Aging of the Immune System: Causes, Related Diseases, and New Treatments

Inflammation represents complex protective immune response when confronted with infection, injury, or harmful stimuli, aiming to eliminate pathogens, repair tissues, and restore homeostasis. However, chronic inflammatory pressure induced by diverse pathologies propels aging and age‐related diseases through multiple molecular mechanisms, including remodeling the senescence‐associated epigenetic landscape. For example, cyclic GMP–AMP synthase (cGAS)–stimulator of interferon genes (STING)–type I interferon (IFN‐I) cascade, the potent driver axis involved in inflammation and cellular aging, can be activated by depletion of repressive histone marks (e.g., H3K9me3 and H3K27me3) [100, 101]. These aging processes reciprocally exacerbate chronic inflammation and amplify other hallmarks of aging in turn. For instance, systemic or local inflammation induced DNA damage prompts CD8⁺ T cells to enter an aging state, as observed in patients with obesity or chronic obstructive pulmonary disease (COPD) [102, 103]. Similarly, continuous viral antigens exposure (e.g., hepatitis B virus, HBV; and COVID‐19) activates inflammatory signals, which triggers aging phenotypes and impaired functions of T cells [104, 105, 106].

As proved by the cancer immunosurveillance theory, spectrum of inflammatory mediators is initially stimulated by T cells recognized specific tumor antigens, exerting pivotal antitumor role of various aspects, such as IFN‐γ and TNF‐α [107, 108]. Along with tumor development, tumor microenvironment (TME) undergoes evolvement, presented with a subverted and sustained adaptive‐innate immune cell crosstalk. Tumor cells employ sophisticated strategies to subvert immune surveillance, such as release of immunosuppressive mediators (e.g., IL; and angiotensin II [Ang II]) and engagement of immune checkpoint proteins. Concurrently, a cluster of chemokines (e.g., chemokine ligand 22 [CCL22]; and C‐X‐C motif chemokine ligand 12 [CXCL12]) are secreted to promote recruitment and infiltration of immunosuppressive cell populations, creating a permissive environment for tumor proliferation and metastatic dissemination [109].

Within the tumoral chronic inflammatory environment, diverse cytokines (e.g., IL and Ang II) are employed by malignant cells to destroy immune cells by arresting cell cycle or disrupting metabolic processes, ultimately impairing the cytotoxic potential of T cells and other immune effector cells [110, 111, 112]. For example, IL‐2 released by melanoma cells can promote CD8⁺ T cell senescence by activating signal transducer and activator of transcription (STAT)5–5‐hydroxytryptophan–aryl hydrocarbon receptor pathway, labeled with enhanced expression of senescence‐related genes in CD8⁺ T cells (PD‐1), lymphocyte activation gene 3, CD39, Tim‐3, and dysfunctional cytotoxicity (decreased release of IFN‐γ and TNF) [113, 114, 115, 116, 117]. Sustained exposure to interleukins (e.g., IL‐1β, IL‐6, IL‐7) also induces profound mitochondrial dysfunction in CD4⁺ T cells, culminating in cellular senescence by elevating expression of CD57, one of the senescence markers [118, 119, 120]. Apart from interleukin family, senescence of CD4⁺ T cells is under the induction of TNF‐α and IFN‐α, which can be significantly attenuated following pharmacological inhibition of TNF‐α [121]. In breast cancer‐bearing mice models, Ang II derived from breast cancer cells promotes accumulation of C‐X‐C motif chemokine receptor (CXCR)1, which drives expansion of senescent neutrophil population (CXCR4^high CD62L^low) in blood, lung, and liver [122]. Beyond cytokine‐mediated proimmunosenescence effect, direct interactions between T cells and tumor‐expressed immune checkpoint molecules are also confirmed to induce cell cycle arrest and senescence‐associated phenotypes [123]. As is exemplified by in vitro studies, coculture with multiple myeloma cells results in arrested S phase in PD‐1^high T cells, accompanied by dysregulation of cell cycle regulators (CDK6 and E2F3 transcription factors) and enhanced SA‐β‐Gal activity [111].

Updating evidences suggest a more complicated function of tumor‐derived chemokines, such as CCL22 and CXCL12 in orchestrating tumor immune evasion. These chemokines not only recruit diverse immunosuppressive cell populations (e.g., Treg cells and MDSCs), but also exacerbate immune cell senescence and dysfunction of them [124, 125, 126, 127]. Abundant infiltrated Treg cells in TME have been observed in types of tumors, such as ovarian cancer, cervical cancer, breast cancer, which is connected to poorer outcome [128, 129, 130]. Beyond the regular role in cytokine‐mediated immunosuppression, tumor‐infiltrated Treg cells also target on cell cycle dynamics and senescence processes in multiple immune cells. Specifically, Treg cell‐derived transforming growth factor‐β (TGF‐β) and other mediators engage specific surface receptors on dendritic cells (DCs), triggering Janus kinase (JAK)/STAT signaling cascade and subsequent programmed cell death ligand 1 (PD‐L1)–STAT3 signaling to drive senescence of DC ultimately [131]. Furthermore, through depleting of glucose availability in CD8⁺ T cell, Treg cells activate the ataxia telangiectasia‐mutated gene (ATM)‐associated DNA damage response (DDR) pathway to promote cellular senescence and functional impairment of them [132]. Besides, cytokines such as CCL2 and IL‐8 facilitate the recruitment and tumor infiltration of MDSCs, which facilitates senescence of immune cells by both the manner of direct cell‐cell interactions (via PD‐L1/PD‐1 axis) or long‐range communication (through exosomes or soluble factors like TGF‐β) [133, 134, 135]. In colon cancer, MDSCs exploit the PD‐L1/PD‐1 axis to downregulate NK cell‐activating receptors nuclear pore protein p46 (NKp46) and NK group 2 member D (NKG2D), effectively impairing NK cell function and promoting senescence [136]. Additionally, MDSC‐derived membrane‐bound TGF‐β further suppresses NK cell cytotoxicity by reducing surface expression of NKG2D and diminishing IFN‐γ production [137].

