Predictive Value of Substance P for Pulmonary Infection in Patients With Acute Cerebral Infarction During Hospitalization

Abstract

Background and Purpose

This study aimed to evaluate the predictive value of serum substance P (SP) levels at hospital admission for the development of pulmonary infection during hospitalization in patients with acute cerebral infarction. Additionally, the study explored the association between SP levels and the severity of inflammation following pulmonary infection.

Methods

A total of 340 patients with acute cerebral infarction were retrospectively analyzed. Serum SP concentrations were measured upon hospital admission. During hospitalization, 96 patients developed pulmonary infections of varying severity, while 244 patients remained free of pulmonary infection. In patients who developed infection, serum levels of high-sensitivity C-reactive protein, procalcitonin, and interleukin-6 were also measured.

Results

Invasive procedures, dysphagia, admission National Institutes of Health Stroke Scale (NIHSS) score, age, and serum SP concentration at admission were identified as significant risk factors for pulmonary infection in patients with acute cerebral infarction. Receiver operating characteristic curve analysis showed that serum SP concentration at admission had a sensitivity of 65.63% and specificity of 85.66% for predicting pulmonary infection (area under the curve=0.81, p<0.001). Moreover, SP levels at admission were positively correlated with NIHSS scores and the levels of inflammatory markers.

Conclusions

Serum SP concentration at the time of admission has predictive value for the development of pulmonary infection during hospitalization in patients with acute cerebral infarction and may serve as an early indicator of infection risk.

Graphical Abstract

INTRODUCTION

Cerebral infarction is a common neurological disorder, particularly in the elderly, characterized by localized brain tissue damage due to impaired cerebral blood flow resulting from cerebrovascular disease or systemic circulatory dysfunction.¹ It is associated with high morbidity, disability, and mortality rates, placing a substantial burden on both patients and healthcare systems.² Pulmonary infection is among the most frequent and serious complications following acute cerebral infarction, contributing significantly to clinical deterioration, increased mechanical ventilation use, prolonged hospitalization, and elevated mortality.³,⁴,⁵ Approximately 5.6% of patients with cerebral infarction develop pneumonia, and its occurrence can nearly triple the risk of death.⁶ Therefore, early identification of patients at risk for pulmonary infection is crucial for improving outcomes.

Substance P (SP) is a neuropeptide composed of 11 amino acids, widely distributed in both the central and peripheral nervous systems.⁷ It plays a key role in the transmission of nociceptive signals, regulation of autonomic functions, and modulation of neurogenic inflammation.⁸ SP exerts its effects primarily via binding to the neurokinin-1 receptor, which is expressed on neurons, endothelial cells, and various immune cells.⁹ In the immune system, SP promotes cytokine release, increases vascular permeability, recruits inflammatory cells, and enhances the expression of adhesion molecules-mechanisms that contribute to both local and systemic inflammatory responses.¹⁰,¹¹

In the respiratory tract, SP contributes to airway smooth muscle contraction, mucus secretion, and inflammatory cell infiltration.¹² Decreased degradation of SP—due to epithelial damage and reduced neutral endopeptidase activity—has been implicated in several pulmonary inflammatory diseases, including chronic obstructive pulmonary disease and acute respiratory distress syndrome.¹³ Despite these known roles, there is a lack of research exploring the role of SP in the development of pulmonary infection specifically following acute cerebral infarction.

To address this gap, the present study aimed to evaluate whether serum SP concentration measured at hospital admission in patients with acute cerebral infarction could serve as a predictive biomarker for pulmonary infection during hospitalization. In addition, the study investigated the association between SP levels and the severity of inflammation once pulmonary infection occurred. This research seeks to provide new insights into the potential utility of SP as an early indicator of post-stroke pulmonary complications, expanding its clinical relevance beyond previously studied inflammatory and respiratory diseases.

