2.2.1 Oncogene-induced senescence

OIS is primarily caused by the activation of oncogenes, leading to a state of permanent growth arrest, which was first demonstrated in vivo in 1997.^34-36 Subsequently, the concept of OIS was expanded to multiple tumorigenesis models, including lung adenomas, lymphoma, prostate cancer, and melanocytic naevi.^{34, 37-39} OIS acts as a barrier to tumorigenesis by preventing the proliferation of cells that harbor oncogenic mutations. Notably, this process can be triggered by various oncogenes, such as RAS, BRAF, and MYC, each contributing to the complex landscape of cancer biology. For example, melanocytic nevi caused by oncogenic BRAF mutations keep senescent for decades, thus impeding their progression into melanoma.³⁹ PTEN inactivation induces growth arrest through the p53-dependent cellular senescence pathway in primary prostate epithelium.³⁷

The mechanisms underlying OIS have been clarified. The activation of an oncogene triggers the production of ROS, introducing DSBs and DDR, which subsequently initiates cellular senescence.⁴⁰ The activation of oncogenes results in DNA damage, which upregulates the expression of functional tumor suppressor proteins like p53 and p21. This upregulation induces a stable cell cycle arrest by inhibiting the downstream cyclin-dependent kinase (CDK)–cyclin complexes. Thereby, hyperphosphorylation of RB protein blocks S-phase entry and induces cellular senescence.^{35, 41, 42} Of note, only oncogene expression does not drive DDR in the absence of DNA replication; instead, OIS originates from DDR activation triggered by oncogene-induced DNA hyper-replication.⁴³

2.2.2 Therapy-induced senescence

TIS is considered a reaction in tumor cells to an array of anticancer therapies, including chemotherapy and radiotherapy, which introduce DSBs and activate DDR.⁴⁴

The mechanisms of TIS are complex and involve various signaling pathways and cellular responses. One of the key pathways implicated in TIS is the activation of the p53 and p16INK4a tumor suppressor pathways, which play critical roles in mediating cell cycle arrest and cellular senescence. Following DNA damage induced by chemotherapy or radiation, p53 is activated, leading to the transcription of genes that inhibit the cell cycle, thereby promoting senescence.⁴⁵ For instance, cisplatin triggers the p53 pathway to induce cellular senescence.⁴⁶ Additionally, an elevated level of ROS during therapy can further activate these pathways, exacerbating the senescence response.

Several studies have provided insights into the role of TIS in the context of specific cancer therapies. For instance, in the treatment of T-cell acute lymphoblastic leukemia with chemotherapy, the induction of senescence was found to suppress disease progression, suggesting that TIS might be beneficial in this context.⁴⁷ Research has demonstrated that targeting senescent cells can enhance the efficacy of cancer therapies. For example, the use of senolytic drugs, which selectively eliminate senescent cells, has shown promise in reducing the side effects associated with chemotherapy while improving overall treatment outcomes.^{48, 49} These findings underscore the importance of understanding TIS not only as a byproduct of cancer treatment but also as a potential target for therapeutic intervention.

3.1.1 Senescence-mediated growth arrest

Entry of cells into senescence is a natural barrier to tumorigenesis, essential for eliminating cells from proliferation, particularly those with a heightened risk of malignant transformation.²³ OIS operates as a cell-intrinsic mechanism that induces growth arrest of premalignant cells following activated oncogenes (such as N-RAS^G12V and BRAF^V600E), which could be further reinforced by CDK inhibitors (commonly known as p21 and p16^INK4a).^{34, 63, 64} Senescence also mediates tumor suppression in a cell-extrinsic fashion. SASP factors limit the propagation of precancerous cells or fully malignant cells in their vicinity through autocrine or paracrine manners.^{51, 65, 66} IL-1$$\alpha $$ triggers an autocrine inflammatory response via activating NF-κB signaling, driving the production of IL-6 and IL-8 that could reinforce growth arrest through increased ROS production and sustained DDR.^{67, 68} IL-1$$\alpha $$ also stimulates paracrine senescence in adjacent cancer cells to prompt tumor suppression.^{51, 65, 66} Moreover, some SASP factors could induce apoptosis or necrosis in neighboring cells. TNF-$$\alpha $$ triggers ROS-dependent apoptosis in cancer cells.⁶⁹ Conversely, IL-6 drives apoptosis in tumor-infiltrating T cells.⁷⁰

