Back

Revisiting AXL as a therapeutic target in soft tissue sarcoma: dubermatinib and bemcentinib effects are largely independent of AXL inhibition

Garnes-Garcia, C.; Fernandez-Aroca, P.; Cimas, F. J.; Ruiz-Hidalgo, M. J.; Belandia, B.; Mayor-Garnacho, N.; Fernandez-Aramburo, A.; Sanchez-Arevalo Lobo, V. J.; Sanchez-Prieto, R.; Fernandez-Aroca, D. M.

2026-10-05 cancer biology
10.64898/2026.10.02.755908 bioRxiv
Show abstract

Soft tissue sarcomas (STS) are a heterogeneous group of mesenchymal malignancies with generally poor prognosis, raising a continued interest in novel therapeutic targets. The TAM-family receptor tyrosine kinase AXL has been proposed as a therapeutic target across several cancer types, including STS. Here we revisit the clinical relevance of AXL in STS by combining in silico analyses with functional and transcriptomic characterisation of two representative AXL inhibitors, dubermatinib and bemcentinib, in the HT-1080 and SK-LMS-1 STS-derived cell models. STS ranked among the tumour types with the highest median AXL mRNA expression across The Cancer Genome Atlas, yet AXL expression was not associated with patient survival. Both dubermatinib and bemcentinib reduced cell viability and induced apoptosis but, unexpectedly, shRNA-mediated AXL knockdown failed to reproduce the antiproliferative effect of either inhibitor and did not modify cellular sensitivity to drugs treatment. Transcriptomic profiling after AXL genetic knockdown showed weak correlation with drug-induced transcriptional changes, with the large majority of drug-regulated genes being AXL-independent and largely specific to each cell line. Our data indicate that although dubermatinib and bemcentinib show clear antitumour activity in STS models, this activity is largely independent from AXL inhibition, suggesting the rationale for AXL-directed therapy may be reconsidered.

Matching journals

The top 7 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.