IKZF1 alterations are not associated with outcome in 498 adults with B-precursor ALL enrolled in the UKALL14 trial

Rachel J Mitchell, Amy A Kirkwood, Emilio Barretta, Laura Clifton-Hadley, Emma Lawrie, SooWah Lee, Daniel Leongamornlert, David I Marks, Andrew K McMillan, Tobias F Menne, Elli Papaemmanuil, Bela Patel, Pip Patrick, Clare J Rowntree, Nahid Zareian, Krisztina Zuborne Alapi, Anthony V Moorman, Adele K Fielding, Rachel J Mitchell, Amy A Kirkwood, Emilio Barretta, Laura Clifton-Hadley, Emma Lawrie, SooWah Lee, Daniel Leongamornlert, David I Marks, Andrew K McMillan, Tobias F Menne, Elli Papaemmanuil, Bela Patel, Pip Patrick, Clare J Rowntree, Nahid Zareian, Krisztina Zuborne Alapi, Anthony V Moorman, Adele K Fielding

Abstract

IKZF1 deletions (ΔIKZF1) are commonly detected in B-precursor acute lymphoblastic leukemia (ALL; B-ALL) and are widely assumed to have a significant impact on outcome. We compared the ability of multiplex ligand-dependent probe amplification (MLPA) and polymerase chain reaction (PCR) to detect ΔIKZF1 and to determine the impact on event-free survival of patients with precursor B-ALL aged 23 to 65 years recruited to the completed trial UKALL14 (ISRCTN 66541317). From 655 recruits with BCR-ABL1+ and BCR-ABL1- B-ALL, all available diagnostic DNA samples (76% of the recruited population) were screened by multiplex end point PCR covering 4 deletions: dominant-negative (DN) Δ4-7 or the loss of function Δ2-7, Δ4-8, and Δ2-8 (n = 498), MLPA (n = 436), or by both (n = 420). Although patients with BCR-ABL1- ΔIKZF1 were more likely to have minimal residual disease at the end of induction, we did not find any impact of ΔIKZF1 (including subgroup analysis for DN or loss-of-function lesions) or the IKZF1plus genotype on event-free, overall survival, or relapse risk by univariable or multivariable analyses. Consistent with the technical approach, MLPA not only detected a wider range of deletions than PCR but also failed to detect some PCR-detected lesions. The main difference between our study and others reporting an association between ΔIKZF1 and outcome is the older age of participants in our population. The impact of ΔIKZF1 in ALL may be less marked in an older population of patients. Our study underscores the need for analyses in large, harmonized data sets. This trial was registered at www.clinicaltrials.gov as #NCT01085617.