3.1.2 Senescence-mediated immune surveillance

Provided evidence suggests that SASP is an important contributor to cancer immunosurveillance. Explicitly, IL-6, IL-8, and CCL2 facilitate the recruitment of M₁-type macrophages, T helper 1 cells, and NK cells to the TME, driving the immune clearance of senescent cancer cells.^{54, 55} Both OIS and TIS could trigger SASP-mediated immunosurveillance against tumors.

Oncogene-induced senescent cells have the capacity to promote cancer immunosurveillance. For instance, N-RAS^G12V-driven senescent premalignant hepatocytes secrete SASP factors and are subject to immune elimination that depends on CD4⁺ T cell-mediated adaptive immunity.⁵⁵ When NRAS^G12V-induced senescent human fibroblasts are injected into mice, a global remodeling of their super-enhancer landscape occurs. Super-enhancers are particularly powerful enhancer regions in the genome that significantly boost gene expression. This reconfiguration, particularly through the recruitment of chromatin reader bromodomain-containing protein 4 (BRD4) to SASP-gene-adjacent super-enhancer loci within the genome, plays a pivotal role in immunosurveillance. Given that BRD4 is essential for the SASP factors and downstream paracrine signaling, BRD4 inhibition would destroy the immune clearance of premalignant OIS cells.⁷¹ Importantly, except for participating in the induction of cellular senescence, the tumor suppressor p53–p21 axis also modifies the interplay between senescence and immunity.^{72, 73} As a tumor suppressor, p53 is frequently mutated in various tumors, manifesting the “proliferating state” tumor with the worst prognosis.^{74, 75} P53-restoration in mouse models contributes to tumor suppression with senescent phenotype, revealing that p53-mediated senescence paves the way for immune clearance triggered by the SASP factors, such as CSF1, CCL2, CXCL1, and IL-15.^{72, 73} By coordinating with NF-κB signaling, p53 regulates SASP secretion that triggers macrophage activation contributing to a tumor-suppressive microenvironment.^{76, 77} p21-induced senescence also activates immunosurveillance with a particular secretory phenotype involving CXCL14 and IGFBP3 that prompt the recruitment of macrophages and lymphocytes into the TME.⁷⁸ Moreover, oncogenic RAS was found to drive immune surveillance against tumorigenesis via p21-dependent senescence.⁷⁸

TIS also contributes to tumor suppression via immune cell-mediated clearance. In the KRAS-mutant lung cancer model, combinatory inhibitors of MEK and CDK4/6 suppress cancer cell proliferation while inducing RB-mediated cellular senescence and activating immunomodulatory SASP. SASP components, TNF-$$\alpha $$ and ICAM1, are needed for NK cell surveillance leading to tumor regression.⁷⁹ Similarly, SASP factors lead to vascular remodeling that facilitates cytotoxic CD8⁺ T cell accumulation following dual inhibitors of MEK and CDK4/6 in KRAS-mutant pancreatic ductal adenocarcinoma.⁶⁸ Furthermore, CDK4/6 inhibitor-mediated senescence induces potent immunosurveillance via regulatory T (Treg) cell suppression and CD8⁺ T cell enhancement.^{67, 80}

3.2.1 SASP-driven carcinogenesis

Several studies have demonstrated protumorigenic effects of senescent cells mediated by individual SASP factors. Matrix metalloproteinases (MMPs) can mediate the degradation of extracellular matrix (ECM) components associated with growth factor release while prompting tumor-driven angiogenesis, operating as a switch favoring tumor growth and invasion.^{60, 83, 84} Accordingly, GDF-15, FGF1, and IGFBP3 also facilitate tumor dissemination to secondary sites.⁸⁵ Moreover, hepatocyte growth factor (HGF) is reported to collaborate with MMPs to further promote tumor progression.⁶⁰ SASP factors IL-6 and IL-8 are important drivers of cancer proliferation via creating a chronic inflammatory microenvironment, facilitating ECM cleavage by MMPs, and driving epithelial-to-mesenchymal transition (EMT), thereby collectively contributing to tumor invasiveness.^{56, 86, 87} Furthermore, VEGF, IL-6, and CXCL5 stimulate angiogenesis to support tumor metastasis. Overall, different SASP factors mediate tumor growth, proliferation, invasion, and metastasis via multiple mechanisms.