Hypoxia represents a symbolic event in almost solid tumors, originated from their urgent demand for oxygen and nutrient resources, even with the support of tumor neovascular or vascular mimicry. This disparity between oxygen demand and supply generates intratumoral hypoxic regions, which is characterized with abnormal accumulated lactate and hypoxia‐inducible factor‐1α (HIF‐1α), and impaired clearance of intracellular ROS. The hypoxic situation drives metabolic reprogramming in tumor cells, predisposing them into anaerobic glycolysis‐preferred energy production pattern even being placed in normoxic environments, which is defined as the Warburg effect [138]. Although with relatively low energy‐conversion efficiency, Warburg effect of tumor cells endows them with rapid generation of ATP and synthesis of intermediates, to meet the demands of tumor cell proliferation [139]. Along with that, abundant lactate is substantially accumulated to acidize TME, impairing oxygen utilization and exacerbating hypoxic conditions in feedbacks [140, 141].

Lactate, a crucial product of tumor glycolysis, serves as a carbon source to be converted to pyruvate by lactate dehydrogenase to support tricarboxylic acid cycle and proliferation of tumor cells. Clinical and experimental evidence demonstrates high levels of lactate in TME is positively correlated with malignancy and incidence of metastases [142]. Lactate also acts as a potent immunomodulatory signaling molecule, acidifying surrounding TME and eliciting proinflammatory effect [143]. As suggested in renal cell carcinoma, lactate‐rich environment dampened cytokine secretion ability of cytotoxic T lymphocytes (CTLs), through inhibiting phosphorylation of mitogen‐activated protein kinase (MAPK) pathway (p38, c‐Jun N‐terminal kinase [JNK], and c‐Jun), ultimately compromising proliferation and activation of CTLs [144]. Lactate‐induced acidosis activates cyclic adenosine monophosphate (cAMP)‐dependent signals and transfer cAMP into T cells through gap junctions, culminating in DDR activation and T cell senescence [145]. On this basis, IFN‐γ produced by CTLs is heavily blocked by lactate, which potently inhibits NK cell function [146]. The immunosuppressive effect lactate can also be elicited by acidifying intracellular environment, inducing mitochondrial dysfunction and ROS accumulation in immune cells. For instance, in liver metastases of colorectal cancer, elevated lactate levels induce mitochondrial stress in liver‐resident NK cells, impairing their tumoricidal capacity and accelerating NK cell senescence [147]. Furthermore, lactate absorbed by B cells triggers cellular senescence, characterized with increased secretion of SASP factors [148].

In response to hypoxic condition, tumor cells activate HIF‐1α to orchestrate metabolic reprogramming and stimulate angiogenesis. This adaptive response not only enhances anaerobic glycolysis but also boosts excessive secretion of angiogenic factors, resulting in formation of structurally and functionally abnormal vasculature that paradoxically exacerbates tumor hypoxia. Disrupted metabolic processes and activated HIF‐1α signaling pathways within immune cells accelerate cellular senescence and compromise their antitumor immunity. In breast and bladder cancer models, HIF‐1α‐mediated activation of NF‐κB in macrophages upregulates CD47 expression, which in turn promotes macrophage senescence by upregulating senescence‐associated markers p16, p53, and p21 [149, 150, 151]. HIF‐1α also increases quantity of Treg cell, expedites secretion of CTL‐associated antigen‐4 (CTLA‐4) and IL‐10, consequently aborts effector T cell (Teff) from activation and induces senescence [152, 153, 154]. Moreover, HIF‐1α establishes a self‐reinforcing signaling loop via miR‐125a‐5p‐mediated positive feedback pattern, activating NF‐κB pathway to drive secretion of SASP factors, such as IL‐6 and TNF‐α [154].

Hypoxia in TME compromises antioxidant defenses, reducing ROS clearance and creating a redox imbalance. This oxidative stress acts as both an initiator and effector of immune senescence via mitochondrial dysfunction and cell cycle disruption [155]. In melanoma models, ROS directly disrupts immune cell mitochondrial membranes, triggering senescence pathways like mammalian target of rapamycin (mTOR), p53/p21, and NF‐κB [156]. Moreover, ROS stimulates the secretion of SASP components, including matrix metalloproteinase (MMP)‐1, from immune cells. Correspondingly, attenuating ROS levels by ROS scavengers (e.g., N‐acetylcysteine and glutathione) restores proliferative potential and reverses senescence in Treg cells [157].

Age‐related mitochondrial dysfunction and metabolism disorder have been confirmed in various cells and tissues, serving as an accelerator for the aging of multiple immune cell subtypes [158]. In particular, tumor cells undergo profound metabolic remodeling to accommodate their heightened proliferative demands, which limits the availability of nutrients and exacerbates acidosis. Competing with tumor cells for critical nutrients (e.g., glucose) intensifies the physiological stress of T cells, activating ATM‐related DDR and MAPK‐related pathways to drive cell cycle arrest and senescence [132]. Within this hostile metabolic environment, immune cells in TME undergo extensive metabolic reprogramming, which impairs their survival and function [159].

