High Clinical Benefit of Diagnostic RNA Sequencing in Pediatric T-Cell Lymphoblastic Lymphoma for the Identification of Potentially Targetable Risk Groups

T-cell lymphoblastic lymphoma (T-LBL) is a common pediatric malignancy accounting for approximately 20% of the non-Hodgkin lymphomas during childhood. Survival rates of T-LBL are ~80%, but outcome after relapse is dismal, with salvage rates reaching only ~15%. Therefore, there is an urgent need to recognize different risk groups through identification of molecular characteristics and biomarkers. RNA sequencing (RNAseq) has been proven to be valuable in the identification of gene fusions and gene expression profiles in many types of diseases including acute lymphoblastic leukemia and lymphomas. Therefore, RNAseq is arising as an important diagnostic tool, but the added value in T-LBL has not yet been clearly shown. Here, we show that combining fusion data and expression data can lead to identification of potentially targetable risk groups.

We included 41 pediatric T-LBL patients that were diagnosed in the Princess Maxima Center for Pediatric Oncology between 2018-2024 and for whom RNAseq was performed either at diagnosis (n=38) or at relapse (n=3). Informed consent was obtained from all patients included in our cohort. We used RNAseq data at time of first presentation or relapse to perform fusion analysis and gene expression analysis.

It has been shown previously that RNAseq is sufficiently robust for gene fusion detection in routine diagnostics of pediatric cancers with a concordance rate of >98% with traditional diagnostics and an increase in diagnostic yield of 39% (Hehir-Kwa et al., 2022). In total, we detected 19 fusions in 41 patients in our cohort. We have previously discovered a high-risk biological subgroup of children with T-LBL (Kroeze et al., 2024). This subgroup is characterized by NOTCH1 gene fusions and was found in 22% (9/41 patients) of our current cohort (8 primary tumor samples, 1 relapse sample). Fusions partners of NOTCH1 were IKZF2, TRBJ, and miR142. Six patients with a NOTCH1 fusion had a relapse and could not be rescued. Two patients are still under treatment and one patient developed a therapy related AML.

We detected 10 additional fusions at diagnosis that have previously, but only rarely, been described in literature. We detected three HOXA-fusions (HOXA9, HOXA11) and an LMO1 fusion that have proven to play a role the development in T-ALL, although with unknown prognostic relevance. Additionally, we detected two ABL1-fusions that were shown to be high-risk in B-ALL and LBL. Lastly, we found two ZFP36L2 fusions, a MYB fusion, and a MYC fusion with unknown clinical relevance. For three patients only RNAseq at relapse was available and these patients presented with a NOTCH1, an MLLT10 and an PDGFRB fusion, all established high-risk fusions. We consider it likely that these fusions were present at time of initial diagnosis as well.

Together, we show that 46% or our T-LBL cohort expresses a gene fusion, of which at least 58% are classified as high-risk. This stresses the importance of fusion gene detection for diagnostic risk stratification. Although gene fusions can also be detected in a targeted manner by various techniques, RNAseq has the advantage to provide an unbiased detection of both novel and known fusions, or fusions with unknown fusions partners. For example, for NOTCH1, several additional rare fusion partners have been described in literature (e.g. involving IKZF1, EDF1, SEC16A) which might occur as high-risk fusions in T-LBL as well.

Subsequent interpretation of these fusion genes via expression analysis and pathway analysis can result in improved understanding of the pathogenic mechanism of these gene fusions, and the identification of therapeutic targets. For example, patients with a high-risk ABL-class fusion might be good candidates for tyrosine kinase inhibitors. Furthermore, gene expression analysis revealed that the high-risk NOTCH1-fusion positive all presented with highly elevated CD1a expression, representing the top 5% compared to the rest of the T-LBL cohort. CD1a was recently discovered as potential target for CAR T-cell therapy which makes it a promising potential target in these high-risk patients.

In conclusion, our data show that gene fusion detection through RNA sequencing has a very high clinical potential in order to stratify T-LBL, and perform downstream analyses to optimize treatment decisions.