Smoking is unequivocally the most significant and potentially modifiable risk factor, contributing not only to public health issues but also to an economic burden of over $ 193 billion in annual healthcare costs in the United States (Truntzer et al., 2015[92]; Scherubl 2021[75]; CDC, 2008[11]). According to a study conducted by the Robert Koch Institute (RKI) in 2022, approximately 55 % of adults in Germany are smokers, with the group being divided into former (26.7 %) and active smokers (28.9 %) (Starker et al., 2022[87]; RKI[70]). Furthermore, it is common knowledge that smokers have longer hospital stays than non-smokers (Rezaei et al., 2016[69]; Ehnert et al., 2019[24]; Abate et al., 2013[1]). Smoking represents a major risk factor for a plethora of diseases such as cancer, cardiovascular and respiratory conditions and reproductive disorders (Ehnert et al., 2019[24]). In addition, smoking also affects bone metabolism and regeneration, resulting in reduced bone mass and, accordingly, a higher rate of fragility fractures (Daniell, 1976[18]; Scolaro et al., 2014[78]). Moreover, there is strong evidence that smoking causes post-surgical complications including impaired wound and delayed fracture healing (Hess et al., 2020[33]; Mills et al., 2011[53]). These complications include a twofold increase in the risk of non-union and infections (Pearson et al., 2016[61]; Gaston and Simpson, 2007[27]). As a consequence, smokers experience extensive revision surgery, resulting in additional pain and loss of function in the affected limb. Due to the prolonged rehabilitation, smoking represents a critical strain on our healthcare system (Kanis et al., 2005[40]). In 2018, smoking incurred direct costs of approximately € 30 billion in Germany, with smoking-related medical costs alone amounting to € 27 billion (Effertz, 2019[23]).

In general, adults aged 65 and older are much less likely to smoke than younger age groups. As a result, smokers tend to be younger than ex-smokers (Starker et al., 2022[87]). The inflammatory response is a crucial prerequisite for the successful process of wound healing and regeneration (Loi et al., 2016[50]). Interestingly, there is contradicting information on the effects of smoking on the expression of inflammatory markers. Javed et al. (2015[39]) demonstrated that interleukin (IL)-1β and IL-6 levels are significantly higher in the salivary of smokers when compared to non-smokers. Moreover, Suzuki et al. (2016[88]) showed that smokers exhibit higher levels of immunoglobulin A (IgA) and tumor necrosis factor (TNF)-α in the saliva than non-smokers. In contrast, several other studies indicate that smoking inhibits toll-like receptor (TLR)-2 and-4-dependent expression of pro-inflammatory cytokines and compromizes immunity after bacterial infection (Chen et al., 2007[12]; Lugade et al., 2014[51]). This is supported by epidemiological studies indicating that smokers are prone to developing infections after injury (Knapik and Bedno, 2018[43]). However, there is no information on the cytokine and immune checkpoint protein profile of non-smokers and smokers after musculoskeletal surgery and how cytokine and protein levels may serve to predict the healing process.

Therefore, this study aimed at analyzing the blood circulating cytokine and immune checkpoint protein profile of smokers and non-smokers prior to a musculoskeletal surgery and its relation to the clinical outcome, in order to identify a potential marker predictive of complications in smoking patients, as not all smokers develop them.

Smoking can cause inflammatory and autoimmune diseases through various mechanisms, including genetic changes, increased oxidative stress and free radicals (Wang et al., 2022[94]). Furthermore, smoking is in general a well-established risk factor for osteoporosis, fragility fractures, and related post-surgical complications such as infections or delayed wound healing (Mills et al., 2011[53]; Hess et al., 2020[33]; Rudang et al., 2012[72]; Sloan et al., 2010[84]; Scolaro et al., 2014[78]; Abate et al., 2013[1]; Gronkjaer et al., 2014[29]; Sorensen, 2012[85]; Knapik and Bedno, 2018[43]; Porter and Hanley, 2001[66]). The literature contains numerous theories on the impact of smoking on fracture healing including reduced blood supply, increased levels of reactive oxygen species, lack of antioxidants and vitamins, and toxic effects on osteoblasts/osteoclasts and calcitonin (Sloan et al., 2010[84]; Gaston and Simpson, 2007[27]; Abate et al., 2013[1]).

