|
|
| |||||||||||||||||||||||||||||||
Background and Methods A characteristic of acute myeloid leukemia is the frequent ability of the leukemic cells to sustain their own proliferation in vitro. To determine the clinical importance of this property, we measured the uptake of tritiated thymidine by leukemic cells in serum-free and cytokine-free cultures as a means of determining the rate of spontaneous proliferation in 114 patients with newly diagnosed acute myeloid leukemia. Proliferation was then classified according to three quantitative levels of activity and related to overall survival and to treatment outcome (the response to treatment, the actuarial probability of relapse, and disease-free survival) in 91 patients who were treated with chemotherapy to induce remission.
Results Of the 114 patients, 37 had low, 39 had intermediate, and 38 had high levels of proliferation. The probability of survival at three years was 36 percent among patients with low levels of proliferative activity and 3 percent among those with high levels (P<0.001). Among the patients treated with chemotherapy, those with low rates of proliferative activity had a 68 percent rate of complete remission and a 49 percent probability of remaining free of relapse, whereas those with high rates of proliferative activity had only a 39 percent rate of complete remission (P = 0.04) and an 11 percent probability of remaining in complete remission (P = 0.009). The probability of disease-free survival at three years among the patients in complete remission after chemotherapy was 49 percent among those with low rates of proliferative activity and 9 percent among those with high rates (P = 0.004). Accordingly, patients with low rates of proliferative activity also had a significantly higher rate of overall survival (44 percent vs. 4 percent; P = 0.002). Patients whose cells had intermediate levels of proliferation in vitro had intermediate rates of survival, relapse, and disease-free survival.
Conclusions The capacity of leukemic blasts for autonomous proliferation is associated with highly aggressive acute myeloid leukemia. .
In a substantial proportion of patients with AML, autonomous proliferation has been observed as a characteristic feature of blast-cell proliferation in culture, and spontaneous growth in vitro generally results from autocrine stimulation1,2. In these patients, progenitors of AML cells form colonies or proliferate in culture without any growth-factor supplement or serum. The leukemic cells may produce one or more of the principal hematopoietic growth factors (granulocyte-macrophage colony-stimulating factor, granulocyte colony-stimulating factor, and macrophage colony-stimulating factor)1,2,3,4 and growth-enhancing cytokines (interleukin-1, tumor necrosis factor, and interleukin-6)5,6,7,8,9 and secrete these factors as biologically active peptides at effective concentrations. The concomitant expression of functional receptors for several of these molecules on the leukemic cells10,11,12 establishes complete autocrine circuits of hematopoietic growth-factor stimulation. In experimental models of leukemia, autocrine mechanisms substantially alter the kinetics of growth and confer a growth advantage13,14,15. On the basis of these observations, it has been argued that autonomous mechanisms of growth may contribute to the clinical biology of leukemia.
We present an analysis of the spontaneous proliferative activity of AML in relation to disease and the outcome of therapy in a series of 114 patients. The proliferative activity of AML cells from all the patients was assessed by measuring the rate of DNA synthesis and examined under uniform and nonstimulatory conditions; that is, serum or cytokines were not added to the cultures.
Methods
Patients
The study included all patients with previously untreated spontaneous AML who were seen at our institution between 1981 and 1990 from whom we obtained sufficient cells for culture. A total of 114 patients were eligible for study, and their disease was classified according to the system devised by the French-American-British (FAB) Committee16. The clinical, hematologic, and cytogenetic characteristics of the patients are summarized in Table 1. The mean age of the patients was 54 years (range, 16 to 82). All patients gave informed consent.
|
Patients treated with high-dose chemotherapy were assessed for signs of a complete remission after one or two courses of induction therapy. A complete remission was defined as the presence of fewer than 5 percent blast cells (including monocytoid cells) and fewer than 10 percent blast cells plus promyelocytes in a cellular bone marrow smear, peripheral-blood counts of more than 1500 granulocytes per cubic millimeter and more than 100,000 platelets per cubic millimeter, and the absence of circulating blasts and extramedullary leukemic-cell infiltration.