3.2.2 SASP-mediated immune evasion

As a negative effect, SASP factors are critical mediators of TIME suppression leading to immune evasion. For example, IL-6 is proven to recruit myeloid-derived suppressor cells (MDSCs) to the tumor site, which are reported to block immune surveillance by suppressing CD8⁺ T cells and NK cells.^{54, 88} Meanwhile, MDSCs are found to inhibit IL-1$$\alpha $$ signaling and consequently hinder the induction of senescence in tumor cells.⁶¹ IL-6 and IL-8 lead to the upregulation of HLA-E that interacts with Natural Killer Group 2 Member A (NKG2A), which blunts the effector activity of NK cells and CD8⁺ T cells.⁸⁹

3.2.3 Senescence-associated stemness

Pioneering studies pointed out that TIS can induce stemness properties in malignant cells, driving them to escape proliferation arrest and re-enter the cell cycle.^{90, 91} For instance, chemotherapy-induced senescence in transgenic mice with B-cell lymphomas where WNT signaling is activated ultimately showed distinct stem-cell markers.⁹⁰ Oncogene-induced senescent cells also acquire stemness characteristics driving tumor aggressiveness.⁹² In an OIS mouse model of breast cancer, RANK-induced senescence is required for RANK-triggered stemness; despite initially displaying a delayed tumor onset, tumor progression and aggressiveness are observed in the long term.⁹³ Moreover, overexpression of CIP1 enables p53-null tumor cells to gain enhanced stem cell properties following a transient senescent state.⁹⁴ Growing evidence indicates that senescence-associated stemness is triggered by WNT signaling activation depending on SASP factors rather than WNT ligands.⁹⁰ Therefore, senescent tumor cells that regain proliferative capacity exhibit WNT-dependent growth, a process that is particularly common in recurrent tumors.⁹⁰

4.1.1 Abnormal molecular mechanisms underlie immunosenescence

Recent studies have underscored the significance of immunosenescence in tumor development.⁹⁹ Many factors within the TME possess the capability to trigger immunosenescence. As the most powerful immune cells in eradicating tumor cells, T-cell senescence has attracted the most interest. Given the potential of T cells to be continuously activated by antigens and affected by inflammatory cytokines, it is plausible that the TME may be the source of senescent T cells.¹⁰⁰ Signaling pathways involved in T-cell senescence are displayed in Figure 2.

Tumor-derived cyclic adenosine monophosphate (cAMP), driven by hypoxia, has the potential to inhibit tumor-specific effector T cells by inducing DNA damage and senescence.^{101, 102} Glucose competition between T cells and Treg cells could initiate a series of events, including ATM-related DNA damage, activation of ERK1/2 and P38 pathways, and interaction with STAT1/3 accompanied by upregulation of p21–p16–p53, eventually leading to T cells withdraw from the cell cycle.^{103, 104} Furthermore, the activation of AMPK pathways due to glucose deficiency could downregulate the telomerase reverse transcriptase gene binding to TAB1, followed by autophosphorylation of P38. The activation of P38 leads to the downregulation of TERT, thereby inducing DNA damage.¹⁰⁵ The downregulation of T cell receptor (TCR) signaling could activate the P38 pathway and hinder the PI3K/AKT/mTOR signaling pathway, which in turn inactivates autophagy and accelerates ROS production in senescent T cells. Of note, high levels of ROS contribute to inflammaging and immunosenescence.¹⁰⁶

4.1.2 Cellular level mechanisms in immunosenescence

Immune system senescence, also referred to immunosenescence, reflects the senescence process observed in diverse subpopulations of immune cells (Figure 3). Innate immune senescence is characterized by a reduced ability to process and present antigens, leading to a diminished responsiveness to stimuli. Meanwhile, adaptive immune senescence encompasses a decline in the diversity of the TCR repertoire and an impaired formation of immunological memory. Therefore, immune cell senescence dramatically impacts the efficacy of antitumor responses.

