Sensitivity and Resistance to Therapy

Early clearance of peripheral blasts measured by flow cytometry during the first week of AML induction therapy as a new independent prognostic factor: a GOELAMS study

Abstract

An early appreciation of treatment efficacy could be very useful in acute myeloblastic leukemia (AML), and a prognostic value has been suggested for the morphological assessment of decrease in blasts during induction therapy. More sensitive, multiparametric flow cytometry (FCM) can detect far lower blast counts, allowing for a precise and reliable calculation of blast cell decrease rate (BDR). Such a multiparametric FCM four-colours/single-tube protocol, combining CD11b, CD45-ECD and CD16-PC5, was applied to peripheral blood samples from 130 AML patients, collected daily during induction chemotherapy. Normalized blast cell percentages were used to calculate the relevant decrease slopes. Slope thresholds (<−25, −25 to −15 and >−15), or the time required to reach 90% depletion of the peripheral blast load (<5, 5 or >5 days), was strongly associated with the achievement of complete remission (P<0.0001). Log-rank test and Cox model showed that they also carried high statistical significance (P<0.0001) for disease-free survival. The prognostic value of cytogenetic features, confirmed in this series, was refined by BDR, which allowed to discriminate between good- and poor-risk patients among those with intermediate or normal karyotypes. This simple FCM protocol allows for an accurate prognostic sequential approach adapted to the determination of decrease in peripheral blast cells during induction chemotherapy.

Introduction

In spite of the progress of therapeutic approaches in the past few decades, the prognosis of acute myeloblastic leukemia (AML) remains unsatisfactory.1, 2 Progress of diagnostic methods, that is, morphology, cytogenetics, immunophenotyping and molecular biology has led to the identification of prognosis factors3, 4, 5 and has helped to better stratify patients.6 Nevertheless, it is difficult to properly appreciate the efficacy of induction therapy before the first informative examination of bone marrow smears. Yet, for some of the patients who do not achieve complete remission (CR) after the first cycle of chemotherapy, prognosis has been reported to be poorer.7, 8, 9 In the early 1990s, a pathophysiologic model of the evolution of leukemia under chemotherapy was developed by Browman and Preisler,10 showing that the rate of decrease of morphologically identified medullary blasts during induction therapy could constitute an important prognostic indicator. This hypothesis, although elegant, was, however, not pushed further, mostly because repeated bone marrow aspirations were difficult to implement for practical reasons and because precise counts of bone marrow blast cells are extremely difficult to perform by morphological examination.

Increasing sophistication of the technology of flow cytometry (FCM) has, in recent years, considerably improved the sensitivity of this method, and a combination of two morphological signals (forward and side scatter) and four fluorochromes labelling allows for a theoretical detection sensitivity of 10−5 events.11 Such levels of detection allow to envision the possibility of tracking extremely rare events, such as residual blasts in peripheral blood (PB) samples.12 Indeed, recent data suggest that the detection of minimal residual disease by FCM could be successfully applied in the early times of chemotherapy to bone marrow.13 By extension, the possibility of sequentially evaluating PB blast cells during induction chemotherapy could constitute an interesting alternative.

Here, we propose such a one-step FCM protocol allowing to reliably identify all mature white blood cell (WBC) compartments in whole PB samples, and, by use of a subtractive approach, to isolate a region of interest normally devoid of any cell but occupied by blast cells in leukemic patients. Sequential determination of the percentage of PB blasts in this normally ‘empty space’ allows to monitor blast cell decrease (BDR), a new feature that was correlated to the clinical evolution of 130 AML patients and other classical prognostic factors.

