Prognostic factors in acute myeloid leukaemia

Prognostic factors in acute myeloid leukaemia

Best Practice & Research Clinical Haematology Vol. 14, No. 1, pp. 65±75, 2001 doi:10.1053/beha.2000.0116, available online at http://www.idealibrary...

129KB Sizes 0 Downloads 77 Views

Best Practice & Research Clinical Haematology Vol. 14, No. 1, pp. 65±75, 2001

doi:10.1053/beha.2000.0116, available online at http://www.idealibrary.com on

4 Prognostic factors in acute myeloid leukaemia Bob LoÈwenberg

MD, PhD

Professor of Haematology Erasmus University and University Hospital Rotterdam, P.O. Box 2040, 3000 CA Rotterdam, The Netherlands

The prognosis between subgroups of patients with acute myeloid leukaemia (AML) may range considerably. Haematological, genetic and clinical analysis has provided possibilities for de®ning the heterogeneity of prognosis. This furnishes clinically relevant insights into the probability of treatment response and survival in individual cases of AML and it provides a lead in treatment management. Key words: prognosis; prognostic factors; acute myeloid leukaemia; cytogenetics; ¯t-3 receptor; ¯t-3 mutations; molecular genetics.

INTRODUCTION The group of diseases collectively designated as acute myeloid leukaemia (AML) represents a broad variety of distinct entities. AML arises from a malignantly transformed haematopoietic stem cell or progenitor cell. The cell stage (more or less primitive, multipotent or committed) at which transformation occurs and the residual capabilities of maturation of these cells, may contribute to the phenotypic diversity of this disease.1±3 These variations may be re¯ected in the di€erent cytomorphology of the leukaemic blasts, the dysplasia of the other haematopoietic cell lineages and the characteristic sites of presentation of leukaemia (e.g. skin, gingiva etc). The molecular abnormalities leading to full leukaemic transformation vary greatly. Di€erences in the molecular pathogenesis may also dictate variation in clinical presentation as well as variation of response to therapy. The understanding that the term AML lumps together a variety of distinct pathogenic conditions, has led to a general appreciation of the importance of delineating AML disease entities more sharply, e.g. AML with t(8;21) or AML1/eight-twenty-one (ETO) fusion gene etc. Each of the genetically de®ned AML diseases carries a typical prognosis and one may anticipate that in the long run these conditions will require speci®c therapy. Acute promyelocytic leukaemia with t(15;17) and the promyelocytic leukaemia gene/retinoic acid receptor alpha gene (PML/RAR-a) fusion gene, treated with all-trans retinoic acid (ATRA) is a classic example of this approach.4 Meanwhile a thorough understanding of the individual prognostic factors has become of signi®cant importance for treatment choice in today's clinical practice. Should the patient be treated with an allograft from an HLA-identical sibling or should 1521±6926/01/010065‡11 $35.00/00

