Serum interleukin-17 predicts severity and prognosis in patients with community acquired pneumonia: a prospective cohort study

All CAP patients were enrolled from August 2019 to April 2021 in the Department of Respiratory and Critical Care Medicine of Second Affiliated Hospital. At first, all 288 patients were recruited and agreed to take participate in this research. Twenty-eight patients with incomplete information, 12 lost cases and 9 withdrawing cases were excluded from this research. Finally, 239 CAP patients were enrolled in the present research. All CAP patients were eligible for this research. After obtaining informed consent from the patients, we gathered relevant demographic characteristics and clinical information. At the onset of CAP, CAP patients always showed specific symptoms and signs, primarily including fever, cough, expectoration, and even purulent sputum; with or without chest pain; chest tightness and dyspnea in CAP patients with affected multiple pulmonary lobes. All selected 239 CAP patients in this research must meet certain diagnostic standard: In accord with the clinical manifestations of pneumonia and consistent with chest CT scans, mainly including (1) chest radiograph suggesting either a new patchy infiltrate, leaf or segment consolidation, ground glass opacity or interstitial change; (2) at least one of the following signs: the presence of cough, sputum production and dyspnoea; core body temperature higher than 38.0 °C; auscultatory findings of abnormal breath sounds and rales; or the counts of white blood cell more than 10 × 10⁹ L or less than 4 × 10⁹ L; (3) occurred in the community, rather than in a hospital [22]. The inclusion criteria were as follows: pneumonia was occurred in the community or within a definite incubation period after hospitalization; these candidates were not admitted to the hospital for the last three months; not pregnant; no other pulmonary diseases, such as pulmonary malignant tumor, pulmonary tuberculosis or immunodeficiency [23, 24]; no antibiotic treatment or intervention before hospitalization. Peripheral blood samples were collected and anticoagulated with EDTA within 24 h after hospitalization. Blood routine examination and biochemical detection were performed in all CAP patients. Then, the severity of pneumonia was evaluated using well-recognized CAP severity scores, including Pneumonia Severity Index (PSI), CURB-65, CRB-65, SMART-COP and Acute Physiology and Chronic Health Evaluation II (APACHE II). Moreover, CAP patients were divided into mild patients and severe patients through CURXO score. Severe CAP is a group of patients who requires treatment in the intensive care unit (ICU). Our research has obtained approval from Ethics Committee of Second Affiliated Hospital of Anhui Medical University (YX2021-085). Furthermore, written agreement consent from eligible patients or the patient’s next of kin were obtained.

Before any treatment, peripheral blood samples were collected and centrifuged. Then samples were stocked in the − 87 °C ultracold refrigerator until use. IL-17 ELISA kits (CSB-E12819h) were purchased from Cusabio, Wuhan, China (https://​www.​cusabio.​com/​). IL-6 ELISA kits (JYM1942Hu) were obtained from Wuhan Colorful Gene Biological Technology Co. The concentrations of serum inflammatory cytokines were detected following the instructions closely [24, 25].

In this study, 239 CAP patients were recruited. Demographic characteristics and laboratory test results were collected and generalized. The average age was 61.09 years old among CAP patients (Table 1). Among them, female patients accounted for 40.17%. The proportion of smoking in CAP patients was 17.57%. Besides, the comorbidities of CAP patients were analyzed. As shown in Table 1, 64 (26.78%) patients were with hypertension, 22 (9.21%) patients were with diabetes mellitus, 20 (8.37%) patients were with cerebral infarction, 11 (4.60%) patients were with coronary heart disease, 19 (7.95%) patients were with bronchitis and 78 (32.64%) patients were with other diseases. Blood cell content indicated that the count of white blood cell (WBC) was 8.26 × 10⁹/L, the count of neutrophil was 7.79 × 10⁹/L, the count of lymphocyte was 2.00 × 10⁹/L (Table 1). Meanwhile, liver function, renal function and myocardial function were tested among CAP patients. In addition, the median hospital stay is 10.00 days among CAP patients. During hospitalization, 71 cases admitted to the ICU, median hospital stays were 3.21 days, 66 cases were conducted with mechanical ventilation, 34 cases were used with vasoactive agent and 22 cases were died among CAP patients (Table 1). According to CURXO score, CAP patients who met the following two or more criterions were regraded as severe cases: confusion; urea nitrogen > 30 mg/L; respiratory rate > 30/min; abnormal shadows in multilobar or bilateral lungs; PaO₂ < 54 or PaO_2/FiO₂ < 250 mmHg. Severe CAP patients accounted for 27.62% in CAP patients. The mean of CURB-65, CRB-65, PSI, SMART-COP and APACHE II was 1.31, 1.07, 78.72, 1.97 and 7.95, respectively.