Materials and methods

Patients

One hundred and thirty patients were enroled in this study, in three university hospitals (Bordeaux, Nancy and Dijon) and one cancer centre (Marseille). All those who suffered from non-M3 AML were enroled in one of the GOELAMS's clinical trials, LAM2001 or LAMSA02,14, 15 or treated accordingly, and had provided informed consent. LAM2001 was devised for patients lesser than 60 years of age and compared the efficacy of idarubicin (8 mg/m2/d, 5 days) or daunorubicin (60 mg/m2/d, 3 days) and high-dose cytarabine (200 mg/m2/d, 7 days). On examination of a bone marrow aspiration at day 15, patients received or not a second induction chemotherapy, depending on the persistence of more than 5% blasts or Auer rods. Patients over 60 years of age were entered in the LAMSA02 trial, which used a comparable induction therapy of idarubicin (8 mg/m2/day, 5 days) and cytarabine (100 mg/m2/day, 7 days), associated with 1-(2-chloroethyl)-3-cyclohexyl-1-nitrosourea (CCNU) (one take of 200 mg/m2) and G-CSF from day 8 on.

PB samples were obtained from the catheter and collected on EDTA-K on a daily basis from the first day (D1) of induction chemotherapy, until at least day 7. WBC counts were collected at each time point. Characteristics of the patients are summarized in Table 1. At day D0 of treatment, for all PB samples, WBC ranged between 0.5 and 355 G/L and blast cells between 3.1 and 97.8% of leukocytes.

Table 1 Characteristics of the patients’ series

CR and relapse were appreciated on bone marrow samples by morphological examination according to standard criteria.16

Flow cytometry

A mixture of four monoclonal antibodies was prepared extemporaneously before each test. It comprised the following conjugates: CD14-FITC, CD11b-PE, CD45-ECD and CD16-PC5 (Beckman Coulter, Miami, FL, USA).

At each time point, 50 μl of PB was incubated with the antibody mix for 15 min in the dark after gentle vortexing. The samples were then lysed using Versalyse (Beckman Coulter, Bordeaux, France), Immunoprep (Beckman Coulter, Nancy, France) or BDFACSLysis (BD Biosciences (Franklin Lakes, NJ, USA, Dijon, France, and Marseille, France) and immediately processed in FCM.

Events were collected for 300 s or until at least 100 000 events had been acquired. All list modes were saved for further analysis according to the method described in Figure 1 legend. This gating strategy allowed to positively identify monocytes (CD14+/CD11b+), polymorphonuclears (CD16+/CD11b+), immature granulocytes (CD16−/CD11b+) and lymphocytes (CD45hi).17, 18, 19 A subtractive strategy then allowed to evidence peripheral blast cells.

Figure 1
figure1

(Part A) Gating strategy for the identification of blast cells among peripheral blood leukocytes from M4 patient on D0. (a) Definition of live gate (A) eliminating debris; (b) leukocyte staining by CD45. Colour code is as follows: mature polymorphonuclears in red, monocytes in green, lymphocytes in purple, blasts in blue and myelocytes unstained. (c) Positive identification of monocytes by CD14 staining, definition of gate (B). (d) Positive staining of mature polymorphonuclears by CD16, definition of gate (C). (e) Definition of lymphocytes by bright CD45 labelling and CD11b staining, definition of gate (D). (f) Improvement of definition of lymphocytes by bright CD45 labelling and CD16 gating, definition of gate (E). (g) Visualisation of all subsets within gate (A) by CD11b/CD16 staining. (h) Identification of blast cells within gate (A) by the same CD11b/CD16 staining excluding gates (BE), definition of gate Blasts. In this sample, D0 percentage of blast cells was 5.3%, normalized to 100 as indicated in red. (Part B) Blast cell decrease in the same patient over the first 4 days of treatment. The same dot plots as in (A), panels g and h, are shown. Total cell numbers have been adjusted at 40 000 in each plot. For the first 4 days (D1, D2, D3, D4), the remaining blast cell percentages in gate Blasts are shown in black and their relative values normalized to 100 from D0 are shown in red. D90% was reached by D3.

BDR assessment

For each patient, three parameters were recorded at each time point of PB collection: (i) blast cell percentage among leukocytes, (ii) WBC, (iii) calculated absolute blast cell numbers. Data were tabulated on an Excel (Microsoft Corporation) matrix. To compare results, data normalization was performed by attributing a value of 100 to D0 parameters and subsequently adjusting all data (Figure 1).

