved by three different dt~ags: HU, 3-deazauridine, and thymidine. 3-Deazauridine acts as an inhibitor of CTP synthetase, an enzyme converting CTP to UTP, while thymidine blocks CDP reduction. In Raji cells, HU (5mM) depleted dCTP to 44% of control and increased dCK activity by 3.7fold. On the other hand, thymidiee (100/~M) and 3-deazauridine (5/,M) caused a reduction of the intracellular dCTP pool to 50% and 20% respectivly. However, dCTP depletion by the latter agents was not associated with an increase of dCK activity compared to control c.ells. We therefore concluded, that in Raji cells HU-mediated enhancementof dCK activity is not a result of lowered dCTP-feedback on riCK. Enhanced ara-CTP formation may, however, be explained by the observation that HU decreased the intracellular dCyd pool, while this was not achieved by exposure to thymidine or 3-deazauridine. We conclude that HU activates ara-C phosphorylation by dCK predominantly through a depletion of the competing dCyd pool. dCyd depletion is most likely caused by inhibition of the de novo pathway of dCTP synthesis and the associated compensatory activation of the dCTP salvage pathway. Intracellular depletion of dCyd may consequently be regarded as a necessary event which allows HU-mediated enhancement of dCyd analog phosphorylation. Agents which only deplete dCTP without affecting the dCyd pool may not modulate araC phosphorylation.

138

140

DRUG RESISTANCE PROFILE: STRONGEST INDEPENDENT PROGNOSTIC FACTOR IN C H I L D H O O D ACUTE L Y M P H O B L A S T I C LEUKEMIA (ALL) IN A B F M - O R I E N T E D TREATMENT P R O T O C O L I I I Kaspers ~JL , Pieters R , Van Zantwijk 2 CH , Van Wer~gg IER , Van Der Does-Van Den Berg A , Veerman AJP'-. Dept. of Pediatrics, Free U~iv Hospital, De Boelelaan 1117, 1081 HV Amsterdam, --Dutch C h i l d h o o d L e u k e m i a ' S t u d y Group, The Hague, The Netherlands. We p r o s p e c t i v e l y studied the prognostic value of cellular drug resistance, assessed with the colorimetric MTT assay. Samples of 152 children with newly diagnosed ALL were successfully tested. Patients received B F M - o r i e n t e d treatment according to one of two successive Dutch Childhood Leukemia Study Group protocols, and had a median follow-up of 34 months from diagnosis. At risk-group stratified, multivariate analysis (including age, white blood cell count, immunophenotype, and DNA p!oidy), in vitro r e s i s t a n c e to prednisolone (PRD), dexamethasone, vincristine (VCR), l-asparaginase (ASP) and doxorubicin each were independently related (pi0.03) to the p r o b a b i l i t y of d i s e a s e - f r e e survival (pDFS). In vitro r e s i s t a n c e to vindesine, daunorubicin, mitoxantrone, mercaptopurine, thioguanine, cytarabine and t e n i p o s i d e were not of independent prognostic significance. Beside drug resistance, only age was a significant p r o g n o s t i c factor in this group of patients. Multiple regression analysis showed that combining the data for PRD, ASP and VCR p r o v i d e d a drug r e s i s t a n c e profile with additional p r o g n o s t i c information t o that of the single best p r e d i c t i v e drugs, PRD and ASP. The 3-years pDFS d e c r e a s e d from 100% for the group with the most sensitive profile to 30% for the group of patients with the most resistant profile (p<0.001, multivariate analysis). This drug resistance profile was by far the most significant prognostic factor in this group of patients. The profile can be used for risk group stratification. Supported by the Dutch Cancer Society (IKA 89-06).

IN, VITRO DRUG SENSITIVITY OF AML CELLS USING THE MI-I" ASSAY A. Mestice, G. Specchia, M.R. Coppi, I. Attolico, A. Ricco, D. Pastore, A. Vaire, V. Caretto, V. Liso Hematology - University of Bad- italy Different assays (clonogenie, dye exclusion, 3H-Thymidine uptake) have been developed to assess the chemosensitivity of malignant cells. The MTT assay, in particular, based on the reduction by living cells, of MTT (3-[4,5-dimethylthiazot-2yl]-2,5-diphenyl tetrazolium-bromide) to a formazan product, provides a simple, automated and efficient method for chemosensitivity testing in Leukemias. In this study, in vitro drug sensitivity was assessed in cells from 32 untreated adult Acute Myeloid Leukemias. Dose response curves were obtained for Cytosine Arabinoside (ARA-C), Daunorubicin (DNR), Idarubicin (IDA), Mitoxantrone (MITOX), Etoposide (VP16). There were marked interiodividual differences of LCS (Leukemic cell survival) values (expressed as percentage of cell survival of the untreated controfs). LCS was less than 50% in 20/29 samples (68.9%) tested with ARA-C, in 26/32 (81.3%) with IDA, in 29/32. (90.6%) with DNR, in 23/'29 (79.3%) with MITOX and in 21/28 (75%) with VP-16. The in vitro results were compared with the clinical response in 23 patients treated by combination chemotherapy according to G IMEMA protocol (schedule: ARA-C + VP-16 + DNR or IDA or MITOX). The M'I-F test correlated well in 13 (57%) out of 23 patients. 10 patients showed chemotherapeutic response sensitive in vitro and resistant in vivo. No AML cases showing chemasensitivity resistant in vitro and sensitive in vivo were observed. Our preliminary resuits suggest that the M'l'r assay may be useful in evaluating chemosensitivity in some AML patients, although further studies are needed to assess the clinical utility of these in vitro chemosensitivity tests.

A 130 141

143

Does P-glycoprotein predict response to chemotherapy? Expression of P-glycoprotein in children's and adults' leukemia correlation with clinical outcome.

Lack of correlation between multi-drug-resistance (MDR) expression and clinical response to chemotherapy in patients with acute myeloid leukaemia

Slawomir Kaczorowski 1, Maria Ochocka2, Maria KaczorovJska3, Michal Matysiak2 1 Department of Immunology, The Maria Sklodowska-Curie Memorial Cancer Center and Institute ofOncology, Wawelska 15, 02-034 Warsaw, Poland 2 Warsaw University Medical School, Department of Pediatric Hematology -Oncology, Dzialdowska 1.0t-184 Warsaw, Poland 3 Department of Internal Medicine - Family Medicine, Postgraduate Medical Center, Czerniakowska 231 00-416 Warsaw, Poland Chemotherapy is still an important treatment method for cure of acute and chronic leukemias. P-glycoprotein (P-gp) is one of the factors responsible for lack of tumor sensitivity to cytotoxic drags, potentially limiting their effectiveness in cancer chemotherapy. In the present study the expression of P-gp was assessed in 45 children and 33 adult leukemia cases as well as in 27 healthy donors by means of immunocytoehemieal (IC) APAAP method and three monoclonal antibodies directed to intra- (C219, JSB-1) and extra-ceUular (MRK-16) part of the protein. As positive staining for P-gp we have considered cases positive with at least two MAb's. Blasts and peripheral blood mononuclear cells (PBMC) obtained from healthy probands were separated by FicoU gradient centrifugation. As negative and positive controls served human myelogeanus leukemia cell line (K562) and Vincristine resistant (K562VCR) subline, respectively. Cell lines and PBMC from healthy donors were as well investigated by means of flow cytometry with C219 MAb. The K562 and K562(VCR) cell line showed consistent reactivity with three antibodies. The expression of P-gp was found in 31% children's and in 33% adults' specimens. Complete remission (CR) was noticed in 12/14 (85%) children's and 9/11 (81%) adults' P-gp(+) leukemia cases. In the P-gp(-) cases CR was observed in 24/29 (82%) and 18/22 (81%), respectively. Partial remission, relapse, resistance and death were noticed in 14% children's and 18% adults' P-gp(+) samples. In the P-gp(-) eases these parameters were observed in 17% and 18~ respectively. We have not observed any reactivity on healthy donors PBMC by means of IC and three MAb's. Flow cytometry analysis revealed reactivity of C219 MAb with lymphocytes (0.21%-6.6%) and monocytes (1.53%-13.6%). Summing up IC is a sensitive method for detection of P-gp(+) cells. Since reactivity and specificity of the MAb's are not fully characterized the minimal panel of three antibodies should be used for detection of MDR phenotype. We can conclude, based on our immuanchistochemieal staining results and clinical data, that expression of P-gp, as detected by panel of monoclonal antibodies, could not serve as alone prognostic factor.

142 CLINICAL RELEVANCEOF P-GLYCOPROTEIN-RELATEDRESISTANCEIN PATIENTS WITH ACUTE LEUKEMIA VOLKMAR NUSSLER 1, RENATE PELKA-FLEISCHER, HEINZ ZWIERZINA, CHRISTOPH NERL, FRANK GIESELER, EDITH GALLIS, BETTINA BECKERT, DIETER HOLZEL, GEORG LEDDEROSE, HANSJORG SAUER, WOLFGANG WILMANNS 1) MedizinischeKlinikIII, KlinikumGroShadern,Marchioninistr.15, 80377Munich, Germany P-glycoprotein (P-gp) is a crucialfactor in thedevelopment of chemotherapy resistance in malignant disorders. Between 1989 and 1994 P-gp expression was prospectively studied in mononuclear bone marrow cells of 304 (221 AML; 83 ALL) acute leukemia patients. In 282 patients P-gp was investigated before and after therapy and in 22 patients only before therapy: 148 AML patients with AML-6 protocol (EORTC), containing daunorabicin, vinaristine and conventional-dose cytarabine (am-C), and 63 AML patients were treated with intermediate-dose am-C plus amsacrine. Further 71 ALL patients were treated according to a German standard polychemntherapy protocol (BMFT04/1989). P-gp was determined by using monoeinnal antibodies C219 and 4E3, and the cutoff point for P-gp overexpressioo was set at> 10%. A significant (p<0.05) difference in P-gp overexpression was demonstrated between AML (21.7%) and ALL (10.2%) patients at primary diagnosis and between primary diagnosis and relapse/refractoriness in AML (21.7%;45A%) and ALL (10,2%;27.2%) patients. According to FAB classification P-gp overexpression was detected in AML patients significantly (10<0.05)more frequently in classes M4, M5a and M5b and tess frequently in M3, as compared to other types. Of AML patients 35% after therapy according to AML-6 protocol and 11.5% after intermediate-dose ara-C plus amsacrine as well as 7% of ALL patients after treatment developed P-gp overexpression. For AML patients with P-gp overexpression at primary diagnosis or early relapse/refractoriness, the predictive value for nonresponse to AML-6 protocol were 90% and 94% respectively, while late-relapsed AML patients with P-gp overexpression had a significantly (p<0.05) lower predictive value of 73% for nonresponse. Additionally, in refractory and late-relapsed P-gp-overexpressing AML patients treated with intermediatedose ara-C plus amsacrine the predictive values for nonresponse were 44% and 38%, respectively, significantly (9<0.05) lower as compared to AML-6 protocolqreated refractory or late-relapsed AML patients. In P-gp-overexpressing treated ALL patients the predictive values of 50% and 55% for nonrespoose were caleulated at primary diagnosis and late relapse, respectively. We conclude that P-gp overexpression is a common phenomenon in AML patients at primary diagnosis or relapse, has an inverse influence on AML-6 treatment outcome and should be taken into consideration in the development of new therapy strategies.

A Schulz, P Engling, Q Cao, C Scheid, V Diehl and PD Wickramanayake 1. Dept. of Medicine, University Hospital Cologne, 50924 Cologne, Germany Expression of the multi-drug-resistance (MDR) protein gp170 encoded by the MDR1 gene has been implied to play a role in acute myeloid leukaemia (AML) refractory to chemotherapy. We investigated MDR expression by flowcytometry and immunocytologically stained slides using the monoclonal antibodies MRK16 and C219 respectively, as well as RT-PCR in 15 patients (7 male, 8 female) with AML. The median age was 49,5 years (range 22-75 years), the distribution according to the FABclassification was MI: t case, M2:5 cases, M3:1 case, M4:3 cases, M5: 2 cases, M6:1 case and 1 biphenotypic case. Induction chemotherapy consisted in TAD/HAM (9 cases), Ida-FLAG (5 cases) or AVA-7 (2 cases). The proportion of leukaemic blasts in the bone marrow varied between 30 and 100 % with a median of 95%. In 7 patients who subsequently achieved complete remission (CR) 5-47% (median 27%) of bone marrow cells were positive for MRK16 compared to 9-47% (median 29%) in 8 patients without CR. The flowcytometry results were confirmed by immunocytochemistry and RT-PCR.A high percentage of CD34+ cells was correlated with failure to achieve OR. Otherwise there was no correlation with age, FAB classification or chemotherapy protocol. In conclusion MDR expression does not seem to predict treatment failure in AML, however it has to be considered that expressed MDR protein may not necessarily be functional.

