CAR-T therapy in non-small cell lung cancer: Clinical prospects, potential, and strategies for cardiotoxicity management

Highlights

• •CAR-T Therapy Shows Promising Prospects for NSCLC Treatment: The application targets and potential targets of CAR-T in NSCLC, as well as the combination of CAR-T with immunotherapy (PD-L1 inhibitors) or chemotherapy, can achieve better efficacy and overcome resistance or immune evasion.
• •Molecular Structure Research of CAR-T, Progress from 1st to 5th Generation: More structural designs are of research value, making CAR-T therapy more precise and cost-effective.
• •Limitations of CAR-T in NSCLC Treatment: High tumor heterogeneity and the lack of effective targets lead to immune evasion and off-target effects.
• •Exploration of Adverse Reactions in CAR-T Treatment for Solid Tumor Patients: Focus on CRS and cardiotoxicity, while also discussing countermeasures.
• •Recommendation for CAR-T Therapy in Solid Tumor Patients: Developing a CAR-T treatment risk model through a cardiotoxicity scoring system.

Abstract

Lung cancer ranks first among all malignancies in incidence, with current treatment strategies including surgery, chemotherapy, immunotherapy, and targeted therapy. Despite these advances, drug resistance in advanced non-small cell lung cancer (NSCLC) remains a major obstacle and innovative therapeutic approaches are imperative to address it. Chimeric antigen receptor T-cell (CAR-T) therapy has shown impressive and long-lasting results in blood cancers, but its success in solid tumors such as lung cancer remains limited. This review summarizes recent advances and future directions of CAR-T therapy in NSCLC, focusing on major therapeutic targets such as EGFR, MSLN, PD-L1, MUC1, CEA, and ROR1, as well as on the efficacy and potential of combining CAR-T therapy with other treatment modalities. Additionally, we discuss adverse events in NSCLC patients undergoing CAR-T therapy, emphasizing cytokine release syndrome (CRS) and cardiovascular complications—their incidence, pathophysiology, interrelation, and management strategies.

Graphical abstract

Introduction

Lung cancer continues to be the most prevalent and deadliest form of cancer across the globe, responsible for a significant proportion of cancer-related deaths each year. Clinically, it is classified into two primary categories—small cell lung cancer (SCLC) and non-small cell lung cancer (NSCLC)—based on their distinct cellular characteristics and biological behavior [[1], [2], [3], [4]]. Current therapeutic approaches encompass surgery, radiotherapy, chemotherapy, targeted therapy, and immunotherapy [5,6]. For patients diagnosed with early-stage and localized NSCLC, surgical intervention continues to be the primary and most effective treatment choice. In recent years, the adoption of minimally invasive surgical techniques has transformed clinical practice, offering reduced trauma and faster recovery [7]. Approaches such as video-assisted thoracoscopic surgery (VATS), robotic-assisted systems like the Da Vinci platform, and percutaneous ablation procedures including radiofrequency (RFA) and microwave ablation (MWA) have become increasingly common in routine management [[8], [9], [10], [11]]. In cases where tumors exhibit PD-L1 expression levels of 1 % or higher and lack targetable oncogenic mutations, immune checkpoint inhibitors (ICIs)—notably pembrolizumab and atezolizumab—are frequently administered in combination with chemotherapy as the current first-line therapeutic standard.t [12]. With the advent of next-generation sequencing (NGS), actionable oncogenic alterations—including EGFR mutations, ALK rearrangements, and ROS1 fusions—can be identified, enabling personalized targeted therapies that have markedly improved survival outcomes in advanced disease [[13], [14], [15], [16]]. Despite these advancements, acquired resistance to chemotherapy, immunotherapy, and targeted therapy continues to represent a major therapeutic challenge in advanced NSCLC.

