Increased incidence of spontaneous apoptosis in the bone marrow of hyperdiploid childhood acute lymphoblastic leukemia

Experimental Hematology 30 (2002) 333–339

Increased incidence of spontaneous apoptosis in the bone marrow of hyperdiploid childhood acute lymphoblastic leukemia Yili Zhanga, Jinhua Lub, Janette van den Berghec, and Szu-Hee Leea a Department of Pathology and bNational University Medical Institutes, National University of Singapore, Singapore; cRoyal University Hospital, Saskatoon, Saskatchewan, Canada

(Received 24 September 2001; revised 12 December 2001; accepted 21 December 2001)

Objective. Hyperdiploidy of 51–65 chromosomes is associated with a good prognosis in childhood B-lineage acute lymphoblastic leukemia (ALL). Blasts from childhood ALL patients with a hyperdiploid karyotype have a tendency to apoptosis when cultured on stromal layers in vitro. In this study, we apply a novel method to investigate the relationship between apoptosis and hyperdiploidy in lymphoblasts of childhood ALL. Materials and Methods. The DNA content of individual ALL blasts in Feulgen-stained archival bone marrow smears can be determined by static cytometry. TUNEL (TdT-mediated dUTP-biotin nick end labeling) detects the DNA degradation associated with apoptosis. We performed TUNEL in situ sequential to DNA ploidy analysis in archival bone marrow smears from 12 patients with childhood ALL. Results. Five patients were diploid and seven were hyperdiploid (51–65 chromosomes) by conventional cytogenetic analysis. In the five diploid cases, the percentage of TUNEL-positive blasts ranged from 1.0% to 1.3%; in the seven hyperdiploid cases, the percentage of TUNELpositive blasts ranged from 3.6% to 9.0%. Comparing TUNEL and corresponding Feulgen images, we found that apoptotic blasts were predominantly of high DNA ploidy in both diploid and hyperdiploid cases. The mean DNA value of apoptotic blasts was larger than that of the total blast population in each case. Conclusions. The results demonstrate an increased incidence of spontaneous apoptosis in situ of hyperdiploid blasts in ALL bone marrow and indicate that this phenomenon is not restricted to in vitro cultures. The findings provide a possible rationale for the good prognosis associated with hyperdiploid childhood ALL. © 2002 International Society for Experimental Hematology. Published by Elsevier Science Inc.

Childhood B-lineage acute lymphoblastic leukemia (ALL) is the most common form of cancer in children [1]. The identification of prognostic factors has permitted the use of risk group stratification and treatment assignment, which has contributed to a marked improvement in the long-term prognosis of childhood ALL [2]. Approximately one third of childhood ALL patients have a hyperdiploid karyotype at diagnosis. The longer survival time after treatment of children in this subgroup was first reported by Secker-Walker et al. [3] and later found to be associated with leukemic karyotypes having over 50 chromosomes [4]. The explanation for

Offprint requests to: Szu-Hee Lee, M.B., Ph.D., Division of Haematology, National University Hospital, Singapore 119074; E-mail: [email protected]

the prognostic significance of this cytogenetic feature is largely unknown, but the low leukemic cell burden associated with this finding [5], increased accumulation of methotrexate polyglutamates in hyperdiploid ALL blasts [6,7], their increased sensitivity to antimetabolites [8], and their propensity for spontaneous apoptosis in vitro [9,10] are all thought to contribute to the favorable prognosis. Several methods are available to identify the hyperdiploid clone. Karyotyping a bone marrow sample is an accurate but time-consuming procedure and requires dividing cells. Fluorescence in situ hybridization [11], flow cytometry, static cytometry, and FICTION (fluorescence immunophenotyping and interphase cytogenetics as a tool for investigation of neoplasms) can all be conducted on interphase cells. Of these methods, only static cytometry permits morphologic identification of the cells utilized. It does not require a fresh sample

0301-472X/02 $–see front matter. Copyright © 2002 International Society for Experimental Hematology. Published by Elsevier Science Inc. PII S0301-472X(02)0 0 7 7 1 - 3

