The Thioredoxin and Glutaredoxin Systems in Smoking Cessation and the Possible Relation to Postoperative Wound Complications

Abstract: Background. Smoking cessation prior to surgery has proven to be an effective means of reducing wound complications. The underlying mechanisms are poorly characterized. The present study is the first to investigate how short-term smoking cessation affects the thioredoxin and glutaredoxin systems. Methods. A total of 21 daily smokers scheduled for surgery were enrolled and randomized to smoking cessation intervention or control group. The main outcome measure was changes in plasma levels of thioredoxin (Trx) and glutaredoxin (Grx) following smoking cessation. Secondary outcomes were changes in plasma levels of TNF-a, IL-1b, and IL-6. Results. Glutaredoxin levels in plasma increased after smoking cessation. The mean Grx difference between sample 2 and sample 1 among abstainers was 11.4 ng/mL; among smokers it was 0.6 ng/mL (P = 0.05). Among those with a postoperative complication there was a slight decrease (-0.9 ng/mL in mean) in glutaredoxin, which was not statistically significant (P = 0.27). A change in thioredoxin levels in plasma after smoking cessation was not evident. Changes in cytokine levels before and after smoking cessation were not significant. Conclusion. Short-term smoking cessation seems to increase plasma levels of Grx. The Trx system seems to be more robust against smoking toxicity. Fewer wound complications occurred in the smoking cessation group and glutaredoxin might be involved in the underlying mechanism.
Address correspondence to: David Lindström, MD, PhD Department of Clinical Science and Education Södersjukhuset, Karolinska Institute 118 83 Stockholm, Sweden Phone: +46 736 200636 Email: David.lindstrom@ki.se
Disclosure: The Swedish National Institute of Public Health and the Regional County Council of Stockholm (Stockholms Läns Landsting) funded this study. Pfizer financed the nicotine replacement therapy. Neither of the funding parties took part in the design/conduct of the study; nor in collection, management, analysis, and interpretation of the data; nor the preparation, review, or approval of the manuscript. None of the authors has any financial interest in the study.

     Smoking cessation initiated prior to surgery has been shown to reduce the postoperative complication rate by nearly 65%. Two randomized, controlled trials have shown a clear benefit for smoking patients undergoing a smoking cessation program prior to surgery.^1,2 These two trials offered an intervention that started 4 to 8 weeks prior to surgery. Another small trial that investigated a shorter preoperative cessation period of 2–3 weeks did not find any reduction in postoperative complications.³ The reduction in complications is mainly found among wound healing complications, but the pathophysiological mechanisms behind this have not been elucidated.      Cigarette smoke contains more than 4700 substances including gaseous oxidants and organic radicals.⁴ Since the lungs are directly exposed to oxygen in high concentrations, it is crucial that they possess an effective defense to oxidative stress. These defense mechanisms are activated in cigarette smoking.^5,6 Thioredoxin-1 (Trx) and glutaredoxin-1 (Grx) are two major cellular antioxidant enzymes.^6–8 Trx is secreted from cells upon oxidative stress. Plasma levels are increased in inflammatory disease, acute exacerbation of asthma, heart failure, and coronary angina.^9–12 Extracellular Trx acts as a cytokine and chemokine, and high levels of plasma Trx can attenuate the local inflammatory response to an infection.^13,14 Grx is present in plasma and is secreted from the cell upon oxidative stress.15 Smokers with chronic obstructive pulmonary disease (COPD) have decreased intracellular levels of Grx in the lung compared to healthy smokers.¹⁶      It was hypothesized that changes in Trx, Grx, and cytokine levels after smoking cessation could be involved in the reduction of postoperative complications. Therefore, Trx and Grx plasma levels and the inflammatory cytokines TNF-α, IL-1B and IL-6 in smokers before and after smoking cessation were analyzed. The a priori hypothesis was that smoking cessation 4 weeks before planned surgery would normalize levels of thioredoxin, glutaredoxin, and proinflammatory cytokines.

