Impact of IKZF1 Deletions and IKZF1 Plus in Pediatric Acute Lymphoblastic Leukemia Outcome

Introduction: Survival of children with acute lymphoblastic leukemia (ALL) has improved in the last decades, achieving approximately 80% in Argentina. However, relapses remain the most frequent adverse event and the identification of patients with higher risk need to be refined. Deletions in IKZF1(IKZF1del) in addition with deletion of CDKN2A, CDKN2B,PAX5 or PAR1 region define a new subgroup of patients (IKZF1plus) with higher relapse rate and poor survival.

Objectives: To analyze the characteristics of patients with IKZF1del and IKZF1plus, assessing the impact of the copy number alterations in several genes on survival of pediatric ALL treated with ALLIC strategy.

Methods: This is a retrospective analysis performed in the population of patients admitted from October 2009 to May 2018. Samples of 432 patients with diagnosis of ALL were collected and analyzed by MLPA P-335 (MRC-Holland) for copy number alterations of IKZF1, EBF1, JAK2, CDKN2A, CDKN2B, PAX5, ETV6, BTG1, RB1 genes and PAR1 region. IKZF1plus cases were defined as those with IKZF1del with at least one additional deletion in: PAX5, CDKN2A, CDKN2B, PAR1 region. Patients were treated with 2 consecutive ALLIC protocols, according to studies stratification. Patient characteristics were compared with chi-square and Wilcoxon-sum-rank-test. Survival probability was analyzed with Kaplan-Meier calculation and survival results compared with Log-rank-test.

Results: IKZF1 was not deleted in 345 cases, IKZF1del was detected in 87 cases and 47 of them were defined as IKZF1plus. Statistically significant higher WBC, MRD+ positivity on day 15, day 33 and week 12, more BCR-ABL+ and high-risk group cases, null response and higher relapse rate were observed in the IKZF1del group (total) when comparing with IKZF1 not del, and also when comparing IKZF1plus vs IKZF1 not del + IKZF1del only. EFS (SE) and DFS (SE) probabilities were: 73 (4)% and 75 (3)% for IKZF1 not del, 66 (9)% and 70 (9)% for IKZF1del, and 20 (10)% and 21 (10)% for IKZF1plus, respectively (p<0.00001).DFS of the standard-risk group was not influenced by the presence of only 1 case of IKZF1del detected in this risk-group of patients. However, DFS of intermediate-risk patients was 41 (11)% for IKZF1plus while 70 (7)% and 73 (4)% were achieved for IKZF1del and IKZF1 not del respectively (p=0,0083). Therefore, high-risk patients with IKZF1plus achieved DFS of 12 (19)% vs 65 (7)% and 50 (21)% for IKZF not del and IKZF1del respectively (p=0.0085). DFS of patients with IKZF1del + CDKN2Adel was 30 (10)% and CDKN2A not deleted 67 (9)% (p=0.0433). DFS of patients with IKZF1del + CDKN2Bdel was 42 (12)% and 66 (9)% for CDKN2B not del. DFS of cases with IKZF1del in addition to deletion of ETV6, BTG1, EBF1 orJAK2 did not show statistically significant differences when comparing with IKZF1del + normal copy number of these genes. In addition, DFS of cases with RB1del was 36 (13)% while cases without RB1del showed 70 (3)% (p=0.0071).

Conclusions: 1- Patients with IKZF1del and IKZF1plus disclosed biological features related to poor outcome. 2- IKZF1plus was associated with a poor outcome in intermediate and high-risk groups according to ALLIC stratification. 3- The addition of CDKN2Adel to IKZF1del influence the outcome. However, CDKN2Bdel did not show the same effect. 4- Copy number alterations of ETV6, BTG1, EBF1 or JAK2 did not demonstrate prognostic impact. 5- RB1 showed negative influence in survival. 6- Identification of patients with IKZF1plus at diagnosis could be very useful for improving risk-group stratification of pediatric ALL patients.