c 2001 Harcourt Publishers Ltd. *

66 B. LoÈwenberg

the transplant be withheld and reserved in case of relapse? Who should receive an allogeneic HLA matched transplant from an unrelated donar and when? Who should receive high-dose cytarabine as part of their chemotherapy? Who should be entered onto experimental protocols? In addition to cytogenetics in recent years new leukaemia-related prognostic factors (e.g. ¯t-3 mutations) have been recognized, which may have an important impact on prognosis. Here we will review the current knowledge of factors that are predictive of the response to therapy and outcome in adults with AML (Table 1). AGE AND GENDER Age has a strong prognostic impact on the outcome of AML. There is a gradual but consistent decline in the prognosis for patients with AML with increasing age. This negative relationship is continuous from infant to 85 years of age. The latter relationship is apparent as a progressive decrease in treatment response and survival as a function of decade of age (Oxford Collaborative AML Intergroup, pers. comm.). For practical reasons of protocol design, a cut-o€ point of 60 years is usually used to distinguish older patients. Young and middle-aged adults, i.e. those less than 60 years old, have response rates to remission induction chemotherapy of approximately 75% and approximately 35±40% of them survive following diagnosis for 5 years or more.5±11 In contrast, patients with AML who are more than 60 years old have an overall response probability to induction therapy of 45±55%, and fewer than 10% of these survive for a minimum of 5 years.12,13 While the results of treatment have improved steadily over the last 20 years in younger adults, due to more intensive cytotoxic treatment, the use of stem cell grafts and improvements in supportive care, no signi®cant change in outcome has been noted among individuals of 60‡ years of age. The outcome for older subjects may even be worse than these estimates would indicate, since many older patients are not referred to hospitals for treatment, and are not registered or evaluated for outcome. Thus these outcome data for aged individuals probably represent overestimations. Why is outcome so unsatisfactory in the aged? The reasons for the poorer outcome of patients of higher age most probably relate to the increased frequency of unfavourable cytogenetics among older patients with AML, a greater frequency of antecedent myelodysplasia, as well as a greater frequency of drug resistance phenotypes. Also, because of the reduced general health of older patients, they cannot withstand intensive chemotherapy as well as younger individuals do.13,14 Several large databases of phase III studies in untreated patients with AML would also suggest that female sex is an independent favourable prognostic parameter for disease-free survival, although of moderate impact only. The intergroup analysis of the Oxford Collaborative AML Intergroup (pers. comm.) has con®rmed that disease-free survival of male patients is somewhat less than that of females. CYTOGENETICS AND MOLECULAR GENETICS Cytogenetic abnormalities are seen in approximately 60% of cases in AML and are highly predictive of response to therapy as well as the probability of relapse.15±18 The translocations t(8;21), inv16 and t(15;17) generally carry a relatively favourable prognosis and these are more commonly seen among younger patients with AML. The fusion genes of each of these translocations have been identi®ed and can be detected

Prognostic factors in acute myeloid leukaemia 67

using molecular probes in reverse transcriptase-polymerase chain reactions (RT-PCR). Among patients with these chromosomal abnormalities or the corresponding molecular abnormalities (AML1-ETO, CBFb-MYH11 (core binding factor beta genesmooth muscle myosin heavy chain gene), PML-RARa) the response to induction therapy is 80% or greater and for those entering remission the probability of relapse is 30±40% resulting in 5 years survival rates of 60±70%. The latter molecular or cytogenetic subsets of AML correlate with certain cytomorphological categories. For instance, t(15;17) or PML/RAR correlates with the French±American±British (FAB) group subtype M3, and inv16 with cytological FAB subtype M4 with eosinophils. However, the cytogenetics and molecular genetics rather than the cytological FAB classes, determine the overriding prognostic signi®cance. In contrast patients with t(6;9), 11q23 (MLL-gene) abnormalities19, monosomies of chromosomes 7 (-7) or 5 (-7) and complex cytogenetic abnormalities (three di€erent cytogenetic aberrations or more) generally have a distinctively poor prognosis. In these individuals (when 60 years or younger), the average complete response to induce treatment is 60% with a majority (approximately 80%) relapsing within 2 years so that their survival at 5 years is approximately 15% only.11,20 The cytogenetic and molecular indicators have led to the de®nition of prognostic groups, i.e. favourable risk (with favourable molecular genetics or cytogenetics), unfavourable risk (those with poor-risk cytogenetics) and intermediate risk (all others). Usually, these risk classi®cations are re®ned by considering selected additional prognostic determinants that enhance the value of these cytogenetic or molecular distinctions. The risk-scoring systems that are based upon multiple prognostic factors may permit a more precise approximation of prognosis in individual cases. To keep this practically simple, most collaborative groups have restricted the use of these additional parameters, and only taken into account a limited number of covariables that have a considerable impact. As an example, the Dutch HOVON (Haemato-Oncology Collaborative Group for Adults±Netherlands) group has required a white blood cell (WBC) count of less than 20  109/l in addition to favourable t(8;21) cytogenetics for the good risk category assignment (risk of relapse is 24% at 5 years: Figure 1). The Medical Research Council (MRC) has included the rapid versus late attainment of complete response (i.e. after cycle 1 versus after cycle 2) in their risk score in order to enhance the separation between good risk and poor risk.20

IMMUNOPHENOTYPING Cell stage and cell lineage speci®c markers have been used for immunophenotyping of AML. While the cytogenetic and molecular genetic markers carry a strong prognostic impact, the value of immunophenotyping for the assessment of prognosis has not been ®rmly established. This is true for the myeloid markers CD13, CD14, CD15, CD11b, HLA-DR, CD117 (c-kit) as well as CD34 for which controversial results regarding the prognostic value for response have been obtained in a variety of studies.21 The choice of antibodies, the threshold values for positivity (e.g. percentage positive cells) and clinical variations may explain some of the discrepancy between di€erent studies. More importantly, most positive correlations between immunophenotype and prognosis do not hold up in multivariate analysis indicating that other associated determinants exert a greater e€ect on prognosis.