Follow-up information was gathered for all patients until July 1, 1992. At that time 25 patients were still alive (median follow-up, 41 months; maximal follow-up, 7 years).
Collection and Culture of Bone Marrow
Cells were obtained from bone marrow aspirated from the iliac bones and collected in heparin-treated bottles containing Hanks' balanced-salt solution. The cells were isolated and cultured as described22. T lymphocytes and adherent cells were always removed from the suspension of AML blasts. Mononuclear cells were recovered from the interface after centrifugation in the presence of Ficoll and metrizoic acid (Isopaque) and then depleted of T lymphocytes by E-rosetting22. Cells adhering to the plastic culture dish (monocytes) were removed after one hour of incubation at 37 °C in 5 percent carbon dioxide (plating density, 4 x 106 cells per milliliter in serum-free medium). Finally, the purified AML blast cells22 were plated (density, 0.2 x 106 per milliliter) into 96-well microtiter plates with round-bottomed wells (Greiner, Nurtingen, Germany) (100 microl per well). Cells were cultured in serum-free and cytokine-free medium for three or seven days to determine maximal DNA synthesis. Under these conditions, the majority of the inoculated cells remained blastic in appearance, with some monocytic maturation and rare signs of granulocytic maturation23. After two days (for cells cultured for three days) and six days (for cells cultured for seven days), 0.1 micro Ci of tritiated thymidine (2 Ci per millimole) (Amersham, Buckinghamshire, United Kingdom) was added to each well, and the cultures were allowed to incubate overnight. The cells were then harvested on nitrocellulose filters with a cell harvester, and radioactivity was measured22. All experiments were performed in triplicate, and data were expressed as log-transformed mean values (expressed as disintegrations per minute [dpm]). There was a strong positive correlation between the values obtained on day 3 and those obtained on day 7, so that values for both days produced identical effects on regression analysis. To prevent problems with collinearity in multivariate analysis, we combined the two values. The maximal value recorded in cells from each patient (obtained on either day 3 or day 7) was entered into the analysis. The values were obtained on day 3 in samples from 70 patients and on day 7 in samples from 44 patients. Irradiated (30 Gy) AML cells were always cultured in parallel with cultures of cells from each patient to assess the background level of incorporation of tritiated thymidine into the blasts.
Statistical Analysis
The probability of survival and disease-free survival was calculated according to the actuarial method of Kaplan and Meier24. Overall survival was determined for all 114 patients, regardless of whether they received chemotherapy or whether they had a complete remission. In addition, to assess the prognostic value of the level of spontaneous proliferation as an indicator of the outcome of treatment, the rate of complete remission, the actuarial probability of a relapse after a complete remission, and disease-free survival after a complete remission were determined for the patients who received induction chemotherapy. Logistic-regression analysis was used to determine the prognostic factors for complete remission. Cox regression analysis25 was applied to determine the prognostic factors for overall survival, the incidence of relapse after a complete remission, and disease-free survival. The independent variables considered in the multivariate analyses were age, FAB subtype, presence of specific chromosomal abnormalities, sex, percentage of leukemic blasts in the bone marrow at diagnosis, white-cell count at diagnosis, and the rate of proliferation in vitro. The log-transformed value of the rate of proliferation was the main variable in the regression analysis for tests of trend. The range of values for thymidine uptake was also divided into three classes in which values of less than 7.50 dpm were deemed indicative of a low rate of proliferative activity, values ranging from 7.50 to 29.59 dpm an intermediate rate of activity, and values of 29.60 dpm or more a high rate of activity. There were 37 patients in the lowest group, 39 in the intermediate group, and 38 in the highest group. Spearman's rank correlation, Pearson's chi-square test, and the Kruskal-Wallis test were used to assess the association between spontaneous proliferation and the characteristics of the patients and their treatment regimens.
