Leukemia (2006) 20, 645–649 & 2006 Nature Publishing Group All rights reserved 0887-6924/06 $30.00 www.nature.com/leu

ORIGINAL ARTICLE Dendritic cell deficiencies in pediatric acute lymphoblastic leukemia patients B Maecker1, D Mougiakakos1, M Zimmermann1, M Behrens1, S Hollander1, A Schrauder2, M Schrappe2, K Welte1 and C Klein1 Department of Pediatric Hematology/Oncology, Hannover Medical School, Hannover, Germany and 2Department of Pediatrics, University Medical Center Schleswig-Holstein, Kiel, Germany 1

Acute lymphoblastic leukemia (ALL) cells are particularly poor at generating anti-leukemia immunity, despite residing in lymphoid organs. To assess a potential role of dendritic cells (DC) in poor anti-leukemia immunity, we analyzed peripheral blood DC in 55 pediatric ALL patients at the time of initial diagnosis and 19 age-matched healthy controls. Dendritic cells were identified by their expression of HLA-DR, lack of B, T, NK, and monocyte markers, and expression of CD11c (myeloid DC(mDC)) or BDCA-2 (plasmacytoid DC(pDC)) using flow cytometry. We found that in children with B-lineage ALL, numbers of both mDC and pDC were significantly reduced (P ¼ 0.0001). In contrast, T-lineage ALL patients showed normal pDC and significantly elevated mDC (P ¼ 0.003) levels, with normal expression of HLA-DR and co-stimulatory molecules. A decrease in DC could not be explained by general impairment of myelopoiesis, as we could not demonstrate a correlation of DC numbers with granulocyte/monocyte numbers in patients with B-lineage ALL. However, aberrant expression of myeloid surface markers on leukemic blasts was frequent in patients lacking myeloid DC indicating a potential block of DC differentiation. Thus, depletion of DC in B-lineage ALL patients may contribute to poor anti-leukemia immune responses. Leukemia (2006) 20, 645–649. doi:10.1038/sj.leu.2404146; published online 23 February 2006 Keywords: acute lymphoblastic leukemia; dendritic cell; flow cytometry; tumor immunology

Introduction Acute lymphoblastic leukemia (ALL) represents the most common malignancy in children. Despite advances in risk stratification and risk-adapted chemotherapeutic treatment, about 20% of patients eventually suffer from disease relapse.1–3 Experience from allogeneic bone marrow transplantation in pediatric ALL indicates that strong anti-leukemia immune responses, such as graft-versus-leukemia effects, are associated with increased disease-free survival.4,5 However, several investigators have demonstrated poor in vivo anti-leukemia immune responses in ALL patients, even in the bone marrow as the primary site of disease.6 Strategies to induce anti-leukemia T-cell responses ex vivo have utilized activation of ALL blasts,7 addition of cytokines and priming of T cells using professional or engineered antigen-presenting cells.8 Dendritic cells (DC) represent the key initiators of immune responses in vivo. Defects in DC function have been demonstrated in several other diseases of either viral infection (chronic hepatitis C,9 HIV infection10) or leukemic malignancies (acute Correspondence: Dr B Maecker, Paediatrische Haematologie/Onkologie, Medizinische Hochschule Hannover, Carl-Neuberg-Street 1, 30625 Hannover, Germany. E-mail: [email protected] Received 31 December 2005; accepted 16 January 2006; published online 23 February 2006

myeloid leukemia,11 chronic myelogenous leukemia,12 hairy cell leukemia13). Recently, Mami et al.14showed reduced DC in a group of adult and pediatric patients with B-cell precursor ALL. Thus, defects in dendritic cells may impact on poor antileukemia immune responses in patients with childhood ALL. We reasoned that quantitative and/or qualitative aberrations in DC might reflect impaired immunity in children with ALL. Here, we show a significant reduction of peripheral blood DC (pbDC) in pediatric patients with B-precursor ALL. Furthermore, we demonstrate that aberrant expression of myeloid surface markers on leukemic blasts is common in patients with lack of myeloid DC (mDC) and may point toward a block in DC differentiation in pB-ALL patients. In contrast, DC numbers in T-precursor ALL patients are increased and the cells do not exhibit functional alterations.

