Search

We used polymerase chain reaction amplification of minisatellite sequences or of a Y chromosome-specific sequence and Southern blotting to analyse long-term engraftment (12-82 months) after bone marrow transplantation (BMT) for familial haemophagocytic lymphohistiocytosis (FHL). Six children aged from 1 to 18 months were transplanted with bone marrow from an HLA-identical sibling in five cases and from an HLA-nonidentical related donor (one mismatched HLA antigen) in one. The conditioning regiment included VP 16-213 (900 mg/m2), busulfan (16 mg/kg), cyclophosphamide (200 mg/kg) and, in one case, aracytine (2 g/m2). Four patients are alive without therapy more than 3 years after BMT; the other two relapsed 1 year after BMT. DNA was extracted from separated polymorphonuclear cells and mononuclear cells, as well as from separated E+ and E- cells in one case and CD16+ (natural killer) and CD16- cells in two cases. Engraftment was partial in the four long-term survivors. Recipient cells were largely predominant in three of them as well as in one of the patients who relapsed (the donor also developed FHL 18 months after BMT). E+, E-, CD16+ and CD16- cells presented the same pattern of chimaerism. Engraftment failed to occur in the patient who received an HLA-nonidentical bone marrow. These results indicate that partial engraftment is compatible with long-term remission of FHL and that the presence of a small proportion of cells of donor origin can prevent FHL-related lymphocyte and macrophage activation.

Objectifs L’existence d’une hyperdiploidie est retrouvee dans 25 %-30 % des LAL de la lignee B de l’enfant, et peut etre associee a des rearrangements geniques de la voie Ras. L’objectif de cette etude retrospective est de decrire ces anomalies et leur role pronostic eventuel dans ce groupe phenotypique. Patients et Methodes 113 pts pris en charge pour une LAL avec hyperploidie a l’hopital Trousseau entre 2000 et 2008 ont ete repertories. Une analyse de mutation concernant les genes FLT3, NRAS et PTPN11 a ete realisee associee a l’etude de l’expression du gene FLT3 par RQ-PCR. Resultats Des mutations des genes FLT3, NRAS et PTPN11 ont ete retrouvees chez 24,3 % (n = 25) des patients. Il n’a pas ete mis en evidence d’association entre ces mutations et âge, sexe, hyperleucocytose, corticosensibilite, MRD et EFS. Nous avons retrouve une surexpression de FLT3 dans ce groupe comparees aux autres groupes cytogenetiques ( p p = 0,04). Conclusion Notre etude confirme la relative frequence des mutations de FLT3, NRAS et PTPN11 dans les LAL hyperdiploides et le role de la voie RAS dans les evenements genomiques successifs associes a la leucemogenese des LAL de l’enfant.

Abstract 1559 Poster Board I-582 Introduction Median age of Hodgkin's Lymphoma (HL) cases in children is 12-14 years. Epidemiological and histological features are similar to HL occurring in young adults. In contrast, HL under the age of 10 has specific features; it predominates strongly in boys and it is more frequently associated with Epstein-Barr virus (EBV). There is accumulating evidence for an inherited susceptibility to HL based on many reports of familial aggregation of the disease in adults and in childhood. Cohort studies have shown that patients affected by several immune deficiency syndromes - e.g. X-linked lymphoproliferative syndrome (XLP), functional deficit of Fas/FasL pathway and common variable immuno-deficiency (CVID) - are risk factors of HL. They are currently thought to explain only few cases of HL. However, their frequency may be underestimated since, to our knowledge, no study has systematically looked for these disorders among patients with HL. Methods We intend to search for qualitative and quantitative immune deficiencies as susceptibility factors to child's HL in a prospective study related to Euronet –PHL C1 protocol. Patients and their parents are invited to participate to the present study at inclusion time. Lymphocyte sub-populations are analysed by flow cytometry using various panels of monoclonal antibodies directed against the following markers: CD45, CD3, CD4, CD8, TCRαβ, TCRγΔ, CD19, CD5, kappa, lambda, CD27, IgM, IgD, CD21, CD38, CD16, CD57 and CD56-CD16. Immune study is completed by immunoglobulin and IgG subclasses quantification and anti-tetanus serology. Results Sixty patients at diagnosis of HL have been analysed. Median age of the study population is 13 years, (5-18 years). Gender-ratio M/F is 1.1. It increases to 6 below the age of 10. An unexpected high frequency of B-cell lymphopenia has been detected in 20 out of 60 patients (33.3%) (median 68 cells/μl, range 19-90) (Table 1). Immunoglobulin and anti-tetanus IgG levels are normal in all these cases. T-cell lymphopenia is observed in a low proportion of cases (3/60 (5%), HIV infected patients being excluded) (Table 1). These B and T-cell quantitative abnormalities do not correlate with age or gender. In contrast to the high frequency of quantitative defects, we have not detected any qualitative defect of specific B- or T-cell subpopulations. Moreover, the analyses of 26 parents of the lymphopenic patients has not shown any decreased lymphocytic population. Finally, our analyses exclude XLP, functional deficit of Fas/FasL pathway and CVID in all cases. Discussion The present detailed immune analysis of 60 children diagnosed with HL has not detected any characteristic immune deficiency syndrome confirming their low prevalence among children with HL. However, we have found a high frequency of B-cell global lymphopenia. Two studies had previously reported severe B-cell lymphopenia in patients with HL. The pathophysiological mechanisms leading to lymphopenia remain largely unknown. Since we did not find any qualitative immune defect – in B-cell subpopulation or in serological immune analysis - nor any defect among the parents we analysed, we propose that this peripheral blood lymphopenia is likely to be due to lymphocyte trafficking and homing rather than due to an absolute quantitative deficiency. Disclosures No relevant conflicts of interest to declare.

