The active role of pulmonary immunity in ischaemic stroke Authors: Junyi Wu, Yikun Gao, Yilin Li, Jiayan Gu, Xiaoting Li, Jin Wang, Xiaoxing Xiong, Lijuan Gu Published: 2026 ABSTRACT Previous stroke studies have mostly regarded the lungs as passively damaged target organs, and in clinical practice in the context of stroke, the role of the lungs has been habitually limited to secondary complications such as stroke-associated pneumonia or neurogenic pulmonary oedema. However, recent studies have indicated that the lung plays a more active and central role in the pathophysiology of stroke. In this review, we propose and systematically elaborate on an emerging concept: the lung, as one of the largest immune organs in the body, functions as an active initiator and amplifier of systemic inflammation and neuroinflammation following stroke. Stroke “remotely primes” the unique immune reservoir of the lung (such as marginated pool neutrophils and resident macrophages) via neural and humoral pathways, leading to the rapid activation and release of these cells into the circulation. Activated pulmonary immune cells not only possess an enhanced capacity for brain homing but also produce large amounts of proinflammatory mediators, establishing a self-amplifying inflammatory cycle that significantly worsens neurological outcomes. Furthermore, we explore how advanced age and metabolic comorbidities, such as diabetes, pathologically remodel the pulmonary immune baseline, thereby increasing susceptibility to this self-amplifying cascade. By highlighting the current translational gap between preclinical models and human clinical cohorts, we propose that elevating the lung from a therapeutic bystander to a core intervention target provides a novel “brain–lung coprotection” strategy, which may offer a new perspective for optimizing the comprehensive clinical management of ischaemic stroke. Introduction As one of the leading causes of mortality and disability worldwide [1], the pathological damage caused by ischaemic stroke (IS) is not confined solely to neuronal necrosis within the ischaemic core but also involves a multitude of complex secondary injury processes. An excessive inflammatory response serves as the core pathological mechanism exacerbating blood–brain barrier disruption, cerebral oedema, and neuronal death [2]. Owing to the presence of the blood–brain barrier (BBB), previous studies have often separated the central nervous system (CNS) from the peripheral immune system [3]. However, increasing evidence suggests that IS can induce significant systemic immune responses and that the factors driving these responses often originate from organs outside the CNS [2, 4]. Clinical data show that approximately 20%-35% of IS patients develop stroke-associated pneumonia (SAP), primarily aspiration pneumonia, within 7 days of onset [5]. Some patients also develop neurogenic pulmonary oedema (NPE); compared with IS patients without pulmonary complications, IS patients with NPE have a threefold higher mortality rate and a significantly lower level of neurological functional recovery [6, 7]. For a long time, these complications have been attributed to a decrease in mechanical defences after stroke, such as aspiration, bed rest, and weakened swallowing reflexes [8]. While aspiration after stroke increases the risk of pneumonia sevenfold, aspiration alone is insufficient to explain the high incidence of acute stroke pneumonia. The incidence of pneumonia remains high even in patients without obvious aspiration [9]. Recent research has confirmed that the lung is not a passively damaged organ after stroke but rather an important organ involved in the immune response after stroke [10]. As the largest peripheral immune organ in the human body, the lung can undergo rapid activation of its immune reserves under the stimulation of stroke-related signals [11]. The ensuing pulmonary inflammation does more than just harm local tissue; by releasing migratory immune cells and inflammatory mediators into the bloodstream, it also worsens ischaemic brain injury [10]. This establishes a vicious cycle in which brain injury triggers pulmonary immune activation, which in turn intensifies neurological damage [12]. Indeed, the latest theoretical framework of the “brain‒lung axis” explicitly suggests that a bidirectional communication network exists between the respiratory and central nervous systems, comprising multiple pathways, including microbiological, immunological, neural, and metabolic routes [13]. While systemic inflammation is a hallmark of various stroke subtypes, IS and haemorrhagic stroke (ICH) exhibit fundamental differences in their neuroinflammatory evolution and peripheral immune mobilization [14, 15] (Table 1). The unique pathophysiological cascade of IS, particularly the paradigm of reperfusion injury and the sustained release of damage-associated molecular patterns (DAMPs), provides a critical substrate for prolonged and intricate “brain–lung” immune crosstalk [2]. In this context, stroke-induced autonomic dysfunction precipitously disrupts pulmonary immune homeostasis; subsequently, the infiltration of lung-derived immune cells and inflammatory mediators serves as a primary driver of secondary brain injury [16]. Therefore, to elucidate this complex interaction, this article focuses on ischaemic stroke. By thoroughly exploring the mechanisms of poststroke pulmonary immune activation, we aim to highlight the core regulatory role of the lung, potentially providing highly valuable intervention targets for clinical treatment. Table 1Comparison of pathological mechanisms and systemic immune responses between ischaemic stroke and haemorrhagic strokeComparisonIschaemic Stroke, ISHaemorrhagic Stroke, ICHPrimary CNS Injury MechanismCaused by the occlusion of cerebral arteries leading to a reduced oxygen and glucose supply, which triggers energy failure, culminating in multiple forms of neuronal death [17].Caused by haematoma formation within the brain parenchyma, resulting in the mechanical compression of adjacent tissue and direct neurotoxic effects from blood‑derived products including thrombin, haemoglobin, and iron [18].CNS Neuroinflammatory MicroenvironmentMicroglia are progressively activated by DAMPs from the ischaemic core, and the NLRP3 inflammasome assembles in the penumbra, aggravating injury [19, 20].Within 1–3 h after haemorrhage, microglia are rapidly activated by thrombin. This activation induces the release of proinflammatory factors (TNF-α, IL-1β) and promotes neuroinflammation [21].Autonomic Storm CharacteristicsStorm onset is relatively delayed (usually hours to 1–2 days) but can persist for several weeks [22].Focal haemorrhagic injury will immediately cause autonomic nerve dysfunction within minutes after the bleeding occurs, manifesting as sudden increases in heart rate and fluctuations in blood pressure [23].Reperfusion InjuryUnique to ISNo classical ischaemia‑reperfusion injuryRemote Impact on Pulmonary ImmunityA persistent brain-lung vicious cycle tends to establish. The continuous influx of brain-derived inflammatory factors leads to a persistently high incidence of delayed SAP [24].Severe acute secondary lung injuries, such as NPE and ARDS, are triggered, driven by an abrupt early sympathetic storm [8]. Table 1Comparison of pathological mechanisms and systemic immune responses between ischaemic stroke and haemorrhagic strokeComparisonIschaemic Stroke, ISHaemorrhagic Stroke, ICHPrimary CNS Injury MechanismCaused by the occlusion of