0

25

50

75

100

0

0 39 4

34 4

27 3

20 2

23 2

Months 60

WBC>20

WBC<=20

24 3

Months 60

WBC>20

WBC<=20

0 96 79

0 297 206

N O WBC<=20 686 422 WBC>20 570 344 Log rank P = 0.27

152 100

(D) No favourable cytogenetics

At risk: WBC<=20 686 WBC>20 570

0

25

50

75

100

182 147

N O WBC<=20 546 269 WBC>20 450 203 Log rank P = 0.82

At risk: WBC<=20 546 WBC>20 450

0

25

50

75

100

(B) No favourable ctyogenetics

91 71

65 58

56 49

Months 60

WBC<=20

WBC>20

39 41

Months 60

WBC<=20

WBC>20

Figure 1. E€ect of white blood cell count on prognosis of AML (8;21). Probability of relapse in complete remittors with AML t(8;21) (panel A) and complete responders with AML without favourable cytogenetics (panel B). Overall survival for patients with AML t(8;21) is depicted in panel C, and for those without favourable cytogenetics in panel D. A white blood cell count of 20  109/l or more has a negative impact in the AML t(8;21) subgroup but not in other patients with AML. Thus the use of a WBC count criterion separates a relatively poor risk subset (20% of cases) from a cytogenetically good risk AML t(8;21). N, number of cases; O, number of events.

48 8

N O WBC<=20 63 21 WBC>20 14 9 Log rank P = 0.02

(C) t(8;21)

At risk: WBC<=20 63 WBC>20 14

0

25

50

75

100

41 6

N O WBC<=20 58 12 WBC>20 13 5 Log rank P = 0.05

At risk: WBC<=20 58 WBC>20 13

Relapse free (%)

Overall survival (%)

Relapse free (%) Overall survival (%)

(A) t(8;21)

68 B. LoÈwenberg

Prognostic factors in acute myeloid leukaemia 69

WHITE BLOOD CELL COUNT In several large prospective studies of previously untreated patients with AML, the WBC count stands out as an independent prognostic factor. There is a subgroup of patients presenting at diagnosis with hyperleucocytosis (WBC more than 100  109/l) who generally have a signi®cantly reduced complete response rate and a greater rate of relapse.22 The WBC count also provides a practically useful parameter for distinguishing the prognostic heterogeneity among patients with favourable cytogenetics. Thus those with t(8;21), inv(16) and t(15;17) as well as a WBC of 20  109/l or greater do signi®cantly worse than those with the same cytogenetic subtypes and low WBC counts (Figure 1). FAB CLASSIFICATION In multivariate analysis, the cytomorphological classi®cation also appears to add some prognostic information, independent of cytogenetics and molecular genotyping. In particular AMLs with the FAB types M0, M6 and M7 appear to correlate with an inferior outcome in most of the large phase III prospective trials. Since patients with M0, M6 and M7 represent minorities, the basis of the evidence for their prognostic value has remained limited and in clinical practice FAB classi®cation is not generally applied to risk assessment in AML. In small series of patients with minimally di€erentiated AML (FAB M0), AML-M0 was associated with a poor prognosis.23±27 Also AML with the natural killer (NK) cell immunophenotype appears to carry a poor prognosis.28 Notably, a signi®cant proportion of AML-M0 appear to carry point mutations in the runt domain of the AML1 gene. The functional consequences of these mutations in critical myeloid transcription factors are currently under investigation.29 DRUG RESISTANCE Resistance of AML to the cytotoxic e€ects of chemotherapy has remained the main stumbling block in obtaining a cure. Thus one would assume that insights into the molecular pathways leading to drug resistance might provide particularly powerful prognostic markers for response to therapy. Multidrug resistance type I (MDR1) or classical drug resistance is associated with the enhanced expression of P-glycoprotein (Pgp), an ATP binding drug transporter in the plasma membrane of leukaemic cells. This transporter may bind a variety of substrates (including anthracyclins and epipodophylotoxins). It acts as an e‚ux pump for these drugs. High levels of MDR1 have been associated with reduced intracellular concentrations of chemotherapy within leukaemic cells (see Chapter 12). The frequency of MDR1 positive AML markedly increases with age.14,30 In several studies MDR1 positivity has been shown to have a negative prognostic value for response to induction chemotherapy.14,30±35 This was seen in studies in which MDR1 was assessed by Pgp staining or by mRNA measurements. Other studies did not ®nd such correlations.36 It should be noted that a lack of agreement between some studies may relate to technical variations, e.g. the use of di€erent endpoints for positivity, di€erences in cell sample preparation (e.g. fresh versus cryopreserved) and analysis, and the use of di€erent immunological reagents.37 In some studies functional assays of drug export (e.g. rhodamine export inhibitable by an MDR blocker) have