334

Y. Zhang et al./Experimental Hematology 30 (2002) 333–339

and can be conducted on archived material. It has been of value in estimating the ploidy levels in several malignant conditions [12–16]. Apoptosis, or programmed cell death, is a process whereby cells die in a controlled manner, in response to specific stimuli, apparently following an intrinsic program [17]. Characteristic morphologic changes, identified by electron microscopy, that accompany apoptosis include nuclear chromatin condensation, compactness of cytoplasmic organelles. and the appearance of pedunculated protuberances on the cell surface [18]. Apoptosis is associated with DNA degradation and thought to be caused by endonuclease activity that can cleave DNA between nucleosomes to give fragments that resolve on electrophoresis as multiples of about 180 bp [19]. The 180-bp fragments may be detected by the appearance of a ladder of nucleosomal DNA fragments after gel electrophoresis of extracted genomic DNA. An enzymatic approach to detect apoptosis in individual cells relies upon the use of terminal deoxynucleotidyltransferase (TdT) to add labeled deoxynucleotides to 3-OH ends of DNA fragments [20]. In TUNEL (TdT-mediated dUTPbiotin nick end labeling), the enzymic signal is amplified by fluorescence or peroxidase-conjugated avidin, for the detection of apoptosis in situ or by flow cytometry. Recent studies demonstrated that hyperdiploid childhood ALL has poor cell recovery in vitro after 7 days of culture on stromal cell layers, which led the authors to conclude that lymphoblasts from hyperdiploid childhood ALL patients have a marked propensity to undergo apoptosis in vitro [9,10]. However, the DNA ploidy of the apoptotic blasts had not been ascertained, and it was unclear whether hyperdiploid ALL lymphoblasts require more stringent culture conditions [10] and whether the findings reflect in vivo phenomena, because hyperdiploid ALL lymphoblasts apparently thrive in the bone marrow. Therefore, the relationship between hyperdiploidy and apoptosis in childhood ALL lymphoblasts in the bone marrow is of considerable relevance and interest. In this study, we analyzed this relationship by sequential determination of DNA ploidy by static cytometry and detection of apoptosis by in situ TUNEL in archival bone marrow smears from cases of childhood ALL.

Materials and methods Patients Archival bone marrow smears from 12 children (age 22 months to 4 years) presenting with childhood ALL at the National University Hospital, Singapore, between 1994 and 1998 were studied. The study was approved by the local Institutional Review Board (Ethics). Unfixed, unstained smears of bone marrow aspirates obtained at first presentation, before treatment, and stored at room temperature (RT; 24C) were analyzed. Cases were included if G-banding showed a mainline karyotype of more than 50 chromosomes or a diploid karyotype and, if unstained, unfixed presentation bone marrow slides were available. The percentage of blast cells in the

