Leukemia (2002) 16, 1685–1690  2002 Nature Publishing Group All rights reserved 0887-6924/02 $25.00 www.nature.com/leu Infant acute lymphoblastic leukemia – combined cytogenetic, immunophenotypical and molecular analysis of 77 cases A Borkhardt 1 *, C Wuchter 2 *, S Viehmann 1 *, S Pils 1 , A Teigler-Schlegel 1 , M Stanulla 3 , M Zimmermann 3 , W-D Ludwig 2 , G Janka-Schaub 4 , M Schrappe 3 and J Harbott 1 1 Children’s University Hospital, Department of Hematology and Oncology, Giessen, Germany; 2 Robert-Ro ¨ssle Cancer Center, Charite ` `, Humboldt University Berlin, Department of Hematology, Oncology and Tumor Immunology, Berlin, Germany; 3 Medical School Hannover, Department of Pediatric Hematology and Oncology, Hannover, Germany; and 4 University Hospital Hamburg, Pediatric Hematology and Oncology, Hamburg, Germany We used karyotyping, fluorescence in situ hybridization (FISH), Southern blotting, and RT-PCR in order to analyze prospec- tively 77 infants (less than 1 year of age) with acute lymphobl- astic leukemia for the occurrence of 11q23/MLL rearrange- ments and/or other cytogenetic abnormalities. Out of the 69 informative samples we found an 11q23/MLL rearrangement in 42 cases (61%). Regarding only pro-B ALL cases, the incidence of 11q23/MLL rearranged cases, however, reached more than 90% The infants were treated within the therapy studies ALL- BFM90, ALL-BFM95 and CoALL-05–92. For patients with an adequate follow-up of 4 years the event-free survival of the 11q23/MLL-positive and 11q23/MLL-negative group was 0.2 or 0.64, respectively (P = 0.024). The monoclonal antibody 7.1. (moab 7.1) does not react with normal hematopoetic precursors or mature blood cells but was shown to specifically react with leukemic cells bearing a rearrangement of chromosome 11q23 or the MLL gene, respectively. We, therefore, specifically addressed the question whether the reactivity of moab 7.1, as determined by flow cytometry, may substitute for molecular testing of an 11q23/MLL rearrangement in this cohort of infant ALLs. Reactivity of moab 7.1 indicated a 11q23/MLL rearrange- ment with a specificity of 100%. However, five of the 11q23/MLL-positive cases did not react with moab 7.1 indicat- ing a sensitivity of 84% only. Three of these five moab 7.1-nega- tive but 11q23/MLL-positive cases could be identified by their unique expression pattern of CD65s and/or CD15. Thus, 95% of all 11q23/MLL-positive ALL cases in infancy may be ident- ified by flow cytometry based on their expression of CD15, CD65s and/or moab 7.1. Leukemia (2002) 16, 1685–1690. doi:10.1038/sj.leu.2402595 Keywords: infant ALL; 11q23/MLL rearrangement; prognosis; immunophenotype; moAb 7.1; flow cytometry Introduction Thirty years ago, the prognosis of children suffering from acute lymphoblastic leukemia (ALL) was poor and most of them sur- vived only 3 or 4 months after diagnosis. Several years later, in the mid-1970s, the 5-year event-free survival (EFS) for older children approached nearly 50% while the cure rate for infants with ALL (diagnosed within the first 12 months of life) still remained extremely poor. At that time, the few survivors of infant ALL usually had substantial neurologic damage from cranial irradiation. Although the EFS of infants who were treated with current chemotherapy protocols has dramatically improved over the last years, their EFS is still much lower as compared to older children. Aside from children with ALL and BCR/ABL rearrangement, infants have the worst prognosis of all pediatric ALL patients with an EFS of about 35–50%. 1,2 Correspondence: A Borkhardt, Children’s University Hospital, Depart- ment of Hematology and Oncology, Feulgenstr. 12, 35392 Giessen, Germany; Fax: 641–9943429 *The first three authors contributed equally to this study Received 22 November 2001; accepted 5 April 2002 Infant ALL accounts for about 5% of all leukemic cases in childhood and is characterized by an initial high white blood cell count, bulky extramedullary disease, an immature immu- nophenotype and translocations involving the MLL gene at chromosome 11q23. The spectrum of recombination partners of 11q23 comprise more than 35 loci, with a predominance of a translocation t(4;11) and t(11;19). 3 The detection of these 11q23 aberrations is clinically most important as their occur- rence has frequently been associated with poor treatment out- come. 4–6 The careful evaluation of all infants with ALL for the presence of a rearrangement within the 11q23/MLL gene is therefore highly desirable. Previous studies have shown that the expression of the human homologue of the rat 220 kDa chondroitin sulfate proteoglycan (NG2) is highly correlated with 11q23/MLL abnormality. 7,8 As compared to the cyto- genetic or molecular methods for detection of 11q23/MLL abnormalities, the determination of the chondroitin sulfate expression by flow cytometry may be an attractive alternative approach with regard to cost and speed. The latter analysis is facilitated by the monoclonal antibody (moab) 7.1 which does not react with normal hematopoietic precursor cells but specifically reacts with 11q23/MLL-rearranged leukemic cells. 7 In a previous series, we evaluated the reactivity of moab 7.1 in a large series of adults and older children suffer- ing from various leukemias. 9 We extended our former analysis and report herein the immunophenotypic, cytogenetic and molecular data of 77 infants with ALL. Apart from determi- nation of the overall frequency of MLL/11q23 aberrations using karyotyping, FISH, Southern blotting and RT-PCR, we addressed the question whether flow cytometry of infant ALL with moab 7.1 and a panel of additional antibodies may sub- stitute for molecular testing of 11q23/MLL rearrangements. The infants were enrolled within three German multicenter therapy trials ALL-BFM90, ALL-BFM95 and CoALL-05–92. The prognostic impact of 11q23/MLL abnormalities and the immunophenoytpe for patients with an adequate follow-up of 4 years is also presented. Materials and methods Patient samples Between January 1994 and April 2000, 77 bone marrow and/or peripheral blood samples from infants were sent in by mail to the Oncogenetic Laboratory of the Children’s Univer- sity Hospital, Giessen, Germany. Informed consent from the guardians for each patient was obtained. The diagnosis of a previously untreated ALL was based on standard morphologi- cal studies, and cytochemical staining of leukemic cells. 10