70 B. LoÈwenberg

been used and have shown prognostic signi®cance for response.30,38,39 Of particular interest are the results of a study indicating a strong prognostic impact of simultaneous functional measurements of multidrug resistance associated protein (MRPI) and Pgp activity with regard to response.38 Combined MRP1 and Pgp resistance was mainly seen among the poor cytogenetic subgroup of AML. Notably, in most of these studies the prognostic signi®cance of MDR1 has been evaluated in the context of daunomycin chemotherapy. One study of patients treated with induction chemotherapy using idarubicin did not demonstrate a correlation between MDR1 expression and response.40 It is not unlikely that multiple resistance mechanisms co-operate in the same patient and in synergy determine clinical resistance. This is supported by the recent observation of two distinct resistance phenotypes that in combination added to the prognostic impact of the individual marker.38 Finally, it would probably make more sense to relate resistance phenotyping to speci®c subpopulations of AML cells, i.e. AML progenitor cells, rather than the overall AML cell population. In one study the assessment of MDR1 expression on the subpopulation of CD34‡ cells by dual surface marker analysis, signi®cantly enhanced the predictive value for response.34 Other genes involved in the cellular redistribution of chemotherapeutic agents are MRP (MDR associated protein), a transporter of the glutathione complex and the lung resistance protein (LRP). Some studies have indicated that LRP may predict for response as well as for leukaemia-free survival39,42,43, but other studies did not reveal a correlation between the expression of LRP and response.30,43

PROLIFERATIVE ABILITIES IN VITRO AND GROWTH FACTOR RECEPTOR MUTATIONS It has been shown that autonomous proliferation of AML cells in short term culture predicts for a comparatively high probability of relapse and poor survival.44 Patients with AML and high spontaneous proliferative activity, also more frequently express CD34 positivity and a multidrug resistance phenotype.34 More recently mutations in the receptors for the haematopoietic growth factors have been detected, in particular in fms-like tyrosine kinase 3 (¯t-3, receptor for FL or ¯t ligand) and kit (receptor for stem cell factor or kit-ligand).22,45±51 Flt-3, kit and fms (receptor for M-CSR) are class III receptor tyrosine kinases, which play a role in haematopoiesis as growth factor receptors and share similar structural domains. In a signi®cant proportion of cases of AML, c-kit50,52,53 and c-fms54 are expressed. A subset of cases of AML carry mutations of c-kit and c-fms. Kit mutations are mainly seen in patients with good risk cytogenetics.55 Internal tandem duplicates (ITDs) of ¯t-3 have been reported in the part of the ¯t-3 gene coding for the juxtamembrane (JM) in 15±20% of AML patients.22,51 The elongation of the JM domain causes ligand independent dimerization thus resulting in constitutive activation and a reduced responsiveness to stromal support in long term culture.49 Flt-3 mutations have been shown to confer a grave prognosis both in young and older adults, independent of other prognostic factors. Of particular note is that the large subset of cytogenetically de®ned patients of intermediate risk can be split into good risk and poor risk according the absence or presence of ¯t-3 mutations51 (Figure 2).