144

PROGNOSTIC SIGNIFICANCE AND THE RELATIONSHIP OF MYO AND I~)R-1 -GENE EXPRESSION IN ADULT ACUTE LEUKEMIA A. Zaritskey, T. Bykova, I.Stuif, A. Sominskaya, N. Anikina, N. Medvedeva, B. Afanasiev BMT CENTRE, CENTRE FOR ADVANCED MEDICAL

TECHNOLOGIES, PAVLOV ~,IEDICAL UN1VE,r~ITY, PETROV INSTITUTE OF ONC.'OLOGY, ST. PETERSBURG, RUSSIA

The myo gene (proliferation related) ~ d ~R-i gene (responsible for drup; resistance) a~e the subjects of" a nulr~oer of laboratory studie-3 dealing with the biology of leukemic ceils (it), neanwhile the clinical data are poorly evalueted. We have. studied 23 patient:':;,#it,h acute myelogenous leukemia (AML) before ~ d durin~ chemotherapy. Expression of above mentioned genes '#as :studied by in situ hybridization method. N.~SR-Igene expression in io depended upon previous therapy. Regimens including vinca alkaloids / epipodophylline + ~thracyolines (A) ve~us only A resulted in higher percentage of patients w i t h }r gene expressing leukemic cells (P< 0,05). lh~ i r r a r y or drug induced Ir gene

expression was si.~nificant, according to prognosis only on standard dose therapy whereas high dose Ara-~C regin~ns overcame these differences. Myc gene expression in lc negatively correlated to complete remission primarily therapy in

rate (P
Further analysis has shown this correlation to be valid only when using induction regimens more inLensive than "7+3" based. Coexpression of" ~3R-I and. myc genes correlated to resistant/relapsed disease (P
A131 145

147

Treatment results of the EORTC-Children's Leukemia Cooperative Group (CLCG) protocols 58831/2 and 58881 in 177 children with T-cell acute lymphoblastic leukemia (ALL) or non Hodgkin lymphoma (NHL). E.BAHNER-ENNASSIH, S.SUCIU, Y.BERT, RAND, AM.MANEL, JP.MAGAUD, A.THYSS, A.BABIN-BOIIzLETOT, A.ROBERT, M.MAZINGUE, J.OTTEN and N.PH1LIPPE. Of the 1624 newly diagnosed patients entered in the two studies between October 1983 and December 1992 151 were identified by immunophenotyping as T-Cell ALL, 26 as T-Cell NHL (< 25 % blasts in bone marrow). 70 Children received the therapy protocol EORTC-CLCG 58831/2, 107 the 58881 (81 T-Cell ALL, 26 T-Cell NHL). These BFM-like protocols consist of induction/consolidation phase, interval therapy for CNS prophylaxis, reinduction/reconsolidation phase and maintenance therapy to complete 2 years of treatment. In protocol 58831/2 patients were randomized for cranial irradiation after reconsolidation. 97 % (study 58831/2), 93 % (T-Ceil-ALL) and 100 % (T-Cell NHL) in study 5888 l achieved complete remission (C.R.) ; all others patients died, Four year event-free survival (EFS) and overall survival (S) were respectively 65 % and 73 % in 58831/2 protocol, 56 % and 69 % for T-Cell ALL, 83 % and 90 % for T-Cell NHL in 58881 protocol. The fi)lhiwing factors are analyzed for their effect on EFS : age, mediastinal mass, stage of Murphy classification, immunophenotype (early T, intermediate T, m~tture T), caryotype, WBC count, hemoglobin, calculated risk factor RF (0,2 tog (blasts/mm3 + t) + 0,06 hepatomegaly (cm) + 0,04 splenomegaly (cm)), cortico-sensitivity and cranial irradiation (DFS). The only significant predicting factors in EFS were RF (< 1.7 : 75 % ; > 1.7 : 42 % in 58831/2 study) WBC coant (< 100 x 109/1 : 82 %, > t00 x 109/1 : 50 % in 58831/2 study) and stage of Murphy classification 0II : 92 %, IV : 76 % in study 58881). Initial nivoh,ement was present in 13 % ofT-Cell ALL in 58881 study. Relapses occured during the second year of treatment by preference in CNS (> 50 %). These results suggest the importance of separately intensive therapeutic strategy for T-CeU ALL according to the RF, to improve their outcome ; this could be include reintroduction of cranial irradiation and increasing of intra-thecal chemotherapy.

INTENSIVE CHEMOTHERAPY IN POOR PROGNOSIS RELAPSED ACUTE LYMPHOBLASTIC LEUKEMIA IN CHILDHOOD : A pilot Study of EORTC Children's Leukemia Cooperative Group. Y.Bertrand E.Plouvier, H.Rubie, A.Bobin, A.Ferster, P.Latz, G.Souillet, A.Thvss, AM.Manel, J.Ottea and N.Phitippe for the EORTC Children's leukemia cooperative group (CLCG). Between 7/90 and 4/94, 63 patients (pts) with poor prognosis relapsed (PPR) acute tymphoblastic leukemia (ALL) underwent intensive therapy followed by bone marrow transplantation (BMT). PPR was defined by CRI < 18 months duration and/or T-cell ALL relapse and/or relapse occuring after very intensive first line therapy (bad response to prednisone prephase therapy, no CR after the end of induction therapy, t (9 ; 22), t (4 ; 11), near-haploidy). The induction regimen included two courses cytosine arabinoside (2 g/m2 x 2/d-dl, d2), mitoxantrone (8 mg/m2/d-d3, d4), etoposide (150 mg/m2/d-d3, d4, d5), dexamethasone (20 mg/m2/d, dl-5), asparaginase (10.000 U/m2/d-d7, d9, dl 1, d13) and intrathecal methotrexate + cytosine arabinoside (d5 if no CNS involvement, twice a week if CNS involvement). Thirty-seven pts had isolated bone marrow relapses, 14 pts had combined bone marrow relapses and 12 pts had extrmnedallary relapses. Fifty-five pts are evaluable for response to induction : a complete remission was obtained in 43 pts (78 %).Ten pts had refractory disease and 2 pts died (toxic death). Time to 0.5 x 10WL neutrophil recovew was 21 d (9-60 d) for the first course and 15 d (9-20 d) for the second. Time to 50 x 109/L platelet was 18 d (8-40 d) and 10 d (4-16 d).Toxicity included nausea, vomiting, mucositis, diarrhea, gastro intestinal hemorrhage and neurologic toxicity (in 2 cases with isolated CNS relapse). Twenty-two pts received aUogeneic BMT in CR2 (9 genoidentical, 10 haploideatical, 3 unrelated donor BMT) : 8 pts are in CCR (6-40 m). Twelve pts underwent autologous BMT (the majority of then had extramedullary relapses) : 8 pts are alive and well (11-30 m). Among the last 9 pts in CR after intensive induction therapy, 3 relapsed before BMT and 6 are ongoing.With a 78 % CR rate, we conclude that this combination is effective in induction of a second complete remission in PPR childhood ALL with an acceptable toxicity. The pts achieving CR are in good condition for subsequent therapy including BMT (allogeaeic or autologous). BMT should be performed as soon as possible to avoid a subsequent relapse.

146

148

PFtELtMINARY TREATMENT RESULTS IN CHILDHOOD B-NHL WITH A MODIFIED BFM-90 PROTOCOL IN ST. PETERSBURG

SIMILAR O U T C O M E IN BOYS WITH ISOLATED AND C O M BINED TESTICULAR ALL RELAPSE AFTER STRATIFIED BFM SALVAGE T H E R A P Y

M. Belogurova, B. Kolygin, G. Radulesku, M. Kirichenko. (Centre for Advanced Medical Technologies, St. Petersburg, ~ussia) From Feb. 1992 to May 1994, 21 newly diagnosed B-NHL children (19 boys, 2 girls) aged 3-14 years were treated according to a modified B-NHL BFM-90 protocol (A. Reiter). 19 of the 21 pats had abdominal tumors, and 2 had lymphomas at the neck. Diagnosis of Bcell NHL was histologically confirmed in 19 pats and was based on FAB L3 morphology of lymphoblasts in malignant effusions in 2 pats. Stages according to Murphy classification (1988) were: 3 pats st. II, 9 pats st. III, 9 pats st. IV. All children were allocated to the risk groups according to the original BFM-protocoh 2 pats in risk group 2, 19 pats in risk group 3. The original BFM-~O protoc,QI was modified by reducing the dosage of MTX from 5g/m'r to Ig/m z. Results: Complete remission was achieved after 2 blocks of chemotherapy in 13 pats, and after the third or fourth block in 5 pats. Due to severe treatment-related blood cell depletion transfusions of red blood cells and platelets were frequently used. All pats developed neutropenia, 4 0 % of them had various infectious complications (pneumonia, enterocolitis, septicemia). 1 pat developed a peritonitis after the first AA block due to intestinal perforation. One toxic death was observed: a five year old boy with abdominal B-NHL developed severe enterocolitis after the first AA block and died on the 18th day after beginning of treatment. - 18/21 pats achieved complete remission, 2 pats were non-responders and died, 1 pat died from toxic complication, 1 pat developed early local relapse and died due to tumor progression. The projected overall survival at 30 months (Kaplan-Meier estimate) is 80% (SD 9%), - Conclusions: The preliminan/ results obtained with the modified B-NHL BFM-gO protocol are clearly superior compared to the historical controls of our centre. A high percentage of children has achieved remission with tolerable toxicity.

C. Wolfroml, R. Hartmannl, S. Br0hm011erl,R. Fenglerl, A. Reiter2,J. Ritter3,and G. Henzel; for the BFM RelapseStudy Group. UniversityChildren's Hospitalsof Bedinl (UKRV), Hannover2(MHH), M~nster3;Germany Objective: To evaluate the efficacy of tumor-load adapted multi-

agent chemotherapy, 114 boys with first overt isolated or combined testicular relapse of acute lymphoblastic leukemia (ALL) were treated according to four consecutive ALL-REZ BFM protocols during 1983 and 1993. Methods: Systemic treatment consisted of alternating chemotherapy courses followed by conventional maintenance therapy. Local treatment of unilateral relapse comprised orchiectomy and irradiation of the contralateral testis, and irradiation in bilateral relapse (24 Gy). Patients (pts) were stratified according to site and time of relapse. Boys with isolated testicular relapse received 6 courses. Depending on time of relapse and study design, 2 or 3 additional courses (a total of 8 or 9) were given to pts with concomitant bone marrow (BM) involvement. Duration of maintenance was 1 year or 2 years for pts with isolated or combined relapse, respectively. Results: Event-free survival (EFS) rates at 8 years were .53_ + .07 in 59 pts with isolated testicular relapse and .52_+.07 in 55 pts with combined relapse. Pts with late relapses (beyond 6 months after cessation of front-line therapy) had a significantly better outcome than pts with early relapses (EFS 0.65_+.06 vs 0.33_+.07, P<.001). EFS rates of pts with early isolated (0.32+.09) and early combined (0.37_+.13) relapse were virtually identical. Likewise, BM involvement was not of statistical significance in pts with late relapses (EFS 0.78-+ .09 for isolated and 0.57_+ .08 for combined relapse, P > .05). Conclusion: Treatment according to the concept of the BFM relapse trials is effective in boys with testicular relapses. Irrespective of time of relapse, results are comparable in pts with and without concomitant BM involvement. We conclude that pts with combined relapses benefit from the additional courses of chemotherapy given in respect to the higher leukemic cell burden. - Supportedby grantsfrom the "DeutscheKrebshilfe" -

A 132 149

151

I M P A C T OF E A R L Y T R E A T M E N T I N T E N S I T Y ON O U T C O M E AFTER FIRST R E L A P S E OF C H I L D H O O D A L L

OUTCOME OF 14 PATIENTS WITH PHILADELPHIA-CHROMOSOME-POSITIVE ALL RECEIVING AUTOLOGOUS OR ALLOGENEIC BONE MARROW TRANSPLANTATION M.Stockschlader a, S.Hegewisch ~, W.Krtigera, W.Zeller~, B.Kohlschfittera, H.Kabiseh~, S.Peters ~, H.Martin', HJ Weh ~, DK Hossfeld ~, A.Zander~, Bone Marrow Transplantation, Dept. of Hematology/Oncology, Universit~llskrankenhaus Eppendorf, Hamburg a, Bone Marrow Transplantation, Frankfurt*. 14 patients (pts)(median age 30 [2 - 48] years) with ALL (12 Ph+, 2 bcr/abl+) received TBI (12 Cry) (n=13) or busulfan (16 mg/kg) (n=l), VPI6 (30 - 45 mg/kg) (n=]4), and cyclophosphamide (120 mg/kg) (n=14) followed by autologous (n=4) or allngeneic (n=10) BMT. Autologous marrow was purged (n=2) with mafosfamide (n=l) or immunologically (n=l). Two pts received unpurged autologous marrow only. One patient receiving purged marrow received an unpurged marrow boost because of impending graft failure. All pts receiving autologous marrow were in CR (3 in 1.CR, 1 in 2.CR) at the time of transplantation. For allogeneic BMT (9 related, 1 unrelated donor) 7 pts were in I.CR, 1 patient in second relapse, and 2 pts in first refractory relapse. Results: The median follow-up for all pts is 15 (1 50) months. 8/14 pts (57 %) are alive in CR [6/10 (60 %) receiving allogeneic and 2/4 (50 %) receiving autologous BMT]. Major causes of death for both groups were relapse (n=3) and transplant-related toxicity (VOD; sepsis)(n=3). All pts tested initially (4 wks post BMT) negative for the bcr/abl rearrangement. 2/3 pts who relapsed were repeatedly bcr/abl+ prior to relapse (test not done in the third). Conclusion: Considering the negligibly low curability of Ph+-ALL with conventional chemotherapy regimens our data support the concept of early (> 1.CR) BMT (allogeneic > autologous [purged]) following triple therapy with TBI, VP16, and CTX for pts suffering from this form of high-tisk ALL.

R. Hartmann, D. Hubalek, R. Fengler, and G. Henze; for the BFM Relapse Study Group. Rudolf Virchow Medical Center, University Children's Hospital, Bedin, Germany Objective: One third of patients (pts) with relapse of non-B/non-T acute lymphoblastic leukemia (ALL) can be cured with intensive retreatment like SFM salvage chemotherapy consisting of alternating multidrug courses for remission induction and consolidation. To assess the Influence of dose intensitity, data from 229 pts of the multieenter trial ALL-REZ BFM 90 have been analyzed. Methods: Relative dose intensities (RDI) expressed as percentage of applied to planned dosages per time unit have been calculated according to standard formulas for all courses, for each individual drug, and for drug combinations. Using RDI instead of absolute amounts of doses allows to include all pts in survival time analyses, except for induction deaths. Results: Differences in RDI were mainly caused by treatment delays reflecting slow clinical and hematological reconstitution rather than by significant drug dosage modifications which were seen in < 10% of courses. Outcome of pts was strongly dependent on RDI during induction therapy. All 65 pts receiving a second course within 20 days after the first one achieved complete remission (CR); 4 of 100 pts (4%) continuing therapy between days 21 and 25 failed to achieve CR as did 8 of 64 pts (13%) with more than 25 days delay. In these groups, event-free survival rates at 4 years were .75+-.06, .37+-.07, and .26+-.08, respectively (P<.O01). Biases by known risk factors did not completely explain differences in outcome. In multivariate analyses controlling for time and site of relapse, induction treatment delays were independently correlated with survival and event-free survival times (P <.05). Conclusion: These findings emphasize the importance of early intensification of salvage therapy. Stdctty time-scheduled administered induction courses increase RDI but may not necessarily improve overall results as toxicity may be increased as well. A randomized prospective study would be required to answer the question whether prognosis could be improved by intensified supportive therapy like use of G-CSF during induction treatment. - Supported by grants from the "Deutsche Krebshilfe" -

150

152

ACUTE CHOLECYSTITIS COMPUCATING INTENSIVE CHEMOTHERAPY OF CHILDHOOD ACUTE LYMPHOBLASTIC LEUKEMIA.

