Activating mutations in RAS and receptor tyrosine kinases such as KIT and FLT3 are hypothesized to cooperate with chimeric transcription factors in the pathogenesis of acute myeloid leukemia (AML). To test this hypothesis, we genotyped 150 pediatric AML samples for mutations in KIT (exons 8, 17), NRAS and KRAS (exons 1, 2) and FLT3/ITD. This is the largest cohort of pediatric AML patients reported thus far screened for all four mutations. Of the children with AML, 40% had a mutation in KIT (11.3%), RAS (18%) or FLT3/ITD (11.1%), and 70% of cases of core-binding factor (CBF) leukemia were associated with a mutation of KIT or RAS. Mutations in RAS or FLT3/ITD were frequently found in association with a normal karyotype. Patients with a FLT3/ITD mutation had a significantly worse clinical outcome. However, the presence of a KIT or RAS mutation did not significantly influence clinical outcome. We demonstrate that KIT exon 8 mutations result in constitutive ligand-independent kinase activation that can be inhibited by clinically relevant concentrations of imatinib. Our results demonstrate that abnormalities of signal transduction pathways are frequent in pediatric AML. Future clinical studies are needed to determine whether selective targeting of these abnormalities will improve treatment results.

No abstract

Summary The methyl-thiazol-tetrazolium (MTU) assay is a drug resistance assay which cannot discriminate between malignant and non-malignant cells. We previously reported that samples with > 80% leukaemic cells at the start of culture give similar results in the MTT assay and the differential staining cytotoxicity assay, in which a discrimination between malignant and non-malignant cells can be made. However, the percentage of leukaemic cells may change during culture, which might affect the results of the MTT assay. We studied 106 untreated childhood acute lymphoblastic leukaemia (ALL) samples with > 80% leukaemic cells at the start of culture. This percentage decreased below 80% in 28%, and below 70% in 13%, of the samples after 4 days of culture. A decrease below 70% occurred more often in case of 80-89% leukaemic cells (9/29) than in case of > 90% leukaemic cells at the start of culture (5/77, P = 0.0009). Samples with <70% leukaemic cells after culture were significantly more resistant to 6 out of 13 drugs, and showed a trend towards being more resistant to two more drugs, than samples with > 80% leukaemic cells. No such differences were seen between samples with 70-79% and samples with > 80% leukaemic cells after culture. We next studied in another 30 ALL samples whether contaminating mononuclear cells could be removed by using immunomagnetic beads. Using a beads to target cell ratio of 10:1, the percentage of leukaemic cells increased from mean 72% (s.d. 9.3%) to mean 87% (s.d. 6.7%), with an absolute increase of 2-35%. The recovery of leukaemic cells was mean 82.1% (range 56-100%, s.d. 14.0%). The procedure itself did not influence the results of the MTT assay in three samples containing only leukaemic cells. We conclude that it is important to determine the percentage of leukaemic cells at the start and at the end of the MTT assay and similar drug resistance assays. Contaminating mononuclear cells can be successfully removed from ALL samples using immunomagnetic beads. This approach may increase the number of leukaemic samples which can be evaluated for cellular drug resistance with the MTT assay or a similar cell culture drug resistance assay.

