Targeted therapies and chemoimmunotherapy have transformed outcomes for non–small cell lung cancer (NSCLC), yet relapse remains common. Resistance is increasingly recognized to include an early, largely reversible phase in which a minor subpopulation survives lethal therapy through non-genetic adaptation. These drug-tolerant persister (DTP) cells may be quiescent or cycling, and provide a reservoir from which stable, genetically resistant clones can later emerge. In parallel, late recurrence may reflect tumor dormancy, in which disseminated or residual cells persist for prolonged periods under microenvironmental constraint and/or immune surveillance. This review integrates DTP and dormancy frameworks in NSCLC, summarizes mechanisms that sustain persistence (chromatin and transcriptional plasticity, stress signaling, metabolic rewiring, and stromal/immune protection), and highlights experimental models and translational readouts, including circulating tumor DNA (ctDNA)–based minimal residual disease (MRD) monitoring. We also discuss potential therapeutic concepts to prevent DTP formation, exploit persister liabilities, or enforce dormancy in minimal-disease settings. A mechanistically grounded understanding of these survival programs is essential for rational combinations and biomarker-guided trials aimed at durable remission.

Kaposi Sarcoma (KS) is a vascular tumor originating from endothelial cells and is associated with human herpesvirus 8 (KSHV) infection. It disproportionately affects populations facing health disparities. Although antiretroviral therapy (ART) has improved KS control in people with HIV (PWH), treatment options for advanced KS remain limited. This study investigates the tumor microenvironment (TME) of KS through whole-transcriptomic profiling, analyzing changes over time and differences based on HIV status. The TME was categorized into four subtypes: immune-enriched (IE), non-fibrotic, immune-enriched/fibrotic (IE/F), fibrotic (F) and immune-depleted (D). Nine KS patients (four HIV-negative and five HIV-positive) were enrolled in the study. Longitudinally collected KS samples from three patients (one HIV-negative and two HIV-positive) allowed for the investigation of dynamic TME changes within individual patients. The immune cellular composition was determined using deconvolution and compared to a cohort of non-KS patients. Our findings revealed that all KS samples, regardless of HIV status, were enriched in endothelial cells. Compared to non-KS tissues, the KS samples contained a higher percentage of NK and CD8+ T cells. HIV-negative KS samples displayed the IE and IE/F TME subtypes, while HIV-positive samples exhibited IE, IE/F, and F subtypes. Over the course of the disease, a decrease in angiogenic signatures was observed in two HIV-positive KS patients. Notably, HIV-negative KS samples showed alterations in NK cell-mediated immunity and cytotoxic response pathways, whereas HIV-positive samples exhibited changes in growth regulation and protein kinase activity pathways at the time of initial diagnosis. The gene expression of immune checkpoints, including CD274 (PD-L1) and PDCD1LC2 (PD-L2), was comparable between HIV-positive and HIV-negative KS samples at diagnosis. Furthermore, sequencing identified a shared TCRβ chain in all patients analyzed, indicating a T-cell immune response to a common antigen. This study demonstrates unique transcriptomic features and TME subtypes in KS that differ based on HIV status. Additionally, it illustrates longitudinal dynamic changes in the gene signatures and TME subtypes in individual patients. The identification of a shared TCRβ chain suggests that immune T cells in KS patients may target a common antigen. Future studies should further explore the immune microenvironment and unique T cell clonotypes, which could pave the way for the development of novel therapeutic strategies for KS patients.

Brain metastases (BMs) are a relatively common and severe complication in advanced non-small cell lung cancer (NSCLC), significantly affecting patient prognosis. Metastatic tumor cells can alter the brain tumor microenvironment (TME) to promote an immunosuppressive state, characterized by reduced infiltration of tumor-infiltrating lymphocytes (TILs), diminished expression of programmed death-ligand 1 (PD-L1), and changes in other proinflammatory factors and immune cell populations. Microglia, the resident macrophages of the brain, play a pivotal role in modulating the central nervous system (CNS) microenvironment through interactions with metastatic cancer cells, astrocytes, and infiltrating T cells. The M2 phenotype of microglia contributes to immunosuppression in BM via the activation of signaling pathways such as STAT3 and PI3K-AKT-mTOR. Recent advances have enhanced our understanding of the immune landscape of BMs in NSCLC, particularly regarding immune evasion within the CNS. Current immunotherapeutic strategies, including immune checkpoint inhibitors, have shown promise for NSCLC patients with BM, demonstrating intracranial activity and manageable safety profiles. Future research is warranted to further explore the molecular and immune mechanisms underlying BM, aiming to develop more effective treatments.