|
Advertisement |
|
|
||
 |
|
|||||
|
|||||
|
 |
||||||
|
||||||
|
||||||
|
Journal of Clinical Oncology, Vol 24, No 22 (August 1), 2006: pp. 3686-3692 © 2006 American Society of Clinical Oncology. DOI: 10.1200/JCO.2005.05.4312 Residual Disease Monitoring in Childhood Acute Myeloid Leukemia by Multiparameter Flow Cytometry: The MRD-AML-BFM Study Group
From the Department of Pediatric Hematology and Oncology, Hannover Medical School, Hannover; Department of Pediatric Hematology and Oncology, University Children's Hospital, Muenster; Germany; Children's Cancer Research Institute, St Anna Children's Hospital, Vienna, Austria; Department of Immunology/Pediatric Hematology/Oncology, Charles University, Prague, Czech Republic; and the Department of Hematology and Oncology, University of Goettingen, Goettingen, Germany Address reprint requests to Claudia Langebrake, PhD, Department of Pediatric Hematology and Oncology, Hannover Medical School, Carl-Neuberg-Strasse 1, D-30625 Hannover, Germany; e-mail: langebrake.claudia{at}mh-hannover.de
PURPOSE: Monitoring of residual disease (RD) by flow cytometry in childhood acute myeloid leukemia (AML) may predict outcome. However, the optimal time points for investigation, the best antibody combinations, and most importantly, the clinical impact of RD analysis remain unclear. PATIENTS AND METHODS: Five hundred forty-two specimens of 150 children enrolled in the AML-Berlin-Frankfurt-Muenster (BFM) 98 study were analyzed by four-color immunophenotyping at up to four predefined time points during treatment. For each of the 12 leukemia-associated immunophenotypes and time points, a threshold level based on a previous retrospective analysis of another cohort of children with AML and on control bone marrows was determined. RESULTS: Regarding all four time points, there is a statistically significant difference in the 3-year event-free survival (EFS) in those children presenting with immunologically detectable blasts at 3 or more time points. The levels at bone marrow puncture (BMP) 1 and BMP2 turned out to have the most significant predictive value for 3-year-EFS: 71% ± 6% versus 48% ± 9%, PLog-Rank = .029 and 70% ± 6% versus 50% ± 7%, PLog-Rank = .033), resulting in a more than two-fold risk of relapse. In a multivariate analysis, using a combined risk classification based on morphologically determined blasts at BMP1 and BMP2, French-American-British classification, and cytogenetics, the influence of immunologically determined RD was no longer statistically significant. CONCLUSION: RD monitoring before second induction has the same predictive value as examining levels at four different time points during intensive chemotherapy. Compared with commonly defined risk factors in the AML-BFM studies, flow cytometry does not provide additional information for outcome prediction, but may be helpful to evaluate the remission status at day 28.
Minimal residual disease (MRD) monitoring in childhood and adult acute myeloid leukemia (AML) using flow cytometry is still under discussion in terms of the prognostic impact, the optimal time points for analysis, and the best antibody combinations. AML blast cells do not express specific antigens that could serve as single and unambiguous markers for RD in regenerating bone marrow. It is therefore necessary to carefully characterize combinations of antigens that are able to sensitively detect residual blast cells among normal hematopoietic cells during treatment. Another obstacle for MRD monitoring in AML is the instability of the blast cell antigen expression pattern. As previously shown by us1 and others,2-4 the vast majority of AML cases undergo a shift of antigen expression pattern between diagnosis and eventual relapse. It is therefore indispensable for MRD evaluation to monitor a wide range of leukemia-associated immunophenotypes (LAIP). Until now, there are only a few reports about the prognostic relevance of RD monitoring in pediatric5,6 and adult7-10 AML. These investigations were based on different therapy regimens and have employed divergent technical approaches in terms of utilized LAIP, time points of analysis during therapy, and the positivity thresholds used for outcome prediction. The objective of this study was to determine the following assessment criteria for the AML-Berlin-Frankfurt-Muenster (BFM) treatment strategy: the sensitivity of specific LAIP based on a retrospective analysis of children with AML; the appropriateness of RD assessment by multidimensional flow cytometry for outcome prediction in children with AML; the most predictive time point during therapy; and the additional value of flow cytometric assays for outcome prediction in comparison to known risk factors.
