With contemporary treatment, 80% to 90% of children with acute

Transcription

1 157 Prognostic Factors and Outcome of Recurrence in Childhood Acute Myeloid Leukemia Jeffrey E. Rubnitz, MD, PhD 1,2 Bassem I. Razzouk, MD 1,2 Shelly Lensing, MS 3 Stanley Pounds, PhD 3 Ching-Hon Pui, MD 1,2,4 Raul C. Ribeiro, MD 1,2 1 Department of Oncology, St. Jude Children s Research Hospital, Memphis, Tennessee. 2 Department of Pediatrics, College of Medicine, University of Tennessee Health Science Center, Memphis, Tennessee. 3 Department of Biostatistics, St. Jude Children s Research Hospital, Memphis, Tennessee. 4 Department of Pathology, St. Jude Children s Research Hospital, Memphis, Tennessee. BACKGROUND. Outcome after recurrence of childhood acute myeloid leukemia (AML) is poor. We performed this study to identify prognostic factors for recurrence and for survival after recurrence of AML. METHODS. The clinical characteristics, biological features, treatment modalities, and outcomes of children with de novo AML who were enrolled on 3 consecutive clinical protocols from 1987 to 2002 at St. Jude Children s Research Hospital were studied. Regression modeling was used to identify prognostic factors for recurrence and for survival after recurrence. RESULTS. The outcome after recurrence was poor, with a 5-year survival estimate of only 23.3% 6 5.7%. Multivariable analysis indicated that male sex (P ¼.005), autologous stem cell transplant before recurrence (P ¼.097), each additional month from diagnosis to recurrence (P ¼.041), and stem cell transplant after recurrence (P <.001) were associated with a better survival after recurrence, whereas M5 or M7 morphology (P ¼.001) were significantly predictive of a lower survival estimate after recurrence. CONCLUSION. Survival after recurrence was poor in children with AML. Novel therapies are urgently needed to prevent or to treat recurring AML. Cancer 2007;109: Ó 2006 American Cancer Society. Supported in part by Cancer Center Support (Core) grant P30 CA from the National Institutes of Health by a Center of Excellence Grant from the State of Tennessee, and by the American Lebanese Syrian Associated Charities (ALSAC). C.-H. Pui is an American Cancer Society Professor. Presented, in part, at the Annual Meeting of the American Society of Hematology, December 12, 2005, Atlanta, GA. We thank Sharon Naron for expert editorial review and Ramona Ratliff for assistance with data collection. Address for reprints: Jeffrey E. Rubnitz, MD, PhD, Department of Oncology, Mail Stop 260, St. Jude Children s Research Hospital, 332 N. Lauderdale St., Memphis, TN ; Fax: (901) ; jeffrey.rubnitz@stjude.org Received September 6, 2006; revision received October 9, 2006; accepted October 11, KEYWORDS: acute myeloid leukemia, recurrence, childhood. With contemporary treatment, 80% to 90% of children with acute myeloid leukemia (AML) experience complete remission (CR), but 30% to 40% of these patients subsequently suffer recurrence, and the long-term survival rate is only about 50%. 1 7 After recurrence, the likelihood of survival is poor, ranging from 21% to 33% in recent reports In these reports the length of first remission was the best predictor of survival It is not clear whether other factors are prognostically important after recurrence. We reviewed the outcome of all patients treated on successive AML clinical trials at our institution from 1987 to 2002 to investigate factors predictive of hematologic recurrence, second remission, or overall survival (OS) after recurrence. MATERIALS AND METHODS Patients From 1987 through 2002, 191 patients with de novo AML who were 21 years of age or younger (excluding those with Down syndrome or acute promyelocytic leukemia) were enrolled on 3 successive institutional clinical trials (AML-87, AML-91, and AML-97). 6 The median time since last follow-up for surviving patients was 0.6 years (range, years) at the time of this study, and 72.9% had received follow-up during the past year. Of the 160 patients who attained CR, 60 suffered hematologic ª 2006 American Cancer Society DOI /cncr Published online 28 November 2006 in Wiley InterScience (