Moreover, smoking has been proven to prolong the healing process of fractures and wounds (Abidi et al., 1998[2]; Schmitz et al., 1999[76]) and it is well established that both, direct and indirect exposure to smoke has a detrimental impact on bone mineral content (Abate et al., 2013[1]). It is suggested that smoking reduces bone strength by increasing the size of the diaphyseal medullary cavity and decreasing the size of the epiphyseal trabecular bone, although the exact cause remains unknown (Wust et al., 2010[102]). Smoking notoriously impairs wound healing (Mills et al., 2011[53]), but the mechanisms behind this phenomenon are not yet understood.

As demonstrated in this study, there is a correlation between smoking and postoperative complications. Such occurrences are also significantly more prevalent among individuals with a low socioeconomic status. Furthermore, an elevated risk of morbidity and mortality is observed (Bartley 1994[7]; de Jager et al., 2022[19]; Mehaffey et al., 2020[52]; Reames et al., 2014[68]; Bennett et al., 2010[8]). Nevertheless, a 2012 study demonstrated that there was no association between educational status and mortality when income was taken into consideration and that lower income was positively associated with mortality after adjustment for education (Sabanayagam and Shankar 2012[74]). However, a review of the literature did not identify any evidence of an increase in inflammatory markers.

Obesity, which is associated with a low socioeconomic status (de Jager et al., 2022[19]; Bartley, 1994[7]; Bambra and Eikemo, 2009[5]) is associated with an increase in several inflammatory markers, leading to chronic low-grade inflammation. Adipose tissue in obese has been shown to secrete pro-inflammatory markers due to a shift in immune cells from anti-inflammatory M2 macrophages (producing IL-10, IL-5 and interferon-γ) to pro-inflammatory M1 macrophages (TNF-α, IL-17, IL-6 and IL-1β). This leads to altered serum levels of the aforementioned inflammatory cytokines, which have been shown to exert significant effects on diabetes, hypertension, and other chronic conditions. The increasing inflammatory reaction in obesity can be explained at least in part as follows: the enlargement of fat cells leads to an increase in the distance to the nourishing vessels, creating hypoxic conditions. This could lead to fibrosis of the adipose tissue and the resulting infiltration of macrophages, which trigger the local inflammatory reaction. Furthermore, adipose tissue can secrete approximately 50 different hormones, affecting the immune system and metabolism (Khanna et al., 2022[41]; Ellulu et al., 2017[25]).

Although several factors have been shown in the literature to be associated with an increased risk of complications, we decided to focus on smoking, since our previous study cohort of 817 participants confirmed that smoking is a major risk factor for complications like infection, wound healing, need for revision surgery, thrombosis, and/or death. Compared with participants with risk for malnutrition or daily alcohol intake, the incidence of complications after total joint arthroplasty was higher in former and active smokers (Ehnert et al., 2019[24]).

For the healing process to occur, growth factors and cytokines need to be secreted for instance by fibroblasts migrating into the wound area (Barrientos et al., 2008[6]; Werner and Grose, 2003[98]; Wang et al., 2024[95]). However, smoking inhibits the biological function of fibroblasts directly (Wong and Martins-Green 2004[100]; Sloan et al., 2010[84]; Nakamura et al., 1995[56]). The importance of these cells is clearly demonstrated by a study conducted by Oe et al. (2007[57]), in which they isolated fibroblast-like progenitor cells from the hematoma of tibial fractures, which could be differentiated in vitro into osteocytes, chondrocytes and adipocytes. These findings underscore the critical role of these cells in the regeneration of bone tissue.

Other studies have shown that nicotine directly alters a number of cytokine genes associated with neovascularization and osteoblast differentiation, such as vascular endothelial growth factor (VEGF), bone morphogenetic protein (BMP)-2, -4 and -6, basic fibroblast growth factor (bFGF) and collagen I/II (Theiss et al., 2000[90]). Smoking thus disrupts the homeostasis of the immune system by modulating immunoregulatory activities leading to inflammation (Wang et al., 2022[94]), which is controlled by a complex network of secondary messengers, such as cytokines and chemokines, being part of both innate and adaptive immunity (Wong and Fish, 2003[101]; Wang et al., 2022[94]; Commins et al., 2010[17]).

