566FN2 Introduction: Despite improved outcome of current therapy, up to 20% of pediatric patients with B-cell precursor Acute Lymphoblastic Luekemia (BCP-ALL) still relapse. Recent genome-wide analysis showed that IKZF1, JAK2 and CRLF2 gene alterations correlate with poor prognosis in these patients. We analyzed theses gene alterations in Japanese patient cohort. Materials and Methods: Two hundred patients with BCR-ABL1-negative BCP-ALL between 1 and 18 years old, who were enrolled in Japanese Children's Cancer & Leukemia Study Group (JCCLSG)-2004 study from 2004 to 2008, were analyzed in this study. DNA in 179 patients and RNA in 147 patients, respectively, were extracted from the initial diagnostic bone marrow or peripheral blood samples. Using DNA samples, JAK2R683 point mutation was analyzed by direct sequencing, and the copy number alterations of IKZF1 by multiplex ligation-dependent probe amplification (MLPA). CRLF2 expression in RNA samples were measured by real-time quantitative polymerase chain reaction (RQ-PCR). At the same time, we detected IKZF1 transcript of Ik6 isoform (del. exon 4 to 7) and P2RY8-CRLF2 fusion gene expression by reverse transcribed (RT) - PCR. Event-free survival (EFS) and relapse-free interval (RFI) was calculated by Kaplan-Meier method. Event was defined with induction failure, relapse, death of any cause in EFS, and with relapse in RFI. Result: Twenty-two (12%) of 179 patients had some exon deletions in IKZF1. Seven (32%) of 22 patients had only one copy of all exon (exon 1 to 8) and 6 of 22 (27%) patients had deletion from exon 4 to 7, which expressed Ik6 isoform. The older the age was, the higher the frequency of IKZF1 deletion was. The frequency of IKZF1 deletions (12/59) in the patients of NCI-HR group was significantly higher than that (10/120) of the patients of NCI-SR group (p=0.029). In outcome, EFS for the patients with IKZF1 deletions was significantly worse than that for without IKZF1 deletions (EFS; 66.4%±12.1% vs. 89.2%±2.7% respectively, p=0.007). And RFI for the patients with IKZF1 deletions was also worse than that for without it, (RFI; 73.1%±12.4% vs. 90.9%±2.6%, respectively, p=0.047). IKZF1 deletions seemed to be poor prognostic factor in NCI-HR patients, while they are not in NCI-SR, as shown in Table. High CRLF2 expression (10 times higher than median value, 14.4 copies) was detected in 15 (10%) of 147 patients. Only 2 of 15 CRLF2-high patients had P2RY8-CRLF2 fusion gene by RT-PCR. We also detected 3 of 135 CRLF2-low patients had this fusion. EFS for CRLF2-high expression patients had also significantly worse than that for CRLF2-low patients (EFS; 62.2%±13.7% vs. 90.6%±2.7%, p=0.003). And RFI for CRLF2-high patients was 73.1%±12.4%, while RFI for CRLF2-low patients was 90.9%±2.6% (p=0.069). EFS for CRLF2-high showed significantly worse than that for CRLF2-low in NCI-HR patients (53.3%±17.3% vs. 87.0%±6.1%, p=0.028), but not significantly in NCI-SR (p=0.445). Regardless of patients with or without CRLF2-high expression, they showed similar RFI in NCI-HR and SR. Five patients had IKZF1 deletions and CRLF2-high expression, simultaneously, and showed extremely poor prognosis (4 of 5 showed induction failure or relapse). None of 179 patients had JAK2R683 mutation in this study. Discussion: While IKZF1 and CRLF2 alterations were detected in similar frequency to previous reports, P2RY8-CRLF2 fusion was found only 13% of CRLF2-high patients. Moreover, JAK2R683 mutation was not detected in any patients in this study. These findings suggest that there might be racial difference in genetic mechanism of leukemogenesis. IKZF1 and CRLF2 alterations might be a poor prognostic factor in NCI-HR patients, but not in NCI-SR. These independency would be disclosed in next large-scale national study. Disclosures: No relevant conflicts of interest to declare.