BDR was evaluated using blast cell percentages and absolute numbers; yet more significant results were obtained using the former (data not shown). For BDR evaluation, two points were considered: D0 and D90%, the latter being the first day when at least 90% of the initial blast load had disappeared from PB. BDR was evaluated using two modalities: (a) calculation of the blast cell decrease slope by linear regression between D0 and D90% and (b) determination of the time lapse to reach D90%, further dubbed BDR90. An example of slope calculation is given in Figure 2a from the data shown in Figure 1.

Figure 2
figure2

(a) Slope of blast cell decrease (BDR) calculated for the patient shown in Figure 1 using normalized blast percentage data from D0, D1, D2, D3 and D4. The resulting slope is of −32.4. (b). The slope of peripheral BDR strongly correlates with the time lapse needed to reach 90% clearance (BDR90) when the 130 patients are considered. Both BDRslope (c) and BDR90 (d) are significantly associated with CR achievement and risk of relapse.

Slope analysis and BDR90 calculation are compared in Figure 2b. This shows an excellent correlation between these two means of BDR assessment (r=0.913). BDR90 could be easier to use for clinical considerations.

Cytogenetics

Karyotypic information was collected in each centre, reviewed by EL and IL, and finally considered available for 118 patients. Karyotypes, classified into three categories as described elsewhere,20 were of favourable risk (t(8;21) or inv(16)) for 13 patients, unfavourable (complex karyotype or inv(3) or t(6;9) or t(6;11) or t(11;19) or del (5q) or t(9;11) or del(11q) or del(20q) or −5 or −7) for 40, and of intermediate (normal karyotype or del(7q) or del(9q)) or numerical aberrations (−Y or +8 or +11 or +13 or +21) for 65. In this intermediate group, 49 patients had no chromosomal anomaly (Table 1).

Statistical analysis

Descriptive statistics, including proportions, means, standard deviations and medians, were used to analyze parameter distributions within subgroups. Patients’ characteristics and CR rate comparisons were performed using two-sided Fisher's exact test or χ2 test for categorical variables. Mean comparisons were performed using t-test after F-test verification of equal variance and the Mann–Whitney test for median comparisons.

Pearson's correlation coefficient was used to evaluate the correlation between paired values of BDRslope and BDR90. Slopes were compared between patients who failed to reach remission, relapsed or remained in CR, using two-way analysis of variance.

Logistic regression was used to evaluate the relationship between the results of induction (CR or not) and clinico-biological parameters. All variables were simultaneously entered into the model to check their respective contribution to the outcome.

Disease-free survival (DFS) was calculated from the date of CR achievement until first relapse or death in CR, and patients alive in CR were censored at the last known information. Event-free survival (EFS) was calculated from the date of initial treatment until first relapse or death, and patients alive in CR were censored at the last known information. DFS and EFS were estimated by the Kaplan–Meier method21 and compared using the log-rank test22 or log-rank test for linear trend in the case of ordered categorical variables (BDRslope and BDR90).

Thresholds used in statistical analyses were those that maximized the differences in statistical tests and were carefully chosen by examining raw data.

A multiparametric analysis of prognosis was performed using Cox's proportional hazard model.23 All variables were first added and sequentially deleted at P>0.05.

All statistical computations were performed using the Medcalc software (Mariakerke, Belgium).

Results

Patients’ outcome

Of the 130 patients, 111 (85.4%) reached CR, among whom 47 (36.2%) relapsed. CR was obtained with one round of induction chemotherapy for 76% of the patients enroled in the LAM2001 trial. In all 19 (14.6%) patients failed to reach CR. Median follow-up was 347 (max 1250) days at updating. Consequently, three groups could be constituted (Table 1): (1) patients achieving CR without relapse in the course of the follow-up period (n=64), (2) patients relapsing after achieving CR (n=47) and (3) patients failing to achieve CR (n=19). Median CR durations were of 610 and 176 days, respectively, for patients in groups 1 and 2.

