Comparative analysis of genome-scale, base-resolution DNA methylation profiles across 580 animal species.

Methylation of cytosines is a prototypic epigenetic modification of the DNA. It has been implicated in various regulatory mechanisms across the animal kingdom and particularly in vertebrates. We mapped DNA methylation in 580 animal species (535 vertebrates, 45 invertebrates), resulting in 2443 genome-scale DNA methylation profiles of multiple organs. Bioinformatic analysis of this large dataset quantified the association of DNA methylation with the underlying genomic DNA sequence throughout vertebrate evolution. We observed a broadly conserved link with two major transitions-once in the first vertebrates and again with the emergence of reptiles. Cross-species comparisons focusing on individual organs supported a deeply conserved association of DNA methylation with tissue type, and cross-mapping analysis of DNA methylation at gene promoters revealed evolutionary changes for orthologous genes. In summary, this study establishes a large resource of vertebrate and invertebrate DNA methylomes, it showcases the power of reference-free epigenome analysis in species for which no reference genomes are available, and it contributes an epigenetic perspective to the study of vertebrate evolution.

Detection and Monitoring of Lineage-Specific Chimerism by Digital Droplet PCR-Based Testing of Deletion/Insertion Polymorphisms.

Analysis of specific leukocyte subsets for post-transplantation monitoring of chimerism provides greater sensitivity and clinical informativeness on dynamic changes in donor- and recipient-derived cells. Limitations of the most commonly used approach to chimerism testing relying on PCR-based analysis of microsatellite markers prompted us to assess the applicability of digital droplet (dd) PCR amplification of deletion/insertion polymorphisms (DIPs) for lineage-specific chimerism testing in the related stem cell transplantation setting, where the identification of informative markers facilitating the discrimination between donor-derived and recipient-derived cells can be challenging. We analyzed 100 genetically related patient-donor pairs by ddPCR analysis using commercially available DIP kits including large sets of polymorphic markers. At least 1 informative marker was identified in all related pairs analyzed, and 2 or more discriminating markers were detected in the majority (82%) of instances. The achievable detection limit is dependent on the number of cells available for analysis and was as low as 0.1% in the presence of ≥20,000 leukocytes available for DNA extraction. Moreover, the reproducibility and accuracy of quantitative chimerism analysis compared favorably to highly optimized microsatellite assays. Thus, the use of ddPCR-based analysis of DIP markers is an attractive approach to lineage-specific monitoring of chimerism in any allogeneic transplantation setting.

Gliomasphere marker combinatorics: multidimensional flow cytometry detects CD44+/CD133+/ITGA6+/CD36+ signature.

Glioblastoma is the most dangerous brain cancer. One reason for glioblastoma's aggressiveness are glioblastoma stem-like cells. To target them, a number of markers have been proposed (CD133, CD44, CD15, A2B5, CD36, CXCR4, IL6R, L1CAM, and ITGA6). A comprehensive study of co-expression patterns of them has, however, not been performed so far. Here, we mapped the multidimensional co-expression profile of these stemness-associated molecules. Gliomaspheres - an established model of glioblastoma stem-like cells - were used. Seven different gliomasphere systems were subjected to multicolor flow cytometry measuring the nine markers CD133, CD44, CD15, A2B5, CD36, CXCR4, IL6R, L1CAM, and ITGA6 all simultaneously based on a novel 9-marker multicolor panel developed for this study. The viSNE dimensionality reduction algorithm was applied for analysis. All gliomaspheres were found to express at least five different glioblastoma stem-like cell markers. Multi-dimensional analysis showed that all studied gliomaspheres consistently harbored a cell population positive for the molecular signature CD44+/CD133+/ITGA6+/CD36+. Glioblastoma patients with an enrichment of this combination had a significantly worse survival outcome when analyzing the two largest available The Cancer Genome Atlas datasets (MIT/Harvard Affymetrix: P = 0.0015, University of North Carolina Agilent: P = 0.0322). In sum, we detected a previously unknown marker combination - demonstrating feasibility, usefulness, and importance of high-dimensional gliomasphere marker combinatorics.

Flow-cytometric minimal residual disease monitoring in blood predicts relapse risk in pediatric B-cell precursor acute lymphoblastic leukemia in trial AIEOP-BFM-ALL 2000.