cerebral arteries leading to a reduced oxygen and glucose supply, which triggers energy failure, culminating in multiple forms of neuronal death [17].Caused by haematoma formation within the brain parenchyma, resulting in the mechanical compression of adjacent tissue and direct neurotoxic effects from blood‑derived products including thrombin, haemoglobin, and iron [18].CNS Neuroinflammatory MicroenvironmentMicroglia are progressively activated by DAMPs from the ischaemic core, and the NLRP3 inflammasome assembles in the penumbra, aggravating injury [19, 20].Within 1–3 h after haemorrhage, microglia are rapidly activated by thrombin. This activation induces the release of proinflammatory factors (TNF-α, IL-1β) and promotes neuroinflammation [21].Autonomic Storm CharacteristicsStorm onset is relatively delayed (usually hours to 1–2 days) but can persist for several weeks [22].Focal haemorrhagic injury will immediately cause autonomic nerve dysfunction within minutes after the bleeding occurs, manifesting as sudden increases in heart rate and fluctuations in blood pressure [23].Reperfusion InjuryUnique to ISNo classical ischaemia‑reperfusion injuryRemote Impact on Pulmonary ImmunityA persistent brain-lung vicious cycle tends to establish. The continuous influx of brain-derived inflammatory factors leads to a persistently high incidence of delayed SAP [24].Severe acute secondary lung injuries, such as NPE and ARDS, are triggered, driven by an abrupt early sympathetic storm [8]. Comparison of pathological mechanisms and systemic immune responses between ischaemic stroke and haemorrhagic stroke Strategic immune cell reserves in the lungs and rapid immune cell mobilization mechanisms after stroke The lung has the largest internal surface area of any organ, with an adult human alveolar surface area of approximately 140 m² [25]; it also contains the most densely distributed capillary network in the human body [26]. In parallel, the lung serves as one of the largest reservoirs of immune cells in vivo [27–29]; its unique architecture and cellular composition enable it to respond extremely rapidly to circulating danger signals [30]. The speed of the pulmonary immune response far exceeds that of traditional peripheral immune organs such as the spleen and bone marrow [31], providing a structural foundation for the rapid initiation of systemic inflammatory responses following stroke [24]. Pulmonary immune cell reservoir The prominent advantages of the pulmonary immune cell reservoir are its abundant cell quantity, widespread distribution, and simplified activation process [31]; within this reservoir, neutrophils and macrophages are the primary effector cells, and lung capillary endothelial cells are the structural and signalling platform that enables their rapid deployment. This integrated system allows the lung to respond to stroke‑related signals within minutes [28, 32–34]. The marginal pool of neutrophils is the most rapidly responding cell population in the lung immune system [30]. Circulating neutrophils remain in the pulmonary microvascular system, forming a reservoir called the “marginal pool of the lung”. This cell population is much larger than the population of neutrophils in the blood and is in a preactivated state for a long time, thus having the ability to mount a highly efficient and rapid innate immune response [27, 35]. After stroke, the function of neutrophils in pulmonary vessels changes rapidly. Wen et al. used lung intravital imaging to track neutrophil behaviour in pulmonary microvessels after stroke. They reported that in younger mice, stroke promoted neutrophil arrest in pulmonary microvessels; however, this response was not observed in older poststroke mice. Thus, ageing weakens the normal pulmonary neutrophil response to stroke, which may explain why elderly patients have a greater risk of poststroke pneumonia [34]. Mai et al. proposed the concept of the “lung–brain axis” and provided direct experimental evidence in a global cerebral ischaemia–reperfusion mouse model. They showed that lung tissue-specific overexpression of the antioxidant enzyme SOD3 can reduce blood–brain barrier damage and neuronal injury by regulating the activation state of circulating neutrophils [33]. These findings indicate that the local redox state in the lung can remotely influence the severity of brain injury by regulating peripheral neutrophil functions. Recent studies have confirmed that haematopoietic progenitor kinase 1 (HPK1) can increase neutrophil activation levels; it does so by enhancing the NF-κB/STAT3/p38-MAPK signalling pathway and gasdermin D cleavage, which promotes the formation of neutrophil extracellular traps (NETs) and their infiltration into lung and brain tissues. The genetic inhibition of HPK1 can concurrently attenuate poststroke lung injury and neurological deficits in murine models (Fig. 1) [36]. These findings verify that stroke can remotely regulate the behaviour and functional state of neutrophils within the lung, suggesting that targeting pulmonary neutrophils or their key signalling pathways may emerge as a potential therapeutic strategy to achieve “lung–brain coprotection”. Fig. 1Mechanisms of strategic lung immune cell reserves and rapid immune cell mobilization after stroke. (Left) Physiological basis of the rapid immune response: To pass through narrow pulmonary capillaries, neutrophils must deform, and this deformation delay leads to the formation of storage space in the pulmonary marginal pools. In the lung, neutrophils have an exchange rate of seconds to minutes, which is much shorter than the exchange rate in the spleen. (Right) Pathological changes after stroke: (1) Neutrophils: HPK1 promotes the formation of NETs through the NF-κB/STAT3/p38-MAPK signalling pathway, and the overexpression of SOD3 can reduce BBB and neuronal damage. (2) Resident macrophages: Brain-derived alarmin activates the TLR4/RAGE-EGFR signalling pathway and directly drives lung inflammation. (3) Endothelial cells: Stroke induces endothelial cell ultrastructural damage, basement membrane thickening, upregulated IL-6 mRNA expression, and immune cell recruitment. Figure 1 was created with https://BioRender.com Mechanisms of strategic lung immune cell reserves and rapid immune cell mobilization after stroke. (Left) Physiological basis of the rapid immune response: To pass through narrow pulmonary capillaries, neutrophils must deform, and this deformation delay leads to the formation of storage space in the pulmonary marginal pools. In the lung, neutrophils have an exchange rate of seconds to minutes, which is much shorter than the exchange rate in the spleen. (Right) Pathological changes after stroke: (1) Neutrophils: HPK1 promotes the formation of NETs through the NF-κB/STAT3/p38-MAPK signalling pathway, and the overexpression of SOD3 can reduce BBB and neuronal damage. (2) Resident macrophages: Brain-derived alarmin activates the TLR4/RAGE-EGFR signalling pathway and directly drives lung inflammation. (3) Endothelial cells: Stroke induces endothelial cell ultrastructural damage, basement membrane thickening, upregulated IL-6 mRNA expression, and immune cell recruitment. Figure 1 was created with https://BioRender.com Resident macrophages are primarily responsible for clearing pathogens and apoptotic cells and suppressing excessive systemic inflammation [37]. Pulmonary resident macrophages are divided into alveolar macrophages (AMs) and interstitial macrophages (IMs); they differ in distribution, phenotype, and function but