To the editor: In 2007, the germline loss-of-function mutations in SPRED1 were reported to originate a new autosomal dominant human disorder with multiple cafe-au-lait spots, axillary freckling, macrocephaly, and learning difficulties.[1][1],[2][2] This disorder belongs to the recently identified

Pediatric AML still represent an unfavourable disease resulting from the heterogeneous clonal expansion of malignant transformed haematopoietic stem or progenitor cell. The leukemia cell population is continuously replenish by rare, functionally distinct “leukaemia stem cells” (LSC) endowed with the capacity to self renew as well with the ability to generate clonogenic leukemic progenitors The AML-LSCs have been well documented and seem to behave like quiescent or slowly dividing hematopoietic stem cells. Therefore, LSC are considered less sensitive to treatments, which rather target actively dividing cells, and responsible for relapse. Recently, Y. Ma et al. suggested a major role of SALL4 gene both in stemness activity and leukemia transformation of normal hematopoietic stem cells. We sought to evaluate the expression of SALL4 gene in a panel of 88 pediatric AML, 60 Acute Lymphoblastic Leukemia (T and B ALL) and a few hematopoietic normal tissues. SALL4 expression was determined by quantitative RT-PCR in pre-treatment bone marrow samples (BM) (median blasts: 80%) and in normal tissues. SALL4 expression was much higher in AML compared to ALL (p 16, n= 12) had the worst outcome compared to the three others. Once stratified on MRC groups, MRC2 patients in the upper SALL4 quartile had 3.2 times more risk of relapse (HR= 3.2, CI95%: 1.3–7.8, P=0.02) and 5.4 more risk to die (HR=5.4, CI95%: 1.8–7.6; P= 0.0005) than MRC2 patients in the three others quartiles. In conclusion, SALL4 expression level may define an important risk factor in AML, particularly among patients with cytogenetic intermediate risk.

From 06/93 to 12/99, 1395 children and adolescents were included in the FRALLE 93 protocol: group A (Very Low Risk): 182, B (Standard risk): 672 and C (High risk including T-ALL): 541. The median FU is 9.5 yr. Inclusion criterias in group B were: age between 1 and 15 yr, WBC< 100,000, BCP phenotype and no poor-prognosis cytogenetic feature (MLL rearrangement or Ph1). Group B pts were randomised in a 2-steps schedule: at inclusion: daunorubicin (DNR: 40 mg/m2 × 2) vs idarubicin (IDA: 8 mg/m2 × 2), and at consolidation: high-dose IV methotrexate (HD-MTX): 4 × 8g/m2 at D1, D15, D29 and D42, vs low-dose po MTX (LD-MTX): 4 × 25 mg/m2; the number of triple IT was the same for both arms (N = 18). Pts with CNS leukemia were not randomized for MTX. During induction, pts were also classified on D21 bone marrow aspiration according to the level of residual blasts: M1 (< 5%): n=555, M2 (6–25%): n=71 and M3 (> 25%): n=41; M2/M3 pts received 1 more antracyclin infusion at D22 and M3 pts were subsequently treated according to the HR group and not randomized for MTX. Both M1 and M2 pts who reached CR were randomized for MTX. The overall group B EFS, DFS and OS were respectively 80±2, 81±2, and 91±1%. Globally there is no statistically significant impact of 1st or 2nd randomization on the outcome of patients. When we look at the outcome according to the D21 status, the EFS are very different: M1: 84±2%, M2: 63±6% and M3: 74±7%. The very poor EFS of M2 pts, who can not actually be classified as SR pts, and the fact that M1 pts represent 90% of pts randomized for MTX led us to analyse separately M1 pts. For M1 pts there is no interaction between 1st and 2nd randomization for survival, EFS and relapse rate (Gail and Simon test, p = 0.51); therefore pts were pooled for 2nd randomization analysis. Among M1 pts, 271 pts were randomized for HD-MTX and 264 for LD-MTX; there is no difference regarding age or leucocytosis between the 2 arms. Results are as follows: 5-yr EFS Relapses CI iBM relapses CI Other relapses CI 5-yr OS (EFS: event-free survival, CI: cumulative incidence, iBM: isolated bone-marrow, OS: overall survival) HD-MTX 87.82±1.99% 0.12±0.04 7.01±0.02 5.17±0.02 95.20±1.30 LD-MTX 79.92±2.47% 0.20±0.06 10.6±0.03 9.10±0.03 88.64±1.95 Test log-rank Gray Gray Gray log-rank p value 0.028 0.037 0.16 0.18 0.07 At the contrary, among M2 pts there is no difference in EFS for pts treated with HD-MTX (61±7%) and LD-MTX (65±9%) Conclusion: HD-MTX seems to benefit to pts with standard-risk BCP-ALL and good early response to chemotherapy, and the benefit is associated with a reduction in the number of relapses. No benefit is demonstrated in M2 pts with SER who have a very poor outcome and do require a more intensive treatment; the better EFS of M3 patients treated with double delayed intensification is in favour of this hypothesis.

L’association d’un syndrome lymphoproliferatif et de manifestations auto-immunes a ete surtout etudiee dans le cadre des lymphomes non Hodgkiniens de l’adulte. Rares sont les publications concernant la maladie de Hodgkin, en particulier chez l’enfant. Les objectifs de cette etude etaient de definir les caracteristiques de cette association et de rechercher des facteurs predictifs et pronostiques de l’evolution afin d’ameliorer la prise en charge therapeutique de la maladie de Hodgkin chez l’enfant. Une etude retrospective portant sur les vingt cinq dernieres annees est realisee aupres des centres membres de la Societe Francaise des Cancers de l’Enfant. Nous rapportons dix cas d’enfants porteurs d’une maladie de Hodgkin presentant des manifestations dysimmunitaires apparues avant le diagnostic ou au decours du traitement. Il s’agit de quatre purpuras thrombopeniques idiopathiques, un syndrome lupique, un syndrome des anticorps anti-phospholipides avec accident ischemique transitoire, un syndrome d’Evans, une vascularite leucocytoclasique, une anemie hemolytique auto-immune et une thyroidite auto-immune. Trois enfants ont au moins rechute une fois et deux sont decedes, dont l’un directement d’une pathologie auto-immune associee (syndrome lupique). Nous constatons une morbidite importante et des difficultes therapeutiques liees a l’auto-immunite. Les mecanismes physiopathologiques pouvant expliquer cette association sont egalement discutes. Afin de definir une population a risque pour laquelle des examens complementaires et des adaptations de traitement seraient necessaires, une etude comportant un plus grand nombre de patients est indispensable.