Anaerobic glycolysis‐dependent catabolism pattern of tumor cells aggravates consumption of nutrients (e.g., glucose, glutamine, arginine) in TME [140, 160]. In particular, restricted glucose directly hinders proliferation, metabolism and function of immune cells, especially T cells [161]. In contrast, exogenously supplementation with glucose restores glucose‐deprived CD8⁺ T cell proliferation, cytokine production, and metabolic function [162]. In CD4⁺ T cells, depriving glucose directly promotes phosphorylation of MAPK and p38, leading to cell cycle arrest and proliferative impairment [163]. As further revealed by in vitro study, being placed in glucose‐restricted conditions impedes glutamine uptake a critical substrate for mitochondrial respiration while disordered mitochondrial metabolism stimulates senescence related signaling pathways (e.g., p16^INK4A/pRb) in CD4⁺ T cells [164]. Treatment with 2‐deoxy‐d‐glucose, an inhibitor of glycolysis, also induces G1/S phase arrest accompanied by increased senescence of CD8⁺ T cells [165, 166]. Analogously, high consumption of glucose in natural regulatory T (nTreg) cells or tumor‐associated γδTreg cells leads to intratumoral glucose scarcity [132]. Inhibiting glucose transporters and glycolysis in Treg cells attenuates their suppression on Teff cells proliferation, as well as Treg cells‐induced SA‐β‐gal⁺ senescent T cells [132]. In addition to glucose depletion, polyamine metabolic pathways are highly active in tumor cells such as liver cancer cells significantly enhancing the synthesis and excretion of N1‐acetylspermidine. Excreted N1‐acetylspermidine activates SRC signaling pathway in macrophages, driving the polarization toward CCL1⁺ macrophages with immunosuppressive properties [167].

Pathological metabolic dysregulation perturbs core metabolites and metabolic enzymes, thereby profoundly reshape the epigenetic landscape in immune cells especially via those pivotal epigenetic switches, such as DNA methylation and histone modifications. In elderly individuals, CD4⁺ T cells exhibit reduced mitochondrial function and one‐carbon metabolism compared to younger counterparts, a defect mediated by impaired methylation of mitochondrial transfer RNA that compromises energy metabolism [168, 169]. Pervasive mitochondrial dysfunction depletes α‐KG in senescent T cells, inhibiting tet methylcytosine dioxygenase 2 (TET2)‐mediated DNA demethylation and culminating in epigenetic silencing of CD28 and CD27 genes [121]. Furthermore, deficiencies in glucose‐metabolizing enzyme, particularly those governing the nonoxidative pentose phosphate pathway and transketolase, elevate oxidative stress and exacerbate mitochondrial dysfunction, contributing to DNA hypermethylation and functional gene suppression during progressive immunosenescence [170, 171].

NK cells serve as critical effectors of innate immunity, which are responsible for recognizing and eliminating aberrant cells, such as senescent or tumor cells. Their activity is governed by both surface activating and inhibitory receptors, such as CD16 and CD56, as well as immune checkpoint molecules. Based on the expression of CD16 and CD56, NK cells are conventionally categorized into two functional distinct subsets: immature (CD16⁺CD56^dim) and mature subtypes (CD16⁺CD56^bright). CD56^bright NK cells primarily secrete cytokines such as IFN‐γ and TNF‐α to act immunomodulatory role. In contrast, the CD56^dim subset constitutes the predominant subgroup in peripheral blood, eliminating tumor cells at immune synapses and secreting cytotoxic molecules, such as perforin and granzyme.

Organismal aging profoundly alters NK cell biology, manifested with skewed subset distribution (e.g., declined CD56^bright/CD56^dim ratio), phenotypic alterations (e.g., reduced expression of activating receptors like nuclear pore protein p30, NKp30), expansion of CD56^dimCD57⁺ NK cells, reduced cytokine secretion, and cytotoxicity against target cells (Figure 3) [186, 187]. In particular, NK cells with CD56^dimCD57⁺ are described as a reliable symbol of senescent NK cells, which has been detected in patients with atherosclerosis, Parkinson disease (PD), or underwent human cytomegalovirus infection [188, 189, 190]. These age‐related changes are conserved in mice, with significant reductions in mature NK cell populations in lymphoid organs of aged animals [80]. In patients with Alzheimer's disease (AD), the number of NK cells also significantly decreases [191]. The altered subset composition in elderly individuals correlates with diminished TNF‐α and IFN‐γ secretion capacity, which persists even following IL‐2 or IL‐12 stimulation [192]. Furthermore, in vitro studies demonstrate that aging compromises the cytotoxic machinery of NK cells. In a series of NK cell–leukemia cell coculture experiments, perforin produced by NK cells from elderly donors are reduced compared with those from younger counterparts. This perforin deficiency directly contributes to the age‐associated decline in NK‐mediated tumor cell killing, representing a key mechanism of immune senescence in the NK compartment [193].

In addition to subset redistribution, senescent NK cells exhibit unique characteristic phenotypics. A hallmark feature of NK cell senescence is the downregulation of surface activating receptors, particularly natural cytotoxicity receptors (NCRs; including NKp30 and NKp46) and NKG2 family members, which significantly compromises their antigen recognition capacity. This receptor deficiency leads to impaired cytokine production and diminished cytotoxic function against target cells. Mechanistically, these critical receptors mediate NK cell activation through distinct ligand interactions: NKG2D (a C‐type lectin‐like receptor) recognizes major histocompatibility complex class I‐related molecules expressed on tumor cells, while NCRs (NKp30 and NKp46) bind to immunoreceptor tyrosine‐based activation motif (ITAM)‐bearing ligands [194]. Accumulating evidence indicates that tumor‐derived factors downregulate NKp30, NKp46, and NKG2D expression, resulting in diminished cytotoxic activity of NK cells against tumors and contributing to NK cell senescence. Clinical observations consistently demonstrate this phenomenon across various malignancies. In patients with non‐Hodgkin lymphoma or acute myeloid leukemia, significant downregulation of NKp30, NKp46, and NKG2D has been well documented [195, 196]. Similarly, Kaposi's sarcoma patients exhibit markedly reduced expression of NKG2D and NKp30 compared with healthy controls, with receptor levels being restored following successful treatment‐induced remission [64].

