Kinase Inhibitors Beat Aggressive Skin T-Cell Lymphomas—Until Tumors Learn to Fight Back

For patients with the rarest and most lethal forms of cutaneous T-cell lymphoma, the arrival of tyrosine kinase inhibitors felt like the first genuine breakthrough in decades. These skin-homing T-cell malignancies, which include primary cutaneous aggressive epidermotropic cytotoxic T-cell lymphoma and primary cutaneous gamma delta T-cell lymphoma, have long defied conventional chemotherapy, with median survival often measured in months rather than years. Now a comprehensive review published in Frontiers in Hematology synthesizes the genomic, clinical, and structural evidence behind a sobering pattern: while kinase inhibitors can produce dramatic initial responses in these tumors, resistance emerges with remarkable predictability, driven by secondary mutations that tumors acquire under the selective pressure of treatment itself.

The review, led by Julia Yescas and colleagues at Northwestern University Feinberg School of Medicine, together with collaborators at Yale, Case Western Reserve, and MD Anderson Cancer Center, examines how aggressive cutaneous T-cell lymphomas become addicted to dysregulated kinase signaling and how that addiction becomes both a therapeutic vulnerability and an escape route. The authors argue that understanding the molecular architecture of resistance is essential for designing next-generation inhibitors and rational combination strategies that could finally deliver durable remissions in these devastating diseases.

Primary cutaneous aggressive epidermotropic cytotoxic T-cell lymphoma, first recognized as a distinct entity in 1999, is characterized by epidermotropic CD8-positive cytotoxic T cells expressing granzyme B, perforin, and TIA-1. Despite its rarity, the disease carries a median overall survival of just 32 months and a five-year survival of only 31 percent. Recent high-throughput genomic studies have revealed that this lymphoma harbors recurrent and targetable kinase fusions at striking frequency. In one study, fusion genes involving JAK2 were detected in three of twelve patients, while another reported kinase gene fusions in six of six samples, with five harboring a JAK2 fusion and one carrying a novel ABL1 fusion. A separate report described a fusion joining SATB1 to FGFR1, a receptor tyrosine kinase for fibroblast growth factors. In all cases, the tyrosine kinase domains are conserved at the C-terminus and fused to a diverse set of partner genes including STAT3, KHDRBS1, PICALM, CAPRIN1, PCM1, and TFG. The biological basis for the recurrent involvement of JAK2, rather than the JAK1 and JAK3 family members typically utilized by CD8-positive T cells, remains unknown, though the authors speculate it may reflect lineage infidelity, with transformed cells deviating from canonical cytokine signaling to exploit JAK2-mediated proliferative pathways.

Primary cutaneous gamma delta T-cell lymphoma presents a different genomic landscape. This rare and highly aggressive subset, with a median overall survival of approximately 31 months and a five-year disease-specific survival of only 19.9 percent, is characterized by marked genomic instability, with a median of 166.5 somatic copy number variants per sample. Rather than kinase fusions, PCGDTL harbors recurrent point mutations in key oncogenic signaling pathways, including the MAPK pathway (KRAS, NRAS, MAPK1), the MYC pathway (MYC, MYCN, FBXW7), and the JAK/STAT pathway (STAT3, STAT5B, JAK3, SOCS1). Among these, KRAS mutations were most common, while approximately 21 percent of samples carried JAK/STAT point mutations. Notably, clinical and histopathologic features correlate with TCR delta-chain usage—epidermal and dermal cases typically express TRVδ1, while subcutaneous cases express TRδV2—yet next-generation sequencing has not identified major mutational differences between these subsets.

The clinical experience with targeted inhibitors in both subtypes tells a consistent story of initial sensitivity followed by adaptive escape. In the first reported clinical use of a tyrosine kinase inhibitor in PCAETL, a patient harboring the SATB1-FGFR1 fusion was treated with the FGFR inhibitor pemigatinib. The response was remarkable: approximately 90 percent regression of ulcerated lesions within two weeks, with partial remission maintained for around four months. However, adverse events including hyperphosphatemia, nail changes, and ocular irritation necessitated a dose reduction. The patient’s lesions relapsed within one month, and sequencing revealed secondary point mutations within the FGFR1 kinase domain. The patient expired two months later. In PCGDTL, a patient with SOCS1 loss treated with the JAK1/2 inhibitor ruxolitinib achieved near-complete regression after four weeks, but disease progression occurred after eight weeks, accompanied by newly acquired lesions. Sequencing of the relapse sample revealed a STAT5B p.N642H mutation, a well-characterized gain-of-function driver associated with T-cell neoplasms and drug-resistant leukemias. Another PCGDTL patient harboring a JAK3 p.A573V mutation was enrolled in a phase II trial of the dual SYK/JAK inhibitor cerdulatinib, achieving an 86 percent reduction in skin disease burden after seven weeks. But therapy was discontinued due to grade 3 adverse events, and shortly after, relapse occurred with sequencing revealing acquired JAK3 p.M511I, JAK1 p.L783F, and JAK1 p.T901A mutations in circulating tumor cells.