BACKGROUND: Flow-cytometric monitoring of minimal residual disease (MRD) in bone marrow (BM) during induction of pediatric patients with acute lymphoblastic leukemia (ALL) is widely used for outcome prognostication and treatment stratification. Utilizing peripheral blood (PB) instead of BM might be favorable, but data on its usefulness are scarce. PROCEDURE: We investigated 1303 PB samples (days 0, 8, 15, 33, and 52) and 285 BMs (day 15) from 288 pediatric ALL patients treated in trial AIEOP-BFM ALL 2000. MRD was assessed by four-color flow cytometry and evaluated as relative, absolute, and kinetic result. RESULTS: In B-ALL only, PB measures from early time points correlated with relapse incidence (CIR). Best separation occurred at threshold <1 blast/µL at day 8 (5-year CIR 0.02 ± 0.02 vs 0.12 ± 0.03; P = 0.044). Patients with highest relapse risk were not distinguishable, but PB-MRD at days 33 and 52 correlated with prednisone response and postinduction BM-MRD by PCR (P < 0.001). Kinetic assessment did not convey any advantage. In multivariate analysis including day 15 BM-MRD, PB-MRD measures lost statistical power. CONCLUSIONS: In summary, PB-MRD in pediatric B-ALL correlates with outcome and risk parameters, but its prognostic significance is not strong enough to substitute for BM assessment in AIEOP-BFM trials. It might, however, be valuable in treatment environments not using multifaceted risk stratification with other MRD measures.

Differential DNA Methylation Analysis without a Reference Genome.

Genome-wide DNA methylation mapping uncovers epigenetic changes associated with animal development, environmental adaptation, and species evolution. To address the lack of high-throughput methods for DNA methylation analysis in non-model organisms, we developed an integrated approach for studying DNA methylation differences independent of a reference genome. Experimentally, our method relies on an optimized 96-well protocol for reduced representation bisulfite sequencing (RRBS), which we have validated in nine species (human, mouse, rat, cow, dog, chicken, carp, sea bass, and zebrafish). Bioinformatically, we developed the RefFreeDMA software to deduce ad hoc genomes directly from RRBS reads and to pinpoint differentially methylated regions between samples or groups of individuals (http://RefFreeDMA.computational-epigenetics.org). The identified regions are interpreted using motif enrichment analysis and/or cross-mapping to annotated genomes. We validated our method by reference-free analysis of cell-type-specific DNA methylation in the blood of human, cow, and carp. In summary, we present a cost-effective method for epigenome analysis in ecology and evolution, which enables epigenome-wide association studies in natural populations and species without a reference genome.

Human adenovirus-specific γ/δ and CD8+ T cells generated by T-cell receptor transfection to treat adenovirus infection after allogeneic stem cell transplantation.

Human adenovirus infection is life threatening after allogeneic haematopoietic stem cell transplantation (HSCT). Immunotherapy with donor-derived adenovirus-specific T cells is promising; however, 20% of all donors lack adenovirus-specific T cells. To overcome this, we transfected α/ß T cells with mRNA encoding a T-cell receptor (TCR) specific for the HLA-A*0101-restricted peptide LTDLGQNLLY from the adenovirus hexon protein. Furthermore, since allo-reactive endogenous TCR of donor T lymphocytes would induce graft-versus-host disease (GvHD) in a mismatched patient, we transferred the TCR into γ/δ T cells, which are not allo-reactive. TCR-transfected γ/δ T cells secreted low quantities of cytokines after antigen-specific stimulation, which were increased dramatically after co-transfection of CD8α-encoding mRNA. In direct comparison with TCR-transfected α/ß T cells, TCR-CD8α-co-transfected γ/δ T cells produced more tumor necrosis factor (TNF), and lysed peptide-loaded target cells as efficiently. Most importantly, TCR-transfected α/ß T cells and TCR-CD8α-co-transfected γ/δ T cells efficiently lysed adenovirus-infected target cells. We show here, for the first time, that not only α/ß T cells but also γ/δ T cells can be equipped with an adenovirus specificity by TCR-RNA electroporation. Thus, our strategy offers a new means for the immunotherapy of adenovirus infection after allogeneic HSCT.

Robust multi-parameter single-platform quantification of myeloid and B-lymphoid CD34 progenitor cells in all clinical CD34 cell sources and in thawed PBSC.