Methods

     Setting. This study was part of a randomized clinical trial in Sweden described in detail elsewhere.^2,17 The study took place in two university-affiliated hospitals in Stockholm, Sweden. The Ethics Committee of the Karolinska Institute (Ref. No. 03-214, 215) Stockholm approved the study and it was registered at Clinicaltrials.gov (ID NCT00533000). Patients scheduled for elective surgery were asked to participate in the study. Eligible procedures were laparoscopic cholecystectomy, inguinal hernia repair, and hip or knee arthroplasty. To be eligible, patients had to be daily smokers       (> 2 cigarettes daily for at least 1 year prior to inclusion) and 18- to 79-years-old. Patients with overt alcohol or drug abuse, pregnancy, severe mental illness, dementia, and poor Swedish language proficiency were excluded.      Assignment and intervention. Patients were enrolled after giving their written informed consent. The nurse administrating the smoking cessation intervention performed the randomization on the day of inclusion. Patients were randomized in a 1:1 ratio to a control group or intervention group. The intervention was intended to start 4 weeks prior to surgery and last 4 weeks after surgery. The intervention included weekly contact with a nurse professionally trained in smoking cessation therapy and free nicotine substitution administered as self-adhesive patches, chewing gum, or microtabs (Nicorette®, GlaxoSmithKline, London) based on patient preference. Nicotine replacement therapy was the only drug offered. The control group received standard care, which aside from the neutral information given in the consent form included little to no information about smoking cessation or the potential harm of tobacco smoking.      Clinical data. Each patient filled out a questionnaire upon inclusion providing background information on the presence of comorbidities, medications, and smoking history. Questionnaires were administered throughout the study to evaluate the smoking status of all participants. Measurements of carbon monoxide (CO) in the expired air were recorded at inclusion and 10 days postoperatively² (Micro™ Smokerlyzer®, Bedfont Scientific Ltd, Rochester, UK).      Blood samples. Venous blood was drawn from an antecubital vein at the time of inclusion in the study. A second sample was drawn in the morning at the day of surgery 4 weeks later. The blood samples were immediately centrifuged at 400 g for 15 minutes. The plasma fraction was removed and divided into aliquots and stored at -70˚C until use.      Trx, Grx, TNF-a, IL-1b, and IL-6 measurements. Trx was measured in duplicate plasma samples as described previously.¹⁸ Briefly, a sandwich ELISA was conducted using a mouse monoclonal anti-Trx as the primary antibody, thus ensuring no unspecific binding to truncated Trx (Trx80).¹⁹ Subsequently, Trx standards and plasma samples were added in duplicate, and a secondary biotinylated polyclonal goat anti-Trx antibody was added. Finally, alkaline phosphates-conjugated streptavidin was added and the reaction was developed by ρ-nitrophenyl phosphate. Absorbance was recorded at 405 nm by a microplate reader (SpectraMax 250, MDS Inc, Canada). Data were analyzed using the accompanying SoftMax Pro 2.6 software. Detection limit for the Trx sandwich ELISA was 0.5 ng/mL. Intra-assay and inter-assay coefficients of variation were 4.5% and 4.7%, respectively.      Plasma levels of Grx were measured in a sandwich ELISA with a primary goat polyclonal anti-Grx1 antibody.¹⁵ The secondary antibody was the same as the primary, but biotinylated. The Grx sandwich ELISA apart from this was carried out in the same manner as was done for Trx. The detection limit for the Grx sandwich ELISA was 0.2 ng/mL. Intra-assay and inter-assay coefficient of variation were 7.3%, respectively.      Sandwich ELISAs for TNF-a, IL-1b, and IL-6 were purchased from Mabtech AB (Nacka, Sweden) and performed according to the manufacturer’s instructions. The detection limits were 1.4 pg/mL for TNF-a, 2.25 pg/mL for IL-1b, and 4.3 pg/mL for IL-6. Intra-assay coefficient of variation was < 5% for all three ELISAs. Absorbance was measured and data were analyzed in the same manner as for Trx and Grx ELISAs. Hemolysis in blood samples was negligible. Those with nondetectable values of TNF-a and IL-6 were given a value of half the detection limit in the analyses: 0.7 pg/mL and 2.15 pg/mL, respectively. Laboratory analyses were done blinded to allocation and smoking status.

Statistical Analysis

     All laboratory values were continuous and smoking status at the time of the second sample was assessed as a dichotomous variable (yes/no). Abstinence from smoking was defined by reporting of smoking zero cigarettes for the minimum period of 1 week prior to blood sample number two (and surgery). The difference between sample 1 and sample 2 was calculated and the difference was then compared according to smoking status. Additional testing was done using nicotine replacement therapy as a dichotomous variable. Normality of the difference was assessed with Shapiro-Wilk test. Mann-Whitney test was used for nonparametric data and t-test was used when applicable. The level of statistical significance was set at P < 0.05 and tests were always two-sided. Due to the small number of patients, a post-hoc power analysis was made to estimate the numbers needed to evaluate if the effect found would be true in a larger population. Outliers were defined as values between 1.5 and 3 times the interquartile range, and extreme values more than 3 times the interquartile range (Figures 1–3). Data were analyzed using SPSS version 16.0.2 (SPSS, Chicago, IL).

