Faura et al. Journal of Neuroinflammation (2021) 18:127 https://doi.org/10.1186/s12974-021-02177-0 REVIEW Open Access Stroke-induced immunosuppression: implications for the prevention and prediction of post-stroke infections Júlia Faura1, Alejandro Bustamante2*, Francesc Miró-Mur3 and Joan Montaner1,4 Abstract Stroke produces a powerful inflammatory cascade in the brain, but also a suppression of the peripheral immune system, which is also called stroke-induced immunosuppression (SIIS). The main processes that lead to SIIS are a shift from a lymphocyte phenotype T-helper (Th) 1 to a Th2 phenotype, a decrease of the lymphocyte counts and NK cells in the blood and spleen, and an impairment of the defense mechanisms of neutrophils and monocytes. The direct clinical consequence of SIIS in stroke patients is an increased susceptibility to stroke-associated infections, which is enhanced by clinical factors like dysphagia. Among these infections, stroke-associated pneumonia (SAP) is the one that accounts for the highest impact on stroke outcome, so research is focused on its early diagnosis and prevention. Biomarkers indicating modifications in SIIS pathways could have an important role in the early prediction of SAP, but currently, there are no individual biomarkers or panels of biomarkers that are accurate enough to be translated to clinical practice. Similarly, there is still no efficient therapy to prevent the onset of SAP, and clinical trials testing prophylactic antibiotic treatment and β-blockers have failed. However, local immunomodulation could open up a new research opportunity to find a preventive therapy for SAP. Recent studies have focused on the pulmonary immune changes that could be caused by stroke similarly to other acquired brain injuries. Some of the traits observed in animal models of stroke include lung edema and inflammation, as well as inflammation of the bronchoalveolar lavage fluid. Keywords: Stroke, Immunosuppression, Biomarkers, Inflammation, Infection, Pneumonia Background infections after stroke, with pneumonia and urinary tract Stroke-induced immunosuppression (SIIS) is a set of pro- infections (UTI) being the most common forms, both hav- cesses that lead to a peripheral suppression of the immune ing a frequency of 10% [2]. system after the occurrence of stroke. One of the main Among these infections, stroke-associated pneumonia and direct consequences of this SIIS is that it makes stroke (SAP) is usually the most acute type of SAI and has the patients more susceptible to bacterial infections. Stroke- worst impact on functional outcome [3]. It increases associated infections (SAIs) represent one of the major mortality for up to 1 year, prolongs hospital stays, and complications post-stroke, which worsens the functional worsens the functional outcome at discharge [4]. The outcome of patients and increases their mortality rates [1]. clinical definition of pneumonia after stroke has differed Approximately 30% (24–36%) of patients develop in many studies in both the terminology and the diagno- sis of the complication [5]. To address this issue, the * Correspondence: abustamanter.germanstrias@gencat.cat Pneumonia in Stroke Consensus (PISCES) Group pro- 2 Stroke Unit, Hospital Universitari Germans Trias i Pujol, Carretera de Canyet, posed the term SAP, to encompass all terms referring to s/n, 08916 Badalona, Barcelona, Spain lower respiratory tract infections in stroke patients Full list of author information is available at the end of the article © The Author(s). 2021 Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data. Faura et al. Journal of Neuroinflammation (2021) 18:127 Page 2 of 14 within the first 7 days after stroke onset [6]. In the same that were replicated in antibiotic-depleted microbiota publication, new diagnostic criteria were proposed based animals [14]. In experimental stroke, it has been ob- on the criteria for the definition of healthcare-associated served that the depletion of gut microbiota by antibiotics infection from the Centers for Disease Control and Pre- worsens the outcome of the ischemic mice, decreasing vention (CDC) of the United States of America, which their survival, and developing acute colitis [15]. How- were the most used criteria by that point [7]. However, ever, Benakis et al. found that antibiotic-induced alter- SAP assessment is still challenging, especially due to the ations in the gut microbiota decreased ischemic injury in limited role of chest radiography. For this, chest com- the brain. Concretely, dysbiosis reduced the trafficking puted tomography has been proposed as a complement of effector T cells from the gut to leptomeninges after to PISCES criteria in the screening of SAP in stroke pa- stroke. In the brain, these cells enhance neuroinflamma- tients. In a small cohort, thorax high-resolution com- tion [16]. Further studies are needed to elucidate the ef- puted tomography (THRCT) was able to differentiate fects and interaction of antibiotics with stroke, to between bronchopneumonia and other low respiratory determine the impact of the antibiotic treatment in the tract infections in SAP patients, demonstrating a high brain lesion in SAI patients. accuracy in the diagnosis of SAP [8]. More recently, The failure of these clinical trials could make the Kishore et al. [9] have also addressed this issue, but they stroke units rethink therapeutic strategies for SAP pre- have observed that conventional chest X-rays have lim- vention or early treatment. Diverse reasons might be re- ited accuracy for the diagnostic of SAP when compared sponsible for the lack of success of these trials, such as with THRCT. Similarly, they found in their work some the inclusion of patients with a relatively low risk for in- discrepancies between PISCES criteria and THRCT. Lar- fection, the wrong choice of the antibiotic regime, or the ger studies, including perhaps serial imaging, are needed time-window for patient inclusion. But another plausible in this field, to validate whether PISCES criteria and explanation is that despite being useful for SAP treat- THRCT might be combined for the diagnosis of SAP. ment, antibiotics might not be effective in the prevention Nowadays, the clinical strategies against SAP are based of SAP, thus opening a new opportunity for alternative on wide-spectrum antibiotics once an infection is diag- preventive measures, with SIIS as an alternative thera- nosed through clinical criteria, along with prevention by peutic target. dysphagia screening in stroke units. Recently, the PISC Along these lines, immunomodulation of the periph- ES consortia launched a recommendation for a stan- eral immune system has also been explored as an alter- dardized approach to antibiotic therapy in post-stroke native therapy for the prevention of these infections, but pneumonia [10]. Nonetheless, early treatment before the to a lesser extent than preventive antibiotics and mostly development of clinical signs could prevent the onset of in experimental stroke. Developing new immunomodu- SAP or ameliorate its consequences. This would have lating preventive therapies against the development of benefits for patients and saving also a great number of SAP will require a deep knowledge of the mechanisms resources for health care systems [11]. Several clinical that lead to SIIS and the resulting susceptibility of stroke trials have explored this idea through the administration patients to bacterial infections. of prophylactic antibiotics to acute stroke patients. How- Hence, this review intends to present those mecha- ever, the meta-analysis performed by Vermeij et al. [12] nisms implicated in SIIS, focusing on which clinical im- showed that preventive antibiotics neither reduce the plications these pathways could have, as the finding of risk of pneumonia nor the risk of death or poor outcome biomarkers capable of predicting the onset of SAI and after stroke. The treatment did reduce the occurrence of SAP and the use of immunomodulatory therapies (either overall infections and UTIs, but did not show any effect systemic or local) to prevent them. regarding the occurrence of SAP. These results were translated into recommendations against the routine use Cellular and molecular mechanisms of stroke- of prophylactic antibiotics in stroke guidelines. induced immunosuppression In addition, antibiotics can have both protective and After the occurrence of stroke, an activation of the in- detrimental effects on the central nervous system (CNS) flammatory cascade due to the release of damage- and could have modulatory effects on the immune sys- associated molecular patterns (DAMPs) by the injured tem. The administration of antibiotics affects human and dead cells causes a neuroinflammatory state in the bacterial communities and may lead to a dysbiosis of in- brain. The release of inflammatory mediators, together testinal microbiota, which could remain altered for 6 with oxidative stress and other factors, increases the per- months after antibiotic therapy [13]. Studies in mice liv- meability of the brain-blood barrier (BBB), which in turn ing in germ-free conditions showed the importance of facilitates leukocyte infiltration into the brain [17]. gut microbiota in the development and function of the Conversely, focusing on the peripheral immune sys- CNS. These germ-free mice developed brain diseases tem, immunosuppression has been reported early within Faura et al. Journal of Neuroinflammation (2021) 18:127 Page 3 of 14 the first hours after stroke. One of the main processes of et al. [21] observed the same pattern of circulating cyto- SIIS is the shift of the T cell response from a T-helper kine expression in human stroke patients. In the same (Th) 1 response, characterized by the