Minimal Residual Disease and IKZF1 As Predictors of Relapse, and Increased Treatment Related Mortality in Down Syndrome Acute Lymphoblastic Leukemia: A Unique and Large International Matched Case-Control Study

INTRODUCTION Down syndrome patients with acute lymphoblastic leukemia (DS ALL) are less likely to have a favorable (cyto)genetic subtype and are at higher risk of relapse and treatment-related mortality (TRM) than non-DS ALL (Buitenkamp et al., Blood 2014). At present, the independent predictive value of minimal residual disease (MRD) is unclear and may be biased by an unequal distribution of (cyto)genetic risk groups among DS ALL and non-DS ALL patients. This study was aimed to decipher the prognostic implications of MRD and IKZF1 deletions and the frequency of TRM in a matched cohort of DS ALL cases and non-DS ALL controls.

METHODS Each DS ALL patient was matched to 3 non-DS ALL patients based on treatment protocol, induction treatment, cytogenetic subtype, IKZF1 status, age (cutoff at 10 years), and white blood cell count (cutoff at 50,000 cells/µl). For MRD analysis, matching was only on induction treatment and excluded the MRD-guided treatment arm; for survival analyses, matching included the MRD-guided treatment arm, thus resulting in two separately matched cohorts. Patients who died during induction were excluded from matching. Absolute MRD levels were measured with RQ-PCR, log-transformed and analyzed with a multilevel mixed-effects linear regression model. Matched Cox proportional hazard regression models were used to analyze event-free survival (EFS), overall-survival (OS), relapse-free survival (RFS) and mortality in remission as surrogate for TRM.

RESULTS Patients treated between 2002 and 2018 on Dutch DCOG-ALL10/11 trials, Australian ANZCHOG-ALL8 and AIEOP-BFM-ALL2009 trials, and UKALL2003 trial were included, resulting in 160 DS ALL and 5313 non-DS ALL patients. Out of these 160 DS ALL patients, 13 died during induction versus 42/5313 non-DS ALL patients (8.1% versus 0.8%, p<0.0001). Exclusion of induction deaths and patients with missing IKZF1 or MRD data, resulted in 110 DS ALL matched to study MRD differences (1 DS ALL could not be matched and was excluded), and 93 DS ALL matched to analyze survival (ALL11 was excluded due to short follow-up; 3 DS ALL could not be matched and were excluded). In the matched cohort, 22% (24/110) had favorable cytogenetics, 19% (21/110) had an IKZF1 deletion, 26% (29/110) were ≥10 years, and 17% (19/110) had a WBC ≥50,000 cells/µl. Only 3 cases (2 DS, 1 non-DS) did not achieve complete remission (excluding induction deaths). Median follow-up time of survivors was 7.1 years.

The MRD levels did not differ between DS ALL and matched non-DS ALL patients at end of induction (p=0.96) nor when analyzed over time (p=0.43). In accordance, the 5-yr RFS did not differ between DS ALL and matched non-DS ALL patients (85 ± 4% versus 89 ± 2%, p=0.09). The 5-yr TRM of patients in remission was higher in DS ALL compared to non-DS ALL (11 ± 3% versus 3 ± 1%, p=0.001). In line, 5-yr EFS was lower in DS ALL compared to non-DS ALL patients (74 ± 5% versus 86 ± 2%, p=0.0007), as was 5-yr OS (77 ± 5% versus 93 ± 2%, p=0.0001). Multivariable analysis revealed that IKZF1 deletion can discriminate DS ALL patients at high risk of relapse (RFS: HR>3, 95% CI=1-12, p=0.05). The effect of IKZF1 deletion was stronger in DS ALL patients than in the non-DS ALL patients in our cohort.

CONCLUSION The MRD levels did not differ between DS ALL and non-DS ALL patients when matched for (cyto)genetics and other risk factors. In accordance, the overall relapse rate of DS ALL patients did not differ from that of matched non-DS ALL patients. Similar to non-DS ALL, IKZF1 deletion is an adverse risk factor for DS-ALL, indicating the need for treatment aimed at reducing the high relapse risk. DS ALL patients suffer more frequently from death in induction and from treatment while in remission, which jeopardizes treatment intensification. Therefore, the efficacy of targeted, less toxic therapies such as immunotherapies should be assessed in DS ALL.