However, the precise mechanism by which smoking affects inflammation remains unclear. Thus, we examined the impact of smoking on cytokine and protein levels prior to musculoskeletal surgery to identify a blood parameter that can be measured in a routine laboratory and accurately predict complications.

The present study involved 897 patients with various musculoskeletal diseases who had to undergo surgery and were categorized into three groups, based on their smoking status: non-smokers, ex-smokers, and active smokers. The data were then analyzed, including the rate of complications, gender distribution, length of hospital stay and the score of the systemic immune inflammatory response. Moreover, the levels of various cytokines and proteins in the patients' serum were measured via arrays and more specified by ELISAs.

We were able to show, that over 50 % of the patients who attend our clinic are smokers, either former smokers (31.6 %) or current smokers (22.4 %), which is in accordance with the 55 % smoking rate reported by the Robert Koch Institute (RKI) (Starker et al., 2022[87]; RKI[70]). Moreover, our study confirms the link between smoking and a higher rate of complications, consistent with the existing literature (Scolaro et al., 2014[78]). Due to the higher complication rate, the hospital stay was longer compared to non-smokers and former smokers, which is also supported by previous research (Abate et al., 2013[1]).

Upon closer examination of the hospital stay, it is evident that active smokers experience longer hospitalizations than non-smokers and former smokers, even in the absence of complications, which suggests a delayed post-surgical recovery. This is the case even though smokers are, on average, younger than ex-smokers and have a lower BMI and FINDRISC, as demonstrated by our data. However, the hospitalization of ex-smokers is also prolonged to some extent, which is likely to be due to their older age, higher BMI and FINDRISC compared with the rest of the cohort. Nevertheless, with shorter hospital stays and lower rates of complications, former smokers seem to have a better recovery than young and active smokers.

A more detailed look at the active smokers in our study cohort shows, that the proportion of men in this subgroup is higher than that of women. The calculated ratio of women to men, approximately 0.62, is consistent with the approximately 0.6 ratio reported by the World Health Organization for smokers in Germany (Hitchman and Fong, 2011[34]). Moreover, smokers are more prone to be hospitalized for traumatological and infectious events than non-smokers, which indicates both a compromized immune system and poorer bone quality. In contrast, a significant number of non-smokers were hospitalized for elective surgeries, such as total joint arthroplasty.

Since not all smokers develop a complication, we then measured cytokine and protein levels in the patients' blood to identify a potential predictive factor.

The RayBiotech® Human Cytokine and Immune Checkpoint Array were performed to provide an overview of possible alterations and discrepancies in cytokine or protein levels between non-, ex- and active smokers. The control and complication groups were matched for gender, BMI, alcohol consumption, diabetes risk and age in order to ensure that any observed differences were not attributable to these factors. Subsequently, some of these molecules were quantified using ELISA. A preliminary literature search was therefore conducted with the objective of identifying existing findings on the molecules analyzed here. Following the identification of 13 different molecules in the literature, such as CD28, B7-1, CTLA-4, MIG, TGFβ2 and IL-1α/β, either shown to alter in smokers or affecting bone metabolism and remodeling, the study proceeded to quantify these molecules by ELISA (Wang et al., 2022[94]; Ehnert et al., 2019[24]; Haytural et al., 2015[32]; Miyamoto et al., 2009[54]; Onoe et al., 1996[58]; Axmann et al., 2008[4]; Ahmadi et al., 2020[3]; Xie et al., 2014[103]).

None of these cytokines have been investigated in relation to smoking status and the development of postsurgical complications. Consequently, this study was undertaken in order to address this issue and proceeded to quantify the aforementioned molecules in serum by ELISA.