Study Design The AML-BFM MRD study is composed of two phases. Phase A includes the establishment and standardization of a consensus panel for four-color immunophenotyping. The focus was to define the sensitivity and specificity of different LAIP in normal and regenerating bone marrow specimens11 and to identify clinically relevant threshold-levels for each LAIP at defined time points during treatment in a retrospective approach. Phase B was designed to prospectively apply these thresholds to evaluate the impact of RD monitoring for outcome prediction.
Patients
Phase B. Children enrolled in the AML-BFM 98 study and diagnosed for de novo AML between January 1, 2002 and July 31, 2004 were eligible for the prospective evaluation. Excluded were children with acute promyelocytic leukemia (AML French-American-British [FAB] M3), with t(15;17)/PML-RAR  and children with trisomy 21 because they exhibit biologically different leukemias and receive slightly different chemotherapy. Altogether, 542 samples from 150 children (Table 1) were available. This cohort is representative as compared with the overall study collective in terms of age, sex, WBC, FAB classification, and risk-group allocation. From these 150 children, five children with detectable blast cells by flow cytometry, but an antigen expression pattern that was not covered by the LAIPs used for RD monitoring (CD33/7/117/56, CD33/56, CD7/33) were excluded. The standard risk group (SR) definition that is generally used for stratification in the AML-BFM studies comprises FAB subtype (M1/M2+ Auer rods, M3, M4eo), favorable cytogenetics, and morphologically determined bone marrow blasts of less than 5% at day 15 (not required for FAB M3).12 For this analysis, the morphologic evaluation at day 28 of treatment (< 5% blasts in bone marrow) was included as an additional parameter for the SR group (herein referred to as extended AML-BFM risk). The date of each bone marrow puncture was correlated to the courses of intensive chemotherapy of the AML-BFM 98 study (Fig 1). Only specimens that were obtained at one of the first four scheduled time points (bone marrow puncture [BMP]1: day 15; BMP2: before second induction; BMP3: before third therapy course; BMP4: before fourth therapy course) were included in the study (Table 2).
Bone marrow specimens were obtained after informed consent from each patient or each patient's guardian. All children were treated according to the German AML-BFM 98 study (as to the treatment schedules see Creutzig et al13). All investigations performed had been approved by the local ethics committees and were in accordance with an assurance filed with and approved by the Department of Health and Human Services.
Diagnosis Cytogenetic and moleculargenetic data were obtained from the reference laboratory of the AML-BFM study (J. Harbott, Giessen, Germany).
Multiparameter Flow Cytometry A wide antibody panel based on a CD33/CD34 backbone, independent of the initial immunophenotype including fluorescence conjugated myeloid markers CD13-PE (SJ1D1; Immunotech, Krefeld, Germany), CD15-FITC (MMA; Becton Dickinson, Heidelberg, Germany), CD33-PC5 (D3HL60.251; Immunotech), CD33-APC (D3HL60.251; Immunotech), and HLA-DR-FITC (L243, Becton Dickinson), lymphoid markers CD7-PE (8H8.1; Immunotech), CD10-FITC (ALB2; Immunotech), CD19-FITC (J4.119; Immunotech), CD56-PE (NCAM 16.2; Becton Dickinson), the activation and proliferation marker CD38-PE (HB7; Becton Dickinson) as well as the progenitor-associated markers CD34-APC (8G12; Becton Dickinson), CD34-PC7 (581; Immunotech) and CD117-FITC (95C3; Immunotech) was applied (Table 3). Syto 16 (Molecular Probes, Eugene, OR) was used for staining nucleated cells to exclude debris and not completely lysed erythrocytes from analysis.
After incubating the bone marrow samples with monoclonal antibodies for 15 minutes, erythrocytes were lysed for 7 minutes using FACS Lysing Solution (Becton Dickinson) or Versa Lyse (Beckman Coulter, Krefeld, Germany). Afterwards, the specimens were washed twice with 2 mL phosphate-buffered saline (PBS)-buffer (pH 7.4) and centrifuged (5 minutes, 20°C, 600 g) to remove excess antibodies and lysed RBCs. Specimens were measured using the FACS-Calibur (Becton Dickinson) or EPICS (Beckman Coulter), analyzing at least 30,000 events. Extensive interinstrumental comparisons were performed to ensure that either site could detect similar percentages of positivity.
Data Analysis/Gating Strategy
Statistics
Definition of Cut Off Levels by Retrospective Analysis The determination of LAIP specificity has been described in detail previously.11 In brief, bone marrow specimens of 39 children with acute lymphoblastic leukemia, Ewing sarcoma, non-Hodgkin's lymphoma, or without a malignant disease were evaluated for the presence and amount of different LAIP. Three groups of specificity could be defined according to the median percentage of LAIP in regenerating bone marrow: low specificity with 1.0% or more, medium specificity from 0.1 to 1.0%, and high specificity with less than 0.1%. Based on these results, clinically prognostic relevant thresholds for each specificity group and time points were defined by retrospectively analyzing 25 children with relapse and 40 children without relapse at a median follow-up of 1.5 years. Low specificity LAIP are only informative at BMP1, while high and very high specificity LAIP can be utilized for discrimination at all four time points. According to these data, the thresholds given in Figure 2 were calculated that are able to unambiguously discriminate between those children who relapsed and those children in continuous complete remission.
Serial Assessment of RD at Three or More Time Points Identifies Children With Poor Prognosis Children with at least three specimens until BMP4 (n = 95) have been evaluated to investigate the impact of serial immunologic monitoring for outcome prediction. In 34 children, all measured LAIP levels were below the threshold at each time point. In 13 children, the measured LAIPs in at least three specimens were above the determined threshold. In 48 children, one or more LAIPs were above the threshold at one or two time points (Table 4).
Using this approach, it was possible to identify a poor-risk group, characterized by positive flow cytometric assays at three or more consecutive time points during chemotherapy, with an EFS of 31% ± 11% (Fig 3). However, EFS was not significantly different between children who were negative by flow cytometry at all analyzed time points and those with one or two positive time points (73% ± 8% v 61% ± 7%, P = .43). Combining the good and the intermediate group, the difference in contrast to the poor group was statistically significant: PEFS 65% ± 5% v 31% ± 13%; P = .02. Shifting the cut off levels, which were used for group allocation, did not result in a better distinction in terms of EFS (data not shown).
Single Time Point RD Assessment Before Second Induction Is Most Informative for Outcome Prediction When investigating the four different time points separately, we found that only at BMP1 and BMP2, there is a statistically significant difference between RD-positive and RD-negative children in the 3-year EFS (Fig 4; 71% ± 6% v 48% ± 9%; PLog-Rank = .029; 70% ± 6% v 50% ± 7%: PLog-Rank = .033). This is also true, if only high and very high specificity LAIPs are regarded (70% ± 6% v 37% ± 11%; PLog-Rank = .031; 66% ± 6% v 45% ± 8%; PLog-Rank =.045). Similarily, only at BMP1 and BMP2 the RD-positive children differed significantly from the RD-negative children regarding their cumulative nonresponse and relapse incidence: 48% ± 10% v 25% ± 6%; PGray = .02; 48% ± 7% v 25% ± 6%; PGray = .01. Regarding 3-year overall survival, only BMP2 turned out to be a statistically significant discriminator (P = .036).
MRD Monitoring Has No Additional Prognostic Impact Compared With Known Risk Factors Using the univariate COX regression model for FFS risk assessment, both BMP1 and BMP2 had statistically significant impact: RR, 2.35 (95% CI, 1.13 to 4.89; P = .021) and risk ratio, 2.21 (95% CI, 1.18 to 4.14; P = .013), respectively. Regarding only those children, who responded to treatment as determined by morphology at day 15, there is a statistically significant difference in 3-year EFS between MRD-negative and MRD-positive children: 69% ± 6% v 40% ± 15% (P < .05). A multivariate analysis controlling for AML-BFM risk classification, including FAB subtype, cytogenetics, and morphologically determined blasts at day 15, was performed. At BMP1, both flow cytometry and AML-BFM risk show almost the same risk ratio for FFS with similar 95% CIs (2.09; 1.00 to 4.39; 2.06; 0.87 to 4.88). The influence of BMP2 on the risk of failure is less than the impact of the AML-BFM risk, however, the differences are not significant. Using an extended risk group classification including morphologically determined blasts at day 28 as a covariate, this turned out to have more impact on FFS with a RR of 2.8 for both time points (Table 5).
We further analyzed, whether the inclusion of flow cytometry in the risk group assessment could help to define more precisely risk groups for additional treatment stratification. Therefore, the 3-year EFS according to immunologically determined RD was calculated separately for the SR and high-risk group. There was no difference in both groups at BMP1 or BMP2 (Table 6).
Due to the fact that the antigen expression pattern of AML blasts differs significantly between diagnosis and relapse1-4 and that no specific antigens exist which clearly can identify leukemic blasts, we have developed an antibody panel that allows us to detect residual blast cells independently of the initial immunophenotype. Based on a CD33/CD34 basis, the LAIPs utilized for MRD assessment comprise the commonly accepted antigen expression patterns in AML.18-20 We are the first group applying time-dependent prognostically relevant cut off levels that have been determined by the retrospective analysis of children treated within the AML-BFM studies for the occurrence of 12 different LAIPs at defined time points, instead of empirically defining cut off level. In our international prospective study, we were able to show that the detection of residual blast cells by flow cytometry at early time points of follow-up (until day 84) is a significant predictor of treatment outcome regarding 3-year EFS. Especially in the very early course of therapy—before the start of the second induction, at day 28 from diagnosis—multidimensional flow cytometry can help to differentiate between children with good and poor prognosis. Similar results were obtained by the exclusive analysis of LAIPs with high or very high specificity, indicating that antigene combinations with low or medium specificity are not relevant for RD monitoring. Furthermore, flow cytometry is also able to detect "minimal" RD in those children with morphologically undetectable blasts after first induction. Our results are consistent with reports on pediatric and adult AML regarding the prognostic impact of immunological blast detection after first induction.5,7,9 The series of Coustan-Smith et al5 comprises the analysis of residual blasts at the end of remission induction therapy by defined marker combinations dependent on the initial immunophenotype. In the RD-negative group, their results show less than 0.1% residual AML cells: six children (21%) relapsed, and three children (10%) died in CR, whereas in the RD-positive group the proportion of children who relapsed and of those in CR is equal (both relapse and CCR, n = 5; 38%; death in CR, n = 3). The probability of 2-year overall survival is statistically different, but the assimilation of the two curves after that time indicates that a stable situation has not yet been achieved. Other investigator groups found that only the monitoring at later time points6,8,10 (after consolidation therapy) is significant for outcome prediction, which may limit the clinical usefulness of these data for risk-adapted therapy tailoring. The most recent report by Kern et al21 even revealed that only time points longer than 1 year after diagnosis were independently related to EFS and overall survival. During this period of therapy follow-up, RD positivity may rather have represented resurgent leukemia (occult relapse) than a surrogate marker of initial therapy response usable for treatment stratification. The individual differences in the kinetics of leukemic recurrences may therefore impede a prospective application of this approach for the early diagnosis of relapse. Although we could show that flow cytometry is a reliable and objective method to detect residual blast cells in regenerating bone marrow specimens, and that it is therefore appropriate for outcome prediction, we further wanted to know whether these results bear additive values for treatment stratification as compared with conventional risk factors. Notably, the AML-BFM risk group classification12 is based on initial cytogenetics, FAB classification, and morphologically detectable blast cells of more or less than 5% at day 15. Applying this binary risk group classification to the children analyzed, a highly significant difference in terms of 3-year EFS is achieved without using immunological information: 78% ± 6% versus 49% ± 6% (P = .0025). When including information on morphological blast counts from day 28 (BMP2) in addition to the conventional risk classification (extended AML-BFM risk classification), the difference in 3-year EFS between the two risk groups even increased to 85% ± 5% versus 48% ± 5% (P = .0002). The impact of flow cytometry results on FFS at either time point was found equivalent to that of the AML-BFM risk in terms of risk ratio and 95% CIs. When including the extended AML-BFM risk classification in the COX-model, it turned out that immunological RD as a covariate does not contribute to a better risk group separation. It has to be mentioned that none of the recently published studies of other groups included information of RD monitoring as compared with morphologically determined blast cells at the analyzed time points as part of the risk classification system. Considering our results, this comparison has to be recommended in order to interpret potential additional values of flow cytometric investigations correctly. Applying the covariates used by other groups (age at diagnosis, leukocyte count at diagnosis, karyotype) to our data, we obtain similar significant results for the influence of BMP1 and BMP2 on outcome. However, in our study the more sophisticated AML-BFM risk group classification has been taken as the basis to determine the true additional value of MRD monitoring. In conclusion, for the AML-BFM studies, risk group stratification based on FAB subtype, cytogenetics, and morphologically determined bone marrow blasts before second induction does not benefit from inclusion of RD data, as assessed by multidimensional flow cytometry. Further investigations will focus on an improved risk group stratification including blast percentage at day 28. In a prospective study, we will evaluate whether the discrimination of blast cells and regenerating bone marrow cells can be improved in terms of accuracy and objectivity by flow cytometry as compared with morphological interpretation alone.
Principal investigators of the Acute Myeloid Leukemia-Berlin-Frankfurt-Muenster 98 studies in Germany. R. Mertens, Kinderklinik RWTH, Aachen; A. Gnekow, I. Kinderklinik des KZVA, Augsburg; Th. Rupprecht, Universitäts-Kinderklinik GmbH, Bayreuth; G. Henze/R. Fengler, Charité Campus Virchow-Klinikum, Berlin; M. Schöntube/A.-K. Liebeskind, Helios Klinikum Berlin-Buch, Berlin; N. Jorch, Ev. Krankenhaus Bielefeld gGmbH, Bielefeld; U. Bode/G. Fleischhack, Universitäts-Kinderklinik, Bonn; H.G. Koch/W. Eberl, Städt. Klinikum, Braunschweig; A. Pekrun, Prof.-Hess-Kinderklinik, Bremen; I. Krause, Klinikum Chemnitz gGmbH, Chemnitz; E. Holfeld, Carl-Thiem-Klinikum, Cottbus; W. Andler/Th. Wiesel, Vestische Kinderklinik, Datteln; C. Niekrens, Kinderklinik der Städtischen Kliniken, Delmenhorst; H. Olschewski, Kinderklinik der Städt. Kliniken, Dortmund; M. Suttorp/ I. Lauterbach, Universitäts-Kinderklinik Carl-Gustav-Carus, Dresden; U. Göbel, Universitäts-Kinderklinik, Düsseldorf; A. Sauerbrey/G. Weinmann, Helios Klinikum Erfurt GmbH, Erfurt; J.D. Beck, Universitäts-Kinderklinik, Erlangen; B. Kremens, Universitäts-Kinderklinik, Essen; Th. Klingebiel/Th. Lehrnbecher, Klinikum d. J. W. Goethe-Universität, Frankfurt; C.M. Niemeyer, Universitäts-Kinderklinik Freiburg; A. Reiter/R. Blütters-Sawatzki, Universitäts-Kinderklinik, Gießen; M. Lakomek/L. Schweigerer, Georg-August-Universität, Göttingen; J.F. Beck/H. Weigel, Universitäts-Kinderklinik, Greifswald; E. Pretel, Kreiskrankenhaus, Gummersbach; D. Körholz/R. Schobeß, Martin-Luther-Universität Halle-Wittenberg, Halle; R. Schneppenheim/H. Kabisch, Universitätsklinikum Hamburg-Eppendorf, Hamburg; K. Welte, Zentrum f. Kinderheilkunde der Med. Hochschule, Hannover; A. E. Kulozik, Universitäts-Kinderklinik, Heidelberg; N. Graf, Universitäts-Kinderklinik, Homburg/Saar; J. Hermann, FSU Jena, Klinik f. Kinder- und Jugendmedizin, Jena; J. Kühr/A. Leipold, Städtische Kinderklinik, Karlsruhe; M. Rodehüser, Städt. Kinderklinik, Kassel; M. Schrappe/A. Claviez, Universitätsklinikum Schleswig-Holstein, Campus Kiel, Kiel; M. Rister, Städt. Klinikum Kemperhof, Koblenz; F. Berthold, Universitäts-Kinderklinik, Köln; W. Sternschulte, Städt. Kinderkrankenhaus Riehl, Köln; S. Völpel, Städt. Krankenhäuser, Krefeld; U. Bierbach, Universitäts-Kinderklinik, Leipzig; S. Selle, Kinderklinik St. Annastift, Ludwigshafen; P. Bucsky, Universitäts-Kinderklinik, Lübeck; U. Mittler/U. Kluba, Otto-v. Guericke-Universität, Magdeburg; P. Gutjahr, Klinikum d. Joh. Gutenberg-Universität, Mainz; M. Dürken, Universitäts-Kinderklinik, Mannheim; H. Christiansen, Klinikum d. Philipps-Universität, Marburg; A. Borkhardt, Kinderklinik und Poliklinik im Dr v. Haunerschen Kinderspital (Klinikum der Universität München), München; St. Burdach/A. Wawer, Kinder- und Poliklinik des Klinikums rechts der Isar der Technischen Universität München Kinderklinik Schwabing, München; R. Roos/T. Papousek, Städt. Krankenhaus Harlaching, München; H. Jürgens, Universitäts-Kinderklinik, Münster; W. Scheurlen/A. Jobke, Cnopf'sche Kinderklinik, Nürnberg; H. Gröbe/U. Schwarzer, Klinikum Nürnberg, Nürnberg; H. Müller/R. Kolb, Klinikum Oldenburg gGmbH, Oldenburg; N. Albers, Kinderspital, Osnabrück; F.J. Helmig/O. Peters, Klinik St. Hedwig, Regensburg; G. Eggers, Universitäts-Kinderklinik, Rostock; R. Dickerhoff, Asklepios Klinik St. Augustin GmbH, St. Augustin; St. Bielack, Olgahospital, Stuttgart; W. Rauh, Krankenanstalt, Mutterhaus der Borromäerinnen e.V., Trier; R. Handgretinger/H. Scheel-Walter, Universitäts-Kinderklinik Tübingen; K.-M. Debatin, Universitäts-Kinderklinik, Ulm; M. Albani/G. Beron, Dr Horst-Schmidt-Kinderklinik, Wiesbaden; T. Liebner, Reinhardt-Nieter-Krankenhaus, Wilhelmshaven; H. Sopnik, Stadtkrankenhaus, Worms; P.-G. Schlegel/St. Rutkowski, Universitäts-Kinderklinik, Würzburg; K. Sinha/B. Dohrn, Klinikum Barmen, Wuppertal. Principal investigators in Austria. Ch. Urban, Universitätsklinik für Kinder- und Jugendheilkunde, Graz; F.-M. Fink/ B. Meister, Univ. Klinik für Kinder- und Jugendheilkunde, Innsbruck; K. Schmitt, Landes-Kinderklinik, Linz; O. Stöllinger, Krankenhaus der Barmherzigen Schwestern, Linz; N. Jones, St. Johanns Spital / Landeskrankenhaus, Salzburg; H. Gadner/M. Dworzak, Zentrum für Kinder- u. Jugendheilkunde im St. Anna Kinderspital, Wien. Principal investigators in the Czech Republic. H. Hrstkova/J. Sterba, University Hospital, Brno; K. Tousovska, University Hospital, Hradec Kralove; J. Stary, University Hospital Motol, Prague.
The authors indicated no potential conflicts of interest.
We thank Carolin Augsburg, Jutta Meltzer, Elisabeth Kurzknabe, Gertraud Fröschl, and Angela Schumich for their excellent technical assistance. We also thank all the hospital personnel and clinicians participating in the Acute Myeloid Leukemia-Berlin-Frankfurt-Muenster study group for providing bone marrow samples and clinical data. This article is dedicated to Dr Guenther Schellong in honor of his 80th birthday.
Supported by grants from the Leukemia and Lymphoma Society (LLS 6180-02), Competence Network Pediatric Oncology and Haematology (KPOH, 01GI9963/2), and the José Carreras Foundation (SP 01/04). Authors' disclosures of potential conflicts of interest and author contributions are found at the end of this article.