Unlike traditional immunotherapies, CAR-T therapy involves the genetic modification of a patient’s own T cells to express specialized chimeric antigen receptors (CAR) that recognize specific tumor-associated antigens, enabling these reprogrammed cells to precisely target and destroy malignant cells within the body [[17], [18], [19]]. Common CAR-T targets in NSCLC include EGFR, PD-L1, MUC1, CEA, and ROR1. Rational optimization of receptor design can enhance therapeutic potency; for instance, EGFR-targeted CAR-T cells co-expressing CXCR5 or B7-H3-targeted CAR-T cells have demonstrated promising antitumor activity in preclinical studies [20,21].

While CAR-T has achieved breakthrough success in hematologic malignancies, its application and efficacy in solid tumors—especially lung cancer—remain limited and require substantial clinical validation. Major obstacles include the scarcity of tumor-specific antigens and the occurrence of severe toxicities, notably cardiotoxicity and cytokine-mediated inflammation. Consequently, achieving an optimal balance between therapeutic efficacy and safety remains a central challenge for CAR-T–based immunotherapy.

Current progress in CAR-T

The concept of CAR-T therapy emerged nearly three decades ago, when the first chimeric antigen receptors were designed to recognize specific tumor antigens [22]. Unlike native T-cell receptors, CARs recognize antigens independently of major histocompatibility complex (MHC) presentation, thereby circumventing tumor immune evasion through MHC downregulation [23]. Structurally, a CAR consists of three main components: an extracellular antigen-recognition domain, a transmembrane segment, and one or more intracellular signaling regions [24,25]. The extracellular portion, typically a single-chain variable fragment (scFv), is engineered to bind selectively to tumor-associated antigens by connecting the variable heavy and light antibody chains through a flexible peptide linker [26]. The hydrophobic α-helical transmembrane region stabilizes signaling across the cell membrane, while intracellular signaling domains initiate T-cell activation upon antigen engagement.The process of CAR-T cell preparation involves collecting peripheral blood mononuclear cells and isolating and enriching T cells [26,27]. The T cells are then activated to enhance their function and proliferative capacity. Next, the CAR gene is transduced into the T cells, enabling them to recognize tumor antigens. After transduction, the CAR-T cells are expanded in vitro, cryopreserved, and prepared for transportation. Finally, after thorough quality control testing, the CAR-T cells are infused back into the patient. Throughout the process, quality control is essential, and the entire procedure typically takes 10 to 25 days.

First-generation CAR-T cells incorporated only the CD3ζ signaling domain, whereas second-generation constructs introduced additional costimulatory molecules such as CD28 or 4–1BB to enhance activation and persistence. Third-generation CAR-T cells—typically introduced via lentiviral vectors—combine multiple costimulatory domains (e.g., CD28 + 4–1BB or CD28 + OX40). Fourth-generation, or “armored,” CAR-T cells are paired with constitutive chemokine/cytokine expression (e.g., IL-12) to recruit other effector cells (DCs, macrophages, NK cells) for cooperative tumor killing. Fifth-generation, or “universal,” CAR-T cells are engineered for allogeneic use by disrupting endogenous T-cell receptors to prevent host rejection. These next-generation constructs integrate IL-2 receptor β-chain (IL-2Rβ) and STAT3/5 binding domains into the second-generation backbone to improve proliferation and persistence. Fig. 1A illustrates the general process of CAR-T cell preparation and infusion, while Fig. 1B outlines the evolution of CAR-T design across five generations.

Fig. 1.

Preparation process and structural evolution of CAR-T cells. (A) Simplified schematic of the CAR-T cell manufacturing and infusion process. (B) Molecular structures of the five generations of CAR-T cells.