P R O G N O S T I C S I G N I F I C A N C E OF P E A N U T A G G L U T I N I N (PNA) B I N D I N G IN C H I L D H O O D A C U T E L Y M P H O B L A S T I C L E U K E M I A 1 12 , 2 Kaspers GJL , V e e r m a n ~ J P " , Van Werlng I ER , Van Der L i n d e n - S c h r e v e r ~EM ~, Van Zan~wij~ CH-, Van Der D o e s - V a n Den Berg A , Pieters R . Department of Pediatrics, Free U n i v e r s i t y Hospital, De Boelelaan 1117, 1081 HV Amsterdam, and --Dutch C h i l d h o o d L e u k e m i a Study Group, The Hague, The Netherlands. W e p r e v i o u s l y r e p o r t e d a favorable p r o g n o s i s a s s o c i a t e d with p o s i t i v e P N A - b i n d i n g in c h i l d h o o d T-cell ALL (Veerman et al., Cancer Res 1985,45: 1890). We now p r o s p e c t i v e l y s t u d i e d , w h e t h e r PNAb i n d i n g was r e l a t e d to the prognosis and to the e x t e n t of in vitro r e s i s t a n c e to 2 g l u c o c o r t i c o i d s and 12 other drugs in 202 c h i l d r e n with n e w l y d i a g n o s e d ALL. M e t h o t r e x a t e was not t e s t e d b e c a u s e of technical problems. In v i t r o drug r e s i s t a n c e was a s s e s s e d using the c o l o r i m e t r i c MTT assay. P N A - p o s i t i v i t y was more f r e q u e n t (p<0.001) in Tcell ALL (65%) than in p r o - B (0%), common (17%) and p r e - B (16%) ALL. P N A - p o s i t i v i t y was a s s o c i a t e d w i t h a 4 . 6 - f o l d lower risk of any event then PNAnegativity, at i m m u n o p h e n o t y p e - s t r a t i f i e d analysis (p=0.025). P N A - b i n d i n g was not r e l a t e d to the e x t e n t of in vitro r e s i s t a n c e to any drug. However, P N A - p o s i t i v i t y was r e l a t e d to the c l i n i c a l r e s p o n s e to a one week p r e d n i s o n e therapy, given together with one intrathecal injection with methotrexate. Poor responders were found e x c l u s i v e l y in the P N A - n e g a t i v e group. W i t h i n TALL, 9 out of 13 P N A - n e g a t i v e p a t i e n t s showed a poor response, while none of 25 P N A - p o s i t i v e cases s h o w e d a poor r e s p o n s e (p<0.0001). We c o n c l u d e that P N A - b i n d i n g is a p r o g n o s t i c factor in c h i l d h o o d ALL, and seems useful in Tcell ALL. The p r o g n o s t i c s i g n i f i c a n c e of PNAb i n d i n g c o u l d not be e x p l a i n e d by a relation w i t h in vitro resistance to any of 14 drugs tested, but m e t h o t r e x a t e was not i n c l u d e d in the drug panel. (Supported by the Dutch C a n c e r S o c i e t y KWF 89-06).

T.Urasifiski, A.Brodkiewicz, A. Walecka*. Dept. of Pediatrics and Dept. of Radiology*, Pomeranian Medical Academy, Szczecin-Poland. Acute cholecystitis is a rare disorder in the pediatric age group with age over 10 years, female gender, obesity and chronic hemolysis considered as predisposing factors. We report on 3 cases of acute cholecystitis complicating intensive chemotherapy of childhood acute lymphoblastie leukemia. Two girls and a boy aged 3, 4 and 10 years, aH with newly diagnosed acute lymphoblastic leukemia treated according to BFM 86 protocol developed abdominal pain localized in right abdominal quadrant and epigastic region, nausea and vomiting accompanied by septic fever with chills. Diffuse abdominal tenderness and guarding were seen in one patient. A tender mass below the right costal margin was felt in another patient. Murphy sign was positive in all of children. Abdominal ultrasound revealed distended gallbladder with thickened walls in all of patients. Gallbladder stones were seen in none of children. Blood cultures were negative. Therapy with antibiotics (cephalosporins in combination with semisynthetic penicillins plus metronidazole in one of children) resulted in rapid resolution ( 3-5 days ) of all clinical and ultrasonographic signs and symptoms. All three episodes occurred during granulocytOpenia caused by intensive chemotherapy ( protocol ! in 2 patients, protocol il in 1 child), while on l-asparaginase administration. We would like to draw attention that acute cholecystitis can be a septic complication in granuioeytopenic children treated for acute lympho-

blastic leukemia. Authors address: I Pediatric Department, Pomeranian Medical Academy, ul. Unii Lubelskiej 1, 71-344 Szczecin - Poland

A 133 153

155

LONGITUDINAL STUDY OF COGNITIVE, MOTOR, AND BEHAVIORAL DEVELOPMENT IN CHILDREN DURING AND AFTER TREATMENT OF ACUTE LYMPHOBLASTIC LEUKEMIA (ALL): AN INITIAL REPORT FROM THE CHILDRENS CANCER GROUP

Childhood ALL was considered a fatal condition 40 years ago, but is curable in more than 70% of patients today. The death rate is being lowered by higher chemotherapeutic intensity early in treatment, while risk factors for late developmental sequelae are defined. We present preliminary results from the first large scale longitudinal study of neurobehavioral d6velopment (CCG-105N P) in childhood ALL patients from a randomized clinical trial (CCG-105). More than 200 patients with intermediate-risk ALL were examined at 9, 21, and 48 months after diagnosis. These children had been randomized to: a) one of four systemic chemotherapy approaches, comparing CCG versions of BFM four drug induction/consolidation (t/C) and delayed intensification (DI) versus standard three-drug induction; and b) CNS-directed therapy using intrathecal methotrexate (IT MTX) alone throughout treatment versus 1800 cGy cranial irradiation and IT MTX during I/C and DI. During treatment, standardized test results indicated no significant differences in cognitive and motor development associated with DI o r type of CNS-directed therapy. After completion of ALL therapy, patients 1-2.5 years at diagnosis who received IT MTX alone scored significantly higher on 5 of 6 McCarthy Scales indices. For patients ->6 years at diagnosis, there were no significant differences on any of the Wechsler Scale mean scores associated with either CNS treatment regimen. Findings from parent-completed questionnaires indicated behavior problem rates in the study sample comparable to those in the US population. Additional clinical and statistical analyses of study data and of the interactions among patient, disease, and therapy-specific factors which mediate developmental outcome are currently in progress.

RESULTS OF INTENSIVE CHEMOTHERAPY OF A M L IN C H I L n R E N A C C O R D I N G BFM STRATEGY: A SINGLE INSTITUTION (TRIAL) A.A. Maschan, L.V. Baydun, A.I. Karachunsky, M.]. Kharit, O.I. Kryzhanovsky, E.V. Samochatova, T.V. Zhiganova, A.G. Rumjantsev Instituo of Children Hematology of Russia, Moscow In order to improve the poor results of treatment of childhood AML we applied the BFM strategy for patients with AML admitted to the clinic of our Institute. Reference group included 29 patients treated with courses "7+3" and "5+2" (conventional doses of Ara-C and daunorubicine). The results of such treatment were as follows: remission was.achieved in 58,6 % of pts, relapse-free survival at 42 months was 21,4 %. Between February 1990 and May 1994 a total o f 46 consecutive patients with AML were treated according to BFM strategy (compilation from AML-BFM-83 and AML-BFM-87 studies). 17 males and 29 females. Mean age was 8 years (115 years). The FAB-subtypes distribution was as follows: M0-4,25 %: M16,45 %: M2-19,1%: M3-10,65 %: M4-23,4 %: M5-27,75 %: M6-6,4 %. All pts reveived an 8-day induction course: conventional doses of AraC, datmorubicin 30 mg/sqm bidaily - 6 doses and etoposide 150 mg/sqm a day 3 doses. Six drugs (prednisolon, 6-MP, vincristin, AraC, eyelophosphamide, doxorubicin) containing consolidation course lasted 6 weeks. 23 pts received intensification therapy: 20 pts with high doses of AraC + etoposide, 3 pts - I-ID AraC + mituxantrone. 12 pts did not receive intensification course. CNS prophylaxis included AraC intrathecally and cranial irradiation (18 C-y). As a result of the treatment 35 pts (76 %) achieved complete remission. There were 11 induction failures: 6 toxic deaths (13 %) and 5 primary refractoriness (10 %). 1 patient died in CR aider intensification course. Of 34 pts completed intensive phase of therapy 5 patient relapsed (15 %). The aEFS is 63 % at 51 months. Conclusions: our cohort of AML patients is charactefised by peculiar distribution of FAB subtypes with prevalence of M4 and M5 subtypes and need intensive regimens of therapy. The BFM strategy of intensive treatment showed high effectiveness for the presented group &patients with AML.

154

156

CHILDHOOD ACUTE NONLYMPHOBLASTIC LEUKEMIA - RESULTS O F T H E AUSTRIAN-HUNGARIAN AML-STUDIES IGCI..84 AND 90 Fink F.M?, Mann G?, Masfit P?, Panzer-Griimayer E.R.=, Kajtar P., Kardos G., Urban Ch., Schuler D?, Gadner H?, for the IGCI Pediatric Study Group. Dept. of Pediatrics, UniversityInnsbmck1, St.Aana Children's Hospital,Vienna2, Austria Teaching Hospital Szombathely~, SemmelweisUniv. Med: SchoolBudapest',Hungary

GM-CSF IN THE THERAPY OF DE-NOVO AML PATIENTS: AN UPDATE OF A DOUBLE-BLIND RANDOMIZED, PLACEBO CONTROLLED TRIAL

Between 1984 and 1992 135 children with de novo ANLL - 68 boys, 67 girls, median age 7;06 years (1 day - 17;10 years) - were treated in two consecutive prospective cooperative Austrian-Hungarian AML-studies (AML-IGCI-94 and 90, respectively). FAB subtypes inoluded M 1 (n=46), M2 (n=t9), M3 (n=2), M4 (n=34), M5 (n=24), M6 (n=2) and M7 (n=8). The median initial white blood count was 15.8 G/I (0.4 - 1350.0), 5 patients showed CNS-involvement. Study design IGCI-84: Induction course I-1: eytosin arabinoside (CA) 100mg/mVday continuous i.v. infusion on days 1 & 2; CA 100mg~mVl2hours i.v., aclarubicin (ACR) 25mg/m=/dayi.v. and etoposide (E) 100mg/mZ/day i.v. on days day 3 - 7. Second induction course I-2 (BFM-83 [1]: CA, daunorabiein, E) only for patients not in complete remission (CR) after I-1. Consolidation, maintenance (both BFM-83 [1]) for 2 years from diagnosis, and cranial irradiation. Study design IGCI-90: Induction I-1 with increased ACR-dose (30mg/mZ/day),1-2 (ARA-C 6x2g/m2, idarubicin 3xl0mg/m=), 1-3 (ARA-C 4x3g/m2, mitexaotrone 2xl0mg/mZ), consolidation and maintenance (both BFM-87 [2]) for 1.5 years from diagnosis, no irradiation. Results (med. follow-up 6;05 years [IGCI-84] and 2;08 years [ICrCI-90], resp.):

1Dept. Internal Medicine III, University of UIm; 2Dept. Haematology/Oncology, Universityof Frankfurt;3Dept. Haematology/OneologyUniversity of Mainz;4Dept. Haematology/Oncology University of Freiburg; SBehringwerke AG,Marburg, Germany

Kaleita T., Sather H., Ruymann F., Noll R., Stehbens J., O'Brien R., MacLean W., H a m m o n d D. Childrens Cancer Group, Arcadia, CA, USA

Study n CR CRbyl-1 DeathCR1 BMTCRI pEFS* pEFI* 84 95 64 (67%) 47 (73%) 5 7 .30 .44 90 40 33 (83%) 24 (73%) 5 3 .37 .46 total 135 97 (72%) 71 (73%) 10 10 * BMT-patieniswere censoredfrom all analysesat the date of grafting in CRI. Conclusion: Induction chemotherapy including aolatubicin as first line anthracycline is effective. Improved management of the indaction-related problems resulted in an improved remission rate in the second study. Remission duration, however, was not improved. The problem of deaths in CR remains to be solved in future protocols. References: [1] Blood 75:1932-1940,1990, [2] J ClanOncel 11:279-286,1993

G. Heil ~, L.Chadid 1, D. Hoelzer2, G. SeipeIC, P. Mitrou 2, Ch. Huber3, K. Kolbe3, R. Mertelsmann 4, A. Lindemann 4, J. Frisch s, U. Nicolay S, W. Gaus1 and H. Heimpel1,

GM-CSF given concomitant and after chemotherapy (CT) should improve the outcome of AML patients by increasing the efficacy of CT and by reducing the rate of infectious complications. Induction and early consolidation therapy included cytarabine (Ara-C, 100 mg/m 2, day 1 -8 CIVI), daunorubicin (60 mg/m 2, IV, day 3 - 5,) and etoposide (100 mg/m 2, day 4 - 8, 2 h IV infusion) with reduced dosages in the second induction and early consolidation course. Late consolidation included one cycle with high-dose Ara-C (3g/m 2, 12 doses) and daunorubicin (30 mg/m 2, day 7 9) for patients aged -< 50 years, whereas patients > 50 years received a reduced dose of Ara-C (0.6g/m 2, 12 doses). Patients were randomized after the first induction course to receive either rhu GM-CSF (E. coil, 250 pg/ m2/ day, s.c.) or placebo starting 48 hours prior to the second induction and the subsequent courses and given throughout chemotherapy until absolute neutrophil count had recovered > 500//11. The overall complete remission rate of the 82 patients included into the study was 78% with 81% in the GM-CSF versus 79% in the placebo group (p = 0.57). After a median follow-up of 35 months the probability of relapse-free survival (RFS) at 41 months was 42 % for the GM-CSF patients versus 41% for the placebo group (p = 0.89). Patients aged <50 years did better under GM-CSF with a probability of RFS of 65% versus 58% (p = 0.31 ), whereas the GM-CSF patients aged > 50 years did worse (20% versus 31% in the placebo group, p = 0.28). The survival rate of the GM-CSF patients was 45% versus 49% of the placebo group (p = 0.66). Again outcome of patients aged -< 50 years appeared to be improved by GM-CSF (70% versus 50%, p = 0.26) whereas survival rate of the patients aged > 50 years seemed to be impaired under GM-CSF (24% versus 50%, p = 0.08). Taken together, GM-CSF as an adjunct to AML therapy is feasible, while it's effect on treatment outcome might correlate with the age of the patients.