METHODS

Participants

We retrospectively analyzed 340 eligible patients with acute cerebral infarction who were admitted to the Affiliated Suzhou Hospital of Nanjing Medical University. This retrospective study included patients admitted to our hospital between January 2022 and December 2023. During hospitalization, 96 patients developed lung infection of different degrees, and 244 patients did not develop lung infection. The study was approved by the Affiliated Suzhou Hospital of Nanjing Medical University (#SZMJYY2024032002), and informed written consent was waived since this was a retrospective analysis.

Inclusion criteria: 1) all participants met the diagnostic criteria for cerebral infarction as outlined in the “Guidelines for the Diagnosis and Treatment of Acute Ischemic Stroke in China 2018”; 2) MRI and CT scans of the patients confirmed the diagnosis of acute cerebral infarction; 3) patients with a first onset, acute onset, and onset time ≤48 hours; 4) patients aged over 18 years; 5) patients without coagulation disorders; 6) patients with no history of drug allergies; 7) patients without severe kidney or liver dysfunction; and 8) patients without aphasia or hearing impairments.

Exclusion criteria: 1) patients with onset for more than 48 hours; 2) patients without cerebral infarction, transient ischemic attack or old cerebral infarction; 3) patients with pulmonary infection before cerebral infarction or before admission; 4) patients with unstable vital signs, kidney and liver, severe heart, blood, immune diseases and tumor; and 5) patients with a history of brain trauma or stroke.

Pulmonary infection diagnosis

Patients with pulmonary infection were required to meet the diagnostic criteria outlined in either the “Guidelines for the Standard Diagnosis and Treatment of Community-Acquired Pneumonia (CAP)” or the Hospital Acquired Pneumonia (HAP) diagnostic criteria established by the Respiratory Disease Branch of the Chinese Medical Association in 2006.¹⁴ The diagnosis was based on the following criteria¹⁵: 1) the onset of a new cough, expectoration, and elevated respiratory secretion (for example, purulent sputum) with or without chest pain; 2) a fever with a body temperature greater than 37.5℃; 3) physical examination revealing signs of pulmonary consolidation and/or moist rales in the lungs; 4) peripheral blood findings showing a white blood cell count >10×10⁹/L or <4×10⁹/L, with or without a left shift in the neutrophil count; and 5) imaging examination (chest X-ray or CT) demonstrating patchy, sheet-like, globular infiltrative changes, or interstitial changes in the lungs, with or without pleural effusion. A diagnosis of pulmonary infection was made if a patient exhibited any one of the symptoms listed in items 1–4, along with item 5, after excluding other conditions such as lung cancer, pulmonary edema, aspiration or foreign body aspiration, pulmonary tuberculosis, pulmonary infarction, or non-infectious pulmonary interstitial disease. Additionally, the serum concentrations of high-sensitivity C-reactive protein (hs-CRP), procalcitonin (PCT), and interleukin (IL)-6 were measured at the time of pulmonary infection diagnosis, and only in patients who developed infection during hospitalization.

National Institutes of Health Stroke Scale

National Institutes of Health Stroke Scale (NIHSS) scoring was used to assess the neurological impairment of the participants. The NIHSS scores range from 0 to 42, with the following classifications: 0–1 for normal patients, 2–4 for mild impairment, 5–15 for moderate impairment, 16–20 for severe impairment, and 21–42 for very severe impairment.

SP level assessment

Blood samples were obtained within 24 hours of hospital admission, after a fasting period of at least 8 hours. Serum was obtained after centrifugation, and ELISA kits were used to detect the levels of SP (EH4245; sensitivity: 4.688 pg/mL, coefficient of variation: 5.17%), IL-6 (EH0201; sensitivity: 2.813 pg/mL, coefficient of variation: 4.72%), PCT (EH0341; sensitivity: 18.750 pg/mL, coefficient of variation: 5.59%), provided by Wuhan Fine Biotech Co., Ltd. and hs-CRP (CSB-E08617h; sensitivity: 0.156 ng/mL, coefficient of variation: 4.93%), provided by Cusabio Technology LLC.