12 bone marrow aspirates ranged from 84 to 98%. The peripheral total white blood cell count was less than 50  109/L in all cases. All cases were determined to be of L1 or L2 ALL phenotype according to the morphologic and cytochemical criteria of the FrenchAmerican-British Co-operative Group [21]. The karyotype and DNA ploidy analysis for six of the hyperdiploid cases have been previously reported by us [22]. Cells The HL-60 cell line (ATCC, Manassas, VA, USA) was maintained in RPMI 1640 tissue culture medium (Sigma Chemical Corp., St. Louis, MO, USA) supplemented with 10% fetal calf serum (FCS; Life Technologies, Inc., Gaithersburg, MD, USA) and 2 mM L-glutamine (Sigma) in 5% CO2 at 37C. Cell suspensions were treated with the protein synthesis inhibitor cycloheximide (CHX) and/or the DNA topoisomerase inhibitor camptothecin (CAM) as described previously [23]. Briefly, cell suspensions were washed in phosphate-buffered saline (PBS), resuspended in RPMI with 10 mM HEPES (Sigma), and aliquoted at a concentration of 5  105 cells/mL into 12 mm  75 mm polypropylene tubes. CAM 0.15 M CAM (Sigma), 5 M CHX (Sigma), and 0.15 M CAM mixed with 5 M CHX were added to cell aliquots. The same volume of RPMI was added to control cell aliquots. Cell suspensions were incubated for 4 hours at 37C in 5% CO2, then washed three times with PBS and resuspended in RPMI with 5% FCS. Aliquots of 2  105 cells were cytospun onto glass slides. Slides were air dried, fixed in freshly prepared 4% paraformaldehyde (BDH, Poole, England) for 30 minutes at RT, then washed three times in PBS and air dried. Sequential DNA ploidy determination and in situ TUNEL assay Static cytometric ploidy analysis of Feulgen-stained cytospun HL-60 cells or patient bone marrow smears was carried out exactly as previously described [22] and conducted by an operator who did not have prior knowledge of the karyotype. Mitotic cells, overlapping cells, and nonlymphoblasts (e.g., nucleated red cells, myeloid cells, lymphocytes, and red cells) were excluded from analysis. More than 200 lymphoblasts from each patient were identified by their morphologic features of contour, size, nuclear to cytoplasmic ratio, chromatin pattern, and presence of nucleoli. The position of each cell was recorded electronically by Scoposcan relocation equipment (Leica, Wetzler, Germany) and the image captured and digitally stored by LBase software (Leica). The intensity of Feulgen stain was converted to gray scale and the integrated optical density (IOD) of each cell measured. The ploidy level was established by comparison with the 2c value, determined by sex-matching patients to cases whose blast cells were known to have a normal karyotype. To avoid any variation from staining or processing, all slides were stained at the same time and all images were captured under the same calibration settings of the automated light microscope. Subsequent to Feulgen image capture, the slides were destained in 5N HCl at room temperature for 1 hour and rinsed in PBS three times. Apoptosis was detected by the TUNEL assay using the Fluorescence In situ Cell Death Detection kit (Boehringer Mannheim, Fremont, CA, USA) according to the manufacturer’s instructions. Briefly, unfixed bone marrow slides or slides that had been destained after Feulgen staining were permeabilized with 0.1% Triton X-100 in 0.1% sodium citrate for 2 minutes on ice. The slides then were rinsed with PBS twice and air dried. TUNEL reac-

Y. Zhang et al./Experimental Hematology 30 (2002) 333–339

335

tion mixture (50 L) was added to each slide. The slides were incubated in a humidified chamber in the dark for 60 minutes at 37C. After incubation, the slides were viewed under a fluorescence microscope (40 objective; Leica) at absorption and emission maximum wavelengths of 492 and 515, respectively. After incubation, slides were rinsed with PBS three times and mounted in Antifade (Oncor, Gaithersburg, MD, USA). The lymphoblasts previously captured after Feulgen staining were electronically relocated and the fluorescent TUNEL image captured and stored using QFish software (Leica). By comparing corresponding Feulgen and TUNEL images, the percentage of TUNEL-positive lymphoblasts in each patient and the DNA ploidy value of each TUNEL-positive cell could be ascertained. Mitotic index The mitotic index in each case was measured by counting the number of mitotic blasts per 1,000 lymphoblasts in Giemsa-stained marrow smears.

Results Cytogenetic data The bone marrow karyotypes of the 12 childhood ALL cases are listed in Table 1. The numerical additions of chromosomes 4, 6, 10, 14, 17, 18, 21, and X, together with the structural alterations to the long arm of chromosome 1, which are observed in cases 1 to 7, are consistent with those previously reported in the good prognosis group of childhood ALL patients [24]. Cases 9 to 12 have a normal diploid karyotype, but case 8 is pseudodiploid, having deletions of four normal chromosomes and additions of four unidentified marker chromosomes. Determination of DNA ploidy by static cytometry Histograms of the DNA ploidy of Feulgen-stained blast cells obtained were in good agreement with those previously published [22]. They are presented in Figure 1, subdivided into TUNEL-negative blasts, as indicated by the large plot, and TUNEL-positive blasts, as indicated by the small

Figure 1. Histograms of DNA ploidy of the TUNEL-negative and TUNELpositive blasts in 12 cases of childhood ALL. For each patient, the large plot represents the TUNEL-negative blasts and the small plot the TUNEL-positive blasts. The white histograms represent the DNA ploidy of control diploid cases.

plot for each patient. The diploid control population is shown alongside the test population in each case, wherever the diploid histogram does not obscure the test population. The control populations are represented by a normal distri-