Materials and methods

Donor and patient samples Peripheral blood from healthy children and ALL patients was obtained by phlebotomy following informed consent and approval by our institution’s review board. All patients included in this study were treated on the ALL-BFM 2000 chemotherapy protocol. Clinical and disease parameters were retrieved from the ALL-BFM 2000 database. Peripheral blood mononuclear cells (PBMC) were purified by Ficoll-density centrifugation.

Flow cytometry Peripheral blood mononuclear cells were preincubated with human IgG to block Fc receptors and stained with fluorochrome-conjugated antibodies against CD3, CD4, CD8 (Dako Corporation, Hamburg, Germany), CD11c, CD14, CD16, CD19, CD25, CD28, CD40, CD56, CD86, CD123, HLA-DR (BD Pharmingen, Heidelberg, Germany) and BDCA-1, BDCA-2, BDCA-3 (Miltenyi Biotech, Bergisch Gladbach, Germany). A cocktail of CD3, CD14, CD16 and CD19 was used to discriminate lineage-negative cells. Myeloid DC were characterized as lineage-/HLA-DR þ /CD11c þ and plasmacytoid DC (pDC) as lineage-/HLA-DR þ /BDCA2 þ . Samples were analyzed on a FACS Calibur flow cytometer using cellquest software (Becton Dickinson, Heidelberg, Germany). At least 10 000 events were acquired in healthy children and 50 000 in ALL patients.

Statistical analysis The median proportion of pbDC was compared using Mann– Whitney test. Univariate analysis within the group of pB-ALL patients was performed using U-test, Fischer’s test or Spearman’s Rho correlation as indicated. A result is considered significant if Po0.05.

Dendritic cell deficiencies in pediatric ALL patients B Maecker et al

646 Results

Peripheral blood dendritic cells are markedly reduced in B-precursor acute lymphoblastic leukemia patients but elevated in T-precursor acute lymphoblastic leukemia patients As published experience with respect to normal levels of pbDC in children is limited,15,16 we first analyzed pDC and mDC numbers in a cohort of 19 healthy children. No significant correlation of either mDC (median (range); 20 (2.3–52) mDC/ml, P ¼ 0.40) or pDC (16 (1.7–64) pDC/ml, P ¼ 0.79) with age could be demonstrated in these children (data not shown). Next, we enumerated pbDC counts in pediatric patients presenting with ALL (Table 1), analyzing patients with T-cell precursor ALL (pT-ALL, n ¼ 17) and patients with B-cell precursor ALL (pB-ALL, n ¼ 38) separately. Compared to healthy donors, patients with pB-ALL revealed a significant decrease in the number of mDC (median 2.2 (0–78) mDC/ml, P ¼ 0.0001) and pDC (0.5 (0–68) pDC/ml, P ¼ 0.0001) in peripheral blood

Table 1

(Figure 1a and b). Within the group of B-cell precursor ALL patients, we found a group of patients characterized by a complete absence of mDC (n ¼ 6) or pDC (n ¼ 12). Interestingly, whereas numbers of mDC and pDC correlated well in healthy children (Figure 1c), no correlation could be established in pBALL patients: three pB-ALL patients had neither mDC nor pDC detectable in their blood, whereas nine patients had normal mDC counts in the absence of pDC, and three patients had no detectable mDC; while pDC counts were within normal range (Figure 1d). In striking contrast to patients with pB-ALL, pT-ALL patients showed significantly elevated levels of mDC (median 98.5 (6.8– 711) mDC/ml, P ¼ 0.003; Figure 1a) and slightly elevated counts of pDC (28.4 (0–330) pDC/ml, P ¼ 0.022; Figure 1b). Except for two patients with no detectable mDC in the peripheral blood, numbers of mDC and pDC correlated well (Figure 1e). Thus, pbDC are normal or increased in children with pT-ALL whereas reduced or absent in pediatric pB-ALL patients. In the following steps we further characterized pbDC in pTALL patients and sought to elucidate potential mechanisms of DC depletion in pB-ALL patients.