The use of serum was deemed appropriate, given that studies on the effects of sample treatment on the stability of cytokines had demonstrated that serum and plasma samples yielded comparable results, particularly when frozen without delay following collection (Flower et al., 2000[26]; Skogstrand et al., 2008[82]). As part of the study, Skogstrand et al. were able to demonstrate that prolonged storage of blood before separation into plasma or serum leads to an increase in the measurable concentration of several inflammatory markers, depending on the storage time and temperature. Therefore, a treatment at 4 °C and isolation of plasma and serum immediately after sample collection is recommended (Skogstrand et al., 2008[82]). It is important to note that the procedure for processing the blood samples in this case was highly standardized, with the samples being handled within the first hour of collection, during which time they were stored on ice. In contrast, other studies have demonstrated that the concentrations of cytokines in plasma differ from those in serum, which was attributed to the blood-clotting processes, given that a number of cytokines are released during this process (Guo et al., 2013[31]; Leng et al., 2008[46]). Other researchers have proposed that these discrepancies between plasma and serum concentrations can be explained by the plasma processing, whereby proteins such as fibrinogen and others are removed. This process also results in the removal of circulating proteins, such as cytokines, which frequently bind to these blood components (Leng et al., 2008[46]). Furthermore, the thawing of frozen samples was observed to result in a notable alteration in the concentrations of cytokines, while serum appeared to be less susceptible to this phenomenon compared to plasma (Parkitny et al., 2013[60]). Due to the aforementioned and in light of the previous findings conducted by our research group indicating a correlation between smoking and reduced levels of pro-inflammatory cytokines in serum, we elected to utilize the same test material in this investigation (Ehnert et al., 2019[24]).

CD28, B7-1 and B7-2 play a crucial role in the activation and differentiation of T-cells as essential costimulatory molecules (Chen and Flies, 2013[14]; Slavik et al., 1999[83]) and it is widely known that bone cells are under the influence of immune cells, particularly T-lymphocytes (Weitzmann and Pacifici, 2005[97]). Active T-cells, for example, can be destructive to bone and are a major source of RANKL (receptor activator of NF‐κB ligand) (Grassi et al., 2011[28]). Furthermore, Grassi et al. (2011[28]) defined osteoclasts as immunocompetent cells that attract and retain T-cells on the bone surface. B7-1, for instance, is significantly upregulated during osteoclast maturation, showing a fivefold increase, similar to APC (antigen-presenting cells) (Grassi et al., 2011[28]).

In contrast, CTLA-4 (cytotoxic T lymphocyte antigen 4) is a co-inhibitory receptor that regulates the surface expression of B7-1 and B7-2 (Chen and Flies, 2013[14]; Qureshi et al., 2011[67]). CTLA-4 downregulates both B7-1 and B7-2 through trans-endocytosis (Qureshi et al., 2011[67]) and it promotes the activity of regulatory T-cells by binding to B7-1 and B7-2 and inhibits conventional T-cells (Chen and Flies 2013[14]; Rowshanravan et al., 2018[71]). Previous studies have shown increased levels of CTLA-4 have been found in patients with chronic obstructive pulmonary disease (COPD), a condition commonly associated with tobacco smoking (Shen et al., 2013[79]; Christenson et al., 2022[15]). Our study demonstrated that non-smokers with 0 pack-years and former smokers exhibit a reduced concentration of CTLA-4 when complications arise, compared to the corresponding control group. In contrast, smokers showed an increase in CTLA-4 levels, indicating impairment of the immune system in smokers (Zimmermann et al., 1997[109]; Simone et al., 2014[81]). For CD28, B7-1, and B7-2, active smokers showed slightly lower levels compared to non-smokers. Also, no significant difference was detected in patients with complications.

Furthermore, we measured serum levels of MIG (monokine induced by interferon-γ (IFN-γ). In 2022, a study investigating the effects of smoking on the level of inflammation-related cytokines, found a significant increase in MIG, in human serum in both active and former smokers (Wang et al., 2022[94]).

In the current study, we repeated this experiment and analyzed MIG levels in relation to complications. Unlike Wang et al. (2022[94]) we clearly demonstrated that smoking cessation leads to a decrease in serum levels of MIG. Also, MIG levels are lower in active smokers when complications occur compared to non-smokers with complications.