A134 157

159

HIGH-DOSE CYTOSINE ARABINOSIDE AND DAUNORUBICIN POSTREMISSION THERAPY IN ADULT DE-NOVO AML PATIENTS: LONG-TERM RESULTS OF A PROSPECTIVE MULTICENTER TRIAL

In this p r o s p e c t i v e t r i a l t h e i n f l u e n c e of a h i g h - d o s e c y t o s i n e arabinoside/daunorubicin (HD-Ara-C/DNR) consolidation therapy on treatment outcome was studied in patients with de-novo AML aged 50 years or less. A total of 149 patients entered the study. All patients had received the same induction and early consolidation therapy with DNR, Ara-C and etoposide (DAV). HD-Ara-C/DNR therapy included Ara-C at 3g/m 2, 12 doses (HD-Ara-C/DNR I) respectively 8 doses (HD-Ara-C/DNR II) followed by DNR 30 mg/m 2, on 3 days each. A complete remission (CR} was achieved in 104 (70%) patients. 61 complete responders received at least one cycle with HD-Ara-C/DNR. The median follow-up was 80 months. The probability of relapse-free survival (RFS) of all 104 CR patients at 100 months was 29% (95% confidence interval, t 7% to 40%) with a median relapse free survival (MRFS) of 19.9 months. The probability of RFS for those 61 CR patients, who had received at least one HDAra-C/DNR consolidation course, was 33% (95% CI, t 9 % to 48%) with a MRFS of 32 months. The median survial time (MST) of all 149 study patients was 23 months and the probability to be alive at 103 months was 28% (95% CI, 18% to 37%). MST of the 61 CR patients with HD-Ara-C/DNR consolidation therapy was 56 months with a probability of survival at 103 months of 49% (95% CI, 34% to 63%). The median duration of the event-free survival (EFS) of all 149 patients was 9.7 months with a probability of EFS of 20% (95% CI, 12% to 28%) at 103 months. These data confirm, that a HD-Ara-C/DNR based postremission therapy is suitable to improve the Iongterm outcome of a subgroup of de-hove AML patients. Despite this progress relapse is still common, indicating a necessity for further therapy intensification and/or identification of risk groups, which might benefit from alternate treatment strategies.

7+3 w 7+3+VP-16 as remission induction in AML patients under the age of 60: preliminary results of Russian multicenter trial H. Parovitchnikova, V. Savchenko, V. Isaev, H. Maslova for Russian Leukemia Research Group The goal of the study was to determine the rol6 of VP-16 as an additional component to the standard 7-3 regimen in de novo AML treatment. VP-16 was infused for 5 days at the dose fo 120 mg/m2 started on day l0 after finishing 7+3 (ara-r 100 mg/m2 bid 7 days, daunorubiein 45 mg/m2 3 days). Treatment for each patient consisted o f 4 cycles 7+3 or 7+3+VP-16 according to the ann o f randomization. All patients with M3 were treated with 7+3 regimen. The type of maintenance was also defined according to randomization: repetition 0f7+3 with 56 weeks interval or rotation programme (5 days ara-e with daano or cyclophosphoraideor 6-MP) with 4 weeks interval for 3 years of CR. Since September 1993 till October 1994 181 patients less than 60 yrs old with de novo AML were enrolled in the study from 19 participating hematological centers. 91 pts were randomized to 7+3 arm and 90 pts to 7+3+VP-16. 45 pts were excluded from the analysis (too early to evaluate - 9, wrong diagnosis - 5, death before treatment- 6, secondary or pretreated AML - 7, protocol deviation - 10, loss from follow-up - 5, pts refusals - 3). 9 M3 patients were analysed separately. Thus, 69 pts (In - 31, f - 38, median age - 33 yrs, MI-10, M2-27, M4-29, M5-3) on 7+3+VP-16 and 58 pts (in - 32, f - 26, median age - 35 yrs, MI-10, M2-23, M418, M5-6, M6-1, M7-1) on 7+3 were finally evaluated. CR rate was 65 % and 58 %, N R 10 % and 19 %, ED 25 % and 22 % respectively. Death rate in consolidation phase was 6 % on 7+3 and 16 % on 7+3+VP-16. Among 9 M3 pts 5 died during induction (cranial hemorrhages), 4 achieved CR. Event-free survival in two years on 7+3 arm was 52 % and 45 % on 7+3+VP-16 (p=O,8). 23 pts were randomized to 7+3 maintenance and 28 pts to rotation programme. Event-free survival according to the type of maintenance was 75 % on 7+3 and 40 % on rotation programme (p
158

160

G. HeiP, P.S. Mitrou 2, D. Hoelzer 2, M. Freund 3, H. Litlk a, G. Ehninger 4, B. Steinke +, S. 0hi s, H. Wandt 5, E. Fackler-Schwalbe 6, G. Schlimok 6, A. L6sch 7, W. Quei6er ?, B. L6ffler 8, W. Gaus 9, J. H69eP, H. HeimpeP and E.Kurrle ~ 1Dept. Int. Med. IlL Uni. Ulm, ~Dept. Haematologie/ Ontology, Uni. Frankfurt, 3Oept. Haematology / Oncology, Uni. Hannover, ~ Dept.Haematology,Uni. Tgbingen, s Dept. Int. Med. and Institut for Medical Oncology/ HaematologyNgrnberg, e Dept. Int. Mnd. Zentralklinikum Augsburg, 7 Dept. Onkology, Klin~kumMannheim, s Dept. Internal Medicine. Robert-Boseh-Krankenhaus,Stuttgart, o Dept. B[ometryand Medical Documentation,Uni. Ulrn, FRG

IDAROBICIN, ARA-C, ETOPOSIDE (ICE) vs RISK ADJUSTED DAUNORUBICIN plds ARA-C (3+7 days, 9 HD ARA-C ) FOR INDUCTION TREATMENT OF AML.; A RANDOMISED MULTICENTER TRIAL. J.Holowiecki, S.Krzemien, Z.Kuratowska, A.Hellman, E.Krykowski, A.Dmoszynska, A.Skotnicki, L.Konopka, S.Maj for the Polish Adult Leukaemia Group. The aim of the study was to evaluate the efficacY, and tolerabUity of the ICE remission induction protocol in comparison to the modified, risk adjusted DNR+AraC 3+7 schedule. Fifty six adults with primary AML (med. age 36y.,range 16-60, males 46%) were centrally randomised to obtain one of the following two induction treatments: 1) ICE 7/10: Idarubicin 10 mg/m2 iv on days 1,3,5 cytarabine (Ara-C) 100 mg/m2/d c.i. days1 to 7/10, etoposide 100 mg/m2/d days 1-5. 2) DA 3+7, :rl-IDAra-C: Daucombioin 60 mg/m2/d iv on days 1,2,3, Ara-C 100 mg/m2/d ci on d.1-6 and then modified dependable on the day 6 bone marrow biopsy; in respenders displaying >30% blast reduction and <20% blast in the specimen, the initial dose was continued up to day 10 in pts <30y, day 9 in pts 30-40y, day 6 in pts 40-50 and day 7 in pts aged 50-60y. In non respenders with lower cytoreduction, HD Ara-c 1.5 g/m2 inf. every 12 h was administered on days 7 to 10 in pls <30 y, and on days 7,8 in pts older than 30y. In addition Etoposide 65 mg/m2/d iv was given in M4/5 pts. Non respenders obtained a second cycle before evaluation. Patients roaching CR obtained two Mitoxantrone+ARA-C consolidation courses and thereafter were submitted to BMT, ABMT or maintenance treatment. Results. Both groups were well balanced with comparable prognostic factors distribution. The results were as follows: ICE DA 3+7 +/- HD 63% 61% +CR rate 44 -T- 13,/48 days 31 T- 6,/29.5 days Time to CR mean/median 1 in 46%, 2 in 54% 1 in 78%, 2 in 22% No of cycles 1 in 55*/o, 2 in 45% 1 in 93%, 2 in 7~ No of cycles to CR 12.4 -T-2 day, reed.13 14~ 4 day, reed 12 Nadir cytopenia 22 + 4, med. 22 days 26 9 5, med. 24 days : time to gran.>O.5 G/I 6:1: 8, meal. 2 (0-31) 8 -T-6, reed. 6 (0-26) subst.thrombocytes-units 2.3 ~ 1.6, med 2 (1-7) 3 9 2, med 2.5 (0.6-11) ; subst.erythrocytes-liters 7% 14% early mortality (
Stratification of Postremission Therapy in Adult Acute Myeloid Leukemia According to the Karyotype. Preliminary Results of a Multicenter Treatment Trial (AML HD93).

H. DOhner, g. Schlenk, J.Th. Fischer, F. del Valle, A+'Pezzutto, M. Pfreundschuh, W. Weber, A. Distelrath, M. Biirger, R. Haas, K. Fischer, W. Hunstein. Specific chromosome aberrations are one of the most important predictive factors for response rote and remission duration in acute leukemia and are now increasingly being used for selecting postremission therapy. In December 1993, we initiated a multicenter treatment txial for adult acute myeloid leukemia [AML] where the karyotype is used for stratifying intensive consolidation therapy. Cytogenetic analysis [including conventional G-banding and interphase cytogenetics using fluorescence in-situ hybridization] is performed in a central reference laboratory. Induction chemotherapy consists of two cycles of ICE lICE I: idarubicin 12 rag/m2 i.v. d 1,3,5; cytarabine 100 rag/m2 cont. i.v. d 1-7; etoposidc 100 rag/m2 i.v. d 1-3; ICE II: start on d 28, +G-CSF 5 ~tg/kg/d s.c. from d 9] followed by a first consolidation therapy with HAM [cytarabine 3 g/m2/ql2hr i.v. d 1-3; mitoxantroec 12 rag/m2 i.v. d 2+3, +G-CSF from d 5]. For second consolidation, patients <55 yrs in CR arc stratified into three treatment arms: 1. low-risk [t(8;21); inv(16)]: HAM; 2. intermediate-risk [normal karyotypc]: allogenetc BMT [patients with HLA-identical sibling] versus S-HAM [cytarabine 3 g/m2/q12hr i.v. d 1+2 and 8+9; mitoxantrone 12 rag/m2 d 3+4 and 10+11, +G-CSF from d 12]; 3. high-risk lull other chromosome aberrations]: allo/autoBMT in first CR. Autulugon~ I.~ne u,~'ow is harvested after first consolidation with HAM and pinged with adapted dose mafosfamid. Patients >55 yrs are treated according to the low-risk arm, patients in whom cytogenetic analysis is not successful in analogy to the intermediate-risk arm. At present, 42 patients [median age: 46, range 17-60] have been registered. Cytogenetic analysis was successful in 27133 (82%) analyzed patients. Remission rates in these 33 patients are as follows: low-risk In=2], CR 100%; intermediate-risk [n=ll], CR 82%, ED 18%; high-risk [n=14], CR 57%, RD 43%; no successful cytogenetic analysis In--6], CR 83%, RD 17%. The overall response rate was 70%. So far, all patients <-55yrs have received the allocated treatmenL including autoBMT in one pt. [+8,t(9;11)] and alloBMT in two patients [both with complex karyotype], in conclusion, these preliminary results confirm the predictive value of the haryotype for remission induction. A high nonlber of patients and a longer follow-up will be no~ssary to answer the questions whether intensive remission induction and intensive first consolidation chemotherapy followed by auto/allo BMT will increase disea~-frec survival in cytogeamtically defined high-risk AML patients, and whether alloBMT is superior to intensive consolidation with S-HAM in patients with normal haryotypo. Karlsruhe, O l d e n b u r g , Berlin-Bueh, H o m b u r g , Trier, Wiirzburg, Giegen, and Medizinischc Klinik und Poliklinik V, Universitiit Heidelberg, Hospitalstr. 3, 69115 Heidelberg

