ALBERT

Description: Key Points A comprehensive study of 19 gene mutations and their cooperation, including the first report of ASXL1 and TET2 mutations in pediatric AML. The development of pediatric AML requires fewer gene mutations than adult AML.

Description: Contemporary treatment of pediatric acute lymphoblastic leukemia (ALL) requires the assignment of patients to specific risk groups. We have recently demonstrated that expression profiling of leukemic blasts can accurately identify the known prognostic subtypes of ALL, including T-cell lineage ALL (T-ALL), E2A-PBX1, TEL-AML1, MLL rearrangements, BCR-ABL, and hyperdiploid karyotypes with more than 50 chromosomes. As the next step toward developing this methodology into a frontline diagnostic tool, we have now analyzed leukemic blasts from 132 diagnostic samples using higher density oligonucleotide arrays that allow the interrogation of most of the identified genes in the human genome. Nearly 60% of the newly identified subtype discriminating genes are novel markers not identified in our previous study, and thus should provide new insights into the altered biology underlying these leukemias. Moreover, a proportion of the newly selected genes are highly ranked as class discriminators, and when incorporated into class-predicting algorithms resulted in an overall diagnostic accuracy of 97%. The performance of an array containing the identified discriminating genes should now be assessed in frontline clinical trials in order to determine the accuracy, practicality, and cost effectiveness of this methodology in the clinical setting.

Description: Transcription factor CCAAT/enhancer binding protein alpha(C/EBPα) is essential for granulocyte differentiation. CEBPα mutations have been described in 7–10% of adult patients with acute myeloid leukemia (AML) at initial diagnosis and conferred a favorable prognosis. CEBPα mutations in childhood AML has not been reported yet. In this study, we sought to assess the frequency of CEBPα mutration, its clinicohematologic correlation and to analyze the cooperating mutations, including FLT3 and N-ras mutations in childhood AML. CEBPα mutations were analyzed in 117 children (age one day to 17 years, median 5 years) with de novo AML. CEBPα mutation status was examined by performing DNA polymerase chain reaction (PCR) followed by direct sequencing for each PCR product. CEBPα mutations were detected in 7 (6.0%) of 117 patients. Of the 7 patients with CEBPα mutations, 4 had FAB subtype of M2, 2 M1 and 1 M4. All had intermediate cytogenetics and comprised 11.5% (7/61) of the intermediate cytogenetic risk group. Five of the 7 mutations occurred in both N-terminal part and bZIP domain, one had N-terminal frameshift mutation and the remaining one had an inframe insertion in the bZIP domain. Clonal analysis using expand long template PCR assay with a primer pair covering entire coding sequence was performed in all 5 patients carrying two different mutations. The results demonstrated that one had homozygous combined mutation and 4 had heterozygous biallelic mutations with the cloned alleles carrying single mutations at N-terminal or bZIP domain in the different alleles or harboring combined mutations in the same alleles. None of the remission marrow samples obtained from 5 patients who had CEBPα mutations at diagnosis carried the mutations. FLT3-ITD mutations were detected in 2 of the 7 patients with CEBPα mutations compared to 15 of the 109 patients without CEBPα mutations. None of the patients carrying CEBPα mutations had FLT3-TKD mutations. Two of the 7 CEBPα(+) patients also harbored N-ras mutations compared to 10 of the 110 patients without CEBPα mutations. Taken together, 4 of 7 CEBPα(+) patients had cooperating mutations including FLT3-ITD or N-ras as compared to 27 in 109 CEBPα(−) patients (P=0.081). There were no statistical differences in age, WBC count, LDH level, 5-year overall survival and event-free survival between patients with and without CEBPα mutations; similar results were obtained if we only analyzed patients with intermediate cytogenetics. The present data showed that the frequency of CEBPα mutations in our series of childhood AML was similar to that in adults. CEBPα mutations were frequently associated with FLT3-ITD or N-ras mutations supporting the two-hit hypothesis for the pathogenesis of AML.