La nephroblastomatose (NBL) est un processus proliferatif qui s’associe volontiers avec des nephroblastomes bilateraux. La SFCE a propose une approche therapeutique homogene combinant chimiotherapie et chirurgie. La chimiotherapie etait systematique et prolongee, et la chirurgie indiquee seulement en cas de nephroblastome, certain ou suspect, sur les donnees de l’imagerie, lors du diagnostic ou de la surveillance. Cette etude retrospective a pour but d’evaluer les indications et les techniques chirurgicales dans l’evolution de la NBL. Patients En 29 ans (01/1979-12/2007), 23 enfants ont ete pris en charge pour NBL, 10 garcons et 13 filles avec un âge median de 12 mois (6 mois–5 ans). Il s’agissait d’une NBL bilaterale (20) ou unilaterale (3). L’indication chirurgicale etait portee uniquement en presence d’un nephroblastome, certain ou suspect, associe a la NBL, soit d’emblee au moment de la prise en charge initiale (29 reins), soit lors de la surveillance (6 reins : 2 masses stables pendant la chimiotherapie et 4 apparues ou progressant apres la chimiotherapie). Resultats Sur 35 reins operes, une chirurgie conservatrice a ete realisee pour 27 reins (77 %) et une nephrectomie totale pour les 8 autres. Les 23 enfants ont du etre operes, d’un seul rein (11 enfants) ou des 2 reins (12); il s’agissait histologiquement de 31 nephroblastomes et 4 nodules residuels de restes nephrogeniques. Parmi les nephroblastomes, 7 etaient de haut risque, 18 de risque intermediaire et 6 de bas risque. Avec un suivi median de 4,5 ans (4 mois–15 ans), une recidive est survenue chez 6 enfants (26 %) dont 4 haut risque et 2 intermediaire, tous apres chirurgie conservatrice. Une recidive locale isolee est survenue chez 3 enfants et une recidive avec metastases chez 3 autres. Quatre enfants sont decedes (debut de traitement avant 1983 pour 3). Une insuffisance renale moderee (1) et/ou une hypertension arterielle (2) etaient presentes dans 2 cas. Conclusion L’incidence du nephroblastome dans l’evolution de la NBL est elevee dans notre experience de centre chirurgical. Les masses apparaissant ou progressant sous chimiotherapie sont fortement suspectes de nephroblastome. La recidive parait frequente apres chirurgie conservatrice dans les nephroblastome de haut risque.

Objectifs repertorier les lymphomes de Burkitt (LBu) survenant apres transplantation d’organe et analyser leurs particularites. Sujets/materiels et Methodes Apres contact avec les services pediatriques de greffe d’organe en France et avec le Registre National des Hemopathies de l’Enfant, 13 cas de LBu prouves histologiquement, de moins de 18 ans (a), ont ete repertories entre 1997 et 2007, puis analyses retrospectivement. Resultats 11 cas etaient evaluables : 9 garcons, 2 filles. Les transplantations etaient renales [2 patients (pts)] et hepatiques (9 pts). L’âge median a la greffe etait de 14 mois (m) (foie), de 8a 7 m (rein). La seroconversion EBV est survenue a une mediane de9 m apres la greffe, et de 20 m avant le LBu. Au diagnostic, les greffes renaux avaient une clairance moyenne de 50 ml/min/1,73 m2, 3 greffes hepatiques presentaient un rejet chronique. Le delai de survenue median du LBu etait de 2 a 9 m (foie), de 1a 11 m (rein). Les stades du lymphome etaient : stade III (6 pts), stade IV (1 pt), LAL L3 (4 pts). 10/11 pts avaient une charge virale sanguine EBV elevee au diagnostic. Le traitement immunosuppresseur a ete diminue dans tous les cas (pas de rejet observe), 3 pts ont recu du rituximab sans efficacite. 10/11 pts ont ete traites selon le protocole LMB 96 ou LMB 2001. 3 pts ont ete dialyses au diagnostic (1 greffe renal) mais un est decede de syndrome de lyse a J2 de traitement. Les complications infectieuses comprennent des septicemies bacteriennes documentees (5 pts), des atteintes fongiques (aspergillose pulmonaire, candidose hepatique : 2 pts), 1 abces pulmonaire (1 pt). 1 pt a presente une toxicite neurologique severe de l’aracytine a haute dose. Les 9 pts effectivement traites par le LMB sont en remission complete avec un recul median de 4 a. Le pt traite par un protocole moins intensif (prednisone, cyclophosphamide) est decede en rechute. Conclusions Le lymphome de Burkitt de l’enfant transplante peut etre gueri par un traitement intensif de type LMB mais les complications infectieuses peuvent etre plus frequentes et severes. Le lymphome et sa prise en charge semblent avoir peu d’influence sur le greffon (toxicite directe, rejet). Nous remercions vivement les medecins et chirurgiens des centres de greffe pour leur collaboration .