secretion of pro- study, they were able to locate this shift in the pattern of inflammatory cytokines such as tumor necrosis factor-α circulating cytokine expression, which could occur be- (TNF-α) and interferon-γ (IFN-γ), to a T-helper 2 (Th2) tween days 1 and 3 after stroke. anti-inflammatory response, with the secretion of inter- As mentioned earlier, the shift from Th1 to Th2 is asso- leukin (IL)-10 and IL-4 among others. An increased ratio ciated with a reduction in lymphocyte counts at the of Th2-type cytokines over Th1-type cytokines charac- spleen, blood, and thymus. The decrease was observed in terizes this shift. Another characteristic trait of SIIS is MCAO mice 12 h after surgery, and it is due to enhanced the lymphocyte and splenic impairment, which consists apoptosis in these organs [18]. Spleen atrophy character- mainly of lymphocytopenia in blood, spleen, and lymph ized by a drastic loss of splenocytes due to apoptosis was nodes [18]. The molecular and cellular processes that observed at 4 days in experimental stroke [22]. In line with drive to SIIS are described in this chapter, and summa- this, Yan and Zhang [19] observed a contraction of the rized in Fig. 1. spleen volume in MCAO rats beginning at 12 h after the Various studies have reported a decrease in the pro- occlusion. This reduction of the splenic volume has also duction of IFN-γ and the secretion of TNF-α in blood been observed in humans [23, 24]. samples from mice whom underwent a middle cerebral Although susceptibility to infection after stroke could artery occlusion (MCAO), as well as an increase of the be eased by SIIS, the latter may represent an evolved circulating levels of IL-10 [18, 19]. IL-10 is a well-known mechanism to prevent immune reactions, mainly auto- anti-inflammatory cytokine that inhibits the production reactive immune responses against CNS molecular com- of TNF-α and IFN-γ [20]. A time-dependent increase in ponents by infiltrating lymphocytes. The efferent IL-10 as well as a decrease in TNF-α/IL-10 ratio was ob- neuronal pathways through CNS regulates immune sys- served in stroke patients developing infection [18]. Jiang tem are presented below. Fig. 1 The interplay of local and systemic processes leading to pneumonia in stroke patients. Stroke-associated pneumonia is influenced by systemic and local mechanisms. Locally, there are pulmonary alterations due to stroke itself and the inflammatory processes that develop. BALF inflammation, lung inflammation, and edema seem to be the principal alterations, although there are some discrepancies between studies. On a systemic level, various represented processes lead to 3 main alterations that cause systemic immunosuppression after stroke: an increase of the T- helper (Th) 2/Th1 cytokine ratio; a reduction of the lymphocyte counts in the spleen, thymus, and blood; and a decrease of the antimicrobial defense mechanisms of neutrophils and monocytes. SNS sympathetic nervous system, iNKT invariant natural killer T cells, PNS parasympathetic nervous system, ACh acetylcholine, HPA hypothalamic–pituitary–adrenal, HMGB-1 high motility group box-1, sCD163 soluble cluster of differentiation 163, MZ marginal zone, FasL Fas ligand, BALF bronchoalveolar lavage fluid. Parts of this figure were supported by Servier Medical Art with permission under the Creative Commons Attribution 3.0 Unported License Faura et al. Journal of Neuroinflammation (2021) 18:127 Page 4 of 14 SIIS is mediated by the autonomic nervous system bacterial infections, unlike propranolol treatment. These The shift from the Th1 response to the Th2 response results reveal a role of the HPA axis in SIIS, but perhaps seems to be led by the autonomic nervous system not as pivotal as the SNS. (ANS), especially the sympathetic nervous system (SNS). The other division of the ANS, the parasympathetic Catecholamines like epinephrine and norepinephrine are nervous system (PNS), also seems to play a part in SIIS. SNS mediators that are secreted when the SNS is acti- The vagus nerve and acetylcholine (ACh) suppress the vated. They act through β-adrenergic receptors of im- secretion of proinflammatory cytokines through the mune cells, suppressing Th1 activities and cellular binding of acetylcholine to the α7 nicotinic acetylcholine immune responses through the inhibition of IFN-γ se- receptor (α7nAChR) in activated macrophages. Specific- cretion and the stimulation of IL-10 production by im- ally, the inhibition of TNF-α production in spleen is mune cells, among other processes [25]. driven by ACh-producing T cells [30]. Engel et al. [31] After MCAO surgery in mice, higher levels of meta- observed a rapid increase of parasympathetic activity in nephrine (MN) and normetanephrine (NMN) (metabo- experimental stroke mice through an analysis of heart lites of epinephrine and norepinephrine, respectively) rate variability and blood pressure. Moreover, MCAO were observed from 6 h after surgery until 2 weeks. surgery was performed on two types of mice: wild-type Treating MCAO mice with the beta-blocker propranolol (WT) mice who suffered a vagotomy and α7nAChR blocks the catecholaminergic receptors and partially in- knockout (KO) mice. Both cases showed a decrease in hibits the activation of the SNS in these animals. The ef- the bacterial load of the lungs when compared with fects of this blockade in the circulation were an increase MCAO WT mice on day 3 after ischemia. of IFN-γ levels; a decrease of IL-10, MN, and NMN; and These results suggest that increased parasympathetic an increase of the splenic volume [19]. The inhibition of activity plays some role in the increased susceptibility to the SNS through propranolol has also been proven to re- bacterial infections after stroke, possibly through the duce lymphocyte dysfunction and bacterial infections, as suppression of innate immune responses of the lungs well as improve the survival of MCAO mice [18], thus due to the expression of α7nAChR in activated lung supporting the relevance of the SNS in SIIS. macrophages and alveolar epithelial cells. The results ob- β-arrestin2 (ARRB2) has been studied to elucidate the tained by Lafargue et al. [32] are in line with this hy- endogenous factors and pathways of the SNS that are in- pothesis. They observed attenuation of the effect of volved in SIIS. ARRB2 protein is involved in multiple stroke in lung injury and mortality when α7nAChR was pathways. Wang et al. [26] found elevated splenic levels pharmacologically inhibited or genetically depleted in of ARRB2 in MCAO rats and a correlation of the pro- MCAO mice with pneumonia. Conversely, pretreatment tein with sympathetic activity. They also observed dys- with an α7nAChR activator before MCAO surgery in- function of splenic monocytes, which was reversed with creased lung injury. ARRB2 deficiency in the same manner as propranolol. Hence, these results suggest ARRB2 as an effector of the Other immunomediators of SIIS apart from ANS SNS for the mediation of SIIS. The Th2/Th1 cytokine ratio is also favored by the hep- In fact, ARRB2 seems to be mediating a loss of splenic atic invariant natural killer T (iNKT) cells, producing marginal zone B cells in MCAO animals. McCulloch et al. greater amounts of IL-10, but not IFN-γ. The direct [27] demonstrated that stroke induces a substantial loss of immunomodulation of iNKT cells through the adminis- splenic B cells, especially in those from the marginal B tration of a specific activator (α-galactosylceramide) in zone. This cell loss is causing impairment to some im- MCAO mice reduced stroke-induced neutrophil pul- mune functions, such as those related with the early monary influx, lung edema, and infections. This reduc- defense against bacterial microorganisms, and this is asso- tion was observed similarly in MCAO animals treated ciated with spontaneous lung infections. The treatment with propranolol [33]. with propranolol reversed these impairments and reduced A decrease in the lymphocyte proliferation also seems the susceptibility to lung infections in MCAO mice. to cause the reduction of lymphocyte counts and could It is well known that the activity of the hypothalamic- be partially explained by a release of arginase I (ArgI) pituitary-adrenal (HPA) axis increases after stroke, with from neutrophils. The enzyme ArgI is constitutively an early increase in the concentrations of cortisol [28, expressed in neutrophil granules, and its main function 29]. In a similar manner to catecholamines, glucocorti- is to metabolize L-arginine into L-ornithine and urea. coids stimulate the secretion of anti-inflammatory cyto- Higher levels of this enzyme were found in the blood of kines. Prass et al. [18] observed that the inhibition of the MCAO mice on day 4 after ischemia, and the supple- HPA axis through the glucocorticoid receptor blocker mentation of L-arginine in vitro in splenic T cells in the RU486 prevented lymphopenia, lymphocyte apoptosis, same animals restored T cell proliferation and increased and monocyte dysfunction. However, it did not reduce IFN-γ production [34]. In humans, ARG1 mRNA Faura et al. Journal of Neuroinflammation (2021) 18:127 Page 5 of 14 expression and ArgI serum activity correlate with count after stroke, and this process is directly led by neutrophil-to-lymphocyte ratio (NLR) and stroke sever- SNS and HPA [24]. In the same study, the administra- ity in stroke patients [35]. These two clinical parameters tion of propranolol and RU486 significantly increased are associated with SAI [36, 37], thus suggesting a role the NK-cell-mediated immune response against post- of ArgI in immune dysfunction. stroke pneumonia, which improved the mortality rates The release of ArgI could be mediated by high mobil- and reduced the bacterial burden in MCAO mice