Results

     Baseline data. Twenty-one patients were enrolled into the study between February 2005 and November 2006. Baseline characteristics according to smoking status the last week before surgery are shown in Table 1. In total, 8 individuals became completely smoke free at the time of sample 2, while the remaining 13 were still smoking.      Postoperative complications. Among abstainers 12.5% (1/8) suffered any type of complication postoperatively, among those who continued to smoke 30.8% (4/13) had any postoperative complication (P = 0.6). One abstainer suffered from a seroma in the wound, smokers developed wound infection and hematoma, blistering of the skin surrounding the wound, urinary tract infection, and urinary retention. Glutaredoxin increase was higher among those who did not suffer any postoperative complication, 6.5 ng/mL in mean. Among those with a postoperative complication, there was a slight decrease (-0.9 ng/mL in mean) in glutaredoxin. This difference was not statistically significant (P = 0.27; Figure 3).      Glutaredoxin. Glutaredoxin levels in plasma increased after smoking cessation (n = 8) from 7.7 ng/mL (range 3.7–44.2) to 29.2 ng/mL (range 3.7–62) in median. Corresponding values for those who continued to smoke (n = 13) were 9 ng/mL (range 3.1–169.5) and 18.5 ng/mL (range 1.6–143.3), respectively. The mean difference between sample 1 and sample 2 among abstainers was 11.4 and among smokers 0.6 (P = 0.05; Figure 1).      Thioredoxin. There was no evidence of a change in plasma levels of thioredoxin after smoking cessation. The median value of thioredoxin was 3.2 ng/mL before and 4.1 ng/mL after smoking cessation (n = 8) and corresponding values for those who continued to smoke (n = 12) were 3.3 ng/mL and 3.2 ng/mL. One patient had a thioredoxin level that was not within the detection range on both samples. The mean Trx-difference between sample 2 and sample 1 among abstainers was 0.0 ng/mL and among smokers 0.2 ng/mL (P = 1.0; Figure 2).      Cytokine levels. There were no significant changes in cytokine levels before and after smoking cessation. There was a non-significant decrease in IL-6 levels in plasma after smoking cessation, IL-6 median decreased from 14.4 pg/mL to 6.2 pg/mL. Corresponding values for those who continued to smoke were 4.9 pg/mL and 7.9 pg/mL (P = 0.7, when comparing the difference between the two groups). Levels of TNF-a did not change after smoking cessation. IL-1b was not detectable in any patient (assay sensitivity 1.5 pg/mL) either during smoking or after smoking cessation.      Effects of nicotine replacement therapy. Nicotine replacement therapy was used by 5 (62%) of the abstainers and by 1 (8%) of those still smoking. Nicotine did not seem to affect the result, no significant changes in glutaredoxin, thioredoxin, IL -6 or TNF-a were found among users versus non-users (data not shown).

Discussion

     This study indicates that smoking cessation increases glutaredoxin levels in plasma. It also shows that short term smoking cessation does not affect thioredoxin levels in plasma. Several possible mechanisms are involved in defect wound healing and increased susceptibility to infection in smokers. The finding that smoking cessation is a powerful tool to reduce complications,²⁰ indicates a causal relationship between abstaining and fewer complications. The strong effect of smoking cessation could partly be explained by a higher oxygen delivery to the healing tissues. Oxygen delivery is vital to wound healing.²¹ Oxygen delivery is impaired among smokers. This hypothesis is supported indirectly by the positive effect seen with supplementary oxygen delivery.^22,23 Smoking also alters the immune system²⁴ and the elevated levels of white blood cells has been shown to be a reversible condition.²⁵ Also, collagen synthesis and protein deposition is disturbed among smokers.^26,27 The elevated levels of Grx immediately preoperatively after 4 weeks of smoking abstinence may be connected to enhanced defense capabilities regarding surgical trauma. This possible connection warrants further investigation.      The present study’s findings align with previous findings that the Trx system is more robust against oxidative stress compared to Grx.^7,28 Little is known about the effects of smoking on the Grx system and the possible importance of the Grx system in humans. Some data indicate that the high expression of Grx in the lungs may be a part of a defense mechanism against hazardous agents.⁵ In-vitro and animal studies have been conducted regarding the effect of smoking on the Trx system. When transgenic mice over-expressing Trx were exposed to cigarette smoke, the inflammatory response was suppressed compared to normal mice.²⁹ Exogenously added Trx suppressed the inflammatory response in normal mice subjected to cigarette smoke.³⁰      Any possible effects of smoking cessation on IL-1b, IL -6, and TNF-a were not detectable in the present study, which addresses one important limitation of this trial—the small number of patients increases the risk of making a type II error, that is, not being able to detect a possible true effect. A post-hoc power analysis revealed that 20 patients in each group would have been needed to detect a Grx-difference at the same magnitude that we found with a power of more than 80%. Another limitation of this study is the possibility that some patients might have lied about their cigarette consumption. On the other hand, smoking data in this trial were consolidated by CO measurement in exhaled air pre- and postoperatively. Even if a smoking patient were misclassified as a non-smoker at the time of sample two, this would bias our estimates versus zero and weaken the effect shown. In an analysis of self-reported smoking status, 10% in the control group falsely reported non-smoking status versus 6% of those included in a smoking cessation program.³¹ Another limitation in the study is the relatively low sensitivity in the immunoassays for TNF-a, IL-6, and IL-1b, but in contrast, previous findings corroborate the results of the present study regarding Grx-levels.^15,32

Conclusion

     Smoking cessation increases plasma levels of glutaredoxin. The thioredoxin system seems to be more robust against smoking toxicity. Fewer wound complications occurred in the smoking cessation group and glutaredoxin might be involved in the underlying mechanism.

Acknowledgments

     The authors thank the Perioperative Smoking Cessation Study Clinical Centers, Danderyd Hospital, Department of Surgery and Orthopedics, Södersjukhuset, and the Department of Surgery and Orthopedics. We are very grateful to the nurses who worked on this study. We also thank Lotta Larsson and Christina Häll at the Metabolic laboratory (Södersjukhuset) who collected, centrifuged, and froze the samples.