CAR-T cell therapy is more widely applied in hematologic malignancies, with significant success in the treatment of acute lymphoblastic leukemia (ALL), particularly in children and young adults. Leukemia cells often express the specific antigen CD19, so CD19-targeted CAR-T cells have shown extremely high therapeutic response rates in treating CD19-positive ALL [28]. Due to relapsed/refractory Acute Myeloid Leukemia (R/R AML) being a major cause of treatment failure in adult leukemia, patients who are ineligible for allogeneic hematopoietic stem cell transplantation (alloHSCT) or chemotherapy have an extremely low five-year overall survival rate. The heterogeneity of AML presents challenges in treatment, with various genetic mutations such as TP53, EZH2, and DNMT3A playing a role. Chromosomal translocations are also one of the main factors affecting the prognosis of AML. Recently, CAR-T cell therapies targeting AML markers such as CD33 and CD123 have entered clinical trials. Although there are still challenges, these studies demonstrate the potential for treating AML [29].

To date, most clinically approved CAR-T products remain based on second-generation designs [30]. Further advances must improve target specificity, reduce toxicity, lower manufacturing costs, and be supported by robust clinical data before broad application in lung cancer.

CAR-T applications in lung cancer

Major targets for CAR-T in lung cancer

The key to the efficacy of solid tumor therapies lies in the presence of specific antigens. The target antigens need to be predominantly expressed in NSCLC, rather than in normal tissues. Therefore, exploring critical targets is essential for treating NSCLC. This review highlights several of the most studied targets for CAR-T therapy in lung cancer.

Epidermal growth factor receptor (EGFR) remains important frequent oncogenic drivers in NSCLC, particularly among Asian, non-smoking female patients. Iterative EGFR-TKIs and combination strategies have significantly extended survival; however, acquired resistance is nearly inevitable. Barbie et al. had concluded that TROP2-directed CAR-T therapy can sustain the therapeutic benefits of osimertinib and eradicate minimal residual disease in NSCLC, thereby prolonging the time to relapse [31,32]. CXCR5, a chemokine receptor highly expressed in NSCLC, interacts with its ligand CXCL13 to regulate immune cell chemotaxis. Engineering T cells into CXCR5-EGFR-CAR-T cells enhances both chemotactic ability and antitumor efficacy in NSCLC [33]. Early-phase clinical trials are currently investigating CXCR5-CAR-T therapy in advanced NSCLC (NCT05060796) and EGFR-TGFβR-KO CAR-T constructs (NCT04153799), although conclusive outcomes are not achieved so far. In another EGFR-CAR-T study (NCT03182816), nine patients achieved a median progression-free survival (PFS) of 7.13 months and a median overall survival (mOS) of 15.63 months, suggesting moderate clinical benefit. ICIs, such as pembrolizumab or atezolizumab, have become standard treatments for advanced NSCLC, significantly prolonging survival in specific patient populations, particularly those with high PD-L1 expression, with a median progression-free survival (mPFS) of 6–8 months and a median overall survival (mOS) of 15–20 months (KEYNOTE-042). However, resistance to these therapies, especially in later-line treatments, still necessitates novel therapies, such as CAR-T.

Mucin-1 (MUC1) is a heavily glycosylated transmembrane protein found on the apical surface of epithelial cells in many organs. In lung cancer, MUC1 is often overexpressed and contributes to tumor recurrence and metastasis via the MAPK–PI3K/Akt signaling pathway, making it a promising target for immune-based and precision therapies. Zhou et al. reported that the truncated glycoform TnMUC1 functions as a tumor-specific antigen, although its expression in lung cancer is typically low. Interestingly, EGFR-TKI treatment reduces MUC1 glycosylation while increasing TnMUC1 expression, and combining CAR-T cells targeting TnMUC1 with EGFR-TKIs significantly enhances therapeutic efficacy in EGFR-mutant NSCLC [34].

