Relationship between transient abnormal myelopoiesis and acute megakaryoblastic leukaemia in Down's syndrome.

Abstract

We read with interest the recent review article (Lange, 2000) on neoplastic haematopoietic disorders in children with Down's syndrome (DS); this suggests that acute megakaryoblastic leukaemia (AMKL) in DS, having an antecedent history of transient abnormal myelopoiesis (TAM), may have evolved from the original TAM clone. The relationship between TAM and AMKL in DS is intriguing. Whereas, the clinical course of TAM is that of spontaneous resolution, around 25% of these patients will later relapse as AMKL, usually within 3 years (Avet-Loiseau et al, 1995). Conversely, short of a prospective study, most of AMKL in DS reportedly arise in children who have had a history of TAM in the neonatal period (Zipursky et al, 1997). Leukaemic blasts in TAM and AMKL shared the same immunophenotype. Taken together, these clinical observations raise the possibility that the neoplastic clone in TAM and AMKL is related, and that the acquisition of additional genetic events may be responsible for subsequent relapse of the TAM clone manifesting as AMKL. We analysed four consecutive cases of AMKL in DS diagnosed at our hospital, and the results are given in Table I. The blast count of less than 30% in two patients (cases 3 and 4) may warrant a diagnosis of myelodyplastic syndrome. Nevertheless, immunophenotyping of leukaemic blasts in all cases uniformly showed expression of platelet markers, consistent with a diagnosis of AMKL as seen in DS (Kojima et al, 1990). Three patients (75%) had a history of TAM during the neonatal period in which chromosomal analysis showed trisomy 21 (+21) only, without additional changes. In contrast, all patients with or without TAM phase showed cytogenetic aberrations in addition to + 21 at AMKL. This supports the contention that cytogenetic abnormalities are indeed the rule in AMKL phase (Ma et al, 1997) and may play a pivotal role in leukaemogenesis. To further detect genetic imbalance not evident on conventional cytogenetics, comparative genomic hybridization (CGH) was performed on bone marrow samples of patients 3 and 4 at AMKL presentation. Results showed no genomic amplification or deletion apart from +21, most probably reflecting small clone size and, therefore, escaping detection by CGH. However, the high mitotic activity enabled karyotypic aberrations to be detected by conventional cytogenetic study. To prove that TAM and AMKL in DS represent different phases of the same disease requires the demonstration of clonal identity at TAM and AMKL in the same patient. This was not possible in our series because of a lack of TAM specimens. The deciphering of genetic events that cause re-emergence of the TAM clone is important because they may be involved in the de novo leukaemic process as well. We have previously shown by fluorescence in situ hybridization that trisomy 8, a common cytogenetic abnormality in AMKL phase, is only present in a proportion of blast cells and, hence, merely represents a secondary or evolutionary change (Ma et al, 1999).