Patient characteristics

Total number of patients Sex Male Female Age (years) Median (range)

Healthy donors

pB-ALL patients

pT-ALL patients

19

38

17

16 3

21 17

11 6

10.6 (2.5–18)

4.6 (2.1–15.9)

10.0 (3.9–17.7)

Correlation of peripheral dendritic cell numbers with monocyte and neutrophil counts In view of these significant numerical alterations in the DC compartment, we reasoned that reduced DC numbers might reflect decreased myelopoiesis in patients with ALL. To elucidate this question, we correlated neutrophil and monocyte counts as surrogate markers of myelopoesis with DC counts in pT-ALL and pB-ALL patients, respectively. In pT-ALL patients, mDC numbers correlated well with neutrophil counts (P ¼ 0.03; Figure 2a) and monocytes (P ¼ 0.004; data not shown). For pDC, there was only a slight

Figure 1 Absolute numbers of myeloid and plasmacytoid dendritic cells (pDC) in the peripheral blood of patients with pB- or pT-ALL and healthy donors. Dendritic cells (DC) were measured by flow cytometry. P-values were calculated using Mann–Whitney-test. (a) Myeloid DC (mDC); (b) plasmacytoid DC (pDC). Correlation of mDC and pDC within the same individuum is shown for healthy children (c), B-precursor ALL patients (d) and T-precursor ALL patients (e). Leukemia

Dendritic cell deficiencies in pediatric ALL patients B Maecker et al

647 trend of correlation with neutrophils (P ¼ 0.15; Figure 2b) and monocytes (P ¼ 0.05; data not shown). In contrast, mDC in pBALL patients showed a slight trend of negative correlation with neutrophils (P ¼ 0.17; Figure 2c) and monocytes (P ¼ 0.06; data not shown). No correlation at all could be demonstrated for pDC in pB-ALL patients (P ¼ 0.97 for neutrophils, Figure 2d; P ¼ 0.99 for monocytes, data not shown). Similarly, in univariate analysis, reduced numbers of DC were not associated with hemoglobin levels, platelet counts or blast infiltration of bone marrow

(Table 2a). In summary, whereas levels of pbDC in pT-ALL patients seem to be correlated with intact myelopoeisis, decreased DC numbers in patients with pB-ALL cannot readily be explained by impaired bone marrow function.

Dendritic cells are phenotypically normal in pT-ALL patients To further characterize pbDCs in pT-ALL patients, we next determined their surface marker expression by flow cytometry. Both mDC and pDC showed equivalent expression levels of costimulatory molecules CD40 and CD86 (Figure 3a and b) as well as MHC class II (not shown) in pT-ALL patients and healthy controls, respectively. Activation marker CD83 was almost absent on pbDC in both groups. Dzionek et al.17 described different subsets of mDC in peripheral blood characterized by expression of BDCA-1 or BDCA-3. In our analysis, there was no difference in expression of either marker on pbDC between pTALL patients and healthy control donors (Figure 3c). Based on our phenotypic analysis, we conclude that DCs in pT-ALL patients do not show altered activation/maturation levels.

Absence of mDC is associated with aberrant expression of myeloid markers CD13, CD33, and myeloperoxidase on leukemic blasts in pB-ALL patients

Figure 2 Correlation of peripheral blood dendritic cell and absolute neutrophil count in pT-ALL (a, b) and pB-ALL patients (c, d). Myeloid DC levels (a, c) and plasmacytoid DC (b, d) are shown.

Table 2

Owing to the limited numbers, DC could not be further analyzed in pB-ALL patients. We therefore sought to assess whether any biological features of ALL cells might be associated with the lack of DC in pB-ALL patients. No association of low DC numbers was found with patient’s age or sex (data not shown). In univariate analysis, patients with lack of mDC were significantly more likely to have an immunophenotype of common ALL (P ¼ 0.01; Table 2a), whereas there was no association between absence of pDC and immunophenotype.

Univariate analysis of mDC and pDC numbers and patient parameters in pB-ALL patients

(a) Analysis of parameters at presentation using Fisher’s test expressing myeloid markers CD13, CD33 and myeloperoxidase with mDC and pDC numbers mDCo2/ml mDCX2/ml P-value pDC ¼ 0/ml pDC40/ml P-value Parameters at initial diagnosis Immunophenotype (common vs pra¨B) (%) Peripheral WBC count (mean) (  103/ml) Hemoglobin (mean) (g/dl) Platelet count (mean) (  103/ml) Peripheral blast count (mean) (%) Blasts in bone marrow (mean) (%) Lactat dehydrogenase (mean) (U/l) Expression of myeloid antigens on blasts (%)