Mesothelin (MSLN) is a differentiation antigen on mesothelial cells; it shows low expression in normal tissues but is upregulated in several solid tumors, including lung cancer. Abnormal MSLN expression activates NF-κB, MAPK, and PI3K pathways, thereby promoting tumor proliferation, invasion, and survival. In a phase I trial, Adusumilli et al. evaluated MSLN-targeted CAR-T with pembrolizumab in patients with malignant pleural disease, including metastatic lung cancer and malignant pleural mesothelioma. Among 18 patients treated with the combination, the median overall survival reached 23.9 months, with six grade3 adverse events—supporting the potential clinical utility of MSLN—CAR-T therapy [35]. Ye and colleagues further demonstrated that MSLN—CAR-T cells effectively suppressed tumor growth in mice, though substantial cytotoxicity was observed [36]. However, Haas et al. reported severe lung injury in two patients following MSLN—CAR-T infusion, likely due to off-tumor recognition of upregulated MSLN in normal lung tissue, resulting in diffuse pulmonary damage [37]. These findings underscore the need for additional trials and long-term safety evaluation to determine the therapeutic window of MSLN-targeted CAR-T cells.

Programmed cell death protein1 (PD-1), expressed on activated immune cells, modulates immune tolerance, while its ligand PD-L1, expressed on tumor cells, facilitates immune escape and metastatic progression [38,39]. “In preclinical research, Liu et al. reported that CAR-T cells engineered to target PD-L1 effectively eliminated NSCLC cells, and their activity was further strengthened when combined with radiotherapy [40]. Similarly, Ma and co-workers constructed PD-L1-specific chimeric switch receptor CAR-T cells that exhibited strong antitumor effects against pleural and peritoneal metastases in murine models, findings that supported the launch of a phase I clinical trial (NCT04684459) to evaluate their safety and efficacy in humans [41].

In addition to these well-established targets, other tumor-associated antigens—including carcinoembryonic antigen (CEA), HER2, and receptor tyrosine kinase-like orphan receptor 1 (ROR1)—have emerged as promising candidates for CAR-T therapy in lung cancer, with the potential to overcome treatment resistance. Moreover, several novel targets such as the CD98 heavy chain, tissue factor (TF), protein tyrosine kinase 7 (PTK7), EphA2, and TROP2 have shown encouraging preclinical results, warranting further investigation [31].”

Limitations and challenges of CAR-T in lung cancer

Despite its success in hematologic malignancies, CAR-T cell therapy faces substantial limitations in the treatment of solid tumors. One of the major challenges is the scarcity of tumor-specific antigens (TSAs), which restricts CAR-T cells from selectively targeting malignant cells without affecting normal tissues. In contrast, tumor-associated antigens (TAAs)—which are shared between tumors and normal cells—pose a risk of off-target cytotoxicity, leading to tissue damage and diminished therapeutic efficacy [42]. In NSCLC, antigenic heterogeneity remains a critical barrier. Antigen expression can vary widely not only between patients but also among distinct lesions within the same individual, including primary tumors, metastases, and recurrent nodules. Furthermore, intratumoral spatial heterogeneity contributes additional complexity, as different regions of a single tumor may express divergent antigenic profiles [43,44]. This variability contributes to immune escape, resistance, and tumor relapse, posing a major challenge for consistent CAR-T efficacy. Physical barriers also impede CAR-T infiltration in solid tumors. The encapsulating tumor stroma forms a dense matrix, while irregular tumor vasculature limits effective CAR-T trafficking. Additionally, the immunosuppressive tumor microenvironment (TME) recruits fibroblasts and deposits extracellular matrix components that hinder T cell penetration [45]. Chemokines play a crucial role in directing CAR-T migration; for instance, CXCR1 and CXCR5 expression enhances T -cell homing and cytotoxicity [[46], [47], [48]]. CAR-T cells also modulate the immune milieu by recruiting endogenous NK cells, dendritic cells, and native T cells. Lanitis et al. reported that inhibition of vascular endothelial growth factor (VEGF) promotes vascular remodeling and improves T cell infiltration in solid tumors, while VEGF-A upregulation attenuates CAR-T efficacy [49,50].