Results
Autonomous Activation of DNA Synthesis of AML Blasts in Culture
The distribution of spontaneous proliferative activity is shown in Figure 1. In most patients, the rate of thymidine uptake (mean, 34 dpm) exceeded the (irradiated) background value (mean, 2.6 dpm). Cells from patients with stage M4 or M5 AML had significantly (P<0.001) greater levels of activity than cells from patients with stage M1 or M2 AML (Table 2). Neither the absence of cytogenetic abnormalities nor the presence of one, two, or multiple abnormalities correlated with the rate of thymidine uptake. There was also no apparent correlation between the rate of autonomous proliferative activity and age, sex, white-cell count, or the percentage of leukemic blasts in the marrow at diagnosis. When the patients were classified according to the chemotherapeutic regimens they received, they were found to be evenly distributed among the groups with low, intermediate, and high rates of proliferative activity (data not shown).
|
|
All 114 patients, irrespective of whether they received induction chemotherapy, were included in the survival analysis (Figure 2). All relapses and deaths, except one, occurred within three years of the start of treatment. Eighty-nine patients had died and 25 were still alive (16 of whom had been in continuous complete remission) at the time of the last follow-up visit, with a median survival of 41 months. A high rate of autonomous proliferation (P<0.001) and older age (P<0.001) were independent unfavorable prognostic variables for overall survival. Patients with high rates of proliferation had a 3 percent probability of survival at three years, as compared with a 36 percent probability of survival among patients with low rates of proliferation (P<0.001). None of the other patient characteristics analyzed -- sex, percentage of leukemic blasts in the marrow at diagnosis, white-cell count, FAB subtype, and the presence of cytogenetic abnormalities -- were predictive of survival in this series.
|
|
|
|
|
|
Leukemogenesis is thought to be a process in which successive transformation events alter the abilities of the hematopoietic progenitor cells to proliferate, differentiate, and survive. In recent years, several mechanisms of altered growth causing autonomous proliferation of the cells have been implicated in the origin and progression of myeloid leukemia in a variety of experimental systems13,14,15,26,27. The proliferation of AML cells in humans generally is due to stimulation by exogenous hematopoietic growth factors. Nevertheless, it has become apparent that in the majority of cases, the leukemic cells have an appreciable level of autonomous proliferative activity in vitro. The cells may be able to produce and release hematopoietic growth factors and growth-modifying cytokines. As a consequence, the autocrine and paracrine cascades among AML cells often cause various levels of stimulation1,2,3,4,5,6.
In the present study, the capacities of AML blasts for autonomous activation of DNA synthesis in culture were assessed for their prognostic importance. Cell specimens were prospectively included in the analysis irrespective of whether chemotherapy was used and independently of the different treatment protocols used in adult and elderly patients. The studies were carried out under reproducible conditions: we used culture medium that was not supplemented with specific growth factors or serum, and we cultured the AML blast cells after adherent monocytic cells and E-rosetting lymphocytes had been removed to eliminate the stimulating influences of admixed cells. The autonomous activation of DNA synthesis of AML blasts in vitro was evident in most specimens. The heterogeneous degrees of autocrine capacity are in accordance with the scattered values of autonomous thymidine uptake that were observed in the study (Figure 1). Cells from patients with stage M4 or M5 AML were associated with higher levels of autonomous proliferative activity than were cells from patients with stage M1 or M2 AML (Table 2). There were too few patients with stage M3 or M6 AML for a separate analysis. The results indicate that monocytic leukemias are particularly active in releasing autocrine growth factors, a characteristic that these leukemias appear to share with normal monocytes28. A high rate of proliferation was a particularly unfavorable prognostic factor of AML in terms of overall survival. Among patients who received induction chemotherapy, those whose cells had high rates of autonomous proliferation in culture had a poor response to chemotherapy, resulting in a significantly reduced probability of complete remission (Table 3). In addition, the patients with high rates of proliferation who had a complete remission tended to relapse with greater frequency. A high rate of autonomous proliferative activity was the single most unfavorable prognostic factor for relapse after remission. These data suggest that the ability of neoplastic cells to undergo active autonomous proliferation characterizes biologically aggressive AML.