In Children’s Langerhans cell histiocytosis, hematological dysfunction refractory to standard regimen, is the major cause of death until recently. In 2005, (Eur J Cancer, Bernard F et al, 2005) we have reported the results of a pilot study of 2-CdA and Ara-C in this extremely rare subset of patients. Among 10 patients enrolled between 1996 and 2004 (group 1), seven were cured, but three died, two of toxicity, and one after subsequent bone marrow transplantation. We report here A) the results of the 9 additional patients treated in France between mid 2004 and july 2007 (group 2) and B) the results of the national registry which had collected at the national level all cases of LCH in france since 1983. A) The group 2 comprised 9 patients who have all received at least two courses of Ara-C (1000 mg/m2/d) and 2-CdA (9 mg/m2/d) administered for 5 days. Maintenance therapy usually involved 6 courses of 2-CdA (5mg/m2 for 3 days every 3 weeks). Group 2 was comparable to group 1 for median diagnosis age (group 1: 0.78 years vs group 2: 0.73 years), as well as the number of organs involved. The delay between LCH diagnosis and 2-CdA Ara-C onset was shorter in the group 2 (group 1: 0.98 years- group 2: 0.35 years). The initial regimen to treat LCH was the same in the two groups (Vinblastine + steroid +- Methotrexate = LCH III protocol). Group 1 had received several second line therapies before 2-CdA Ara-C, while group 2 had received only the 2-Cda Ara-C as a second line therapy. Grade IV WHO haematological toxicities were observed in all patients, but no toxic death was observed in the group 2. One patient in group 2 underwent a 2-CdA overdose (injection at 10-fold the dose) but no major side effect was observed and latter the protocol was completed. After 2 courses, a decrease in disease activity was observed in all patients, but none had achieved a complete remission status after 2 courses. One patient had received a subsequent HSCT with an attenuated conditioning regimen with a very short hematological recovery at day 15, while the other patients had received only standard chemotherapy regimen as maintenance. Complete remission was achieved in the 8 assessable patients (including the patient who had received HSCT) and one patient had still an active disease but only 2 months after 2-Cda Arac onset. B) In the national French LCH register, 103 cases with hematological dysfunction, among 827 patients (

Purpose Nephroblastomatosis (NBis) is associated with a high risk of developing a Wilms tumor (WT). An homogeneous therapeutic approach was adopted combining chemotherapy (CT), and surgery only in case of WT, certain or suspected, on imaging features and possibly histology by transcutaneous needle biopsy, at diagnosis or during follow-up.This retrospective study evaluates indications and techniques of surgery in the evolution of NBis. Material and methods From 1979-oct.2006, 19 children were cared for in our surgical department,(median age :11 monthes; range 8m-5y). Nbis was bilateral in 17 cases. CT was given in all cases:Vincristine+Actinomycin, as maintenance treatment in case of isolated NBis, and in association with Doxorubicine for 4 to 8 weeks in case of WT. 19 patients were operated on, 10 on both kidneys and 9 on one. Operative procedure consisted in nephron sparing surgey (NSS) for 24 kidneys (83%) and a total nephrectomy in 5 .Surgical indication was decided either “on diagnosis” of a certain or suspected WT (24 kidneys), or later during follow-up for a mass suspect of WT(5 :1 mass stable during CT and 4 masses progressive or appearing after CT). Results No kidney was lost by ischemia. After histological review of the 29 specimens, 25 (86%) corresponded to WT in 17 children, and 4 to residual mass of NBis (3 operated “on diagnosis” and 1 for a progressive mass after CT). For the 25 WT, 5 were low risk tumor, 14 intermediate and 6 high. With a median follow-up of 5 years (4m-15y), recurrence occurred in 6 children (31%), 4 with high risk tumour and 2 with intermediate, all following NSS. Four patients died, among whom 3 who began their treatment before 1983. Mild renal insufficiency (1) and/or HT(2) were found in 2 patients. Conclusions Occurence of WT in NBis is very frequent in our surgical experience. In case of a high risk WT developing in NBis, local recurrences are frequent after NSS. Progressive or new masses after CT are strongly suspected of being a WT unlike masses that are stable during CT.

Introduction: Lymphocyte predominant Hodgkin’s lymphoma (LP) is a rare CD20-positive subtype of Hodgkin’s lymphoma which was only recently recognised as a separate disease entity. Most patients present with localised stage IA or IIA disease and usually, the disease has a relatively indolent clinical course. Up until recently, children with LP were treated alike classical Hodgkin’s lymphoma and received chemotherapy +/− radiotherapy. Two published reports of surgical resection alone in 19 children with LP (Pellegrino 2003, Murphy 2003) suggest that this may be an effective therapeutic option for a select group of children with limited stage disease. Study objectives: Can a proportion of children with limited LP be cured by surgery alone if resection is complete? Is a watch and wait strategy after surgery alone safe for those patients with residual disease after resection? What is the risk of relevant upstaging at relapse and of transformation into an aggressive B-cell lymphoma? Methods: European study groups participating in the EuroNet-PHL inter-group were requested to report their experience of surgery alone in children with LP. Surgery as single treatment modality has been used on a case by case basis for several years by the SFCE (France), the DAL / GPOH (Germany), and more recently by the UKCCSG (UK). Individual data of 57 patients were collected using a common CRF. 11 cases already published by Pellegrino were updated and are included. Results: 49 patients with initial stage IA achieved a CR after surgery. In 8 patients (5 in IA, 1 in IIA, 2 in IIIA) resection was incomplete. At a median observation time of 43 months (max. 172 months) all patients are alive. In the CR group 13 relapses were observed, all occurring early within 26 months of resection. Freedom from progression is estimated to be 67% (95% CI [59%; 75%]). Of the 8 patients who had incomplete resection 6 patients relapsed (p=0.008), of whom one (initial stage IIIA) relapsed as non-Hodgkin lymphoma. Of the 16 stage IA patients who experienced a relapse, 9 had a local relapse alone and the remaining 7 relapsed with stage II A. The rate of upstaging (B-symptoms or relapse stage > II) was 0% (95%-CI: [0%; 21%]). Conclusion: If complete resection is achieved, a substantial proportion (about 2/3 in our series) of surgically treated stage IA patients experience long-term remission and may be actually cured. Nevertheless, as most of the relapses occurred in the initially involved lymph node region, a better evaluation of the remission status after surgery may be obtained if FDG-PET is combined with CT/MRI. Based on these data we have designed a Europe-wide study in FDG-PET-negative stage IA-IIA LP to prospectively confirm these promising results in a larger cohort.

