Prognostic Importance of Cytogenetic Subgroups in Relapse Risk Stratification in Children and Young Adults with First Relapsed B-ALL: A Children's Oncology Group Study AALL1331

Background Cytogenetic (CG) abnormalities play a pivotal role in risk stratification to guide post induction therapy in B-lymphoblastic leukemia (B-ALL) at initial diagnosis, but their prognostic relevance on post-relapse outcomes/survival is not clearly known. To understand the impact of relapse CG, data was collected from patients (pts) enrolled on Children's Oncology Group (COG) AALL1331 trial for first relapse B-ALL patients 1-30 years (yr) old; pts with Ph+ ALL were excluded. After a common chemotherapy reinduction (Block 1), pts were risk assigned based on site of relapse, Complete Remission 1 duration, and minimal residual disease (MRD) response at end Block 1 (EB1-MRD). Risk groups were defined as: High risk (HR): Early bone marrow (BM) relapse <36 months from diagnosis or isolated extramedullary (IEM) relapse <18 months; Intermediate risk (IR): Late BM (>36 months) or late IEM (>18 months) relapse with EB1-MRD ≥0.1%; Low risk (LR): Late BM or late IEM relapse with EB1- MRD <0.1%; or Treatment failure (TF): M3 BM or failure to clear central nervous system at EB-1. The primary objective of AALL1331 was to compare survival of HR/IR pts receiving post-induction therapy with 2 blocks of chemotherapy (arm A) vs blinatumomab (arm B), followed by allogeneic transplant and to compare survival in LR pts treated with chemotherapy alone (arm C) or chemotherapy plus blinatumomab (arm D).

Methods CG data (karyotype and FISH) at relapse were centrally reviewed. CG groups were defined as: favorable (fav) (double trisomies (DT) or ETV6-RUNX1), unfavorable (unfav) (hypodiploidy, KMT2A-rearranged (KMT2Ar) or iAMP21), and neutral (all others). Of 661 eligible pts, 129 IEM relapses were excluded. CG data were available for 455/532 (85%) BM relapses, which were used to evaluate the association between CG and clinical features and Block 1 response. At EB-1, 119 pts with CG results were HR, 76 were IR and 174 were LR. These subgroups were then used for associations of CG with response to post-reinduction therapy. CG distribution at relapse was compared to those from over 25,000 pts enrolled on COG newly diagnosed B-ALL trials (1996-2014). Statistical comparisons used Pearson's chi-squared, Fisher's exact, Wilcoxon rank sum, or logrank test with two-sided p values.

Results There were a significantly higher proportion of unfav CG pts compared to CG distribution at initial diagnosis (16% vs 6.6%, p<0.001), as opposed to fav CG at first relapse (25% vs 44% p<0.001). Similar to trends in newly diagnosed pts, fav CG and KMT2Ar was observed in younger pts (median 8 yr and 2 yr, respectively) compared to the median age of 10.5 yr among all 532 first relapse pts. WBC at relapse was higher in KMT2Ar (median 9.7k/uL), but similar in all others (median 5k/uL, p=0.057). Time to relapse differed by CG, with late BM relapses comprising 79% of fav vs. 45% of unfav (p<0.001) (Table 1).

EB-1 MRD-negative (<0.01%) rates were highest for fav (53%), intermediate for neutral (37%) and lowest for unfav CG (25%, p<0.001). CG significantly influenced EB-1 risk assignments, with fav/neutral/unfav CG ranges of 49%/30%/21% for LR, 12%/24%/25% for HR, and 2%/4%/7% for TF (p<0.01) (Table 2).

For HR/IR pts, while there was a trend towards better 4-yr disease-free survival (DFS) for fav CG on arm B as compared to neutral and unfav CG pts, there was a significant improvement in overall survival (OS) for arm B fav CG pts (p=0.041). However, for arm A, DFS/OS was similar and consistently poor in all three CG groups. Likewise, for LR pts, 4 yr DFS/OS was better for fav CG, as compared to unfav CG, though not statistically significant (p= 0.72 for DFS, p= 0.26 for OS) in arm C, but further improvement in DFS was seen for fav CG pts in arm D as compared to arm C (85% vs. 61%, p=0.047). Interestingly, unfav CG pts in LR group did better with blinatumomab (arm D) compared to unfav CG pts in arm C, although numbers were small in each arm (Table 2).

Conclusions CG at relapse remains prognostic similar to those in newly diagnosed pts. Fav CG pts have higher rates of LR risk assignment as they relapse later and are more likely to achieve MRD negativity after reinduction chemotherapy. Blinatumomab helped improve the outcomes of fav CG pts with first BM relapse in both HR/IR and LR groups, and for unfav CG pts in LR group. Thus, the impact of CG on outcomes after first relapse may differ based on whether chemotherapy or immunotherapy is used as post-reinduction consolidation.

Gore:Janssen: Membership on an entity's Board of Directors or advisory committees; Dura: Membership on an entity's Board of Directors or advisory committees; Onkure: Membership on an entity's Board of Directors or advisory committees; Syndax: Membership on an entity's Board of Directors or advisory committees; Takeda Development Center Americas, Inc.: Research Funding; Millennium: Research Funding; Sanofi-Paris: Current equity holder in publicly-traded company; Mirati: Current equity holder in publicly-traded company; Amgen: Current equity holder in publicly-traded company, Membership on an entity's Board of Directors or advisory committees. Hunger:Servier: Honoraria; Jazz: Honoraria; Amgen: Current equity holder in private company, Honoraria. Brown:Bristol Myers Squibb: Current Employment.