Prognostic factors in acute myeloid leukaemia 71

1.0

(A)

0.8

0.6 n=95 0.4

0.2 Event-free survival

n=29 0.0

0

1.0

12

24

36

48

60

(B)

0.8

0.6 n=69

0.4

0.2

0.0

n=26 0

12

24

36

48

60

Time after diagnosis (months) Figure 2. Kaplan±Meier event-free survival curves of the e€ect of Flt-3 mutations in AML. Continuous lines indicates Flt-3/ITD (internal tandem duplication) negative samples. Dotted lines refer to Flt-3/ITD positive populations. (A) gives event-free survival of the total AML population. (B) depicts event-free survival of the AML population with intermediate-risk cytogenetics.

SECONDARY LEUKAEMIA AND ADVANCED DISEASE There is a considerable body of data indicating that patients with AML subsequent to an antecedent haematological disorder, e.g. myelodysplasia, polycythemia vera, congenital neutropenia have a reduced probability of attaining a complete response, and disease-free survival and overall survival are also reduced.14,56

72 B. LoÈwenberg Table 1. Major prognostic factors in acute myeloid leukemia. For response

For relapse

Cytogenetics/molecular genetics

Cytogenetics/molecular genetics Time towards complete response White blood cell count ¯t-3 mutations Autonomous proliferation Secondary AML Age

White blood cell count MDR phenotype Secondary AML Age

Patients with AML relapsing after an initial complete response, have a notoriously bad prognosis overall. By and large, less than 10% of these will have the prospect of long-term leukaemia-free survival. Usually these patients are entered on to transplant protocols or, if these options are not available, they may be o€ered experimental therapy.57 A small subset (5±10%) of patients with a distinctly better prognosis may be identi®ed among cases of recurrent AML. These are those who have favourable cytogenetics (t(8;21), inv 16, t(15;17)), relapse following a ®rst remission of at least 12 months duration, have not had a prior stem cell transplant and are less than 35 years old. In particular, patients with recurrent AML with at least three of the latter characteristics have been estimated to have a probability of 5 years disease-free survival of approximately 40%.

CONCLUSION The understanding of dominant prognostic determinants of response to therapy in patients with AML is rapidly evolving. Cytogenetics has become part of the essential and standard work-up of patients with AML and currently furnishes distinctive insights into the nature of the disease and provides a useful clue to the prognosis of individual patients. However more precise distinctions are needed. New parameters with powerful prognostic signi®cance are emerging. When these latter variables are considered in one comprehensive prognostic model and when their value has been validated in large series of patients, these parameters are likely to provide more exact quantitative estimations of the prognosis. These prognostic distinctions are likely to provide the elementary foundations for treatment choice in the near future.

REFERENCES 1. Fialkow PJ, Singer JW, Adamson JW et al. Acute non-lymphocytic leukemia: heterogeneity of stem cell origin. Blood 1981; 57: 1068±1073. 2. LoÈwenberg B & Bauman JG. Further results in understanding the subpopulation structure in AML: clonogenic cells and their progeny identi®ed by di€erentiation markers. Blood 1985; 66: 1225±1232. 3. Grin JD & LoÈwenberg B. Clonogenic cells in acute myeloid leukemia. Blood 1986; 68: 1185±1195. 4. Chen SJ, Zhu YJ, Fong JH et al. Rearrangements in the second intron of the RAR alpha gene are present in a large majority of patients with acute promyelocytic leukemia and are used as molecular markers for retinoic acid induced leukemia cell di€erentiation. Blood 1981; 58: 584±590.