Senescence of NK cell also performed with metabolic reprogramming, especially alterations in glucose and lipid metabolism mediated by tumor‐derived cAMP [197, 198]. Collected from murine breast cancer models or when cocultured with breast cancer cells, uptake of tumor‐derived cAMP markedly increased SA‐β‐Gal⁺ NK cells, along with impaired expression of effector molecules like granzymes A/B (GZMA/B), perforin, and IFN‐γ. Through high‐throughput sequencing, elevated transcriptional activity of metabolism‐associated genes flux and glucose/lipid metabolic level are unearthed in those senescent NK cells. Besides, treatment of cAMP inhibitor substantially attenuated breast cancer cell‐induced senescence in both murine and human NK cells. Parallel phenomena were observed in melanoma models. For instance, tumor‐derived cAMP in melanoma‐bearing mice upregulated glycolipid metabolism‐related genes in murine NK cells, which is repeatable in human NK cells cocultured with melanoma cell, accompanied with elevated senescent SA‐β‐Gal⁺ NK cells [86].

Macrophages are regarded as primary effectors of adaptive immunity, orchestrating the recognition, phagocytosis, and degradation of cellular debris and pathogens through intracellular degradation mechanisms including autophagy. Senescent macrophages exhibit sustained metabolic impairment that alters polarization dynamics and drives functional decline, manifesting as diminished antigen presentation and loss of phagocytic capacity. Macrophages acquire senescent characteristics during both physiological aging and pathological conditions, exhibiting imbalanced polarization states and elevated expression of senescence markers (p16^INK4A, SA‐β‐gal). As for macrophages in aged murine models, there is tissue‐specific polarization patterns. For instance, immunosuppressive M2 macrophages are observed predominantly accumulated in bone marrow, spleen, lymph nodes, lungs, and muscle tissues of aged mice [199, 200]. In comparably, proinflammatory M1 macrophages are enriched in adipose tissue and liver [201, 202]. Except for natural aging processes, p16^INK4A‐positive senescent macrophage populations expand under diverse pathogenic conditions, including PD, osteoarthritis, streptozotocin‐induced diabetes, and ionizing radiation exposure [203, 204]. Radiation‐induced senescence specifically increases SA‐β‐gal⁺ bone marrow‐derived monocytes/macrophages [205].

Tumor‐associated macrophages (TAMs) represent a critical cellular constituent of TME, constituting one of the most prevalent immune cell populations across diverse malignancies [206]. During initial tumor development, malignant cells recruit macrophages through paracrine secretion of cytokines and chemokines, including granulocyte–macrophage colony‐stimulating factor (GM‐CSF), CCL2, and CXCL4. TAMs polarization exhibits microenvironment‐dependent plasticity, with functional specialization dictated by local signaling milieus. Proinflammatory M1‐like TAMs emerge in response to GM‐CSF and IFN‐γ stimulation, characterized by STAT1 and suppressor of cytokine signaling three pathway activation, which promotes antitumor immunity [207]. Conversely, exposure to colony‐stimulating factor 1 (CSF‐1), IL‐4, and IL‐13 induces M2‐like polarization, associated with immunosuppressive and tumor‐promoting functions [208]. This dynamic polarization capacity enables TAMs to perform content‐dependent roles throughout tumor progression and in response to therapeutic interventions.

When being confronted with tumor, senescent signatures always endow TAM populations with discordant polarization and immunosuppressive capacity to assistant tumor progression, as evidenced by clinical observations. For example, markers of M2 macrophage (e.g., CD163; V set and Ig domain‐containing 4, VSIG4) are upregulated pronouncedly in elderly (>60 years) or prostate cancer patients [59]. This phenomenon is recapitulated in KRAS‐driven lung cancer models, where CD206⁺ M2‐polarized TAMs coexpress the senescence marker p16 [57]. In accompany with those TAMs, a subset of tissue‐resident alveolar macrophages (AMs) also exhibit senescence‐associated signatures, including elevated expression of cell cycle inhibitory gene (e.g., Ctnnb1), suggesting an indirect tumor‐mediated induction of macrophage senescence [56]. Similar evidence has been collected in glioblastoma models, where tumor‐derived IL‐6 activates STAT3 signaling to initiate senescence in peritumoral macrophages rather than TAMs [209]. The presence of senescent TAMs has been confirmed to impair antitumor immunity by suppressing infiltration and function of CD8⁺ T cells. Senescent AMs in lung cancer models restrict accumulation of CD8⁺ T cells, while depletion of AM successfully restores infiltration of CD8⁺ T cells [57]. Similarly, senescent macrophages in glioblastoma attenuate response of T cells by producing arginase‐1 (Arg1) to downregulate expression of CD3ζ in T cells [209].

As an essential component of adaptive immune system, B lymphocytes undergo a tightly regulated differentiation process beginning with pro‐B and pre‐B cell stages before maturing into naive B cells characterized by high IgD and low IgM expression. Upon tumor antigen encounter, naive B cells experience clonal expansion and differentiate into matured plasma cells or memory B cells respectively. Specific antibodies are synthesized and secreted by matured plasma cells to facilitate effector cells recruitment (e.g., NK cells, macrophages) and mediate their antibody‐dependent cytotoxic effect [210]. Memory B cells can be divided into CD27^dull and CD27^bright subpopulations, while the former one promptly differentiate into plasma cells when recognizing tumor cells. Comparably, upon antigen re‐exposure, CD27^bright B cells can also be differentiated into plasma cells and secreted specific antibodies toward tumor cells in a relatively tardive mode [211].

During the process of aging, there are significant alterations in both quantity and function of B lymphocytes, manifesting as decreased pro‐B cell numbers alongside expanded memory B cell compartments. Major changes of this age‐associated lymphoid‐myeloid shift include reduced CD27^dull/CD27^bright B cell ratio, and diminished antibody repertoire diversity. The myeloid‐biased differentiation of senescent HSCs is responsible for dampened production of pro‐B cells [95]. TNF‐α secreted by aged B cells suppresses proliferation and induces programmed cell death of pro‐B cells [212]. The nuclear structure of aging pro‐B cells also experiences architectural changes, reduced topological domain on Ig heavy chain (IgH) sites related to B cell development hindered the differentiation of progenitor B cells and caused cell apoptosis [213]. Similar phenomenon has also been observed in aged canine models [83]. In comparison, the number of memory B cells exhibits age‐related expansion, and the ratio of CD27^dull /CD27^bright B cells in memory B cells decreases during aging [214]. Comparative analyses reveal significant reduction in CD27^dull subsets with concomitant dominance of CD27^bright populations in elderly individuals, which is a phenotypic shift associated with impaired neoantigen responsiveness [211, 215].