It is important to note that MIG has been reported to have an osteoimmuno-modulating function (Phan et al., 2022[63]). Along with its receptor CXCR3, which is expressed on activated T-, B-cells and endothelial cells (Kwak et al., 2005[45]), it has been detected in osteoblasts and osteoclast precursors and has an inductive function on osteoclastogenesis leading to bone resorption (Phan et al., 2022[63]). It also impedes the proliferation, differentiation and mineralization of bone marrow stem cells (Huang et al., 2016[37]). Interestingly, MIG expression is directly induced by RANKL during osteoclastogenesis (Kwak et al., 2005[45]), which is an essential ligand for osteoclastogenesis (Yasuda, 2021[105]).

Our study and others have unequivocally demonstrated that smoking increases the proinflammatory response, as demonstrated by the rise in MIG levels (Wang et al., 2022[94]). Moreover, inflammation is associated with osteoclastic bone resorption (Kwak et al., 2005[45]), which implies that smokers with increased inflammation and higher serum MIG levels are at an elevated risk of fractures. Phan et al. (2022[63]) demonstrated this in a study by identifying an increased risk of hip fracture in men with high MIG levels, but not in women, though they would be expected to be at increased risk in the context of postmenopausal osteoporosis. However, comparing MIG levels between men and women, there were no differences within our cohort (data not shown).

Blocking MIG in the bone marrow with a neutralizing antibody prevented bone loss by inhibiting resorption and increasing bone formation (Liu et al., 2020[49]). Furthermore, it led to a significant reduction in CD4⁺ T-lymphocyte infiltration, which demonstrates the immunomodulatory effect of MIG in cardiac allograft vasculopathy (Yun et al., 2002[106]). Interestingly, scientists have shown that MIG interacts with vascular endothelial growth factor (VEGF), preventing it from binding to endothelial cells and osteoblasts, which in turn inhibits bone angiogenesis and osteogenesis (Huang et al., 2016[37]). Moreover, MIG appears to reduce the formation of type H blood vessels in bone, which is related to a loss of osteoprogenitor cells and a reduction in bone mass (Huang et al., 2016[37]; Liu et al., 2020[49]; Kusumbe et al., 2014[44]).

In the present study, we found a previously undescribed downregulation of MIG in the occurrence of complications compared to the respective control group. This was unexpected as the complication that has occurred is expected to lead to MIG-induced chemotaxis and hence recruitment of T-lymphocytes. Furthermore, our data suggest, that smoking cessation leads to a decrease of MIG. In order to assess the diagnostic potential of the measured proteins, receiver operating characteristics (ROC) were calculated. Notably, MIG exhibited the highest AUC of approximately 0.7. The analysis revealed a high sensitivity of 0.963, indicating a strong ability to accurately identify true positive cases. However, the specificity was only 0.337, suggesting a relatively high rate of false positive results. Due to this discrepancy, the analysis resulted in a Youden index of 0.3, indicating a low precise benefit of the diagnostic test. Similar situations were observed for the other proteins analyzed.

The implementation of a predictive marker could facilitate a more accurate estimation of the individual risk of postoperative complications, thereby enabling evidence-based clinical decision-making and the initiation of suitable preventive measures. In the most favorable circumstances, this could result in cost savings and the more effective utilization of resources.

In order for a marker like MIG to be considered a valid predictor, it must satisfy the following criteria: firstly, a correlation between the biomarker, the patient's smoking behavior and the occurrence of complications must be identified. Furthermore, in order to enhance the diagnostic accuracy of the test, the marker should be capable of exhibiting both high sensitivity and specificity. Finally, in order for the results of the test to be reliable, it must be guaranteed that there is no inconsistency in the results of repeated measurements conducted under different conditions and circumstances. It is therefore essential to ensure that other parameters have minimal impact on the marker.

Provided the aforementioned criteria are met, it would be possible to predict complications, which would be of great clinical relevance. This could facilitate the adaptation of therapy, reducing the burden on hospitals and healthcare systems in terms of personnel and financial resources. Nevertheless, it is important to consider that the measurement of the marker must also be economically justifiable. Unfortunately, MIG was found to be unsuitable as a predictive indicator, exhibiting an extremely low level of specificity (0.337).