As B-lymphoid progenitor cells do not give rise to in vitro colony formation and are unlikely to support myeloid engraftment, we validated a five-color extension of the single platform Stem Cell Enumeration (SCE) kit, to routinely quantify myeloid and B-lymphoid progenitor cells. Fresh samples (n > 20 each) of granulocyte colony stimulating factor mobilized blood (peripheral blood (PB)), cord blood (CB), bone marrow (BM), and apheresis products (APs) were stained in TruCOUNT™ tubes and the results were compared with those from the two-color CD45/CD34 reagent combination and the three-color SCE kit. To address repeatability, 10 samples from one AP were prepared by four technicians. Aliquots (n = 15) of four frozen AP were analyzed after thawing. Excellent correlations were observed between the three kits (R(2) > 0.99), for the quantification of white blood cells and total CD34. The extended kit showed considerable amounts of B-lymphoid progenitors in all CD34 sources (0-20% of all CD34 in PB, AP, and CB; 3-90% in BM). Very similar results were obtained when the same sample was prepared by different technicians. After thawing of frozen AP, the recovery of viable cells varied depending on the freezing medium employed, but the results from the different quantification methods were identical. Most non-viable cells were clearly identified with 7 Aminoactinomycin D (7AAD) but an additional gate in the forward scatter/side scatter was necessary to address dead cells negative for 7AAD. The extended SCE kit allows rapid and exact quantification of viable B-lymphoid and myeloid CD34(+) cells in all cell sources and in thawed stem cell harvests, and may thus improve the correlation between CD34 number and engraftment kinetics.

Novel single-platform multiparameter FCM analysis of apoptosis: Significant differences between wash and no-wash procedure.

FCM is a generally accepted tool to analyze apoptosis. Unfortunately, the cell preparation of all commercial kits available includes cell washing known to cause cell loss which is most likely to affect apoptotic cells in particular. To address this, we developed a seven-color single-platform no-wash analysis technique and compared the results with those from an analogous procedure including cell washing. A five-color mAb cocktail was employed to address target cells by surface labeling, Yo-PRO-1® and DAPI were used to discriminate apoptotic and necrotic from viable cells. Cells were quantified on the basis of internal-standard fluorescent beads. Jurkat cells ACC 282 treated with camptothecin were employed to establish the staining procedure, which was then applied to blood cells collected by extracorporeal apheresis and treated with UV irradiation. Data evaluation showed that although each method by itself was highly reproducible (R(2) = 0.973), the numbers of apoptotic cells detected with the no-wash procedure were significantly higher than those obtained after cell washing (P = 6.6 E(-5), Wilcoxon Test). In addition, the observed differences increased with higher cell numbers (Bland and Altmann). We conclude that the described test is a feasible and reliable tool for apoptosis measurement and it provides results that are definitely closer to the truth than those obtained from kits that require cell washing.

Death induction by CD99 ligation in TEL/AML1-positive acute lymphoblastic leukemia and normal B cell precursors.

Our study was performed to examine the role of CD99 in normal and leukemia BCPs. CD99 is strongly expressed by certain pediatric cancers including BCP-ALL. Modulation of the antigen in ETs and T cells induces apoptosis, hence implicating CD99 as a potential target for anti-cancer therapy. However, nothing is known about these aspects in BCPs. We investigated BCP-ALL cases and normal BCP cells from pediatric BM for CD99 protein and RNA expression as well as for effects of CD99 modulation by mAb. Immunophenotypes, recovery, apoptosis, and aggregation were assessed. Flow cytometry, light microscopy, and qRT-PCR were used in our experiments. An association of CD99 expression levels with the cytogenetic background of pediatric BCP-ALLs was found. Highest CD99 levels were observed in hyperdiploid, followed by TEL/AML1 and random karyotype leukemias. CD99 ligation moderately induced cell death only in TEL/AML1 cases. Stroma cell contact mitigated this effect. Very immature normal BCPs were the most sensitive to CD99-mediated death induction. Type I CD99 mRNA was the main isoform in ALLs and was expressed differentially during BCP maturation. Our data suggest that clinical targeting of CD99 may be effective in BCP-ALL-bearing TEL/AML1 but also may elicit negative effects on normal B-lymphopoiesis. We consider our results as an indication that CD99 may play a physiologic role in the clonal deletion processes necessary for B-lymphoid selection.

CD20 up-regulation in pediatric B-cell precursor acute lymphoblastic leukemia during induction treatment: setting the stage for anti-CD20 directed immunotherapy.