together regulate the pulmonary immune response [38, 39]. After a stroke, however, the functions of these cells change substantially. In their experiments, Samary et al. observed that the phagocytic capacity of alveolar macrophages in rats decreased significantly 24 h after focal cerebral ischaemia, which was accompanied by lung tissue damage and increased expression levels of inflammatory factors [40]. In addition, a recent study by Zhong et al. revealed that tissue-resident macrophages can directly initiate an inflammatory response after stroke. Circulating alarm molecules released by brain injury bind to RAGE receptors on the membrane of lung macrophages, activating the downstream EGFR signalling pathway and directly driving the inflammatory response in the lungs; clearing these resident macrophages or blocking the RAGE-EGFR pathway can alleviate stroke-induced lung injury in mice (Fig. 1) [41]. Farris et al. reported that the proportion of macrophages in the lungs of mice significantly changed after stroke and that the levels of various chemokines decreased; these findings further indicate that stroke can directly reshape the immune microenvironment of the lungs [32]. Thus, lung macrophages not only maintain homeostasis after stroke but also become initiators of inflammation, forming an important link in the lung–brain axis that connects peripheral injury to central responses. Lung capillary endothelial cells not only serve as a barrier for gas exchange but also play a role in immune regulation. Although they are not traditional immune cells, lung capillary endothelial cells are a critical component of the pulmonary immune reserve because they provide a functional platform for rapid neutrophil and macrophage responses [27, 28]. After a stroke, these endothelial cells are quickly activated and help recruit other immune cells [34]. Samary et al. reported that IL-6 mRNA expression was significantly upregulated in isolated lung endothelial cells 24 h after focal cerebral ischaemia in rats; moreover, electron microscopy revealed ultrastructural damage and basement membrane thickening of endothelial cells. These changes directly indicate that stroke activates lung endothelial cells [40]. Rapid pulmonary immune response The ability of the lung to rapidly initiate a poststroke immune response is closely related to its anatomical structure and cellular composition [42, 43]. The unique geometry of pulmonary capillaries provides a physical marginal pool of storage space for neutrophils. Wiggs et al. reported that the average diameter of pulmonary capillaries in rabbit lungs is only 5.0 to 6.0 micrometres, whereas the diameter of neutrophils is between 6.4 and 8.0 micrometres. Because neutrophils must change shape to pass through pulmonary capillaries, this deformation delay causes the physiological retention of neutrophils in the lungs [42]. Schwab et al. quantified the kinetics of this rapid response using canine models; they reported a transfer coefficient of 0.195 s⁻¹ for neutrophils moving from the circulation into the reversible marginal pool and a return rate of 0.054 s⁻¹. This rapid exchange capacity at the second to minute level enables the lung to release a large number of neutrophils from the capillaries into the circulatory system in a very short time after a stroke, thereby rapidly responding to distal inflammatory signals [43]. In addition, a special type of pulmonary vascular macrophage is distributed in pulmonary blood vessels. These cells attach directly to the surface of the capillary endothelium and can quickly identify danger signals in the circulatory system and release inflammatory mediators [44]. Using intravital imaging, Juzenaite et al. compared how neutrophils in the lung and spleen respond to systemic stimuli within the same experimental system. They found that neutrophils in pulmonary vessels cooperate with endothelial cells to capture pathogens and form cell clusters within seconds. In contrast, splenic neutrophils take minutes to hours to migrate to the marginal zone and form clusters, highlighting a difference in the underlying molecular mechanisms [31]. Therefore, owing to its unique geometric structure and the vast number of prestored cells within the vasculature, the lung achieves a faster and more direct immune response (Fig. 1). The impact of age and comorbidities on pulmonary immunity Ageing is a key regulatory factor for the immune response after stroke [45, 46]. Compared with younger individuals, older patients have a weaker immune system reserve and exhibit a phenomenon known as “inflammaging”, characterized by a chronic, sterile proinflammatory state [47]. In older patients, proinflammatory cytokines remain elevated, and this altered immune baseline makes them more prone to excessive or dysregulated systemic inflammatory responses after stroke [45]. Wen et al. revealed that advanced age weakens the antibacterial capacity of pulmonary vascular neutrophils, manifested as defects in bacterial phagocytosis and weakened chemotaxis; additionally, using advanced techniques such as single-cell RNA sequencing, they found that the expression of 79 genes in the lungs of aged mice was selectively altered, with these genes mainly controlling the chemotactic movement of neutrophils [34]. In line with this, a series of classic studies by Roy-O’Reilly, Ritzel and others further revealed that neutrophils in aged stroke animals overrelease reactive oxygen species (ROS) and matrix metalloproteinase-9 (MMP-9) [48, 49]. This dysregulated immune state not only fails to effectively eliminate pathogens but also causes severe oxidative stress damage to the alveolar‒capillary barrier, increasing susceptibility to infections (such as SAP). Ageing can also amplify poststroke pulmonary infections through the gut. In aged mice, stroke reduces the expression of intestinal mucins and tight junction proteins. This disruption creates a TNF-α-driven proinflammatory microenvironment in the colon that promotes the translocation of gut-derived bacteria to the lungs, thereby increasing the susceptibility of elderly individuals to SAP [50]. Unlike the immune decline associated with advanced age, in diabetes, hyperglycaemia activates trained immunity by expanding myeloid progenitors and releasing proinflammatory monocytes and neutrophils [51], thereby significantly exacerbating systemic inflammation following ischaemic stroke [52]. Hyperglycaemia leads to the accumulation of advanced glycation end products (AGEs) in lung tissue [53, 54]. By activating the RAGE signalling pathway, AGEs induce a subclinical proinflammatory state in alveolar macrophages and endothelial cells [41]. Under this metabolic background, the pulmonary immune system becomes pathologically amplified [55]. The diabetes-mediated pulmonary inflammatory response not only aggravates acute lung injury but also releases peripheral inflammatory mediators into the bloodstream, which may further exacerbate secondary damage to the central nervous system [3]. The clinical relevance of diabetes-induced pulmonary immune priming in the context of stroke is reinforced by large-scale epidemiological evidence. In a cohort of 608,890 hospitalized stroke patients, compared with non‑diabetic individuals, individuals with pre‑existing diabetes had 19% greater odds of developing pneumonia (OR 1.19, 99% CI 1.15–1.24; p < 0.001) [56]. Both clinical and experimental studies have shown that in diabetic patients after stroke, the recovery of neurological function is worse and the risk