Statistical methods

Data normality was assessed using the Anderson–Darling test. Continuous variables were expressed as mean±standard deviation or median (interquartile range), and categorical variables as number (%). Group comparisons were conducted using the unpaired t-test with Welch’s correction, Mann–Whitney U test, chi-square test, or Fisher’s exact test, as appropriate. Binary logistic regression analysis was used to identify independent risk factors for pulmonary infection (0=no, 1=yes). Continuous variables including age, initial NIHSS score, and serum SP concentration were dichotomized based on receiver operating characteristic (ROC)-derived cutoff values (age >66 years, NIHSS >10, SP >69.91 pg/mL). Stepwise backward elimination was applied, retaining only variables with p<0.05 in the final model. ROC analysis was also used to evaluate the predictive performance of serum SP, with the optimal cutoff determined by the maximum Youden index. Correlations between SP and inflammatory markers were analyzed using Spearman correlation. A p value <0.05 was considered statistically significant.

RESULTS

Baseline characteristics of the participants

A total of 340 eligible patients with acute cerebral infarction were included in this study. During hospitalization, 96 patients developed varying degrees of pulmonary infection, while 244 patients did not experience any pulmonary complications. Table 1 presents a comparison of baseline characteristics between the infection and non-infection groups at the time of admission. Prior to statistical analysis, data normality was assessed using the Anderson–Darling test. As a result, age and initial serum SP concentration were found to be normally distributed. In contrast, the initial NIHSS score was not normally distributed and is presented as median (interquartile range). Statistically significant differences were observed between the two groups in age (p=0.003), smoking status (p=0.039), invasive procedures (p=0.002), dysphagia (p=0.005), initial NIHSS score (p<0.001), and serum SP concentration at admission (p<0.001). Notably, the mean serum SP level at admission was significantly higher in the infection group compared to the non-infection group (85.43±36.71 pg/mL vs. 48.86±22.04 pg/mL), suggesting a possible association between elevated SP levels and the development of pulmonary infection. Additionally, the average onset of pulmonary infection is 4.2±1.6 days after hospital admission.

Table 1. Baseline characteristics of patients with and without pulmonary infection during hospitalization following-up acute cerebral infarction.

Characteristics		Non-infection (n=244)	Infection (n=96)	p
Age (yr)		62.31±9.13	65.52±10.26	0.003
Sex				0.548
	Male	120 (49.2)	51 (53.1)
	Female	124 (50.8)	45 (46.9)
Smoke		45 (18.4)	28 (29.2)	0.039
Invasive operation		87 (35.7)	53 (55.2)	0.002
Complication
	Diabetes mellitus	42 (17.2)	14 (14.6)	0.628
	Hypertension	137 (56.1)	59 (61.5)	0.395
	CHD	43 (17.6)	20 (20.8)	0.536
	Dysphagia	58 (23.8)	38 (39.6)	0.005
Cerebral infarction subtypes				0.287
	Cerebral thrombosis	133 (54.5)	56 (58.3)
	Cerebral embolism	60 (24.6)	27 (28.1)
	Lacunar cerebral infarction	51 (20.9)	13 (13.6)
Cerebral infarction locations				0.798
	Brainstem	38 (15.6)	16 (16.7)
	Lobe	51 (20.9)	23 (23.9)
	Cerebellum	44 (18.0)	19 (19.8)
	Basal ganglia region	111 (45.5)	38 (39.6)
Initial NIHSS		8 (5–12)	12 (8–17)	<0.001
Initial serum SP (pg/mL)		48.86±22.04	85.43±36.71	<0.001

Data are mean standard deviation or median (interquartile range) or n (%). The comparisons of data between the non-infection and infection groups were done by Mann-Whitney test or unpaired t-test with Welch's correction or Fisher's exact test or chi-square test.

CHD, coronary heart disease; NIHSS, National Institutes of Health Stroke Scale; SP, substance P.

Risk factors studies

Multivariate logistic regression analysis was performed to identify independent risk factors associated with the development of pulmonary infection during hospitalization in patients with first-onset acute cerebral infarction. The dependent variable was the occurrence of pulmonary infection (0=no, 1=yes). Independent variables included age, sex, smoking status, invasive procedures, comorbidities (diabetes, hypertension, coronary artery disease, dysphagia), cerebral infarction subtypes and locations, initial NIHSS score, and serum SP concentration at admission.

