SEARCH · PubMed Health

Results for “Bootstrap resampling”

Explore indexed PubMed citations for clinical trials, systematic reviews and public health research. Read source abstracts and follow each citation to its original PubMed record.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 199 records · Page 11Linked to original sources

PubMed2006

Time to discontinuation of atypical versus typical antipsychotics in the naturalistic treatment of schizophrenia.

BACKGROUND: There is an ongoing debate over whether atypical antipsychotics are more effective than typical antipsychotics in the treatment of schizophrenia. This naturalistic study compares atypical and typical antipsychotics on time to all-cause medication discontinuation, a recognized index of medication effectiveness in the treatment of schizophrenia. METHODS: We used data from a large, 3-year, observational, non-randomized, multisite study of schizophrenia, conducted in the U.S. between 7/1997 and 9/2003. Patients who were initiated on oral atypical antipsychotics (clozapine, olanzapine, risperidone, quetiapine, or ziprasidone) or oral typical antipsychotics (low, medium, or high potency) were compared on time to all-cause medication discontinuation for 1 year following initiation. Treatment group comparisons were based on treatment episodes using 3 statistical approaches (Kaplan-Meier survival analysis, Cox Proportional Hazards regression model, and propensity score-adjusted bootstrap resampling methods). To further assess the robustness of the findings, sensitivity analyses were performed, including the use of (a) only 1 medication episode for each patient, the one with which the patient was treated first, and (b) all medication episodes, including those simultaneously initiated on more than 1 antipsychotic. RESULTS: Mean time to all-cause medication discontinuation was longer on atypical (N = 1132, 256.3 days) compared to typical antipsychotics (N = 534, 197.2 days; p < .01), and longer on atypicals compared to typicals of high potency (N = 320, 187.5 days; p < .01), medium potency (N = 140, 213.5 days; p < .01), and low potency (N = 74, 208.7 days; p < .01). Among the atypicals, only clozapine, olanzapine, and risperidone had significantly longer time to all-cause medication discontinuation compared to typicals, regardless of potency level, and compared to haloperidol with prophylactic anticholinergic treatment. When compared to perphenazine, a medium-potency typical antipsychotic, only clozapine and olanzapine had a consistently and significantly longer time to all-cause medication discontinuation. Results were confirmed by sensitivity analyses. CONCLUSION: In the usual care of schizophrenia patients, time to medication discontinuation for any cause appears significantly longer for atypical than typical antipsychotics regardless of the typical antipsychotic potency level. Findings were primarily driven by clozapine and olanzapine, and to a lesser extent by risperidone. Furthermore, only clozapine and olanzapine therapy showed consistently and significantly longer treatment duration compared to perphenazine, a medium-potency typical antipsychotic.

Adult↗

PubMed2005

Assessing the reliability of ordered categorical scales using kappa-type statistics.

Methods for the analysis of reliability of ordered categorical scales are discussed, focussing on the limitation of the single summary-weighted kappa coefficients. A symmetric matrix of kappa-type coefficients is suggested as an alternative. The method is proposed as being suitable for ordinal scale where there is no underlying continuum. Their application is illustrated using two data sets from reliability studies. If, instead, distances between categories can be specified, a weighted mean of the matrix terms can be used as a summary measure. This is equal to a weighted kappa coefficient with squared weights, provided distances between adjacent categories are equal. When a study design corresponds to a one-way random effects model, estimates of precision of kappa-type coefficient, including the coefficients described here, can be obtained using the delta-method, bootstrap resampling by subjects or jack-knifing by subjects. In the case of interobserver reliability studies, where there may be systematic differences between observers, the investigator may wish to generalise to a population of observers and subjects. In this case, jack-knifing by observer and subject is suggested. Empirical comparisons are made between standard error estimates based on the delta-method, on jack-knifing by subjects and a two-way jack-knife by subjects and observers. The results suggest that standard errors based on the delta-method or jack-knifing by subject alone may be overly precise.

Biomedical Research↗

PubMed2006

Checking linearity of non-parametric component in partially linear models with an application in systemic inflammatory response syndrome study.

Two tests are proposed for checking the linearity of nonparametric function in partially linear models. The first one is based on a Crámer-von Mises statistic. This test can detect the local alternative converging to the null at the parametric rate 1/square root n. A bootstrap resample technique is provided to calculate the critical values. The second one is constructed in a penalized spline framework along with linear mixed-effects (LME) modeling. This is an extension likelihood ratio test for testing zero variance of random effects in LME models. Simulation experiments are conducted to explore the numerical performance of two tests. It is observed that two tests have good level properties, and the first test has a substantially superior power property over the second test in a variety of cases. A real data set is analysed with the proposed tests.

