Hyperoxia during colon surgery is associated with a reduction of xanthine oxidase activity and oxidative stress in colonic mucosa

Abstract

Background

Gut manipulation during surgery may induce local oxidative stress. Supplemental inspired oxygen increases arterial and tissue oxygen partial pressures. This study examined whether a 0.80 fraction of inspired oxygen (FiO₂) attenuates the oxidative stress, reducing xanthine oxidase (XO) activity in colonic mucosa as a possible mechanism during colon surgery.

Methods

Twenty-four cancer patients who underwent elective colon resection were randomly assigned to either 0.30 FiO₂ or 0.80 FiO₂ during surgery. Malondialdehyde (MDA) and oxidized glutathione (GSSG) levels were measured in blood plasma. We also determined the enzymatic activities of XO and xanthine dehydrogenase (XDH) in the colonic mucosa after resection of the piece of colon.

Results

Oxygen partial pressure in arterial blood was higher in the 0.80 FiO₂ group than in the 0.30 FiO₂ group (P < 0.001). MDA and GSSG levels measured in blood plasma as well as MDA measured in colonic mucosa were lower in the 0.80 FiO₂ group than in the 0.30 FiO₂ group. Otherwise, XO enzymatic activity and XO/(XO + XDH) ratio in colonic mucosa were lower in the 0.80 FiO₂ group than in the 0.30 FiO₂ group.

Conclusions

XO may be a major source of reactive oxygen species in patients during colon surgery. Inspiring 0.80 oxygen during colon surgery increases arterial partial pressure and this treatment was associated with reduced XO activity and levels of oxidative stress in colonic mucosa.

Keywords:

• Colon surgery
• Supplemental oxygen
• Oxidative stress
• Xanthine oxidase

Introduction

Intestinal tissue is metabolically very active and has a poor tolerance for even short periods of hypoxia or ischaemia.¹ The special architecture of intestinal villi's microcirculation may increase the susceptibility of intestinal epithelium to hypoxic damage due to insufficient perfusion, which increases the dependency on oxygen uptake.^2,3 Other factors may increase the relative intestinal hypoxia that can occur with normal arterial oxygen saturation during abdominal surgery, including: surgical stress, surgical retractors, increased intra-abdominal pressure, intestinal mobilization, operative disruption of blood vessels, and splanchnic vasoconstriction.⁴ On the other hand, laparotomy and mild intestinal manipulation leads to activation of xanthine oxidase (XO) in enterocytes. In brief, in ischaemic tissues the phosphorylation potential fall enhances ADP catabolism with increased production of hypoxanthine and xanthine. Intracellular proteases are stimulated by Ca²⁺ overload and acidosis and convert xanthine dehydrogenase (XDH) into XO. This results in oxidative damage to the intestinal mucosa by generation of superoxide anions.^5–7

Fractional inspired oxygen (FiO₂) is a major determinant of arterial oxygen partial pressure (PaO₂) and tissue oxygen tension (as well as haemoglobin levels and cardiac output). Consequently, a high PaO₂ improves oxygenation during hypoxia in intestinal epithelial cells in poorly perfused areas.^8–10 Perioperative administration of highly concentrated inspired oxygen (0.80 FiO₂ throughout surgery and for a postoperative period) may enhance PaO₂ and oxygen tissue tension and reported a reduction in the risk of wound infection in patients undergoing colon surgery.^4,11,12 In contrast, other studies hypothesized that supplemental perioperative oxygen increased the production of reactive oxygen species,¹³ or that it did not result in a difference in risk of surgical site infection after abdominal surgery.¹⁴ However, we recently found a decrease in oxidative stress levels in patients undergoing colon surgery with supplemental perioperative oxygen (0.80 FiO₂), with a decrease in blood levels of oxidative stress markers from lipoperoxidation (malondialdehyde, MDA) and glutathione oxidation (oxidized glutathione, GSSG).¹⁵ We do not have data on the possible oxidative stress mechanisms in these patients. For this reason, this work is a complement of the previously cited article.¹⁵

We initiated this study to determine the effect of a high concentration of oxygen in the inspired gas mixture on oxidative stress mechanisms during colon surgery. The objective was to establish the effect of two different oxygen concentrations (0.30 FiO₂ or 0.80 FiO₂) on oxidative stress marker levels. Specifically, to determine whether supplemental oxygen with 0.80 FiO₂ protected against oxidative stress during colon surgery. We have analysed the MDA levels, as a lipoperoxidation marker, as well as the XDH and XO activities in the colonic mucosa, as these activities represent a major source of free radicals during intestine manipulation or ischaemia–reperfusion.