Cellular drug resistance is thought to be an important cause of the poor prognosis for children with relapsed or refractory acute lymphoblastic leukemia (ALL), but it is unknown when, to which drugs, and to what extent resistance is present. We determined in vitro resistance to 13 drugs with the MlT assay. Compared with 141 children with initial ALL, cells from 137 children with relapsed ALL were significantly more resistant to glucocorticoids, L-asparaginase, anthracyclines, and thiopurines, but not to vinca-alkaloids, cytarabine, ifosfamide, and epipodophyllotoxins. Relapsed ALL cells expressed the highest level of resistance to glucocorticoids, with a median level 357-and >24-fold more resistant to prednisolone and dexamethasone, respectively, than ini-OWADAYS, using intensive front-line multiagent chemotherapy along with improved supportive care, more than 95% of children with acute lymphoblastic leukemia (ALL) can achieve a complete remission (CR), of whom 70% will remain in continuous CR and be considered cured.'-3 These results show the tremendous improvement in the development of more effective chemotherapy regimens for childhood ALL, which was once a fatal d i~e a s e .~ Nevertheless, current protocols fail in the remaining 30% of the children with newly diagnosed ALL, with bone marrow relapses representing the most common treatment failures.'-3 Second remissions can be induced with intensified chemotherapy in more than 90% of the children with ALL whose disease relapsed while they were on modem protocols, but their long-term prognosis is p~o r . ' .~.~ Long-lasting second hematologic remissions can be expected in about 10% of the children with early relapses and in up to 30% of those with late relapses, despite even more intensive second-line therapy that includes the effective front-line drugs used in an alternative Knowledge about the nature of relapsed ALL is limited, but it is assumed that regrowth of drug-resistant leukemic cells plays an important role.6 It is unknown to which drugs and to what extent relapsed leukemic cells express resistance, mainly because a suitable drug-resistance assay was lacking until recently. The poor growth capacity of ALL cells in vitro, limiting the use of long-term clonogenic assays, could be circumvented with the introduction of so-called shortterm cell culture drug-resistance assay^.^ The 3-[4,5-dimethyl-thiazol-2,5-diphenyl] tetrazolium bromide (MTT) assay, first described by Black and Speer in 1954' and revised by Mosmann in 1983: has been adapted by us for testing ALL cells."." The 4-day semiautomated MTT assay is an efficient tool for large-scale drug-resistance testing and results showed a good correlation with the prognosis in childhood We present here the results of in vitro drugresistance testing on samples from 141 children with relapsed or refractory ALL.

Acute lymphoblastic leukemia (ALL) in infants under 1 year is strongly associated with translocations involving 11q23 (MLL gene), CD10-negative B-lineage (proB) immunophenotype, and poor outcome. The present study analyses the relationship between age, MLL rearrangements, proB-lineage, and in vitro drug resistance determined using the MTT assay. Compared to 425 children aged over 1 year with common/preB (c/preB) ALL, the 44 infants were highly resistant to steroids (for prednisolone (PRED) more than 580-fold, P ¼ 0.001) and L-asparaginase (L-ASP) (12-fold, P ¼ 0.001), but more sensitive to cytarabine (AraC) (1.9-fold, P ¼ 0.001) and 2-chlorodeoxyadenosine (2-CdA) (1.7-fold, Po0.001). No differences were found for vincristine, anthracyclines, thiopurines, epipodophyllotoxines, or 4-hydroperoxy (HOO)-ifosfamide. ProB ALL of all ages had a profile similar to infant ALL when compared with the group of c/ preB ALL: relatively more resistant to L-ASP and PRED (and in addition thiopurines), and more sensitive to AraC and 2-CdA. Age was not related to cellular drug resistance within the proB ALL group (o1 year, n ¼ 32, vs X1 year, n ¼ 19), nor within the MLL-rearranged ALL (o1 year, n ¼ 34, vs X1 year, n ¼ 8). The translocation t(4;11)(q21;q23)-positive ALL cases were more resistant to PRED (47.4-fold, P ¼ 0.033) and 4-HOO-ifosfamide (4.4-fold, P ¼ 0.006) than those with other 11q23 abnormalities. The expression of P-glycoprotein, multidrug-resistance protein, and lung-resistance protein (LRP) was not higher in infants compared to older c/preB ALL patients, but LRP was higher in proB ALL and MLL-rearranged ALL of all ages. In conclusion, infants with ALL appear to have a distinct in vitro resistance profile with the proB immunophenotype being of importance. The role of MLL cannot be excluded, with the t(4;11) being of special significance, while age appears to play a smaller role.