Humans↗

PubMed2005

The combined percentage of Gleason patterns 4 and 5 is the best predictor of cancer progression after radical prostatectomy.

PURPOSE: Clinical outcome is variable in prostate cancer patients treated with radical prostatectomy. The Gleason histologic grade of prostatic adenocarcinoma is one of the strongest predictors of biologic aggressiveness of prostate cancer. We evaluated the significance of the relative proportion of high-grade cancer (Gleason patterns 4 and/or 5) in predicting cancer progression in prostate cancer patients treated with radical prostatectomy. PATIENTS AND METHODS: Radical prostatectomy specimens from 364 consecutive prostate cancer patients were totally embedded and whole mounted. Various clinical and pathologic characteristics were analyzed. All pathologic data, including Gleason grading variables, were collected prospectively. RESULTS: A multiple-factor analysis was performed that included the combined percentage of Gleason patterns 4 and 5, Gleason score, tumor stage, surgical margin status, preoperative prostate-specific antigen (PSA), extraprostatic extension, and total tumor volume. Using Cox regression analysis with bootstrap resampling for predictor selection, we identified the combined percentage of Gleason patterns 4 and 5 (P < .0001) and total tumor volume (P = .009) as significant predictors of PSA recurrence. CONCLUSION: The combined percentage of Gleason patterns 4 and 5 is one of the most powerful predictors of patient outcome, and appears superior to conventional Gleason score in identifying patients at increased risk of disease progression. On the basis of our results, we recommend that the combined percentage of Gleason patterns 4 and 5 be evaluated in radical prostatectomy specimens. The amount of high-grade cancer in a prostatectomy specimen should be taken into account in therapeutic decision making and assessment of patient prognosis.

Adenocarcinoma↗

PubMed1996

PBPK modeling/Monte Carlo simulation of methylene chloride kinetic changes in mice in relation to age and acute, subchronic, and chronic inhalation exposure.

During a 2-year chronic inhalation study on methylene chloride (2000 or 0 ppm; 6 hr/day, 5 days/week), gas-uptake pharmacokinetic studies and tissue partition coefficient determinations were conducted on female B6C3F1, mice after 1 day, 1 month, 1 year, and 2 years of exposure. Using physiologically based pharmacokinetic (PBPK) modeling coupled with Monte Carlo simulation and bootstrap resampling for data analyses, a significant induction in the mixed function oxidase (MFO) rate constant (Vmaxc) was observed at the 1-day and 1-month exposure points when compared to concurrent control mice while decreases in glutathione S-transferase (GST) rate constant (Kfc) were observed in the 1-day and 1-month exposed mice. Within exposure groups, the apparent Vmaxc maintained significant increases in the 1-month and 2-year control groups. Although the same initial increase exists in the exposed group, the 2-year Vmaxc is significantly smaller than the 1-month group (p < 0.001). Within group differences in median Kfc values show a significant decrease in both 1-month and 2-year groups among control and exposed mice (p < 0.001). Although no changes in methylene chloride solubility as a result of prior exposure were observed in blood, muscle, liver, or lung, a marginal decrease in the fat:air partition coefficient was found in the exposed mice at p = 0.053. Age related solubility differences were found in muscle:air, liver:air, lung:air, and fat:air partition coefficients at p < 0.001, while the solubility of methylene chloride in blood was not affected by age (p = 0.461). As a result of this study, we conclude that age and prior exposure to methylene chloride can produce notable changes in disposition and metabolism and may represent important factors in the interpretation for toxicologic data and its application to risk assessment.

Administration, Inhalation↗

PubMed2001

Criticality of predictors in multiple regression.

A new method is proposed for comparing all predictors in a multiple regression model. This method generates a measure of predictor criticality, which is distinct from and has several advantages over traditional indices of predictor importance. Using the bootstrapping (resampling with replacement) procedure, a large number of samples are obtained from a given data set which contains one response variable and p predictors. For each sample, all 2p-1 subset regression models are fitted and the best subset model is selected. Thus, the (multinomial) distribution of the probability that each of the 2p-1 subsets is 'the best' model for the data set is obtained. A predictor's criticality is defined as a function of the probabilities associated with the models that include the predictor. That is, a predictor which is included in a large number of probable models is critical to the identification of the best-fitting regression model and, therefore, to the prediction of the response variable. The procedure can be applied to fixed and random regression models and can use any measure of goodness of fit (e.g., adjusted R2, Cp, AIC) for identifying the best model. Several criticality measures can be defined by using different combinations of the probabilities of the best-fitting models, and asymptotic confidence intervals for each variable's criticality can be derived. The procedure is illustrated with several examples.