A 135 161

163

POOR RISK AML UNDERGOING FLAG REGIMEN: MDREXPRESSION, G-CSF PRIMINIG ACTIVITY AND CLINICAL RESPONSE. MC. Petti, MP. MarteUi, S. Tosti, L. De Felice, T. Valentini, A. Tafuri, MT. Petrncci, F. Mandelli. Hematology, Department of Human Biopathatagy, University "La Sapienza", Rome, ITALY. The use of G-CSF combined with fludarabine and .~ra-C (FLAG) has been recently investigated in acute myeloid leukemia (AML). The aim of our ongoing study was to evaluate the activity of this regimen in poor risk AML patients (pts) correlating clinical response with expression of multidrug resistance (MDR) and G-CSF priming ability. Therefore we have treated 12 pts, eleven poor risk AML (6 refractory to first line of treatment, 2 secondary to MDS, 3 third relapses) and one CMMoL using G-CSF 400 ug/sqm/daily prior to and during chemotherapy; in five cases G-CSF administration was extended until neutrophil recovery. Fludarabine was administered at dosage of 30 ug/sqm over 30 minutes days 1-5 + Ara-C 2gr/sqm over 4 h days 1-5. Detection of MDR expression was made using the flow cytometric Rhadamine-123 functional test performed in presence or absence of Cyclosporin A (CsA), used as MDR-reversing agent. Within each sample the percent decrease of mean flunrescence with and without CsA was compared. Proliferative changes induced by G-CSF were investigated by clonogenic assay and cell cycle studies (flow cytometric DNA/RNA) on peripheral blood (PB) and bone marrow (BM) aspirates. Samples were studied b'.fore growth factor administration and immediately before starting chemotherapy (after 24 h of G-CSF). The mean Rhd-Etthix value (Rhd-E) was = 17.2% (range 0-51.2%) showing in 10/12 cases a Rhd-E >5% efficiently blocked by the MDR-reversing agent. Among the six refractory cases five showed Rhd-E > 5% (range 6-51.2%). Cell cycle studies, performed in BM samples showed after 24 h of G-CSF administration a sig~fificant mean increase of S-phase cells (from 11% to 16.5%, p=.004). The clonogenic celtl growth at time O and after 24 h of in rive G-CSF showed only minor changes in standard culture conditions; whereas, at the same time points, a 5.02 fold increase in the spontaneous proliferation of cloaogeaic leukemic cells was detected. WBC mean value increased from 4,6 to 13,9 X 109/lt although no significant changes were seen in PI3 and BM blast percentage. When we analyzed individual samples heterogeneous prolifc.rative response was observed from case to case: in 7/12 pts was detected an increase in ctonogenic leukemic cell growtl~ paralleled by cell cycle recruitment induced by G-CSF. Eleven cases were evaluable for response. Among the 6 refractory pts five (83%) achieved complete remission (CR). In the other cases with secondary and relapsed AML the remission was achieved rispectively in 1/2 and 1/3 pts, and in the pt with CMMoL. No significant correlation was found between CR and proliferative response: four cases in whom G-CSF failed to increase proliferation went in CR whereas 3/7, that showed recmitultent into the cell cycle and increased significantly the spontaneous clonogenic proliferation, did not. In conclusion FLAG regimen seems to be effective in induce remission in poor risk AML pts expressing MDR. The role of growth factor priming by G-CSF in these pts seems no to be correlated with achievement of CR and will be further evaluated in the ongoing study.

GM-CSF DURING AND AFTER REMISSION INDUCTION TREATMENT FOR ELDERLY PATIENTS WITH ACUTE MYELO]I) LEUKEMIA (AML). F. Witz, J.L. Harousseau, J.Y. Cahn, J.F. Abgrall, J. Bri~re, A. Sadoun, D. Guyotat, P. Hurteloup, M. Mercier, P. Berthaud, on behalf of the GOELAM Group. Departments of Hematology, Nancy and Nantes, France. The remission induction treatment of AML in elderly patients remains unsatisfactory because of a high toxicity of aggressive regimens and of a high failure rate due to a high incidence of adverse prognostic factors. Hematopoietic growth factors have been used after induction chemotherapy in order to reduce the time to neutrophil recovery and the infectious mortality. Current results of randomised studies are controversial. They have also been tested in combination with chemotherapy in order to recruit leukemic cells into the cell cycle and to increase their sensivity to cytotoxic drugs. As of November 15, 1994, 240 patients (pts) aged 55 to 75 (median 68) were enrolled in a multicenter, double blind randomized study comparing placebo and GM-CSF (E.coli derived) (5gg/K/D 6 Hr infusion) during induction chemotherapy (Idarubicin 8 mg./m2/D IV D1 - D5 plus ARA-C 100mg/m2/D CI D1-D7) and until recovery of a neutrophil count > 0.5x109/L. Post remission therapy was identical in the 2 arms.The 2 groups were comparable regarding the following parameters : age, sex, performance status, fever, organomegaly, initial blood cell counts, FAB classification, presence of myelodysplastic features, incidence of caryotypic abnormalities. Of the 209 pts currently evaluable, 130 (62%) achieved CR with no significant difference between pts aged 55 to 64 (57/90 = 63%) and pts aged 65 to 75 (73/119 = 61%). There were 17 (8%) early deaths, 18 (9%) deaths in aplasia and 45 (21%) failures. No significant difference was observed between GM-CSF group (n = 103) and placebo group (n = 106) regarding the CR rate (63 % vs 61%), the incidence of early deaths (10 % vs 6%), of deaths in aplasia (8 % vs 9%), of failures (19 % vs 24%). The duration of neutropenia was shorter in the GM-CSF group (median 22 days vs 26 days p= 0.002). However the incidence of febrile episodes and of bacteriemias and the duration of hospitalization were similar in the 2 groups. With a median follow-up of 19 months, the overall survival of all eligible pts was similar in both groups (39% actuarial survival at 2 years in GM-CSF group vs 33% in placebo group, p = 0.45 log rank test). Nonetheless, the disease free survival was longer for pts who achieved CR in the GM-CSF group (44% continous CR rate at 2 years vs 19%, p = 0.024). The study medication was prematurely stopped for toxicity in 17 cases (14 GM-CSF, 3 placebo, p = 0.003). We conclude that 1) The induction treatment with Idarubicin and conventional doses of ARA-C obtains a high CR rate in elderly pts. 2) The administration of GM-CSF during and after induction chemotherapy results in a faster recovery of neutrophils but does not reduce infectious toxicity and does increase neither the efficacy of this chemotherapy nor the overall survival. However the disease free survival was significantly longer for pts who received GM-CSF before achieving CR.

162

164

P H A S E II T R I A L O F IDARUBICIN, FLUDARABINE, CYTIDINE ARABINOSIDE (ARA-C), AND FILGRASTIM (G-CSF) (IDA-FLAG) F O R T R E A T M E N T O F R E F R A C T O R Y , RELAPSED AND SECONDARY ACUTE M Y E L O I D L E U K E M I A (AML). H.T. Steinmetz, M. Schwonzen, A. Glasmacher 1, P. Staib, I. Katay, A. Neufaag, A. Schulz, V. Diehi a n d P.D. W i c k r a m a n a y a k e . First Dept. of Internal Medicine, University Hospital, Cologne, 1Bonn, FRG. Introduction and Methods: The combination of Fludarabine and Ara-C with a simultaneous infusion of G-CSF (FLAG) is an effective chemotherapy to induce remission in poor prognosis A M L (Estey Proc. ASCO 12:301 1993). W e started a phase 2 trial adding Idarubicin to FLAG in order to improve response rate and duration. Patients (pts.) refractory to or with first relapse after standard chemotherapy, pts. with secondary A M L and the history or signs of trilineage myelodysplasia at diagnosis were included. Ida-FLAG consisted of Idarubidn 8mg/m 2 dl,3,5, Flndarabine 25mg/m 2 1h-iv dl-5, Ara-C 1000mg/m 2 ql2h lh-iv dl-5 and Filgrastim 400gg/m 2 eontinous infusion dO until recovery of ANC > 1000/incl. Results: 31 pts. were evaluated after the first course of I d a - F L A G from 110/1994. 9 pts. (7 men/2 women) with refractory A M L had a median age of 48y (22-68), 8 pts. (3m/Sw) with A M L in first relapse of 563, (30-75), and 14 pts. (10m/4w) with secondary or ]t,H)S-AML of 593' (45-67). CR was achieved in 18 pts. (58%) overall, in 1/9 pts. with refractory AML, 7/8 with relapsed AML, 10/14 pts. with MDS-AML. 7 pts. died within 28 days from severe infection. 10/18 pts. with C R received a 2 ~. course of I d a - F L A G for consolidation. 3 of them died from sepsis, 6 are in continoos CR, 1 pt had a relapse. F r o m the 8 CR-pts. without consolidation therapy 3 h a d a relapse, 5 are in eontinoos CR. Toxicity was evaluated in 24 pts. surviving the first course of Ida-FLAG. The median (range) of the duration of G-CSF infusion was 24(! (17-66); of W B C <3000/#1 20.5d (11-44); of W B C <500/#1 16.5d (6-31); of ANC < 1000/#1 18d (11-33); of ANC <500/#! 17d (9-31); of platelets <50(O0/#1 24d (13-68), of days with fever > 3 8 . 0 ~ 4d (1-33). 7 of 10 pts. survived the 2 ~. course. In 4 of 7 the time till recovery of haematopoiesls was longer than in the first course, in 3/7 it was shorter. Conclusion: Ida-FLAG is a well tolerated chemotherapy and induces a high rate of complete remission in patients with relapsed or secondary AML. The rate of early death from sepsis seems to increase in the 2 ~d. course.

PROSPECTIVE COMPARISON OF ALLOGENIC BONE MARROW TRANSPLANTATION (ALLO-BMT), INTENSIVE CONSOLIDATION CHEMOTHERAPY (ICC) AND UNPURGED AUTOLOGOUS BONE MARROW TRANSPLANTATION (ABMT) AS POST REMISSION THERAPY IN ADULT ACUTE MYELOID LEUKEMIA (AML). J.L. Harousseau. B. Pianon. * F. Witz. * P. Linassier. * B. Licure. * P.Y. Le Prise. * J.Y. Cahn. B. Desablens. * D. (~eiIIot. * N. Ifrah. * J.F. Aberall. * F. Guilhot. * D. Guvotat. * A. Mors. * P. Casassus. * J. Bri~re. * V. Polin. * P. Berthaud. * Z. Tellier. * P. Hurtelou#. * Service d'H6matologie H6tel Dieu, 44045 Nantes 01, France. From November 1987 to April 1994, 522 adult (t 5-50 years) patients (pts) with de novo AML were included in the GOELAM 1 protocol comparing allo BMT, ICC and ABMT. For induction treatment, pts were randomized to receive a combination of Cyteaine-Ambiooside (ARA-C) (200 mg/m2/D cont. IV infusion D1 D7) and either idarubicine (IDR) (8 mg/m2/D IV D1 -D5) or Rubidazone (RBZ) (200 mg/m2/D D1-D4). After complete remission (CR) achievement, an allo BMT was proposed to pts up to the age of 40 with a HLA identical sibling. Other pts in CR had to receive a first course of ICC (ICC1) with high dose ARA-C (3 g/m2/q 12 H, 3 hr infusion, D1-D4, 8 doses) and either IDR (10 mg/m2/D IV D5-D6) or RBZ (200 mg/m2/D IV D5 - D6). Bone morrow was collected after ICC1 and cryopreserved without any in vitro manipulation. If the hematopoietic quality of the collected marrow was adequate, pts were then randomly assigned to receive a second course of ICC (ICC2) with m AMSA (150 mg/m2/D IV D1-D5) and VPt6 (100 mg/m2/D IV D1-D5) or an ABMT after a preparative regimen with Busulfan (4 mg/K/D 4 days) and Cyclophosphamide (50 mg/K/D 4 days). As of July 1, 1994, 490 lots were evaluable and 361 (74%) achieved CR with no significant difference between IDR and RBZ. An alia BMT was planned in 83 cases and was actually performed in 67. Out of the 278 other pts, 227 did receive ICC1. The median duration of neutropenia after ICC1 was 19 D and there were 9 toxic deaths (4%). 171 pts were randomized between ICC2 (84) and ABMT (87) and 128 have been currently analyzed (61 ICC2, 67 ABM'r). The reasons for exclusion were toxicity, refusal, poor hematologic reeanstitution post ICC1, relapse. With a median followup of 44 months, the overall survival of the entire cohort of pts is 37% at 6 years (median 22 months) with no difference between the 2 induction treatment arms. The 4 year relapse rate, event free survival and overall survival (of the pts in CR who actually received the assigned treatment) were respectively 30%, 48%, 56%, for allo BMT, 43%, 56%, 59% for ICC2 and 44%, 52%, 56% for ABMT. When considedng intention to treat there was no significant difference in event free survival or survival between alia BMT and other forms of post remission therapy or between ICC2 and ABMT. We conclude that 1) A significant improvement of survival can be obtained for pts with de novo AML up 50 years of age with 3 different modalities of intensive consolidation, 2) The 3 approaches give comparable results. After ICC1, ICC2 appears to be as effective as unpurged ABMT and is easier to perform, 3) New strategies are needed to reduce the exclusion rate.

A 136

165

167

LATE RELAPSES IN ADULT ACUTE MYELOGENOUS LEUKEMIA (AML): A RETROSPECTIVE STUDY OF CHARACTERISTICS AND OUTCOME IN 76 CASES. X.Thomas, L.Vila, J.Troney, D.Fiere, E.Arehimbaud. H6pital Edonard Herrint, Lyon, France.