92.9 vs 7.1 56.9 7.10 64.6 51.1 92.9 809 57.1

Parameters during therapy Peripheral blast count day 8 (mean) (  103/ml) Prednisone good response (%)

0.25 92.9

47.6 vs 52.4 107.2 7.87 51.7 80.4 92.3 2405 19.0 1.05 85.7

0.01 0.01 0.59 0.48 0.01 0.89 0.03 0.03 0.77 0.64

69.2 vs 30.8 106.8 8.48 52.3 74.5 94.7 1982 30.8 0.35 76.9

60.9 vs 39.1 73.2 7.03 57.4 4.1 90.2 1364 34.8 0.81 95.7

0.73 0.19 0.18 0.50 0.16 0.24 0.67 1.00 0.62 0.12

(b) Spearman’s Rho correlation of percentage of blasts expressing myeloid markers CD13, CD33 and myeloperoxidase with mDC and pDC numbers mDC Spearman P-value pDC Spearman P-value patients (n) coefficient patients (n) coefficient Expression of CD13 on blasts Expression of CD33 on blasts Expression of MPO7 in blasts

26 26 20

0.55 0.56 0.54

0.003 0.003 0.014

27 27 21

0.07 0.17 0.03

0.73 0.40 0.29

Abbreviations: mDC, myeloid dendritic cells; pDC, plasmacytoid dendritic cells. Bold values indicate a significant difference (Po0.05). Leukemia

Dendritic cell deficiencies in pediatric ALL patients B Maecker et al

648

Figure 3 Phenotypic analysis of surface molecules on dendritic cells from pT-ALL patients and healthy children. Mean fluorescence intensity of CD40, CD83 and CD86 was determined on myeloid DC (mDC) (a) and plasmacytoid DC (pDC) (b) using flow cytometry. (c) Subgroups of mDC expressing BDCA1 or BDCA3 were compared in pT-ALL patients and healthy children; mean percentage of positive mDC is shown.

Interestingly, aberrant expression of myeloid surface markers on leukemic blasts was significantly more frequent in patients lacking mDC in peripheral blood (P ¼ 0.03; Table 2a). In particular, increasing percentage of blasts expressing CD13, CD33 and, to a lesser extent, myeloperoxidase was strongly associated with absence of mDC in pB-ALL patients (P ¼ 0.003 and 0.014, respectively; Table 2b). This unexpected finding raises the hypothesis that in some cases of pB-ALL the physiological differentiation of a common myeloid progenitor cell might be blocked and differentiation might be skewed toward leukemia development. Furthermore, statistical analysis showed that patients with high peripheral leukocyte counts and high peripheral blast counts had significantly more mDC in the peripheral blood (P ¼ 0.01; Table 2a). This phenomenon may just be related to high general cell turnover, as these patients also had higher levels of lactate dehydrogenase. However, the analysis did not reveal any statistically significant relation between low numbers of DC and response to initial prednisone treatment (Table 2a). Similarly, there was no association to stratification into risk groups by minimal residual disease evaluation, and time of follow-up is currently too short to assess disease-free and overall survival.

Discussion Dendritic cells play a critical role in the generation of antitumor immunity. Several studies have demonstrated numerical as well as functional abnormalities of DC in patients with cancer (for review see Gabrilovich18), in particular acute leukemias.11,14 Among functional defects, there are decreased expression of MHC or co-stimulatory molecules,19,20 insufficient T-cell stimulatory capacity20,21 and accumulation of immature DC in cancer patients.21 As patients with ALL display particularly poor antitumor immune responses,6 we sought to investigate the DC compartment in pediatric patients. Leukemia