The structural basis of these resistance mutations reveals elegant molecular logic. In the FGFR1 case, two secondary mutations were detected following pemigatinib treatment: E462V in the juxtamembrane region and V561L in the kinase domain. The E462V substitution, despite its location in a critical signaling hub, did not confer resistance—structural modeling showed no clashes or modulation of the drug-binding state. In contrast, V561L represents a canonical gatekeeper mutation that sterically impedes inhibitor binding and enhances ATP affinity, thereby reducing drug potency. Structural modeling of the analogous V561M substitution revealed that replacement with a bulky methionine side chain induces steric clashes within the inhibitor-binding pocket, accounting for observed resistance. This mutation has been previously described in FGFR1-driven myeloproliferative neoplasms, where it confers broad resistance to multiple FGFR inhibitors.

In the JAK pathway, the resistance architecture is equally instructive. The STAT5B p.N642H mutation occurs in the Src homology 2 domain and stabilizes STAT5B dimers, increases phosphorylation, and renders the protein constitutively active. Structural studies indicate that this substitution alters the hydrogen-bonding network within the phosphotyrosine pocket, favoring inter-protein interactions in the STAT5B dimer. The JAK3 mutations follow a different logic: M511 lies within the SH2-pseudokinase linker, which typically interacts with the kinase domain active site, and the M511I substitution destabilizes this linker, promoting kinase domain activation. A573 is located within the alphaC helix of the pseudokinase domain, a structural element that toggles between active and inactive conformations. Through structural modeling with cerdulatinib, the combination of M511I and A573V causes steric hindrance, suggesting these variants confer resistance through inter- or intramolecular interaction alterations. Remarkably, the JAK1 p.L783F mutation, which also causes steric hindrance with cerdulatinib, occurs in a different JAK family member than JAK3, raising important questions about how cross-kinase cooperation contributes to drug resistance. The combined presence of JAK3 p.A573V and JAK1 p.L783F conferred greater resistance to cerdulatinib compared to single mutants, indicating cooperative resistance across paralogous kinases.

Toxicity emerges as an indirect but powerful facilitator of resistance. Pemigatinib’s most common adverse events include hyperphosphatemia, nail toxicity, alopecia, diarrhea, anemia, dry eyes, and serous retinal detachment—manifestations that reflect the broad physiological role of FGFR signaling in the corneal epithelium, photoreceptors, iris, and hair follicles. Ruxolitinib, evaluated in a phase II trial of 53 patients with CTCL, produced anemia in 15 patients, diarrhea in 7, thrombocytopenia in 9, and neutropenia in 10, reflecting the critical role of JAK2 signaling in erythropoietin and thrombopoietin receptor function. Cerdulatinib, targeting both SYK and JAK families, produced frequent grade 3 or higher adverse events in a phase II trial of relapsed or refractory peripheral T-cell lymphoma and CTCL, including increased lipase in 29.2 percent, increased amylase in 43.1 percent, diarrhea in 43.1 percent, and anemia in 32.3 percent. Serious adverse events occurred in 64.6 percent of patients, and tragically, two deaths not related to disease progression were attributed to the drug. When dose reductions or discontinuations become necessary, resistant subclones may gain a selective advantage under reduced drug exposure.

The review argues that durable responses will require moving beyond single-agent therapy toward rational combination regimens. Preclinical and clinical evidence across malignancies supports combining kinase inhibitors with immune checkpoint blockade. In non-small cell lung cancer, the JAK1 inhibitor itacitinib combined with pembrolizumab achieved a 12-week overall response rate of 62 percent with a median progression-free survival of 23.8 months. In Hodgkin lymphoma, ruxolitinib combined with nivolumab yielded an overall response rate of 53 percent, with ruxolitinib appearing to rescue exhausted T cells and synergize with checkpoint blockade in reshaping the myeloid compartment. Although no clinical trials have yet evaluated such combinations in aggressive T-cell lymphomas, the authors suggest these approaches warrant investigation given the immune-evasive nature of these tumors. Topical application of inhibitors might also circumvent systemic toxicity and provide a window for prolonged therapeutic efficacy.

Aggressive cutaneous T-cell lymphoma thus presents both a challenge and an opportunity for precision dermato-oncology. The discovery of recurrent kinase alterations reveals that subsets of these lymphomas are mechanistically dependent on constitutive kinase activity for survival and proliferation, making them vulnerable to targeted inhibition. Yet the predictable emergence of adaptive resistance—through gatekeeper mutations, cooperative cross-kinase substitutions, and toxicity-driven dose reductions—underscores that single-agent approaches are insufficient. Future progress, the authors conclude, relies on molecular stratification based on kinase dependency, longitudinal genomic monitoring to detect resistance as it emerges, rational combination therapy, and the development of mutation-selective next-generation inhibitors capable of overcoming the adaptive mechanisms that currently limit treatment durability.

Subject of Research: Kinase inhibitor resistance in cutaneous T-cell lymphomas

Article Title: Mechanisms of kinase inhibitor resistance in subtypes of cutaneous T-cell lymphomas

Article References: Yescas, J. A., Bunick, C. G., Grada, A., Guitart, J., Torres-Cabala, C. A., Le Poole, I. C., & Zhou, X. A. (2026). Mechanisms of kinase inhibitor resistance in subtypes of cutaneous T-cell lymphomas. Frontiers in Hematology, 5, Article 1812908. https://doi.org/10.3389/frhem.2026.1812908

Image Credits: AI Generated

DOI: 10.3389/frhem.2026.1812908

Keywords: cutaneous T-cell lymphoma, kinase inhibitors, drug resistance, JAK/STAT pathway, FGFR1, precision oncology, hematology, Mechanisms, kinase, inhibitor, resistance, subtypes