AUTHORS
Jack H. Peplinski, MSc, Rhonda E. Ries, MA2, Derek H. Janssens, PhD3, Katherine Tarlock I, MD4, Soheil Meshinchi, MD, PhD5, Danielle C. Kirkey, MD6
AFFILIATIONS
1 Fred Hutchinson Cancer Center, Seattle, Washington, United States
2 Fred Hutchinson Cancer Center, Seattle, Washington, United States
3 Van Andel Institute, Grand Rapids, Michigan, United States
4 Seattle Children's Hospital, Seattle, Washington, United States
5 Fred Hutchinson Cancer Center, Seattle, Washington, United States
6 Fred Hutchinson Cancer Center, Seattle, Washington, United States, Seattle Children's Hospital, Seattle, Washington, United States
ABSTRACT
KMT2A-rearranged (KMT2A-r) acute myeloid leukemia (AML) is characterized by chromosomal translocations involving KMT2A, accounting for 20–25% of newly diagnosed pediatric AML (pAML). Unlike lymphoid malignancies, prognosis in KMT2A-r AML is largely determined by the fusion partner—more than 100 of which have been identified, with five classified as high-risk. KMT2A::MLLT3 is the most common KMT2A fusion in pAML (~40%) and is considered a standard-risk fusion with a 5-year event-free survival (EFS) of ~40% and overall survival (OS) of ~60%. CLEC2A, a recently identified member of the C-type lectin superfamily located at chromosome 12p13.31, has emerged as a therapeutic target in KMT2A-r AML, with CLEC2A-directed CAR-T cells entering clinical trial for relapsed/refractory AML. Although CLEC2A is predominantly expressed in high-risk KMT2A fusions such as KMT2A::MLLT4 and KMT2A::MLLT10, nearly half of KMT2A::MLLT3 cases aberrantly express CLEC2A. To define the clinical and molecular significance of CLEC2A expression in KMT2A::MLLT3 AML, we analyzed genetic, transcriptomic, and clinical data from pAML patients.
We evaluated 102 patients with KMT2A::MLLT3 pAML treated on prior phase III Children's Oncology Group studies with diagnostic RNA sequencing and clinical outcome data available. CLEC2A+ was defined as transcripts per million (TPM)>4 (>75th percentile) and CLEC2A− as TPM<0.2 (<50th percentile), yielding 35 CLEC2A+ and 67 CLEC2A- cases. Between these groups, we performed differential gene expression and then survival analysis based on the resulting supervised k-means clusters.
We first investigated differential gene expression in CLEC2A+ vs. CLEC2A- KMT2A::MLLT3 patients. Nearly all CLEC2A+ patients showed increased expression of neighboring genes on chromosome 12, including KLRF2, CLEC2B, and CLEC9A. We have previously demonstrated direct binding of the KMT2A fusion complex to the CLEC2A promoter and coordinated expression of adjacent genes may reflect fusion-mediated transcriptional regulation at this locus. Among the most highly upregulated genes in CLEC2A+ patients was MECOM (Log2FC=5.34, padj=2.46e-11), a hematopoietic stem cell transcription factor and repressor of myeloid differentiation. Additional stem/progenitor markers, including SALL4 (Log2FC=2.19, padj=3.06e-06), were also enriched. Supervised k-means clustering identified two CLEC2A+ subgroups: Cluster A (CLEC2A+/MECOM+, n=17) and Cluster B (CLEC2A+/MECOM–, n=16), whereas Cluster C comprised CLEC2A-negative patients (n=69). CD34 expression was strongly associated with Cluster A but largely absent from Clusters B and C. xCell deconvolution demonstrated enrichment of a hematopoietic stem cell signature in CLEC2A+ versus CLEC2A− patients (p=0.00047), along with higher stromal (p=0.0014) and endothelial (p=0.0015) scores. Conversely, gene set enrichment analysis showed suppression of T-cell receptor signaling, T-cell differentiation, and T-cell activation pathways in CLEC2A+ cases.
We next evaluated clinical outcomes by cluster. Patients in Cluster A (CLEC2A+/MECOM+) had markedly inferior 5-year EFS of 11.8% compared with Cluster B (56.3%, p=0.0077) and Cluster C (51.7%, p=0.0017). Similarly, 5-year OS was significantly worse in Cluster A at 28.2% than in Cluster B (62.0%, p=0.02) or Cluster C (73.5%, p=0.0023). Multivariate Cox proportional hazards analysis identified Cluster A as an independent predictor of inferior EFS (p=0.025), whereas hematopoietic stem cell transplantation in first complete remission independently predicted improved EFS (p=0.008).
These findings underscore a previously unrecognized high-risk subset of KMT2A::MLLT3 AML defined by CLEC2A expression and a stem-like transcriptional program characterized by MECOM, SALL4, and CD34 expression, with 5-year EFS below 15% and OS below 30%. While we’ve previously reported CD34-positive KMT2A::MLLT3 AML to have poor outcomes, this data supports refinement of risk stratification to identify newly diagnosed patients who may benefit from treatment intensification, including hematopoietic stem cell transplantation. Although transcriptomic profiling is not routinely incorporated into diagnostic risk assessment, flow cytometric detection of CLEC2A and/or CD34 on leukemic blasts provides a clinically feasible approach to identify this high-risk subgroup. In addition, CLEC2A-directed CAR-T therapy represents a promising targeted strategy for patients with relapsed or refractory KMT2A-r AML.
DISCLOSURES
Jack H. Peplinski, MSc: No relevant conflicts of interest to declare.