1. ¹Center of Pediatric Hematology Oncology, Azienda Policlinico University of Catania, Italy
2. ²Cancer Cytogenetic Laboratory, Department of Pediatric Hematology Oncology Istituto "Gianina Gaslini", Genova, Italy

Correspondence: Luca Lo Nigro, Center of Pediatric Hematology Oncology, University of Catania, via Santa Sofia 78, 95123 Catania, Italy. Fax: 39-095-222532

Received 30 October 2002; Accepted 8 November 2002.

TO THE EDITOR

We read with considerable interest the recent report by Rubnitz et al,¹ regarding the occurrence of t(9;11)(p22;q23) as the most favorable genetic factor for children with acute myeloid leukemia (AML) who were treated at their institution. Although we agree with the authors' conclusions, we would emphasize the unfavorable prognostic value of t(9;11) occurrence in association with additional chromosomal alterations and/or with a FAB different from M5. In order to support this concept, we report here a case of a 2-year-old boy presenting with an AML FAB M1 and low white blood cell (WBC) count who was diagnosed and treated with a BFM protocol at the Center of Pediatric Hematology Oncology, University of Catania. The cytogenetic analyses, performed at the Gaslini Hospital, showed the presence of the t(9;11)(p22;q23) with additional chromosomal abnormalities [46,Y,der(X) (9qter->9q34::Xp11.2->Xqter) (der(9)(11qter->11q23::9p22->9q34::17q2?4->17qter), der(11) (11pter->11q23::9p22>9pter), der(17) (17pter->17q2?4::Xp 11.2->Xpter). The child achieved morphological remission after two courses of induction therapy. Since he withdrew the consolidation phase, he suffered from a chronic thrombocytopenia. The bone marrow (BM) aspirate showed a low number of blasts. We consistently detected the fusion transcript MLL-AF9 using a recently described RT-PCR protocol.² The child then underwent a reinduction therapy (IDA-FLAG), achieving a partial morphological remission. As a last resort, we subsequently performed an allogeneic BM transplantation (BMT) from a sibling HLA-matched donor using low doses of cyclosporin as graft-versus-host (GVH) disease prevention in an attempt to induce a graft-versus-leukemia phenomenon. Interestingly, the patient achieved remission after a cutaneous grade III GVH. However, at 60 days from BM infusion, he presented with second isolated (BM) relapse. As a result of the high expression of CD33 antigen in patient's blasts, we decided to treat him with mylotarg followed by a second BMT. Unfortunately, after the second administration of anti-CD33 drug, the child died of infectious complication (aspergillosis).

Here we report a single case that in addition to others previously reported in the literature,^1,3,4 indicates that the occurrence of t(9;11) in childhood AML associated with additional chromosomal alterations and/or a FAB different from M5 could be considered an adverse prognostic factor. Even if in only a few pediatric cases, the trend of this supposition found in the authors' findings showed that the t(9;11) positive patient 18 (FAB M7) and patient 21 (FAB M5) both presented with additional translocations have died and relapsed, respectively.¹ In addition, the authors showed that 14 out of 17 cases with a FAB different from M5 associated with 11q23 abnormality and/or MLL gene rearrangement have died.¹ Although having a low WBC at diagnosis, which was demonstrated by the authors to be an independent good prognostic factor, 13 of these 17 cases died.¹ To date, recent evidence has showed that acute leukemias with MLL gene rearrangement seem to represent a different disease from both acute lymphoblastic leukemia (ALL) and AML, presenting with a specific gene expression profile.⁵ Reporting our case, we suggest that the occurrence of t(9;11) at an immature stage of myeloid committed stem cell is associated with a resistant disease and a poorer prognosis. In contrast, the use of a specific chemotherapy drug (2-chlorodeoxyadenosine) is effective against FAB M5 AML, as recently reported.⁶ Thus, it is conceivable that the type of hematopoietic cells targeted by MLL rearrangements influences the drug sensitivity of transformed blasts. According to recent evidence on childhood ALL with 11q23 alterations,^7,8 it is possible that leukemic cells with MLL rearrangement involve early progenitor cells that are more likely to exhibit drug resistance than are cells from patients older than 1 year or, in case of AML with FAB M5, harboring the same genetic change.