It is with great interest we read the scientific letter by Pigazzi et al and their finding that high ERG expression is an independent unfavorable prognostic marker for both event-free and overall survival in pediatric acute myeloid leukemia (AML) with MLL (11q23) rearrangement treated within the AIEOP LAM-2002 protocol.1 As the authors also point out, our recent study on pediatric AML patients enrolled in the NOPHO-1993 or NOPHO-2004 protocols supports this finding.2 We could show that high ERG expression is an unfavorable prognostic marker for event-free survival in this cohort of pediatric AML, where patients with MLL rearrangement were included.2 And indeed, this observation was even more pronounced if we analyzed the group of patients with MLL rearrangement separately for event free survival (Figure 1A, P = .03). However, there was no significant difference between low and high ERG expression for overall survival in this group (P = .16). High ERG expression was not identified as an independent marker in our study but this was mainly because BAALC and ERG were often found coexpressed at high levels, where high BAALC expression came out as a stronger predictor for prognosis in the multivariate analysis when the whole pediatric AML cohort was included. But, importantly, high BAALC was not statistically significant for event-free survival within the subgroup with MLL rearrangement (Figure 1B, P = .43) in contrast to ERG. There were too few patients in the MLL group (n = 40) in our study to perform a meaningful multivariate analysis, so we cannot comment on the independence of the ERG expression. However, we think that these 2 separate studies now have strengthened the evidence that also high expression level of ERG is an unfavorable prognostic marker in pediatric AML, in particular for the important group of patients with MLL rearrangements. The remaining challenge before gene expression levels can be used for prognostic stratification for clinical use is to establish standardized methods for quantitative measurement and to define relevant threshold values for high ERG and BAALC expression.
Acknowledgments: This work was performed on behalf of the Nordic Society of Pediatric Hematology and Oncology (NOPHO) and supported by grants from the Swedish government under the LUA/ALF agreement, the Swedish Childhood Cancer Foundation, the Åke Wiberg Foundation, Stockholm, and Lion's Cancer Foundation, Uppsala, Sweden.
Contribution: A.S., M.K., R.H., R.R., G.L., E.F., H.H., L.P., and H.E. contributed to the design and analysis of the study; L.P. wrote the manuscript and all authors agreed on the final version; A.S. and M.K. carried out laboratory-based research and A.S. performed statistical analysis; and J.A., E.F., K.J., O.G.J., B.Z., J.P., G.L., and H.H. were involved directly or indirectly in the care of patients and/or sample procurement.
Conflict-of-interest disclosure: The authors declare no competing financial interests.
Correspondence: Lars Palmqvist, Institute of Biomedicine, Departmentof Clinical Chemistry and Transfusion Medicine, University of Gothenburg, SE-413 45 Gothenburg, Sweden; e-mail: firstname.lastname@example.org.