5.1.1 Senolytic therapies

The persistence of therapy-induced senescent cells can lead to adverse outcomes, including an increased incidence of secondary tumors, the recurrence of more aggressive cells, and complications related to treatment.^{204, 205} This argues for a strategy to eliminate senescent tumor cells by senolytic agents, which can selectively induce apoptosis in senescent cells but spare other nonsenescent cells.²⁰⁶ Notably, a key feature of senescent cells is a change in chromatin structure that alters gene expression. These changes can affect crucial processes like apoptosis regulation, creating new vulnerabilities in senescent cells that can be specifically targeted by senolytic drugs.⁸ Therefore, applying senolytic agents in combination with or following senescence-inducing therapy may generate unexpected clinical benefits.

BCL-X_L-modulating drugs have shown the greatest promise in obliterating senescent tumor cells by targeting antiapoptotic BCL-2 pathways.²⁰⁷ For example, navitoclax (ABT-263), selectively targeting BCL-2, BCL-X_L, and BCL-W proteins, could effectively eliminate senescent cells by inducing programmed cell death.^{172, 208} As reported, navitoclax could kill breast and ovarian cancer cells after application of PARP inhibitors in aged animal models and in vitro culture.²⁰⁹ Exposure to navitoclax following etoposide or doxorubicin showed strong effectiveness in slowing tumor progression.^{204, 210} Consistently, in a phase-II study, the combination of navitoclax and rituximab showed higher effectiveness than rituximab monotherapy accompanied by well tolerance in patients with chronic lymphocytic leukemia.²¹¹ Other BCL-2 family inhibitors, such as ABT-737, A1155463, A1331852, and EF24, have been developed as promising senolytic drugs.^{165, 173, 174}

However, the side effects of navitoclax primarily arise from its off-target effects on hematological cells, including thrombocytopenia and neutropenia, which significantly limits its broader use.²¹² Several attempts have been made to address this challenge. Galacto-conjugated navitoclax prodrug (nav-gal) has been designed to improve delivery accuracy.²¹³ This prodrug is processed by SA-$$\beta $$-gal in senescent cells, thereby specifically killing senescent cells and limiting off-target effects in normal cells. The combination of nav-gal and cisplatin has demonstrated efficient clearance of lung cancer cells. Moreover, nanocarrier-encapsulated doxorubicin is another strategy to induce specific cytotoxicity in senescent cells, which has been validated in various studies.^{214, 215} Alternatively, the proteolysis-targeting chimera (PROTAC) drug PZ15227 is reported to limit platelet toxicity of navitoclax by hijacking the cereblon (CRBN) E3 ligase for BCL-XL degradation. Given the minimal expression of CRBN in platelets, PZ15227 could eliminate senescent tumor cells without inducing severe thrombocytopenia.²¹⁶

The mTOR inhibitor AZD8055 also showed potent senolytic effects against cancer cells with a senescent phenotype induced by DNA-replication kinase CDC7 inhibitors.^{177, 217} Similarly, the dual treatment of docetaxel and mTOR inhibitor temsirolimus results in a remarkable reduction of tumor growth in prostate and breast cancer animal models.¹⁷⁵ Sertraline, a selective serotonin reuptake inhibitor, is frequently used to treat psychological disorders. By suppressing mTOR signaling, sertraline was identified as a senolytic agent to kill hepatocellular carcinoma cells that have been rendered senescent through CDC7 inhibition.¹⁷⁷ Nevertheless, the senolytic effect of mTOR inhibitors still requires further exploitation and tests in more tumor types.