2 158 CANCER January 1, 2007 / Volume 109 / Number 1 TABLE 1 Univariate Analysis of Factors Prognostic of Event-Free Survival (EFS) Characteristics at diagnosis n Percentage 5-y EFS 6 SE P* Sex Male Female Age, y 0-< Race White Black Other Leukocyte count < /L /L Platelet count < /L /L FAB M M M M M Other y Cytogenetics t(9,11) <.001 inv(16) t(8;21) Other 11q23/MLL Other Study AML AML AML SE indicates standard error; FAB, French-American-British; AML, acute myeloid leukemia. * Exact log-rank test. y M0, M6, and unclassifiable. or combined recurrences and are the subject of this report. Patients with isolated extramedullary recurrence were not included in analyses of outcome after recurrence. Written informed consent was obtained from the legal guardians or patients (as appropriate), and all studies (therapeutic and diagnostic) were approved by our Institutional Review Board. Diagnoses were based on standard techniques and cases were classified by morphologic, immunologic, and genetic testing. TABLE 2 Multivariable Analysis of Factors Prognostic of Event-Free Survival Variable* Hazard ratio y 95% Confidence limits Age 10 years Leukocyte count /L FAB M <.001 Favorable karyotype { AML87 protocol AML91 protocol FAB indicates French-American-British; AML, acute myeloid leukemia. * Reference categories age <10 years, leukocyte count < /L, all other FAB categories combined for M7 FAB, all other categories combined for favorable karyotype, and AML97 protocol. y A hazard ratio greater than 1 is unfavorable, indicating increased risk of recurrence or death. { Favorable karyotype includes t(9;11) and inv(16). Statistical Methods The method of Kaplan and Meier 13 was used to estimate the probability of event-free survival (EFS) and to estimate the probability of survival of patients after recurrence. Standard errors were calculated by the method of Peto et al. 14 The duration of EFS was calculated from the on-study date to the date of recurrence or death from any cause. The EFS was recorded as 0 for patients who did not achieve a CR and was censored at the date of last follow-up for those who were alive and in remission. The duration of postrecurrence OS was calculated from the date of recurrence to the date of death and was censored at the date of last follow-up for surviving patients. Various potential prognostic factors were compared by using exact log-rank analyses. The independent effect of prognostic variables was investigated by using proportional hazards modeling. 15 Age, hemoglobin concentration, platelet count, and leukocyte count were investigated as continuous and categorical variables in separate models. Given the number of variables under consideration, stepwise variable selection was performed using SAS Release 9.1 software (SAS Institute, Cary, NC) to select variables included in the models. For variable selection, the variable entry and removal significance levels were To identify prognostic factors for a second remission, exact x 2 tests were used for univariate comparisons. Logistic regression modeling was then used to explore joint effects with stepwise variable selection similar to that described above. The final model was fitted using exact logistic modeling methods as implemented in SAS Release 9.1. P-values <.05 were considered to indicate a statistically significant association, and all P-values were 2-sided. No adjustment was made for multiple comparisons. RESULTS Response to Initial Therapy The characteristics of the 191 patients are shown in Table 1. Overall, 160 (84%) patients attained CR and the 5-year EFS estimate (6SE) was 41.8% 6 3.9%. In P

3 Outcome After Recurrence of Childhood AML/Rubnitz et al. 159 TABLE 3 Univariate Analysis of Factors Prognostic of Second Remission and Survival After Recurrence Characteristics at diagnosis and treatment No. No second remission N ¼ 21 No. (%) Second remission N ¼ 39 No. (%) P* Percentage 5-y OS 6 SE P y Sex Male 30 5 (16.7) 25 (83.3) Female (53.3) 14 (46.7) Age, y 0-< (35.1) 24 (64.9) (34.8) 15 (65.2) Race White (31.6) 26 (68.4) Black 14 6 (42.9) 8 (57.1) Other 8 3 (37.5) 5 (62.5) Leukocyte count, 10 9 /L < (39.0) 25 (61.0) (26.3) 14 (73.7) Platelets, 10 9 /L < (27.3) 16 (72.7) (39.5) 23 (60.5) Hgb, 10 9 /L < (38.9) 11 (61.1) (28.1) 23 (71.9) > (55.6) 4 (44.4) CNS status CNS3, 5 WBC/mL with blasts 7 2 (28.6) 5 (71.4) CNS2, <5 WBC/mL with blasts 12 5 (41.7) 7 (58.3) CNS1, no blasts (38.7) 19 (61.3) Traumatic 8 2 (25.0) 6 (75.0) FAB subtype M (10.0) 9 (90.0).045 { { M (34.6) 17 (65.4) M4 8 2 (25.0) 6 (75.0) M5 9 4 (44.4) 5 (55.6) M7 6 5 (83.3) 1 (16.7) 0 Other { 1 0 (0.0) 1 (100.0) Cytogenetic features t(9,11) or inv(16) 3 1 (33.3) 2 (66.6) t(8;12) 19 5 (26.3) 14 (73.7) Other 11q23/MLL 11 4 (36.4) 7 (63.6) Other (40.7) 16 (59.3) Study AML (26.7) 11 (73.3) AML (36.4) 14 (63.6) AML (39.1) 14 (60.9) Treatment prior to recurrence Chemotherapy only 31 8 (25.8) 23 (74.2) Autotransplant 20 6 (30.0) 14 (70.0) Allotransplant 9 7 (77.8) 2 (22.2) 0 Treatment after relapse Palliative/none 4 0 <.001 Chemotherapy 17 0 Transplant Time from diagnosis to recurrence Recurrence <1 y (43.2) 21 (56.8) Recurrence 1y 23 5(21.7) 18(78.3) OS indicates overall survival; SE, standard error; CNS, central nervous system; WBC, white blood cell count; FAB, French-American-British; AML, acute myeloid leukemia. * Exact x 2 test. y Exact log-rank test. { Other category is 1 M0 patient, who was not included in the calculation of P. This patient was alive at 12.6 years after recurrence.