IDARUBICIN OR MITOXANTRONE, VP-16 AND CYTARABINE FOR INDUCTION/CONSOLIDATION THERAPY FOLLOWED BY AUTOLOGOUS STEM CELL TRANSPLANTATION IN ELDERLY PATIENTS W I T H ACUTE MYELOID LEUKEMIA (AM[L): A FEASIBILITY STUDY. E. Archimband, U. Jehn, F. De Cataldo, C. Martin, X. Thomas. H6pital Edduard Herriot, Lyon, Klinikum Grosshadern, Miincben, Ospedale Maggiore Ca Granda, Milano, and Centre Hospitalier, Annecy. Since April, 1993, 74 patients (pts) of median age 70 years (range 61 to 83 years), with AML at diagnosis, 56 of them with d e novo AML and 18 with AML secondary to primary myelodysplastic syndrome or toxic exposure (sAML), received induction therapy with idarubicin (IDA), 8 rag/me/day IV on days 1, 3 and 5 or, on a randomized basis, mitoxantrone (MIT), 7 mg/m2/day IV on days 1, 3, and 5, both associated to VP16, 100 mg/m2/day IV on days 1 to 3 and araC, 100 mg/m2/day as a C[, on days 1 to 7. GCSF, 5 #g/kg/day, was administered after chemotherapy in pts aged more than 70 years. Pts in complete remission (CR) received one course of consolidation using the same schedule as for induction except for the shortening of alaC administration to 5 days. Pts aged less than 70 were then scheduled to undergo peripheral blood stem cell (PBSC) harvest on days 5 to 7 of GCSF, 5 #g/kg/day, initiated after hematopoietic recovery from consolidation. PBSC transplantation was performed after conditioning with BCNU, 800 mg/m2 IV on day -3, and followed by GCSF administration when PMN were < 0.1 x109/l. 71 pts are evahtable for indticdot~ and 39 (55%) acl'fieved CR, 3 of thcm aftcr 2 courses of induction. CR rate was 58% in d e novo AML and 44% in sAML (p > 0.05). There was no significant difference in CR rate between IDA (51% CR) and MIT (58 % CR), nor between pts aged - 70 years (58% CR) and > 70 years (53% CR). Among CR pts, median times to recovery of PMN > 0.5 x109/l and platelets > 50 x109/l were 24 days (13 to 56 days) and 24 days (17 to 50 days) respectively, without difference between IDA and MIT. Median time to PMN recovery was 22 days (18 to 40 days) in pts who received GCSF and 25 days (13 to 56 days) in pts who did not (p > 0.05). Severe toxicities of induction did not differ between the 2 arms and included sepsis (42 %), diarrhea ( 10 %), hyperbilirubinemia (7 %), hemorrhage (6 %) and vomiting ( 1% ). Overall 4 pts (6%), 2 in each arm, died from toxicity of induction. First consolidation is evaluable in 28 pts. Median time to recovery ofPMN > 0.5 x109/1 and platelets > 50 x109/1 were 22 days (16 to 39 days) and 20 days (13 to 39 + days) respectively. 3 pts (11%) died from toxicity of consolidation. At this time, 7 pts have received PBSC transplantation. This procedure was well tolerated, with a median duration of PMN < 0.5 x 109/1 of 7 days (0 to 10 days) and of platelets < 50 x109/1 of 11 days (3 to 21+ days) and no toxic death. We conclude that this regimen is well tolerated and has a good efficacy to induce CR. Intensive postinduction is feasible, including transplantation up to the age of 70 years.

Relapses in AML are generally considered of poor prognosis. Late relapses are known to represent a special entity with better prognosis. Whether late relapses occur from the original leukemic clone or not is still questioned. In order to bring clues, the characteristics and outcome of 76 patients with AML who experienced a first relapse after a first complete remission (CR) duration longer than 18 months (late-relapse AML) were analyzed. Median duration of first CR was 27 months (range: 18 - 118 months). A second CR was achieved in 44 of the 68 retreated patients (65%). Median duration of CR was 8 months (range: 1 -34 + months). Major differences, reflecting exclusively the tumor burden, were noted between presentations at initial diagnosis and after first relapse. Indeed 25 patients presented with tumoral syndrome at initial diagnosis versus 16 patients at first relapse (p = 0.1); mean percentages (+ SD) of blast ceils in peripheral blood and in bone marrow were respectively 42% + 36% and 77% + 17% at initial diagnosis, versus respectively 26% + 34% and 58% + 27% at first relapse (p = 0.0004 and p = 0.0001); mean values (+ SD) of LDH was 576 u/1 + 416 u/t at initial diagnosis versus 431 u/l • 415 u/1 at first relapse (p = 0.0001). Sixteen patients were classified initially as M1, 22 as M2, 12 as M3, 12 as M4, 12 as M5 and 2 as M6. At the time of first relapse, morphologie changes of blast cells caused assignment to another FAB subgroup in only 3 patients by comparison with diagnosis. Cytological dysbemopoiesis features, involving at least one lineage, were observed in 29% of the cases at diagnosis and 33% at relapse. Rapid first CR achievement (p = 0.01), young age at diagnosis (p = 0.0001), low WBC count (p = 0.01) and low percentage of peripheral blood blast ceils (p = 0.01) at relapse were associated with probability of response to second line chemotherapy. Only low WBC count at relapse appeared correlated with the length of second CR duration (p = 0.04). Metaphase chromosomes were analyzed both at time of diagnosis and at time of first relapse only in 15 cases. No difference were noted in karyotypie patterns except in 3 cases: one patient with normal karyotype at diagnosis presented at relapse with -Y, +8 and + 15; one presented at relapse with t(15;17) not detected at diagnosis; one with del(21) at diagnosis and t(1; 17) at relapse. Overall none of the features at time of relapse allow us to suggest a relationship between theses late relapses and the development of new lenkemias.

166

168

METHODS TO DETECT PROGNOSTIC FACTORS IN AML A. Heinecke, M. C. Sauerland, T. Bfachner, Dept. o f Med. Biostatistics, University o f Mfinster, Germany, for the A M L C G

HIGH DOSE CHEMOTHERAPY WITH BUSULFAN, CYCLOPHOSPHAMIDE AND VP 16 (BU, CY, VP) AS CONDITIONING FOR ALLOGENEIC AND AUTOLOGOUS BONE MARROW TRANSPLANTATION FOR PATIENTS WIiH ACUTE MYELOID LEUKEMIA A. R. Zander 1 , W. Kr0ger 1, M. Stockschi~der 1 , U. v. EisenhartRothe 1, W. Zeller 1, M. Hoffknecht 1, G. Kroschke 1 , H. Kabisch 1 , P. SchSnrock 2, R. Kuse 2, H. J. Weh 1, K. D. Hossfeld 1, Universit&tskrankenhaus Eppendorf, Hamburg, 2AK St. Georg, Abtl. f. H&m./Onkologie, Hamburg.

Risk adapted treatment in A M L requires a clear and practical pmgnostic subgroups.

definition o f

At present their are two biometrical methods to identify such subgroups: the method o f classification and regression trees (CART) and CoxRegression. Using C A R T the considered population is successively spirted into two subgroups which internally are most homogeneous and externally are most different. The results o f C A R T are convenieofly displayed in a d i a g r a m m containing the prognostic subgroups as terminal nodes o f a binary tree. Cox analysis is based on the assumption o f proportional hazard model. The results o f Cox analysis are formulated in a regression equation with the logarithme o f the hazard function as dependent variable. This is interpreted as a prognostic index by which patients are classified into prognostic subgroups. Both methods are compared analysing the data o f 1042 adult patients with A M L recruited between 1986 and 1992 and treated by intensive and prolonged chemotherapy according to AMLCG-protocol. Subject o f the analysis is disease flee survival (DFS) defined as time from complete remission (CR) to relapse o f A M L or death. Age, WBC, platelets, serum L D H , sex, FAB-subtype and karyotype categorized as favourable (t(8;21), inv(16), t(15;17)), unfavourable (-5/5q-, -7/7q-, abnormal 11q23, complex abnormalities), other karyotype, not evaluable, and not done are introduced as independent covariates. Both methods revealed age, serum L D H , and FABM3 as the most powerful prognostic factors. In addition: with C A R T favourable karyotype had good prognosis, with Cox-analysis unfavourable karyotype had poor prognosis.

51 patients (median age: 28 [9 - 54] years) received Busulfan 16 mg/kg, Cyclophosphamide 120 mg/kg and VP16 30 - 45 mg/kg body weight followed by autologous (16) cr allogeneic (37) bone marrow. Autologous marrow was purged with Mafosfamide. GVHD prophylaxis in aliogeneic transplants consisted of: Cyclosporin A (3 mg/kg) and Methotrexate or Prednisone. The median follow up is 26 (2 - 53) months. Results: 28/51 are alive in CR. 17/20 (85 %) of aIIogeneic patients and 4/5 of autologous patients transplanted in 1. CR are alive in CCR. Major causes of death for all groups were: GVHD: 5, Infections: 5, VOD: 2 and Relapse: 11. Conclusiens: BU, CY, VP 16 is well tolerated and effective in AML. The low mortality and low relapse rate in first CR justify autologous or allogeneic transplant in first CR independent of prognostic features.

A137 169 A~t_OGENEIC BMT FOR TREATMENT OF AML: PROGNOSTIC FACTORS FOR SURVIVAL R. Arnold, B. Hertenstein, D. Bunjes, M. Hafner, J. Novotny, M. Stefanic, M. Wiesneth, H. Heimpel Dept. Internal Medicine III and Transfusion Medicine, University of UIm, 89070 UIm, FRG Between 1980 and November 1994 126 patients with a median age of 35 (15-51) years underwent allogeneic brat for treatment of AM[_ 90/126 patients were transplanted in 1. CR, 18/126 patients in 2. CR and 18/126 patients in more advanced disease. Conditioning consisted mainly of Cyclophosphamide/TBI (n=83). Bone marrow donors were HLA identical siblings (n=t 13) or 1 mismatch family members (n=9). GvHD prophylaxis depended on disease status and changed with time (MTX, ex vivo T cell depletion, CSA/M'I'X and in viw~/ex vivo T cell depletion). For the whole group the most important factor for survival was status of disease at BMT (1. CR 0,56 vs 2, CR 0.?.6 vs > 2rid 0.0). Furthest analysis is therefore restricted to patients transplanted in 1. CR. FAB type, leucocyte count at diagnosis, age and primary or secondary AML are of no prognostic significance for survival after brat. Post brat the incidence of acute gvhd or chronic gvhd is of prognostic importance for survival. Patients with agvhd have a probability of survival of 0.24 vs 0.73 for patients without agvhd, for cgvhd probability of survival is 0.0 vs 0.67. Different methods for GvHD prophylaxis over time recuded the incidence and severity of agvhd and cgvhd and is the strongest prognostic factor for survival in our patient group. For the MTX group probability of survival is 0.44, for ex vivo T cell depletion 0.39, CSA/MTX 0.61 and for the in vivo/ex vivo T cell depletion 0.82. The same trend is seen for DFS. In conclusion: Prognostic factors for survival in patients transplanted for AML are status of disease at brat and in patients with AML, 1.CR GvHD.

171

D o w n ' s S y n d r o m e and Megakaryoblastic Leukemia u. creutzigl, J. Ritterl, J. Harbott2,Ch. Niemeyer3,B. Stollmann-Gibbels4 University Children's Hospital M(insterl, Giel~en2, Freiburg3, Essen4, FRG Recent reports estimate an approximately 20fold higher incidence of leukemia in children with Down's syndrome (DS) compared to non-DS children. There is also a marked increase (600x) of the megakaryoblastic subtype (FAB M7) (Zipursky 1994). Twenty of 34 children with DS enrolled in our studies since t987 were classified as FAB M7 or presented with a preleukemic phase (MDS) with megakaryoblasts. All patients were under 3 years (.6 - 2.7; median 1.8 yrs.); and showed generally a low leukocyte count (1.6 - 177; median 7.1 x t0~lmm 3) and platelet count (4.0 - 2t0; median 28.5 x 10Slmm=). Karyotype abnormalities other than +21c were found in 7/9 patients analyzed (Sx complex karyotypes, 2x +8). Only 7 children presented with >30% blasts in the bone marrow (BM). In 6 patients diagnosis was based solely on morphology, and in t4 patients confirmed by immunophenotyping. A preleukemic phase with thrombocytopenia for at least 2 months was recorded in 7 of the 12 patients presenting with a low blast count (<30%) and myelodysplasia in the BM. Three children had a history o f neonatal transient leukemia. Nine o f the 11 untreated or minimally treated children subsequently died (12 days to 6 months) due to progressing leukemia. In 2 of the untreated patients a spontaneous improvement or a stable disease (with retinoids) lasting for several months, was observed. Nine patients were treated. 3/9 had an endstage disease and received inadequate dosages of therapy and died early. 5/6 children treated according to the AML-BFM protocols including high-dose cytosine-arabinoside are in continuous complete remission for .8 - 4.6 years. - We conclude, that M7 leukemia in DS is often preceded by a preleukemic phase with thrombocytopenia. Patients treated according to the AML protocol showed good results, and further studies are warranted to clarify whether DS patients with a preleukemic phase o f M7 would benefit also from an early and adequate treatment.

170

172

YNK01, AN ORAL ARA-C DERIVATIVE IN AML AND CML P. Heugner 1, R. Willemze, A. Ganser, A. Hanauske, S. Amadori,Q. Heil, 1~. Schleyer, W. Hiddemarm, J. Selbach, M. Schiigler~, M. Freund~ (The European YNK'01 Study Group). 1Dept. of Hematology ar/,d Oncology, Hannover Medical School, 2ASTA Medica, Frankfurt and ~ of Hematology and Oncology, Medical School at the University of Rostock, Germany 27 patients have been treated in a phase I/II multicenter trial and a pilot monocenter trial with AML and CML with YNK01, an oral ara-C derivative. In contrast to ara-C YNK01 is resistant to cytidine deaminases. Therefore YNK01 is converted to ara-C in the liver and released into blood slowly. It has been shown in ongoing pharmacoldnetic studies that a mean of 16% of YNK01 is secreted as ara-U into the urine. 21 patients with AML (12 pts. with relapse, 5 pts. with secondary AML, 4 pts. primary not qualifying for intensive chemotherapy) were included (median age 64 (range 22-79) yrs., 13 pretreated/11 with ara-C). In the AML trial the doses of YNK01 were escalated intedndividually from daily 100 mg per body up to 1200 mg per body for 14 days. Cycles were repeated every 21 to 28 days. Major toxicities at the 900 and 1200 mg dose levels have been nausea grade 3 (WHO) in 1 pt., diarrhea grade 3 in 5 pts., grade 4 in 1 pt., exanthema grade 3 in 1 pt., and stomatitis grade 3 in 1 pt., grade 4 in 1 pt.. At the lower dose levels no grade 3 or 4 organ toxicities have been observed. Six pts. (median age 50 (26-56) yrs., no pretreatment) have been included in the CML pilot trial. Treatment was started with IFN ct-2b 5 MU SC daily. After one week YNK01 600 mg daily continuously was added. IFN and YNK01 were modified according to toxicity and effectivity. Maximum toxicities were diarrhea grade 3 in 1 pt. and bone pain grade 3 in 1 patient. Results: In AML-patients CR has been observed in 2/21 pts., PR in 1/21 pt. and stable disease for up to 7 months in 3/21 patients. In CML 1/6 pts. achieved a CHR after 1 month of continous treatment and a PCR after 2 months, 2/6 pts. are in CHR after each 2 months of treatment and 2/6 are in PHR after 8 weeks. We conclude that YNK01 has a mild toxicity profile in patients with hematological malignancies. Diarrhea seems to become the dose limiting toxicity. The maximum tolerable dose of YNK01 seems to be reached at the 1200 mg dose-level in AML. Phase If-studies will be performed to further evaluate the effectivity of the drug in AML patients as a maintenance treatment and in CML following the pilot trial.

