Impact of Late Relapses in the Outcome of Childhood Acute Lymphoblastic Leukemia IKZF1del and IKZF1plus: An Update Findings Previously Reported

Acute Lymphoblastic Leukemia (ALL) in children with IKZF1plus has been described as a group with a poorer prognosis than IKZF1 deletion (IKZF1del) by referral groups. However, we have found and reported similar results for both these two ALL subsets, and other groups confirmed this finding later. We have updated the outcome of IKZF1del and IKZF1plus ALL, focusing on the relapse patterns of both groups.

Patients and Methods: From October 2009 to March 2023, 1304 consecutive patients were diagnosed by SAHOP group. Of them, 37 (2.8%) presented early deaths and 88 (6.8%) deaths in complete remission (CR), defining a study group for the analysis of 1179 patients. They were classified as Group 1: cases without IKZF1 deletion, Group 2: IKZF1del, and Group 3: IKZF1plus (IKZF1del with additional gene deletions). LFS probability was analyzed using the Kaplan-Meier calculation and compared with the log-rank test. The comparison of the median time of CR was calculated with Wilcoxon rank sum test. All patients were accrued at the ALLIC-BFM 2009 Protocol and the Pilot study for the ALLIC-BFM 2022 Protocol.

Results: Group 1 included 967 (82%) cases, Group 2: 108 (9.2%) and Group 3: 104 (8.8%). We detected 168 relapses in Group 1, 36 in Group 2, and 37 in Group 3. Relapses in Group 1 occurred at a median time after achieving CR of 25 (r:1-64) months, and 91/168 (54%) of them at 24 months from CR (44 of them later than 36 months from CR). In Group 2, the median time for relapses was 29 (r:2-62) months, and 22/36 (61%) relapses occurred after 24 months from CR (10 after 36 months). In Group 3, the median time for relapses was 26 (r:3-60) months, and 21/37 (57%) patients relapsed after 24 months from CR (8 of them at >36 months). The p-value of the median time of CR comparing Group 1 vs. Group 2 and 3 was 0.0011, and 0.566 for Group 2 vs Group 3. In Group 2, 1 of the relapsed patients was treated as Standard Risk and 9 as Intermediate Risk (IR). In Group 3, 12 of the relapsed patients received treatment as IR. The cumulative incidence of relapses (CIR) was, in Group 1: 21.1%, Group 2: 39.8% and Group 3: 40.5%. We also detected 4 second malignant neoplasms in Group 1, 2 in Group 2, and 1 in Group 3. Group 1 achieved LFSp (SE) of 79 (2)%, Group 2 achieved 60 (5)% and Group 3, 58 (5)% The p-value for CIR and LFS comparing Group 1 vs. Group 2 and 3 was <0.00001; however, the p-value between Group 2 vs Group 3 was 0,55.

Discussion: Deletions in the IKZF1 gene have been defined as an adverse prognostic factor for childhood ALL, and efforts have been made to define different subsets of patients within this group. Initial reports from the BFM and AEIOP groups have defined a subset of IKZF1plus as the poorest prognosis group. However, other groups define as well as HR-ALL ALL with IKZF1del, regardless of additional deletions. We confirmed this in our setting, disclosing almost the similar CIR and LFS probability for IKZF1del and IKZF1plus subgroups, and a longer follow-up allowed us to observe later relapses in both subsets of IKZF1 family. Nevertheless, in both IKZF1 subsets, almost half and a third of the relapsed patients had been treated as IR, based on protocol criteria.

Conclusions: 1-IKZF1del ALL achieved a similar outcome when compared with IKZF1plus cases, in our setting. 2-Patients with ALL IKZF1del and IKZF1plus relapsed significantly later than patients without deletion of IKZF1; however, no difference was observed when comparing IKZF1del and IKZF1plus. 3-These patients could benefit from the intensification of the chemotherapy schedule. 4-More data is needed for defining a more accurate therapy for these patients, considering both the low incidence of these genetic abnormalities, and the different results among referral groups.