The concentration of serum IL-17 was measured and compared among various groups of CAP patients. The severity degrees of pneumonia were assessed by using CAP scoring system, such as CURXO, PSI, CURB-65, CRB-65, SMART-COP and APACHE II. Serum IL-17 was decreased in 0 score than those in the groups of 1–2 and ≥ 3 scores based on CRB-65 score (Fig. 1A). In accordance with SMART-COP score, serum IL-17 was lower in the group of 0–2 and 3–4 scores than those in the group of 7–8 score (Fig. 1B). IL-17 was higher in the group of 5–6 score than these in the group of 0–2 score (Fig. 1B). According to APACHE II score, IL-17 was highest in the group of > 10 score than other groups (Fig. 1C). The level of IL-17 was elevated in the group of 6 ~ 10 score compared with the group of < 4 score (Fig. 1C). Based on CURB-65 score, IL-17 was gradually increased in line with CURB-65 score (Fig. 1D). Additionally, IL-17 was lower in severe group than these in mild group (CURXO score) (Fig. 1E). Based on PSI score, serum IL-17 was gradually elevated with PSI score elevation (Fig. 1F).

The correlations between serum IL-17 and blood routine indices were analyzed among CAP patients. As shown in Table 2, serum IL-17 was positively correlated with WBC (r = 0.28, P < 0.01) and neutrophil (r = 0.31, P < 0.01), inversely associated with lymphocyte (r = − 0.13, P = 0.02) and eosinophil (r = − 0.12, P = 0.04) in CAP patients. Moreover, the relationships of IL-17 with renal function, liver function and myocardial function were evaluated among CAP patients. The results revealed that IL-17 was positively linked with uric acid (r = 0.16, P < 0.01), urea nitrogen (r = 0.13, P = 0.03), alanine aminotransferase (ALT) (r = 0.19, P = 0.02), alanine transaminase (AST) (r = 0.25, P < 0.01) and creatine kinase isoenzyme (CKMB) (r = 0.16, P = 0.03) in CAP patients (Table 2). In addition, we found that there were positive correlations between serum IL-17 with platelet count (PLT) (r = 0.13, P = 0.03), d-dimer (r = 0.22, P < 0.01), B-type natriuretic peptide (BNP) (r = 0.21, P = 0.01) and fibrinogen (FIB) (r = 0.15, P = 0.02) among CAP patients. Finally, the associations between serum IL-17 and inflammatory cytokines were accessed. As shown in Table 2, the results indicated that serum IL-17 was positively and obviously correlated with C-reactive protein (CRP) (r = 0.36, P < 0.01) and interleukin-6 (IL-6) (r = 0.37, P = 0.02).