Description: Contemporary treatment of pediatric acute myeloid leukemia (AML) requires the assignment of patients to specific risk groups. To explore whether expression profiling of leukemic blasts could accurately distinguish between the known risk groups of AML, we analyzed 130 pediatric and 20 adult AML diagnostic bone marrow or peripheral blood samples using the Affymetrix U133A microarray. Class discriminating genes were identified for each of the major prognostic subtypes of pediatric AML, including t(15;17)[PML-RARα], t(8;21)[AML1-ETO], inv16 [CBFβ-MYH11], MLL chimeric fusion genes, and cases classified as FAB-M7. When subsets of these genes were used in supervised learning algorithms, an overall classification accuracy of more than 93% was achieved. Moreover, we were able to use the expression signatures generated from the pediatric samples to accurately classify adult de novo AMLs with the same genetic lesions. The class discriminating genes also provided novel insights into the molecular pathobiology of these leukemias. Finally, using a combined pediatric data set of 130 AMLs and 137 acute lymphoblastic leukemias, we identified an expression signature for cases with MLL chimeric fusion genes irrespective of lineage. Surprisingly, AMLs containing partial tandem duplications of MLL failed to cluster with MLL chimeric fusion gene cases, suggesting a significant difference in their underlying mechanism of transformation.

Description: Abstract 3064 Poster Board III-1 Both ETV6-RUNX1 (TEL-AML1)fusion and hyperdiploidy (〉50 chromosomes) of lymphoblasts are favorable outcome predictors in childhood acute lymphoblastic leukemia (ALL). In 433 children with ALL diagnosed at our hospitals between 1997 and 2007 in Taiwan, the frequency of ETV6-RUNX1 fusion was 15.8%, and the frequency of hyperdiploidy (〉50 chromosomes) was 14.1%, both were lower than those of the West. While ETV6-RUNX1 fusion had borderline favorable impact on outcome (p=0.053-0.061), hyperdiploidy showed significant favorable impact on event-free survival (91.1% vs 76.6 %, p= 0.016) in our patients. A meta-analysis from literature enrolled reports in which the case numbers and frequency of ETV6-RUNX1 fusion or hyperdiploidy in childhood ALL were described. It revealed that the frequency of ETV6-RUNX1 fusion in childhood ALL in Far East (Japan, Korea, Hong Kong, Chinese in Singapore, and Taiwan) was 14.2% (127/893, range 10-17%), significantly lower than 21.8% (152/697, range 19-27%) in the West (USA, Germany, Italy, France and Chile) (p 〈 0.0001). The frequency of hyperdiploidy in Japan and Taiwan was 15.2% (140/921, range 13-20%), significantly lower than 31.6% in the West (977/3,158, range 19-34%) (USA, UK and Germany) (p 〈 0.0001). So far as we know, there were several articles, including ours, addressing that the frequency of ETV6-RUNX1 fusion in childhood ALL was lower in a Far East country. This is the first meta-analysis to demonstrate that the frequency of ETV6-RUNX1 fusion in childhood ALL in Far East was lower than that in the West. There was no report on that the frequency of hyperdiploidy in Far East was lower than that in the West. This is also the first meta-analysis to demonstrate that the frequency of hyperdiploidy in childhood ALL in Far East is significantly lower than that in the West. The nature of these differences, probably due to racial, needs further study. In Far East, with both a lower frequency of ETV6-RUNX1 fusion, and a lower frequency of hyperdiploidy, it warrants renewed effort to cure a higher proportion of children with ALL. Disclosures No relevant conflicts of interest to declare.