in- ity group box 1 (HMGB1), a well-known DAMP, via the fected with Listeria monocytogenes (LM). receptor of advanced glycation end-products (RAGE). Some functions of the peripheral innate immune cells Liesz et al. [38] found that this pathway seems to stimu- are altered after stroke, and they could also contribute to late the production of myeloid-derived suppressor cells SIIS. The mechanisms related to bacterial killing, like (MDSCs), which could include neutrophils. They also oxidative burst and NETosis, are significantly impaired observed that mice with RAGE deficiency after MCAO in the monocytes and neutrophils of stroke patients. surgery had reduced splenic lymphocyte apoptosis and This impairment was observed upon admission and nor- increased activated T cell counts. These results suggest malized after day 5 for NETosis. In contrast, the inhib- that HMGB1 could stimulate the release of ArgI by ition of the oxidative burst persisted until day 7 [42]. MDSCs, in turn causing a decrease in leukocyte Recently, it has also been described a reduced expres- production. sion of CD11b in the neutrophils of stroke patients when Soluble CD163 (sCD163) is another molecule that compared with healthy controls. CD11b recognizes and seems to be implicated in the reduction of lymphocyte binds to opsonized bacteria and facilitates the internal- counts. CD163 is a receptor expressed in the membrane ization of pathogens, among other functions, and it is of monocytes and macrophages and is shed through considered a leukocyte activation marker [43]. However, matrix metalloproteinase ADAM17 into sCD163. in this same study, van Gemmeren et al. did not found O’Connell et al. [39] reported higher blood levels of an impairment of the neutrophil oxidative burst of sCD163 and ADAM17 in stroke patients, which nega- stroke patients. The authors ascribe this finding to the tively correlated with post-stroke lymphocyte counts. In small sample size. vitro, the authors related the elevations of sCD163 with Regarding monocytes, Li et al. [44] found overexpres- the suppression of lymphocyte proliferation. sion of microRNA-4445 in this leukocyte type. This Intense research is being conducted on new signaling overexpression seems to be implicated in SIIS, as it pathways implicated in SIIS. Last year, a role of the clus- causes a suppression of the tumor necrosis factor ter of differentiation (CD) 200-CD200R1 signaling axis receptor-associated 4 (TRAF4)/inhibitor of kappa B in SIIS was identified. CD200R1 is an inhibitory immune alpha (IκBα)/nuclear factor kappa B (NF-κB) signaling receptor that is expressed in myeloid cells, and it has pathway, activating, in turn, the expression of anti- been observed that CD200R1-KO MCAO mice devel- inflammatory cytokines in monocytes. oped more spontaneous bacterial lung infections than WT MCAO mice, as well as lymphocytopenia and worse functional outcomes, among other effects [40]. These Sex hormones play a part in SIIS findings suggest that CD200-CD200R1 could be a thera- Differences between sexes in stroke have been observed peutic target for immunomodulating preventive therap- in many studies, and male sex is a risk factor for its inci- ies of SAI due to its possible protective role against dence and outcome. The difference between sexes seems them. to diminish with menopause among women, thus sug- Another mechanism that seems to be implicated in gesting a neuroprotective role of endogenous estrogens the reduction of lymphocyte counts through the increase against stroke. Male sex has been also described as a risk of T cell apoptosis is the activation of the AIM2 factor for SAP [45]. inflammasome-driven signaling cascade in monocytes For this reason, experimental stroke studies and the [41]. This inflammasome is activated by sensing cell-free study of SIIS are almost always performed with males. DNA, and through the secretion of IL-1β, induces the However, Zang et al. [46] performed a novel study on expression of FasL in monocytes, leading to the T cell SIIS in female mice and described a protective role of es- apoptosis. Inhibiting this pathway, Roth et al. have ob- trogens (concretely, estradiol (E2)) in the peripheral im- served a reduction of bacterial infections after experi- munosuppression of MCAO mice, which improved mental stroke and an increase of the T cell survival. splenocyte counts. Furthermore, a similar protective role to that of E2 was observed for an agonist of G1: a syn- Innate immune cells are also modified after stroke thetic agonist of the membrane estrogen receptor G Apart from lymphocytes, splenic NK cells have also been protein-coupled receptor 30. This suggests a potential reported to undergo atrophy and a reduction in the treatment for SIIS. Faura et al. Journal of Neuroinflammation (2021) 18:127 Page 6 of 14 Regarding male sexual hormones, Dziennis et al. [47] NLR is a well-known marker of systemic inflammation. reported an exacerbation of the peripheral immunosup- Higher levels of NLR have been found in SAP patients, pression in castrated MCAO mice that were treated with and there seems to be an association between NLR levels dihydrotestosterone (DHT) replacement. The authors and the severity of SAP [37]. described the role of DHT as a peripheral immunosup- Monocyte to high-density lipoprotein (HDL) choles- pressor following stroke. terol ratio (MHR) has emerged as a novel inflamma- Altogether, these antagonistic roles of the sex hor- tion marker. In a recent study, Sun et al. [51] showed mones in immunosuppression could be involved in the the value of MHR and monocyte count as predictors fact that male patients have a higher risk of suffering of SAP at admission in a study with more than 800 SAP. However, it is also needed to take into account that SAI patients. most women who undergo a stroke are post- menopausal, so these differences among sexes could be Clinical routine and nutritional-related molecules lower than expected. Blood biomarkers that are measured routinely in most patients in clinical practice are also good candidates Clinical implications (1): the use of cellular and for SAI prediction due to their accessibility, in a simi- molecular changes as biomarkers for the lar manner to leukocyte counts. In non-diabetic prediction of stroke-associated infections stroke patients, hyperglycemia upon admission was As mentioned in the “Background” section, there is a found to be independently associated with SAI, SAP, lack of preventive therapies for SAI and SAP in clinical and UTI, as well as a poor functional outcome at 3 practice. The patient selection could be one of the rea- months after the stroke episode [52]. Rodriguez-Sanz sons for the failure of the clinical trials with prophylactic et al. [53] showed an inverse correlation between antibiotics. Nowadays, apart from dysphagia screening, HDL cholesterol levels and the development of SAI. there are no routine measures to identify patients who After establishing a cut-off point for SAI prediction have the highest risk of experiencing SAI in clinical (38.5 mg/dl), higher HDL was significantly associated practice. The ability to select such patients could repre- with a lower risk of developing SAI. sent an improvement in the experimental design of fu- Biomarkers related to nutritional status have also been ture clinical trials, either with the administration of explored. Serum albumin levels were identified as an in- preventive antibiotics or through immunomodulation. dependent predictor of SAP in a cohort of 705 patients For this reason, research on biomarkers for the predic- [54]. Furthermore, lower levels of serum prealbumin tion of SAI development has been active in the last dec- were associated with a greater risk of SAI in a smaller ade. Different cell and molecular changes related to the cohort of 104 ischemic stroke patients [55]. Huang et al. mentioned pathways have been explored and could act [56] recently found an association between lower levels as predictive biomarkers for SAI or SAP (table S1). of vitamin D and SAP. In this study, patients were di- vided into 3 categories according to their vitamin D sta- Cell counts tus: deficiency (< 25 nmol/L), insufficiency (25–50 nmol/ As a result of leukocyte recruitment to the brain, im- L), and sufficiency (> 50 nmol/L). Patients with vitamin munological changes can be observed in circulating im- D deficiency had a higher incidence of SAP than patients mune cells, and systemic inflammation is observed after with vitamin D insufficiency and sufficiency. Moreover, stroke. In the first hours following stroke, there is an ex- when adjusting by cofounders, vitamin D remained an ponential increase in the peripheral neutrophil count independent predictor of SAP. and a decrease in the circulating levels of lymphocytes [48]. Due to these changes in peripheral blood cells, dif- ferent leukocyte counts are being studied as biomarkers Acute-phase reactants in stroke research, such as white blood cell (WBC) count C-reactive protein (CRP) has been widely used in clinical and, recently, the NLR. The leukocyte counts and ratios practice as an acute-phase marker, which is why it has are easily detectable and they make the early determin- been also studied deep in the field of SAI. Fluri et al. ation a possibility, which means they could potentially [49] investigated the role of CRP and other markers in be useful biomarkers. SAI in more than 300 patients. They found that CRP is WBC count was explored as a predictor of SAI, SAP, an independent predictor of SAI, both alone and in and UTI, and the results indicated it could be an inde- combination with clinical factors and other markers. pendent predictor of these complications in acute stroke Our group reviewed the studies related to CRP