Partner-agnostic targeted RNA sequencing confirms a rare TPM3::PDGFRB fusion in MLN-TK with eosinophilia: a three-phase clinical course under imatinib

Abstract Title
From Cytogenetics to RNA: A Pragmatic Diagnostic Workflow for TPM3::PDGFRB-Rearranged MLN-T
Abstract body:
Objectives
To illustrate a cytogenetically driven diagnostic workflow for PDGFRB-rearranged myeloid/lymphoid
neoplasms with eosinophilia (MLN-TK), and to explore the concordance and routine applicability of
complementary transcriptomic approaches for the characterization of rare PDGFRB fusion partners.
Methods
A 77-year-old male presented with marked hypereosinophilia (~3.0×10⁹/L). Bone marrow morphology
supported chronic eosinophilic leukemia without blast excess. Conventional cytogenetics revealed a t(1;5)
(q21;q33) involving the 5q33 region, prompting PDGFRB break-apart FISH, which confirmed rearrangement
in approximately 70% of nuclei. Fusion partner identification and breakpoint characterization were established
using transcriptomic analysis and targeted DNA sequencing at diagnosis. Whole-transcriptome RNA
sequencing and optical genome mapping were initiated to assess cross-platform concordance and to further
refine structural characterization.
Results
A TPM3::PDGFRB fusion was identified, with breakpoints involving TPM3 and PDGFRB exons consistent
with previously reported rearrangements. The cytogenetic and molecular findings enabled prompt initiation of
targeted therapy. Treatment with low-dose imatinib (100 mg/day) resulted in rapid hematologic improvement
and complete hematologic and cytogenetic remission, with normalization of PDGFRB FISH within five
months. Ongoing analyses aim to evaluate concordance between transcriptomic methodologies and to
assess their respective contributions to fusion detection and routine diagnostic implementation.
Conclusions
This case highlights the central role of cytogenetics as an entry point for PDGFRB-rearranged MLN-TK and
confirms the reliability of transcriptomic and DNA-based approaches for definitive fusion partner identification.
Evaluating concordance between complementary RNA-based methods is critical to define scalable, partneragnostic
strategies suitable for routine diagnostics and long-term disease monitoring.
Abstract Stream/Category
Abstract - Biology
Authors of the abstract:
Lambert, Frederic1; Koopmansch, Benjamin1; Fernandez Carazo, Rafael1; Dardenne, Elisa1; Vanstraelen,
Gaëtan2; Collins, Patrick1; Somja, Joan1; Keutgens, Aurore1; Simon, Germain1; Catarin, Nicolas1; Menten,
Catherine1; Vertenoeil, Gaëlle1
1CHU de LIEGE, LIEGE, Belgium ;
2CHR East Belgium, VERVIERS, Belgium ;