BACKGROUNDDeletions in IKZF1 are found in approximately 15% of children with B-cell precursor acute lymphoblastic leukemia (BCP-ALL). There is strong evidence for the poor prognosis of the most common IKZF1 deletions affecting exons 4-7 (DEL 4-7) and exons 1-8 (DEL 1-8), but evidence for the remaining 33% of cases harboring other variants of IKZF1 deletions is lacking. In an international multi-centre study we analyzed the prognostic value of these rare variants.METHODSMultiplex ligation-dependent probe amplification (MLPA) assays were performed on genomic DNA from patients’ bone marrow aspirates at diagnosis by the national study groups. Each IKZF1-deleted case was matched to three wild-type controls based on cytogenetic subtype, treatment protocol, stratification arm, white blood cell count and age at diagnosis. Known high-risk factors age <1 year (infants), BCR-ABL1-positive, and MLL-rearranged cases were excluded. We compared the cumulative incidence of relapse with death as competing event (CIR) between cases and their controls using Gray’s test. Matched pair Cox regression was used for event-free survival (EFS) analysis, and the hazard ratio (HR) with 95% confidence interval (CI) was reported.RESULTSWe included 134 BCP-ALL cases with a rare IKZF1 deletion and 402 matched controls. Of these cases, 26 (19%) had a deletion in exon 2 to 3 (DEL 2-3), 32 (24%) in exon 2 to 7 (DEL 2-7), 15 (11%) in exon 2 to 8 (DEL 2-8), 27 (20%) in exon 4 to 8 (DEL 4-8), and 34 (25%) belonged to the remaining group (DEL-Other). All rare IKZF1 deletion variants together had a higher 5-year CIR compared with the matched wild-type controls (40% vs. 22%, p<0.001), and a lower matched pair EFS (HR 1.8, 95% CI: 1.4-2.3; p<0.001). Analysis of cases and matched controls within their own risk group (56 standard risk, 33 intermediate risk and 45 high risk cases), showed an unfavorable effect for rare IKZF1 deletions in all stratification groups.Rare IKZF1 deletions were found in all BCP-ALL subtypes. The frequency of ETV6-RUNX1-positive (12 cases, 9%), high-hyperdiploid (21 cases, 16%), and unclassified BCP-ALL (13 cases, 10%) was relatively low among rare IKZF1-deleted cases. Most cases were found in the B-other group (88 cases, 66%). These B-other cases had a higher 5-year CIR compared with wild-type controls (47% vs. 27%, p<0·001), which translated into a lower EFS (HR 1·8, 95% CI: 1·3-2·4, p=<0·001). CIR and EFS analysis of high-hyperdiploid cases revealed a weak trend for an adverse outcome associated with rare IKZF1 deletions (5-year CIR 29% vs. 18%, p=0·1 and HR 2·4, 95% CI: 0·8-6·7, p=0·1). No prognostic impact was seen for rare IKZF1 deletions in ETV6-RUNX1-positive BCP-ALLSeparate analyses per IKZF1 deletion type showed a higher 5-year CIR for DEL 2-7 (38% vs. 18%, p=0.05), for DEL 2-8 (60% vs. 31%, p=0.02), and for DEL-Other cases (45% vs. 24%, p=0.04). Matched pair analysis of EFS revealed a poor prognosis for DEL 2-7 (HR 2·0, p=0·03), DEL 2-8 (HR 2·2, p=0·002), and DEL-Other (HR 2·2, p<0·001). The CIR and EFS of DEL 2-3 cases displayed a trend for unfavorable outcome (5-year CIR 28% vs. 17%, p=0.06; HR 1.8, p=0.1) but not for DEL 4-8 (34% vs. 26%, p>0.1; HR 1.0, p>0.1). The prognosis of each rare variant, including DEL 2-3 and DEL 4-8, was equal or worse compared with the most frequently observed and unfavorable prognostic DEL 4-7 and DEL 1-8 variants.CONCLUSIONSAll types of rare IKZF1 deletions, with the possible exception of DEL 4-8 cases, had a significantly increased risk of relapse and poorer EFS compared with their matched wild-type controls. The prognosis of DEL 4-8 cases was as poor as those of the other rare variants and that of the known high-risk variants DEL 4-7 and DEL 1-8. We therefore conclude that all variants of IKZF1 deletions are equivalent in terms of their prognostic impact. DisclosuresNo relevant conflicts of interest to declare.