Table 1. Clinical details and karyotype of twelve cases of childhood ALL Case no. 1 2 3 4 5 6 7 8 9 10 11 12

Age

Sex

Karyotype

FAB subtype

4 years

M

L1

3 years 22 months 2 years 3 years 2 years 4 years 4 years 4 years 4 years 3 years 4 years

F M F F F M F F M M M

61,XY, X, Y, dup(1)(q11q44), 4, 5, 6, 8, 10, 12, 14, 17, 18, 21, 21, 22, mar [4]/46, XY [3] 50–56, XX, ?, ?, ?, ?, ?, [cp9]/46, XX [9] 57, XY, X, 4, 6, 8, 10, 14, 17, 18, 21, 21, 22 [6]/46, XY [4] 56, XX, X, dup(1)(q21q32), 4, 6, ?9, 10, 14, 17, 18, 21, 21 [8]/46, XY [7] 51, XX, del(1)(q25), 6, 8, 21, 21, mar [14]/45, XX, del(1)(q25), 13, 16, mar[3]/46, XY [7] 53, XX, X, 6, 8, 10, 14, 21, 21[10] 57, X, Y, X, dup(1)(q23q42), 4, 6, 10, 11, 11, 12, 14, 17, 18, 21, 21 [24]/46, XY [6] 46, X, X, 4, 10, del(11)(q23), 18, 4mar 46, XX 46, XY 46, XY 46, XY

L2 L1 L1 L1 L2 L1 L1 L1 L2 L2 L2

336

Y. Zhang et al./Experimental Hematology 30 (2002) 333–339

bution curve around 2c; this also is observed in cases 8 to 12, which were diploid by cytogenetics. In cases 1 to 7, where a hyperdiploid karyotype had been established by conventional cytogenetics, the blast population is represented by a normal distribution curve with a peak DNA value greater than 2c. Sequential Feulgen and TUNEL procedure: preliminary experiments A method by which DNA ploidy assay and detection of apoptosis could be performed sequentially was tested to determine whether the percentage of TUNEL-positive blasts in HL-60 cells or childhood ALL marrow smears was affected by prior Feulgen staining and destaining. Apoptosis in HL-60 cells was induced by CHX and CAM, as described previously [23]. Treated HL-60 cells were cytospun onto glass slides and fixed. The slides were Feulgen stained, then destained, followed by TUNEL in situ. Feulgen and TUNEL images were captured at the appropriate stages. In control replicate slides, TUNEL was carried out first, followed by Feulgen staining. The percentage of TUNEL-positive cells in each slide was obtained by comparison of TUNEL and corresponding Feulgen images. Results showed good agreement between the percentages of TUNEL-positive cells detected in the destained slides and control slides (Table 2). The percentage of TUNEL-positive cells in these experiments also showed good agreement with the data previously reported by Gong et al. [23], who used TUNEL by flow cytometry to detect apoptosis in similarly treated HL-60 cells. A similar experiment that was carried out on pairs of bone marrow smears from six childhood ALL patients also showed good agreement of percentages of TUNEL-positive lymphoblasts in destained and control slides (not shown). These data indicate that the percentage of TUNEL-positive lymphoblasts was not altered by Feulgen staining and destaining.

Sequential determination of DNA ploidy and apoptosis by TUNEL in childhood ALL lymphoblasts Representative TUNEL and corresponding Feulgen images in childhood ALL smears are shown in Figure 2. TUNELpositive lymphoblasts did not have a morphology that suggested they were necrotic. The percentages of TUNEL-positive blasts in the 12 cases of childhood ALL are listed in Table 3. In the five diploid cases, the percentage of TUNELpositive blasts ranged from 1.0 to 1.3% (mean 1.22%); in the seven hyperdiploid cases, the percentage of TUNELpositive blasts ranged from 3.6 to 9% (mean 4.96%). The median (interquartile range) number of TUNEL-positive blasts (%) was 1.3 (0.2) in the diploid cases (n  5) and 4.4 (1.9) in the hyperdiploid cases (n  7). By Mann-Whitney U-test, p  0.003, which indicated that there was a highly significant difference between the two groups. Furthermore, the mean DNA value of TUNEL-positive blasts was larger than that of the total blast population in each case. Case 1,