Materials and methods

Study subjects

Twenty-four consecutive cancer patients undergoing elective colon resection were randomly assigned to receive 0.30 FiO₂ (n = 12) or 0.80 FiO₂ (n = 12) after obtaining informed consent. Surgeons and data analysers were blinded to the treatment regimen. Ages ranged from 18 to 80 years. The study was carried out in accordance with the Helsinki Declaration of 1975 and subsequent reviews. The Ethical Committee of the hospital approved the study. Exclusion criteria were fever, existing signs of infection, diabetes mellitus (type 1 or 2), pregnancy, drug addiction, intake of antioxidants vitamins, and malnutrition (serum albumin under 3.3 g/dl, white cell count under 2500 cells/mm³, or body weight loss over 20% in the last 2 months).

Protocol

The night before surgery all patients underwent a standard bowel preparation with electrolyte solution.^11,12 In addition, 2 days before surgery patients were on a liquid diet without residues or supplemental antioxidant vitamins. All patients were given prophylactic intravenous antibiotics (metronidazole combined with amoxicillin-clavulanic acid) 60 minutes before the surgical incision and for 48 hours postoperatively (every 8 hours), according to the hospital protocol.

Anaesthesia

Anaesthesia was induced intravenously with sodium thiopental (3–5 mg/kg) and fentanyl (2 µg/kg). Endotracheal intubation was facilitated with rocuronium bromide (0.6 mg/kg). Anaesthesia was subsequently maintained with sevoflurane (2%), fentanyl, and rocuronium bromide. After induction of anaesthesia, patients were randomly assigned to receive either 0.30 FiO₂ or 0.80 FiO₂ throughout surgery. During surgery, the lungs were ventilated in a volume-controlled mode (8–10 ml/kg tidal volume) and patients were hydrated with Ringeŕs solution (15 ml/kg/hour).^11,12 Standard anaesthesia monitoring, including arterial oxygen saturation, blood pressure, heart rate, and end tidal CO₂ were measured at 5 minutes intervals during anaesthesia.

Measurements

Arterial blood samples for analysis were obtained from a radial artery line 20 minutes after induction of anaesthesia (‘before surgery’ in Table 3) and immediately after the surgery procedure ended (‘after surgery’ in Table 3). The arterial blood gases: pH, partial oxygen pressure (PaO₂), and partial carbon dioxide pressure (PaCO₂) were analysed with an ABL-3 (Radiometer, Copenhagen, Denmark).

Once the surgeon removed the piece of colon, a sample was taken from the healthy proximal border that included the colonic mucosa. These samples were frozen in liquid nitrogen and stored at −80°C until they were processed. All frozen colonic mucosa samples were extracted at ice temperature. A 200 mg aliquot was weighed, homogenized (Ultraturrax T8, Labortecnik, Staufen, Germany) in 4 ml/g phosphate buffer, and centrifuged (15 000 g for 30 minutes at 4°C). The supernatant was removed and eluted through a 5 × 1.44 cm² Sephadex G-25 column equilibrated with phosphate buffer. The eluted was fractionated into aliquots, which were individually tested for XO and XDH activities and protein concentration in colonic mucosa.

All chemicals were provided at the highest available purity degree from Sigma-Aldrich (St Louis, MO, USA), Merk (Darmstad, Germany), and Boehringer Mannheim (Mannheim, Germany).

MDA levels were measured in blood plasma and colonic mucosa. Briefly, lipoperoxides from lipid peroxidation of cellular polyunsaturated fatty acid were hydrolysed by boiling in dilute phosphoric acid (0.44 M). MDA is one of these hydrolysis products. Then, thiobarbituric acid (TBA) was added to the hydrolysis products, which reacts specifically with MDA to form MDA-(TBA)₂ adduct. Next, plasma (or colonic mucosa homogenate) proteins were precipitated with methanol and removed by centrifugation. The protein-free extract was fractioned by a high-performance liquid chromatography (HPLC) method to separate the MDA-(TBA)₂ adduct from interfering chromogens and quantified spectrophotometrically (Shimadzu SPD 10 AV, Shimadzu Scientific Instruments Inc., Columbia, MD, USA) with UV light at 532 nm.¹⁶