Humans↗

PubMed2001

Phylogeny of Steinernema travassos, 1927 (Cephalobina: Steinernematidae) inferred from ribosomal DNA sequences and morphological characters.

Entomopathogenic nematodes in Steinernema, together with their symbiont bacteria Xenorhabdus, are obligate and lethal parasites of insects that can provide effective biological control of some important lepidopteran, dipteran, and coleopteran pests of commercial crops. Phylogenetic relationships among 21 Steinernema species were estimated using 28S ribosomal DNA (rDNA) sequences and morphological characters. Sequences of the rDNA internal transcribed spacers were obtained to provide additional molecular characters to resolve relationships among Steinernema carpocapsae, Steinernema scapterisci, Steinernema siamkavai, and Steinernema monticolum. Four equally parsimonious trees resulted from combined analysis of 28S sequences and 22 morphological characters. Clades inferred from analyses of molecular sequences and combined datasets were primarily reliably supported as assessed by bootstrap resampling, whereas those inferred from morphological data alone were not. Although partially consistent with some traditional expectations and previous phylogenetic studies, the hypotheses inferred from molecular evidence, and those from combined analysis of morphological and molecular data, provide a new and comprehensive framework for evaluating character evolution of steinernematids. Interpretation of morphological character evolution on 6 trees inferred from sequence data and combined evidence suggests that many structural features of these nematodes are highly homoplastic, and that some structures previously used to hypothesize relationships represent ancestral character states.

Animals↗

PubMed1997

Evolution of the nuclear receptor superfamily: early diversification from an ancestral orphan receptor.

From a database containing the published nuclear hormone receptor (NR) sequences I constructed an alignment of the C, D and E domains of these molecules. Using this alignment, I have performed tree reconstruction using both distance matrix and parsimony analysis. The robustness of each branch was estimated using bootstrap resampling methods. The trees constructed by these two methods gave congruent topologies. From these analyses I defined six NR subfamilies: (i) a large one clustering thyroid hormone receptors (TRs), retinoic acid receptors (RARs), peroxisome proliferator-activated receptors (PPARs), vitamin D receptors (VDRs) and ecdysone receptors (EcRs) as well as numerous orphan receptors such as RORs or Rev-erbs; (ii) one containing retinoid X receptors (RXRs) together with COUP, HNF4, tailless, TR2 and TR4 orphan receptors; (iii) one containing steroid receptors; (iv) one containing the NGFIB orphan receptors; (v) one containing FTZ-F1 orphan receptors; and finally (vi) one containing to date only one gene, the GCNF1 orphan receptor. The relationships between the six subfamilies are not known except for subfamilies I and IV which appear to be related. Interestingly, most of the liganded receptors appear to be derived when compared with orphan receptors. This suggests that the ligand-binding ability of NRs has been gained by orphan receptors during the course of evolution to give rise to the presently known receptors. The distribution into six subfamilies correlates with the known abilities of the various NRs to bind to DNA as homo- or heterodimers. For example, receptors heterodimerizing efficiently with RXR belong to the first or the fourth subfamilies. I suggest that the ability to heterodimerize evolved once, just before the separation of subfamilies I and IV and that the first NR was able to bind to DNA as a homodimer. From the study of NR sequences existing in vertebrates, arthropods and nematodes, I define two major steps of NR diversification: one that took place very early, probably during the multicellularization event leading to all the metazoan phyla, and a second occurring later on, corresponding to the advent of vertebrates. Finally, I show that in vertebrate species the various groups of NRs accumulated mutations at very different rates.

Animals↗

PubMed2026

Development and internal validation of a six-gene prognostic model based on galactose metabolism for overall survival in lung adenocarcinoma.