Description: Background: Discontinuation of E. coli- asparaginase in patients with acute lymphoblastic leukemia (ALL) upon severe allergic reactions is unavoidable. We aimed to examine the outcomes following E.coli- asparaginase discontinuation upon severe allergic reactions in ALL. Patients and methods : In Taiwan Pediatric Oncology Group (TPOG)-2002-ALL protocol (enrolled 2002-2012), intramuscular E. coli- asparaginase (Kyowa Hakko, Japan) was given at 5000 IU/m2 per dose thrice weekly for 3 weeks during the remission induction therapy of high-risk (HR) and very-high-risk (VHR) groups. During the first 20 weeks of continuation therapy, HR patients received weekly intramuscular E. coli- asparaginase 10,000 IU/m2 per dose every week. They also received two reinduction treatments during which E.coli- asparaginase was given at 5000 IU/m2 per dose thrice weekly for 3 weeks. Patients in VHR groups received one reinduction treatment during which E. coli- asparaginase was given at 5000 IU/m2 per dose thrice weekly for 2 weeks. Patients in standard-risk (SR) group, received no E. coli- asparaginase in induction, but were randomized to receive one or two reinduction phases, during which E. coli- asparaginase was given at 5000 IU/m2 per dose thrice weekly for 2 weeks. The scheduled cumulative doses of E.coli- asparaginase in each risk group were 30,000 IU/m2 or 60,000 IU/m2 in SR, 265,000 IU/m2 in HR and 75,000 IU/m2 in VHR group. We evaluated outcome of children enrolled in TPOG-2002-ALL protocol who had E. coli- asparaginase discontinued due to severe allergic reactions (marked swelling and redness at the injection site or anaphylaxis) between 2002 and 2012, and compared outcomes between those who with Erwinase continued and those who without after the discontinuation of E. coli- asparaginase. The distributions of the Kaplan-Meier estimates of event-free survival (EFS) and overall survival (OS) were compared using log-rank test. Chi-square test was used to compare each parameter between groups. Results: In 700 patients from 10 hospitals retrospectively studied, 52 patients had E. coli- asparaginase treatment discontinued due to the development of severe pancreatitis in 17 patients, severe thrombosis in 2, and severe allergic reactions in 33. In Taiwan, Erwinase has been available since 2012, and could be purchased from foreign countries before. PEG-asparaginase is not available in Taiwan. In the 33 patients had E. coli- asparaginase discontinued due to severe allergic reactions, 17 continued Erwinase and 16 did not. The parameters between these two groups were similar. They were of more HR group reflecting that HR patients received more E. coli- asparaginase than other patients. The 5-year OS did not differ significantly among the 648 patients without discontinuation (81±1.6%, mean±S.E.), the 17 with allergic reactions and continued with Erwinase (88±7.8%) and the 16 with allergic reactions not treated with Erwinase (87±8.6%). The P value for the difference between the latter two groups was 0.96. In the 16 patients who did not receive Erwinase, all the 10 patients, who had received 〉= 50% scheduled dose of E.coli- asparaginase before discontinuation, survived without events. Conclusions: We suggest that patients who had received 50% or more of the scheduled doses of E. coli- asparaginase before the development of severe allergic reactions do not need to continue treatment with Erwinase. This may be helpful in those who cannot afford Erwinase or in the vast majority of the world where Erwinase is not available. It also raises a question on that how much asparaginase is enough for the treatment of childhood ALL. Disclosures No relevant conflicts of interest to declare.