Prognostic factors in acute myeloid leukaemia 73 5. Vogler WR, Velez-Garcia E, Weiner RS et al. A phase III trial comparing idarubicin and danorubicin in combination with cytarabine in acute myelogenous leukemia: A Southeastern Cancer Study Group study. Journal of Clinical Oncology 1992; 10: 1103±1111. 6. Wiernik PH, Banks LC, Case DC Jr et al. Cytarabine plus idarubicin or daunorubicin as induction and consolidation therapy for previously untreated adult patients with acute myeloid leukemia. Blood 1992; 79: 313±319. 7. Mayer RJ, Davies RB, Schi€er CA et al. Intensive postremission chemotherapy in adults with acute myeloid leukemia. New England Journal of Medicine 1994; 331: 896±903. 8. Zittoun RA, Mandelli F, Willemze R et al. Autologous or allogeneic bone marrow transplantation compared with intensive chemotherapy in acute myelogenous leukemia. New England Journal of Medicine 1995; 332: 217±223. 9. Weick JK, Kopecky KJ, Appelbaum FR et al. A randomized investigation of high-dose versus standarddose cytosine arabinoside with daunorubicin in patients with previously untreated acute myeloid leukemia: a South West Oncology Group Study. Blood 1996; 88: 2841±2851. 10. Burnett AK, Goldstone AH, Stevens RMF et al. Randomised comparison of addition of autologous bonemarrow transplantation to intensive chemotherapy for acute myeloid leukaemia in ®rst remission: results of MRC AML 10 trial. Lancet 1998; 351: 700±708. 11. Cassileth PA, Harrington DP, Appelbaum FR et al. Chemotherapy compared with autologous or allogeneic bone marrow transplantation in the management of acute myeloid leukemia in ®rst remission. New England Journal of Medicine 1998; 339: 1649±1656. 12. LoÈwenberg B, Suciu S, Archimbaud E et al. Mitoxantrone versus daunorubicin in induction-consolidation chemotherapy. The value of low-dose cytarabine for maintenance of remission, and an assessment of prognostic factors in acute myeloid leukemia in the elderly: ®nal report of the Leukemia Cooperative Group of the European Organization for the Research and Treatment of Cancer and the Dutch-Belgian Hemato-Oncology Cooperative HOVON Group. Randomized phase III study AML-9. Journal of Clinical Oncology 1998; 16: 872±881. 13. LoÈwenberg B. Treatment of the elderly patient with acute myeloid leukemia. BallieÁre's Clinical Haematology 1996; 9: 147±161. 14. Leith CP, Kopecky KJ, Godwin J et al. Acute myeloid leukemia in the elderly in assessment of multidrug resistance (MDR1) and cytogenetics distinguishes biologic subgroups with remarkably distinct responses to standard chemotherapy. A South West Oncology Group Study. Blood 1997; 89: 3323±3329. 15. Yunis JJ, Brunning RD, Howe RB & Lobell M. High-resolution chromosomes as an independent prognostic indicator in adult acute nonlymphocytic leukemia. New England Journal of Medicine 1984; 311: 812±818. 16. Keating MJ, Smith TL, Kantarjian H et al. Cytogenetic pattern in acute myelogenous leukemia: a major reproducible determinant of outcome. Leukemia 1988; 2: 403±412. 17. Mrozek K, Heinonen K, de la Chapelle A & Bloom®eld CD. Clinical signi®cance of cytogenetics in acute myeloid leukemia. Seminars in Oncology 1997; 24: 17±31. 18. Bloom®eld CD, Lawrence D, Byrd JC et al. Frequency of prolonged remission duration after high-dose cytarabine intensi®cation in acute myeloid leukemia varies by cytogenetic subtype. Cancer Research 1998; 58: 4173±4179. 19. Dimartino JF & Cleary ML. MLL-rearrangements in haematological malignancies: lessons from clinical and biological studies. British Journal of Haematology 1999; 106: 604±626. 20. Wheatly K, Burnett AK, Goldstone AH et al. A simple, robust, validated and highly predictive index for the determination of risk-directed therapy in acute myeloid leukemia derived from the MRC AML 10 trial. British Journal of Haematology 1999; 107: 69±79. 21. Sanz MA & Sempere A. Immunophenotyping of AML and MDS and detection of residual disease. BallieÁre's Clinical Haematology 1996; 9: 35±55. 22. Kiyoi H, Naoe T, Nakano Y et al. Prognostic implication of FLT 3 and N-RAS gene mutations in acute myeloid leukemia. Blood 1999; 93: 3074±3080. 23. Lee EJ, Pollak A, Leavitt RD et al. Minimally di€erentiated acute nonlymphocytic leukemia: a distinct entity. Blood 1987; 70: 1400±1406. 24. Bennett JM, Catovsky D, Daniel MT et al. Proposals for the recognition of minimally di€erentiated acute myeloid leukaemia. British Journal of Haematology 1991; 78: 325±329. 25. Stasi R, Del Poeta G, Venditti A et al. Analysis of treatment failure in patients with minimally di€erentiated acute myeloid leukaemia (AML-M0). Blood 1994; 83: 1619±1625. 26. Segeren CM, De Jong-Gerrits GCMM & Van't Veer MB. AML-M0: clinical entity or waste basket for immature blastic leukaemias? A description of 14 patients. Annals of Hematology 1995; 70: 297±300. 27. Cuneo A, Ferrant A, Michaux JL et al. Cytogenetic pro®le of minimally di€erentiated (FAB-M0) acute myeloid leukaemia: correlation with clinicobiologic ®ndings. Blood 1995; 85: 3688±3695.