INTRODUCTION. Since the discovery of the cryptic t(12; 21) translocation, many secondary genetic abnormalities have been described in association with TEL-AML1 fusion gene. Extensive studies of karyotypes in a few series of TEL-AML1 positive leukaemia revealed a heterogeneous pattern of chromosomal abnormalities. Numerical and structural abnormalities are often present together. The modal chromosome number does not exceed 49 so that high hyperdiploidy and TEL-AML1 fusion are so far considered mutually exclusives. Here we reported that such association is found in a small group of patients and that relapse may still occur in those patients despite the good prognostic impact commonly attributed to each lesion. PATIENTS and METHODS. Between April 1994 and April 2006, 105 children were consecutively diagnosed with TEL-AML1 positive B-ALL. TEL-AML1 expression was detected by RT-PCR in Bone Marrow (BM) diagnostic samples and the t(12;21) explored by FISH with LSI TEL-AML1 dual-color probe (Vysis) in cases with low level of fusion gene expression or lacking molecular study. Conventional cytogenetics with G and R banding was performed on BM cells after overnight and 24 hours culture RESULTS and DISCUSSION. We detected numerical and/or structural abnormalities in 82/105 (78%) of the karyotypes. Aneuploidy alone was found in 4/105(3.8%): two cases had an extra chromosome 21, one an extra chromosome 10 and one lost 1 chromosome X. Structural abnormalities alone were present in 18/105 (17.1%) and up to 58 /105 (55.2%) presented both. The most frequent structural change was observed on chromosome 12p13 (50% of all structural abnormalities) whereas the most frequent chromosome gain was +21 (14%) and +10 (6.6%). Interestingly, in four cases (3.8%) we detected high hyperdiploidy with classical supplementary chromosomes; in two of these cases a partial trisomy of chromosome 1 was also present. Karyotypes are presented: Pt1: 52,XX,+10,+16,+18,+21,+21,+22 [22]/46, XX [4] Pt 2: 54,XY,+X,+4,+ 6?[del(6q)],+9,+14,+15,+18,+21,+21,-22 [17]/46,XY [3] Pt 3: 55,XX,+X,dup(1)(q21q31),+4,+6,+10,+14,+17,+18,+21,+21 [1]/56,idem+19 [4]/56,idem,+14 [5] / 46,XX [10] Pt 4: 7,XX,+der(X)(t(X;1)(q26;q12),+4,+5,+6,+8,+10,+14,+17,+18,+21,+21 [17]/46,XX [3] Pt1 present CNS relapse at 24 months follow up without involvement of BM which tested negative for TEL-AML1 expression. Patients 2,3,4 are in continuous complete remission at respectively 6, 6 and 4 years follow up. Since t(12;21) translocation and high hyperdiploidy seem to be primary events in leukaemia our observation raises the question if the two lesions coexist in a same cell or are independent events in different clones. In two cases (Pt 3,4), FISH analysis indicated the presence of TEL-AML1 gene fusion in 7% and in 5% of nuclei respectively. The distribution and the number of the signals on AML1 locus let think that t(12;21) occurred independently from hyperdiploidy. Microarrays studies have revealed distinct gene expression signatures associated to TEL-AML1 fusion and hyperdiploidy over 50 chromosomes, our observation points out the existence of cases which could be difficult to assign to one or the other of these classes.

Leukemia is the most common type of cancer in pediatric patients (pts). Treatment options for Philadelphia-chromosome positive (Ph+) acute lymphoblastic leukemia (ALL) and chronic myeloid leukemia (CML) include: risk-adapted chemotherapy, imatinib (im) and stem cell transplant (SCT). Relapsed leukemia has few therapeutic options. Dasatinib (SPRYCEL®, formerly BMS-354825) is a novel, oral, multi-targeted kinase inhibitor of BCR-ABL, KIT, and SRC kinases that was recently granted approval for adults with CML, and Ph+ ALL with resistance or intolerance to prior therapy. CA180018 is the first trial evaluating dasatinib in pediatric patients pts and Ph+ and in Ph- leukemias. CA180018 is a phase I/II dose-finding study of dasatinib in pts aged 1–20 with CML or relapsed Ph+ ALL resistant or intolerant to im, or Ph- ALL or AML in ≥2nd relapse performed in collaboration with 12 centers (6 countries) from the ITCC Consortium. Preliminary data are available on the first 15 pts treated from March-July 2006 (2 chronic phase (CP) CML, 7 Ph+ ALL, 1 accelerated phase (AP) CML, 3 Ph- ALL, and 2 Ph- AML) at a starting dose of 60 mg/m2 daily (course = 3 weeks). Intra-patient dose escalation was allowed for lack of initial response, and a 3+3 design for maximum tolerated dose (MTD) determination. Hematologic, cytogenetic, and molecular responses, as well as plasma and cerebrospinal fluid (CSF) pharmacokinetic (PK) analysis and BCR-ABL mutational analysis are ongoing. Median age was 11 yrs (range 4–17), median time from diagnosis of leukemia was 19.5 months (range 1.4–89.9). Prior therapy included chemotherapy, im, and stem cell transplant. Median duration on study is 0.69 month (range 0.03–3.45). Intra-patient dose escalation to 80 or 100 mg/m2 occurred in 7 pts. Six pts remain on study. There have been 7 responders: 4 complete hematologic responses (CHR): 1 CP CML, 1 AP CML, 2 Ph+ ALL; 5 cytogenetic responses (CyR): 2 Ph+ ALL complete CyR (CCyR), 1 AP CML CCyR, 1 CP CML partial CyR (PCyR), 1 Ph+ ALL minor CyR who never achieved a CHR, and then progressed; and 2 Ph+ ALL pts with CNS disease who cleared the CSF of leukemic blasts with single agent dasatinib. Preliminary PK in 7 pts showed rapid absorption with a median Tmax of 1.0 h, and mean terminal phase half-life (SD) of 2.7 (1.1) h. There is one dose-limiting toxicity (DLT): grade 4 anaphylactic shock which occurred 5 hours after the first dose. Dasatinib has otherwise been well tolerated up to 100 mg/m2 with toxicities including grade 3/4 thrombocytopenia, and sepsis. Nine pts are off study: 8 for progressive disease (2 Ph+ ALL pts with a resistant T315I BCR-ABL mutation), and 1 DLT. These preliminary results provide evidence supporting the safety and efficacy of dasatinib in pediatric pts with relapsed or refractory leukemia. An updated analysis including further enrollment, safety data, PK, cytogenetic and molecular responses, and mutational analyses will be presented.