4 160 CANCER January 1, 2007 / Volume 109 / Number 1 TABLE 4 Multivariable Analysis of Factors Prognostic of Second Remission and Survival After Recurrence Second remission variables* Odds ratio y 95% Confidence limits P Male M Allogeneic transplant before relapse Overall Survival Variables { Hazard Ratio y 95% Confidence Limits P Male FAB M5 or M Autologous transplant before relapse Transplant after relapse <.001 Months to relapse from diagnosis FAB indicates French-American-British. * Reference categories are female, all other FAB categories combined for M1 FAB, chemotherapy only or autologous transplant for therapy prior to relapse. y An odds ratio greater than 1 is favorable, indicating increased odds of achieving second remission. A hazard ratio greater than 1 is unfavorable, indicating increased risk of death. { Reference categories are female, all other FAB categories combined for M5 and M7 FAB group, chemotherapy only and allogeneic transplant for therapy prior to relapse, and palliative care/none or chemotherapy only for therapy after recurrence. For months to recurrence from diagnosis, every month from diagnosis conveyed a 3% (95% CI, 0.1 6%) reduction in the risk of death. univariate analysis, age, French-American-British (FAB) subtype, and karyotype were predictive of outcome (Table 1). Proportional hazards modeling (Table 2) demonstrated that leukocyte count /L at diagnosis (5-year EFS 6 SE, 34.7% 6 7.2%), FAB M7 subtype (5- year EFS 6 SE, 20.8% 6 9.3%), and age older than 10 years (5-year EFS 6 SE, 32.4% 6 5.8%) each conferred an increased risk of recurrence or death, whereas a favorable karyotype (5-year EFS 6 SE, 71.1% 6 8.0%) conferred a decreased risk that was statistically significant after adjustment for other variables. For the purposes of this study, favorable karyotype was defined as t(9;11) or inv(16) based on the excellent outcomes for patients with either of these characteristics (Table 1). Outcome After Recurrence Of the 160 patients who attained a CR, 60 suffered hematologic recurrences at a median of 9.8 months (range, months) after diagnosis. Thirty-seven (62%) patients had a recurrence younger than 1 year after diagnosis and 23 (38%) experienced recurrence later. After recurrence, 4 patients received palliative care, 17 received chemotherapy alone, and 39 underwent hematopoietic stem cell transplant (HSCT). The 17 patients who received only chemotherapy did not undergo stem cell transplant because of a lack of a suitable donor or because of progressive disease before FIGURE 1. Survival after recurrence according to treatment before recurrence. The 5-year survival estimates (6standard error [SE]) were 45% 6 11% for patients who had recurrence after undergoing autologous stem cell transplantation in first remission, 16% 6 7% for patients who had recurrence after receiving only chemotherapy, and 0% for patients who received allogeneic stem cell transplantation in first remission. transplantation could be performed. To prevent any bias that might be introduced by removing the 4 patients who received only palliative care, most of whom had very aggressive disease after recurrence, all analyses described below included these patients. Notably, these results were similar to those of separate analyses that excluded these patients (data not shown). Thirty-nine (65%) of the patients attained a second remission. At the time of last follow-up, 14 patients were alive, 30 died of progressive disease, and 16 died of regimen-related toxicity (14 after stem cell transplant). The 5-year OS estimate 6 SE after recurrence for the entire cohort of 60 patients was 23.3% 6 5.7%. In a univariate analysis (Table 3), factors associated with attainment of second remission were sex (males, 83%; females, 47%; P ¼.006),thetypeofinitialtherapy (chemotherapy alone, 74%; autologous transplant, 70%; allogeneic stem cell transplant, 22%; P ¼.012), and FAB subtype (eg, FAB M7, 17%). Logistic regression analysis (Table 4) demonstrated that male sex (P ¼.005) and M1 subtype (P ¼.056) were associated with attainment of CR2, whereas allogeneic stem cell transplant during first remission (P ¼.023) was associated with failure to achieve second remission. Univariate analysis of survival after recurrence indicated significant differences according to sex (P ¼.002), type of treatment before recurrence (P ¼.020, Fig. 1), and type of treatment after recurrence (P<.001, Table 3). There was a trend toward improved survival among patients who had a recurrence older than 1 year from diagnosis as compared with those who had recurrence earlier(5-yearos6 SE, 26.1% 6 8.5% vs 21.6% 6 7.2%,

