A molecular classification of leukaemia reveals MDS as a disease continuum with non-leukaemia and AML sub groups Prof. K. Mills, Cardiff, United Kingdom    - Biography English - 2007-06-09 Topic: Myelodysplastic Syndromes
Speaker Disclosure I act as a consultant to Roche Molecular Systems who have supported this research

PLAY PRESENTATION	( 36 slide(s) )
PODCAST

VIEW THE SLIDE LIBRARY FOR THIS PRESENTATION

A molecular classification of leukaemia reveals MDS as a disease continuum with non-leukaemia and AML sub groups.

Ken I. Mills1,2,7, Amanda F Gilkes1, Jesus M. Hernandez4,7, Wolf-Karsten Hofmann,5,, Alexander Kohlmann6, Mickey Williams6, Lothar Wieczorek6, Torsten Haferlach3,7 . 1Department of Haematology, Cardiff, United Kingdom; 2CCRCB, Queens University Belfast, United Kingdom, 3MLL Munich Leukemia Laboratory, Munich, Germany; 4Hematología, Salamanca, Spain; 5Charité, Berlin, Germany; 6Roche Molecular Systems, Pleasanton, United States; 7MILE study, on behalf of European Leukemia Network (ELN), WP13

Robust gene expression signatures associated with distinct sub-classes of paediatric and adult leukaemias have been identified from microarray studies. Recently, the MILE (Microarray Innovations in LEukemia) study group has compared the accuracy of gene expression profiling to gold standard diagnostic workup for ~2000 patients within 16 acute and chronic leukaemia subclasses, MDS, and non-leukaemia as control group in 11 centres (ELN: 7, USA: 3, Singapore: 1). The overall cross-validation accuracy was very high for the leukaemia subclasses: >95%. Based on this data set a customized AmpliChip Leukemia microarray has been designed for leukaemia classification.
However, only 49.1% of the 173 MDS samples included in the study were correctly called as MDS from their underlying gene expression profiles. The remainder were approximately equally split between an algorithm classification call of “non-leukaemia” (24%) and “AML” (24.6%). Our analysis showed that neither centre nor age were a factor in differentiating between “MDS”, “MDS with an AML-like signature” or “MDS with a non-leukaemia like signature”. WHO classification was highly correlated with the microarray classification result; specifically RAEB (I or II) was associated with “AML-like signature” (p < 0.0001) whilst, RA/RARS was highly correlated with “MDS” or “non-leukaemia-like signature”. Furthermore, IPSS was significantly correlated with classification call (p>0.0001): 65% of patients with an IPSS score of Int-2 or above were classified as “AML”. Individually, the blast, karyotype and cytopenia contributions were highly significant (p