To improve interpretability and address potential non-linearity in the logistic regression model, continuous variables—age, initial NIHSS score, and serum SP concentration—were dichotomized based on ROC-derived optimal cutoff values: age >66 years=1; ≤66 years=0; initial NIHSS score >10=1; ≤10=0; serum SP >69.91 pg/mL=1; ≤69.91 pg/mL=0.

A stepwise backward elimination method was used for variable selection, retaining only those predictors with p<0.05 in the final model. The results indicated that age >66 years (p=0.008), undergoing invasive procedures (p=0.002), presence of dysphagia (p=0.004), initial NIHSS score >10 (p=0.001), and serum SP concentration >69.91 pg/mL (p<0.001) were all independent risk factors for developing pulmonary infection (Table 2). Notably, patients with serum SP levels above the cutoff had a 2.318-fold greater odds of developing pulmonary infection compared to those with lower levels.

Table 2. Multivariate logistic analysis for pulmonary infection during hospitalization following-up acute cerebral infarction.

	OR	95% CI	p
Age >66 years old	1.096	1.025 to 1.177	0.008
Invasive operation	1.482	1.295 to 1.613	0.002
Dysphagia	1.724	1.451 to 1.982	0.004
Initial NIHSS >10	2.271	1.783 to 2.734	0.001
Initial serum SP >69.91 pg/mL	2.318	1.532 to 3.119	<0.001

Cl, confidence interval; NIHSS, National Institutes of Health Stroke Scale; OR, odds ratio; SP, substance P.

These findings suggest that early identification of high-risk patients using clinical and biochemical indicators may facilitate timely preventive measures.

Predictive value of serum SP

Fig. 1A illustrates the comparison of serum SP concentrations at admission between 96 patients who developed pulmonary infection and 244 who did not during hospitalization following acute cerebral infarction. Serum SP levels at admission were significantly higher in the infection group than in the non-infection group (p<0.001). To assess the predictive performance of SP, a ROC curve analysis was conducted. The optimal cutoff value of 69.91 pg/mL was determined using the maximum Youden index, which identifies the point on the ROC curve offering the best balance between sensitivity and specificity. At this threshold, the sensitivity was 65.63% (95% confidence interval [CI]: 55.69%–74.36%) and the specificity was 85.66% (95% CI: 80.70%–89.50%) (Fig. 1B). The area under the curve was 0.81 (95% CI: 0.75–0.86, p<0.001), indicating good overall discriminative ability. The positive predictive value was 64.29% (95% CI: 53.17%–73.59%), and the negative predictive value (NPV) was 86.36% (95% CI: 82.23%–89.87%). Despite moderate sensitivity, the high specificity and NPV suggest that serum SP concentration at admission may serve as a useful tool for early identification of patients at lower risk of developing pulmonary infection.

Fig. 1. Serum SP concentrations and its predictive values. A: Comparison of serum SP concentrations at admission between patients with (n=96) and without (n=244) pulmonary infection during hospitalization following-up acute cerebral infarction. Data were shown with mean±standard deviation. ^***p<0.001 from unpaired t-test with Welch's correction. B: ROC analysis of predictive values of serum SP at admission for pulmonary infection during hospitalization following-up acute cerebral infarction. AUC, area under the curve; ROC, receiver operating characteristic; SP, substance P.

Correlation studies

Fig. 2A is a comparison of NIHSS score at admission between 96 patients with pulmonary infection and 244 patients without during hospitalization after treatment for the first acute cerebral infarction. The initial NIHSS scores were significantly higher in the infection group compared to the non-infection group, suggesting that patients with more severe disturbance of consciousness have a higher probability of lung infection. Spearman correlation analysis showed that the concentration of SP in the serum at admission was significantly positively correlated with the NIHSS score at admission in all acute cerebral infarction patients (r=0.38, p<0.001) (Fig. 2B). Taking 69.91 pg/mL as the cutoff value, all acute cerebral infarction patients (n=340) were divided into serum SP concentrations ≤69.91 pg/mL group (n=242) and >69.91 pg/mL group (n=98). Our results showed that patients with high serum SP concentrations at admission had significantly higher NIHSS scores (Fig. 2C).