Background: Nodular Lymphocyte Predominant Hodgkin’s Lymphoma (NLPHL) is a distinct clinicopathologic subtype of Hodgkin’s lymphoma [HL]. Reports in published literature suggest that very little, and occasionally, no chemotherapy is sufficient for long term survival in patients with LPHL and that this disease appears to have a more indolent course than classical HL. The optimum therapy for early stage NLPHL is not known, but it is likely that current therapy for classical Hodgkin’s lymphoma is unnecessarily toxic and intensive for this indolent, slowly progressive, CD 20 positive disease. Adverse treatment related late effects are the major causes of morbidity and death. Consequently children with early stage NLPHL represent ideal candidates for low intensity treatment. Study Objectives: To assess whether a non-intensive chemotherapy regimen consisting of cyclophosphamide [500mg/m2 iv- day 1], vinblastine [6 mg/m2 iv-day1 & 8] and oral prednisolone [40 mg/m2 - days 1–7] (CVP) every 14–21 days, count dependent, could replace standard chemotherapy protocols used for classical HL without compromising efficacy in children and young people with NLPHL. Patients and Methods: Between May 2004 and April 2006 18 patients with stages IA [n=12] and IIA [n=6] biopsy proven NLPHL were treated with 3 cycles of CVP chemotherapy. Three of the eighteen patients were initially treated with surgical resection alone and received CVP at first relapse. Ten patients were males and the median age at diagnosis was 10 years (range 7–15 years). Staging investigations at diagnosis included both conventional cross sectional as well as 18 fluro-deoxyglucose [FDG] PET imaging. Remission status at the end of 3 cycles of CVP was confirmed by both conventional cross sectional and PET imaging. Results: 17 patients achieved a complete remission [CR] after 3 courses of CVP and 1 patient a very good partial remission [VGPR]. To date, only 1 patient has relapsed while all the remaining patients remain in continuous CR. Median duration of follow up is 12 months (range 2–26 months). The overall survival is 100% and event free survival is 94%. No significant early or late toxicity has been observed to date. Conclusions: Nothithstanding the relatively short follow up period, CVP is an effective non toxic chemotherapy regimen with 95% of patients with early stage NLPHL achieving a CR after 3 courses. Based on these preliminary results, we have designed a prospective European study for patients with stage IA & IIA NLPHL to confirm these results in a larger cohort of patients.

9003 Background: Imatinib mesylate inhibits selectively specific activations of the platelet-derived growth factor receptor (PDGFR), c-KIT and BCR/ABL tyrosine kinases and is approved for the treatment of chronic myeloid leukemia and gastro-intestinal stromal tumors (GIST). This study evaluated efficacy of imatinib in solid childhood tumors. Methods: Phase II study of imatinib as single agent in children and adolescents with refractory or relapsing solid tumor expressing at least one of the receptors. Patients were to be treated at 340 mg/m2, a dose escalation allowed to 440 mg/m2 after 2 months in case of insignificant improvement. C-KIT, PDGFRα and β expression was determined on archive tissue sections by immunohistochemistry prior to study entry. Gene mutations, pharmacokinetics, pharmacogenetics, and positron emission tomography imaging were assessed. Results: 36 patients, 21 boys, median age 13.7 years (2.2–22.5 y), 12 with brain tumors, 6 fibromatosis, 8 mesenchymal/bone tumors, and 10 other solid tumors, including 1 GIST and 3 chordoma, were treated at 340 mg/m2 daily during a total of 168 months (median 1.9 month/patient, range 0.5–19). 18/36 expressed c-KIT, 10 PDGFRα, 21 PDGRβ; 12 expressed more than one receptor. Ten patients were escalated to 440 mg/m2 due to lack of efficacy. During the 1st month, 17 patients experienced mild toxicity (grade 1 and 2) related to study treatment: gastro-intestinal (n=22), face edema (n=7), asthenia (n=5), tumor induration (n=2), skin toxicity (n=2), thrombocytopenia (n=1). No partial or complete response was observed; 5 patients (2 fibromatosis, 1 GIST, 1 medulloblastoma, 1 pseudo-inflammatory tumor) experiencing durable stable disease have been under treatment for more than 12 months. Interesting tumor stabilization during 10 and 7 months, respectively, was achieved in a brain stem glioma and a renal carcinoma. Glucose uptake on 18FDG PET scan was reduced in a chordoma, although the child progressed and died due to disease. Pharmacokinetic and genetic data are currently evaluated. Conclusions: Imatinib as single agent was well tolerated, but—as used in our study —failed to show measurable anti-tumor effects according the standard criteria in the pediatric malignancies studied. No significant financial relationships to disclose.