BACKGROUND: Lung cancer remains a leading cause of cancer incidence and mortality globally. Metabolic reprogramming promotes tumor progression and shapes an immunosuppressive tumor microenvironment. Galactose metabolism is involved in multiple malignancies, but its prognostic value in lung adenocarcinoma (LUAD) remains unclear. This study aimed to develop and internally validate a galactose metabolism-related multigene prognostic model for LUAD. METHODS: A retrospective prognostic model development and internal validation study was performed using RNA sequencing (RNA-seq) and clinical data from 585 LUAD patients in The Cancer Genome Atlas (TCGA). Differential expression, functional enrichment, univariate and multivariate Cox regression were applied to construct a prognostic gene signature. Internal validation was performed using bootstrap resampling. Model performance was evaluated by time-dependent receiver operating characteristic (ROC), C-index, calibration, and Kaplan-Meier analysis. Associations between the model and immune infiltration, immunotherapy responsiveness, and tumor stemness were also analyzed. RESULTS: A six-gene prognostic model (GALT, GANC, PGM1, GALM, B4GALT1, PGM2) was developed. The model showed good discrimination with 1-, 3-, and 5-year area under the curve (AUC) values of 0.719, 0.693, and 0.684, respectively. The low-risk group exhibited significantly longer survival, increased antitumor immune infiltration (CD8+ T cells, M1 macrophages, activated CD4+ memory T cells), higher expression of T cell proliferation-related genes, lower immune checkpoint expression, better predicted immunotherapy response, and lower tumor stemness compared with the high-risk group. CONCLUSIONS: We developed and internally validated a six-gene prognostic model for LUAD based on galactose metabolism. The model shows moderate prognostic performance and is associated with antitumor immunity and tumor stemness. It may be used for prognostic risk stratification and to guide personalized immunotherapy in LUAD.

Galactose metabolism↗

PubMed2026

Coagulation activation is associated with genomic-instability-related features in TP53-mutated AML and MDS: routine laboratory patterns beyond classical disseminated intravascular coagulation.

BACKGROUND: Disseminated intravascular coagulation (DIC) is a serious complication of acute myeloid leukemia (AML) associated with poor prognosis. In TP53-mutated AML and myelodysplastic syndrome (MDS), however, the classical ISTH criteria rarely identify overt DIC, although bleeding and thrombotic complications are well documented in acute leukaemia. We hypothesized that these patients exhibit a lower-grade, subclinical coagulation activation that is associated with the underlying genomic-instability-related features of TP53-mutant disease. METHODS: We retrospectively analyzed 107 consecutive patients with TP53-mutated AML (n = 52) or high-risk MDS (MDS, n = 55), median age 65 years, diagnosed and initially evaluated at our centre between 2018 and 2025. Seven routine coagulation markers and 46 co-mutated genes were evaluated for associations with overall survival (OS) using univariate and multivariable Cox regression, continuous dose-response modeling, and unsupervised k-means clustering. Internal validity was assessed by 1000 bootstrap resamples. RESULTS: Overt DIC according to ISTH criteria was rare (15%). Subclinical activation was common: 50% of patients had a D-dimer ≥1 μg/mL, 41% a fibrinogen ≥4 g/L, and 29% an INR ≥1.2. In univariate analysis, D-dimer, fibrinogen, INR, prothrombin time, and activated partial thromboplastin time were each associated with OS (HR 1.33-1.38 per SD; all p < 0.05). Complex karyotype correlated with higher D-dimer (median 1.39 vs. 0.60 μg/mL, p = 0.022) and fibrinogen (3.91 vs. 2.53 g/L, p = 0.007), while TP53 variant allele frequency (VAF) showed modest positive correlations with D-dimer (ρ = 0.21), INR (ρ = 0.27), and PT (ρ = 0.27; all p < 0.05). Clustering identified three coagulation phenotypes: Silent (51%), Thrombo-inflammatory (31%), and Consumption-like (18%), showing a graded but statistically non-significant gradient in molecular features and a stepwise decline in median OS (14, 10 and 8 months; log-rank p = 0.041). After adjustment for complex karyotype, TP53 VAF, and favorable co-mutation count, the Consumption-like phenotype was associated with a non-significant increased risk (HR 1.83, 95% CI 0.92-3.65, p = 0.084), whereas favorable co-mutation pathways remained independently protective (HR 0.56, 95% CI 0.35-0.90, p = 0.016). CONCLUSION: In TP53-mutated AML/MDS, coagulation activation intensity is associated with the degree of genomic instability. The three phenotypes may add biological resolution beyond classical DIC and cytogenetic risk groups, but represent laboratory patterns rather than validated bleeding or thrombosis prediction tools. However, after accounting for genomic features, phenotypes were not independent predictors of outcome, with complex karyotype, TP53 VAF, and favorable co-mutation count driving prognosis. Because treatment intensity and other clinical confounders were not available, these survival associations are hypothesis-generating. Coagulation profiling remains inexpensive, widely accessible, and offers a practical window into disease biology that warrants prospective validation.