74 B. LoÈwenberg 28. Paietta E, Gallagher RE & Wiernick PH. Myeloid/natural killer cell acute leukaemia: a previously unrecognized form of acute leukaemia potentially misdiagnosed as FAB-M3 acute myeloid leukaemia. Blood 1994; 84: 2824±2825. 29. Ferrao P, Gouda TJ & Ashman LK. Expression of constitutively activated human c-kit in myb transformed early myeloid cells leads to factor independence, histiocytic di€erentiation, and tumorigenicity. Blood 1997; 90: 4539±4552. 30. Leith CP, Kopecky KJ, Ming Chen I et al. Frequency and clinical signi®cance of the expression of the multidrug resistance proteins MDR1/P-glycoprotein, MRP1 and LRP in acute myeloid leukemia. A South West Oncology Group Study. Blood 1999; 94: 1086±1098. 31. Sato H, Gottesmann MM, Goldstein LJ et al. Expression of the multidrug resistance gene in myeloid leukemias. Leukemia Research 1990; 14: 11±22. 32. Marie JP, Zittoun R & Sikic BI. Multidrug resistance (MDR1) gene expression in adult acute leukemias: correlations with treatment outcome and in vitro drug sensitivity. Blood 1991; 78: 586±592. 33. Campos L, Gyotat D, Archimbaud E et al. Clinical signi®cance of multidrug resistance P-glycoprotein expression on acute nonlymphoblastic leukemia cells at diagnosis. Blood 1992; 79: 473±476. 34. Te Boekhorst PAW, De Leeuw K, Schoester M et al. Predominance of functional multidrug resistance (MDR-1) phenotype in CD34‡ leukemia cells. Blood 1993; 82: 3157±3162. 35. Ino T, Miyazaki H, Isogai M et al. Expression of P-glycoprotein in de novo acute myelogenous leukemia at initial diagnosis: results of molecular and functional assays, and correlation with treatment outcome. Leukemia 1994; 8: 1492±1497. 36. Broxterman HJ, Sonneveld P, Pieters R et al. Do P-glycoprotein and major vault (MVP/LRP) expression correlate with in vitro daunorubicin resistance in acute myeloid leukemia? Leukemia 1999; 13: 258±265. 37. Broxterman HJ, Sonneveld P, Feller N et al. Quality control of multidrug resistance assays in adult P-glycoprotein expression and activity. Blood 1996; 87: 4809±4816. 38. Legrand O, Simonin G, Beauchamp-Nicoud A, Zittoun R & Marie JP. Simultaneous activity of MRP1 and Pgp is correlated with in vitro resistance to daunorubicin and with in vivo resistance in adult acute myeloid leukemia. Blood 1999; 94: 1046±1056. 39. Borg AG, Burgess R, Green LM, Scheper RJ & Lin Yin JA. Overexpression of lung-resistance protein and increased P-glycoprotein function in acute myeloid leukemia cells predict a poor response to chemotherapy and reduce patient survival. British Journal of Haematology 1998; 103: 1083±1091. 40. Broxterman HJ, Sonneveld P, Van Putten WJL et al. P-glycoprotein in primary acute myeloid leukemia and treatment outcome of idarubicin/cytosine arabinoside-based induction therapy. Leukemia 1999; 14: 1018±1024. 41. Filipits M, Pohl G, Stranzl Th et al. Expression of the lung resistance protein predicts poor outcome in de novo acute myeloid leukemia. Blood 1998; 91: 1508±1513. 42. List AF, Spier CS, Grogan TM et al. Overexpression of the major vault transporter protein lungresistance protein predicts outcome in acute myeloid leukemia. Blood 1996; 87: 2464±2469. 43. Michaeli M, Damiani D, Ermacora A et al. P-glycoprotein, lung-resistance-related protein and multi-drug resistance associated protein in de novo acute non lymphocytic leukaemias: biological and clinical implications. British Journal of Haematology 1999; 104: 328±335. 44. LoÈwenberg B, Van Putten WWLJ, Touw IP et al. Autonomous proliferation of leukemic cells in vitro as a determinant of prognosis in adult acute myeloid leukemia. New England Journal of Medicine 1993; 328: 614±619. 45. Nakao M, Yokota S, Iwai T et al. Internal tandem duplication of the Flt-3 gene found in acute myeloid leukemia. Leukemia 1996; 10: 1911±1918. 46. Yokota S, Kiyoi H, Nakao M et al. Internal tandem publication of the Flt-3 gene is preferentially seen in acute myeloid leukemia and myelodysplastic syndrome among various hematological malignancies: a study on a large series of patients and cell lines. Leukemia 1997; 11: 1605±1609. 47. Horiike S, Yokota S, Nakao M et al. Tandem duplications of the Flt-3 receptor gene are associated with leukemic transformation of myelodysplasia. Leukemia 1997; 11: 1442±1446. 48. Kiyoi H, Naoe T, Yokota S, Nakao M & Minami S. Internal tandem duplication of Flt-3 associated with leukocytosis in acute promyelocytic leukemia. Leukemia 1997; 11: 1447±1452. 49. Rombouts WC, Broyl A, Martens ACM, Slater R & Ploemacher R. Human acute myeloid leukemia cells with internal tandem duplications in the Flt-3 gene show reduced proliferative ability in stroma supported long-term cultures. Leukemia 1999; 13: 1071±1078. 50. Gari M, Goodeye A, Wilson G et al. c-Kit protooncogene exon 8 in-frame deletion plus insertion mutations in acute myeloid leukemia. British Journal of Haematology 1999; 105: 894±900. 51. Rombouts WJC, Blokland I, LoÈwenberg B & Ploemacher RE. Biological characteristics and prognosis of adult acute myeloid leukemia with internal tandem duplications in the Flt3 gene. Leukemia 2000; 14: 675±683.