No significant difference was observed between these three groups for sex, age, WBC and PB blast percentages. Conversely, BDR was significantly different between the three groups using BDRslope (P<0.0001) or BDR90 (P<0.0001) (Figures 2c and d). Similarly, the frequency of favourable, intermediate and unfavourable karyotypes was significantly different within the three groups (P<0.0001).

Four patients in group 3 died during induction therapy and had a very bad BDRslope (−3.25±3.67) compared with the other 15 patients of group 3 who completed induction (−8.72±4.34).

CR achievement

Among the two clinical (sex, age) and four biological (WBC, PB blast %, karyotype, BDR) parameters summarized in Table 1, in a multivariate logistic regression, only BDR influenced significantly CR achievement (P<0.001). Among 40 patients with unfavourable karyotype, 15 (37.5%) failed to achieve CR compared with only 4/78 (5.1%) with favourable or intermediate karyotype. No significant difference was observed between favourable and intermediate karyotypes for CR achievement.

Using either slope or BDR90, BDR was the other major parameter determining CR achievement. Thus, all patients with BDR90 <6 (n=81) achieved CR, whereas 19/49 (38.8%) patients with BDR90 >5 did not achieve CR (P<0.0001).

Among 118 patients with both karyotype and BDR90 parameters available, 19 did not achieve CR. All patients (n=74) with a BDR90 <6 days achieved CR. The respective values of karyotype and BDR in CR achievement are interesting to analyze in the remaining 44 patients with BDR90 >5 days. Two patients with a favourable karyotype were unable to reach BDR90 before day 6. Fourteen patients with an unfavourable karyotype and BDR90 <6 days achieved CR. Among 26 patients with an unfavourable karyotype and BDR90 >5 days, 15 (57.7%) did not achieve CR. For the 56 patients with either BDR90 >5 days or unfavourable karyotype, the data can be summarized and stratified in three groups. The first group (n=14) (BDR90<6 days and unfavourable karyotype) presents 14 CR and 0 failure, the second group (n=16) (BDR90>5 days and intermediate karyotype) presents 12 CR and 4 failures and the third group (n=26) (BDR90>5 days and unfavourable karyotype) presents 11 CR and 15 failures. Applying the same logistic regression as used earlier to the group with intermediate karyotype, BDR remained the only parameter linked with CR achievement (P<0.05). The combination of BDR and karyotype thus seems to provide a powerful tool to significantly predict CR achievement (P<0.0008).

Survival analysis

Considering either relapse onset or patient's death as events, two EFS curves are shown in Figures 3a and b according to BDRslope or BDR90, respectively. Log-rank comparisons were established according to the threshold values identified for the three groups shown in Figures 2c and d (that is, <−25, between −25 and −15, >−15 for BDRslope and <day 5, =day 5 and > day 5 for BDR90).

Figure 3
figure3

Log-rank comparisons of event-free survival (EFS) according to BDRslope (a, upper line <−25, middle line between −25 and −15, lower line>−15), BDR90 (b, upper line <day 5, middle line=day 5 and lower line >day 5 (P<0.0001)). Log-rank comparisons of disease-free survival (DFS) according to BDRslope (c, upper line <−25, middle line between −25 and −15, lower line>−15) and BDR90 (d, upper line <day 5, middle line=day 5 and lower line > day 5 (P<0.0001)), age (e, upper line <65 years of age, lower line over 65 years of age (P<0.005)), karyotypic risk (f, upper line favourable karyotype, middle line, intermediate and lower line, unfavourable (P<0.0001)).

Once CR is achieved, a major component of AML prognosis is CR duration or DFS analysis. There was no difference in DFS among patients aged over 60 years enroled in LAMSA02 or treated accordingly. Similarly, younger patients enroled in LAM2001 or treated accordingly had similar DFS. Moreover, the discriminative powers of BDRslope and BDR90 were similar whatever the age group (data not shown).

Only three parameters were found to significantly contribute to DFS length in this series: age, karyotype and BDR. DFS curves are not shown for nonsignificant parameters, that is, sex, WBC and PB blast percentage.

Although age did not significantly influence CR achievement, it was found to contribute to DFS duration, 65 years of age being the best statistical threshold (Figure 3a, P<0.005).