of infection is greater, suggesting that diabetes greatly enhances the role of the brain–lung axis. Neuro-humoral dual pathways in the remote regulation of pulmonary inflammation initiation by stroke There is no direct anatomical connection between the brain and the lungs; however, the poststroke brain rapidly transmits danger signals associated with tissue injury to distant pulmonary tissues via well-established neural and humoral pathways [10, 13, 57]. Neural pathways Following a stroke, excitotoxic substances such as glutamate released from the ischaemic core and penumbra can activate cardiovascular centres in the hypothalamus and brainstem, inducing profound hyperexcitability of the sympathetic nervous system (SNS). Sympathetic nerve fibres extensively innervate pulmonary blood vessels, airways, and immune cells; abnormal sympathetic excitation poststroke can trigger the massive release of norepinephrine, and these catecholamines can reach various organs throughout the body via the blood circulation [24, 58, 59]. Sympathetic activation can precipitate a series of pathological changes in the lung, including endothelial dysfunction and a systemic inflammatory response accompanied by pulmonary neutrophil infiltration and cytokine release [60]. Winklewski et al. reported that stroke-induced SNS overactivation is a primary cause of systemic immunosuppression, which manifests as decreased lymphocyte counts, impaired monocyte function, increased lymphocyte apoptosis, and a shift from Th1 cytokines to Th2 cytokines [24]. Crucially, this SNS-driven pathological mechanism has been extensively validated in experimental settings. The pharmacological inhibition of peripheral SNS activation using β‑blockers effectively ameliorates the immunosuppressive state and limits bacterial infections in animal models of stroke [11]. While experimental animal models provide compelling evidence that the pharmacological inhibition of the SNS can reverse poststroke immunosuppression, real-world clinical data yield promising but not entirely consistent results. Large-scale cohort studies, such as the analysis of the VISTA database by Sykora et al., have confirmed that the prophylactic or early use of β-blockers is significantly associated with a reduced incidence of poststroke pneumonia and lower early mortality in patients with ischaemic stroke [61]. However, subsequent meta-analyses and other large cohort studies have failed to uniformly corroborate these protective effects, with some even suggesting a potential increase in infection risk depending on patient subgroup [62, 63]. A detailed summary of the clinical evidence is provided in Table 2. Table 2Summary of clinical studies investigating the effects of β‑blocker (BB) therapy on poststroke outcomesStudy Design (Sample Size)Key Findings on Infection/PneumoniaKey Findings on MortalityStudy (Year)/ReferencesRetrospective analysis of VISTA database (n = 5,212)Prestroke or on‑stroke BB use inversely associated with incidence of nosocomial pneumonia (RR 0.77 and RR 0.49, respectively)On‑stroke BB use associated with reduced 3‑month mortality (aRR = 0.63, 95% CI 0.42–0.96)Sykora et al. 2015/ [61]Retrospective analysis of PASS cohort (n = 2,538)Prestroke BB use independently associated with an increased risk of any infection (aOR = 1.61, 95% CI 1.19–2.18) and pneumonia (aOR = 1.56, 95% CI 1.05–2.30)No significant association with mortalityWestendorp et al. 2016/ [63]Historical cohort study (n = 625)BB therapy did not reduce poststroke pneumonia (RR = 1.00, 95% CI 0.77–1.30) but significantly reduced urinary tract infections (RR = 0.65, 95% CI 0.43–0.98)BB therapy was associated with higher 30‑day mortality (HR = 1.93, 95% CI 1.20–3.10)Maier et al. 2015/ [64]Systematic review & meta‑analysis (6 studies)Meta‑analysis did not support a preventive effect of BBs against stroke‑associated infections; subgroup analyses even suggested an increased risk of pneumonia (OR = 1.85) or UTI (OR = 1.69) in certain patientsNot reportedYang et al. 2022/ [62]BB β-blocker, RR Relative Risk, aRR adjusted Relative Risk, aOR adjusted Odds Ratio, HR Hazard Ratio, CI Confidence Interval, UTI Urinary Tract Infection Table 2Summary of clinical studies investigating the effects of β‑blocker (BB) therapy on poststroke outcomesStudy Design (Sample Size)Key Findings on Infection/PneumoniaKey Findings on MortalityStudy (Year)/ReferencesRetrospective analysis of VISTA database (n = 5,212)Prestroke or on‑stroke BB use inversely associated with incidence of nosocomial pneumonia (RR 0.77 and RR 0.49, respectively)On‑stroke BB use associated with reduced 3‑month mortality (aRR = 0.63, 95% CI 0.42–0.96)Sykora et al. 2015/ [61]Retrospective analysis of PASS cohort (n = 2,538)Prestroke BB use independently associated with an increased risk of any infection (aOR = 1.61, 95% CI 1.19–2.18) and pneumonia (aOR = 1.56, 95% CI 1.05–2.30)No significant association with mortalityWestendorp et al. 2016/ [63]Historical cohort study (n = 625)BB therapy did not reduce poststroke pneumonia (RR = 1.00, 95% CI 0.77–1.30) but significantly reduced urinary tract infections (RR = 0.65, 95% CI 0.43–0.98)BB therapy was associated with higher 30‑day mortality (HR = 1.93, 95% CI 1.20–3.10)Maier et al. 2015/ [64]Systematic review & meta‑analysis (6 studies)Meta‑analysis did not support a preventive effect of BBs against stroke‑associated infections; subgroup analyses even suggested an increased risk of pneumonia (OR = 1.85) or UTI (OR = 1.69) in certain patientsNot reportedYang et al. 2022/ [62]BB β-blocker, RR Relative Risk, aRR adjusted Relative Risk, aOR adjusted Odds Ratio, HR Hazard Ratio, CI Confidence Interval, UTI Urinary Tract Infection Summary of clinical studies investigating the effects of β‑blocker (BB) therapy on poststroke outcomes BB β-blocker, RR Relative Risk, aRR adjusted Relative Risk, aOR adjusted Odds Ratio, HR Hazard Ratio, CI Confidence Interval, UTI Urinary Tract Infection Studies at the cellular level have revealed that pulmonary immune cells, particularly macrophages and neutrophils, widely express β2-adrenergic receptors on their surfaces. Catecholamines can activate the downstream cAMP‒PKA signalling pathway through these receptors. Following ischaemic stroke, this sympathetic hyperactivation leads to profound pulmonary immunosuppression, promoting macrophage polarization towards an anti-inflammatory (M2-like) phenotype while impairing neutrophil degranulation and antibacterial responses (Fig. 2) [65, 66]. A recent study by Xu et al. revealed that the administration of the β2-AR agonist clenbuterol poststroke in mice significantly alters immune cell composition, resulting in a decrease in total T cell numbers and an increase in anti-inflammatory cytokine levels. While this intervention can reduce the risk of pneumonia, it concurrently causes the abnormal expansion of the cerebral infarct volume, suggesting that the activation of the β2-AR signalling pathway may involve a specific balance between neuroprotective effects and peripheral immune regulation [58]. Excessive excitation of the sympathetic nervous system can also indirectly cause pulmonary vasoconstriction and increased vascular permeability. Clinical studies have shown that owing to insular involvement, patients with right-sided hemisphere stroke exhibit significant autonomic imbalance and sympathetic hyperactivity, which may increase the risk of NPE among susceptible individuals [67]. Fig. 2Neural and humoral pathways synergize to trigger lung inflammation after