Abstract
Introduction: The translocation, t(8;16)(p11;p13), results in the fusion between KAT6A and CREBBP and has been associated with a poor prognosis in both pediatric and adult acute myeloid leukemia (AML). This lesion has therefore been re-classified as high risk on the active Phase 3 Children's Oncology Group (COG) trial for de novo AML, AAML1831 (NCT04293562). Less is known about the prognostic significance of CREBBP sequence variants.
Methods: CREBBP variant status was determined in patients with AML enrolled on 4 successive COG trials for de novo pediatric AML (NCT00003790, NCT00070174, NCT01407757, NCT01371981). Fusions involving CREBBP were prospectively obtained via conventional cytogenetics and retrospectively confirmed via RNAseq. Insertions and deletions (indels) leading to frameshift mutations and single nucleotide variants (SNVs) were retrospectively interrogated via next generation sequencing.
Results: Of 2216 patients (age: 0-29.8 years), 55 (2.5%) patients had an alteration involving CREBBP. Sixteen (29%) of these were a fusion involving CREBBP (CREBBP/fus), with KAT6A being the most common translocation partner (n=15) and the remaining translocation involving ANK1. The remaining 39 patients (71%) had a CREBBP mutation (CREBBP/mut), including 19 with an indel (CREBBP/indel) leading to a frameshift mutation and 20 with a SNV (CREBBP/SNV).
We compared clinical and biologic characteristics between the three cohorts. CREBBP/fus patients were significantly younger than CREBBP/indel and CREBBP/SNV patients (median ages of 2.6 vs. 7.8 vs. 11.9 years; p=0.027). There was a higher prevalence of t(8;21)/RUNX1-RUNX1T1 in CREBBP/indel patients compared to CREBBP/SNV patients (42.1% vs. 5%; p=0.008). In contrast, CREBBP/SNV patients were more likely to be associated with a normal karyotype (40% vs. 5.3%; p=0.02). There was a similar prevalence of co-occurring high-risk lesions in CREBBP/indel (n=5; CBFA2T3-GLIS2, KMT2A-AFF1, KMT2A-MLLT4, MLLT10-PICALM, NUP98-HOXA9) and CREBBP/SNV (n=7; DEK-NUP214, ETV6-FOXO1, FUS-ERG, NUP98-NSD1, ETV6-MNX1, FLT3-ITDx2) patients. There was otherwise no difference between presenting WBC count, FLT3-ITD, NPM1, CEBPA, remission rates or MRD status after Induction 1 therapy.
Patients with any CREBBP alteration had a significantly worse 5-year event free survival (EFS) compared to patients without (25.9% vs. 45.2%; p=0.002) and this inferior EFS overlaps with contemporarily defined high-risk patients (Figure 1a). Evaluation of outcomes based on type of alteration demonstrated a similar 5-year EFS of 33.3% and 23.1% between CREBBP/fus and CREBBP/mut patients, respectively (Figure 1b; p=0.832). This poor EFS was maintained in the CREBBP/indel patients with a co-occurring t(8;21) (n=8, 5-year EFS 12.5%). When patients with co-occurring high-risk lesions were excluded from analysis, the remaining CREBBP/mut (n=27) patients maintained their poor EFS (29.6%). Despite their poor EFS, CREBBP/mut patients had an analogous overall survival (OS) to non-CREBBP patients (57.4% vs. 62.3%; p=0.499, Figure 1c), demonstrating that these patients could be successfully salvaged following relapse. In contrast, all patients with CREBBP/fus that relapsed subsequently died from their disease (OS 33.3%).
Conclusions: In a large study of CREBBP alterations in pediatric patients with de novo AML, we show that these patients have a dismal EFS, regardless of alteration type. Further, despite enrichment of t(8;21), the favorable prognosis typically conferred by this alteration was abrogated by the co-occurrence of CREBBP/indel. Similarly, by excluding patients with co-occurring high-risk lesions from analysis, we show that these poor outcomes persist in a cohort of patients that would otherwise be considered low risk. Translocations between CREBBP and KAT6A in patients over 90 days of age are considered high risk on the active COG phase 3 trial. Given the inferior EFS and high salvage rates associated with other CREBBP alterations, intensification of upfront treatment, including hematopoietic stem cell transplant, should be considered in this population.
The authors would like to acknowledge Astellas Pharma Global Development, Inc.
No relevant conflicts of interest to declare.