5 Outcome After Recurrence of Childhood AML/Rubnitz et al. 161 P ¼.112). There were no survivors after recurrence among patients with M7 AML or patients who underwent allogeneic stem cell transplant in first remission. Relatively good outcomes after recurrence were observed among patients who received autologous transplant during first remission (5-year OS 6 SE, 45.0% %) and among patients who underwent stem cell transplant after recurrence (5-year OS 6 SE, 35.9% 6 8.0%). In a multivariable Cox regression model (Table 4), M5 or M7 morphology (P ¼.001) was associated with a lower survival estimate after recurrence. In contrast, male sex (P ¼.005), autologous HSCT before recurrence (P ¼.097), HSCT after recurrence (P <.001), and each additional month from diagnosis to recurrence (P ¼.041) were significantly predictive of improved survival. Among the 14 patients who had experienced recurrence and who survived at the time of this report, all but 2 had significant late effects of therapy. Nine patients have endocrine disorders, including growth hormone deficiency in 6, gonadal dysfunction in 2, and hypothyroidism in 7. Two patients have seizures or encephalopathy, 4 patients have cataracts, 4 have school difficulties, 3 have chronic renal insufficiency, and 1 has hepatitis C. DISCUSSION We previously demonstrated that among patients with AML treated at St. Jude, the t(9;11) conferred a favorable outcome, 16 the t(8;21), an intermediate outcome, 17 and M7 morphology, a dismal outcome. 18 Inv(16) was associated with a good prognosis only during the recent era. 19 In the present study we sought to update and expand our analysis of prognostic factors for recurrence and to identify factors predictive of outcome after recurrence. In this study, better outcomes were observed among patients younger than 10 years old and among patients with the t(9;11) or inv(16), whereas an initial leukocyte count greater than /L, M7 morphology, and treatment on an earlier protocol were associated with a greater risk of recurrence. After recurrence, about two-thirds of patients achieved a second remission. Patients with M7 AML and patients who underwent allogeneic stem cell transplant during first remission had particularly low second remission rates of only 17% and 22%, respectively. Overall, 23% of patients who had recurrence are long-term survivors, similar to recent reports In concordance with their low rates of second remission, patients with M7 AML and patients who underwent allogeneic stem cell transplant during first remission had dismal outcomes after recurrence. Although survival estimates were statistically better in certain subgroups, including males, patients who had late recurrences, and patients who underwent stem cell transplant after recurrence, outcomes were clinically unsatisfactory in all groups. For example, only 45% of patients who underwent allogeneic stem cell transplant after recurrence are longterm survivors. In addition, serious late effects were observed in the majority of survivors. What is the likelihood of survival after recurrence in AML? Investigators in the Berlin-Frankfurt-Munster (BFM) study group reported a 51% second remission rate and a 5-year survival estimate of 21% for children with AML after first recurrence. 9 In that report, patients whose recurrence occurred less than 1.5 years after diagnosis had a 5-year survival estimate of only 10%, compared with 40% for those whose recurrence occurred later. In a report from the Children s Cancer Group, the survival after bone marrow recurrence was 21% overall and only 8% in patients who had recurrence less than 1 year from diagnosis. 12 Similarly, the 3-year survival estimate after recurrence was 24% for children treated on the Medical Research Council AML10 trial but was only 11% for children whose initial remission lasted less than 1 year. 8 Children whose recurrences occurred later fared significantly better, with a 3-year survival estimate of 49%. Recently, the Leucamie Aique Myeloide Enfant study group reported a 5-year survival estimate of 33% and confirmed the prognostic impact of time to recurrence. 