Description: Abstract 3564 Background The introduction of CNS-directed therapy had successfully improved the outcome of children with acute lymphoblastic leukemia (ALL). Pui et al. (N Engl J Med, 2009) demonstrated that with effective risk-adjusted systemic chemotherapy and intrathecal therapy initiated at diagnosis (Dx), cranial irradiation can be safely omitted from CNS-directed treatment in all children with newly diagnosed ALL. Manabe et al. (J Clin Oncol, 2001) reported a delay of diagnostic lumbar puncture (LP) and triple intrathecal therapy (TIT) after 1 week of prednisolone monotherapy. Starting 1999, in order to eliminate the adverse prognostic impact of traumatic LP with blasts, we conducted a prospective study of delayed TIT after the disappearance of blasts from peripheral blood (PB) on around 10 days of multi-drug induction therapy. Methods Since June 1999, children with ALL were treated with TIT alone without cranial irradiation, and chemotherapy with TPOG-ALL-93 or TPOG-ALL-2002 (activated in January 2002) protocols. These protocols had been reported previously (Leukemia, 2010). B-precursor ALL patients were stratified into standard risk (SR), high risk (HR) and very high risk (VHR) groups, according to age, WBC count, cytogenetics and molecular analysis at Dx. T-cell ALL was designated as VHR. The first TIT was performed until the disappearance of blasts from PB. If patients with persistent PB blasts on the treatment day 10 (D10), the TIT was planned. Furthermore, SR patient with detectable blasts on D10, the treatment was upgraded to HR group. Cranial irradiation was totally omitted in all patients. For patients with CNS-1 status, SR group received TIT 20 times in TPOG-ALL-93 and 14 times in TPOG-ALL-2002 protocols, respectively. HR patients with CNS-1 status received 23 and 17 times of TIT in TPOG-ALL-93 and 2002 protocols, respectively. In VHR group, TIT was administered 19 times for CNS-1. Additional TIT doses were given if patients with CNS-2 or CNS-3. Results From June 1999 to August 2010, a total of 158 ALL patients were consecutively diagnosed at our hospital. There were 86 patients in SR group, 41 in HR and 31 in VHR, respectively. Six patients were excluded: 2 refusals to further treatment and 4 protocol violations. There were 86 boys and 66 girls enrolled. The median age at Dx was 4.1 years (range, 0.3–17.7 years) with 7 infants. The median WBC count at Dx was 12,100/μL (range, 800–596,000/μL). Fourteen had counts 〉100,000/μL. There were 141 patients of B-precursor ALL and 11 of T-cell ALL. At Dx, 17 patients (16 in SR and 1 in VHR) did not have detectable blasts on PB. In the remaining 135 patients, the blasts of 101 patients (74.8%) can be eliminated from PB after 10-day induction therapy, those included 77.9% of SR, 75.6% of HR and 65.4% of VHR, respectively. Six patients of SR group were upgraded to HR protocol since the detection of PB blasts on D10. Bleeding into the CSF at the time of LP was apparent in 4 patients (2.6%), but no leukemic blasts were identified in any patients. As of July 31, 2012, the follow-up duration of 133 surviving patients ranged from 24 to 153 months (median, 88 months). The 7-year event-free survival and overall survival rates were 83.0 ± 3.2% (± SE) and 86.5 ± 3.1%, respectively. No isolated CNS relapse occurred. One infant with MLL rearrangement experienced combined CNS, BM and testicular relapses and one child with t(12;21) had combined CNS and molecular BM relapses. The cumulative risk of any CNS relapse was 1.5 ± 1.0% at 7 years. Conclusions A delay of TIT after disappearance of PB blasts with multi-drug induction therapy can effectively prevent CNS disease in children with ALL. Cranial irradiation can be completely omitted for CNS prophylaxis in our treatment protocols. Disclosures: No relevant conflicts of interest to declare.