Here, we report results of a study investigating frequency and phenotype of pbDCs in pediatric ALL patients and healthy agematched controls. Whereas patients with T-precursor ALL have elevated numbers of pbDC expressing normal levels of MHC, costimulatory, and activation markers, patients with B-precursor ALL show decreased or absent pbDC numbers. This dichotomy between B- and T-cell ALL clearly illustrates that B-lineage and T-lineage ALL are characterized by distinct tumor–host characteristics, a finding that has also been revealed by the analysis of genetic polymorphisms in cytokine genes.22 Decreased numbers of DC could not be explained by mere bone marrow dysfunction owing to infiltration of leukemic blast cells, as we were unable to document a correlation between DC numbers and other myeloid cell numbers such as neutrophils and monocytes. Our data are in accordance with observations published by Mami et al.,14 who detected normal numbers of DC in adult and few pediatric pT-ALL patients but significantly reduced DC numbers in pB-ALL patients. This group was able to demonstrate apparently normal cytokine-secreting function of DC in T-ALL patients, which is paralleled by our findings of unchanged phenotypic characteristics of DC in pediatric pT-ALL patients. Several lines of evidence demonstrate that lack of DC in pBALL patients is not merely a result of a general hypoplasia of bone marrow cell production: patients with lack of mDC or pDC did not differ from those with measurable DC with respect to neutrophil or platelet counts, hemoglobin levels or blast infiltration in the bone marrow. Interestingly, Mami et al.14 were unable to generate either mDC or pDC from CD34 þ precursor cells from pB-ALL patients. In our studies, lack of mDC was significantly associated with aberrant expression of myeloid markers CD13, CD33, and myeloperoxidase on Bprecursor leukemic blasts. Moreover, the percentage of blasts expressing CD13, CD33, and myeloperoxidase, respectively, was inversely correlated with numbers of mDC in pB-ALL patients. Even though decreased numbers of DC were not due to global bone marrow dysfunction, this finding is consistent with the hypothesis that in some patients with B-precursor ALL leukemic cells or leukemia-derived factors may contribute to a block in mDC development. Furthermore, lack of mDC was more frequent in patients with an immunophenotype of common ALL, which is thought to arise from a more immature precursor than pce B ALL. In acute myeloid leukemia, Mohty11 and colleagues demonstrated leukemia-associated cytogenetic alterations in pbDCs, a finding consistent with a common developmental origin of DCs and leukemia cells. Further confirmatory experimental studies are necessary to examine whether similar mechanisms may be operative in ALL patients. In summary, we provide evidence for distinct DC aberrations in patients with ALL at time of diagnosis, bearing relevance for a refined understanding of tumor immunity and future development of immunotherapeutic strategies. In concert with other observations, such as cytokine gene polymorphisms in IFN-g22 and IL-10 genes,23 this finding substantiates the concept of immunosurveillance in B-precursor ALL.

Acknowledgements We are indebted to our patients and healthy blood donors, their parents and physicians for the commitment to this study. We thank A Frank-Hoppe for excellent study support. This work was supported by a grant of the Kind-Philipp-Stiftung fu¨r Leuka¨mieforschung (BM, DM) and Hannover Medical School (HiLF).

Dendritic cell deficiencies in pediatric ALL patients B Maecker et al

649 References 1 Schrappe M, Camitta B, Pui CH, Eden T, Gaynon P, Gustafsson G et al. Long-term results of large prospective trials in childhood acute lymphoblastic leukemia. Leukemia 2000; 14: 2193–2194. 2 Schrappe M, Reiter A, Ludwig WD, Harbott J, Zimmermann M, Hiddemann W et al. Improved outcome in childhood acute lymphoblastic leukemia despite reduced use of anthracyclines and cranial radiotherapy: results of trial ALL-BFM 90. German–Austrian–Swiss ALL-BFM Study Group. Blood 2000; 95: 3310–3322. 3 LeClerc JM, Billett AL, Gelber RD, Dalton V, Tarbell N, Lipton JM et al. Treatment of childhood acute lymphoblastic leukemia: results of Dana-Farber ALL Consortium Protocol 87-01. J Clin Oncol 2002; 20: 237–246. 4 Woolfrey AE, Anasetti C, Storer B, Doney K, Milner LA, Sievers EL et al. Factors associated with outcome after unrelated marrow transplantation for treatment of acute lymphoblastic leukemia in children. Blood 2002; 99: 2002–2008. 5 Gustafsson Jernberg A, Remberger M, Ringden O, Winiarski J. Graft-versus-leukaemia effect in children: chronic GVHD has a significant impact on relapse and survival. Bone Marrow Transplant 2003; 31: 175–181. 6 Cardoso AA, Seamon MJ, Afonso HM, Ghia P, Boussiotis VA, Freeman GJ et al. Ex vivo generation of human anti-pre-B leukemia-specific autologous cytolytic T cells. Blood 1997; 90: 549–561. 7 Cardoso AA, Veiga JP, Ghia P, Afonso HM, Haining WN, Sallan SE et al. Adoptive T-cell therapy for B-cell acute lymphoblastic leukemia: preclinical studies. Blood 1999; 94: 3531–3540. 8 Stripecke R, Skelton DC, Gruber T, Afar D, Pattengale PK, Witte ON et al. Immune response to Philadelphia chromosome-positive acute lymphoblastic leukemia induced by expression of CD80, interleukin 2, and granulocyte–macrophage colony-stimulating factor. Hum Gene Ther 1998; 9: 2049–2062. 9 Auffermann-Gretzinger S, Keeffe EB, Levy S. Impaired dendritic cell maturation in patients with chronic, but not resolved, hepatitis C virus infection. Blood 2001; 97: 3171–3176. 10 Chehimi J, Campbell DE, Azzoni L, Bacheller D, Papasavvas E, Jerandi G et al. Persistent decreases in blood plasmacytoid dendritic cell number and function despite effective highly active antiretroviral therapy and increased blood myeloid dendritic cells in HIV-infected individuals. J Immunol 2002; 168: 4796–4801. 11 Mohty M, Jarrossay D, Lafage-Pochitaloff M, Zandotti C, Briere F, de Lamballeri XN et al. Circulating blood dendritic cells from myeloid leukemia patients display quantitative and cytogenetic