The most frequent oncogenic activation events characterized in childhood T acute lymphoblastic leukemia (T-ALL) result in the transcriptional activation of genes coding for transcription factors. The main genes are TAL1/SCL, a member of the basic region helix-loop-helix gene family, HOX11L2 a member of the homeobox-containing protein family. Conflicting results have been reported concerning molecular epidemiology and prognostic values of these markers (Cavé Blood2003:103442–445, Ferrando Cancer Cell 2002, Dicciani Blood 2003 abs 67) we therefore analysed retrospectively 200 pts treated in the French protocol FRALLE 93 for T-ALL between 11/93 and 12/99. Pts were stratified according to prednisone response at D8 ( good or poor : GPR or PPR) and bone marrow at D21. Pts with D8 PPR or M3 received an intensified treatment with genoidentical or autologous transplant in CR1. Molecular analysis was possible for 79/200 T-ALL samples. Clinical caracteristics were not significantly different between population with or without molecular analysis male (n=121) (69% vs 70%), median age 8.4y (range1.1–19.5) vs 9.2y, median leucocytosis 140.109 (0.6–736) vs 171.915 109 (100 n=48), mediastinal involvement 72% vs 71% ,CNS+ 4% vs 5%, CD 10 neg 54% vs 52%. Steroid response PPR n=37/73, GPR n=36/73 and D21 bone marrow status (M3 n=10, M2 n=11) were similar. CR was obtained in 72/79 pts (91%) after first induction therapy and 2 deaths occurs during induction treatment. With a median follow-up of 63 months (2–123), 5 y OS , EFS and DFS is 62 %± 9 and 54% ±10. SIL-TAL1/SCL fusion was detected in 20/79 (25%) pts; expression of HOX11L2 was observed in 14/79 (17%) pts. These activations are mutually exclusive and they allow the subclassification of 42 % of the patients. Median leucocytosis was significantly higher in SIL-TAL1/SCL pts (p=0.01) but other significant features ( ie median age, D8 response, D21 status) were not significantly different between each group. OS and EFS for TAL1/SCL , HOX11L2+ and none of these were respectively 81%±10, 43%±13, 62%± 7 and 80%±10 , 42%± 13, 55%± 7. OS and EFS D8 PPR/GPR were 47%± 9 vs 78% ±7 (p=0.009) and 41%±8 vs 71%±7 (p= 0.008); OS and EFS M2M3 vs M1 D21 bone marrow status were 45%± 15 vs 70% ±6 (p=0.05) and 36%±10 vs 66%± 6 (p= 0.018). In multivariate Cox model analysis, D8 steroid response and HOX11L2 expression were significantly associated with adverse event (failure or relapse) Risk Ratio :3 and 2.6 - p=0.008 and 0.03 and a higher risk of death Risk Ratio : 4.1 and 3.4 -p=0.061 and 0.05. Finally HOX11L2 expression and D8 PPR are independently associated with poor outcome in FRALLE 93 protocol analysis which confirms our previous report. Discrepancies among series may be explained by confounding factors such as differences in median leucocytes value(140.109 in our study) and treatment.

From Dec 2000 to Dec 2003, 390 children with SR-BCP-ALL (age: 1–9, WBC 10−3, or according to the exact level of positivity. Results: 1) 15% (51/336 pts) of the SR-BCP ALL have a detectable MRD at EOI 2) As expected, whatever the threshold, pts with D21 M2/M3 marrow are more likely to have a detectable MRD (p=.0017). But 43/316 M1 pts (14%) have a highly +ve (n=19;6%) or weakly +ve MRD (n=24 ; 8%). Surprisingly, only 1 out 20 M2M3 pts had a MRD > 10−2 while it is the case for 8 out 316 M1 pts (p=NS). If only pts receiving DNR are considered (DNR+ M1 pts and all M2/M3 pts), again pts with D21 M2/M3 marrow are more likely to have a detectable MRD (p=.0015). 3) If we compare the MRD levels in the 2 arms (DNR+ve or neg) in the 315/343 M1 pts evaluable for MRD: 136 pts in each arm had no detectable MRD; 16 and 8 have a weak positivity in the DNR− and DNR+ arm respectively while 10 and 9 have a weak positivity in the DNR− and DNR+ arm respectively: p= .29). If the exact level is considered, this absence of difference remains at all levels considered. Conclusions: 15% of the SR-BCP ALL have a detectable MRD at EOI after a three or four-drug induction. D21 M2/M3 pts are more likely than M1 pts to have a detectable MRD at EOI but 14% of the D21 M1 pts have a high (> 10−3) or very high MRD (> 10−2) which confirms the added value of MRD detection to classical morphology. The MRD level at EOI is not different in SR-BCP-ALL after a three or four-drug induction regimen. This is of paramount importance since EOI MRD is a surrogate marker for probability of relapse.

It is hypothesized that AML arises from two cooperative types of mutations: type I mutations mainly induce proliferation and type II mutations involved in the maturation arrest. AML is a rare disease in children and few molecular data are available on pediatric AML. We therefore studied N-RAS, K-RAS, FLT3-ITD, FLT3 , C-KIT mutations (type I), and CEBPA mutations (type II) as well as FLT3, EVI-1 and WT1 gene expression in 77 de novo AML. Patients and methods: All the patients (aged 1 month-17 years, median age: 6.9 years, male/female ratio 1.26) were treated for de novo AML between 1995 and 2003 in two French institutions and prospectively enrolled in LAM91, LAM01 and APLs French collaborative protocols. According to the FAB classification the repartition was: M0:6.5%, M1: 5.2%, M2: 22%, M3: 13%, M4 :14.3%, M5 :30%, M7: 6.5% and unclassified :2.5%. Cytogenetics features according to the MRC classification were favorable, intermediate or poor in 25% (t(8;21) n=5; t(15,17) n=8, inv(16) n=5), 65% (normal n=20 and 11q23 abnormalities n=15) and 10% (−7, n=4) respectively. With a median follow-up of 26 months (range 2–98 months), Complete Remission was obtained in 92% (71/77) of patients, OS was 71% and EFS 61%. CEBPA, N-RAS, K-RAS, C-KIT and FLT3 mutations detection was performed by direct sequencing. FLT3, EVI-1 and WT1 transcripts were quantified by RQ-PCR. Results: (1) Frequency of N-RAS and K-RAS mutations were 11% (8/75) and 16% (12/75) respectively. RAS-mutated patients belonged to favorable (30%), intermediate (60%) and poor (10%) cytogenetic subgroups. In univariate analysis only N-RAS mutations is associated with adverse outcome (OS 37% vs 79%, p Conclusion: In total, 48% of de novo AML in children had a mutation in N-RAS, K-RAS, FLT3-ITD, FLT3 Asp835, C-KIT or CEBPA with a high frequency of RAS mutations (27%) compared to adult AML and a significantly bad survival. Additional gene expression quantification of EVI-1, FLT3 and WT1 allows MDR detection in 95% of patients.