Table 2. Comparison of the percentages of TUNEL positive HL-60 cells in destained and in control slides

Apoptosis-inducing agent None 0.15M CAM 5M CHX 0.15M CAM  5M CHX

Percent TUNEL-positive cells in Feuglen-destained slides (percent SE)

Percent TUNEL-positive cells in control slides (percent SE)

3.5 0.6 38.0 1.1 33.6 1.1 41.8 0.3

3.7 0.6 37.8 0.5 33.0 2.2 40.4 2.0

Percentages are the mean values obtained from counting more than 1,800 cells in three experiments. CAM  camptothecin; CHX  cycloheximide.

Figure 2. Comparison of TUNEL-positive lymphoblast images (A) and their corresponding Feulgen-stained images (B) in childhood ALL bone marrow smears. Image set 6 is a negative control, where TdT was omitted from the TUNEL reaction mixture. Image set 7 is a positive control, where 10 units of DNAse I were added to the bone marrow smear prior to the TUNEL reaction mixture.

Y. Zhang et al./Experimental Hematology 30 (2002) 333–339

337

Table 3. DNA content, TUNEL positivity, and mitotic indexes of childhood ALL cases

Case no. 1 2 3 4 5 6 7 8 9 10 11 12

Abbreviated karyotype 61[4]/46[3] 50–56[9]/46[9] 57[6]/46[4] 56[8]/46[7] 51[14]/45[3]/46[7] 53 57[24]/46[6] 46 46 46 46 46

Mean DNA value of total blasts (c) SD

Percent TUNEL-positive blasts

Mean DNA value of TUNEL-positive blasts (c) SD

Mitotic index (%)

3.68 0.86 2.78 0.67 3.29 0.73 2.95 0.51 3.76 1.43 3.04 1.16 2.88 0.49 2.08 0.43 2.07 0.49 2.01 0.49 2.02 0.40 2.05 0.37

9.0 4.7 3.6 5.5 3.9 3.6 4.4 1.3 1.2 1.3 1.3 1.0

3.75 1.01 3.54 1.22 3.50 1.16 3.08 0.33 4.00 0.55 3.83 0.92 3.20 0.71 3.25 0.62 2.81 0.77 3.08 0.33 3.33 0.88 2.67 0.06

0.35 0.48 0.27 0.45 0.30 0.54 0.33 0.31 0.65 0.69 0.33 0.45

Greater than 200 lymphoblasts from each patient were analyzed.

with a karyotype of 61 chromosomes, had the highest percentage of TUNEL-positive blasts (9%). However, the percentage of TUNEL-positive blasts in the other six hyperdiploid cases did not appear to increase proportionally in relation to chromosome number. The DNA ploidy value of each TUNEL-positive blast was determined and is indicated by the small plot for each patient in Figure 1, with the percentages listed in Table 3. DNA ploidy values of TUNEL-positive blasts indicated that, in all cases, the TUNEL-positive blasts had a DNA content greater than 2c. Interestingly, the mean DNA value of the TUNEL-positive blasts is seen to be larger than that of the total population, even in the diploid samples. However, the possibility that a proportion of the TUNEL-positive blasts were diploid blasts in G2 phase could not be excluded. In the diploid cases, no TUNEL-positive cell was over 4c, but in the hyperdiploid cases, the TUNEL-positive cells had ploidy values of 4, 5, 6, and 7c. Mitotic index The mitotic indices of the cases studied are listed in Table 3. The mitotic index in the 12 childhood ALL cases ranged from 0.27 to 0.69% (mean 0.43%). Discussion Recent studies indicated a strong correlation between in vitro growth potential and treatment outcome in childhood ALL, based on observations that lymphoblasts from hyperdiploid childhood ALL patients were among those least able to grow in 7-day stromal cultures [9]. When flow cytometric labeling with annexin V, a marker of apoptosis, was applied to lymphoblasts from one case of childhood ALL with 55 chromosomes after 2 days in culture, it was found that the lymphoblasts had undergone marked apoptosis and DNA fragmentation [10]. From these findings, the authors hypothesized that hyperdiploid childhood ALL lymphoblasts have a propensity to undergo apoptosis [9,10].