T cell activation can be suppressed by regulatory mechanisms in the TME, including immunosuppressive chemokines and regulatory T cells [51]. Chronic antigen stimulation or hypoxia can lead to CAR-T exhaustion. Persistent PD-1/PD-L1 signaling is known to suppress cytotoxicity; thus, PD-1 knockdown or blockade enhances CAR-T–mediated killing of NSCLC cells [52]. Good et al. demonstrated that sustained antigen stimulation drives CAR-T exhaustion, which can be reversed by modulating transcription factors such as ID3 and SOX4, thereby restoring antitumor activity [53]. These findings suggest that manipulating transcriptional programs, augmenting chemokine expression, or combining CAR-T therapy with immune checkpoint inhibitors may collectively enhance therapeutic efficacy and mitigate tumor immune escape.

TME plays a significant role in hindering the effectiveness of CAR-T cell therapy. Physical barriers, such as abnormal tumor vasculature and a dense extracellular matrix (ECM), limit the infiltration of CAR-T cells. Additionally, the TME is populated with immunosuppressive cells, including regulatory T cells (Tregs), myeloid-derived suppressor cells (MDSCs), and tumor-associated macrophages (TAMs), which outnumber CAR-T cells and actively suppress their function. Furthermore, tumor cells inhibit T cell function by consuming essential nutrients, such as l-arginine, and producing metabolic byproducts, like lactate [54]. The hypoxic and nutrient-deprived conditions within the TME exacerbate this issue, leading to CAR-T cell exhaustion and treatment failure.

Overcoming challenges and resistance mechanisms in CAR-T therapy

To overcome the limitations described above, advancements in CAR structure and multi-targeting strategies are being actively explored. Dual CAR-T systems utilize Boolean logic gates to activate only upon recognition of two distinct TAAs, thereby minimizing off-target effects. Hybrid CAR-T approaches involve the sequential or simultaneous targeting of multiple antigens. Tandem CARs (TanCARs) integrate recognition of two antigens within a single CAR molecule, activating downstream signaling through one intracellular domain. Fifth-generation or "universal" CAR-T cells leverage allogeneic transplantation to broaden antigen coverage, enable large-scale manufacturing, and reduce cost—although these remain in early-phase testing.

Novel switchable designs, such as ON-switch CAR-T cells, permit reversible activation controlled by small molecules, enhancing safety and precision. Wu et al. utilized rapamycin-induced dimerization to finely control CAR-T activity in the presence of both antigen binding and small molecule induction [55]. Traditionally, viral vectors have long been employed for CAR transduction, their transduction efficiency and long-term stability are limited. Although CRISPR/Cas9-mediated PD-1 knockout enhances CAR-T function, it may compromise viability. Recent technologies now enable high-efficiency CAR transduction using DNA-targeted lipid nanoparticles [56].

Combination therapies hold considerable promise for augmenting the antitumor efficacy of CAR-T cells in lung cancer. Chemotherapy has been shown to promote macrophage-mediated chemokine secretion and enhance CAR-T trafficking through exosome-mediated delivery mechanisms. Gao et al. demonstrated that docetaxel upregulates CXCL11 and HMGB1 expression, thereby improving the efficacy of HER2-targeting CAR-T cells [57]. Similarly, combining CAR-T therapy with immune checkpoint inhibitors (ICIs) can significantly improve therapeutic outcomes in solid tumors. A recent study revealed that engineering CAR-T cells to co-express αPD-L1 and IL-12 not only increased tumor infiltration but also reduced systemic toxicity [58]. Radiotherapy represents another effective adjunct that can convert immunologically "cold" tumors into "hot" ones, breaking immune barriers. Murty et al. demonstrated that in glioblastoma-bearing mice, GD2-CAR-T cells only achieved complete responses when administered following local or whole-body irradiation—likely due to improved T cell infiltration into the tumor microenvironment [59]. In lung cancer, minimally invasive interventions such as microwave ablation (MWA) are commonly used. By thermally disrupting the tumor structure, MWA facilitates CAR-T cell penetration. Cao et al. reported that MWA significantly enhanced the antitumor efficacy of AXL-targeting CAR-T cells against NSCLC by increasing mitochondrial metabolism, promoting persistence of infused cells, and maintaining a favorable safety profile [[60], [61], [62], [63]].