Prognostic factors in acute myeloid leukaemia 75 52. Schwartz S, Heinecke A, Zimmermann M et al. Expression of the c-kit receptor (CD117) is a feature of almost all subtypes of de novo acute myeloblastic leukemia (AML), including cytogenetically good-risk AML, and lacks prognostic signi®cance. Leukemia and Lymphoma 1999; 34: 85±94. 53. Wang, Curtis JE, Geiseller EN, McCullock EA & Minden MD. The expression of the proto oncogene ckit in the blast cells of acute myeloblastic leukemia. Leukemia 1989; 3: 699±702. 54. Dubreuil P, Torres H, Courcoul MA, Birg F & Mannoni P. C-fms expression is a marker of human acute myeloid leukemias. Blood 1988; 72: 1081±1085. 55. Berghini A, Peterlongo P, Ripamonti CB, Larizza L & Cairoli R. C-kit mutations in core binding leukemias. Blood 2000; 95: 726±727. 56. Keating AJ, McCredie KB, Benjamin RS et al. Treatment of patients over 50 years of age with acute myelogenous leukemia with a combination of rubidizone and cytosine, arabinoside, vincristine and prednisone (ROAP). Blood 1981; 58: 584±590. 57. LoÈwenberg B, Downing JR & Burnett A. Acute myeloid leukemia. New England Journal of Medicine 1999; 341: 1051±1962. 58. Osato M, Asou N, Abdalla E et al. Bi-allelic and heterozygous point mutation in the runt domain of the AML1/PEBP2aB gene associated with acute myeloblastic leukemia. Blood 1999; 93: 1817±1824. 59. Cahn JY, Labopin M, Mandelli F et al. Autologous bone marrow transplantation for ®rst remission acute myeloblastic leukemia in patients older than 50 years; a retrospective analysis of the European Bone Marrow Transplant Group. Blood 1995; 85: 575±579.