stroke. (Left) Neural pathways: Stroke-induced hyperactivation of the sympathetic nervous system (SNS) results in the release of catecholamines. Catecholamines bind to the β2-adrenergic receptor (β2-AR) on lung immune cells, promoting the polarization of macrophages towards an anti-inflammatory (M2-like) phenotype while inhibiting the degranulation of neutrophils and their antimicrobial response. Excessive excitement of the SNS can also cause pulmonary vasoconstriction through indirect effects, leading to pulmonary oedema. (Right) Humoral pathway: Brain-derived DAMPs (e.g., HMGB1, S100, and CNAs) enter the systemic circulation and bind to pattern recognition receptors (PRRs) on lung cells, activating the HPK1 pathway and downstream inflammatory cascades. The synergistic effect of the two pathways leads to the remodelling of the lung immune microenvironment. Figure 2 was created with https://BioRender.com Neural and humoral pathways synergize to trigger lung inflammation after stroke. (Left) Neural pathways: Stroke-induced hyperactivation of the sympathetic nervous system (SNS) results in the release of catecholamines. Catecholamines bind to the β2-adrenergic receptor (β2-AR) on lung immune cells, promoting the polarization of macrophages towards an anti-inflammatory (M2-like) phenotype while inhibiting the degranulation of neutrophils and their antimicrobial response. Excessive excitement of the SNS can also cause pulmonary vasoconstriction through indirect effects, leading to pulmonary oedema. (Right) Humoral pathway: Brain-derived DAMPs (e.g., HMGB1, S100, and CNAs) enter the systemic circulation and bind to pattern recognition receptors (PRRs) on lung cells, activating the HPK1 pathway and downstream inflammatory cascades. The synergistic effect of the two pathways leads to the remodelling of the lung immune microenvironment. Figure 2 was created with https://BioRender.com Notably, the sympathetic nervous system does not operate in isolation. Engel et al. reported that the parasympathetic nervous system is also activated after stroke. This parasympathetic activation inhibits the production of proinflammatory factors by acting on alveolar macrophages and epithelial cells, but in doing so, it impairs pulmonary innate immunity, increasing susceptibility to pneumonia [57]. Once the balance between the sympathetic and parasympathetic nervous systems is disrupted, these systems participate in the remodelling of the pulmonary immune microenvironment after stroke [10]. Therefore, the excessive activation of the autonomic nervous system is not only the initiation point for systemic immunosuppression after stroke but also provides an important structural link between brain injury and the pulmonary immune response (Fig. 2). Humoral pathways Following stroke, necrotic and apoptotic brain cells release many DAMPs. These endogenous molecules can cross the damaged blood–brain barrier and enter the systemic circulation, thereby transmitting danger signals from brain tissue to peripheral organs [68–70]. After the onset of ischaemic stroke, the polarized expression of aquaporin-4 (AQP4) on astrocyte endfeet is disrupted, which in turn impairs waste clearance function within the cerebral glymphatic system and contributes to oedema and neurotoxic protein accumulation [71]. This dysfunction leads to the accumulation of large amounts of brain-derived DAMPs in the ischaemic brain parenchyma [72]. With the destruction of the BBB and the abnormal activation of the meningeal lymphatic vessels, a pathway connecting the central nervous system and the peripheral immune system, these high concentrations of inflammatory mediators and DAMPs flood into the systemic circulation (Fig. 3) [73]. This not only further aggravates local ischaemic cerebral oedema but also, more importantly, directly transmits danger signals from the CNS to the periphery, becoming a key humoral trigger for pulmonary microvascular endothelial injury [74]. Fig. 3Brain–pulmonary vicious cycle after ischaemic stroke. (Top right) Brain injury disrupts AQP4 polarity, leading to the impaired clearance of lymphoid waste and the release of DAMPs into the circulation. (Bottom right) Activation of the HPA axis results in the release of glucocorticoids (GCs), which inhibit the phagocytosis of alveolar macrophages. (Bottom left) The lungs activate neutrophils. After activated neutrophils reach the brain, they can adhere to vascular endothelial cells and directly block capillaries. (Top left) Lung-derived extracellular vesicles (EVs) enter the circulation and transfer active caspase-1 to the brain, triggering GSDMD-mediated pyroptosis. Figure 3 was created with https://BioRender.com Brain–pulmonary vicious cycle after ischaemic stroke. (Top right) Brain injury disrupts AQP4 polarity, leading to the impaired clearance of lymphoid waste and the release of DAMPs into the circulation. (Bottom right) Activation of the HPA axis results in the release of glucocorticoids (GCs), which inhibit the phagocytosis of alveolar macrophages. (Bottom left) The lungs activate neutrophils. After activated neutrophils reach the brain, they can adhere to vascular endothelial cells and directly block capillaries. (Top left) Lung-derived extracellular vesicles (EVs) enter the circulation and transfer active caspase-1 to the brain, triggering GSDMD-mediated pyroptosis. Figure 3 was created with https://BioRender.com The most classical DAMP is HMGB1. HMGB1 can be generated through either passive release from necrotic neurons or active secretion by activated microglia, mediating neuroinflammation via receptors such as TLR2/4 and RAGE [75]. More importantly, HMGB1 can be released from the brain into the bloodstream, directly affecting peripheral immune cell function [76]. Asavarut et al. confirmed that HMGB1, a key DAMP, plays a vital role in ischaemic injury across various organs, including the heart, kidneys, lungs, and brain [77]. Upon reaching the lung, circulating HMGB1 can directly activate pulmonary immune cells [78]. The results of a recent clinical study by Yan et al. support this mechanism. An analysis of 420 acute ischaemic stroke patients revealed that serum HMGB1 levels were significantly elevated in those with SAP. Patients in the highest HMGB1 quartile had a 2.70-fold greater risk of SAP than those in the lowest quartile did, with a linear dose‒response relationship, providing clinical evidence that brain-derived HMGB1 induces pulmonary complications via humoral pathways [79]. Conversely, the S100 protein family drives the inflammatory response and immune cell recruitment process through binding to TLR4 and RAGE receptors on pulmonary immune cells [80]. In addition to proteinaceous DAMPs, nucleic acid DAMPs also participate in the transmission of signals from brain tissue to distant organs. The levels of circulating cell-free nucleic acids (CNAs) are elevated under acute pathological conditions such as stroke and trauma [81]. Furthermore, extracellular nucleic acids released after brain injury can enter the circulation in forms such as extracellular vesicles (EVs), regulating gene transcription and translation processes after being transported into target cells [69]. The diverse DAMPs released from injured brain tissue poststroke can reach the lung via the blood circulation and bind to pattern recognition receptors (PRRs), such as TLR4 and RAGE, on the surface of pulmonary immune cells, directly initiating the pulmonary inflammatory response (Fig. 2) [41]. In addition to the proinflammatory response pathway