Chronic Myelomonocytic Leukemia (CMML) in Childhood: A Retrospective Analysis of 110 Patients C. Niemeyer, M. Aric6, A. Biondi, G. Basso, A. Cantt~ Rajnoldi, U. Creutzig, O. Haas, J. Harbott, H. Hasle, G. Kemdrup, G. Mann, B. Stollmann-Gibbels, E. v a n ' t Veer, E. van Wering, M. Zimmermann and Members of the European Working-Group on Myelodysplastic Syndromes in Childhood. C M M L is a rare malignant hematopoietic disorder in childhood. Alternate terms for the disease include juvenile chronic myelogenous leukemia and subacute myelomonocytic leukemia. To define the clinical course we performed a retrospective analysis on 110 children with CMML diagnosed between 1975 and 1993 in Austria, Denmark, Italy, Germany and the Netherlands. The median age at the time of diagnosis was 1.9 years. There was a male predominance (68%). Neurofibromatosis Type 1 was known in 13 patients (pts) (12%). Enlargement of spleen and liver of more than 5 cm below the costal margin was noted in 54% and 33% of the pts, respectively. 76% of the pts had lymphadenopathy, 55% pulmonary symptoms at presentation. Karyotypic abnormalities in hematopoietic cells were noted in 30/95 pts (32%) with successful chromosome studies reported. 17 pts had monosomy 7 only. At presentation the median white blood count was 32,982/gL (max. 259,380/pL), the median hemoglobin 9.2 g/100mL and the median platelet count was 44.000//iL. There were no significant differences in the median values for age, liver or spleen size at presentation for pts with monosomy 7 as compared to those with a normal karyotype. However, hemoglobin F was greatly elevated in patients with normal karyotype (median 28%), but normal or only slightly increased in patients with monosomy 7 (median 1.0%). 23 pts (23%) received an alIogeneic bone marrow transplant. The survival at 8 years was 0.54 for the transplanted pts and 0.08 for the non-transplanted pts. An analysis of prognostic factors for duration of survival will be presented. Univ.-Kinderklinik Freiburg, Mathilenstr. 1, D-79106 Freiburg

A138 173

175 Childhood MDS treated with A M L regimens

Henrik Hasle and Gitte Kerndrup, Department of Pediatrics and Pathology, Odense University Hospital, 5000 Odense C, Denmark Myelodysplastic syndromes (MDS) in children are often considered as variants of AML and are frequently treated as such. However, there are very few reported data of the outcome after AML treatment in children with MDS. During the period 1984 - 1991 20 children with MDS were treated according to the Nordic AML protocol in Denmark. All children had de novo MDS. The mean age at presentation was 4.1 years. The mean interval from first admission to the first course of AML treatment was 155 days (range 2 - 961). At presentation the children were classified as: 4 RA, 4 RAEB, 5 RAEB-T, 7 CMML and when the AML treatment was initiated as: 4 RAEB, 4 RAEB-T, 7 CMML, 5 had progressed to AML. Only three children attained a complete remission following the first induction course, a further four children attained remission after the second induction. Eight died during the cytopenia that followed the induction therapy (seven from infection and one from hemorrhage). Two of the children were later treated with allogeneic BMT, both died of procedure-related toxicity. Only two of the 20 children are still alive; an infant showed a spontaneous remission several months after the end of all therapy, the only child who entered remission on therapy and is still alive presented with RAEB and chloroma. During the same period 8 children with MDS were treated with allogeneic BMT without previous chemotherapy, five of them are still alive. Conventional AML regimens in childhood MDS is associated with a low rate of complete remission, a high risk of death in cytopenia, and have a very limited curative potential.

SECONDARY MYELODYSPLASTIC SYNDROMES (sMDS): M O R P H O L O G I C A L FINDINGS, CYTOGENETICS AND CLINICAL COURSE IN 55 PATIENTS H. MINNING, C. AUL, V. RUNDE, U. GERMING, W. SCHNEIDER. Department of Internal Medicine, Hematology-and Oncology, HeinrichHeine-University, Diisseldorf, Germany Between 1974 and 1994, 55 patients with MDS following previous antineoplastic or immunsuppressive treatment were identified at the University of Diisseldorf. In a retrospective study, the clinical and hematological features of these patients were examined and compared with those in 735 patients with primary MDS, diagnosed during the same time period. Morphological subtypes of patients with sMDS were RA in 21%, RARS in 21%, RAEB in 26%, RAEBfF in 19% and CMML in 13%. On cytogenetic analysis, 9 of 11 patients had clonal chromosomal abnormalities including monosomy 7, monosomy 5 and multiple aberrations. The most frequent primary diseases were multiple myeloma in 21%, malignant lymphoma in 20% and breast cancer in 19% of sMDS patients. 25 patients had been treated with chemotherapy alone, with cyclophoshamide, melphatan or procarbazinc being the most frequent administered agents. 12 patients had been treated with high-dose radiotherapy alone and 12 patients with combined radiochemotherapy. 5 patients with sMDS had been treated with immunsuppressive therapy because of nonneoplastic diseases such as rheumatoid arthritis, chronic polymyositis or kidney transplantation. The median latency for development of MDS was 41 months (range 12-174 months) following combined radiochemotherapy, compared with 60 months (17-369) after chemotherapy and 132 months (71-528) after ionizing radiation. Prognosis of patients with secondary MDS was very poor. Their actuarial median survival was 9 months, as compared with 23 months for patients with primary MDS. 11 patients with sMDS died after transformation to AML, whereas 21% and 22% of patients succumbed to infectious or hemorrhagic complications, respectively, without leukemic transformation.

174

176

Remission rates, survival and prognostic factors in 94 patients with advanced MDS treated with intensive chemotherapy C. AUL, V. RUNDE,U. GERMING, M. BURK, A. HEYLL, B. HILDEBRANDT,R. WlLLERS Department of Internal Medicine, Hematology and Oncology Division, Heinrich-Heine-University, Dfisseldorf, Germany Previous studies have suggested that intensive chemotherapy in MDS is associated with low rates of complete remissions and considerable toxicity. These assumptions, however, are based on small patient series. We retrospectively analyzed the outcome of 94 patients (median age 54 yrs, range 16-72 yrs) with advanced primary MDS who were treated with intensive antileukemic regimens at our hospital between 1979 and 1994. The median time from diagnosis of MDS to initiation of chemotherapy was 2 months (range, 0-41 mo). FAB types at treatment begin were RAEB in 11, CMML in 4, and RAEB/T in 38 cases; 41 patients had already transformed to AML. The Kamofsky score ranged from 30 to 100% (median, 80%). Chromosome studies were performed in 67 patients of whom 27 (40%) presented with an abnormal karyotype. 64 patients received TAD, 9 patients idarubicin + Ara-C, 7 patients idarubicin, Ara-C + etoposide, 9 patients a double-induction regimen, and 5 patients other protocols for remission induction. 57 patients (61%) entered CR, and 10 patients (11%) had a partial response. Early death occurred in 13 cases (14%), and 14 patients (15%) had refractory disease. No unusual toxicities of chemotherapy were noted. The median duration of bone marrow aplasia (leukocytes < lx109/1 and/or platelets < 20x109/1) for patients achieving CR after TAD was 17 days. After entering CR, 46 patients received I to 2 cycles of consolidation chemotherapy and 18 patients monthly myelosuppressive maintenance chemotherapy over a maximum period of 3 years. Median follow-up of the patients is now 16 months. Median overall survival after the start of chemotherapy is 13 months. Disease-free survival estimated by life-table analysis is 34% at 2 years and 20% at 5 years. Using multivariate analysis, the following parameters were found to be significantly correlated with successful remission induction therapy: I. medullary blast cell count < 30% (p=0.02) and 2. female sex (p--0.04). Presence of pseudo-Pelger cells was the only factor significantly associated with the duration of DFS (p---0.01). From these date we conclude that intensive AML-type chemotherapy can be successfully administered to patients with high-risk MDS. In selected patients, this approach offers a chance of long-term remission and potential cure.

IV~YELODYSPLASTIC SYNDROMES (MDS) OR LEUKEMIA FOLLOWING MDS (sAML) TREATED WITH ALLOGENEIC BONE MARROW TRANSPLANTATION (BMT): A SURVEY OF THE WORKING PARTY ON CHRONIC LEUKEMIA OF THE EBMTG. V. RUNDE, T. DE WITTE, C. AUL, A. GRATWOHL, D. NIEDERWIESER, A. VAN BIEZEN, J. HERMANS, J. VERNANT, H. KOLB, J. VOSSEN, B. LONQVIST, D. BEELEN, A. FERRANT, R. ARNOLD, J. CAHN, M. VAN LINT, L. VERDONK, AND J. APPERLEY. Heinrich Heine University, Moorenstr. 5, D-40225 Diisseldorf, Germany. Altogeneic BMT offers potentially curative treatment for younger patients with MDS or sAML when a histocompatible sibling is available as a marrow donor. Until now approximately 450 patients from more than 45 centers have been reported to the EBMTG. We retrospectively analyzed those patients who received BMT without prior remission induction chemotherapy (n=107) as well as patients transplanted in first complete remission (CR) after intensive remission induction treatment. Among patients undergoing BMT without prior remission induction chemotherapy (median age 31 years, range 2-58) 9t marrow donors were HLA-identical siblings. Five years after transplant, patients with RA/RARS, RAEB, RAEB/T, or sAML had a disease-free survival (DFS) of 46%, 35%, 27%, and 0%, respectively. Multivariate analysis showed that bone marrow blasts > 30%, age > 40 years, and an interval > 3 months between diagnosis of MDS and BMT were negatively correlated with overall survival. BMT as first line treatment can, therefore, only be recommended for low risk MDS patients early in the course of their disease. In contrast, most european centers will only consider BMT for advanced stages of MDS after remission induction chemotherapy. In 64 patients with advanced MDS or sAML transplanted in fLrst CR following intensive remission induction therapy, the 5-year DFS was 38%. Conclusion: Allogeneic BMT produces long term DFS in a considerable proportion of patients with MDS or sAML.

A 139 177

179

Cyclosporlne A In the Treatment of Myeledysplastic Syndrome V.G. Savcinmko,E.A. Mihai]ova,E.N. Parovitchnikova

COMPARISON OF PCR- AND CULTURE-BASED PRE-EMPTIVE THERAPY WITH GANCICLOVIR FOR PREVENTION OF CMV DISEASE AFTER BMT. H.Einsele, G.Ehninger, N.Hebart, U.Schuler, P.Mackes, M.Hertter, Th. Klingebie !, J.L6ffler, S.Wagner, H.D.Waller, C.A.M~Iler Medizinische Klinik und Poliklinik, Abt. II, Sektion f&r Transplantationsimmunologie und Immunhfimatologie, Kinderklinik, Tfibingen Culture-based pre-emptive therapy with ganciclovir was shown to reduce the incidence of C M V d i s e a s e after BMT. But in these studies culture techniques failed to detect CMV in 15423% of patients prior to the onset of CMV disease. Thus, in a prospective study, 66 consecutive patients were randomized to either receive pre-emptive therapy based on culture assays or on the much more sensitive PCR technique. Therapy was continued in both groups until clinical signs disappeared and PCR-negativity was documented. Patients were assessed for the incidence of CMV disease and mortality until day +i00 after BMT. 34 patients received PCR-based pre-emptive therapy (group i), 32 a culture-based therapeutic intervention (group 2). In group 1 22 PCR-positive, in group 2 16 culture-positive patients received antiviral therapy. PCR allowed detection of the virus (median day + 32 compared to day + 47) and introduction of antiviral therapy earlier than in the culture group (median day + 44 compared to day +54). In i0 patients in the group 1 CMV excretion was documented by culture technique coincident or even after the introduction of the PCR-based pre-emptive therapy. The incidence of CMV disease (2/34 vs 9/32, p= 0.005) and CMV-associated mortality (0/34 vs 5/32, p= 0.03) was signficantly lower in group i. The duration of antiviral therapy of the individual treated patient (3.2 weeks vs 4.8 weeks) was significantly shorter and the incidence of severe neutropenia, of severe bacterial or fungal infections (1/34 vs 8/32, p<0.01) lower in the PCR-monitored group. Overall survival until day +100 was significantly better for patients in group 1 (94.1%) compared to the ones in group 2 (71.9%). Thus, pre-emptive therapy based on more sensitive detection methods like the PCR assay reduces the incidence of CMV disease. PCR-monitoring additionally allows to reduce the duration and side effects of antiviral therapy.

