Clonal trisomy 11 in a child with acute leukemia: G banding vs. FISH.

Translocations involving the MLL gene located at 11q23 have been reported in acute lymphoblastic leukemia (ALL), in 5±10% of cases with acute myeloid leukemia (AML), frequently of the monocytic type, and in biphenotypic leukemias expressing early stem cell as well as both myeloid and lymphoid markers. MLL gene rearrangement is the most common cytogenetic abnormality in infant leukemias and it can occur in therapyrelated leukemias as well. Trisomy 11 is the fourth most common acquired trisomy, occurring in 1% of AML/MDS cases [1]. Leukemias with trisomy 11 tend to express CD34, HLA-DR, myeloid antigens CD15, CD13 or CD33, and occasionally CD19. These leukemias are associated with a poor prognosis similar to cases with MLL gene rearrangements [1]. Recently, Schnittger et al. [2], in their study of 387 patients, reported the partial tandem duplication of the MLL gene, leading to the fusion of the proto-oncogene with itself, in 5.7% of AML patients with normal karyotypes, in 37.5% of cases with trisomy 11 with other cytogenetic abnormalities, and in 79% of cases with trisomy 11 as the sole karyotypic abnormality. Patients with this duplication had varying FAB morphologies and a poor outcome. In addition to their possible contribution to malignant transformation individually, the high incidence of partial tandem duplication of the MLL gene in cases with trisomy 11 suggests a link between these two cytogenetic events. Our experience with a 13-year-old south Asian girl is relevant. She had biphenotypic leukemia and at relapse, was found to have trisomy 11 detected by FISH, but not by conventional G banding. She presented with pancytopenia in December 1998. Family history was signi®cant for several cancers (liver, lung, colon, and brain) among close family members. A maternal aunt with Fanconi anemia developed AML at 5 years of age. Physical examination of the patient was not suggestive of Fanconi anemia and her diepoxybutane (DEB)-induced chromosomal breakage studies performed at the time of relapse were negative. At diagnosis, the patient's bone marrow aspirate revealed blasts of predominant L1 morphology; other blasts were large with prominent nuclei. The blasts were strongly positive for CD10, HLA-DR, CD19, CD22, CD34 and showed milder expression of myeloid markers CD15 and CD13. Cytochemical staining revealed the blasts to be positive with Sudan black and 3±5% also were positive with peroxidase. Conventional G banding revealed a small number of metaphase cells that were 46 XX. A FISH study for the MLL gene revealed two probe signals in both metaphase and interphase. The patient had a good response to a high risk ALL protocol that included high-dose methotrexate, cytarabine, anthracyclines, and teniposide. After 17 months of treatment, while on maintenance therapy, the leukemia relapsed in the bone marrow. Analysis of relapse blasts had similar surface markers but had lost CD34 and Sudan black positivity. The blasts showed no clonal Ig-H variable region rearrangements at that time. Again, conventional cytogenetic analysis showed a normal karyotype; however, FISH study using LSI MLL (11q23) dual color DNA probe (5 0 MLL spectrumgreen 3 0 MLL spectrumorange) (Vysis Inc., Downers Grove, IL) revealed three sets of separate fused signals in 89 of 200 interphase cells (44%), but not in any metaphase cells. The results suggested that the non-dividing cells were most likely trisomic for chromosome 11.