12

13

14 15

16 17

18 19 20

21 22

23

abnormalities as well as functional impairment. Blood 2001; 98: 3750–3756. Mohty M, Isnardon D, Vey N, Briere F, Blaise D, Olive D et al. Low blood dendritic cells in chronic myeloid leukaemia patients correlates with loss of CD34+/CD38 primitive haematopoietic progenitors. Br J Haematol 2002; 119: 115–118. Bourguin-Plonquet A, Rouard H, Roudot-Thoraval F, Bellanger C, Marquet J, Delfau-Larue MH et al. Severe decrease in peripheral blood dendritic cells in hairy cell leukaemia. Br J Haematol 2002; 116: 595–597. Mami NB, Mohty M, Chambost H, Gaugler B, Olive D. Blood dendritic cells in patients with acute lymphoblastic leukaemia. Br J Haematol 2004; 126: 77–80. Vakkila J, Thomson AW, Vettenranta K, Sariola H, SaarinenPihkala UM. Dendritic cell subsets in childhood and in children with cancer: relation to age and disease prognosis. Clin Exp Immunol 2004; 135: 455–461. Teig N, Moses D, Gieseler S, Schauer U. Age-related changes in human blood dendritic cell subpopulations. Scand J Immunol 2002; 55: 453–457. Dzionek A, Fuchs A, Schmidt P, Cremer S, Zysk M, Miltenyi S et al. BDCA-2, BDCA-3, and BDCA-4: three markers for distinct subsets of dendritic cells in human peripheral blood. J Immunol 2000; 165: 6037–6046. Gabrilovich D. Mechanisms and functional significance of tumour-induced dendritic-cell defects. Nat Rev Immunol 2004; 4: 941–952. Troy AJ, Summers KL, Davidson PJ, Atkinson CH, Hart DN. Minimal recruitment and activation of dendritic cells within renal cell carcinoma. Clin Cancer Res 1998; 4: 585–593. Nestle FO, Burg G, Fah J, Wrone-Smith T, Nickoloff BJ. Human sunlight-induced basal-cell-carcinoma-associated dendritic cells are deficient in T cell co-stimulatory molecules and are impaired as antigen-presenting cells. Am J Pathol 1997; 150: 641–651. Gabrilovich DI, Corak J, Ciernik IF, Kavanaugh D, Carbone DP. Decreased antigen presentation by dendritic cells in patients with breast cancer. Clin Cancer Res 1997; 3: 483–490. Cloppenborg T, Stanulla M, Zimmermann M, Schrappe M, Welte K, Klein C. Immunosurveillance of childhood ALL: polymorphic interferon-gamma alleles are associated with age at diagnosis and clinical risk groups. Leukemia 2005; 19: 44–48. Lauten M, Matthias T, Stanulla M, Beger C, Welte K, Schrappe M. Association of initial response to prednisone treatment in childhood acute lymphoblastic leukaemia and polymorphisms within the tumour necrosis factor and the interleukin-10 genes. Leukemia 2002; 16: 1437–1442.

Leukemia