Furthermore, senescent B cells display marked reduction in both antibody repertoire diversity and production capacity [216, 217]. Experimental evidence from in vitro stimulation assays demonstrates that IgM and IgA in peripheral blood B cells from the elderly are diminished compared with younger counterparts [215]. This phenomenon is also observed in lymphoid follicles of ileum [217]. The transcription factor E47 in senescent B cells downregulates activation‐induced cytidine deaminase (AID), failing to catalyzing class switch recombination and somatic hypermutation (SHM), blocking conversion from IgM, IgD to distinct secondary isotype antibodies (IgG, IgA, IgE), ultimately reducing antibody diversity and synthesis [218, 219].

Pathological conditions such as chronic infections or malignancies are confirmed to induce sustained inflammation and senescence in B cells. For instance, in the circumstances of autoimmune diseases (systemic lupus erythematosus [SLE] and RA) or AD, B cells are driven to differentiate into age‐related phenotype by abnormally accumulated proinflammatory factors such as IL‐6 and TNF‐α [220, 221]. Senescence of B cells appears after being infected with HIV even in young individuals [222]. Similarly, chronic inflammatory environment of tumor is also found to drive senescence of B cells. As exemplified in triple‐negative breast cancer (TNBC), breast cancer cells provoke myeloid‐biased niche competition between pro‐B cells and myelocytes in bone marrow, ultimately leading to pro‐B cell depletion [223]. Furthermore, increased proportion of memory B cell has been documented across multiple malignancies, including breast carcinoma, non‐small cell lung cancer (NSCLC), and ovarian cancer [224, 225, 226]. As for B lymphocytes infiltrated inside tumor‐associated tertiary lymphoid structures (TLS), CD27^dull/CD27^bright B cell ratios demonstrate significant reduction, which is correlated with unfavorable clinical outcomes in head and neck squamous cell carcinoma (HNSCC) and TNBC [54, 65]. B cells‐related immunosenescence is also marked with impaired humoral immunity, featured with antibody diversity restriction and localized antibody deficiency. For example, in cutaneous melanoma, higher diversity of antibodies production is associated with higher proportion of B cells expressing AID and performing SHM, which is positively related to improved prognosis [227]. Compared with peripheral blood, secretion IgM is attenuated in B cells located in TLS of pancreatic ductal adenocarcinoma [228].

Originating from primordial T cell precursors, mature T lymphocytes are broadly categorized into CD8⁺ and CD4⁺ T cell subsets. Generally, CD8⁺ CTLs mediate direct tumor cell killing through granzyme/perforin‐dependent mechanisms, while CD4⁺ T helper (Th) and Treg cells maintain immune homeostasis via cytokine‐mediated regulation. There is significant alteration of T cell profiles during age, featured with a declined proportion and proliferative capacity of CD8⁺ T cells, along with diminished cytokine secretion (e.g., GZMB), cell cycle arrest, and shifted phenotypes (e.g., decreased CD28 and increased Tim‐3 expression) (Figure 4). The age‐related decline in CD8⁺ T cells has been observed in aged individuals across many species including murine and canine models [84, 229, 230]. Under stimulation of antigens, CD8⁺ T cells mediated immune responses can be initially activated in lymph nodes through binding to IL‐4 derived from CD4⁺ T cells, followed by effector differentiation [231]. However, aging‐associated attrition of IL‐4‐producing CD4⁺ T cells directly compromises proliferation of CD8⁺ T cells [232]. In addition to altered abundance and proportion, there are molecular reprogramming events and function defects in senescent T cells. For instance, cell cycle inhibitor proteins (e.g., p16^INK4A, p21, and p53) are uniformly upregulated [233]. Conversely, CD28, a critical costimulatory molecule in antigen recognition and T cell activation, is markedly downregulated in senescent T cells, disrupting their binding with antigen‐presenting cells to weaken T cell‐mediated cytotoxicity, which has been detected in patients with HIV infection or RA [234, 235]. Furthermore, inhibitory immune checkpoint protein such as Tim‐3 is significantly upregulated in senescent T cells, braking the production and secretion of tumoricidal cytokines, such as IL‐2, TNF‐α, GZMB, and IFN‐γ, as well as restricting cytotoxicity of those senescent CD8⁺ T cells [85, 236].

Beyond aging‐related physiological T cell senescence, TME and other pathologic conditions are confirmed to contributed to T cell senescence through distinct mechanisms. Compared with healthy peripheral blood, a marked reduced population of circulating CD8⁺ T cells is observed in blood from patients with multiple malignancies, including colorectal cancer, Hodgkin lymphoma, breast cancer, and NSCLC [237, 238, 239, 240]. In particular, tumor‐infiltrating T cells are typically presented with classic immunosenescent characteristics, incorporating with impaired production of cytokines (e.g., GZMB) and phenotypic changes (e.g., downregulated expression of CD28 and upregulated expression of Tim‐3). In pancreatic ductal adenocarcinoma, CD8⁺ T cells infiltrated in both tumor tissue or metastatic lymph nodes demonstrate compromised capacity of secreting GZMB, constructing an immunosuppressive TME [241]. Downregulated CD28⁺ T cells have been observed in colorectal cancer, effectively crippling antitumor immune surveillance by disrupting CD8⁺ T cell activation [242]. Not surprisingly, CD28⁻CD8⁺ T cells are found to be abnormally expanded in individuals with NSCLC [243].