10 Survival rates were 24% for children whose first remission lasted less than 12 months and 54% for children with longer first remissions. The heterogeneity of our patient population, the retrospective nature of our study, and the variety of chemotherapy regimens used over a 15-year period preclude our assessment of the efficacy of any 1 treatment regimen. Many of the patients in the present study received cladribine alone 20,21 or in combination with cytarabine 22 or topotecan. Despite the activity of cladribine against recurring AML, 21 the combination of cladribine and cytarabine was ineffective at inducing second complete remission in a Phase II study. 22 The use of this regimen may have contributed to our low second remission rate. On the basis of these results, our current treatment strategies for recurring AML no longer include cladribine. In contrast, the Children s Cancer Group reported a 76% second remission rate for patients with refractory or recurring AML who were treated with mitoxantrone and cytarabine. 11 However, even with the excellent second remission rate, the 2- year survival estimate was only 24%. 11 In the present study, as well as in previous studies, 8 11 second recurrence and progressive disease were the main causes of treatment failure and death, but regimen-related toxicity after recurrence was a significant problem. Support-

6 162 CANCER January 1, 2007 / Volume 109 / Number 1 ive care measures currently in place at our institution, including the use of prophylactic antibiotics and prophylactic voriconazole in all patients with AML, have reduced complications related to infection in newly diagnosed patients with AML. We hope that this finding will extend to recurring AML as well. In the present study, even patients who underwent allogeneic stem cell transplant after recurrence had an unsatisfactory outcome. Most patients underwent transplantation from a matched sibling or matched unrelated donor, as we were not routinely performing haploidentical stem cell transplants during the time period of this retrospective study. Less than half of patients who underwent transplant survived, and most of the survivors had significant late effects. Because of the small numbers of patients and because of changes in conditioning regimens during this study, we cannot comment on differences in graft versus leukemia effects between matchedsiblingandmatchedunrelateddonortransplants. Nevertheless, it seems unlikely that further progress can be made by using conventional chemotherapy or stem cell transplantation alone. Therapies under development include proteasome inhibitors, 23,24 histone deacetylase inhibitors, 25,26 and agents that target leukemia-specific abnormalities, such as constitutively activated tyrosine kinases Recently, cellular therapy with haploidentical natural killer cells has also been shown to exert antitumor activity with minimal toxicity in patients with recurrence AML. 30 We are currently testing the tandem application of haploidentical stem cell transplantation followed by natural killer cell transplantation for patients with high-risk leukemia. Clearly, novel therapies are urgently needed, both to prevent recurrence and to treat patients with recurring AML. REFERENCES 1. Liang DC, Chang TT, Lin KH, et al. Improved treatment results for childhood acute myeloid leukemia in Taiwan. Leukemia. 2006;20: Creutzig U, Zimmermann M, Ritter J, et al. Treatment strategies and long-term results in paediatric patients treated in four consecutive AML-BFM trials. Leukemia. 2005;19: Smith FO, Alonzo TA, Gerbing RB, Woods WG, Arceci RJ. Long-term results of children with acute myeloid leukemia: a report of three consecutive Phase III trials by the Children s Cancer Group: CCG 251, CCG 213 and CCG Leukemia. 2005;19: Lie SO, Abrahamsson J, Clausen N, et al. Long-term results in children with AML: NOPHO-AML Study Group report of three consecutive trials. Leukemia. 2005;19: Ravindranath Y, Chang M, Steuber CP, et al. Pediatric Oncology Group (POG) studies of acute myeloid leukemia (AML): a review of four consecutive childhood AML trials conducted between 1981 and Leukemia. 