Tumor metabolism also plays a pivotal role in modulating CAR-T function. Metformin, a common antidiabetic drug, has been shown to modulate glucose metabolism, increase CD8+ T cell infiltration, and decrease IL-2 and TNF-γ secretion, thereby enhancing antitumor immunity [64]. Emerging evidence further suggests that metformin enhances CAR-T metabolism by upregulating PGC-1α and inhibiting STAT5 phosphorylation. This not only boosts antitumor memory responses but also mitigates T cell exhaustion [65].

Adverse events and cardiovascular toxicity: overall perspective

While CAR-T therapy has demonstrated remarkable efficacy in hematologic malignancies, its associated adverse events warrant careful clinical vigilance. Upon antigen recognition and cytotoxic degranulation, activated CAR-T cells release large quantities of proinflammatory cytokines, triggering systemic immune activation, capillary leakage, and tissue edema—collectively manifesting as cytokine release syndrome (CRS). In more severe cases, this may progress to immune effector cell–associated neurotoxicity syndrome (ICANS). Prolonged CAR-T activity can also result in immune depletion, exemplified by B-cell aplasia accompanied by hypogammaglobulinemia and recurrent infections. ICANS typically presents with aphasia, agraphia, dyscalculia, memory impairment, seizures, and cerebral edema, and may progress to disseminated intravascular coagulation (DIC). Although the mechanisms of ICANS are not fully elucidated, they are thought to involve CAR-T–induced disruption of the blood–brain barrier. Dexamethasone and other corticosteroids remain first-line treatments, while anakinra—an interleukin-1 receptor antagonist capable of crossing the blood–brain barrier—offers therapeutic benefit in refractory cases. B-cell aplasia occurs because CAR-T cell therapies are typically designed to target B-cell surface antigens such as CD19; thus, when these cells are eliminated, the number of B cells is markedly reduced [26]. This depletion can weaken the immune system, making patients more susceptible to external infections. Hypogammaglobulinemia develops as a consequence of B-cell loss. Immunoglobulin replacement is the primary approach for managing hypogammaglobulinemia, and infection monitoring and control are also essential.

Critically, the antigen specificity of CAR-T influences organ toxicity. Cardiac adverse events occur in up to 36 % of cases, including arrhythmias, heart failure, myocardial infarction, and death. A meta-analysis in hematologic malignancies found an incidence of left ventricular dysfunction in 8.68 % and lower incidences of serious cardiac events—arrhythmia (0.66 %), cardiovascular death (0.63 %), and myocardial infarction (0.62 %) [66]. Another meta-analysis that included all CAR-T cohorts reported higher incidences: arrhythmias in 54 %, acute coronary syndrome in 10 %, and cardiac arrest in 7 %, often accompanied by reduced ejection fraction and elevated biomarkers such as BNP and CRP [67]. However, the limited availability of large-scale datasets and long-term follow-up studies makes it difficult to determine whether these cardiac complications stem directly from CAR-T therapy or are confounded by prior chemotherapeutic exposure. Further research is essential to define the mechanisms, incidence, and controllability of CAR-T–related cardiotoxicity in both hematologic and solid tumors.

Mechanisms of CAR-T–induced cardiovascular injury

Our view is that inadvertent “collateral damage” to the heart arises via direct and indirect mechanisms. Direct injury may result from localized immune activation near the heart, as mediastinal lymph nodes are anatomically adjacent to cardiac tissue, allowing inflammatory extension. Nunes categorized these mechanisms into two groups: “on-target, off-tumor” toxicity, in which CAR-T cells attack normal tissues expressing low levels of the intended antigen, and “off-target, off-tumor” toxicity, resulting from T-cell receptor cross-reactivity with unrelated proteins in normal tissues [68]. A classic example involves cross-reactivity between the tumor antigen MAGE-A3 and the cardiac contractile protein titin, leading to immune-mediated myocardial injury [69]. These direct-effect hypotheses still require extensive basic and clinical validation.