Continuous stimulation of tumoral antigens directs to T cell exhaustion through facilitating immune checkpoint factors, which also serves as senescent modulator of T cells. Based on the AI‐mediated trajectory calculation via AI, the dominant role of antigen‐specific epigenetic regulatory patterns in CD8⁺ T cells aging has been revealed [244]. The PD‐1/PD‐L1 axis also represents a prototypical example of this dual regulatory mechanism. Following PD‐1/PD‐L1 ligation, immunoreceptor tyrosine‐based inhibitory motif (ITIM) and immunoreceptor tyrosine‐based Switch motif domains of PD‐1 are phosphorylated, converting Src homology 2‐containing protein tyrosine phosphatase 2 (SHP‐2) into an active conformation. Activated SHP‐2 dephosphorylates CD3ζ and costimulatory signal CD28 to restrain activation of phosphoinositide 3‐kinase (PI3K)/protein kinase B (AKT) signaling pathway, which not only participates in functional failure, but also leads to cycle arrest and senescence in T cells [245]. This is in accordance with the clinical evidence that expression of PD‐L1 is significantly negatively correlated with infiltration of CD8⁺ T cells, such as in ovarian cancer and NSCLC [246, 247]. Contained in extracellular vesicles derived from tumor cells, PD‐L1 activates cAMP‐response element binding protein (CREB) and STAT signaling pathways, which induce high expression of lipid metabolism enzymes in T cells, promoting lipid metabolism to induce T cell senescence [248]. Blocking the binding of PD‐1 and PD‐L1 effectively reverses senescence process of T cells [249, 250]. By the methods of gene knockout or inhibitors, blocking interaction between PD‐L1 and T cells reverses aging state of T cells, through inhibiting CREB and STAT signaling pathways [248]. This phenomenon has also been demonstrated in mice model, where PD‐1 blockade improves cytotoxic capacity and reverses senescent CD8⁺ T cells in old mice [251]. In addition, PD‐1 antibody can also reduce expression of p16 and activate CD8⁺ T cells to improve systematic immunosenescence [249]. Treatment with anti‐PD‐1 therapy limited senescent T cells but recovered expression of CD28 in Hodgkin lymphoma patients [238].

Extensive metabolic reprogramming events in TME also takes charge of senescence of infiltrated T cells, primarily through glucose deprivation and release of cAMP. High glycolytic activity of tumor cells deprives environmental glucose, that hastens senescence of immune cells [252]. Additionally, senescent T cells prefer to produce energy through anaerobic glycolysis, magnifying mitochondrial malfunction and generating ROS to expedite senescent process [253]. Tumor‐secreted TGF‐β drives differentiation of CD4⁺ T cells into Treg cells, which exhibit enhanced glucose uptake and glycolysis capacity, further intensifying glucose scarcity [254, 255]. In addition, metabolites of glycolysis of tumor cells promote senescence of T cells in indirect manners. In a tumor–T cell coculture system, an aging phenotype emerges in healthy donor‐derived human T cells followed with stimulation of tumor‐derived cAMP, including decreased expression of CD28, shortened telomere, increased p53, p21, and p16 [87]. Tumor‐derived cAMP diffuses into T cells through intercellular gap junctions and binds to regulatory subunits of inactive protein kinase A (PKA) tetrimers, resulting in the release and activation of catalytic subunits. The subsequent phosphorylation of CREB activates the PKA–CREB signaling axis, ultimately driving T cell senescence [145, 256]. Other cellular signals like NF‐κB, CCAAT/enhancer‐binding protein (C/EBP‐β), and cGAS–STING are also critical in inducing senescent T cells [257].

Tumor cells have evolved a variety of mechanisms to actively evade immune surveillance, such as downregulation of immunogenicity, secretion of immunosuppressive factors, and metabolic competition. Correspondingly, the restricted immune response in the process of immune senescence also passively promotes tumor immune escape. Immunosenescence fuels formation of malignant microenvironment, intensifying tumorigenesis in a self‐perpetuating cycle. According to epidemiological data, there is a higher risk of cancer in middle‐aged and elderly populations than adolescents or newborns [383, 384]. In orthotopic lung adenocarcinoma and colorectal carcinoma models, there is a larger tumor burden in aged mice than younger controls, revealing a stronger supporting effect on tumor growth in aged organisms [385]. It is generally believed that not only aging‐accumulated oncogenic mutations raise the risk of cancer, but also the dynamic evolved multidimensional immunosenescence plays a vital role in regulating tumor development.

Age‐related decline in immune cell quantity and function, particularly in T cell populations, significantly compromises immunosurveillance and cytotoxic capacity, thereby facilitating tumor development [386]. Comprehensive analyses of tumor‐associated characteristics reveal an inverse correlation between aging and T cell abundance, concomitant with increased cancer incidence [216]. Additionally, senescent immune system promotes chronic inflammation and angiogenesis by producing SASPs such as IL‐1, IL‐6, TNF‐α, and IFN‐γ contributing to creating a tumor‐tolerant microenvironment [119]. For example, senescent T cells and HSCs increase the secretion of IL‐1α of IL‐1β and recruitment of MDSCs and neutrophils, suppressing infiltration and function of CD8⁺ T and NK cells [263, 387, 388]. Furthermore, once bound with IL‐1R, IL‐1α triggers nuclear translocation of NF‐κB, and sequentially activates transcription of inflammatory mediators including IL‐6 and IL‐8 [389]. Senescent cell‐derived IL‐8 activates macrophages, further promoting production of inflammatory cytokines and infiltration of macrophages to be involved in the occurrence of HCC [291]. Under synergistic stimulation of extracellular IL‐6 and IL‐8, breast cancer cells amplify secretion of IL‐6 and IL‐8 to intensify tumor‐related inflammatory environment by self‐enhancing feedback [277].