2005;19: Ribeiro RC, Razzouk BI, Pounds S, Hijiya N, Pui CH, Rubnitz JE. Successive clinical trials for childhood acute myeloid leukemia at St Jude Children s Research Hospital, from 1980 to Leukemia. 2005;19: Gibson BE, Wheatley K, Hann IM, et al. Treatment strategy and long-term results in paediatric patients treated in consecutive UK AML trials. Leukemia. 2005;19: Webb DK, Wheatley K, Harrison G, Stevens RF, Hann IM. Outcome for children with relapsed acute myeloid leukaemia following initial therapy in the Medical Research Council (MRC) AML 10 trial. MRC Childhood Leukaemia Working Party. Leukemia. 1999;13: Stahnke K, Boos J, Bender-Gotze C, Ritter J, Zimmermann M, Creutzig U. Duration of first remission predicts remission rates and long-term survival in children with relapsed acute myelogenous leukemia. Leukemia. 1998;12: Aladjidi N, Auvrignon A, Leblanc T, et al. Outcome in children with relapsed acute myeloid leukemia after initial treatment with the French Leucemie Aique Myeloide Enfant (LAME) 89/ 91 protocol of the French Society of Pediatric Hematology and Immunology. JClinOncol. 2003;21: Wells RJ, Adams MT, Alonzo TA, et al. Mitoxantrone and cytarabine induction, high-dose cytarabine, and etoposide intensification for pediatric patients with relapsed or refractory acute myeloid leukemia: Children s Cancer Group Study JClinOncol. 2003;21: Johnston DL, Alonzo TA, Gerbing RB, Lange BJ, Woods WG. Risk factors and therapy for isolated central nervous system relapse of pediatric acute myeloid leukemia. JClinOncol. 2005;23: Kaplan EL, Meier P. Non-parametric estimation for incomplete observations. Am Stat Assoc. 1958;53: Peto R, Pike MC, Armitage P, et al. Design and analysis of randomized clinical trials requiring prolonged observation of each patient. II. Analysis and examples. Br J Cancer. 1977; 35: Cox DR. Regression models and life-tables. J Res Stat Soc B. 1972;34: Rubnitz JE, Raimondi SC, Tong X, et al. Favorable impact of the t(9;11) in childhood acute myeloid leukemia. JClinOncol. 2002;20: Rubnitz JE, Raimondi SC, Halbert AR, et al. Characteristics and outcome of t(8;21)-positive childhood acute myeloid leukemia: a single institution s experience. Leukemia. 2002;16: Athale UH, Razzouk BI, Raimondi SC, et al. Biology and outcome of childhood acute megakaryoblastic leukemia: a single institution s experience. Blood. 2001;97: Razzouk BI, Raimondi SC, Srivastava DK, et al. Impact of treatment on the outcome of acute myeloid leukemia with inversion 16: a single institution s experience. Leukemia. 2001; 15: Santana VM, Mirro J Jr, Harwood FC, et al. A phase I clinical trial of 2-chlorodeoxyadenosine in pediatric patients with acute leukemia. JClinOncol. 1991;9: Santana VM, Mirro J Jr, Kearns C, Schell MJ, Crom W, Blakley RL. 2-Chlorodeoxyadenosine produces a high rate of complete hematologic remission in relapsed acute myeloid leukemia. JClinOncol. 1992;10: Rubnitz JE, Razzouk BI, Srivastava DK, Pui CH, Ribeiro RC, Santana VM. Phase II trial of cladribine and cytarabine in relapsed or refractory myeloid malignancies. Leuk Res. 2004; 28: Guzman ML, Swiderski CF, Howard DS, et al. Preferential induction of apoptosis for primary human leukemic stem cells. Proc Natl Acad Sci U S A. 2002;99:

7 Outcome After Recurrence of Childhood AML/Rubnitz et al Adams J. The proteasome: a suitable antineoplastic target. Nat Rev Cancer. 2004;4: Insinga A, Monestiroli S, Ronzoni S, et al. Inhibitors of histone deacetylases induce tumor-selective apoptosis through activation of the death receptor pathway. Nat Med. 2005;11: Nebbioso A, Clarke N, Voltz E, et al. Tumor-selective action of HDAC inhibitors involves TRAIL induction in acute myeloid leukemia cells. Nat Med. 2005;11: Levis M, Allebach J, Tse KF, et al. A FLT3-targeted tyrosine kinase inhibitor is cytotoxic to leukemia cells in vitro and in vivo. Blood. 2002;99: Brown P, Meshinchi S, Levis M, et al. Pediatric AML primary samples with FLT3/ITD mutations are preferentially killed by FLT3 inhibition. Blood. 2004;104: Smith BD, Levis M, Beran M, et al. Single-agent CEP-701, a novel FLT3 inhibitor, shows biologic and clinical activity in patients with relapsed or refractory acute myeloid leukemia. Blood. 2004;103: Miller JS, Soignier Y, Panoskaltsis-Mortari A, et al. Successful adoptive transfer and in vivo expansion of human haploidentical NK cells in patients with cancer. Blood. 2005;105: