Smoking increases risk of complication after musculoskeletal surgery: analysis of single immune parameter to predict complication risk

Abstract

Smoking is the most significant and modifiable risk factor for a range of conditions, including cancer, cardiovascular and respiratory diseases. Furthermore, it significantly reduces bone mass and increases the risk of fragility fractures due to its detrimental effects on bone metabolism and regeneration. Moreover, smoking is a known cause of chronic systemic inflammation, leading to an imbalance of cytokines. Comprehending the pathological mechanisms that underlie cytokine production and its impact on post-surgical healing is essential to prevent post-surgical complications. The present study recruited a total of 1144 patients, including 897 patients, among them non-smokers (N = 413), current smokers (N = 201) and ex-smokers (N = 283). Human proteome profiler arrays were used to screen for smoking-dependent differences in the serum cytokine and protein profiles, after matching samples for age, gender, body mass index (BMI), alcohol use, and diabetes risk. Cytokines and immune checkpoint proteins such as CD28, B7-1, MIG, TGFβ2 and IL-1α/β were quantified by ELISA. Our study demonstrates a comprehensive understanding of the relationship between smoking, the development of complications, the systemic immune inflammation index (SII) and cytokine/protein levels. We found that a comparison of non-smokers, former smokers, and active smokers in our study cohort did not exhibit significantly altered cytokine and protein serum levels although other studies reported differences between smokers and non-smokers. We were unable to identify single blood circulating markers that could predict complications in smokers after trauma. However, we found the ratio of women to men to be inverted between non-smokers and active smokers resulting in a ratio of 0.62 in smokers. Furthermore, we demonstrate a higher complication rate, longer hospitalizations and elevated SII values among smokers, indicating an involvement of the immune system. See also the graphical abstract(Fig. 1).

Keywords: SII; complications; smoking; surgery; systemic immune inflammation index; trauma.

Table 1. Overview of complications

Table 2. Overview of the performed enzyme-linked immunosorbent assays (ELISAs)

Figure 1. Graphical abstract

Figure 2. Overview of the study population. (a) CONSORT Flow Diagram: A total of 1144 patients were interviewed for this study during their hospital stay in the trauma surgery department between July 2020 and October 2022. Due to pre-existing/diagnosis- or procedure-related complications, 247 patients were excluded from the study. Based on their smoking behavior, they were divided into three different groups. (b) Patients' age is given in years. (c) Patients' BMI is given in kg/m². (d) Alcohol Use Disorders Identification Test (AUDIT-C). (e) Finnish Diabetes Risk Score (FINDRISC). Data displayed as boxplots (Box and Whiskers-Tukey to visualize outliers), significant differences were determined by one‐way analysis of variance (ANOVA) with a p-value of less than 0.05 considered as statistically significant (* p < 0.05, ** p < 0.01, *** p < 0.001 and **** p < 0.0001).

Figure 3. Complication rate, gender and departmental distribution in the study cohort visualized as a bar plot. (a) The complication rate is given in % and the total number (N), significance was determined by Chi-square test. (b) Gender distribution within the groups is given in % and total numbers (N), significant differences were determined by Chi-square test with a p-value of less than 0.05 considered as statistically significant (* p < 0.05, ** p < 0.01, *** p < 0.001 and **** p < 0.0001). (c) Departmental distribution within the groups is given in % and total numbers (N).

Figure 4. Duration of hospitalization of the study population based on smoking behavior. For the analysis, a distinction was made between non-smokers with 0 pack-years (N=413), ex-smokers (≥ 1-year ex-smoker regardless of PY, N=283) and active smokers (N=201). The active smokers were divided into moderate (< 15 pack-years, N=96) and heavy smokers (≥ 15 pack-years, N=105). Data are summarized as mean ± SEM; significant differences were determined by one‐way analysis of variance (ANOVA) with a p-value of less than 0.05 considered as statistically significant. “*” indicates the significance of the entire group, while “#” denotes the significance of the population without complication and “§” shows the significant differences of the group with complication (*, #, § p < 0.05, **, ##, §§ p < 0.01, ***, ###, §§§ p < 0.001 and ****, ####, §§§§ p < 0.0001).

Figure 5. Correlations between smoking behavior and SII. The Systemic Immune-Inflammation Index (SII) of non-smokers (0 PY), ex-smokers and active smokers is displayed as a boxplot (Box and Whiskers-Tukey). The SII was calculated using the following equation, SII = P x N/L, blood cell counts of P platelets, N neutrophil and L lymphocytes, significance was determined through one-way analysis of variance (ANOVA) with a p-value less than 0.05 being considered statistically significant (*p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001) as the data followed a normal distribution. (a) Comparison of the SII levels (with and without complication) between non-, ex- and active smokers. (b) SII values for non-, ex- and active smokers, distinguishing between complications and no complications.

Figure 6. Effect of smoking on circulating factors in the blood. (a) Relative cytokine and immune checkpoint protein levels in serum samples from non-smokers (0 PY), ex-smokers and active smokers were determined using the RayBiotech® Human Cytokine Array C5 and Human Immune Checkpoint Array C1 (0 PY: N = 32, ex: N = 25, active: N = 26). For the heatmap, signal intensities were normalized using z-score by following the equation x' = (x - μ) / σ, defining μ as the mean and σ as the standard deviation of the samples. Underrepresented cytokines and proteins are colored blue, overrepresented cytokines/proteins are colored red. (b-f) Description of the study population for the C5 and C1 Array measurements based on smoking behavior with analyses of age, BMI, Audit-C score, FINDRISC and gender. Data shown as boxplot (Box and Whiskers) or bar plot, significance was determined by Student's t-test. (b) Patients' age is given in years. (c) Patients' BMI is given in kg/m². (d) Audit-C score (Alcohol Use Disorders Identification Test) for the detection of patients with hazardous alcohol consumption or active alcohol use disorders. (e) FINDRISC (Finnish Diabetes Risk Score) to identify patients at high risk for type 2 diabetes. (f) Gender distribution within the groups, given in % and total numbers (N). P-values are given at the top of the brackets.

Figure 7. Impact of smoking on serum levels of several proteins. Differentiation was made in the analysis between non-smokers, ex-smokers and active smokers. All ELISAs were performed in duplicate. Error bars indicate the median ± 95 % confidence interval and significance was determined by Kruskal-Wallis H-test with a p-value of less than 0.05 considered statistically. The serum levels of the MCP-1 (a), Eotaxin 3 (b), CD28 (c), B7-1 (d), B7-2 (e), CTLA-4 (f), MIG (g), IL-1α (h), IL-1β (i), IL-13 (j), PDGF-BB (k), TGFβ2 (l) and TIMP-1 (m) were measured in both the complication and control groups. (n) Receiver operating characteristics (ROC) curve for MIG (AUC = 0.70), (o) Receiver operating characteristics (ROC) curve for CD28 (AUC = 0.53), IL-1α (AUC = 0.64), IL-13(AUC = 0.59) and PDGF-BB (AUC = 0.64).