Fig. 2. NIHSS score and correlation studies. A: Comparison of NIHSS score at admission between patients with (n=96) and without (n=244) pulmonary infection during hospitalization following-up acute cerebral infarction. Data were shown with median (interquartile range). B: Spearman correlation analysis of serum SP concentrations with NIHSS score at admission in all acute cerebral infarction patients (n=340). C: According to the cutoff value in the ROC analysis, all acute cerebral infarction patients (n=340) were divided into serum SP concentrations ≤69.91 pg/mL (n=242) and >69.91 pg/mL (n=98). NIHSS score at admission was compared between two groups. Data were shown with median (interquartile range). ^***p<0.001 from Mann-Whitney test. NIHSS, National Institutes of Health Stroke Scale; ROC, receiver operating characteristic; SP, substance P.

Correlation between serum SP concentration at admission and inflammatory markers in patients with pulmonary infection during hospitalization after treatment for the first acute cerebral infarction

To explore whether elevated SP levels at admission are associated with subsequent inflammatory response, we analyzed the correlation between serum SP concentrations at admission and the levels of hs-CRP, PCT, and IL-6, which were measured at the time of pulmonary infection diagnosis, but only in patients who developed pulmonary infection during hospitalization after acute cerebral infarction. Spearman correlation analysis demonstrated that admission SP levels were significantly positively correlated with inflammatory marker concentrations at the time of infection diagnosis: hs-CRP: r=0.44, p<0.001 (Fig. 3A); PCT: r=0.42, p<0.001 (Fig. 3B); IL-6: r=0.36, p<0.001 (Fig. 3C).

Fig. 3. Spearman correlation analysis of serum SP concentrations at admission with the concentrations of hs-CRP (A), PCT (B), and IL-6 (C) after infection in patients with pulmonary infection (n=96) during hospitalization following-up acute cerebral infarction. According to the cutoff value in the ROC analysis, patients with pulmonary infection (n=96) during hospitalization following-up acute cerebral infarction were divided into serum SP concentrations ≤69.91 pg/ml (n=33) and >69.91 pg/mL (n=63) The concentrations of hs-CRP (D), PCT (E), and IL-6 (F) were compared between two groups. Data were shown with median (interquartile range). ^**p<0.01; ^***p<0.001 from Mann-Whitney test. hs-CRP, high-sensitivity C-reactive protein; IL, interleukin; PCT, procalcitonin; ROC, receiver operating characteristic; SP, substance P.

Furthermore, based on the ROC-derived cutoff value of 69.91 pg/mL, all patients (n=340) were stratified into a low-SP group (≤69.91 pg/mL, n=242) and a high-SP group (>69.91 pg/mL, n=98). Among those who developed pulmonary infection, patients with higher serum SP levels at admission exhibited significantly elevated expression levels of hs-CRP (Fig. 3D), PCT (Fig. 3E), and IL-6 (Fig. 3F) at the time of infection diagnosis. These findings suggest that early elevation of serum SP may be associated with a more pronounced inflammatory response once pulmonary infection occurs, even though IL-6, PCT, and hs-CRP were not measured at admission and thus cannot be directly compared with SP for predictive value.