We explored the relationship between DNA ploidy and apoptosis in childhood ALL patients by studying primary uncultured lymphoblasts in situ in archival bone marrow smears, which should provide insight into authentic in vivo conditions. Our results showed that there is a higher percentage of apoptotic blasts in the bone marrow of hyperdiploid cases of childhood ALL (3.6–9.0%; n  7) compared to diploid cases of childhood ALL (1.0–1.3%; n  5). By comparing TUNEL and corresponding Feulgen images, we found that apoptotic blasts were predominantly of high DNA ploidy in both diploid and hyperdiploid cases. Further, the mean DNA value of TUNEL-positive blasts was larger than that of the total blast population in each case. These results are consistent with in vitro findings previously described [9,10], but further demonstrate that spontaneous apoptosis, predominantly of blasts of high DNA ploidy, takes place in the bone marrow in hyperdiploid childhood ALL and that this phenomenon is not restricted to in vitro cultures. To exclude the possibility that hyperdiploid lymphoblasts detected in this study were mitotic blasts, cells with obvious mitotic nuclear morphology were not included in image analysis. In general, mitotic lymphoblasts are seen infrequently in the L1 and L2 FAB subtypes of ALL [25]. The mean mitotic index of lymphoblasts in the 12 cases of childhood ALL (0.43%; range 0.27–0.69%) in this study is in agreement with reported data of a mean mitotic index of about 0.54% (range 0.05–1.6%) in childhood ALL lymphoblasts [26]. One drawback to our study was that it was not possible to determine whether the hyperdiploid TUNEL-positive lymphoblasts were hyperdiploid or diploid 4c blasts in G2 phase. That TUNEL-positive blasts in the 4c state were present in the diploid cases (cases 8 to 12) and that the DNA ploidy value of TUNEL-positive blasts was higher than the mean DNA value for that population could indicate that apoptosis was occurring in G2 in these cases. This could suggest

338

Y. Zhang et al./Experimental Hematology 30 (2002) 333–339

that childhood ALL blasts have a greater tendency to spontaneous apoptosis when the cell is in the 4c state in G2. It is of note that in the diploid cases, no TUNEL-positive cell was over 4c, but in the hyperdiploid cases the TUNEL-positive cells had ploidy values of 4, 5, 6, and 7c, suggesting they were from hyperdiploid cells in G2. Thus, although there is the possibility that some of the TUNEL-positive blasts in the hyperdiploid cases included diploid 4c blasts in G2 phase, this is likely to represent the minority due to the increased ploidy levels obtained and the low incidence of diploid blasts found in these patients. Determination of TUNEL-positive cells in triploid or tetraploid cases of childhood ALL would clarify the extent to which the increased apoptosis observed in the hyperdiploid good prognosis group is related to high DNA ploidy. The good agreement between the karyotype determined by conventional cytogenetics and the DNA ploidy determined by static cytometry in all cases [22], together with the elevated ploidy levels of the TUNEL-positive blasts in the hyperdiploid cases, suggests that at least the majority of the apoptotic blasts in the hyperdiploid cases were from hyperdiploid cells. Investigators have proposed that hyperdiploid childhood ALL lymphoblasts may have reduced tumorigenicity resulting from increased gene dosage due to the extra chromosomes of hyperdiploid childhood ALL [27] or a greater tendency towards terminal differentiation [28], but experimental evidence for these hypotheses is lacking. It is pertinent to note that, unlike hyperdiploidy of 51–65 chromosomes, neartriploidy and near-tetraploidy are unfavorable prognostic features in childhood ALL [29]. This argues that hyperdiploid childhood ALL with 51–65 chromosomes constitutes a biologically distinct subtype of childhood ALL. The mechanisms that could lead to apoptosis in hyperdiploid blasts remain obscure. It is possible that the diverse genetic abnormalities associated with hyperdiploid childhood ALL trigger spontaneous apoptosis by a variety of apoptotic pathways. Indeed, apoptosis is a complex physiologic process dependent on the integrated functioning of a large number of gene products, and several families of genes are known to encode positive or negative regulators that are important in the control of apoptosis in leukemic blasts [30]. Characteristic features of hyperdiploid childhood ALL in vivo include lower blood cell counts, lack of bulky extramedullary disease, and higher percentages of proliferating cells [28,31–33]. A possible explanation for these features is that hyperdiploid childhood ALL lymphoblasts lack the ability to expand rapidly and cannot grow outside of the bone marrow microenvironment [10]. However, our findings favor the alternative explanation that lymphoblasts in hyperdiploid childhood ALL possess the intrinsic characteristic of spontaneous apoptosis in the bone marrow, resulting in a poorer survival in vivo. Compared to flow cytometric methods, static cytometry has unique advantages, such as correlation with cell morphology, the ability to analyze small specimens and archival