CD20 is expressed in approximately one- half of pediatric acute lymphoblastic leukemia (ALL) cases with B-cell precursor (BCP) origin. We observed that it is occasionally up-regulated during treatment. To understand the impact of this on the potential effectiveness of anti-CD20 immunotherapy, we studied 237 CD10(+) pediatric BCP-ALL patients with Berlin-Frankfurt-Munster (BFM)-type therapy. We analyzed CD20 expression changes from diagnosis to end-induction, focusing on sample pairs with more than or equal to 0.1% residual leukemic blasts, and assessed complement-induced cytotoxicity by CD20-targeting with rituximab in vitro. CD20-positivity significantly increased from 45% in initial samples to 81% at end-induction (day 15, 71%). The levels of expression also increased; 52% of cases at end-induction had at least 90% CD20(pos) leukemic cells, as opposed to 5% at diagnosis (day 15, 20%). CD20 up-regulation was frequent in high-risk patients, patients with high minimal residual disease at end-induction, and patients who suffered later from relapse, but not in TEL/AML1 cases. Notably, up-regulation occurred in viable cells sustaining chemotherapy. In vitro, CD20 up-regulation significantly enhanced rituximab cytotoxicity and could be elicited on prednisolone incubation. In conclusion, CD20 up-regulation is frequently induced in BCP-ALL during induction, and this translates into an acquired state of higher sensitivity to rituximab. This study was registered at http://www.clinicaltrials.gov as #NCT00430118.

Expression of the putatively regulatory T-cell marker FOXP3 by CD4(+)CD25+ T cells after pediatric hematopoietic stem cell transplantation.

FOXP3 has been proposed to be critical for the regulatory function of CD4(+)CD25+ T cells and it has been reported that its expression correlates with protection from graft-versus-host-disease (GvHD) after allogeneic hematopoietic stem cell transplantation (HSCT). Here, by monitoring 28 pediatric HSCT recipients, we found that the levels of FOXP3-mRNA expression in highly enriched CD4(+)CD25+ cells were identical to those in healthy controls irrespective of GvHD status. Moreover, FOXP3-mRNA was abundant in recently in vitro stimulated CD4(+)CD25+ cells that lacked regulatory function. Together these findings suggest that FOXP3-mRNA expression primarily reflects CD4(+)CD25+ cell frequency rather than defining the regulatory potential of CD4(+)CD25+ T cells and GvHD risk after HSCT.

Evaluation of normal and neoplastic human mast cells for expression of CD172a (SIRPalpha), CD47, and SHP-1.

Signal regulatory proteins (SIRPs) and tyrosine phosphatases have recently been implicated in the control of receptor tyrosine kinase (RTK)-dependent cell growth. In systemic mastocytosis (SM), neoplastic cells are driven by the RTK KIT, which is mutated at codon 816 in most patients. We examined expression of SIRPalpha, SIRPalpha ligand CD47, and Src homology 2 domain-containing protein tyrosine phosphatase-1 (SHP-1), a tyrosine phosphatase-type, negative regulator of KIT-dependent signaling, in normal human lung mast cells (HLMC) and neoplastic MC obtained from nine patients with SM. As assessed by multicolor flow cytometry, normal LMC expressed SIRPalpha, CD47, and SHP-1. In patients with SM, MC also reacted with antibodies against SIRPalpha and CD47. By contrast, the levels of SHP-1 were low or undetectable in MC in most cases. Corresponding data were obtained from mRNA analysis. In fact, whereas SIRPalpha mRNA and CD47 mRNA were detected in all samples, the levels of SHP-1 mRNA varied among donors. To demonstrate adhesive functions for SIRPalpha and CD47 on neoplastic MC, an adhesion assay was applied using the MC leukemia cell line HMC-1, which was found to bind to immobilized extracellular domains of SIRPalpha1 (SIRPalpha1ex) and CD47 (CD47ex), and binding of these cells to CD47ex was inhibited by the CD172 antibody SE5A5. In summary, our data show that MC express functional SIRPalpha and CD47 in SM, whereas expression of SHP-1 varies among donors and is low compared with LMC. It is hypothesized that CD172 and CD47 contribute to MC clustering and that the "lack" of SHP-1 in MC may facilitate KIT-dependent signaling in a subgroup of patients.

CD25 indicates the neoplastic phenotype of mast cells: a novel immunohistochemical marker for the diagnosis of systemic mastocytosis (SM) in routinely processed bone marrow biopsy specimens.