mediated by DAMPs, an equally important neuroendocrine response occurs in the humoral pathway, which is dominated by the hypothalamic–pituitary–adrenal (HPA) axis. Stroke, as a stress stimulus, activates the HPA axis and prompts the adrenal cortex to release large amounts of glucocorticoids into the systemic circulation (Fig. 3) [82, 83]. Upon reaching the lung, high concentrations of glucocorticoids bind to glucocorticoid receptors on alveolar macrophages and lymphocytes. This signalling broadly suppresses adaptive immunity, including T cell proliferation and chemotaxis, and impairs the phagocytic and antimicrobial functions of alveolar macrophages [84]. These humoral pathway mechanisms act synergistically with sympathetic neural pathways, collectively constituting the remote regulatory network of the poststroke “brain–lung axis”. Mechanisms of brain injury exacerbation by poststroke pulmonary immune activation and the pathological cycle Stroke-induced acute pulmonary immune activation can have adverse effects on ischaemic brain tissue via three primary mechanisms—cellular output, inflammatory mediator output, and the global impact of local lung injury [33, 85], further exacerbating brain injury and ultimately leading to a pathological cycle of brain injury → pulmonary immune activation → exacerbated brain injury [86]. Cellular output Immune cells mobilized and activated in the lung undergo functional priming within the pulmonary microenvironment prior to entering systemic circulation, thereby acquiring phenotypic and functional characteristics distinct from those of immune cells originating in other tissues [87]. The lung-specific overexpression of the antioxidant enzyme SOD3 can alter the activation state of circulating neutrophils, diminish their response intensity and locomotive speed to LPS stimulation, and concurrently attenuate neutrophil-mediated neurotoxicity [33]. Further research by the same team confirmed in a mouse model that lung-derived SOD3 not only mitigates local pulmonary inflammation but also reduces peripheral neutrophil activation, decreases blood-brain barrier permeability, and improves neurological function. These findings suggest that “enhancing adaptive lung–neurovascular coupling” may emerge as a novel therapeutic strategy for stroke [88]. How do these neutrophils mobilize and directionally migrate to the brain? Following a stroke, HPK1 promotes the mobilization of CXCR2⁺ bone marrow neutrophils. Using a mouse model of acute ischaemic stroke, researchers found that the genetic knockout or pharmacological inhibition of HPK1 can suppress the overactivation of peripheral neutrophils, decrease neutrophil infiltration into lung and brain tissues, and reduce the aggregation of NETs, thereby concurrently alleviating lung and brain injury [36]. These findings suggest that the HPK1–CXCR2 axis may be a critical regulatory pathway for neutrophil migration to the brain. Through experiments involving mouse bone marrow transplantation and neutrophil transplantation, Wang et al. demonstrated that myosin 1f mediates the migration of circulating neutrophils into ischaemic brain tissue and that myosin 1f knockout significantly reduces intracerebral neutrophil infiltration and decreases infarct volume [89]. Once they reach the brain, neutrophils exacerbate injury through multiple mechanisms. Research by Keuters et al. confirmed in a mouse model of stroke that neutrophils rapidly infiltrate the brain parenchyma poststroke and secrete reactive oxygen species, proteases, cytokines, and chemokines to aggravate injury and that pharmacologically reducing neutrophil infiltration into the brain significantly improves behavioural outcomes [90]. Therefore, after acquiring a specific functional state in the pulmonary microenvironment, lung-mobilized neutrophils migrate to the site of brain injury; they directly disrupt the blood–brain barrier and exacerbate neuroinflammation by releasing mediators such as NETs and reactive oxygen species, ultimately leading to secondary brain injury [91]. Recent studies have indicated that dysregulated NETs formation within ischaemic microvessels drives a highly pathogenic thrombo-inflammatory state [92]. These findings indicate that once neutrophils activated by inflammatory signals in the lungs enter the bloodstream and are recruited to the brain, they are highly likely to promote the formation of thrombi that are rich in NETs and resistant to dissolution, thereby blocking the restoration of blood flow within the ischaemic penumbra at the microcirculation level [93, 94]. Inflammatory mediator output: systemic dissemination of inflammatory factors and amplification of brain injury As a key systemic immune organ, the lung releases inflammatory mediators that can enter the systemic circulation and impact the central nervous system [95]. Huang et al. systematically elucidated the immunoregulatory mechanisms of the lung–brain axis. They explicitly reported that lung-derived immune mediators possess the physiological capacity to cross the blood‑brain barrier, directly inducing inflammatory responses within brain tissue and causing an abnormal decline in cognitive function [85, 96, 97]. Additionally, protein markers such as TLR4, HMGB1, and TREM2 play critical roles in the lung‑brain axis and are closely correlated with disease severity [85]. This mechanism has been directly validated by experimental studies. Kamuf et al. demonstrated for the first time in a porcine model that 18 h after oleic acid-induced acute lung injury, the mRNA expression of TNF-α and IL-6 in the hippocampus was significantly elevated, accompanied by neuronal apoptosis and microglial activation [98]. These findings indicate that isolated lung injury alone is sufficient to remotely induce neuroinflammation, with some regulatory effects mediated by serum cytokines. Xie et al. further emphasized that pulmonary infections and chronic diseases can exacerbate the risk of poor poststroke outcomes through a bidirectional feedback effect; as a peripheral immune organ, the lung releases inflammatory factors that can enter the circulation and affect the CNS [99]. Upon reaching the brain, these circulating inflammatory mediators amplify early neurological injury through multiple mechanisms. Dénes et al. reported that systemic inflammation worsens brain injury regardless of infarct volume. In an animal model, LPS treatment increased poststroke mortality in a dose‑dependent manner (50–100%); it also caused increased plasma cytokines, microglial IL‑1α expression, BBB disruption, and brain oedema. Importantly, even when hypothermia reduced infarct size, inflammation still led to worse neurological outcomes and higher mortality [100]. These findings have important clinical implications, indicating that even if cerebral blood flow is successfully restored and the infarct size is controlled, persistent systemic inflammation can still lead to neurological deterioration through direct effects on cerebral vessels and the brain parenchyma. Therefore, intervening in peripheral local inflammation, such as that in the lungs, has as important a practical value as neuroprotection in the clinical treatment of stroke. In addition, the “brain‒pulmonary thrombo-inflammation axis” hypothesis recently proposed by Chen et al. has provided us with a new idea. They proposed that poststroke systemic inflammation and NETs formation promote in situ thrombosis in the lungs. Simultaneously, these mediators exert both proinflammatory and prothrombotic effects