TP53↗

PubMed2003

Imince: an unrestricted factor-analysis-based program for assessing measurement invariance.

In this article, a Windows program for analyzing measurement invariance in two different populations is described. Factor analysis is a common way of assessing measurement invariance, and restricted factor analysis is now the most popular method. However, applied researchers have usually found that the theoretical advantages of restricted factor analysis do not always apply in practical situations. For example, when the participant sample is large, as is the case in Internet-based questionnaires, the available software for restricted factor analysis might fail to converge on a solution. Our program is based on unrestricted factor analysis and considers the three parameters that define factor invariance: difficulties, discriminations, and residual variances. The statistical significance of the tests for evaluating invariance is obtained using Bootstrap resampling procedures. A real-life example demonstrates the usefulness of the program.

Adult↗

PubMed2001

Survival analysis with time-varying relative risks: a tree-based approach.

A tree-based method for estimating time-varying effects of baseline patient characteristics on survival is introduced. A Cox-type model for censored survival data is used in which the time-varying relative risks are modelled as piecewise constants. The tree method consists of three steps: 1. Growing the tree, in which a fast algorithm using maximized score statistics is utilized to determine the optimal change points; 2. A pruning algorithm is applied to obtain more parsimonious models; 3. Selection of a final tree, which may be either via bootstrap resampling or based on a measure of explained variation. The piecewise constant model is more suitable for clinical interpretation of the regression parameters than the more continuously time-varying models (spline, loess, etc.) that have been proposed previously.

Algorithms↗

PubMed2001

Model selection and parameterization of the concentration-response functions for population-level effects.

As concentration response functions for chronic population-level effects of pollutant chemicals, three mathematical models were presented and examined for goodness of fit to published toxicological data that estimated the population-level effects of chemicals in terms of the intrinsic rate of population growth (r). Among the examined concentration-r functions, the power function model, that is, r(x) = r(0)[1 - (x/alpha)beta], in which x is the exposure concentration and alpha and beta are parameters, performed with the best fit to each data set. The power function model is characterized by two parameters representing the absolute value of toxicity, alpha, and the curvature of responses, beta. The bootstrap simulation, conducted on the entire data set consisting of all published data that we collected, indicated that the observed variance of beta among actual data sets could be mostly explained by the random error variation generated from the bootstrap resamplings. The generic beta value, determined from the entire data set and expected to denote the best estimate of beta if the variability of beta was completely due to random sampling error, was estimated as 1.84. It was implied that the response of the intrinsic rate of natural increase (r) to chemical exposure was nearly quadratic in many cases.

Animals↗

PubMed2002

Selecting a probability distribution for extreme rainfall series in Malaysia.

This paper discusses the comparative assessment of eight candidate distributions in providing accurate and reliable maximum rainfall estimates for Malaysia. The models considered were the Gamma, Generalised Normal, Generalised Pareto, Generalised Extreme Value, Gumbel, Log Pearson Type III, Pearson Type III and Wakeby. Annual maximum rainfall series for one-hour resolution from a network of seventeen automatic gauging stations located throughout Peninsular Malaysia were selected for this study. The length of rainfall records varies from twenty-three to twenty-eight years. Model parameters were estimated using the L-moment method. The quantitative assessment of the descriptive ability of each model was based on the Probability Plot Correlation Coefficient test combined with root mean squared error, relative root mean squared error and maximum absolute deviation. Bootstrap resampling was employed to investigate the extrapolative ability of each distribution. On the basis of these comparisons, it can be concluded that the GEV distribution is the most appropriate distribution for describing the annual maximum rainfall series in Malaysia.

Environmental Monitoring↗

PubMed1999

Dental insurance and use of dental services.

OBJECTIVES: This article examines socioeconomic differences in insurance for dental services among Canadians aged 15 or older and factors associated with the use of dental services. DATA SOURCE: The data on dental insurance coverage and use of dental services are from the cross-sectional file of Statistics Canada's 1996/97 National Population Health Survey. The sample size of respondents aged 15 or older was 70,884. ANALYTICAL TECHNIQUES: Logistic regression analysis was used to model variables related to dental insurance coverage and to dental visits in the past year. A weighted bootstrap resampling procedure was used to derive variance estimates. MAIN RESULTS: In 1996/97, 53% of the population aged 15 or older reported having dental insurance, and 59% said they had visited a dentist in the past year. But even when they had insurance, individuals with low incomes and low educational attainment had much lower odds of visiting a dentist than those with higher incomes and more education.