materials, and detection of rare cells with high ploidy values. The sequential techniques of determining DNA ploidy by static cytometry and apoptosis by in situ TUNEL perhaps can be applied to further analyze and define high- or low-risk patient groups in childhood ALL or other malignant disorders so that more appropriate chemotherapy can be designed to treat specific patient groups. Acknowledgments This work was supported by Grants RP3690032 from the National Medical Research Council, Singapore, and R-179-000-004-593 from the Singapore Cancer Society. The authors thank Jean Chen and Cathleen Teh for excellent technical assistance, and Dr. Dong Fang for statistical analysis.

References 1. Gurney JG, Davis S, Severson RK, Fang JY, Ross JA, Robison LL (1996) Trends in cancer incidence among children in the U.S. Cancer 78:532 2. Smith M, Arthur D, Camitta B, et al. (1996) Uniform approach to risk classification and treatment assignment for children with acute lymphoblastic leukemia. J Clin Oncol 14:18 3. Secker-Walker LM, Lawler SD, Hardisty RM (1978) Prognostic implications of chromosomal findings in acute lymphoblastic leukaemia at diagnosis. Br Med J ii:1529 4. Bloomfield CD, Goldman AI, Alimena R, et al. (1986) Chromosomal abnormalities identify high-risk and low-risk patients with acute lymphoblastic leukemia. Blood 67:415 5. Pui CH, Raimondi SC, Dodge RK, et al. (1989) Prognostic importance of structural chromosomal abnormalities in children with hyperdiploid (greater than 50 chromosomes) acute lymphoblastic leukaemia. Blood 73:1963 6. Whitehead VM, Vuchich MJ, Lauer SJ, et al. (1992) Accumulation of high levels of methotrexate polyglutamates in lymphoblasts from children with hyperdiploid (greater than 50 chromosomes) B-lineage acute lymphoblastic leukemia: a Pediatric Oncology Group study. Blood 80:1316 7. Synold TW, Relling MV, Boyett JM, et al. (1994) Blast cell methotrexate-polyglutamate accumulation in vivo differs by lineage, ploidy and methotrexate dose in acute lymphoblastic leukemia. J Clin Invest 94:1996 8. Kaspers GJL, Smets LA, Pieters R, et al. (1995). Favourable prognosis of hyperdiploid common acute lymphoblastic leukemia may be explained by sensitivity to antimetabolites and other drugs: results of an in vitro study. Blood 85:751 9. Kumagai M, Manabe A, Pui CH, et al. (1996) Stroma-supported culture of childhood B-lineage acute lymphoblastic leukaemia cells predicts treatment outcome. J Clin Invest 97:755 10. Ito C, Kumagai M, Manabe A, et al. (1999). Hyperdiploid acute lymphoblastic leukemia with 51 to 65 chromosomes: a distinct biological entity with a marked propensity to undergo apoptosis. Blood 93:315 11. Ritterbach J, Hiddemann W, Beck JD, et al. (1998) Detection of hyperdiploid karyotypes ( 50 chromosomes) in childhood acute lymphoblastic leukemia (ALL) using fluorescence in situ hybridization (FISH). Leukemia 12:427 12. Fossa SD, Kaalhus O, Scott-Knudsen O (1977). The clinical and histopathological significance of Feulgen DNA-values in transitional cell carcinoma of the human urinary bladder. Eur J Cancer 13:1155 13. Rehardt K, Auer G, Bjorkholm E, et al. (1984) Prognostic significance of nuclear DNA content in serous ovarian tumors. Cancer Res 44:2198 14. Backdahl M, Tallroth E, Auer G, et al. (1985) Prognostic value of nuclear DNA content in medullary thyroid carcinoma. World J Surg 9:980