Earlier described as a significant parameter for CR achievement, karyotype was also an important parameter for predicting DFS length. Taking into account the three classical karyotypic prognostic factors, three different DFS curves could be derived. Their comparison was highly significant by the log-rank test (P<0.0001) and log-rank for linear trend (P<0.0001) (Figure 3b).

BDR, earlier observed as the major parameter determining CR achievement, is also the major factor for discriminating between different DFS groups. Considering BDRslope, three groups could be created to maximize differences in DFS: with slope <−25, between −25 and −15 and slope >−15. Considering BDR90, three groups could also be created in the same way with BDR90 <5 days, BDR90 =5 days and BDR90 >5 days. Whatever the modality of BDR analysis chosen (slope or BDR90), three groups with clearly different DFS curves could be isolated with a high degree of significance (P<0.0001—log-rank and P<0.0001—log-rank for linear trend) as shown in Figures 3c and d.

The intermediate prognosis karyotype is clinically difficult to use. It is noteworthy that normal karyotypes, although integrated to the intermediate karyotype category, could not be differentiated from other intermediate prognosis karyotypes in terms of DFS (Figure 4a). Yet, BDR assessment allowed to override this discrepancy and observe patients with unfavourable intermediate and favourable BDR, both for normal and for intermediate karyotypes (P<0.0001) (Figures 4b and c).

Figure 4
figure4

(a) Absence of significant difference in disease-free survival (DFS) between intermediate non-normal and intermediate normal karyotypes. However, considering BDR90 with the same values as in Figure 3 BDR, patients with either intermediate (b) and normal (c) karyotypes can be stratified with high significance (P<0.0001).

As for CR achievement analysis, two clinical (sex, age) and four biological (WBC, PB blast %, karyotype, BDR) parameters were tested in a multiparametric analysis using a Cox model. Only two covariables were retained as influencing EFS (karyotype (P<0.05) and BDR (P<0.0001)) and DFS (karyotype (P<0.01) and BDR (P<0.0001)).

Discussion

This study reports on a robust and simple FCM assay allowing to accurately appreciate the percentage of peripheral blast cells in patients with AML during the first days of induction therapy. This tool appears to be reliable, as long as a careful analysis of list modes is performed, with an accurate definition of mature WBC, isolating a ‘negative’ gate containing the blast population. The presence of blast cells in the earliest samples, moreover, provides a good positive control to then track the decreasing blast cell population. Calculation of the decrease rate of peripheral blast percentages after normalization of the data allows to directly appreciate the chemosensitivity of individual patients.

PB BDR assessment is more reliable and easy to interpret than the traditional morphological examination of medullary blast cells, especially when their levels are around the threshold value in aplastic bone marrow.10 Moreover, by being performed on sequential PB samples, BDR appreciation is more accurate than a single examination. This FCM approach is also more objective and systematic than that recently reported using morphological examination of blood smears,24 although it reaches similar conclusions. The great difference between this study and our study is that our initial blast counts were as low as 3%. As shown in the example here, the level of detection can accurately go as low as 10−3 or 10−4 during induction chemotherapy in PB samples, a feature making it applicable also to older patients with low peripheral blastosis. Moreover, it seems statistically impossible to appreciate a blast cell decrease by morphology if the initial blast percentage is <10% (see the classical Rümke table25).

From a practical point of view, this assay makes use of the robust FCM definition of mature leukocytes, which is independent of morphological alterations liable to incur owing to the ongoing infusion of cytotoxic drugs. By relying on a positive identification of normal cells, it levers the subjective interpretation of atypical immunophenotypes exhibited by AML cells, which is moreover patient-specific. Therefore, this assay is applicable whatever the initial immunophenotype of the blasts, with the provision of CD11b+ cases, where the CD16/CD11b scattergram used to track blast cell decrease must be adapted.