SASP components (e.g., IL‐1 and IL‐8) also advertise tumor neovascularization, facilitating oxygen and nutrients supply to meet the requirement of rapid tumor proliferation [390]. For example, IL‐1β enhances angiogenesis by upregulating proangiogenic factors while concurrently suppressing angiogenesis inhibitors in senescent fibroblasts [391]. Similarly, a larger extent of angiogenesis is observed in mice injected with breast cancer cells with elevated IL‐8 expression [289]. Conversely, targeted ablation of senescent macrophages and associated SASPs in lung cancer models lowers tumor vascular density and overall tumor burden, effectively attenuating disease progression [57].

Accompanied with in situ growth of tumors, other malignant phenotypes such as invasion and metastasis occur in most tumors. Acquire invasive capacity through epithelial–mesenchymal transition (EMT) is regarded as the initial metastatic event, before migrating into adjacent blood or lymphatic vessels to enter circulation as circulating tumor cells (CTCs) [392]. Changes of cellular senescence pathways in tumor cells are involved in the regulating tumor metastasis. As revealed by tumor samples from relapsed cervical carcinoma patients, there is opposing altered expression of senescent markers and SASP factors between patients with local and distant relapses following radiotherapy [393]. In contrast, abundant SASP factors especially members of IL family are effective in stimulating EMT. For instance, IL‐6 and IL‐8 secreted by breast cancer cells induce EMT and promote acquisition of more aggressive phenotypes [277].

Once entering circumstance, substantial apoptosis occurs in CTCs due to mechanical stress, redox pressure, immune surveillance and tumor killing effect [392, 394]. Aged immune system, particularly marked with significantly declined peripheral CD8⁺ and CD4⁺ T cells, is commonly fatigued to recognize or clear CTCs [395]. During melanoma metastasis, a Wnt antagonist, secreted frizzled‐related protein 2 (sFRP2), can be secreted by senescent fibroblasts to enhance resistance to redox stress of tumor cells [396]. Subsequently, a unique premetastatic niche (PMN) conducive to colonization in target organ is generated, which is under regulation of both tumor cells at primary site and CTCs through paracrine effect [397]. Senescence‐related immune system alterations are reported to shift the microenvironment toward a PMN. For example, there is upregulated expansion of γδT cells and protumor neutrophils in aged mice with melanoma, which limits cytotoxicity of CD8⁺ T cells to build a metastasis‐permissive microenvironment [398].

After reaching target organs, disseminated tumor cells (DTCs) enter a state of temporary dormancy, maintained by immune surveillance and limited blood supply, while retaining the potential for reactivation upon immune system compromise [399]. Immunosenescence declines surveillance ability of CD8⁺ T and NK cells, disrupting the delicate equilibrium between dormant DTCs and the immune microenvironment. This imbalance is evidenced in aged murine models, where bone marrow exhibits reduced frequencies of dormant DTCs concomitant with enhanced tumor cell proliferation [400]. As for hepatic metastatic model, activated HSCs (aHSCs) suppress proliferation of NK cells by secreting CXCL12 to terminate dormancy of DTCs [401]. Moreover, a series of anti‐inflammatory factors (e.g., TGF‐β) are synthesized to recruit immunosuppressive cell populations (e.g., Treg cells, MDSCs), which exacerbates immunosenescence and reinforces an immunosuppressive niche [402]. This reciprocal tumor–immune interactions accelerates tumor metastatic colonization and development.

Recent advances in innovative tumor therapies, particularly immunotherapy, have revolutionized conventional cancer treatment paradigms encompassing surgery, chemotherapy, and radiotherapy. Nevertheless, clinical outcomes remain suboptimal for most patients. Multidimensional computational profiling enables quantitative prediction of therapeutic response and resistance development, thereby optimizing treatment strategies. Patients with elevated immune senescence indices (ISI) scores exhibit poor prognosis across malignancies, characterized by suppressed antigen presentation and ligand‐receptor downregulation [27]. Furthermore, peripheral blood biomarkers of immunosenescence complement traditional tumor assessments (e.g., histopathology and imaging), offering minimally invasive dynamic monitoring. For instance, regression analysis of lymphocyte subsets in 332 nasopharyngeal carcinoma patients identified CD28 downregulation in senescent CD8⁺ T cells as a stratification biomarker and independent prognostic factor for progression‐free survival and distant metastasis‐free survival, longitudinally validated to predict suboptimal radiotherapy response [50].

Critically, immunosenescence contributes to chemoradiotherapy failure and increased posttreatment recurrence risk. High‐dose therapy damages HSCs, depleting immune cell reserves and compromising systemic immunity. Treatment pressure upregulates immunosenescence across malignancies. For example, peripheral T cells in breast cancer exhibit transient downregulation of CD28/CD27/leucine rich repeat neuronal protein 3 (LRRN3) expression postchemotherapy, while TGF‐β induced by chemoradiotherapy promotes senescent CD27⁻CD8⁺ T cells in cervical cancer [53]. These cells directly associate with adverse outcomes and demonstrate elevated frequencies in postchemoradiotherapy recurrent cases [61].

Almost immunotherapeutic strategies targeting immunosuppression to restore antitumor immunity, particularly immune checkpoint blockade (ICB) via small molecules, are confirmed to exhibit age‐dependent efficacy disparities. AI‐based modeling reveals that elevated senescence scores correlate significantly with immune‐suppressive cells (e.g., Treg cells and MDSCs) infiltrated within tumors, whereas linked with diminished responses after ICBs treatment [25]. In elderly patients with advanced NSCLC, impaired ICB efficacy is attributed to T cell senescence in the TME. Terminally differentiated T cells accumulated in these patients demonstrate coupregulation of PD‐1/CTLA‐4/T cell Ig and ITIM domains, which disrupts immune synapse formation [403]. Similarly, nonresponding melanoma patients exhibit higher frequencies of senescent CD4⁺ and CD8⁺ T cells (marked by reduced CD27/CD28 expression) compared with ICI responders, an observation validated in murine colorectal cancer models where PD‐L1 antibody fails to suppress tumor growth in aged mice [404]. Depleting senescent cells with ABT263 (navitoclax) enhances intratumoral CD8⁺ T cell infiltration and prolongs survival [309]. Therefore, resistance to ICBs can be attributed to both the off‐target effects generated from heterogeneity of senescent cell subpopulations and compensatory upregulation of PD‐L1 through immunosenescence‐driven pathways [405, 406]. For instance, senescent T cells overproduce IFN‐γ, promoting internalization of PD‐L1‐rich exosomes that evade lysosomal degradation via chaperone‐mediated autophagy, thereby inducing adaptive resistance [248]. Furthermore, monocyte‐triggered IFN‐γ upregulates the PD‐L1 axis, enabling endogenous monocyte‐derived PD‐L1 to evade conventional antibody targeting and perpetuate treatment resistance [407].