Adolescent↗

PubMed2002

Molecular profiling of bladder cancer using cDNA microarrays: defining histogenesis and biological phenotypes.

This study was designed to characterize the expression profiles of nine bladder cancer cell lines (T24, J82, 5637, HT1376, RT4, SCaBER, TCCSUP, UMUC-3, and HT1197) using cDNA microarrays (8976 genes and expressed sequence tags). Novel targets involved in bladder cancer progression of potential clinical relevance were validated by immunohistochemistry using tissue microarrays of primary bladder tumors (n = 193 cases). Hierarchical clustering classified uroepithelial cells based on their histopathogenesis and cell cycle alterations. Keratin 10 and caveolin-1 transcripts were more abundant in tumor cells from squamous and invasive origin. Their combined expression was shown to stratify bladder tumors and define squamous differentiation. To assess the robustness of the clustering analysis, a bootstrap resampling technique was used. This grouped tumor cell lines based on their biological properties, including cell cycle and cell adhesion features. E-cadherin, zyxin, and moesin were identified as genes differentially expressed in these clusters and related to the p53, RB, and INK4A status of the cell lines. Loss of these adhesion molecules was associated with stage and grade in primary tumors (P < 0.05), and moesin expression was also associated with survival (P = 0.01). Deregulation of cell cycle and apoptotic pathways, such as mutations or altered expression of p53, pRB, and INK4A (p16), is necessary for uroepithelial transformation. However, it appears that deregulation of cell adhesion is a common event associated with tumor progression in uroepithelial neoplasms.

Biomarkers, Tumor↗

PubMed2003

Estimation of urban runoff and water quality using remote sensing and artificial intelligence.

Water quality and quantity of runoff are strongly dependent on the landuse and landcover (LULC) criteria. In this study, we developed a more improved parameter estimation procedure for the environmental model using remote sensing (RS) and artificial intelligence (AI) techniques. Landsat TM multi-band (7bands) and Korea Multi-Purpose Satellite (KOMPSAT) panchromatic data were selected for input data processing. We employed two kinds of artificial intelligence techniques, RBF-NN (radial-basis-function neural network) and ANN (artificial neural network), to classify LULC of the study area. A bootstrap resampling method, a statistical technique, was employed to generate the confidence intervals and distribution of the unit load. SWMM was used to simulate the urban runoff and water quality and applied to the study watershed. The condition of urban flow and non-point contaminations was simulated with rainfall-runoff and measured water quality data. The estimated total runoff, peak time, and pollutant generation varied considerably according to the classification accuracy and percentile unit load applied. The proposed procedure would efficiently be applied to water quality and runoff simulation in a rapidly changing urban area.

Artificial Intelligence↗

PubMed2006

Phenotypic quantitative features of patients with acute myeloid leukemia.

The recent WHO classification for acute myeloid leukemias (AML) separates entities by recurrent cytogenetic abnormalities and immunophenotypic features presenting prognostic impact. We have examined the expression of several lineage and maturation linked antigens used in routine immunophenotyping of patients with de novo AML, using a 3-color two-step panel. Cases were diagnosed by peripheral blood counts, bone marrow cytology, cytochemistry, cytogenetics and immunophenotyping (CD2, CD3, CD7, CD19, CD13, CD33, myeloperoxydase -- MPO, CD14, CD15, HLA-DR, CD34, CD56 and CD45). Antigen expression was measured by mean fluorescence intensity (MFI) by flow cytometry (Paint-a-gate software). Thirty five patients were analyzed. Median age: 51 years (15-79). Predominant FAB types were M2 and M4. In 6 cases more than one phenotypically distinct blast subpopulation could be detected. Although our set was small, we tried to analyze the impact of MFI of the examined antigens on the overall survival of the patients. In Cox univariate analysis, age, peripheral leukocytes (WBC) at diagnosis, MFI of CD45, and MPO were significant for worse a survival. In the multivariate analysis only MFI of CD45 and WBC remained in the model (p=0.018 and p=0.014 respectively). After bootstrap resampling, MFI of CD45 entered the model in 69%, WBCin 60%, age in 42% and MFI of MPO in 35% of the sets. Analysis of antigen expression by MFI permitted to detect cases presenting phenotypically distinct blast subpopulations. This may represent a pitfall in studies of minimal residual disease by flow cytometry, as chemotherapy may select one of these subsets.

Acute Disease↗