This new tool appears to be very useful clinically for an early and accurate appreciation of individual patients’ behaviour during the first stages of chemotherapy, possibly allowing to better adapt to therapeutic strategies. This usefulness is shown by the highly significant discrimination observed when comparing BDR for patients who did or did not reach CR. Therefore, BDR provides a rapid and kinetic approach of individual patients’ chemosensitivity at the very early stage of the first days of induction therapy. Extension of this approach could lead to devise new strategies for the stratification of consolidation therapy.

A second interest of BDR assessment is its strong prognostic value even of remission duration, with highly significant different outcomes based on either BDRslope or BDR90, and early relapses for the patients with the worst BDR. This suggests that an early decrease of the tumoral load may have an impact on the efficacy of further consolidation therapies. Again, this observation could lead to modify the therapeutic management of AML patients, including the decision of allogeneic stem cell transplantation.

A third interesting feature of BDR is its value as an independent prognostic factor, considering BDR measurement using blast percentages. It was proved that WBC decrease rate was not a significant prognosis factor and consequently BDR measurement using absolute blast count was not so good a prognostic factor than BDR measurement using blast percentages (data not shown). Moreover, BDR provides additional information to karyotypic features. Indeed, among the various parameters considered to be of prognostic value in AML, karyotypic investigations have been the focus of many studies, resulting in precise definitions of anomalies of favourable and unfavourable prognosis.25, 26, 27, 28 Outside of these two groups, patients are considered to be of ‘intermediate’ prognosis. Interestingly, blast cell sensitivity to chemotherapy as assessed here brings further prognostic value within this group of patients. As BDR during the first days of chemotherapy is likely to yield results slightly before complete karyotypic investigations are completed, a combined use of these two sets of information could bring a novel and highly useful means for treatment stratification. In a prospective study, it would be of great interest to analyze the prognostic impact of the more recently described molecular markers in comparison with cytogenetics and BDR.

Finally, performed in four different settings with different equipment, our work already suggests that excellent reproducibility can be achieved in various laboratories. By applying this strategy to the samples obtained nevertheless for WBC counts, this one-combination FCM test could rapidly confirm its value.

In conclusion, we propose a simple, robust, specific and clinically useful method allowing to rapidly appreciate chemosensitivity in the early stages of chemotherapy for patients with AML. The strong correlation of blast cell decrease slopes and patients’ outcome, if confirmed in larger scale studies, could make this test a new standard for therapeutic stratification in this disease, even in patients with intermediate cytogenetic features.

References

  1. 1

    Estey E, Dohner H . Acute myeloid leukaemia. Lancet 2006; 368: 1894–1907.

    Article  Google Scholar 

  2. 2

    Krug U, Serve H, Muller-Tidow C, Mesters RM, Steffen B, Buchner T et al. New molecular therapy targets in acute myeloid leukemia. Recent Results Cancer Res 2007; 176: 243–262.

    CAS  Article  Google Scholar 

  3. 3

    Béné MC, Bernier M, Casasnovas RO, Castoldi G, Doekharan D, van der Holt B et al. Acute myeloid leukaemia M0: haematological, immunophenotypic and cytogenetic characteristics and their prognostic significance: an analysis in 241 patients. Br J Haematol 2001; 113: 737–745.

    Article  Google Scholar 

  4. 4

    Casasnovas RO, Campos L, Mugneret F, Charrin C, Bene MC, Garand R et al. Immunophenotypic patterns and cytogenetic anomalies in acute non-lymphoblastic leukemia subtypes: a prospective study of 432 patients. Leukemia 1998; 12: 34–43.

    CAS  Article  Google Scholar 

  5. 5

    Steffen B, Muller-Tidow C, Schwable J, Berdel WE, Serve H . The molecular pathogenesis of acute myeloid leukemia. Crit Rev Oncol Hematol 2005; 56: 195–221.

    Article  Google Scholar 

  6. 6

    Vardiman JW, Harris NL, Brunning RD . The World Health Organization (WHO) classification of the myeloid neoplasms. Blood 2002; 100: 2292–2302.

    CAS  Article  Google Scholar 

  7. 7

    Gajjar A, Ribeiro R, Hancock ML, Rivera GK, Mahmoud H, Sandlund JT et al. Persistence of circulating blasts after one week of multiagent chemotherapy confers a poor prognosis in childhood acute lymphoblastic leukemia. Blood 1995; 86: 1262–1295.