across the lung and brain, driving a vicious cycle of multiorgan damage (Fig. 3) [101]. Stroke activates the lungs, which in turn release large amounts of soluble inflammatory mediators into the circulation [33]. This surge of mediators is often termed an inflammatory storm. By compromising the blood–brain barrier, activating the cerebral vascular endothelium, and remotely inducing microglial activation, these lung‑derived mediators exacerbate neuroinflammation, BBB disruption, and cerebral oedema, ultimately leading to neurological deterioration [95]. Systemic effects of local lung injury and the vicious cycle of inflammation Pathological changes in the lungs triggered by stroke signals do not occur in isolation. The DAMPs and inflammatory mediators released by lung injury itself enter the systemic circulation, which in turn further activates the immune system and aggravates brain injury; this becomes a vicious cycle [99]. Research by Mai et al. provides key evidence for this vicious cycle: in a mouse cerebral ischaemia–reperfusion model, the systemic inflammatory response induced by LPS can activate the immune status of the lungs and significantly aggravate the degree of damage to brain tissue through the coupling mechanism between the lung and the brain [102]. The lung itself is also a target organ for neutrophil infiltration. Even low-dose endotoxin synergizes with cerebral ischaemia to potently activate neutrophils. These activated neutrophils then migrate to multiple organs, including the brain, lungs, liver, and kidneys. The combined effects lead to worsened blood–brain barrier disruption, enhanced microglial activation, expanded cortical damage, and functional decline [102]. These findings indicate that the immune status of the lung can directly affect the severity of brain injury, and vice versa—the two form a mutually amplifying positive feedback loop. The momentum of this vicious cycle originates from multiple positive feedback mechanisms. First, pulmonary oedema triggered by lung injury exacerbates systemic hypoxia; hypoxia further activates systemic inflammation, fuelling the cell cycle [103, 104]. Second, the EV-mediated transmission of death signals is among the key molecular mechanisms involved. EVs released by activated pulmonary immune cells can transmit pyroptosis signals (such as gasdermin pores and damaged mitochondria) to cells in distant organs, inducing a domino-like cell death effect [105–107]. Chavez et al. confirmed that after mechanical ventilation-related lung injury, circulating EVs can activate the caspase-1/GSDMD pathway in the brain and that the transfer of these EVs to normal animals can reproduce neuroinflammation and cell death in the brain (Fig. 3) [106]. These findings indicate that cell death in local lung tissues can have substantial functional effects on other organs, including the brain, through EVs. Third, neutrophil-mediated microvascular dysfunction further amplifies injury. After activated neutrophils reach the brain, they adhere to vascular endothelial cells and directly block capillaries, causing a “no-reflow” phenomenon [108, 109]. Using two‑photon imaging, El Amki et al. provided direct in vivo evidence in mouse models that neutrophils become trapped in brain capillaries after stroke, accounting for approximately 20–30% of microvascular obstructions [110]. Consequently, even after the successful recanalization of large vessels, microcirculatory perfusion fails to be restored, aggravating ischaemia and hypoxia. This hypoxic state, in turn, activates even more immune cells, forming a vicious cycle. This self-amplifying inflammatory cycle has very direct clinical manifestations: if lung injury is complicated after stroke, the patient’s risk of death and the probability of poor neurological prognosis significantly increase [33, 36, 111]. This also explains why treatments that solely target the brain are often not effective. Because the lungs can continuously deliver inflammatory mediators and activated cells into the circulation, completely blocking this continuous input process is difficult regardless of the effectiveness of the brain treatment [112]. Therefore, interrupting this vicious cycle requires a brain‒lung coprotection strategy: on the one hand, protecting the blood‒brain barrier and suppressing neuroinflammation; on the other hand, preserving the pulmonary barrier, reducing the release of lung-derived DAMPs, and inhibiting the hyperactivation of pulmonary immune cells. The “activation threshold” of the vicious cycle between the brain and lung and stroke severity An important but not yet fully studied issue is whether pulmonary immune activation is related to stroke severity. Clinical evidence indicates that a larger infarction volume and higher neurological deficit scores (e.g., NIHSS) are significantly associated with an enhanced systemic inflammatory response and an increased risk of complications such as poststroke pneumonia [113–115]. These observations suggest that a certain degree of brain damage may be necessary to trigger the lung immune response and the aforementioned self-amplifying inflammatory cycle. In contrast, a minor stroke may trigger only a brief immune response, which is insufficient to sustain this vicious cycle [116, 117]. Currently, there is no precisely defined threshold of infarct volume required to trigger a profound pulmonary immune response. This issue may be related to individual heterogeneity, such as age, baseline immune levels, and underlying comorbidities [118]. However, large-scale stroke has been shown to be an independent risk factor for SAP, which is sufficient to demonstrate the crucial role of infarction volume in driving the vicious cycle between the brain and lung [113]. Therefore, conducting a prospective study that combines infarction volume and clinical severity scores to determine this activation threshold is crucial for guiding clinical practice. Conclusion and perspectives Research has revealed that the lung is a peripheral immune organ that actively participates in the pathological process following stroke. Traditional views attributed pulmonary complications to mechanical factors such as aspiration; however, recent studies have indicated that, owing to its specialized capillary architecture and massive immune cell reserve system, pulmonary tissue can mount a high-speed immune response to stroke-related signals within seconds to minutes, a speed that far exceeds that of traditional immune organs such as the spleen. Stroke can remotely initiate pulmonary inflammation through dual neural and humoral pathways. In the neural pathway, the sympathetic nervous system releases catecholamines, which drive pulmonary immune cells towards a anti-inflammatory phenotype via β2-AR receptors; in the humoral pathway, brain-derived DAMPs (such as HMGB1) activate TLR4/RAGE receptors on the surface of pulmonary immune cells through the circulation, eliciting local inflammation. Furthermore, the activated lung “backfires” on the brain via three mechanisms, establishing a vicious cycle: activated neutrophils within the lung can home directionally to brain tissue, causing structural and functional damage to the blood–brain barrier through the release of neutrophil extracellular traps and reactive oxygen species; lung-derived inflammatory factors can exacerbate the degree of neuronal injury in brain tissue independent of other injury factors; and DAMPs and EVs released after lung injury further activate systemic inflammation, worsening cerebral