Moreover, pharmacological interventions that regulate p53 activity have senolytic effects. Given that FOXO4 can retain p53 in the nucleus, the FOXO4-DRI peptide could disrupt the p53–FOXO4 interaction, leading to p53-mediated apoptosis in senescent cells.¹⁷⁸ Despite the FOXO4-DRI peptide displaying well tolerance in mice, its clinical use remains lacking enough evaluation. Notably, compounds that modulate p53 activity are only effective in tumors with p53-wild-type, significantly limiting their therapeutic scope.^{218, 219}

The BET family protein degrader ARV825 provokes senolysis via attenuating nonhomologous end-joining (NHEJ) DNA DSB repair and upregulating autophagic gene expression.^{180, 181} In mice undergoing doxorubicin-induced senescence, ARV825 treatment effectively eliminates senescent hepatic stellate cells, leading to delayed development of liver cancer.¹⁸¹

Cardiac glycosides, including widely used digoxin, could selectively kill senescent cells in multiple cancer models.^{182, 183} Cardiac glycosides that inhibit the Na⁺/K⁺ ATPase pump could drive cell depolarization and acidification, ultimately leading to senescent cell death.¹⁸³ Besides, these compounds can activate the proapoptotic BCL-2 family protein NOXA to trigger intrinsic apoptosis.¹⁸²

Although numerous small molecules have demonstrated encouraging efficacy against cellular senescence, these agents exhibit a deficiency in sensitivity and may result in significant side effects. During the past decades, immunotherapy has gained breakthroughs in treating patients with advanced or drug-resistant malignancies.²²⁰ Immunotherapy could mediate endogenous senolysis to prompt senescent cell eradication. Antiprogrammed cell death protein 1 (PD1) blockades can elicit senolytic effects following TIS. The combination of trametinib and palbociclib induced senescence in a mouse model of KRAS-mutant PDAC, accompanied by increased tumor vascularization and endothelial cell activation, ultimately improving the efficacy of anti-PD1 therapy.⁶⁸ Notably, chimeric antigen receptor T-cell (CAR-T) therapy directed at senescence-specific surface antigens represent a viable treatment strategy, ablating senescent cells in vitro and in vivo. The urokinase-type plasminogen activator receptor (uPAR) as a cell membrane protein is significantly upregulated following senescence induction. uPAR-specific CAR-T cells extended the tumor growth in mice with lung adenocarcinoma previously treated with a combination of MEK and CDK4/6 inhibitors.¹⁸⁴ Other cell surface proteins might be potential targets for cancer immunotherapy, such as DEP1, DPP4, NKG2D, ICAM1, NOTCH1, and NOTCH3.^221-226 Given these promising findings, more clinical studies are required to explore whether the benefits of senolytics outweigh their potential side events and to determine the optimal dose schedule. We have summarized relevant clinical trials in Table 3.

5.1.2 Senomorphic therapies

Senomorphic SASP inhibitors could function as efficacious alternatives to senolytics. Metformin, a widely recognized medication, could prevent the translocation of NF-κB pathway components to the nucleus and inhibit their subsequent activation, thus leading to a decrease in the expression of diverse SASP factors. This underlying mechanism could plausibly elucidate the antiaging and antitumor effects of metformin in both murine models and diabetic patients.¹⁸⁷ mTOR inhibitor, such as rapamycin, could reduce NF-λB activity, suppress inflammatory SASPs at the translational level, and constrain the tumor-prompting effect of senescent bystander fibroblasts.^{52, 188} Inhibitors of the JAK signaling pathway in aged mice, such as ruxolitinib, could reduce inflammation and ameliorate frailty by suppressing SASP factors. Currently, several JAK inhibitors are being tested in clinical trials including patients with acute myeloid leukemia or lymphomas.^{189, 227}

Antibodies against SASP factors also have senormophic effects. For example, siltuximab, an antibody inhibiting IL-6 approved for the treatment of multicentric Castleman disease, has shown activities in diverse oncological contexts.¹⁹⁰ Canakinumab, an anti-IL-1$$\beta $$ monoclonal antibody approved for pyrexia-featured inflammatory syndromes, is evaluated in clinical trials involving patients diagnosed with non-small cell lung cancer.¹⁹¹

Some hormones may hold the potential to limit detrimental SASP factors. As a novel SASP suppressor, melatonin suppresses SASP gene expression via modulating poly (ADP-ribose) polymerase 1 (PARP1).¹⁹⁵ Other hormones, such as androgens, estrogens, and glucocorticoids, also possess the ability to regulate the secretion of proinflammatory SASP factors.^192-194 Notably, further investigation is necessary to thoroughly examine the anticancer efficacy of these senescence-dependent, SASP-suppressing agents in suitable model systems and clinical trials.

5.2.1 Targeting metabolic reprogramming

Senescence is the culmination of the common effect of internal and external stimuli. Metabolic stresses, such as glucose deprivation, mitochondrial dysfunction, imbalanced lipid metabolism, hypoxia, and tumor metabolites, are critical to regulating T-cell senescence.

Excessive glucose consumption by tumor cells and Treg cells leads to a deficiency of glucose supply for effector T cells, ultimately triggering responder T-cell senescence.^{103, 104} The addition of glucose could avoid effector T-cell senescence induced by glucose competition.²²⁸ Molecular inhibitors targeting mTORC1 or p53 hold the potential to rejuvenate mitochondrial fitness, thereby preventing T-cell senescence.^{125, 198, 199} There is a mechanistic link between T-cell senescence and lipid metabolic reprogramming in the TME. Findings identify that senescent T cells display hyperactive glycolysis but imbalanced lipid metabolism.²²⁹ The latter changes the expression of lipid metabolic enzymes, which, in turn, affects the species and amount of lipid droplets accumulated in T cells.²³⁰ Specifically, over-expression of group IVA phospholipase A₂ in malignant and Treg cells could induce T-cell senescence. Inhibition of group IVA phospholipase A₂ activity is reported to alter T-cell lipid metabolism, prevent T-cell senescence in vitro, and boost antitumor immunity in in vivo models of melanoma and breast cancer.²³¹ The hypoxic TME may also induce T-cell senescence.¹⁹⁷ Consequently, hyperbaric oxygen therapy (HBOT) is confirmed to exert senolytic effects by increasing telomere length and reducing immunosenescence in peripheral blood cells.²⁰⁰ Alternatively, downregulating tumor-derived cAMP in the TME by cAMP pharmacological inhibitors or activating the TLR8 signaling pathway by synthetic poly-G3 and natural TLR8 ligand can prevent the induction of senescence in responder T cells while reversing the suppressive effect of senescent T cells.^{103, 196, 197} Ample evidence supports that TLRs directly participate in metabolic reprogramming to interfere with malignant behavior in several solid tumors.^{197, 232-234} Therefore, TLR agonists might be effective agents or adjuvants for cancer immunotherapy.

5.2.2 Manipulating signaling pathways

The importance of the mitogen-activated protein kinase (MAPK) signaling pathway in mediating T-cell senescence has recently been exploited.^{105, 235} Molecular inhibitors targeting ERK, p38, and STAT signaling pathways may reverse or prevent T-cell senescence.^{90, 236-239} Meanwhile, the MAPK signaling pathway plays a vital role in T-cell activation and effector functions.^{235, 240} Hence, it is urgently needed to select the optimal MAPK inhibitors for restraining senescence induction without concession in self-renewal and cytotoxicity of tumor-reactive T cells.

Selective MAPK inhibitors already available for clinical trials are preferential candidates for treating melanoma patients, with improved T-cell recognition of tumor cells without affecting lymphocyte function.^{201, 202} Moreover, the CRISPR/Cas9 gene editing technology could modulate the MAPK signaling pathway more stably and precisely. Similarly, selective ATM inhibitors have been tested in clinical trials for cancer patients.^{203, 241} Furthermore, the role of sestrins in regulating immunosenescence is multifaceted. Increasing evidence supports that sestrins coordinate ERK, JNK, and P38 phosphorylation in CD4⁺ T cells. Sestrin knockdown restores T-cell responsiveness and cytokine production in vivo during ageing.²⁴² Notably, sestrins may induce the programming of senescent CD8⁺ T cells to obtain NK-like killing ability but lose TCR-dependent signaling activity.¹¹⁶ More mechanisms of sestrin-mediated T-cell senescence need further elaboration.