Y. Zhang et al./Experimental Hematology 30 (2002) 333–339 15. Forsslund G, Zetterberg A (1990) Ploidy level determinations in highgrade and low-grade malignant variants of prostatic carcinoma. Cancer Res 50:4281 16. Inghirami G, Macri L, Rosati S, Zhu BY, Yee HT, Knowles DM (1994) The Reed-Sternberg cells of Hodgkin disease are clonal. Proc Natl Acad Sci U S A 91:9842 17. Kerr JFR, Wyllie AH, Currie AR (1972) Apoptosis: a basic biological phenomenon with wide-ranging implications in tissue kinetics. Br J Cancer 26:239 18. Wyllie AH, Morris RG, Smith AL, Dunlop D (1984) Chromatin cleavage in apoptosis: association with condensed chromatin morphology and dependence on macromolecular synthesis J Pathol 142:67 19. Wyllie AH (1980) Glucocorticoid-induced thymocyte apoptosis is associated with endogenous endonuclease activation. Nature 284:555 20. Gavrieli Y, Sherman Y, Ben-Sasson SA (1992) Identification of programmed cell death in situ via specific labelling of nuclear DNA fragmentation. J Cell Biol 119:493 21. Bennett JM, Catovsky D, Daniel M-T, et al. (1976) Proposals for the classification of the acute leukaemia. French-American-British (FAB) Co-operative Group. Br J Haematol 33:451 22. van den Berghe JA, Chen YC, Zhang Y, Kueh YK, Lee SH (2000) Static cytometry identifies hyperdiploid childhood ALL. Cancer Genet Cytogenet 123:123 23. Gong JP, Li X, Darzynkiewicz Z (1993) Different patterns of apoptosis of HL-60 cells induced by cycloheximide and camptothecin. J Cell Physiol 157:263 24. Raimondi SC, Pui CH, Hancock ML, Behm FG, Filatov L, Rivera GK

25. 26.

27.

28.

29. 30.

31.

32.

33.

339

(1996) Heterogeneity of hyperdiploid (51-67) childhood acute lymphoblastic leukaemia. Leukaemia 10:213 Bain BJ (1990) Leukemia diagnosis: a guide to the FAB classification. London: Gower Medical Publishing Sinks LF, Godden JO (1975). Staging of acute lymphoblastic leukemia. In: Murphy SB, Mauer AM (eds.) Conflicts in Childhood Cancer: an Evaluation of Current Management. New York: Liss, pp. 7–29 Tsuchiya H, Matsuda I, Kaneko Y (1991) Why does childhood lymphoblastic leukemia with hyperdiploidy show a favorable prognosis? Cancer Genet Cytogenet 50:273 Smets LA, Slater RM, Behrendt H, Van’t Veer MB, Homan-Blok J (1985) Phenotypic and karyotypic properties of hyperdiploid acute lymphoblastic leukaemia of childhood. Br J Haematol 61:113 Pui CH, Carroll AJ, Head D, et al. (1990) Near-triploid and near-tetraploid acute lymphoblastic leukemia of childhood. Blood 76:590 Wickremasinghe RG, Hoffbrand AV (1999) Biochemical and genetic control of apoptosis: relevance to normal hematopoiesis and hematological malignancies. Blood 93:3587 Williams DL, Tsiatis A, Brodeur GM, et al. (1982) Prognostic importance of chromosome number in 136 untreated children with acute lymphoblastic leukemia. Blood 60:864 Secker-Walker LM, Swansbury GJ, Hardisty RM, et al. (1982) Cytogenetics of acute lymphoblastic leukaemia in children as a factor in the prediction of long-term survival. Br J Haematol 52:389 Look AT, Melvin SL, Williams DL, et al. (1982) Aneuploidy and percentage of S-phase cells determined by flow cytometry correlate with cell phenotype in childhood acute leukemia. Blood 60:959