Conversely, indirect injury appears predominantly driven by CRS, the most frequent and potentially life-threatening toxicity following CAR-T infusion. CRS results from hyperactivated immune responses and the explosive release of cytokines such as IL-6, interferon-γ (IFN-γ), and tumor necrosis factor-α (TNF-α), among others—plausibly the principal mechanism of cardiotoxicity [70,71]. The close temporal association between CRS onset and cardiac complications strongly supports a causal relationship. IFN-γ–activated macrophages release IL-6, which in turn damages the myocardium and microvasculature. Excess IL-6 also triggers catecholamine surges, endothelial activation, vascular permeability, and intravascular volume depletion, ultimately reducing left ventricular ejection fraction and precipitating heart failure or arrhythmias [72]. The cardiac toxicity mechanisms induced by CAR-T therapy in NSCLC are summarized in Fig. 2A.

Fig. 2.

Mechanisms of CAR-T–induced cardiac injury and grading of CRS. (A) Direct and indirect mechanisms underlying CAR-T–induced cardiac injury. (B) Grading criteria for CRS from grade 1 to grade 4.

Cardiotoxicity assessment and management

Currently, there is no unified standard for the risk stratification or management of CAR-T-related cardiotoxicity. In the perspective of this review, (NT-pro)BNP and troponin can be used as criteria for grading. Patients with obvious clinical symptoms or abnormal electrocardiograms may be considered high-risk, and patients undergoing long-term treatment can be considered an independent high-risk factor. These viewpoints need to be validated by extensive retrospective studies. Based on available evidence, patients with NSCLC scheduled to receive CAR-T therapy should undergo a comprehensive baseline cardiovascular evaluation, including medical history, regular electrocardiogram (ECG), transthoracic echocardiography, serum cardiac troponin, and (NT-pro)BNP testing; in high-risk cases, cardiac MRI or functional stress testing should be considered [[73], [74], [75]] . Particular attention is needed for those with comorbid hypertension, diabetes, coronary artery disease, renal failure, or severe prior chemotherapy hypersensitivity [76,77]. These patients require more detailed assessment, explicit risk communication with patients and families, and individualized monitoring and therapy. Notably, in a retrospective study [73], the incidence of heart failure exceeded that seen with conventional chemotherapy, and cardiovascular events showed a strong correlation with CRS—typically emerging around the fifth day post-infusion [78]. Overall, the cardiovascular health of CAR-T recipients is generally poorer than that of the general population, partly due to previous exposure to cardiotoxic treatments such as anthracyclines or thoracic radiotherapy, as well as chronic comorbidities including hypertension and diabetes.

Long-term management should incorporate serial cardiac monitoring in addition to routine assessments of tumor response and hepatic or renal function. Given the strong link between CRS and cardiac complications, prompt identification and effective management of CRS are essential for preventing severe outcomes. CRS follows a progressive clinical course—beginning with fever, fatigue, and myalgia, and in severe cases advancing to hypotension, tachycardia, arrhythmia, shock, or multi-organ failure. According to the American Society for Transplantation and Cellular Therapy (ASTCT), CRS is classified into four grades of severity, summarized in Fig. 2B.

Early detection and intervention remain the cornerstone of CRS prevention. When fever (≥ 38 °C) develops, close monitoring of vital signs, appropriate antipyretic therapy, and fluid resuscitation are recommended; point-of-care ultrasound evaluation of the inferior vena cava may guide fluid management, and infectious causes must be ruled out. Moderate to severe CRS may present with hypotension, cardiac dysfunction, or shock; timely transfer to intensive care for invasive blood pressure monitoring, mechanical ventilation, or circulatory support is advised. Tocilizumab is first-line therapy that rapidly blocks IL-6 signaling; corticosteroids can be added in refractory or life-threatening cases, used cautiously to avoid suppressing CAR-T activity. Neurotoxicity often coexists with CRS; because tocilizumab does not penetrate the blood–brain barrier, corticosteroids are required for treating ICANS. Arrhythmias—particularly atrial fibrillation—are common and should be managed with electrolyte correction and antiarrhythmic therapy. Invasive hemodynamic monitoring may help distinguish distributive from cardiogenic shock when clinical ambiguity exists. In summary, early grading of CRS, timely tocilizumab administration, corticosteroid escalation when indicated, and dynamic cardiac monitoring are key to reducing mortality from CAR-T–related cardiovascular complications.