Description: Backgrounds and Purposes Minimal residual disease (MRD) monitoring has been proved to be the most important prognostic predictor in childhood acute lymphoblastic leukemia (ALL). The nationwide TPOG-ALL-2013 protocol (TPOG-2013), adapted from the St. Jude Total Therapy XV Study and Total Therapy XVI Study, was launched since January 2013. This is the first MRD-directed protocol for treatment of childhood ALL in Taiwan. Here, we report the improved treatment outcomes and the impacts of adherence to MRD time points. Patients and Methods Totally, 402 patients aged between 1-18 years and diagnosed before December 31, 2018, who had MRD monitoring at the major central laboratory (Chang Gung Memorial Hospital-Linkou), were enrolled with the last follow-up on June 30, 2019. According to TPOG-2013, two MRD measurements were scheduled on days 15-19 of induction (MRD1 time point, TP1) and days 35-42, end of induction (MRD2 time point, TP2) to make the definitive risk stratification to guide subsequent therapy. The methodologies of MRD measurement included multicolor flow cytometry for leukemia-associated immunophenotypes (LAIP) (82.3% of TPOG-2013 cohort), qPCR assay for clonally rearranged antigen-receptor genes (Ig/TCR) if no LAIP (12.5%). Since January 2018, reverse transcription real-time quantitative polymerase chain reaction (RT-qPCR) was applied to patients carrying fusion transcripts (5.2%) of TCF3-PBX1, ETV6-RUNX1, BCR-ABL1, KMT2A-AFF1 (AF4) and KMT2A-MLLT3. The clinical features and outcomes of patients treated with TPOG-2013 were compared with those of 1,300 patients treated with the previous TPOG-ALL-2002 protocol (TPOG-2002), which did not integrate the MRD monitoring. Results The median follow-up time of the 402 patients of TPOG-2013 cohort was 32.5 months (range, 1.0-79.2 months). There were no significant differences in gender, age, WBC counts, and lineage at diagnosis between the patients treated with TPOG-2002 and TPOG-2013. However, based on the MRD data, the percentages of patients assigned to each risk group of TPOG-2013 was statistically differed from those of TPOG-2002 (P〈 0.0001). The 5-year event-free survival (EFS) (% ± SE) was significantly improved from 78.1 ± 1.2 of TPOG-2002 to 85.4 ± 2.5 of TPOG-2013 (P〈 0.0001). Further, the cumulative incidences (% ± SE) of isolated CNS relapse and any CNS relapse significantly decreased from 4.0 ± 0.5 to 0.3 ± 0.3 (P= 0.001) and from 5.8 ± 0.7 to 1.2 ± 0.9 (P= 0.001), respectively. The issue of non-adherence to MRD monitoring emerged since the implementation of MRD-directed TPOG-2013. For further analysis, 321 (80%) patients with exact adherence (EA) to both TPs were assigned as MRD EA group; 80 (20%) patients who were non-adherence (NA) to either one of TPs as MRD NA group; and one patient died between the two TPs was excluded for the comparative outcome analysis. The rate of non-adherence decreased significantly from 26.5% in 2013 to 2.4% in 2018. The major causes of non-adherence for both TPs were delaying MRD monitoring due to neutropenic fever and documented infections. In MRD EA group, 12.5% of patients were upgraded to higher-risk treatment groups based on their MRD results. The MRD NA group had older age (≥ 10 years), lower standard-risk and lower incidence of ETV6-RUNX1 compared with MRD EA group. There were significant differences in outcomes between MRD EA and MRD NA groups: the 5-year EFS were 89.4 ± 2.4 and 71.9 ± 7.4, respectively (P= 0.0005), overall survival (OS) were 90.9 ± 2.1 and 75.6 ± 5.8, respectively (P= 0.0003), and the cumulative incidence of isolated CNS relapse were 0 and 1.4 ± 1.3, respectively (P= 0.048) (Figure 1). In multivariate analysis, older age (≥ 10 years), higher WBC count (≥ 50 × 109/L) at diagnosis and MRD non-adherence were independent predictors for inferior EFS. In addition to these three factors, a higher-risk classification also predicted an inferior OS (Figure 2). Conclusions Contemporary MRD-directed therapy has improved the treatment outcomes of childhood ALL in Taiwan. The adherence to MRD time points remains a significantly prognostic predictor in the era of MRD-guided treatment. Disclosures No relevant conflicts of interest to declare.