1. Langebrake C, Brinkmann I, Teigler-Schlegel A, et al: Immunophenotypic differences between diagnosis and relapse in childhood AML: Implications for MRD monitoring. Cytometry B Clin Cytom 63:1-9, 2005[Medline] 2. Baer Maria R, Stewart Carleton C, Dodge Richard K, et al: High frequency of immunophenotype changes in acute myeloid leukemia at relapse: Implications for residual disease detection (Cancer and Leukemia Group B Study 8361). Blood 97:3574-3580, 2001 3. Macedo A, San Miguel JF, Vidriales MB, et al: Phenotypic changes in acute myeloid leukaemia: Implications in the detection of minimal residual disease. J Clin Pathol 49:15-18, 1996 4. Oelschlagel U, Nowak R, Schaub A, et al: Shift of aberrant antigen expression at relapse or at treatment failure in acute leukemia. Cytometry 42:247-253, 2000[CrossRef][Medline] 5. Coustan-Smith E, Ribeiro RC, Rubnitz JE, et al: Clinical significance of residual disease during treatment in childhood acute myeloid leukaemia. Br J Haematol 123:243-252, 2003[CrossRef][Medline] 6. Sievers EL, Lange BJ, Alonzo TA, et al: Immunophenotypic evidence of leukemia after induction therapy predicts relapse: Results from a prospective Children's Cancer Group study of 252 patients with acute myeloid leukemia. Blood 101:3398-3406, 2003 7. Feller N, van der Pol MA, van Stijn A, et al: MRD parameters using immunophenotypic detection methods are highly reliable in predicting survival in acute myeloid leukaemia. Leukemia 18:1380-1390, 2004[CrossRef][Medline] 8. Kern W, Voskova D, Schoch C, et al: Determination of relapse risk based on assessment of minimal residual disease during complete remission by multiparameter flow cytometry in unselected patients with acute myeloid leukemia. Blood 104:3078-3085, 2004 9. San Miguel JF, Vidriales MB, Lopez-Berges C, et al: Early immunophenotypical evaluation of minimal residual disease in acute myeloid leukemia identifies different patient risk groups and may contribute to postinduction treatment stratification. Blood 98:1746-1751, 2001 10. Venditti A, Tamburini A, Buccisano F, et al: Clinical relevance of minimal residual disease detection in adult acute myeloid leukemia. J Hematother Stem Cell Res 11:349-357, 2002[CrossRef][Medline] 11. Reinhardt D, Langebrake C, Creutzig U, et al: Minimal residual disease in acute myeloid leukemia in children: Standardization and evaluation of immunophenotyping in the AML-BFM-98 study. Klin Padiatr 214:179-187, 2002[CrossRef][Medline] 12. Creutzig U, Zimmermann M, Ritter J, et al: Definition of a standard-risk group in children with AML. Br J Haematol 104:630-639, 1999[CrossRef][Medline] 13. Creutzig U, Zimmermann M, Reinhardt D, et al: Early deaths and treatment-related mortality in children undergoing therapy for acute myeloid leukemia: Analysis of the multicenter clinical trials AML-BFM 93 and AML-BFM 98. J Clin Oncol 22:4384-4393, 2004 14. Bennett JM, Catovsky D, Daniel MT, et al: Proposals for the classification of the acute leukaemias: French-American-British (FAB) co-operative group. Br J Haematol 33:451-458, 1976[Medline] 15. Bennett JM, Catovsky D, Daniel MT, et al: Proposed revised criteria for the classification of acute myeloid leukemia: A report of the French-American-British Cooperative Group. Ann Intern Med 103:620-625, 1985 16. Bennett JM, Catovsky D, Daniel MT, et al: Criteria for the diagnosis of acute leukemia of megakaryocyte lineage (M7): A report of the French-American-British Cooperative Group. Ann Intern Med 103:460-462, 1985 17. Cheson BD, Cassileth PA, Head DR, et al: Report of the National Cancer Institute-sponsored workshop on definitions of diagnosis and response in acute myeloid leukemia. J Clin Oncol 8:813-819, 1990[Abstract] 18. Macedo A, Orfao A, Vidriales MB, et al: Characterization of aberrant phenotypes in acute myeloblastic leukemia. Ann Hematol 70:189-194, 1995[Medline] 19. Reading CL, Estey EH, Huh YO, et al: Expression of unusual immunophenotype combinations in acute myelogenous leukemia. Blood 81:3083-3090, 1993 20. Terstappen LW, Safford M, Konemann S, et al: Flow cytometric characterization of acute myeloid leukemia: Part II. Phenotypic heterogeneity at diagnosis [corrected and republished article originally printed in Leukemia 5:757-767, 1991]. Leukemia 6:70-80, 1992[Medline] 21. Kern W, Voskova D, Schoch C, et al: Prognostic impact of minimal residual disease as determined by multiparameter flow cytometry in patients with acute myeloid leukemia. Onkologie 28:260, 2005[CrossRef][Medline] Submitted December 21, 2005; accepted May 17, 2006.
This article has been cited by other articles:
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
|||||||||||
|
|||||||||||
|
|
Copyright © 2006 by the American Society of Clinical Oncology, Online ISSN: 1527-7755. Print ISSN: 0732-183X
|
TOP
ABSTRACT
INTRODUCTION
2 statistic was used. Probabilities of survival were estimated using the Kaplan-Meier method, with SEs according to Greenwood, and were compared with the log-rank test. Cumulative incidence functions of relapse and death in continuous complete remission were constructed by the method of Kalbfleisch and Prentice. Computations were performed using SAS version 6.12 (SAS Institute Inc, Cary, NC). 
3 time points positive; SE, standard error probability of event-free survival).