The patients in this study were treated with different chemotherapeutic regimens, depending on their age and the year in which they were treated. No statistically significant differences in the rates of complete remission and survival between the different protocols were apparent after correction for the effect of age, although the power of these tests was low because of the small numbers of patients in the groups. The inclusion of treatment variables in the model did not influence the effect of proliferative activity on the outcome of the analysis. Notably, the different chemotherapeutic treatments were found to be evenly distributed among patients in the various prognostic groups.
Why the patients with AML who had high rates of proliferation would have unfavorable clinical features is a complex question. The patients with more active cases of autonomously proliferating AML may provide a local stimulatory advantage in the tissue that favors the expansion of the initial deposits of AML cells. Thus, actively growing AML cells may enter complete remission less frequently because of the rapid regrowth of leukemia. It remains to be seen whether autonomously proliferating AML cells have enhanced growth rates in vivo and therefore lead more rapidly to relapse. Alternatively, AML cells with a higher base-line level of growth stimulation may have a greater propensity for (additional) mutations and may become resistant to chemotherapy more frequently during the development of leukemia. On the basis of current models, one may assume that tumors that are curable have a relatively low rate of mutation leading to drug resistance29. If the frequency of mutations during each cell division is constant, then differences in the rate of proliferative activity may strongly influence the probability of spontaneous drug resistance.
In recent years, various factors have been assessed for their prognostic importance in human AML. The majority of prognostic factors have been general hematologic or clinical variables and not specific to leukemia cells. In the present study we included many patients of relatively advanced age. Our results confirm the strong prognostic influence of age on the rate of complete remission and survival, which was independent of the effect of autonomous proliferation. The limited capacities of older patients to tolerate intensive chemotherapy are most likely responsible for the reduced rates of complete remission and survival probabilities. Differences in the nature of leukemia may also contribute to the negative outcome of chemotherapy in older patients21,30,31. However, increased age did not correlate with a greater probability of a relapse after a complete remission, nor with a higher rate of autonomous proliferation.
Cytogenetic abnormalities provide a useful index of prognosis in AML and reflect intrinsic characteristics of the disease. The capacity of AML cells for autonomous proliferation may provide a new and potentially important biologic prognosticator.
Supported by the Netherlands Cancer Society Queen Wilhelmina Fund. Dr. Santini was supported by a fellowship from the European Organization for Radiotherapy and Chemotherapy.
We are indebted to Drs. W. Sizoo, A. Hagenbeek, and M.B. van 't Veer for their contribution to the clinical studies, to Prof. A. Hagemeijer-Hausman (Department of Genetics, Erasmus University, Rotterdam) for cytogenetic analysis, and to Mrs. L. Broeders for laboratory assistance.
Source Information
From Dr. Daniel den Hoed Cancer Center (B.L., W.L.J.v.P., I.P.T., V.S.) and Erasmus University (B.L., R.D.), Rotterdam, the Netherlands.
Address reprint requests to Dr. Lowenberg at Dr. Daniel den Hoed Cancer Center, Groene Hilledijk 301, 3075 EA Rotterdam, the Netherlands.
References
secreted by acute myelogenous leukemia blasts in autocrine and paracrine leukemia growth control. J Clin Invest 1989;84:451-457.
| |||||||||||||||||||||||||||||||
This article has been cited by other articles:
HOME | SUBSCRIBE | SEARCH | CURRENT ISSUE | PAST ISSUES | COLLECTIONS | PRIVACY | TERMS OF USE | HELP | beta.nejm.org Comments and questions? Please contact us. The New England Journal of Medicine is owned, published, and copyrighted © 2009 Massachusetts Medical Society. All rights reserved. |
Previous
29.60 dpm; n = 38). Samples with an intermediate rate of proliferation exceed the background level of stimulation of irradiated cells in more than 90 percent of cases; samples with a high rate of proliferation exceed the control values in more than 99 percent of cases.