IKZF1 deletions and CRLF2-high expression predicted poor outcomes in patients with HR-ALL but not in patients with SR-ALL in our Japanese cohort.

These results suggest that augmented therapy for MRD-positive patients at TP2 contributed to better outcomes of the ALL2000 study.

No significant cardiac dysfunction was detected in long-term survivors who received THP treatment. The use of post-exercise BNP level to indicate high cardiotoxicity risk should be verified by further study.

Many studies have shown the presence of minimal residual disease (MRD) following therapy for childhood acute lymphoblastic leukemia (ALL) to be an important prognostic marker. We have also shown a significant relationship between survival outcomes in patients enrolled in the previous ALL 911 study and molecular MRD levels 5 weeks (time point 1, TP1) and 12 weeks (TP2) following the initiation of chemotherapy (Leukaemia and Lymphoma2002; 43: 1001). The aim of this study was to evaluate if polymerase chain reaction (PCR)-based MRD assay is sufficiently dependable for tailoring therapy, and if augmented therapy can reduce MRD levels to those associated with a favourable outcome. The subjects were under 18 years of age, and had newly diagnosed precursor B or T-cell ALL. Patients below one year old and those with t(9;22) were excluded. Written informed consent was obtained from patients or their legal guardians. The ALL 941-based protocol (45thASH, San Diego, 2003) utilized PCR-based MRD assay using immunoglobulin & T-cell receptor gene rearrangements. MRD was detected by nested PCR, with screening of rearrangements using multiplex PCR primers as described previously (Leukaemia and Lymphoma2002; 43: 1001). Patients were initially stratified into 3 risk groups (in ascending order: SR, HR, and HHR) according to leukocyte count and age at time of diagnosis. The MRD+/+ patients with levels ≥ 10−3 at both TP1 and TP2 received augmented therapy 14 weeks after initiation, and the remainder continued to receive the initial risk-adapted protocols. A total of 311 patients with a median age of 5.3 years (range 1.0–16.8) were eligible for this study. There were 4 (1.3%) non-responders and no deaths in induction. Of the 307 patients stratified, 169 (55%) were SR, 107 (35%) were HR, and 31 (10%) were HHR. The 2nd stratification by MRD level at TP2 was possible for 72.3% (222/307; insufficient DNA=28; missing time-points=25; no marker=32). Out of the 222 patients stratified, 125 (56.3%) were MRD−/−, 58 (26.1%) were MRD+/−, and 38 (17.4%) were MRD+/+. At the point of analysis, the median follow-up time was 63 months (range 33–89). The overall 5-year event–free survival (EFS) rate of the 307 patients was 80.1% (SE 2.5), higher than the EFS of the ALL941 study, which was 76.2% (SE 2.1) (p=0.167). The 5-year EFS rates according to the 1st stratification were 85.5% (SE 4) for SR, 76.1% (SE 4.5) for HR, and 64.6% (SE 9.2) for HHR, while the equivalent rates for the 2nd stratification were 87.0% (SE 3.1) for MRD−/−, 75.5% (SE 7.7) for MRD+/−, and 75.3% (SE 6.4) for MRD+/+. From the 95 patients whose MRD levels were measured at 5 consecutive points from TP1 to TP5 (5, 12, 18, 24, and 30 weeks after the start of therapy), 21 subjects with MRD+/+ received an augmented chemotherapy, and MRD levels became undetectable in 9 patients at TP3, 5 patients at TP4, and 4 patients at TP5. The corresponding cumulative 5-year relapse rates of those patients were 11%, 50%, and 50%, respectively. Thus, negative MRD status at TP3, but not at TP4 or TP5, seems to be associated with a favourable outcome. Our results confirm the strong performance of MRD-based treatment interaction in a multi-institutional study without adversely affecting the outcome in childhood ALL. Moreover, present findings suggest that an augmented therapy could reduce MRD to levels associated with a favourable outcome. To improve the applicability and accuracy of MRD assay, new MRD-PCR targets and RQ-PCR-based MRD detection are needed in subsequent studies. [Acknowledgment: This study was partly supported by grants from the Children’s Cancer Association of Japan (CCAJ)].