and SAI patients [49]. Higher levels of circulating NK cells within to date in a systematic review. CRP was observed to be the first hours after stroke were observed in patients an independent predictor of SAI, and there was an opti- who developed SAI in a study with 59 participants [50]. mal time window between 24 and 48 h after stroke [57]. Faura et al. Journal of Neuroinflammation (2021) 18:127 Page 7 of 14 Serum amyloid A (SAA) is another acute phase- along with lipopolysaccharide-binding protein (LBP) and protein that has been associated with the prediction of CRP. The four proteins were measured at different time SAI in a small cohort of patients [58] and was validated points in 56 stroke patients, and all of them were associ- as an independent predictor in a recent study with two ated with SAI incidence at most of the time points, but new cohorts with more than 250 participants each [59]. only CRP and IL-10 were independent predictors of SAI An association of SAA with SAP was observed in a dif- at the earliest time point (6 h after stroke onset). In a ferent study, but in this case, the association was in com- different study, IL-6 was an independent predictor of bination with another marker, the soluble urokinase- SAI in a cohort of 82 stroke patients and was also asso- type plasminogen activator receptor (suPAR) [60]. When ciated with mortality [67]. In a meta-analysis, Busta- cut-off points were established, the combination of these mante et al. [68] reported an independent association of two markers predicted SAP with high specificity and IL-6 with the overcoming of SAI, but the additional pre- sensitivity at 48 h after stroke onset. Mid-regional pro- dictive value over clinical predictors of SAI was modest. adrenomedullin (MR-proADM) was also associated with IL-10 has also been evaluated as a predictor of SAI in SAP but at an earlier time than the SAA/suPAR combin- other studies. Chamorro et al. [69] reported that IL-10 ation (24 h after stroke). and circulating monocytes are independent predictors of The relationship of suPAR and MR-proADM with SAI SAI in 110 stroke patients, and there were higher levels was first explored by Bustamante et al. [61] when evalu- of IL-6 in patients who developed an infection. In line ating various classical sepsis biomarkers as predictors of with these results, Ashour et al. [36] reported that IL-10 SAI. suPAR and MR-proADM were both found to be in- is an independent predictor of SAI at admission in a dependent predictors of SAI but in two different cohorts study on 60 patients after establishing a cut-off point. with different time points for blood collection. MR- Salat et al. [70] explored various cytokines in SAI pa- proADM was found to be an early predictor of SAI in a tients and assessed the role of IL-10, IL-13, and IFN-γ as cohort of 78 patients (based on blood collection at the independent predictors of SAI. first 6 h after stroke onset). On the other hand, suPAR Another study also explored IFN-γ jointly with was found to be a predictor of SAI in a different cohort, microRNA-21 (miRNA-21) in a cohort of 54 patients. where the time point of the collection was at 24 h after miRNA-21 was demonstrated to be related to IFN-γ de- stroke. ficiency when it is increased. A significant increase in Fluri et al. [49] explored procalcitonin (PCT) and miRNA-21 levels and a significant decrease in IFN-γ copeptin jointly with WBC, and each one by itself was were observed in the peripheral blood of patients who found to be an independent predictor of SAI, SAP, and developed SAI, and there was a correlation between both UTI. Their combination, along with CRP and clinical molecules [71]. Another studied cytokine in this field is variables, improved the prediction of infection in stroke IL-1 receptor antagonist (IL-1ra). IL1-ra was studied patients. Recently, PCT and copeptin were studied as along with IL-10, cortisol, and lymphocyte count, and SAP predictors along with other biomarkers by pooling only IL-1-ra was independently associated with the risk data from two clinical trials (PREDICT [62] and STRA of infection in 112 stroke patients [72]. WINSKI [63]). Of all the studied biomarkers, the results The expression of human leucocyte antigen-DR (HLA- showed that only PCT and copeptin were independent DR) on peripheral monocytes has also been studied as a predictors of SAP when adjusting for clinical variables predictor of SAI and SAP. In the PREDICT study, HLA- such as dysphagia, chronic obstructive pulmonary dis- DR was found as an independent predictor of SAP in ease (COPD), hypercholesterolemia, and the National the presence of dysphagia on day 1 after stroke [62]. Institute of Health Stroke Scale (NIHSS) at admission Similar results were found by Zhang et al. [73] but fo- [64]. In contrast, at earlier time points, Hu et al. [65] did cusing on SAI. Even though there was an increase of ex- not find PCT to be useful for the prediction of SAP, only pression in SAI patients in comparison with non-SAI when the clinical suspicion of SAP is already high. In the patients on days 4 and 6 after stroke and not earlier, the STRAWINSKI clinical trial, PCT was used as a guide for authors reported a predictive value of HLA-DR for the prophylactic antibiotic therapy. The clinical trial was risk of infection after a stroke on days 1 and 2. negative, and PCT was not a predictor of SAI in the study population [63]. Others The ANS seems to have a role in immunosuppression Cytokines and immune-related molecules after stroke, as mentioned in the “Clinical implications Cytokines are important immune mediators in the post- (1): the use of cellular and molecular changes as bio- stroke response, so they have been studied as predictors markers for the prediction of stroke-associated infec- of SAI, including pro-inflammatory IL-6 and anti- tions” section. Heart rate variability (HRV) reflects the inflammatory IL-10. Worthmann et al. [66] studied both, activity of this system, so the possible role of HRV in the Faura et al. Journal of Neuroinflammation (2021) 18:127 Page 8 of 14 prediction of SAI was explored by Günther et al. [74]. Clinical implications (2): moving They found that HRV indices were able to predict the immunomodulation towards brain recovery and development of SAI before the onset of symptoms. protection against infections Suda et al. [75] investigated whether thyroid hormone Prophylactic antibiotic therapy does not seem to affect levels at admission could be associated with the inci- the functional outcomes or the mortality rate in stroke dence of SAI. They found that lower levels of free tri- patients [76]. In previous sections, we hypothesized vari- iodothyronine (FT3) were associated with an increased ous reasons why antibiotic prophylaxis could not be ef- risk of developing SAI. This association remained signifi- fective for the prevention of SAP. Another possible cant after adjusting by clinical cofounders when estab- explanation for this failure could be that the depletion of lishing a cut-off point for FT3. part of the gut microbiota by the antibiotics could be fa- voring the growth of more resistant and aggressive bac- terial strains. This, in turn, could be promoting the Limitations and future perspectives in biomarkers’ studies appearance of pneumonia in stroke patients, not only Even though multiple candidates have been studied as because of the aspiration of digestive content but also possible biomarkers for the prediction of SAI, SAP, or due to the translocation of the gut microbiota to the UTI development in stroke patients, as presented in lung, as has been suggested in experimental studies [77]. this section, none of them have been proven to have This translocation and the subsequent dissemination are enough predictive value to be used in clinical prac- mainly due to an impaired intestinal permeability after tice. This could be due to some limitations of these stroke. Thus, the need for alternative preventive therap- studies. First, only a few studies evaluated the add- ies for SAI and SAP is vital, and intense research is being itional predictive value of the biomarkers for the pre- done in this field. diction of SAI or SAP. Second, some discrepancies The “Cellular and molecular mechanisms of stroke- among studies have been observed, perhaps due to induced immunosuppression” section of this review the variability between them. The blood collection presented the known pathways implicated in SIIS. time might represent one of the most important is- Along with other clinical factors such as dysphagia, sues. Usually, patients develop SAP within the first invasive procedures, and comorbidities, SIIS increases 2–3 days after stroke, so an early detection in the the susceptibility to infections in stroke patients [78]. first hours is needed to predict the onset of the infec- This immunosuppressive response could be attenuated tion and to be able to develop therapies to prevent by modulating one or more of the mechanisms of them. Furthermore, most of the biomarker studies SIIS at the acute phase of stroke, thus reducing the have not been evaluated regarding a gold standard susceptibility of the patients to infection and decreas- test for SAP diagnosis, as there is a lack of a gold ing the mortality of stroke patients derived from this standard test for the clinical diagnosis of SAP, and complication. therefore, its diagnosis might differ between studies, In experimental stroke, immunomodulation has hindering its reproducibility. For this reason, it is im- been explored as prophylactic therapy for SAI and portant to phenotype well all the patients, whether SAP (Table 1). Several studies have studied the action with chest tomography or other tools that could be of the β-adrenoreceptor