The diagnosis of systemic mastocytosis (SM) is based primarily on the histologic and immunohistochemical evaluation of a bone marrow trephine biopsy specimen. Although mast cell (MC) specific antigens like tryptase and chymase are detectable in routinely processed tissue, no immunohistochemical markers that can be used to discriminate between normal and neoplastic MCs are yet available. We have investigated the diagnostic value of an antibody against CD25 for the immunohistochemical detection of MCs in bone marrow sections in 73 patients with SM and 75 control cases (reactive marrow, n = 54; myelogenous neoplasms, n = 21) and correlated the results with the presence of c-kit mutations. While MCs in almost all patients with SM (72 of 73) expressed CD25, none of the control samples contained CD25-positive MCs. Irrespective of the SM subtype, most of neoplastic MCs expressed CD25. In 3 patients with advanced MC disease, pure populations of neoplastic MCs were obtained and found to express CD25 mRNA by RT-PCR analysis. In addition, all patients with CD25-positive MCs contained c-kit mutations, while all control cases exhibited wild type c-kit. CD25 therefore appears to be a reliable immunohistochemical marker for the discrimination of neoplastic from normal/reactive MCs, with potential as a diagnostic tool in SM.

Identification of heme oxygenase-1 as a novel BCR/ABL-dependent survival factor in chronic myeloid leukemia.

Chronic myeloid leukemia (CML) is a stem cell disease in which BCR/ABL promotes the survival of leukemic cells. Heme oxygenase-1 (HO-1) is an inducible stress protein that catalyzes the degradation of heme and has recently been implicated in the regulation of growth and survival of various neoplastic cells. In the present study, we analyzed the expression and role of HO-1 in CML cells. As assessed by Northern and Western blot analysis as well as immunostaining, primary CML cells were found to express HO-1 mRNA and the HO-1 protein in a constitutive manner. Exposure of these cells to the BCR/ABL tyrosine kinase inhibitor STI571 resulted in decreased expression of HO-1 mRNA and protein. In addition, BCR/ABL was found to up-regulate HO-1 promoter activity, mRNA levels, and protein levels in Ba/F3 cells. To investigate the role of HO-1 for survival of primary CML cells, the HO-1 inducer hemin was used. Hemin-induced expression of HO-1 was found to protect CML cells from STI571-induced cell death. In addition, inhibition of HO-1 by zinc-(II)-deuteroporphyrin-IX-2,4-bisethyleneglycol resulted in a substantial decrease of cell viability. Furthermore, overexpression of HO-1 in the CML-derived cell line K562 was found to counteract STI571-induced apoptosis. Together, our data identify HO-1 as a novel BCR/ABL-driven survival molecule and potential target in leukemic cells in patients with CML. The pathogenetic and clinical implications of this observation remain to be elucidated.

Comparison of p53 mutational status with mRNA and protein expression in a panel of 24 human breast carcinoma cell lines.

We analyzed the p53 mutational status, mRNA and protein expression in 24 human breast carcinoma cell lines. Following measurement of their DNA content with flow cytometry, we ascertained the copy numbers of the centromere of chromosome 17 (cen17) and p53 with fluorescence in situ hybridization (FISH). A functional yeast assay (FASAY) was used to screen for inactivating mutations. Positive results were subsequently verified by DNA sequencing. Finally, we assessed the mRNA expression with a competitive reverse transcription-polymerase chain reaction (RT-PCR) assay and the protein expression with immunocytochemical staining, western blot, and quantitative flow cytometry. The DNA content of the cell lines ranged from 0.85 to 2.58. Nine cell lines had concordant copy numbers (between two and four) of p53 and cen17, whereas 12 had more, and three less cen17 than p53 copies. The FASAY was successful in all but one cell line and revealed the presence of mutated alleles in 16 of them, 13 cell lines expressed only the mutated, and three both the mutated and the wild-type alleles. The mutations were comprised of 11 missense, two nonsense, and three frameshift mutations. Immunocytochemical staining, western blot and quantitative flow cytometry yielded comparable p53 protein expression results. However, both the mRNA and the protein expression levels varied considerably in the different cell lines and no consistent pattern with regard to the respective p53 mutational status became evident. The results obtained in these breast carcinoma cell lines indicate that no clear-cut linear relationship exists between the p53 mutational status and the extent of its respective mRNA and protein expression. Therefore, direct DNA analyses and functional assays remain the only methods for the reliable detection of p53 mutations.

Response of Ewing tumor cells to forced and activated p53 expression.

