CD44v3 Emerges as a Resistance Switch in TrkA-Driven Triple-Negative Breast Cancer
TrkA overexpression in triple-negative breast cancer has long carried a mixed therapeutic message: the target is present, yet standard Trk inhibition often fails to deliver the same depth of response seen in tumors with NTRK gene fusions. In this study, that gap was traced to a surface partner with a far less obvious clinical profile, CD44v3. The work links the receptor tyrosine kinase to a membrane glycoprotein axis that appears to preserve TrkA signaling even when an inhibitor is present.
In practical terms, this is a resistance biology problem rather than a simple drug potency problem. Entrectinib remained active in parental MDA-MB-231 cells, but TrkA-overexpressing counterparts showed marked resistance in clonogenic assays. In those engineered cells, phosphorylation of the TrkA kinase domain fell only at Entrectinib concentrations more than 10-fold higher than those effective in NTRK fusion-positive models. The pattern points to a signaling environment that buffers the receptor against pharmacologic suppression.
A Membrane Partner That Lowers the Threshold for TrkA Signaling
The central mechanistic insight comes from the CD44v3-derived C-terminal peptide, P4. In MDA-MB-231 and BT-549 cells overexpressing TrkA, P4 increased TrkA phosphorylation under subnanomolar NGF conditions. The same effect appeared in COS-7 cells engineered to overexpress TrkA but lacking endogenous CD44, strengthening the argument that the peptide itself can facilitate receptor activation in a permissive context.
The biology here is tightly coherent. NGF is the ligand that drives TrkA activation, but the data suggest CD44v3 helps set the activation threshold, especially when ligand concentrations are low. Once that threshold drops, downstream signaling becomes easier to sustain, and the tumor cell keeps transducing growth and motility signals even while exposed to a kinase inhibitor. Entrectinib still suppressed NGF-induced TrkA and Akt phosphorylation, including in the presence of P4, confirming that the pathway remains druggable. The problem is not loss of target dependence. It is the existence of a structural scaffold that helps the receptor keep firing.
RhoA biology adds another layer. NGF increased RhoA activity, while Entrectinib or P4 alone brought it back toward control levels. The combined P4–Entrectinib condition reduced basal RhoA activation even further. In a metastatic breast cancer context, this matters because RhoA signaling is tied to motility, invasion, and cytoskeletal remodeling. A receptor that sustains RhoA tone can help cells move, colonize, and survive in a metastatic niche.
Brain Micrometastases Fell Sharply With the Combination Strategy
The in vivo experiment moved the question from signaling to disease behavior. MDA-MB-231 HA-TrkA cells were transplanted into NOD SCID mice, where they generated lung and brain metastases. P4 alone or Entrectinib alone produced limited effects on metastatic burden. The combination was different.
Brain micrometastatic penetrance fell by almost 80% with Entrectinib plus P4. The signal was strong enough to suggest that disrupting the TrkA-CD44v3 interaction can restore meaningful drug sensitivity in a brain-metastatic setting where the receptor is overexpressed rather than fusion-driven. This is an important distinction for treatment design, because it suggests that receptor abundance alone may create a distinct pharmacologic state with its own resistance architecture.
The lung data need to be handled carefully. Lung lesions were detected in all groups, and the readout was a binary penetrance score that was saturated across arms. No organ-specific quantitative conclusion can be drawn from that assay. The brain signal, however, was clear within the limits of the model used.
Why the Brain May Be a Distinct Vulnerability
Brain metastasis is not simply breast cancer at a new site. The local microenvironment introduces a different growth factor context, extracellular matrix composition, and signaling pressure. Here, the authors also point to single-cell transcriptomic evidence from patient-derived brain metastases showing higher CD44 and RhoA expression in CD45-negative cells than in healthy brain tissue. That finding supports the idea that the axis uncovered in the model is not just an artifact of one cell line.
The clinical database analysis adds a second anchor. NTRK1 expression was enriched in patients with micrometastatic lesions compared with patients without micrometastasis. Even without fusion-positive disease, elevated TrkA expression may identify tumors that are biologically primed for dissemination and harder to suppress with a kinase inhibitor alone. The result does not prove clinical utility, but it does sharpen the rationale for stratifying by receptor state rather than fusion status only.
A New Combinatorial Logic for TrkA-Overexpressing Disease
The therapeutic implication is straightforward: a Trk inhibitor may need a companion strategy that disrupts receptor organization at the cell surface. In this setting, CD44v3 is more than a marker. It appears to be part of the machinery that permits TrkA to remain active under low-NGF conditions and to preserve downstream survival and motility signals. By weakening that scaffold with a CD44v3-derived peptide, Entrectinib regained activity in a preclinical brain metastasis model.
The authors also point to pathways beyond Akt, including MAPK/ERK and PLCγ, as plausible contributors to the phenotype. Those routes were not fully mapped here, and the study does not claim that Akt is the sole effector. Even so, the data are sufficient to support a mechanistic model in which CD44v3 tunes receptor availability, TrkA phosphorylation, and RhoA-linked invasive behavior.
One sentence from the experimental readout captures the whole problem: TrkA overexpression made Entrectinib much less effective, and CD44v3 blockade helped reverse that resistance.
What Still Needs to Be Proven Before Translation
This is preclinical work, and the limits are important. The model does not establish tolerability, dosing feasibility, or how a CD44v3-directed strategy would behave in humans. The in vivo lung assessment was not quantitative enough to compare organ-specific metastatic burden, and the study did not define whether primary tumor growth changes tracked the metastatic effect. Longitudinal imaging, quantitative lung nodule assessment, and additional confirmation in other model systems will be needed before the biology can be translated into a patient-facing strategy.
Even with those caveats, the study advances a useful concept: Entrectinib resistance in TrkA-overexpressing TNBC may depend on a membrane co-receptor network rather than on TrkA alone. If that holds in broader models, the CD44v3-TrkA axis could become a rational target for patients whose tumors express TrkA without harboring an NTRK fusion, especially in the setting of brain metastatic spread.