The concentration of serum IL-17 was divided into lower than 75% quartile and higher than 75% quartile. Then, the associations were accessed between serum IL-17 and CAP severity scores through logistic regression analysis in CAP patients. Univariable logistic regression analysis found that serum IL-17 was positively associated with 3–4 score (OR = 9.16; 95% CI 3.16, 26.57), 5–6 score (OR = 4.39; 95% CI 1.90, 10.12) and 7 ~ 8 score (OR = 23.50; 95% CI 6.07, 90.95) of SMART-COP, > 10 score of APACHE II (OR = 5.30; 95% CI 2.32, 12.09), ≥ 3 score of CRB-65 (OR = 4.79; 95% CI 1.83, 12.51), 2 score (OR = 3.37; 95% CI 1.58, 7.21) and 3–5 score (OR = 5.38; 95% CI 2.47, 11.75) of CURB-65, IV grade (OR = 3.11; 95% CI 1.08, 8.96) and V grade (OR = 22.00; 95% CI 5.86, 82.65) of PSI, severe patients of CURXO (OR = 5.98; 95% CI 3.10, 11.53) (Table 3). After age and sex were adjusted, further multivariate regression analysis revealed that IL-17 was positively associated with SMART-COP, APACHE II, CRB-65, CURB-65, PSI and CURXO in CAP patients (Table 3).

Serum IL-17 was compared among CAP patients with different prognostic outcomes. As shown in Fig. 2A–C, the levels of serum IL-17 on admission were increased in patients with mechanical ventilation, vasoactive agent and ICU admission. Moreover, the levels of serum IL-17 on admission were elevated in patients with ≥ 14 hospital stays compared with patients with ≤ 8 hospital stays and from 8 to 14 hospital stays in CAP patients (Fig. 2D). In addition, serum IL-17 was further compared between alive and dead patients on admission. We found that serum IL-17 levels on admission were remarkably increased in the dead cases (Fig. 2E).

Univariate and multivariate logistical regression was performed to investigate the association between serum IL-17 and the prognosis of CAP patients. Univariable logistic regression identified that serum IL-17 on admission was positively associated with ICU admission (OR = 1.01; 95% CI 1.00, 1.02), mechanical ventilation (OR = 1.12; 95% CI 1.01, 1.42), death (OR = 1.05; 95% CI 1.00, 1.17) and ≥ 14 hospital stays (OR = 1.21; 95% CI 1.00, 1.51) (Table 4). For the sake of controlling some confounding factors, multivariate logistic regression was performed. The findings demonstrated that there were positive correlations between IL-37 with ICU admission (OR = 1.01; 95% CI 1.00, 1.01), mechanical ventilation (OR = 1.10; 95% CI 1.05, 1.37), death (OR = 1.05; 95% CI 1.00, 1.14) and ≥ 14 hospital stays (OR = 1.21; 95% CI 1.04, 1.57) among CAP patients (Table 4).

The predictive capacities of IL-17 for prognostic outcomes were accessed using receiver operating characteristic (ROC) area under the curve (AUC) among CAP patients. As shown in Fig. 3A, the AUCs for death of serum IL-17, CURB-65, CRB-65, PSI, CURXO, SMART-COP and APACHE II were 0.89, 0.87, 0.85, 0.82, 0.15, 0.91 and 0.86. The cut-off concentration of IL-17 for death was 86.80 ng/mL in CAP patients. Moreover, the predictive powers for ICU admission were determined. As shown in Fig. 3B, the AUCs were as follows: IL-17, 0.65; CURB-65, 0.86; CRB-65, 0.86; PSI, 0.79; CURXO, 0.18; SMART-COP, 0.90; APACHE II, 0.76. The cut-off concentration of IL-17 for death was 84.92 ng/mL in CAP patients. As shown in Fig. 3C, the AUCs for mechanical ventilation were as follows: IL-17, 0.74; CURB-65, 0.84; CRB-65, 0.85; PSI, 0.79; CURXO, 0.21; SMART-COP, 0.87; APACHE II, 0.78. The cut-off concentration of IL-17 for mechanical ventilation was 84.92 ng/mL in CAP patients. Finally, the predicative capacities for ≥ 14 hospital stays were evaluated in CAP patients. Except lower predicative capacity of CURXO, there were similar predicative capacities for ≥ 14 hospital stays between serum IL-17 and CAP severity scores in CAP patients. The cut-off concentration of IL-17 for ≥ 14 hospital stays was 60.29 ng/mL (Fig. 3D).