ischaemia. Together, these findings reveal a positive feedback loop of “brain injury → pulmonary immune activation → exacerbation of brain injury”. At the theoretical level, this article breaks through the traditional framework of studying central and peripheral immunity independently, systematically demonstrating the active status of the lung in stroke pathology. At the practical level, this review reveals why isolated neuroprotective therapies struggle to reverse the disease course, suggesting a novel ‘brain–lung coprotection’ strategy. Finally, at the clinical level, this article highlights the importance of poststroke proactive pulmonary monitoring, suggesting that the early identification of high-risk patients may serve as a new starting point for improving prognosis. Although paying attention to pulmonary conditions after stroke does indeed help improve prognosis, several key issues still cannot be avoided at present. Most of the existing basic research uses young and healthy experimental animals, which are quite different from the elderly patient population, which is composed of individuals with various underlying diseases; as shown in Fig. 4, approximately 73% of the current research is based on animal models, while human clinical data account for only a very small proportion of the data, which are often observational. This imbalance highlights the gap in translational research in current stroke studies and raises concerns about the applicability of experimental results in the heterogeneous clinical population of stroke patients. Moreover, this interaction between the lungs and the brain is not consistent across all patients. Factors such as stroke severity, age, and comorbidities (such as diabetes) may have key effects on the intensity and outcome of this response. In particular, whether a certain brain injury threshold is required to fully activate the pulmonary immune reserve remains an important unanswered question. Future research should prioritize the selection of clinically significant models and prospective human studies to verify these mechanisms. Fig. 4Evidence source distribution in research: This stark contrast highlights the current translational gap in stroke–lung axis research. Approximately 73% of the current mechanistic insights are derived from preclinical animal models, whereas only 27% are from human clinical cohorts. Figure 4 was created with https://BioRender.com Evidence source distribution in research: This stark contrast highlights the current translational gap in stroke–lung axis research. Approximately 73% of the current mechanistic insights are derived from preclinical animal models, whereas only 27% are from human clinical cohorts. Figure 4 was created with https://BioRender.com Despite these numerous challenges, existing achievements still offer hope for clinical application, and future research directions could focus on three key areas: (1) developing lung-targeted drug delivery systems to validate the therapeutic potential of novel targets such as the HPK1–CXCR2 axis and the RAGE–EGFR pathway; (2) integrating serum DAMPs, neutrophil activation markers, and autonomic function indices to identify high-risk patients early; and (3) conducting prospective studies to verify the efficacy of the “brain–lung coprotection” strategy (Table 3). Notably, most of the intervention targets summarized in Table 3 have been validated primarily in animal models, with only a limited number supported by clinical data. This review systematically elucidates the central role of the lung poststroke and proposes a novel conceptualization of the vicious cycle between the brain and lung. Transitioning the lung from a therapeutic bystander to a core target is poised to pioneer entirely new comprehensive treatment pathways for patients with ischaemic stroke, ultimately improving patient outcomes. Table 3Expanded summary of intervention targets in the brain–lung axisIntervention TargetRegulatory Mechanism and Mode of ActionOrgan Protective EffectsEvidence TypeReferencesHPK1–CXCR2 axisPharmacological inhibition or genetic knockout of HPK1 blocks the mobilization of CXCR2⁺ bone marrow neutrophilsSuppresses the overactivation of peripheral neutrophils, decreases their infiltration into lung and brain tissues, reduces NETs aggregation, and concurrently alleviates lung and brain injuryAnimal [36]SOD3Lung-specific overexpression of the antioxidant enzyme SOD3 alters the local microenvironmentLimits peripheral circulating neutrophil activation, mitigates blood‒brain barrier disruption, and reduces neutrophil-mediated neurotoxicityAnimal [33]RAGE–EGFR pathwayBlocks circulating alarmins from distant brain injury from binding to RAGE receptors on pulmonary macrophages, preventing downstream EGFR signalling activationPrevents resident macrophage dysfunction and significantly alleviates stroke-induced lung injuryAnimal [41]Myosin 1fInhibits myosin 1f-mediated chemotactic migration of circulating neutrophils into ischaemic tissuesSignificantly reduces neutrophil infiltration into the brain parenchyma and decreases the cerebral infarct volumeAnimal [89]Peripheral Sympathetic Nervous SystemUses pharmacological agents such as β-blockers to inhibit stroke-induced peripheral SNS overactivation.Ameliorates the poststroke immunosuppressive state and reduces the risk of mortalityAnimal [11]HMGB1/PRR PathwayTargets circulating brain-derived DAMPs (such as HMGB1) to prevent their binding to PRRs like TLR4 and RAGE on pulmonary immune cellsInhibits the humoral initiation of the pulmonary inflammatory response, thereby reducing the dose-dependent risk of stroke-associated pneumoniaAnimal [75, 76] Table 3Expanded summary of intervention targets in the brain–lung axisIntervention TargetRegulatory Mechanism and Mode of ActionOrgan Protective EffectsEvidence TypeReferencesHPK1–CXCR2 axisPharmacological inhibition or genetic knockout of HPK1 blocks the mobilization of CXCR2⁺ bone marrow neutrophilsSuppresses the overactivation of peripheral neutrophils, decreases their infiltration into lung and brain tissues, reduces NETs aggregation, and concurrently alleviates lung and brain injuryAnimal [36]SOD3Lung-specific overexpression of the antioxidant enzyme SOD3 alters the local microenvironmentLimits peripheral circulating neutrophil activation, mitigates blood‒brain barrier disruption, and reduces neutrophil-mediated neurotoxicityAnimal [33]RAGE–EGFR pathwayBlocks circulating alarmins from distant brain injury from binding to RAGE receptors on pulmonary macrophages, preventing downstream EGFR signalling activationPrevents resident macrophage dysfunction and significantly alleviates stroke-induced lung injuryAnimal [41]Myosin 1fInhibits myosin 1f-mediated chemotactic migration of circulating neutrophils into ischaemic tissuesSignificantly reduces neutrophil infiltration into the brain parenchyma and decreases the cerebral infarct volumeAnimal [89]Peripheral Sympathetic Nervous SystemUses pharmacological agents such as β-blockers to inhibit stroke-induced peripheral SNS overactivation.Ameliorates the poststroke immunosuppressive state and reduces the risk of mortalityAnimal [11]HMGB1/PRR PathwayTargets circulating brain-derived DAMPs (such as HMGB1) to prevent their binding to PRRs like TLR4 and RAGE on pulmonary immune cellsInhibits the humoral initiation of the pulmonary inflammatory response, thereby reducing the dose-dependent risk of stroke-associated pneumoniaAnimal [75, 76] Expanded summary of intervention targets in the brain–lung axis