blockers propranolol and 6- developed in future studies. hydroxydopamine (6-OHDA) in experimental rodent Although none of the presented candidates had models of stroke. They observed that the inhibition of enough statistical power by itself to be translated to the SNS reduced bacterial infection of the lungs and clinical practice, combining biomarkers with high partially reestablished some of the mechanisms af- sensitivity and high specificity could open up new fected by SIIS [18, 19, 33, 79]. The blockade of the possibilities for researchers. Future studies on SAI HPA axis has also been explored through the admin- and SAP biomarkers should combine the discovery istration of the glucocorticoid receptor blocker and study of new candidates with the screening of RU486, which seemed to restore leukocyte counts but complementary candidates to find a combination did not reduce bacterial lung infections in MCAO ro- with the highest specificity and sensitivity. Hence, dents [18, 33]. These studies were discussed in detail more research needs to be done in this field to find in the “Cellular and molecular mechanisms of stroke- a biomarker or combination thereof that could be induced immunosuppression” section. used to make decisions for the administration of It is worth mentioning that β-adrenoreceptor antago- preventive therapies. Multicentric cohorts collecting nists as propranolol seem to be neuroprotective, redu- blood at early time-points and following standardized cing infarct volumes and improving neurological scores criteria for pneumonia diagnosis are perhaps the de- when being administered after MCAO surgery [86]. sirable study design. Thus, β-blockers therapy after stroke could at the same Faura et al. Journal of Neuroinflammation (2021) 18:127 Page 9 of 14 Table 1 Immunomodulating therapies to prevent stroke-associated pneumonia and infections Mechanism of Reference Drug Time of Type of Major findings action administration study Inhibition of the Prass et al. Propanolol Immediately before Experimental Prevention of lymphocyte apoptosis, lymphopenia, SNS 2003 [18] and also 4 and 8 h (MCAO mice) monocytic deactivation and changes in lymphocyte after MCAO. cytokine production; prevention of bacteremia and pneumonia; ↑ survival rates Wong Propanolol and 24 h after MCAO Experimental Reversion of the iNKT cell phenotype induced by MCAO; et al. 2011 6-OHDA (MCAO mice) ↑ survival rates; ↓ bacterial load in blood, lungs, liver, and [33] spleen Yan and Propanolol Immediately before Experimental ↓ serum levels of MN, NMN and IL-10; ↑ pro-inflammatory Zhang and also 4 and 8 h (MCAO mice) cytokines; ↑ spleen volume 2014 [19] after MCAO. Deng 6-OHDA 3 days before MCAO Experimental Reversion of the expression of MHC class II; ↑ TNF-a and et al. 2016 (MCAO rats) IFN-γ levels in LPS-stimulated macrophages in vitro; ↓ NF- [79] κB activation; ↑ β-arrestin2 expression Sykora β1-selective Before and after Clinical ↓ frequency of pneumonia; association of post-stroke BB et al 2015 BBs, stroke treatment with mortality [80] nonselective BBs Maier BBs (mainly Before and after Clinical No differences in the risk of pneumonia; ↓ mortality. et al. 2015 metoprolol and stroke [81] bisoprolol) Maier BBs Before and after Clinical No differences in the rates of pneumonia nor mortality et al. 2018 stroke [82] Inhibition of the Prass et al. RU486 24 h, 5 h, and Experimental Prevention of lymphocyte apoptosis, lymphopenia, and HPA axis 2003 [18] immediately before (MCAO mice) monocytic deactivation MCAO Immunomodulation Wong α-GalCer 24 h after MCAO Experimental ↑ systemic levels of IFN-γ; ↓ stroke-induced neutrophil of iNKT cells et al. 2011 (MCAO mice) pulmonary influx and lung edema; ↓ bacterial load in [33] blood, lungs, liver and spleen Inhibition of CD147 Jin et al. CD147 4 h after MCAO Experimental ↓ lung damage; ↓ lung leukocyte infiltration; ↓ plasma 2019 [83] antibody (MCAO mice) and lung IL-17A Inhibition of PTEN Guan Bvp 24 h after MCAO Experimental ↓ bacterial loads in lung of bpv-treated mice; restoration et al. 2013 (MCAO mice) of akt activation in the lung; ↓ mortality [84] GM-CSF Dames Recombinant 6, 30, and 54 h after Experimental ↑ leukocyte counts in lung; ↑ WBC count; ↑ long-term et al. 2018 mGM-CSF MCAO (MCAO mice) outcome [85] Experimental and clinical studies are represented in this table. In the Major findings column, all the results are referred to the patients or animals treated with the immunomodulator agent in comparison with their respective non-treated controls. MCAO middle cerebral artery occlusion, NA non-annotated, 6-OHDA 6- hydroxydopamine, iNKT invariant natural killer T cells, NM metanephrine, NMN normetanephrine, IL-10 interleukin-10, MHC major histocompatibility complex, TNF- α tumor necrosis factor-α, IFN-γ interferon-γ, LBP lipopolysaccharide binding protein, BB beta blocker, α-GalCer α-Galactosylceramide, Bvp bisperoxovanadium, GM- CSF granulocyte-macrophage colony-stimulating factor time attenuate brain injury and prevent suffering not reduce the incidence of SAP or SAI in this type of infections. patient [82]. However, the action of β-blockers in the reduction of Experimental studies have explored other immunomo- SAI and SAP in humans has been explored in various dulation strategies as a preventive therapy for SAP. The studies with different degrees of success. In a retrospect- inhibition of CD147 seems to reduce the lungs’ suscepti- ive study, Sykora et al. [80] reported a reduction in bility to bacterial infections in MCAO mice. The protein stroke mortality and SAP ratio in stroke patients under CD147 is expressed broadly in many leukocyte subtypes β-blocker therapy, along with a reduction of the SAP ra- and is involved in some immune processes like T cell ac- tio in patients with pre-stroke β-blocker therapy. How- tivation and Th17 cell differentiation. By inhibiting ever, Maier et al. [81] performed a prospective clinical CD147, Jin et al. [83] observed attenuation of lung dam- trial and did not observe differences in the ratio of SAP age, increased IFN-γ levels in the lungs, and the modula- between patients receiving the post-stroke β-blocker tion of leukocyte subpopulation changes in the lungs of therapy and those that did not. The effect of β-blockers MCAO mice. Another immunomodulatory therapy in- on major stroke patients was also studied, and they did volving the inhibition of PTEN seemed to reduce Faura et al. Journal of Neuroinflammation (2021) 18:127 Page 10 of 14 bacterial lung infections and mortality in MCAO mice, the lungs, and make them more susceptible to bacterial possibly through the restoration of the PI3K/Akt cascade infections. In turn, the organ damage could exacerbate in the lungs [84]. Systemic treatment with the brain damage [91]. pluripotent cytokine granulocyte-macrophage colony- A new approach in the research of preventive therap- stimulating factor (GM-CSF) was also explored as an ies for SAP could be derived from the exploration of the immunomodulating therapy for the prevention of bac- cellular and molecular modifications in the lungs after terial infections in experimental stroke. This treatment stroke that also contributes to the susceptibility to pneu- did not reduce the bacterial burden or mortality in monia. In a hypothetical scenario, the possible lung MCAO mice, but it did augment antibacterial immune damage could be prevented through local immunomo- responses and improved the long-term neurological out- dulation, hence decreasing the susceptibility of stroke come [85]. patients to SAP without hampering neurorestoration in There has been a lack of success in clinical trials on β- the brain after stroke. blockers, and there have been discrepancies among Samary et al. [92] studied this issue in a rat model of them. Furthermore, no other immunomodulation ther- focal ischemia. Lung edema and inflammation were ob- apy has reached a human-based study phase. Thus, new served in ischemic rats, as well as ultrastructural changes approaches are needed to find therapeutic targets. Thus in the lung parenchyma. Regarding cytokine levels, they far, immunomodulation strategies regarding SAP and observed increased expressions of TNF-α and IL-6 in SAI prevention have been systemic, with a possible det- the brain and plasma, as well as higher TNF-α levels in rimental effect on the brain or other side effects. A local the bronchoalveolar lavage fluid (BALF). Furthermore, immunomodulation approach could be a new horizon the phagocytic capability of alveolar macrophages was for the discovery of new preventive therapies for SAP. decreased in stroke animals, along with alterations in different ventilatory parameters, such as the respiratory The lung as a possible target for new immunomodulatory rate and volume tidal. therapies In a different study performed in MCAO mice, Austin One of these new possible targets could be the local lung et al. [93] observed lung inflammation in the animals, defenses. The brain-lung crosstalk after acute brain in- but not ALI. They observed BALF inflammation, but jury is well known, in which the brain and lungs interact they did not observe pulmonary edema or the same pro- closely and bidirectionally. The occurrence of major pul- inflammatory cytokine expression profile as Samary monary injuries has been observed after acute brain in- et al. (2018). jury. In the case of stroke, patients can suffer other types Farris et al. [94] also explored the lung in MCAO of pulmonary injuries apart from pneumonia, such as mice, and they found alterations in the immune cell acute lung injury (ALI). An ALI incidence of 22% has niche in the lungs of the ischemic animals, with a de- been reported in acute stroke patients [87], although the crease of lymphocytes and an increase