Outlook for efficacy optimization and cardiotoxicity prevention

The clinical application of CAR-T therapy to solid tumors, including NSCLC, continues to expand, driven by rapid innovation in receptor engineering and combinatorial strategies. Novel CAR architectures, along with rational integration of chemotherapy, radiotherapy, or immunotherapy, are expected to achieve qualitative improvements in efficacy and durability of response [79,80]. For patients with advanced NSCLC, these advances significantly broaden therapeutic options beyond conventional modalities. From a translational perspective, technological progress in CAR design and manufacturing should gradually reduce production costs and improve accessibility. The development of standardized manufacturing protocols and universal (“off-the-shelf”) CAR-T platforms may further streamline clinical deployment. However, safety must remain the foremost priority. Integrating baseline inflammatory and cardiac biomarkers—including high-sensitivity C-reactive protein (hs-CRP), troponin T, and NT-proBNP—together with cytokine profiles such as IL-6 and IFN-γ, could enable predictive modeling for early identification of high-risk patients. Early initiation of cardioprotective therapy, combined with real-time cardiac imaging and biomarker surveillance, holds promise for minimizing mortality associated with CAR-T–induced cardiac injury.

Conclusions

In summary, CAR-T therapy shows strong potential for NSCLC, enabling the design of antigen-specific receptors guided by oncogenic mutations and allowing synergistic use with chemotherapy or PD-L1 blockade. Despite its clinical promise, adverse events extend beyond common systemic toxicities such as anemia, rash, nausea, and vomiting—requiring particular attention to CRS and cardiotoxicity. The underlying mechanisms of cardiovascular injury remain incompletely understood, though accumulating retrospective evidence supports a close association with CRS pathophysiology. Before CAR-T therapy can be widely implemented in NSCLC, safety assurance and patient selection frameworks must be clearly defined. Off-tumor toxicities may be mitigated through improved antigen selection, refined CAR design, and integration with interventional approaches that minimize exposure to normal tissues. Moreover, a unified risk-stratification and monitoring system is needed to identify patients most suitable for CAR-T therapy and to guide tailored cardioprotective strategies.

Funding

“This study was supported in part by grants from the National Natural Science Foundation of China (82400967), and the Sichuan Provincial Natural Science Foundation Project (2025ZNSFSC1612). Additional support was provided by the Key R&D Program for International Science and Technology Cooperation of the Gansu Provincial Science and Technology Program (23YFWA-0005); the Innovation Fund for Colleges and Universities of Gansu Province (2025A-110); the Collaborative Innovation Center for Traditional Chinese Medicine Prevention and Control of Nutrition- and Environment-Related Diseases in Northwest China (ZYXT-24-02); the Natural Science Project of the Gansu Provincial Department of Science and Technology (22JRSRA582); and the Natural Science Project of the Gansu Provincial Department of Science and Technology (2023A-088).

CRediT authorship contribution statement

Yihao Liu: Writing – original draft, Visualization, Software, Methodology, Data curation, Conceptualization. Yizhu Gao: Writing – original draft, Validation, Conceptualization. Chenyu Huo: Writing – original draft. Tao Zeng: Writing – original draft. Wenjun Meng: Writing – review & editing, Validation, Data curation, Conceptualization. Haoling Zhang: Writing – review & editing, Supervision, Methodology, Funding acquisition, Conceptualization. Qinqin He: Writing – review & editing, Supervision, Methodology, Funding acquisition, Conceptualization.

Declaration of competing interest

All authors declare that they have no conflicts of interest.