    Google Scholar 

  8. 8

    Wheatley K, Burnett AK, Goldstone AH, Gray RG, Hann IM, Harrison CJ et al. A simple, robust, validated and highly predictive index for the determination of risk-directed from the MRC AML 10 trial. Br J Haematol 1999; 10: 69–79.

    Article  Google Scholar 

  9. 9

    Kern W, Haferlach T, Schoch C, Löffler H, Gassmann W, Heinecke A et al. Early blast clearance by remission induction therapy is a major independent prognostic factor for both achievement of complete remission and long-term outcome in acute myeloid leukemia: data from the German AML Cooperative Group (AMLCG) 1992. Trial. Blood 2003; 101: 64–70.

    CAS  Article  Google Scholar 

  10. 10

    Browman G, Preisler H . Predicting remission outcome in acute non-lymphoblastic leukemia: general principles and their application to residual marrow leukemia. Leuk Res 1992; 16: 511–519.

    CAS  Article  Google Scholar 

  11. 11

    Kern W, Schoch C, Haferlach T, Schnittger S . Monitoring of minimal residual disease in acute myeloid leukemia. Crit Rev Oncol Hematol 2005; 56: 283–309.

    Article  Google Scholar 

  12. 12

    Maurillo L, Buccisano F, Spagnoli A, Del Poeta G, Panetta P, Neri B et al. Monitoring of minimal residual disease in adult acute myeloid leukemia using peripheral blood as an alternative source to bone marrow. Haematologica 2007; 92: 605–611.

    Article  Google Scholar 

  13. 13

    San Miguel JF, Vidriales MB, López-Berges C, Díaz-Mediavilla J, Gutiérrez N, Cañizo 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 2001; 98: 1746–1751.

    CAS  Article  Google Scholar 

  14. 14

    Lioure B, Delaunay J, Blaise D, Milpied N, Guardiola P, Cahn JY et al. Allogeneic stem cell transplantation (SCT) with non myeloablative conditioning regimen (NST) following intensive consolidation may be equivalent to conventional alloSCT and superior to autoSCT for patients over 50 with acute myeloid leukemia (AML) in 1st CR: first results of the AML 2001 trial. Blood (ASH Annual Meeting Abstracts) 2006; 108: 319.

    Google Scholar 

  15. 15

    Pigneux A, Himberlin C, Hunault-Berger M, Witz F, Recher C, Harousseau JL et al. Amulticenter randomized comparison of maintenance treatment with androgens in elderly acute myeloid leukemia after ICL regimen as induction therapy: results of the Goelams-2002 Study. Blood (ASH Annual Meeting Abstracts) 2006; 108: 1983.

    Google Scholar 

  16. 16

    Cheson BD, Bennett JM, Kopecky KJ, Büchner T, Willman CL, Estey EH et al. International Working Group for Diagnosis, Standardization of Response Criteria, Treatment Outcomes, and Reporting Standards for Therapeutic Trials in Acute Myeloid Leukemia. Revised recommendations of the International Working Group for Diagnosis, Standardization of Response Criteria, Treatment Outcomes, and Reporting Standards for Therapeutic Trials in Acute Myeloid Leukemia. J Clin Oncol 2003; 21: 4642–4649.

    Article  Google Scholar 

  17. 17

    Fujimoto H, Sakata T, Hamaguchi Y, Shiga S, Tohyama K, Ichiyama S et al. Flow cytometric method for enumeration and classification of reactive immature granulocyte populations. Cytometry 2000; 42: 371–378.

    CAS  Article  Google Scholar 

  18. 18

    Hubl W, Tlustos L, Erath A, Andert S, Bayer PM . Proposed reference method for peripheral-blood monocyte counting using fluorescence-labelled monoclonal antibodies. Cytometry 1996; 26: 69–74.