Description: Introduction The nationwide TPOG-ALL-2002 protocol for treating children with acute lymphoblastic leukemia (ALL) was activated in January 2002 in Taiwan. To eliminate cranial radiation (CrRT)-related sequelae and minimize the adverse prognostic impact of traumatic lumbar puncture with blasts (TLP), we conducted a prospective study beginning in 1999 to modify CNS-directed therapy with delayed first triple intrathecal therapy (TIT) and omission of CrRT for all risk groups of newly diagnosed ALL. This approach improved CNS control at the single-institution study (J Clin Oncol, 2014) and has been used in whole TPOG since January 2009. Patients with non-CNS-1 status (i.e., CNS-2, CNS-3 and TLP) have higher risk of CNS relapse and need intensive CNS therapy. On behalf of TPOG members, we reported the outcomes of these non-CNS-1 patients in the two eras (2002-2008 and 2009-2012) with or without the modified CNS therapy. Methods From 2002 to 2012, all newly diagnosed ALL children in Taiwan were enrolled in TPOG-ALL-2002 protocol (Leukemia, 2010). B-precursor ALLs were stratified into standard risk (SR), high risk (HR) and very high risk (VHR) groups, according to age, WBC count, cytogenetics and molecular analysis at diagnosis. T-cell ALL was designated as VHR. Since 2009, patients had been treated with TIT alone without CrRT. The first TIT was performed until the disappearance of blasts from peripheral blood (PB), no later than 10 days. SR patient with detectable PB blasts on day 8 was upgraded to HR group. SR with CNS-1 received 14 doses of TIT, and those with CNS-2, or TLP got 2 additional TITs during induction and 3 additional TITs during maintenance therapy. Totally, SR with CNS-2 or TLP received 19 doses of TIT. And SR patients with CNS-3 or cranial nerve palsy at diagnosis were classified as HR. For HR with CNS-1, 17 doses of TIT were given. If HR groups with CNS-2, CNS-3, or TLP, they will receive total 25 doses of TIT including 2 additional ones during induction and 6 additional ones during maintenance therapy. Before 2009, only SR and HR with refractory CNS leukemia (failure to clear CSF blasts after 3 consecutive TITs) and those with CNS relapse will receive CNS radiation. For VHR group before 2009, CrRT of 18 Gy was given to all patients at the end of remission induction. Instead, TIT was given for age 〈 2 years. After 2009, CrRT was totally omitted. VHR with CNS-1 received 19 doses of TIT. For VHR patients with CNS-2, CNS-3 or TLP at diagnosis, TIT was administered once/week until achieving CSF remission and then once 4 weeks during maintenance therapy. Results Since January 2009, all patients received TIT alone, irrespective of risk groups. The two cohorts, totally 1,366 patients, 909 treated with CrRT (2002-2008) and 457 without CrRT (2009-2012) were balanced in presenting features and distribution of CNS status. There were no significant differences between two eras in the event-free survival (EFS), overall survival (OS), incidences of isolated CNS relapse and any CNS relapse (Pediatr Blood Cancer, accepted June 2016). The number of patients with CNS-2, CNS-3 and TLP were 35, 26 and 19 before 2009; and 18, 10 and 9 after 2009, respectively. Patients with younger age (〈 1 year), higher WBC counts (〉 100 X 109/L), HR/VHR or MLL rearrangement had significantly more non-CNS-1 status in the era of 2002-2008. And the significant factors prone to be non-CNS-1 were male, T-lineage and HR/VHR with the first TIT delaying. There were no significant differences between the two eras in the 5-year EFS and OS. Though the incidence of 5-year any CNS relapse was not significantly different, fewer relapses were observed in the era of 2002-2008 (Table 1). Before 2009, eleven CNS relapses experienced between 1 month and 3.1 years (median, 10 months) after remission. Two relapses after 2009 occurred on 1 month and 7 months, respectively. Conclusions Delayed first TIT until the clearance of circulating blasts and total omission of CrRT did not compromise the survivals and CNS control of childhood ALL patients with CNS-2, CNS-3 and traumatic lumbar puncture with blasts. Moreover, this approach prevents the adverse sequelae caused by CrRT. Furthermore, a skillful TIT is mandatory to improve results. Disclosures No relevant conflicts of interest to declare.