of alveolar macro- clinical phenotype is mild when compared with other phages and neutrophils. The production of multiple forms of acute brain injury, such as severe traumatic chemokines was reduced, including CCL3, CCL20, and brain injury (TBI) and subarachnoid hemorrhage. Pneu- CCL22 among others. monia, ALI, acute respiratory distress syndrome, and These three studies demonstrate the existence of pul- neurogenic pulmonary edema have also been reported monary alterations after ischemic stroke, but there is after these two conditions [88]. After TBI, a systemic im- some controversy between them. Thus, more research is munosuppression similar to SIIS occurs, which increases needed to elucidate the extent of this lung damage and the risk of nosocomial infections [89], along with other the mechanisms underlying it. In addition, whether the factors in these neurocritical patients, such as mechan- extension and nature of this lung damage is related with ical ventilation. Another reason why the lung could be a the occurrence of lung infections remains to be proved. target for the local immunomodulation is that its epithe- lial permeability could be altered by stroke, similarly to Conclusions the gut epithelium [90]. Despite the recent advances in acute stroke treatment The mechanisms underlying this interaction between with reperfusion therapies, SAI and particularly the brain and lungs are still not well understood, al- pneumonia represent some of the major complications though a “double hit model” has been proposed. After after stroke, which have a high impact on patient out- an acute brain lesion, the first “hit” involves the secre- comes and health care systems. Due to the failure of tion of pro-inflammatory cytokines in the brain, the al- clinical trials on preventive antibiotics, attractive path- teration of the BBB permeability, and the activation of ways for researchers and clinicians against this complica- the CNS. This causes several systemic alterations that tion involve the prediction and early detection with result in local damage to the peripheral organs, such as biomarkers and preventive interventions with Faura et al. Journal of Neuroinflammation (2021) 18:127 Page 11 of 14 immunomodulators. However, the rapidly changing clin- Funding ical scenario of stroke patients creates new challenges This project received funding from Instituto de Salud Carlos III (ISCIII) [PI17/ 02130], co-financed by the European Regional Development Fund (FEDER), for such research. and from Fundació La Marató de TV3 [201706]. Neurovascular Research La- Despite the undoubted clinical benefits, the boratory takes part into the Spanish stroke research network INVICTUS+ generalization of mechanical thrombectomy exposes hy- [RD16/0019/0021]. peracute stroke patients to situations of high risk for as- Availability of data and materials piration, such as long transfers and prolonged Not applicable. interventions. Therefore, the need to move to very early interventions for both risk stratification and preventive Declarations measures points to the identification of earlier signals of Ethics approval and consent to participate the impaired immune response to identify both bio- Not applicable. markers and therapeutic targets. From a therapeutic Consent for publication point of view, researchers also have to consider the po- Not applicable. tential side-effects on the brain of enhancing systemic immunity. In this sense, the lungs arise as a potential Competing interests The authors declare that they have no competing interests. therapeutic target to be explored in the next years, par- ticularly the lung damage caused by brain ischemia, as Author details 1 well as the mechanisms and pathways that are involved. Neurovascular Research Laboratory, Vall d’Hebron Research Institute (VHIR), Universitat Autònoma de Barcelona, Barcelona, Spain. 2Stroke Unit, Hospital Universitari Germans Trias i Pujol, Carretera de Canyet, s/n, 08916 Badalona, Abbreviations Barcelona, Spain. 3Systemic Autoimmune Research Unit, Vall d’Hebron 6-OHDA: 6-Hydroxydopamine; ALI: Acute lung injury; ANS: Autonomic Research Institute (VHIR), Barcelona, Spain. 4Stroke Research Program, nervous system; ArgI: Arginase-1; ARRB2: β-arrestin2; BALF: Bronchoalveolar Institute of Biomedicine of Seville, IBiS/Hospital Universitario Virgen del lavage fluid; BBB: Blood-brain barrier; CD: Cluster of differentiation; Rocío/CSIC/University of Seville & Department of Neurology, Hospital CNS: Central nervous system; COPD: Chronic obstructive pulmonary disease; Universitario Virgen de la Macarena, Seville, Spain. CRP: C-reactive protein; DAMPs: Damage-associated molecular pattern; DHT: Dihydrotestosterone; FT3: Free triiodothyronine; GM-CSF: Granulocyte- Received: 26 March 2021 Accepted: 19 May 2021 macrophage colony-stimulating factor; HDL: High-density lipoprotein; HLA- DR: Human leucocyte antigen-DR; HMGB1: High mobility group box 1; HPA: Hypothalamic-pituitary-adrenal; HRV: Heart rate variability; IFN- References γ: Interferon-γ; IL: Interleukin; iNKT: Hepatic invariant natural killer T; 1. Suda S, Aoki J, Shimoyama T, Suzuki K, Sakamoto Y, Katano T, et al. Stroke- MCAO: Middle cerebral artery occlusion; MHR: Monocyte to HDL cholesterol associated infection independently predicts 3-month poor functional ratio; miRNA-21: MicroRNA-21; MN: Metanephrine; MR-proADM: Mid-regional outcome and mortality. J Neurol. 2018;265(2):370–5. https://doi.org/10.1007/ pro-adrenomedullin; NF-κB: Nuclear factor kappa B; NIHSS: National Institute s00415-017-8714-6. of Health Stroke Scale; NLR: Neutrophil-to-lymphocyte ratio; 2. Westendorp WF, Nederkoorn PJ, Vermeij JD, Dijkgraaf MG, van de Beek D. NMN: Normetanephrine; PCT: Procalcitonin; PISCES: Pnemonia In Stroke Post-stroke infection: a systematic review and meta-analysis. BMC Neurol. Consensus; PNS: Parasympathetic nervous system; RAGE: Receptor of 2011;11(1):110. https://doi.org/10.1186/1471-2377-11-110. advanced glycation end-products; SAA: Serum amyloid A; SAI: Stroke- 3. Bustamante A, Giralt D, García-Berrocoso T, Rubiera M, Álvarez-Sabín J, associated infection; SAP: Stroke-associated pneumonia; SIIS: Stroke-induced Molina C, et al. The impact of post-stroke complications on in-hospital immunosuppression; SNS: Sympathetic nervous system; suPAR: Soluble mortality depends on stroke severity. Eur Stroke J. 2017;2(1):54–63. https:// urokinase-type plasminogen activator receptor; TBI: Traumatic brain injury; doi.org/10.1177/2396987316681872. Th: T-helper; THRCT: Thorax high resolution computed tomography; TNF- 4. Teh WH, Smith CJ, Barlas RS, Wood AD, Bettencourt-Silva JH, Clark AB, et al. α: Tumor necrosis factor-α; TRAF4: Tumor necrosis factor receptor-associated Impact of stroke-associated pneumonia on mortality, length of 4; UTI: Urinary tract infection; WBC: White blood cell; α7nAChR: α7 nicotinic hospitalization, and functional outcome. Acta Neurol Scand. 2018;138(4): acetylcholine receptor 293–300. https://doi.org/10.1111/ane.12956. 5. Kishore AK, Vail A, Chamorro A, Garau J, Hopkins SJ, Di Napoli M, et al. How is pneumonia diagnosed in clinical stroke research? Stroke. 2015;46(5):1202– Supplementary Information 9. https://doi.org/10.1161/STROKEAHA.114.007843. The online version contains supplementary material available at https://doi. 6. Smith CJ, Kishore AK, Vail A, Chamorro A, Garau J, Hopkins SJ, et al. org/10.1186/s12974-021-02177-0. Diagnosis of stroke-associated pneumonia: recommendations from the pneumonia in stroke consensus Group. Stroke. 2015;46(8):2335–40. https:// Additional file 1: Table S1. Studied biomarkers for the prediction of doi.org/10.1161/STROKEAHA.115.009617. SAI, SAP or UTI. This table summarizes the biomarkers presented in 7. Horan TC, Andrus M, Dudeck MA. CDC/NHSN surveillance definition of section 4 alphabetically ordered. health care-associated infection and criteria for specific types of infections in the acute care setting. Am J Infect Control. 2008;36(5):309–32. https://doi. org/10.1016/j.ajic.2008.03.002. 8. Zapata-Arriaza E, Serrano-Gotarredona P, Navarro-Herrero S, Moniche F, Acknowledgments Pardo-Galiana B, Pallisa E, et al. Chest computed tomography findings and Not applicable. validation of clinical criteria of stroke associated pneumonia. J Stroke. 2019; 21(2):217–9. https://doi.org/10.5853/jos.2018.03251. Authors’ contributions 9. Kishore AK, Devaraj A, Vail A, Ward K, Thomas PG, Sen D, et al. Use of AB, JM, and JF defined the research of interest and were involved in topic pulmonary computed tomography for evaluating suspected stroke- selection. JF carried out the literature review and wrote the first draft of the associated pneumonia. J Stroke Cerebrovasc Dis. 2021;30(6):105757. https:// manuscript. AB, JM, and FMM reviewed the manuscript and made doi.org/10.1016/j.jstrokecerebrovasdis.2021.105757. contributions for improvement. All authors read and approved the final 10. Kishore AK, Jeans AR, Garau J, Bustamante A, Kalra L, Langhorne P, et al. manuscript. Antibiotic treatment for pneumonia complicating stroke: recommendations Faura et al. Journal of Neuroinflammation (2021) 18:127 Page 12 of 14 from the pneumonia in stroke consensus (PISCES) group. Eur Stroke J. 2019; facilitating pneumonia after stroke. Stroke. 2015;46(11):3232–40. https://doi. 4(4):318–28. https://doi.org/10.1177/2396987319851335. org/10.1161/STROKEAHA.115.008989. 11. Ali AN, Howe J, Majid A, Redgrave J, Pownall S, Abdelhafiz AH. The 32. Lafargue M, Xu L, Carlès M, Serve E, Anjum N, Iles KE, et al. Stroke-induced economic cost of stroke-associated pneumonia in a UK setting. Top Stroke activation of the α7 nicotinic receptor increases Pseudomonas aeruginosa Rehabil. 