AXL-erating mantle cell lymphoma treatment

In this issue of Blood, Gelebart et al identify a new third isoform splice variant of the tyrosine kinase AXL (AXL3) that is overexpressed and important for tumor cell survival in mantle cell lymphoma (MCL).¹ 

Although biological mechanisms underlying the clinical heterogeneity of MCL remain rather elusive, the B-cell phenotype and B-cell receptor signaling pathways are increasingly exploited by targeted treatment approaches. Bruton tyrosine kinase inhibitors, such as ibrutinib or pirtobrutinib, the BCL2 inhibitor venetoclax, CD19-directed chimeric antigen receptor (CAR) T cells, and CD20-targeting bispecific antibodies or antibody drug conjugates have shown promising results in high-risk patients.^2-5 However, these approaches are also used in other B-cell malignant cancers. MCL-specific pathways or vulnerabilities have so far, however, been rarely identified as potential therapeutic targets. The identification and characterization of AXL3 expression in MCL discussed herein opens up a new avenue to further understand MCL biology while exploring AXL3 as a novel therapeutic target.⁶ 

AXL is a member of the Tyro3-Axl-Mer (TAM) receptor tyrosine kinase (RTK) family that is activated by its ligand growth arrest–specific protein 6 (GAS6) or may be activated ligand independently (eg, heterodimerization with other non-TAM RTKs).⁷ It is expressed in many healthy tissues but not in healthy B cells or lymphoid tissues. So far, two isoforms (AXL1 and AXL2) have been described, and a third isoform was predicted. AXL is a key component in removal of apoptotic material and limits inflammatory response. In malignant cells, AXL is considered to promote cell survival, proliferation, migration, angiogenesis, and invasion because of activation of multiple downstream pathways, such as phosphatidylinositol 3-kinase–protein kinase B (AKT)–mammalian target of rapamycin, MEK–extracellular signal-regulated kinase, NF-κB, and JAK/STAT. Indeed, overexpression of AXL is associated with therapeutic resistance in various solid cancers, such as glioma, melanoma, or non–small-cell lung cancer.⁶ 

The authors first confirmed relevant expression of AXL on the mRNA and protein level in MCL cell lines and primary MCL patient samples. They observed a high expression of a protein of the size of a predicted yet previously undescribed third AXL isoform (AXL3). Presence of AXL3 was confirmed by cloning and alignment of the sequencing result with the predicted exon structure. Interestingly, AXL3 appears to be constitutively activated independent of GAS6, and only knockdown of the AXL3 (and not AXL1 or AXL2) isoform by siRNA and CRISPRi induced apoptosis in MCL cells.

To further explore this potentially targetable dependency, the authors exposed MCL cell lines and primary patient samples to the AXL inhibitor (AXLi) bemcentinib. These in vitro experiments employed multiple readouts and showed significant cell growth inhibition, cell cycle arrest, and apoptosis of MCL and a decrease of AKT and NF-κB activation across experiments. Next, the authors established an MCL xenograft in vivo model, using the JeKo1 cell line with a luciferase reporter. Confirming the in vitro observations, bemcentinib slowed tumor growth and prolonged overall survival (OS) compared with untreated control animals. Interestingly, tumor control and OS were also significantly improved compared with ibrutinib, and combination treatment yielded the longest OS. Together, these observations suggest a relevant vulnerability to AXLi because of a dependency on AXL3 in MCL.

Although the present study suggests a mechanistic role for the AKT and NF-κB pathways, the mechanisms underlying AXLi sensitivity and the role of AXL and, more specifically, AXL3 in lymphomagenesis remain unclear. In addition, AXL expression in nodal MCL patient samples is unknown, and a larger cohort of MCL patient samples needs to be tested to confirm AXL3 expression in the samples. More important, AXL expression is associated with resistance to antiprogrammed death protein 1 (PD1) blockade in melanoma⁶ and likely shapes an immunosuppressive tumor immune microenvironment (TIME) (eg, by distinct cytokine secretion or upregulation of PD1 ligand 1 (PD-L1) expression).^8,9 The association between AXL expression and TIME composition remains to be determined and may further inform AXL-targeting combination strategies, similar to what is already investigated in other cancers.^6,8,9 

Various AXL inhibitors, such as bemcentinib, DS-1205, and dubermatinib, have been already tested in phase 1/2 clinical trials in various solid tumors, acute myeloid leukemia, or myelodysplastic syndrome. In addition, monoclonal antibodies, antibody drug conjugates, and CAR T cells targeting AXL are under investigation in clinical trials.⁶ Although this shows a clear path for clinical translation of the results discussed herein, all of these approaches have in common that they do not target the AXL3 isoform that appears to mediate efficacy of AXLi in MCL.

Direct targeting of the AXL3 isoform by targeted protein degradation using the concept of a proteolysis-targeting chimera molecule appears promising,¹⁰ especially as it would circumvent potential AXLi-associated toxicities. Because AXL overexpression is found in other solid and hematologic malignancies and is frequently associated with worse prognosis because of therapeutic resistance, exploring the presence of the AXL3 isoform and efficacy of novel AXL-directed therapeutic approaches, including immunotherapy combinations, appears warranted in other tumors beyond MCL.

Conflict-of-interest disclosure: P.J.B. reports research funding from BeiGene, BMS, and MSD; advisory role for Takeda and Stemline; honoraria from BeiGene, BMS, MSD, and Takeda; and travel support from BeiGene, BMS, Celgene, MSD, and Takeda.