    CAS  Article  Google Scholar 

  19. 19

    Lacombe F, Durrieu F, Briais A, Dumain P, Belloc F, Bascans E et al. Flow cytometry CD45 gating for immunophenotyping of acute myeloid leukemia. Leukemia 1997; 11: 1878–1886.

    CAS  Article  Google Scholar 

  20. 20

    Mrozek K, Bloomfield CD . Chromosome aberrations, gene mutations and expression changes, and prognosis in adult acute myeloid leukemia. Hematology Am Soc Hematol Educ Program 2006; 2006: 169–177.

    Article  Google Scholar 

  21. 21

    Peto J . The calculation and interpretation of survival curves. In: Buyse ME, Staquet MJ, Sylvester RJ (eds). Cancer Clinical Trials: Methods and Practice. Oxford University Press: Oxford, England, 1984, pp 361–380.

    Google Scholar 

  22. 22

    Breslow N . Comparison of survival curves. In: Buyse ME, Staquet MJ, Sylvester RJ (eds). Cancer Clinical Trials: Methods and Practice. Oxford University Press: Oxford England, 1984, pp 381–406.

    Google Scholar 

  23. 23

    Cox DR . Regression models and life-tables. J R Soc Stat [B] 1972; 34: 187–220.

    Google Scholar 

  24. 24

    Elliott MA, Litzow MR, Letendre LL, Wolf RC, Hanson CA, Tefferi A et al. Early peripheral blood blast clearance during induction chemotherapy for acute myeloid leukemia predicts superior relapse-free survival. Blood 2007; 110: 4172–4174.

    CAS  Article  Google Scholar 

  25. 25

    Rumke CL . Variability of results in differential counts on blood smears. Triangle, The Sandoz Journal of Medical Science. 1960; 4: 156.

    Google Scholar 

  26. 26

    Slovak ML, Kopecky KJ, Cassileth PA, Harrington DH, Theil KS, Mohamed A et al. Karyotypic analysis predicts outcome of preremission and postremission therapy in adult acute myeloid leukemia: a Southwest oncology Group/Eastern Cooperative Oncology Group study. Blood 2000; 96: 4075–4083.

    CAS  Google Scholar 

  27. 27

    Grimwade D, Walker H, Harrison G, Oliver F, Chatters S, Harrison CJ et al. Medical Research Council Adult Leukemia Working Party. The predictive value of hierarchical cytogenetic classification in older adults with acute myeloid leukemia (AML): analysis of 1065 patients entered into the United Kingdom Medical Research Council AML11 trial. Blood 2001; 98: 1312–1320.

    CAS  Article  Google Scholar 

  28. 28

    Marcucci G, Mrozek K, Ruppert AS, Archer KJ, Pettenati MJ, Heerema NA et al. Abnormal cytogenetics at date of morphologic complete remission predicts short overall and disease-free survival, and higher relapse rate in adult acute myeloid leukemia: results from cancer and leukemia group B study 8461. J Clin Oncol 2004; 22: 2410–2418.

    Article  Google Scholar 

Download references

Acknowledgements

We are grateful to Professors Norbert Ifrah (Angers), Jean Yves Cahn (Grenoble) and Francis Belloc (Bordeaux) for fruitful discussion, Alex Briais (Bordeaux) and Marie Robert (Bordeaux) for helpful technical assistance and to cytogeneticists, respectively Drs Marina Lafage Pochitaloff and Joëlle Mozziconacci (Marseille), Francine Mugneret (Dijon) and Marie Josée Grégoire (Vandoeuvre les Nancy). This work benefited grants from the Goelams group, Fondation de France and Laurette Fugain association.

Author information

Affiliations

Authors

Corresponding author

Correspondence to M C Béné.

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Lacombe, F., Arnoulet, C., Maynadié, M. et al. Early clearance of peripheral blasts measured by flow cytometry during the first week of AML induction therapy as a new independent prognostic factor: a GOELAMS study. Leukemia 23, 350–357 (2009). https://doi.org/10.1038/leu.2008.296

Download citation

Keywords

  • AML
  • multiparameter flow cytometry
  • prognosis

Further reading

Search

Quick links