As the most prevalent age‐related neurodegenerative disease, the incidence of AD increases dramatically along with age. There are approximately 50 million individuals are affected by AD globally, with the highest number of patients reaching 9.83 million in China [417]. The major clinical syndromes of AD are manifested as cognitive and behavioral impairments. Current mainstream theory of AD centers on the abnormal accumulation of amyloid‐beta (Aβ) and tau protein pathology, which is characterized by accumulating extracellular Aβ plaques and intracellular neurofibrillary tangles mainly composed of hyperphosphorylated tau protein, accompanied by excessive activation and proliferation of astrocytes. In addition to the well‐documented genetic and environmental factors, disruption of CNS homeostasis and immunosenescence‐related chronic neuroinflammation have also been confirmed to act parallelly in AD.

Microglia are the resident macrophages of CNS, widely involved in CNS immune surveillance, inflammatory responses, and neural development. In AD, microglia serve as the pivotal cells responsible for eliminating Aβ, whereas significant impaired phenotypes of microglia appear during AD. For example, the downregulation of Aβ phagocytic receptors (e.g., scavenger receptor class A member 1, SCARA1; CD36) and the upregulation of inhibitory receptors (e.g., CD33) contribute to impaired phagocytic function of microglia [418]. Occurrence of lysosomal dysfunction in microglia fails to degrade Aβ [419]. Compared with healthy mice, microglia of AD mice display typical aging phenotypes (increased expression of p16^INK4A, p21, and p53) [420]. These senescent characteristics of microglia such as cell cycle arrest and the expression of p16^INK4A and SA‐β‐gal seem to be pathological response after uptake of hyperphosphorylated tau protein (hp‐tau) [421]. Accumulated intracellular Aβ has also been reported to activates p53, resulting in DNA damage and aging of microglial aging [422].

Aging immune cells, especially microglia and T cell, are fatigue to promptly remove neurotoxic substances. Instead, they are modified into proinflammatory phenotype to secret a cluster of SASP cytokines such as IL‐1β, IL‐6, and TNF‐α to exacerbate neuronal damage and synaptic dysfunction. Senescent microglia facilitate the release of proinflammatory factors (e.g., IL‐6 and IL‐8) through lactate–H3K18la–NF‐κB axis, promoting the aggregation of Aβ protein and contributing to neuronal dysfunction and death [423]. IL‐1α and TNF can expand the neuronal damage by transforming adjacent astrocytes into proinflammatory, neurotoxic A1‐type cells, and accelerating the progression of AD [424, 425].

PD is the second most common neurodegenerative disease after AD, affecting approximately 8.5 million people globally and especially in individuals 60 years old [426]. The main clinical symptoms of PD are motor symptoms (e.g., muscle stiffness, resting tremor, bradykinesia, and postural instability) and nonmotor symptoms (e.g., sleep disorders, hallucinations, and dementia) [427]. Pathological hallmark of PD consists of progressive degeneration of dopaminergic neurons in the substantia nigra pars compacta. Meanwhile, misfolded α‐syn is aggregated to form insoluble Lewy bodies, reducing astrocytic uptake of neurotoxic substance to induce neuronal death [428]. In addition to genetic and environmental risk factors, the chronic inflammatory environment mediated by immunosenescence have also been confirmed to play a critical role in PD.

Excessive neurotoxic substances like α‐syn are supposed to be removed by microglia to maintain CNS homeostasis [429]. However, aging‐related morphological changes of microglia in the substantia nigra of PD patients such as reduced branching impairs their ability of clearing neurotoxic substances (e.g., α‐syn) [430]. Accumulated α‐syn interacts with toll‐like receptor (TLR)2 and TLR5 receptors in microglial, triggering activation of NOD‐like receptor family pyrin domain containing 3 (NLRP3) inflammasome, preventing α‐syn uptake and degradation as a feedback [431]. Similar to the situation of AD, senescent microglia also transform into proinflammatory phenotypes, releasing cytokines such as TNF‐α and IL‐1β creating a chronic inflammatory microenvironment to accelerate PD progression. For instance, in aging mice, reduced expression of the triggering receptor expressed on myeloid cells‐2 gene in microglia induces a shift from an anti‐inflammatory to a proinflammatory phenotype, enhancing neuroinflammation in the substantia nigra–striatum pathway [432]. Proinflammatory microglia release TNF‐α, IL‐1β, and IL‐6 via TLR4/NF‐κB pathway, directly damaging dopaminergic neurons in the substantia nigra. Mitochondrial complex I dysfunction in neurons can be induced by proinflammatory factors as well, increasing production of ROS and oxidative stress‐induced neuronal death [433].

Senescent T cell appears to promote PD via compromising the blood–brain barrier (BBB). BBB stands for restricting the entry of peripheral immune cells into brain, protecting dopaminergic neurons from immune‐mediated damage. However, adhesion molecules (lymphocyte function‐associated antigen 1 and very late appearing antigen‐4), overexpressed in senescent T cells bind to vascular endothelial ligands (ICAM‐1 and VCAM‐1), disrupting the integrity of BBB to permit transmigration of peripheral T cells into the substantia nigra [434]. Once infiltrated within brain, perforin/ GZMB derived from those CD8⁺ T cells directly induce apoptosis of dopaminergic neurons and accelerating disease progression [435].