Myelodysplastiesyndrome (MDS) is a r disorder of m~tipotential ~ n cell comprising a variety of different clinical forms with commonmorphoingiealfeatures of dysmyelopoiesisand peripheral blood (PB) cytopenia. Though a lot of different therapeutical approaches have been tested in RA, RAEB, RAEBt, no exact well-dofmedtzzatm~t protocols m'e established so far. Differential,cytostatlc,cytoldnr therapyla'ovidcssome t~,mlly not gable effects in few patlents~ Long term remissions were reported after aggressive chemothcraW followed by BMT. In most cases the tres~ent is directed to the improvementof PB cytopeninand decrease of transfusion r c q ~ t s . Herewe ~ theresultsof a smallstucb' in MDS patients who we~ mintedwith CyclosporinA (CsA). Five patientswith MDS were included in the study. 2 patients were treated with CsA as the first line aeatsaent, 3 patients (N 1, 2, 4) w~rcpreRmte~(predaisene, low-4o~ ARA-C,low-dine IFN 0t2b).AR of them (exceptN3) were tiensfusiondependentand had o median diseose comsr of one year. CsA was administered orally at an initial dose ][0mg/ks/d constantly with modifications according to tolefabilityand ereatinin level. Final CsA dose used in the patients was 4-5mg/kg/& Resultsof the ~ & are reflected in the table. Patients Diagnosis Follow-up Hb Pit Transfusions Freedom from No. on CsA elevation elevation independence progression r~. N l,'3(b/,m RA 16 months yes yes yes yes N 2~65},,m N 3,.42y,m N4, Igy,m N 5, 40yr m

RA RA ,RAEB

5 months 2 months

yes no

yes no

yes --

yes yes

8months yes yes yes yes RAEBt 1 month** no no , no no yes 9* 1reamerw~ gtoppeddueto progremioninthep~tle~twithMDS~elap*eal~ Amolog~ BMTpeffmmedin tst remlm'onofAML We concludethat in some MDS patients CsA can imprr sovm'cPB cytopcniawith tnmsinsions dependence, can stop progremivodecrease of PB ceils. CU%effects can be rather prolonged but usuallya constantuse of the drog is necessa~. No severe side effeotswere noted during Irea~ent Mechanismof CsA aetiou is believed to be connectedwith the apoptoticcell death inhibition, as apoptosisin MDS is essentially imre.as~ and is one of the reasons for PB eytopenia.Igs seems likely llmt in addison to stimulating proliferation and differentiation activity the inhibition of apoptosisis one of the mechanismsof CSI~ efflc~y in MDS. The, r use of CsA and growthfactorscan be recomendedas a new trial design. National Research Center for Hematology, Novezykowkypr. 4a, 125167, Moscow, Russia

178

180

D~TECTION OF VARIOUS FUNGAL PATHOGENS IN BLOOD SAMPLES BY PCR. H~Einsele, H.Hebart, U.Schumacher, G.Roller, G.Ehninger, C.A.Mfiller, Med. Universitatsklinik TObingen and Med. Mikrobiologie der Universitfit T0bingen, 72076 T~bingen, Germany Early diagnosis of fungal infections in neutropenic patients is essential to improve the outcome in this patient cohort. Detection of a variety of antibodies, antigens and metabolites has been extensively studied in neutropenic patients with limited success. A PCR assay amplifying a 486bp DNA segment out of the 18ssurRNA coding region was established for detection of a variety of fungal pathogens including all clinically important Candida and Aspergillus species and for further species differentiation by additional hybridization with species-specific oligonucleotides. A high sensitivity of the assay of 1 CFU/ml blood as well as a good specificity could be demonstrated. 312 b l o o d samples obtained from 63 febrile neutropenic patients were screened blindly for the presence of fungal DNA. Clinical data as well as additional microbiological parameters were collected from these patients. Overall 34/312 blood[ samples revealed positive PCR-results. 20 of them were found to be PCR positive for C.albicans (8 patients), 3 for C.glabrata (i patient), 2 for C.krusei (i patient) and i sample positive for C.tropicalis. 22 samples were obtained from 2 patients with either pos. blood culture (I0) or histopathologically proven candidiasis (6) or responding to amphotericin B (6), 4 samples were from febrile neutropenic patients without additional evidence of invasive fungal disease. 8 samples obtained from 3 patients with histopathologically proven invasive aspergillosis were found to be PCR positive. The other 278 blood samples were PCR-negative. None of PCR-negative patients was found to develop invasive fungal disease. Additionally 6 patients with cultureproven fungaemia were monitored for the presence of fungal DNA during antifungal therapy. The PCR assay revealed the presence of fungal DNA in blood samples for a much longer period of time during therapy. Thus, the PCR assay might allow erlier detection and monitoring of the therapy of invasive fungal infection with different fungal pathogens.

THERAPY OF SEPSIS SYNDROME WITH H U M A N POLYCLONAL IGM-ENRICHED IMMUNOGLOBULINS: INTERIM ANALYSIS OF A R A N D O M I Z E D TRIAL IN NEUTROPENIC C A N C E R PATIENTS

G. Behre (1), H. Ostermann (2), M. Rothenburger (2), X. Schiel (1), D. Boekelmarm (3), S. Geiger (1), M. Dedrungh (1), M. Helmerking (3), B. WOrmann (1), J. Kienast (2), and W. Hiddemaun (1) (1) Dept. of Hematology/Oncology, University of Gotlingen~ Robert-Kcch-Str. 40, 37075 G6tfingen; (2) Dept. of Hematology/Oncology, University of Mfiuster, Albert-Schweitzer-S~r. 33, 48161 Mfiuster; (3) Clinical Research Orga,i~ation, Albert-Schweitzer-Str. 62/1, 81735 Mfinchen; Germany Backm'ound. Sepsis is a major cause of death in neu~openic cancer patients. We have shown the prognostic s i ~ c a n c e of endotoxin plasma levels in neuttopanic patients with severe sepsis, and other groups have suggested the efficacy of polyclonal IgM-enriched imrmmoglobulius in the treammnt of endotoxin-posifive sepsis in non-nenU'openic patients. Meth94~. To evaluate the efficacy of polyclonal IgM-enficbed immonoglobulins in the tream~nt of sepsis syndrome in neuUopanic patients with hematologic malignancies, a randomized, placebo-controlled trial was initiated at oor institutions. The patients were randomly assigned to either the b . m a n polyclonal itmnunoglobulin preparation every 6 h for 3 days (total dose, 1.3 liter with 49.4 g IgG, 7.8 g IgM, 7.8 g IgA) or the equivalent dose of 5% human albumin. Results. In the 52 entered patients with sepsis syndrome followed to death or day 28, there were 10 deaths among the 22 recipients of albumin (46%) and 9 deaths among the 30 recipients of immunoglobulins (30~ P > 0.05). In the subgroup of patients with sepsis syndrome, but without shock at entry the mortafity was 11% (1/9) in the albumin and 12% (2/17) in the immunoglobulin group. However, the immunoglobulin administration reduced mortality from 69~ (9/13) to 54% (7/13) in the subgroup of patients with shock at entry and from 89~ to 67% in the subgroup o f patients with organ failure at entry (acute lung, renal or hepatic failure or disseminated intravascular coagulation). Conclusions, This interim analysis suggests that b , m a n polyelonal lgM.cm'iohed immanoglobulins may reduce the mortality from sepsis syndrome in nentropenic cancer patients. A completation of the ongoing trial is needed to substantiate these findings taking into account that randomization has to be stratified for sepsis syndrome with and without septic shock at entry because of the different risk of death.

A 140 181

183

THE 10/NL TRIGGER FOR PROPHYLACTIC PLATELET TRANSFUSION IN AML: A PROSPECTIVE COMPARATIVE MULTICENTER STUDY H. Wan&, M. Frank, H. Link, C. Schneider,. N. Brack, A. Daoud, G. Ehninger, E. Fackler-Schwalbe, J. Fischer, Gackle, T. G-eer, M. Gramatzki, P. Harms, B. Ltffler, S. Ohl, B. Otremba, M Raab, P. SchOnroek-Nabulsi, Winter. Cooperative AML-Study of the Soddeutsehe H~oblastosegruppe. Supported by the Deutsche Krebshilfe

QUALITY OF LIFE IN PATIENTS WITH ACUTE MYELOID LEUKEMIA, A. Schumacher, Th. Kessler, Th. Biichner, J. van de Loo. University of MUnster, Dept. Internal Medicine A, 48129 Mtinster, FRG. In oncology, quality of life (QL) has become an essential criterion for evaluating the effects of therapy in clinical trials. So far, in adult patients suffering from acute leukemia assessmentof QL has not been considered possible due to poor prognosis and the acute course of the disease. Advances in chemotherapy significantly prolong survival of patients with acute myeloid leukemia (AML) and may cure a significant percentage (25-30%) of patients. Therefore, the assessment of QL of patients undergoing chemotherapy is of growing interest. This study was designed to evaluate QL in patients with AML treated according to the protocol of the German AML-Cooperative Group (MUnster, FRG). The dynamic concept of QL can change as an adaptational process interacting with life circumstances. The assessmentof QL is usually achieved by an evaluation of specific components contributing to the QL-construct. Based on conceptual, methodological and practical eriteria, the EORTC-QLQ C 30 questionnaire was used. Patients" individual perception of their disease and therapy was evaluated by a semi-structured interview. QL will be analysed during induction therapy and maintenance therapy over 3 years, evaluating defined specific parameters at 12 different time points. The questionnaire is administered in the first 4 courses of chemotherapy (double induction, consolidation, first cycle of maintenance therapy) during in-patient treatment, before and after drug-induced myeloid aplasia. Every six months during outpatient treatment patients assess a set of mailed questionnaires up to the end of their treatment. Currently, 74 patients are enrolled in the protocol. Those patients having completed the course of inpatient treatment are evaluated for changes in the conceptually distinct QL domains : Physical functioning (p=,000) and Emotional functioning (p=.001) improve significantly from beginning of chemotherapy to the end of inpatient treatment, Individual assessmentof Global health status and Subjective QL (p=.000) improve significantly over the same time. Patients suffer significantly less from fatigue, nausea/emesis and loss of appetite (p=.000). The content analysis of the interviews shows to what extent aspects of the inpatient setting influence patients" QL. As clarified by the qualitative data, a substantial number of patients experience a secondary benefit from their disease. Although only a minority of patients with AML remain in continous complete remission, the evaluation of QL in patients undergoing treatment shows that subjective benefit outweighs the adverse effects of antileukemic therapy.

From 04/92 to 07/93 we studied the safety and economy of the routine morning 10/hi trigger for prophylactic platelet transfusion in AML-pts. (FAB-M 3 excluded) compared to the traditional 20/rd trigger. Exclusion criteria for routine platdet transfusion below this 10/hi trigger were as follows: DIC or other coagulation disorders, fever >38,5 o c and rapid decrease of platelet count, major bleeding (WHO-grade >2), any kind o f biopsy, visus deterioration by retinal bleeding, use of drugs affecting platelet function. Eight centers of our AML-study group accepted this experimental ~_._~',~fion protoeot ( ~--~p_ ~ Aj , ~ O-e~tex~ were ~c!acta~ and used the traditional 20/nl trigger (group B). In both groups random, singledonor and multiple-donor platelets were transfused with leukocyte falters. We examined prospectively 2198 days o f tln'ombopenia (<25/nl) during 110 treatment cycles in group A and 1645 days in 106 cycles in group B respectively. Mean age o f all pts. was 47 years (17-73 y). The groups are comparable for age, sex and FAB-classiticatiou and treatment intensity. The median time ofthrombopenia was 18 days for group A vs. 13 days for group B. Bleeding WHO-grade 0/1 was 82 % vs. 83 %, grade 2 was 18 % vs. 9 % and grade 3 and 4 was 0 % vs. 8 %. The 9 pts. with grade 3 and 4 bleeding were all in group B. 5 o f these pts. had a platelet count >30/hi at the time of complication. 2 pts. died during thrombopenia but none o f fatal bleeding. The reduction of platelet transfusions in group A compared to group B was 38 % despite the longer median time o f thrombocytopenia_ Our results show that the low 10/nl trigger for prophylactic platelet support according to our protocol is safe and cost effective.

182 TRANSPLANTATION OF ALLOGENEIC PERIPHERAL BLOOD PROGENITOR CELLS ALONE OR IN ADDITION TO BONE MARROW. H Link, L Arseniev, O B~ihre, RJ Berenson #, JG Kadar, K Battmer, R Jacobs, J Casper, H Diedrich, J Schubert, J K0hl, H Poliwoda. BMT Unit, Medical School Hannover, Germany; #CelIPro Inc., Bothell, WA, USA. The aim of the study is the development of a protocol for allogeneic transplantation of PBPC. Three patient (pt) groups were studied: Gr. I received unmanipulated BM (4.5x106 CD34 + and 172.3 xl05 CD3 + cells/kg) and stored immunosetected (IS) PBPC (3.3x106 CD34 § and 3,7x105 CD3 + cells/kg), Gr. II - both IS BM (1.6x106 CD34* and O.6x10 s CD3 + cells/kg) and stored IS PBPC (2.3x10 ~ CD34+ and 2.7x10 s CD3 + cells/kg), and Gr. III - only fresh IS PBPC (6.2x10 s CD34 + and 9.4x105 CD3 § cell/kg). G-CSF-primed (2x5 pg/kg) PBPC were collected at days 4 and 5 (Gr. I & II) or at days 4-7 (Gr. III) of mobilization. BM was harvested at day O. CD34 + cells were selected from pooled PBPC concentrates or BM-harvests by immunoadsorption. A historical control group transplanted with a mean of 5.2x10 ~ CD34 + cells/kg and 156x10 s CD3 § cells/kg from BM alone was assembled for comparison. Pts were conditioned with either busulfan (16 mg/kg) or 12 Gy TBI fotlowed by t 2 0 mg/kg CY. CsA and short MTX were used for GvHD prophylaxis. Group III patients only received CsA. After transplantation, 10ts received G-CSF (5,ug/kg/d, iv) and erythropoietin (150 U/kg/d, civil Median recovery times (days) of neutrophils (Neu), platelets (Pit) and reticulocytes (Ret) were: Groups (n) I (5) II (5) III (3) Control (12) Neu > 100//]1 12 14 9 17 Neu > 500//11 15 15 10 18 Neu> 1,000//11 17 18 11 20 Pit> 50,O00/pl 24 35 15 41 Re~ 10,000//11 14 16 13 20 The rates of acute GvHD in groups I-III were similar to control patients. We conclude, that hematopoietic recovery is accelerated, if immunoselected PBPC are used for allogeneic transplantation.