2018;25(3):214–23. https://doi.org/10.1080/10749357.2017.1398482. lung injury. FASEB J. 2012;26(7):2919–29. https://doi.org/10.1096/fj.11-1973 12. Vermeij JD, Westendorp WF, Dippel DWJ, van de Beek D, Nederkoorn PJ. 84. Antibiotic therapy for preventing infections in people with acute stroke. 33. Wong CHY, Jenne CN, Lee W-Y, Léger C, Kubes P. Functional innervation of Cochrane Database Syst Rev. 2018;1:CD008530. hepatic iNKT cells is immunosuppressive following stroke. Science. 2011; 13. Dethlefsen L, Huse SM, Sogin ML, Relman DA. The pervasive effects of an 334(6052):101–5. https://doi.org/10.1126/science.1210301. antibiotic on the human gut microbiota, as revealed by deep 16S rRNA 34. Sippel TR, Shimizu T, Strnad F, Traystman RJ, Herson PS, Waziri A. Arginase I sequencing. PLoS Biol. 2008;6(11):e280. https://doi.org/10.1371/journal.pbio. release from activated neutrophils induces peripheral immunosuppression 0060280. in a murine model of stroke. J Cereb Blood Flow Metab. 2015;35(10):1657– 14. Erny D, De Angelis ALH, Jaitin D, Wieghofer P, Staszewski O, David E, et al. 63. https://doi.org/10.1038/jcbfm.2015.103. Host microbiota constantly control maturation and function of microglia in 35. Petrone AB, O’Connell GC, Regier MD, Chantler PD, Simpkins JW, Barr TL. the CNS. Nat Neurosci. 2015;18(7):965–77. https://doi.org/10.1038/nn.4030. The role of arginase 1 in post-stroke immunosuppression and ischemic 15. Winek K, Engel O, Koduah P, Heimesaat MM, Fischer A, Bereswill S, et al. stroke severity. Transl Stroke Res. 2016;7(2):103–10. https://doi.org/10.1007/ Depletion of cultivatable gut microbiota by broad-spectrum antibiotic s12975-015-0431-9. pretreatment worsens outcome after murine stroke. Stroke. 2016;47(5):1354– 36. Ashour W, Al-Anwar AD, Kamel AE, Aidaros MA. Predictors of early infection 63. https://doi.org/10.1161/STROKEAHA.115.011800. in cerebral ischemic stroke. J Med Life. 2016;9(2):163–9. 16. Benakis C, Brea D, Caballero S, Faraco G, Moore J, Murphy M, et al. 37. Nam KW, Kim TJ, Lee JS, Kwon HM, Lee YS, Ko SB, et al. High neutrophil-to- Commensal microbiota affects ischemic stroke outcome by regulating lymphocyte ratio predicts stroke-associated pneumonia. Stroke. 2018;49(8): intestinal γδ T cells. Nat Med. 2015;22:516–23. 1886–92. https://doi.org/10.1161/STROKEAHA.118.021228. 17. Iadecola C, Anrathner J. The immunology of stroke: from mechanisms to 38. Liesz A, Dalpke A, Mracsko E, Antoine DJ, Roth S, Zhou W, et al. DAMP translation. Nat Med. 2012;17:796–808. signaling is a key pathway inducing immune modulation after brain injury. J 18. Prass K, Meisel C, Höflich C, Braun J, Halle E, Wolf T, et al. Stroke-induced Neurosci. 2015;35(2):583–98. https://doi.org/10.1523/JNEUROSCI.2439-14.201 immunodeficiency promotes spontaneous bacterial infections and is 5. mediated by sympathetic activation reversal by poststroke T helper cell 39. O’Connell GC, Tennant CS, Lucke-Wold N, Kabbani Y, Tarabishy AR, Chantler Type 1-like immunostimulation. J Exp Med. 2003;198(5):725–36. https://doi. PD, et al. Monocyte-lymphocyte cross-communication via soluble CD163 org/10.1084/jem.20021098. directly links innate immune system activation and adaptive immune 19. Yan FL, Zhang JH. Role of the sympathetic nervous system and spleen in system suppression following ischemic stroke. Sci Rep. 2017;7:1–14. experimental stroke-induced immunodepression. Med Sci Monit. 2014;20: 40. Ritzel RM, Al Mamun A, Crapser J, Verma R, Patel AR, Knight BE, et al. 2489–96. https://doi.org/10.12659/MSM.890844. CD200-CD200R1 inhibitory signaling prevents spontaneous bacterial 20. Moore KW, Malefyt RDW, Robert L, Garra AO. Interleukin-10 and the infection and promotes resolution of neuroinflammation and recovery after interleukin-10 receptor. Mol Cell Biol. 2001;1:683–765. stroke. J Neuroinflammation. 2019;16:1–16. 21. Jiang C, Kong W, Wang Y, Ziai W, Yang Q, Zuo F, et al. Changes in the 41. Roth S, Cao J, Singh V, Tiedt S, Hundeshagen G, Li T, et al. Post-injury cellular immune system and circulating inflammatory markers of stroke immunosuppression and secondary infections are caused by an AIM2 patients. Oncotarget. 2017;8(2):3553–67. https://doi.org/10.18632/oncota inflammasome-driven signaling cascade. Immunity. 2021;54:648–659.e8. rget.12201. 42. Ruhnau J, Schulze K, Gaida B, Langner S, Kessler C, Broker B, et al. Stroke 22. Offner H, Subramanian S, Parker SM, Wang C, Afentoulis ME, Lewis A, et al. alters respiratory burst in neutrophils and monocytes. Stroke. 2014;45(3): Splenic atrophy in experimental stroke is accompanied by increased 794–800. https://doi.org/10.1161/STROKEAHA.113.003342. regulatory T cells and circulating macrophages. J Immunol. 2006;176(11): 43. van Gemmeren T, Schuppner R, Grosse GM, Fering J, Gabriel MM, Huber R, 6523–31. https://doi.org/10.4049/jimmunol.176.11.6523. et al. Early post-stroke infections are associated with an impaired function 23. Chiu NL, Kaiser B, Nguyen YV, Welbourne S, Lall C, Cramer SC. The volume of of neutrophil granulocytes. J Clin Med. 2020;9(3):872. https://doi.org/10.33 the spleen and its correlates after acute stroke. J Stroke Cerebrovasc Dis. 2016; 90/jcm9030872. 25(12):2958–61. https://doi.org/10.1016/j.jstrokecerebrovasdis.2016.08.012. 44. Li S, Lu G, Wang D, He JL, Zuo L, Wang H, et al. MicroRNA-4443 regulates 24. Liu Q, Jin WN, Liu Y, Shi K, Sun H, Zhang F, et al. Brain ischemia suppresses monocyte activation by targeting tumor necrosis factor receptor associated immunity in the periphery and brain via different neurogenic innervations. factor 4 in stroke-induced immunosuppression. Eur J Neurol. 2020;27(8): Immunity. 2017;46(3):474–87. https://doi.org/10.1016/j.immuni.2017.02.015. 1625–37. https://doi.org/10.1111/ene.14282. 25. Elenkov IJ, Wilder RL, Chrousos GP, Vizi ES. The sympathetic nerve—an 45. Zapata-Arriaza E, Moniche F, Blanca P-G, Bustamante A, Escudero-Martínez I, integrative interface between two supersystems: the brain and the immune Uclés O, et al. External validation of the ISAN, A2DS2, and AIS-APS scores for system. Pharmacol Rev United States. 2000;52:595–638. predicting stroke-associated pneumonia. J Stroke Cerebrovasc Dis. 2018; 26. Wang H, Deng QW, Peng AN, Xing FL, Zuo L, Li S, et al. Β-Arrestin2 27(3):673–6. https://doi.org/10.1016/j.jstrokecerebrovasdis.2017.09.059. functions as a key regulator in the sympathetic-triggered 46. Zang B, Subramanian S, Dziennis S, Jia J, Uchida M, Akiyoshi K, et al. immunodepression after stroke. J Neuroinflammation. 2018;15(1):1–11. Estradiol and G1 reduce infarct size and improve immunosuppression after https://doi.org/10.1186/s12950-017-0178-z. experimental stroke. J Immunol. 2010;184(8):4087–94. https://doi.org/10.404 27. McCulloch L, Smith CJ, McColl BW. Adrenergic-mediated loss of splenic 9/jimmunol.0902339. marginal zone B cells contributes to infection susceptibility after stroke. Nat 47. Dziennis S, Akiyoshi K, Subramanian S, Offner H, Hurn PD. Role of Commun. 2017;8:1–15. dihydrotestosterone in post-stroke peripheral immunosuppression after 28. Emsley HCA, Smith CJ, Gavin CM, Georgiou RF, Vail A, Barberan EM, et al. An cerebral ischemia. Brain Behav Immun. 2011;25(4):685–95. https://doi.org/1 early and sustained peripheral inflammatory response in acute ischaemic 0.1016/j.bbi.2011.01.009. stroke: relationships with infection and atherosclerosis. J Neuroimmunol. 48. Gill D, Sivakumaran P, Aravind A, Tank A, Dosh R, Veltkamp R. Temporal 2003;139(1-2):93–101. https://doi.org/10.1016/S0165-5728(03)00134-6. trends in the levels of peripherally circulating leukocyte subtypes in the 29. Franceschini R, Tenconi GL, Zoppoli F, Barreca T. Endocrine abnormalities hours after ischemic stroke. J Stroke Cerebrovasc Dis. 2018;27(1):198–202. and outcome of ischaemic stroke. Biomed Pharmacother. 2001;55(8):458–65. https://doi.org/10.1016/j.jstrokecerebrovasdis.2017.08.023. https://doi.org/10.1016/S0753-3322(01)00086-5. 49. Fluri F, Morgenthaler NG, Mueller B, Christ-Crain M, Katan M. Copeptin, 30. Rosas-Ballina M, Olofsson PS, Ochani M, Valdés-Ferrer SI, Levine YA, Reardon Procalcitonin and routine inflammatory markers-predictors of infection after C, et al. Acetylcholine-synthesizing T cells relay neural signals in a vagus stroke. PLoS One. 2012;7:e48309. nerve circuit. Science. 2011;334(6052):98–101. https://doi.org/10.1126/ 50. De Raedt S, De Vos A, Van Binst AM, De Waele M, Coomans D, Buyl R, et al. science.1209985. High natural killer cell number might identify stroke patients at risk of 31. Engel O, Akyüz L, Da Costa Goncalves AC, Winek K, Dames C, Thielke M, developing infections. Neurol Neuroimmunol Neuroinflamm. 2015;2(2):e71. et al. Cholinergic pathway suppresses pulmonary innate immunity https://doi.org/10.1212/NXI.0000000000000071. Faura et al. Journal of Neuroinflammation (2021) 18:127 Page 13 of 14 51. Sun Y, Lu J, Zheng D, Qian J, Zhang H, Xing D, et al. Predictive value of 69. Chamorro Á, Amaro S, Vargas M, Obach V, Cervera Á, Torres F, et al. monocyte to HDL cholesterol ratio for stroke-associated pneumonia in Interleukin 10, monocytes and increased risk of early infection in ischaemic patients with acute ischemic stroke. Acta Neurol Belg. 2020:7. stroke. J Neurol Neurosurg Psychiatry. 2006;77(11):1279–81. https://doi.org/1 52. Zonneveld TP, Nederkoorn PJ, Westendorp WF, Brouwer MC, van De Beek 0.1136/jnnp.2006.100800. D, Kruyt ND. Hyperglycemia predicts poststroke infections in acute ischemic 70. Salat D, Penalba A, García-Berrocoso T, Campos-Martorell M, Flores A, Pagola stroke. Neurology. 