Metastatic colorectal cancer (mCRC) remains a significant clinical challenge due to the emergence of drug resistance, highlighting the need for new therapeutic strategies. Aberrant regulation of the proteome drives cancer progression, and dissecting these alterations offers opportunities for targeted intervention. We aimed to identify candidate kinase dependencies in mCRC and to test a rational combination strategy in patient-derived models. We performed proteomic and phosphoproteomic profiling of 10 mCRC tissues, their matched patient-derived organoids (PDOs) and normal colon mucosa, integrated with whole-exome sequencing. Kinase activities were inferred by kinase-substrate enrichment analysis. Silmitasertib-mediated casein kinase II (CKII) inhibition, alone and combined with the MEK inhibitor trametinib, was evaluated by ex vivo drug screening, western blot, flow cytometry and drug-interaction analysis across the PDO cohort and in normal colon PDOs. PDOs recapitulated the molecular features of matched tumours. Both tissues and PDOs showed consistently elevated CKII activity. Silmitasertib exerted heterogeneous antitumour activity, and resistance was associated with MAPK and Akt-mTOR signalling activity. Combining silmitasertib with trametinib produced an additive reduction of both resistance-associated pathways with little effect on the viability of normal colon PDOs. The combination also significantly reduced CKII activity, the abundance of active β-catenin and the colorectal cancer stem cell pool. Phosphoproteomic profiling identified CKII as a therapeutic vulnerability in mCRC and provided a rationale for combining silmitasertib with MEK inhibition, offering a candidate combination strategy for mCRC.
LUAD exhibits diversity in the mechanisms of angiogenesis and the tumor environment’s responses to anti-VEGF therapy, while the tumor-cell-intrinsic mechanisms that prevent vascularization are not well understood. SCGB3A2 (secretoglobin 3A2; HIN-2/UGRP1) is a club/secretory marker of pulmonary adenocarcinoma without established angiogenic function. By using an unsupervised integration approach of two public LUAD single-cell RNA-seq cohorts, we described a SCGB3A2-high malignant epithelial state characterized by an angiogenesis-low phenotype as indicated by the low presence of VEGFA, hypoxic conditions, and a low epithelial-mesenchymal transition (EMT), along with a secretory/anti-inflammatory program, which was confirmed in independent bulk cohorts (TCGA-LUAD, GSE72094) showing that SCGB3A2-high tumors experienced lower instances of angiogenesis, hypoxia, and scores related to EMT/invasion. Using spatial transcriptomics, we positioned SCGB3A2-high tumor cells next to endothelial cells and identified a spatially inferred VEGFA/B–FLT1 interaction pattern. In A549 and NCI-H1975 cells, conditioned medium from SCGB3A2-overexpressing cells contained higher levels of extracellular PGK2-immunoreactive signal and angiostatin and was associated with reduced HUVEC tube formation and VEGFR2 phosphorylation. Thus, SCGB3A2 is a marker for LUAD tumors characterized by low levels of angiogenesis, which is associated with an extracellular PGK2/angiostatin signal.
For more than a century after its initial description, glioblastoma (GBM) has remained one of the most therapeutically refractory human malignancies, resembling an almost impregnable fortress against modern oncology. Despite extensive molecular characterization and identification of numerous actionable targets, survival outcomes have improved only marginally. GBM is characterized by profound spatial and cellular heterogeneity, adaptive signaling rewiring, extensive phenotypic plasticity, an immunosuppressive tumor microenvironment, and major pharmacological constraints imposed by the blood–brain barrier. This review examines the resulting disconnect between biological rationale and clinical outcome through three interrelated translational principles that determine whether mechanistically compelling therapies can achieve meaningful clinical benefit in GBM. First, who is treated and what is biologically active: therapeutic efficacy depends on matching patients to functional pathway dependencies beyond static molecular alterations. Second, whether the drug reaches and modulates its intended target: candidate therapies must achieve sufficient intratumoral exposure and demonstrate target modulation. Third, how the tumor adapts under therapeutic pressure: durable efficacy requires accounting for receptor tyrosine kinase switching, phenotypic plasticity, immune adaptation, and spatial escape. By linking these principles to the molecular, spatial, and pharmacological organization of GBM, this review frames the disease as a dynamic tumor ecosystem and provides a framework for interpreting recurrent clinical failure and guiding more biologically informed therapeutic development.
Human epidermal growth factor receptor 2 (HER2) is a well-established oncogenic driver in multiple solid tumors, most prominently in breast and gastric cancers. Beyond gene amplification–driven signaling through canonical PI3K/AKT and RAS/MAPK pathways, emerging evidence reveals that HER2 functions within a highly dynamic and context-dependent tumor ecosystem. Interactions between HER2-positive tumor cells and immune and stromal compartments actively shape the tumor microenvironment, contributing to immune evasion, therapeutic resistance and disease progression. A central challenge in the clinical management of HER2-altered cancers is pronounced spatial and temporal heterogeneity, which undermines both diagnostic accuracy and the durability of targeted therapies. In response, advances in multi-omics profiling, liquid biopsy technologies and artificial intelligence–assisted analytics are redefining HER2 assessment, shifting it from static tissue-based classification towards dynamic, minimally invasive monitoring. In parallel, the therapeutic landscape has evolved from single-agent HER2 blockade to rational combination strategies. In particular, antibody–drug conjugates have emerged as a transformative modality, extending the therapeutic reach of HER2 targeting beyond traditional indications and offering new opportunities to overcome heterogeneity and drug resistance. This review synthesizes current insights into the molecular and microenvironmental biology of HER2-altered solid tumors and discusses how these advances are reshaping next-generation precision oncology.
Timdarpacept (IMM01) is a recombinant signal regulatory protein α–IgG1 fusion protein that exerts antitumor activity by blocking the “don’t eat me” signal while simultaneously activating the “eat me” signal, inducing robust antibody-dependent cellular phagocytosis. We report the final phase II results of treatment-naïve higher risk chronic myelomonocytic leukemia (CMML) patients treated with IMM01 in combination with azacitidine. We assessed the safety and efficacy of IMM01 in combination with azacitidine (AZA) as first-line therapy for newly diagnosed CMML patients. IMM01 was administered intravenously at 2.0 mg/kg weekly, and AZA was administered subcutaneously at 75 mg/m² on days 1–7 of each 28-day cycle. As of the data cut-off on May 13, 2025, 24 patients were enrolled. Of 22 efficacy-evaluable patients, the overall response rate (ORR) was 72.7