On closer inspection of the arrays and ELISA results for CD28, B7-2, MCP-1, IL-13 and TIMP-1, opposite results are apparent. This is due to the fact that the array groups were adjusted for age, BMI, FINDRISC and Audit-C, whereas no adjustment was made for the ELISA analysis. Thus, it is possible that age may influence blood serum levels of cytokines like IL-13, which has been described in a previous study (Li et al., 2014[48]). MCP-1, on the other hand, is known to be elevated in obesity and diabetes mellitus (Catalan et al., 2007[10]; Kim et al., 2006[42]; Piemonti et al., 2009[65]; Simeoni et al., 2004[80]), while high TIMP-1 serum levels were found in adolescents with alcohol intoxication (Zdanowicz et al., 2022[107]). This demonstrates the difficulty and challenges in identifying an appropriate predictive marker, given that cytokine and protein levels can be affected by various external factors, including age, diet, lifestyle, pre-existing illnesses and medication.

Notably, our study revealed that the SII levels of active smokers were significantly higher than those of non-smokers and ex-smokers. This finding is important for patients undergoing musculoskeletal surgery and suffering from infections. We demonstrated a strong correlation between smoking and an elevated systemic inflammatory response. Additionally, our findings indicate that smokers experience a prolonged inflammatory response, potentially increasing their susceptibility to inflammation-related illnesses due to their higher SII.

These higher levels were more frequently found in the group of active smokers, but not in the ex-smokers or non-smokers. The SII index, which can be calculated from platelet, neutrophil and lymphocyte counts, has been associated with various cancers, including colorectal carcinoma, small cell lung carcinoma, and cervical cancer (Hong et al., 2015[35]; Huang et al., 2019[38]; Chen et al., 2017[13]). Elevated levels of SII have also been described in postmenopausal women with osteoporosis (Zhang et al., 2023[108]; Tang et al., 2022[89]; Du et al., 2021[22]). In a study, Zhang et al. (2023[108]) detected a negative correlation between SII and bone mineral density.

In our study, we were able to show that smokers have significantly higher SII levels overall, but that these levels do not differ when a complication occurs. This may be explained by the fact that there is already an immune response to a trauma or a chronic infection, which cannot be further increased due to exhaustion (Yang et al., 2023[104]; Wherry and Kurachi, 2015[99]).

The study's findings confirmed that smoking has a detrimental effect on postoperative recovery and immune response. This is in line with previous research that has demonstrated an increased risk of postoperative complications, including infections, wound healing disorders and bone healing disorders (Hess et al., 2020[33]; Mills et al., 2011[53]; Pearson et al., 2016[61]; Gaston and Simpson, 2007[27]). Smoking may reduce blood flow and oxygen supply to tissues, which can lead to delayed or impaired wound and bone healing, potentially explaining the higher complication rate and longer hospital stay in smokers (Leow and Maibach, 1998[47]; Sorensen et al., 2009[86]; Gullihorn et al., 2005[30]).

Additionally, the study reveals that smoking modifies the expression of various proteins that regulate inflammation and the immune response. One explanation for the change in expression in smokers is that smoking causes chronic systemic inflammation (Caliri et al., 2021[9]; Wang et al., 2022[94]), which leads to immune dysregulation. However, none of the proteins analyzed is really suitable as a predictive factor for complications in order to be used as a predictive blood marker. This is probably because protein levels can be affected not only by smoking but also by other factors such as pre-existing conditions, medication, or alcohol consumption. Therefore, it may be impossible to identify a single predictive factor.

When interpreting the results, some limitations should be considered. We measured a limited subset of proteins involved in inflammation and immune response via ELISA according to our array data. To obtain a more complete picture of immunological changes in smokers, a more comprehensive analysis of the entire cytokine and immune checkpoint protein profile would be recommendable. Secondly, we only considered short-term postoperative outcomes, which may not reflect the long-term effects of smoking on recovery and immune response. A longer follow-up of patients is necessary to assess the long-term consequences and cost-benefit analysis of smoking and its cessation.

In summary, this study demonstrates the impact of smoking on post-musculoskeletal surgery outcomes. The findings indicate that smoking is linked to a higher rate of complications, longer hospital stays, and increased SII levels. The results unequivocally demonstrate the detrimental effects of smoking on post-surgery recovery and highlight the urgent need for smoking cessation interventions. Additionally, the study reveals that smoking affects various proteins. However, none of the analyzed proteins is suitable for reliably predicting complications in smoking patients.