2017;88(15):1415–21. https://doi.org/10.1212/WNL. J, et al. Immunological biomarkers improve the accuracy of clinical risk 0000000000003811. models of infection in the acute phase of ischemic stroke. Cerebrovasc Dis. 53. Rodríguez-Sanz A, Fuentes B, Martínez-Sánchez P, Prefasi D, Martínez- 2013;35(3):220–7. https://doi.org/10.1159/000346591. Martínez M, Correas E, et al. High-density lipoprotein: a novel marker for risk 71. Lin SP, Ye S, Chen XH, Jiang HL, Mao HF, Chen MT, et al. Increased of in-hospital infection in acute ischemic stroke patients? Cerebrovasc Dis. expression of microRNA-21 in peripheral blood mediates the down- 2013;35(3):291–7. https://doi.org/10.1159/000347077. regulation of IFN-γ and increases the prevalence of stroke-associated 54. Dziedzic T, Pera J, Klimkowicz A, Turaj W, Slowik A, Rog TM, et al. Serum infection. J Neurol Sci. 2016;366:235–9. https://doi.org/10.1016/j.jns.2016.03. albumin level and nosocomial pneumonia in stroke patients. Eur J Neurol. 034. 2006;13(3):299–301. https://doi.org/10.1111/j.1468-1331.2006.01210.x. 72. Tanzi P, Cain K, Kalil A, Zierath D, Savos A, Gee JM, et al. Post-stroke 55. Ye S, Lin SP, Wu K, Fan Y, Xu M. Serum prealbumin is a predictive biomarker infection: a role for IL-1ra? Neurocrit Care. 2011;14(2):244–52. https://doi. for stroke-associated infection after an ischemic stroke. Int J Neurosci. 2017; org/10.1007/s12028-010-9490-7. 127(7):601–5. https://doi.org/10.1080/00207454.2016.1218874. 73. Zhang DP, Yan FL, Xu HQ, Zhu YX, Yin Y, Lu HQ. A decrease of human 56. Huang GQ, Cheng HR, Wu YM, Cheng QQ, Wang YM, Fu JL, et al. Reduced leucocyte antigen-DR expression on monocytes in peripheral blood predicts vitamin D levels are associated with stroke-associated pneumonia in stroke-associated infection in critically-ill patients with acute stroke. Eur J patients with acute ischemic stroke. Clin Interv Aging. 2019;14:2305–14. Neurol. 2009;16(4):498–505. https://doi.org/10.1111/j.1468-1331.2008.02512.x. https://doi.org/10.2147/CIA.S230255. 74. Günther A, Salzmann I, Nowack S, Schwab M, Surber R, Hoyer H, et al. Heart 57. Bustamante A, Vilar-Bergua A, Guettier S, Sánchez-Poblet J, García-Berrocoso rate variability - a potential early marker of sub-acute post-stroke infections. T, Giralt D, et al. C-reactive protein in the detection of post-stroke infections: Acta Neurol Scand. 2012;126(3):189–96. https://doi.org/10.1111/j.1600-0404.2 systematic review and individual participant data analysis. J Neurochem. 011.01626.x. 2017;141(2):305–14. https://doi.org/10.1111/jnc.13973. 75. Suda S, Aoki J, Shimoyama T, Suzuki K, Sakamoto Y, Katano T, et al. Low free 58. Azurmendi L, Lapierre-Fetaud V, Schneider J, Montaner J, Katan M, Sanchez triiodothyronine at admission predicts poststroke infection. J Stroke JC. Proteomic discovery and verification of serum amyloid A as a predictor Cerebrovasc Dis. 2018;27(2):397–403. https://doi.org/10.1016/j. marker of patients at risk of post-stroke infection: a pilot study. Clin jstrokecerebrovasdis.2017.09.012. Proteomics. 2017:14–27. 76. Badve MS, Zhou Z, Anderson CS, Hackett ML. Effectiveness and safety of 59. Schweizer J, Bustamante A, Lapierre-Fétaud V, Faura J, Scherrer N, antibiotics for preventing pneumonia and improving outcome after acute Azurmendi Gil L, et al. SAA (serum amyloid A): a novel predictor of stroke- stroke: systematic review and meta-analysis. J Stroke Cerebrovasc Dis. 2018; associated infections. Stroke. 2020;51(12):3523–30. https://doi.org/10.1161/ 27(11):3137–47. https://doi.org/10.1016/j.jstrokecerebrovasdis.2018.07.001. STROKEAHA.120.030064. 77. Stanley D, Mason LJ, MacKin KE, Srikhanta YN, Lyras D, Prakash MD, et al. 60. Zapata-Arriaza E, Mancha F, Bustamante A, Moniche F, Pardo-Galiana B, Translocation and dissemination of commensal bacteria in post-stroke Serrano-Gotarredona P, et al. Biomarkers predictive value for early diagnosis infection. Nat Med. 2016;22(11):1277–84. https://doi.org/10.1038/nm.4194. of stroke-associated pneumonia. Ann Clin Transl Neurol. 2019;6(9):1882–7. 78. Liu DD, Chu SF, Chen C, Yang PF, Chen NH, He X. Research progress in https://doi.org/10.1002/acn3.50849. stroke-induced immunodepression syndrome (SIDS) and stroke-associated 61. Bustamante A, García-Berrocoso T, Penalba A, Giralt D, Simats A, Muchada pneumonia (SAP). Neurochem Int. 2018;114:42–54. https://doi.org/10.1016/j. M, et al. Sepsis biomarkers reprofiling to predict stroke-associated infections. neuint.2018.01.002. J Neuroimmunol. 2017;312:19–23. https://doi.org/10.1016/j.jneuroim.2017.08. 79. Deng QW, Yang H, Yan FL, Wang H, Xing FL, Zuo L, et al. Blocking 012. sympathetic nervous system reverses partially stroke-induced 62. Hoffmann S, Harms H, Ulm L, Nabavi DG, Mackert BM, Schmehl I, et al. immunosuppression but does not aggravate functional outcome after Stroke-induced immunodepression and dysphagia independently predict experimental stroke in rats. Neurochem Res. 2016;41(8):1877–86. https://doi. stroke-associated pneumonia – The PREDICT study. J Cereb Blood Flow org/10.1007/s11064-016-1899-8. Metab. 2017;37(12):3671–82. https://doi.org/10.1177/0271678X16671964. 80. Sykora M, Siarnik P, Diedler J. β-Blockers, pneumonia, and outcome after 63. Ulm L, Hoffmann S, Nabavi D, Hermans M, Mackert BM, Hamilton F, et al. ischemic stroke. Stroke. 2015;46(5):1269–74. https://doi.org/10.1161/STROKEA The randomized controlled STRAWINSKI trial: Procalcitonin-guided antibiotic HA.114.008260. therapy after stroke. Front Neurol. 2017;8:1–10. 81. Maier IL, Karch A, Mikolajczyk R, Bähr M, Liman J. Effect of beta-blocker 64. Hotter B, Hoffmann S, Ulm L, Montaner J, Bustamante A, Meisel C, et al. therapy on the risk of infections and death after acute stroke—a historical Inflammatory and stress markers predicting pneumonia, outcome, and cohort study. PLoS One. 2015;10:1–10. etiology in patients with stroke: Biomarkers for predicting pneumonia, 82. Maier IL, Becker JC, Leyhe JR, Schnieder M, Behme D, Psychogios MN, et al. functional outcome, and death after stroke. Neurol Neuroimmunol Influence of beta-blocker therapy on the risk of infections and death in Neuroinflamm. 2020;7(3):e692. https://doi.org/10.1212/NXI. patients at high risk for stroke induced immunodepression. PLoS One. 2018; 0000000000000692. 13:1–11. 65. Hug A, Mürle B, Dalpke A, Zorn M, Liesz A, Veltkamp R. Usefulness of serum procalcitonin levels for the early diagnosis of stroke-associated respiratory 83. Jin R, Liu S, Wang M, Zhong W, Li G. Inhibition of CD147 attenuates stroke- tract infections. Neurocrit Care. 2011;14(3):416–22. https://doi.org/10.1007/ associated pneumonia through modulating lung immune response in mice. s12028-009-9325-6. Front Neurol. 2019;10:1–15. 66. Worthmann H, Tryc AB, Dirks M, Schuppner R, Brand K, Klawonn F, et al. 84. Guan YT, Mao LL, Jia J, Dong CS, Zhou XM, Zheng SL, et al. Postischemic Lipopolysaccharide binding protein, interleukin-10, interleukin-6 and C- administration of a potent PTEN inhibitor reduces spontaneous lung reactive protein blood levels in acute ischemic stroke patients with post- infection following experimental stroke. CNS Neurosci Ther. 2013;19(12): stroke infection. J Neuroinflammation. 2015;12:1–9. 990–3. https://doi.org/10.1111/cns.12188. 67. Kwan J, Horsfield G, Bryant T, Gawne-Cain M, Durward G, Byrne CD, et al. IL- 85. Dames C, Winek K, Beckers Y, Engel O, Meisel A, Meisel C. 6 is a predictive biomarker for stroke associated infection and future Immunomodulatory treatment with systemic GM-CSF augments pulmonary mortality in the elderly after an ischemic stroke. Exp Gerontol. 2013;48(9): immune responses and improves neurological outcome after experimental 960–5. https://doi.org/10.1016/j.exger.2013.07.003. stroke. J Neuroimmunol. 2018;321:144–9. https://doi.org/10.1016/j.jneuroim.2 68. Bustamante A, Sobrino T, Giralt D, García-Berrocoso T, Llombart V, Ugarriza I, 018.03.005. et al. Prognostic value of blood interleukin-6 in the prediction of functional 86. Goyagi T, Horiguchi T, Nishikawa T, Tobe Y. Post-treatment with selective β1 outcome after stroke: a systematic review and meta-analysis. J adrenoceptor antagonists provides neuroprotection against transient focal Neuroimmunol. 2014;274(1-2):215–24. https://doi.org/10.1016/j.jneuroim.2 ischemia in rats. Brain Res Netherlands. 2010;1343:213–7. https://doi.org/10.1 014.07.015. 016/j.brainres.2010.04.079. Faura et al. Journal of Neuroinflammation (2021) 18:127 Page 14 of 14 87. Bai W, Li W, Ning YL, Li P, Zhao Y, Yang N, et al. Blood glutamate levels are closely related to acute lung injury and prognosis after stroke. Front Neurol. 2018;8:1–11. 88. Mrozek S, Constantin J-M, Geeraerts T. Brain-lung crosstalk: implications for neurocritical care patients. World J Crit Care Med. 2015;4(3):163–78. https:// doi.org/10.5492/wjccm.v4.i3.163. 89. Hu PJ, Pittet JF, Kerby JD, Bosarge PL, Wagener BM. Acute brain trauma, lung injury, and pneumonia: more than just altered mental status and decreased airway protection. Am J Phys Lung Cell Mol Phys. 2017;313(1):L1– 15. https://doi.org/10.1152/ajplung.00485.2016. 90. Wen SW, Wong CHY. An unexplored brain-gut microbiota axis in stroke. Gut Microbes. 2017;8(6):601–6. https://doi.org/10.1080/19490976.2017.1344809. 91. Mascia L. Acute lung injury in patients with severe brain injury: a double hit model. Neurocrit Care. 2009;11(3):417–26. https://doi.org/10.1007/s12028- 009-9242-8. 92. Samary CS, Ramos AB, Maia LA, Rocha NN, Santos CL, Magalhães RF, et al. Focal ischemic stroke leads to lung injury and reduces alveolar macrophage phagocytic capability in rats. Crit Care Critical Care. 2018;22:1–11. 93. Austin V, Ku JM, Miller AA, Vlahos R. Ischaemic stroke in mice induces lung inflammation but not acute lung injury. Sci Rep. 2019;9:1–10. 94. Farris BY, Monaghan KL, Zheng W, Amend CD, Hu H, Ammer AG, et al. Ischemic stroke alters immune cell niche and chemokine profile in mice independent of spontaneous bacterial infection. Immunity, Inflamm Dis. 2019;7(4):326–41. https://doi.org/10.1002/iid3.277. Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.