The Wilms’ s tumor gene (WT1) is a tumor suppressor gene highly expressed in most acute leukemias. To determine whether WT1 gene expression is a valuable and informative marker for minimal residual disease in pediatric AML, we quantified WT1 transcript amount by RQ-PCR in 92 de novo AML and 20 normal controls. The WT1 transcripts obtained were normalized with respect to the number of TBP transcripts and expressed as WT1 copy numbers by the ratio WT1/TBPx1000. The WT1 levels were extremely low in normal controls, and the median number of WT1 copies was 10 (range 4–30 ). A level above 50 copies was considered as significant. All the patients (aged 2 months-18 years, median age:5.9 years, male/female ratio 0.87) were treated for de novo AML between 1995 and 2003 in two French institutions and enrolled in LAM91, LAM01 and APLs French collaborative protocols. According to the FAB classification the repartition was: MO 5.4%, M1 4.3%, M2 18.5%, M3 14.1%, M4 and M4Eo 12%, M5 33.6 %, M7 10.9% and unclassified 1%. Cytogenetics features according to the MRC classification were favourable, intermediate or poor in 27% (23/83), 59% (49/83) and 13% (11/83) respectively. With a median follow up of 24 months (range 8–97months) OS was 75± 8% and EFS 60 ± 6%. At diagnosis WT1 overexpression was detected in 78.3% (72/92) with a median copy number of 2231 (range 50-429200). The WT1 values were significantly higher (p=0.02) in M2-FAB subtype and lower (p=0.01) in M5-FAB subtype while no correlation was found with WBC count or cytogenetic abnormalities. WT1 quantification for MRD was evaluable in 41/72 pts and positive in 9/32 at D40-50, 5/25 at M3-M5, 5/7at M6-M8. At least one analysis above 50 copies after induction therapy is associated with a significant risk of subsequent relapse 8/11 vs 8/30 ( p=0.007) RR=22 (IC 95%:46–118) and death 10/14 vs 3/30 (p=0.001) RR=7.3 (IC95%: 1–34). Although retrospective analysis may include bias, we conclude that WT1 is a useful and informative molecular marker for MRD in pediatric AML and performed as prospective analysis in ELAM02 protocol.

Neurological involvement has been poorly described in autoimmune cytopenias (AIC) and association evokes an underlying primary immunodeficiency (PID). Among the 1,167 patients of the OBS’CEREVANCE nationwide prospective cohort which includes children with an AIC, we have identified 8 patients with a neurological involvement. Data were collected from all centers and radiological review was centralized. Genetic analyses were performed by a research department. With a median (range) follow-up of 12 years (6-26.5), 7 children had Evans Syndrome (ES) (with autoimmune neutropenia in 5) and 1 had autoimmune hemolytic anemia (AIHA). Neurological symptoms appeared were: seizures (n=4), cranial nerve palsy (n=2), Brown-Sequard syndrome (n=2) and/or sensory neuronopathy (n=1). No infectious pathogens were identified. MRI showed multiple (n=6) or unique (n=2) inflammatory lesions, histologically confirmed in 5 patients. In 4 cases, a lymphocytic meningitis was associated. Non-neurological organ involvement was present in all patients, mainly pulmonary nodules (n=6) and lymphoproliferation (n=4). All patients had a lymphocytes deficiency and 7 had a hypogammaglobulinemia. Patients have been given steroids (n=7), intravenous immunoglobulins (n=2) or immunosuppressive treatment (n=3), improving symptomatology and MRI for all. Five patients relapsed and 3 had an asymptomatic radiological progression. Four out of the 6 patients analyzed had a PID: 22q11.2 microdeletion (n=1), heterozygous CTLA4 mutation (n=2) or suspected homozygous LRBA mutation (n=1). In conclusion, neurological involvement is a rare and severe late event in childhood ES, or exceptionally AHAI, that may reveal various PID.; Les atteintes neurologiques ont été peu décrites dans les cytopénies auto-immunes (CAI) et évoquent un déficit immunitaire primitif (DIP). Parmi les 1 167 enfants de la cohorte prospective française de CAI OBS’CEREVANCE, nous avons identifiés 8 patients avec une atteinte neurologique. Les données ont été récupérées auprès de chaque centre et les IRM relues de manière centralisée. Les analyses génétiques ont été effectuées par un laboratoire de recherche. Les patients ont en médiane (extrêmes) ont été suivis 12 (6-26,5) ans. Sept avaient un syndrome d’Evans (SE) et 1 une anémie hémolytique auto-immune (AHAI). L’atteinte neurologique consistait en : convulsions (n=4), paralysie des nerfs crâniens (n=2), syndrome de Brown-Séquard (n=2) et/ou ganglionopathie (n=1). Aucun pathogène n’a été identifié. Les IRM montraient des lésions inflammatoires multiples (n=6) ou unique (n=2), confirmées histologiquement chez 5 patients. Les patients présentaient également une méningite lymphocytaire (n=4), une lymphopénie (n=8) et/ou une hypogammaglobulinémie (n=7). Tous présentaient une atteinte d’organe non neurologique, principalement des nodules pulmonaires (n=6) et une lymphoprolifération (n=4). Les patients ont été traités par corticoïdes (n=7), immunoglobulines polyvalentes (n=2) ou immunosuppresseurs (n=3), améliorant chez tous la clinique et l’imagerie. Cinq patients ont rechuté et 3 ont montré une progression radiologique asymptomatique. Quatre des 6 patients analysés avaient un DIP : microdéletion 22q11.2 (n=1), mutation hétérozygote de CTLA4 (n=2) ou mutation homozygote suspectée de LRBA (n=1). Cette association rare et sévère concerne surtout les SE et peut révéler un DIP.