The EWS-FLI1 transcription factor is consistently expressed in 85% of Ewing tumors (EFT). In heterologous cells, EWS-FLI1 induces p53-dependent cell cycle arrest or apoptosis. It has been speculated that the p53 tumor suppressor pathway may be generally compromised in EFT despite only rare p53 mutations. In order to test for functional integrity of this pathway, we have investigated a series of EFT cell lines that differ from each other with respect to their endogenous p53 and INK4A gene status for their response to ectopic p53 expression and to stimulation of endogenous p53 activity by X-ray treatment. Significant interindividual and intratumoral variations in the apoptotic propensity of EFT cell lines to transient expression of ectopic p53 were observed, which was independent of the level of p53 expression. In cell lines with a low apoptotic incidence, apoptosis was delayed and the surviving fraction showed a prolonged growth arrest. Complete resistance to p53-induced apoptosis in two cell lines established from the same patient was associated with a high BCL2/BAX ratio and low levels of APAF1. Sensitivity to X-rays showed a trend towards a higher apoptotic rate in wild-type (wt) p53 expressing than in p53 mutant cells. However, one wt p53-expressing EFT cell line was completely refractory to irradiation-stimulated cell death despite high apoptotic responsiveness to ectopic p53. No difference in Ser15 phosphorylation and the transcriptional activation of p53 targets was observed in wt p53 EFT cell lines irrespective of the induction of cell death or growth arrest. All together, our results demonstrate that despite significant variability in the outcome, cell death or cell cycle arrest, the p53 downstream pathway and the DNA damage signaling pathway are functionally intact in EFT.

C5a stimulates production of plasminogen activator inhibitor-1 in human mast cells and basophils.

We have recently shown that resting human mast cells (MCs) produce tissue-type plasminogen activator (t-PA) without simultaneously expressing plasminogen activator inhibitor 1 (PAI-1). In the present study we have identified the anaphylatoxin rhC5a as a potent inducer of PAI-1 expression in human MCs and basophils. In primary human skin MCs and primary blood basophils, exposure to rhC5a was followed by an increase from undetectable to significant levels of PAI-1. In addition, rhC5a induced a concentration- and time-dependent increase in PAI-1 antigen in the MC line HMC-1 and the basophil cell line KU-812 and increased the expression of PAI-1 mRNA in HMC-1. In conditioned media of HMC-1 treated with rhC5a, active PAI-1 could be detected. A simultaneous loss of t-PA activity in conditioned media from the same cells indicated that rhC5a-induced PAI-1 was capable of inhibiting the enzymatic activity of coproduced t-PA. Correspondingly, the levels of t-PA-PAI-1 complexes increased in rhC5a-treated cells. When HMC-1 cells were incubated with pertussis toxin or anti-C5a receptor antibodies, the effect of rhC5a on PAI-1 production was completely abolished. Treatment of C5a with plasmin resulted in loss of its ability to induce PAI-1 production in MCs. Considering the suggested role for MCs and components of the complement system in the development of cardiovascular diseases, we hypothesize that MCs, by producing t-PA in a resting state and by expressing PAI-1 when activated by C5a, might participate in the modulation of the balance between proteases and protease inhibitors regulating tissue injury and repair in such disease processes.

Prognostic significance and modalities of flow cytometric minimal residual disease detection in childhood acute lymphoblastic leukemia.

Detection of minimal residual disease (MRD) in acute lymphoblastic leukemia (ALL) predicts outcome. Previous studies were invariably based on relative quantification and did not investigate sample-inherent parameters that influence test accuracy, which makes comparisons and clinical conclusions cumbersome. Hence, we conducted a prospective, population-based MRD study in 108 sequentially recruited children with ALL uniformly treated with the ALL-Berlin-Frankfurt-Münster (ALL-BFM) 95 protocol in Austria (median follow-up of 40 months). Using sensitive, limited antibody panel flow cytometry applicable to 97% of patients, we investigated 329 bone marrow samples from 4 treatment time points. MRD was quantified by blast percentages among nucleated cells (NCs) and by absolute counts (per microliter). Covariables such as NC count, normal B cells, and an estimate of the test sensitivity were also recorded. Presence and distinct levels of MRD correlated with a high probability of early relapse at each of the time points studied. Sequential monitoring at day 33 and week 12 was most useful for predicting outcome independently from clinical risk groups: patients with persistent disease (> or =1 blast/microL) had a 100% probability of relapse, compared to 6% in all others. Absolute MRD quantification was more appropriate than relative, due to considerable variations in total NC counts between samples. Regeneration of normal immature B cells after periods of rest from treatment limited the test sensitivity. In conclusion, MRD detection by flow cytometry is a strong and independent outcome indicator in childhood ALL. Standardization regarding absolute quantification on the basis of NCs and assessment during periods of continuous treatment promise to increase the accuracy, simplicity, and cost efficiency of the approach.