Description: Abstract 2713 Background. The cooperation of gene mutations, especially their impacts on survivals of childhood acute myeloid leukemia (AML) has not been well known. Aims. Our aims were (1) to study the frequency of each gene mutation in childhood AML, (2) to study the impact of each gene mutation on the treatment outcome, and (3) to examine the cooperativity of gene mutations. Materials and Methods. From Feb. 1996 to Jan. 2010, bone marrow samples at diagnosis from 198 children with AML at Chang Gung Children's Hospital, Taoyuan and Mackay Memorial Hospital, Taipei, were analyzed for gene mutations including FLT3-ITD, FLT3-TKD (D835), c-KIT, cFMS, JAK 2V617F, NRAS, KRAS, PTPN11 (Class I mutations), RUNX1, CEBPα, NPM1 (Class II mutations), WT1 and P53 (tumor suppressor genes). The subtypes included: t(8;21) 19.9%, inv(16) 8.9%, t(15;17) 8.4%, t(9;11) 5.2%, t(10;11) 2.6%, trisomy 21 4.2%, intermediate-risk group 40.3% (including 13 patients with other MLL translocations), and poor-risk group 11.0% (including 7 patients with complex chromosomal abnormalities and 4 patients with MLL-PTD). Results. FLT3-ITD occurred in 15.0% of patients, FLT3-TKD 7.2%, c-KIT 11.5%, c-FMS 2.9%, JAK2V617F 3.3%, NRAS 9.1%, KRAS 7.7%, PTPN11 3.3%, RUNX1 2.7%, CEBPα 7.9%, NPM1 4.1%, WT1 3.9% and P53 1.7%. Taken together, 52.5% of patients had Class I gene mutations, 13.1% had Class II gene mutations, and 5.1% had WT1 or P53 mutations. In all, 59.1% of patients had Class I, Class II or tumor suppressor gene mutations. Only one patient (0.5 %) had gene mutations involving all Class I, Class II and tumor suppressor genes. Ninety-eight patients, who were treated with Taiwan Pediatric Oncology Group (TPOG) APL protocols (for acute promyelocytic leukemia) and TPOG 97A protocol (for other AML) (Liang et al, Leukemia 2006), were analyzed for survivals. In patients with t(8;21), the 5-year event-free survival (EFS) was 66±12%; 71±17% for patients with c-KIT mutations and 50±35% for the 2 patients with JAK2V617F. In patients with inv(16), the EFS of 70±15% seemed to be compromised (60±22%) for those with c-KIT mutations. In patients with t(15;17), the EFS of 78±11% was not compromised by FLT3-ITD or FLT3-TKD mutations. In patients with t(9;11), the EFS of 64% seemed to be compromised (50±35%) in the 2 patients with FLT3-TKD mutations. In 3 patients with t(10;11), no gene mutations were found. In trisomy 21, the EFS of 75±22% seemed to be compromised (50±35%) in the 2 patients with CEBPα mutations. Of the 5 patients with complex chromosomal abnormalities, the only one patient carrying RUNX1 survived. Of the 3 patients with MLL-PTD having an EFS of 33±27%, one each patient with c-FMS or WT1 mutation died. The only one patient who had all Class I, Class II and tumor suppressor gene mutations (FLT3-TKD+ CEBPα+ WT1) died in induction therapy. Two of the other 4 patients who had 3 mutations acrossing 2 classes had EFS of 6 and 10 months, respectively. Conclusions. Our study on a large cohort of pediatric AML patients revealed that 59.1% patients had at least one gene mutation. That 3 of 5 patients who had 3 gene mutations soon failed suggested that gene mutations, especially in 3 combinations, might compromise the survival. Further study on more patients is warranted to explore more of the prognostic significance of cooperating gene mutations in pediatric AML. (Supported by grants MMH-E-98009, NSC 96–2314-B-195-006-MY3, NHRI-EX-96-9434SI and DOH99-TD-C-111-006.) Disclosures: No relevant conflicts of interest to declare.