DISCUSSION

Acute cerebral infarction results from the sudden interruption of cerebral blood flow, most often due to thrombosis or embolism, and accounts for approximately 70% of all strokes.¹⁶ It causes localized brain tissue necrosis and significant neurological deficits.¹⁷ Pulmonary infection is one of the most frequent and severe complications following acute cerebral infarction, and has been associated with worsened neurological outcomes, higher mortality, and increased hospitalization costs.¹⁸,¹⁹ Therefore, identifying early predictors and risk factors of pulmonary infection in this population is critical for targeted prevention and improved clinical management.²⁰

In this study, we retrospectively analyzed 340 patients with first-onset acute cerebral infarction, among whom 96 developed pulmonary infection during hospitalization. Multivariate logistic regression revealed that age >66 years, presence of dysphagia, invasive procedures, NIHSS score >10, and serum SP concentration >69.91 pg/mL at admission were independent risk factors for pulmonary infection. These findings are consistent with prior studies showing that more severe strokes—indicated by higher NIHSS scores—are associated with reduced consciousness, impaired airway protection, and greater risk of aspiration pneumonia.²¹,²² Our results further support the use of NIHSS as a clinical indicator of infection risk, and suggest that serum SP may be a valuable adjunct biomarker.

Importantly, our study demonstrated that serum SP levels at admission were significantly higher in patients who later developed pulmonary infection, and were positively correlated with inflammatory markers hs-CRP, PCT, and IL-6 measured at the time of infection diagnosis. SP is a neuropeptide traditionally known for its vasodilatory and pro-inflammatory effects, but its specific role in pulmonary pathology is increasingly recognized.⁸ SP can bind to the neurokinin-1 receptor on various cell types including endothelial cells, macrophages, and airway epithelial cells, thereby activating intracellular pathways that promote vascular permeability, plasma exudation, and edema formation.⁸

Moreover, SP has been shown to synergize with other inflammatory mediators, such as histamine and prostaglandins, amplifying the inflammatory cascade.²³,²⁴ It enhances leukocyte recruitment by promoting chemotaxis of neutrophils, macrophages, and T cells, and stimulates immune cells to release pro-inflammatory cytokines such as tumor necrosis factor-α, IL-1β, IL-6, and interferon γ.²⁵ These actions create a positive feedback loop that can exacerbate lung injury and contribute to the pathogenesis of pulmonary infection, particularly in patients with compromised neurological function.²⁶,²⁷,²⁸ In our cohort, elevated SP at admission may reflect a pre-existing hyperinflammatory or dysregulated neuroimmune state, making patients more susceptible to lung infection once other risk factors-such as dysphagia or invasive procedures-are present. The significant correlation between SP and IL-6, PCT, and hs-CRP supports the hypothesis that SP contributes to systemic inflammatory priming.

Our findings appear to contrast with a previous study, which reported that SP may protect against aspiration pneumonia, particularly in patients receiving angiotensin-converting enzyme inhibitors (ACEi).²⁹ This discrepancy may be due to differences in study design: their study focused on ACEi-induced chronic elevation of SP enhancing protective reflexes, while our study assessed baseline SP levels at admission, which may reflect an acute inflammatory response. In addition, our endpoint was all-cause pulmonary infection, not limited to aspiration pneumonia. As ACEi use was not documented in our dataset, this may also represent a confounding factor. These findings suggest that the role of SP in pulmonary infection may be context-dependent-potentially protective in chronic settings but indicative of risk when acutely elevated post-stroke.

This study has several limitations. First, although we identified an association between elevated serum SP levels at admission and the risk of pulmonary infection, we could not determine whether this relationship is causal. Second, data on ACEi use were not available, which may have influenced serum SP levels and could confound the interpretation of our findings, especially given prior reports suggesting a protective effect of ACEi-mediated SP elevation against aspiration pneumonia. Third, the study focused on pulmonary infection as a general endpoint and did not distinguish between different subtypes, such as aspiration versus non-aspiration pneumonia. Lastly, this was a single-center retrospective study, and the results should be validated in larger, multicenter prospective cohorts.

In conclusion, this study demonstrated that serum SP concentration at hospital admission can serve as a predictor for the development of pulmonary infection during hospitalization in patients with acute cerebral infarction. Furthermore, elevated SP levels at admission were positively correlated with the severity of subsequent inflammation and the expression of inflammatory markers in patients who developed pulmonary infection. These findings may provide a theoretical basis for the early identification, prevention, and timely intervention of pulmonary infections in this high-risk patient population.

Availability of Data and Material

The datasets generated or analyzed during the study are available from the corresponding author on reasonable request.