GSSG concentration was measured by an HPLC method in whole blood, with detection at 365 nm which was developed to measure GSSG in the presence of large excess of reduced glutathione, using N-ethylmaleimide to prevent reduced glutathione (GSH) oxidation.¹⁷

Determinations of XO and XDH activities in the colonic mucosa were carried out by fluorometric assays. In brief, the XO-catalysed conversion of pterin (2-amino-4-isoxanthopterin hydroxypteridine) to isoxanthopterin was the basis for this fluorometric assay (excitation: 345 nm and emission: 390 nm). XO is assayed in the presence of pterin only, while combined XDH plus oxidase activity is determined with methylene blue, which replaces NAD⁺ as an electron acceptor. XDH activity was calculated by subtracting the XO activity from the total activity (XDH + XO).¹⁸ This assay has been successfully applied where XO and XDH activities are too low to detect spectrophotometrically. To express XDH-to-XO conversion, we used the ratio XO/(XO + XDH). The protein concentration in colonic mucosa was measured by the method of Lowry using bovine serum albumin as the standard.¹⁹

Statistical analysis

Results are expressed in tables (as median and range) and figures (as scatter plot and median). Differences were analysed with two-tailed tests for non-parametric data: Mann–Whitney U and Wilcoxon tests. P values <0.05 were considered statistically significant. Pearson's correlation coefficients were determined between different parameters of oxidative stress in colonic mucosa: XO enzymatic activity and XO/(XO + XDH) ratio as independent variables and MDA as a dependent variable. The Statistical Package for the Social Sciences (version 11.0) was used for analysis (SPSS, Chicago, IL, USA) and Excel (Microsoft Office 2003, Redmond, WA, USA).

Results

No differences were found between the study groups in clinical and physical data (Table 1). Only intra-operative oxygen saturation on pulse oxymetry (SpO₂) values were higher (P < 0.001) in patients receiving 0.80 oxygen (0.80 FiO₂ group) compared with patients who received 0.30 oxygen (0.30 FiO₂ group) (Table 2).

Table 1. Demographic data of the two groups of cancer patients

	0.30 FiO2 group	0.80 FiO2 group
Age (year)	67.5 (43–77)	66.5 (42–76)
Sex (Male:female)	9:3	10:2
Weight (kg)	71.5 (61–91)	67.5 (52–112)
Diagnosis: colon-rectum cancer	12	12
Hemicolectomy	8	7
Sigmoid anterior section	4	5

Values expressed as median (range). Differences were analysed with two-tailed tests for non-parametric data: Mann–Whitney U and Wilcoxon tests. P values <0.05 were considered statistically significant. No differences were found between the study groups.

Table 2. Intraoperative values of clinical and physical status of the two groups of patients

	0.30 FiO2 group	0.80 FiO2 group
Duration of surgery (minute)	195 (150–270)	200 (145–260)
Number of patients transfused	3	4
Volume infusion (l)	4.5 (3.0–5.0)	4.5 (3.5–5.2)
Average core temperature (°C)	36 (35–37)	36 (35–37)
Average SpO2 (%)	97 (95–99) 99	(97–100)*
Average mean arterial pressure (mmHg)	83 (72–91)	86 (65–100)

Values expressed as median (range). Differences were analysed with two-tailed tests for non-parametric data: Mann–Whitney U and Wilcoxon tests. P values <0.05 were considered statistically significant. No differences were found between the study groups except in average SpO₂.

*Average SpO₂ (oxygen saturation on pulse oxymetry) values in the 0.80 group were higher than in the 0.30 group values (P < 0.001).

Twenty minutes after the induction of anaesthesia and immediately after the end of surgery, both SpO₂ and PaO₂ values from arterial blood gas analysis were significantly higher in the 0.80 FiO₂ patient group compared with the 0.30 FiO₂ patient group (P < 0.001). No differences were found between the study groups in other parameters, such as PaO₂/FiO₂, PaCO₂, pH, haemoglobin, and haematocrit (see Table 3 and Fig. 1).

Figure 1. Effects of FiO₂ on arterial blood gas analysis (PaO₂ and PaCO₂) in the two groups (0.30 FiO₂ and 0.80 FiO₂) of cancer patients, before surgery and after surgery. Values expressed as scatter plot and median (horizontal bars). Differences were analysed with two-tailed tests for non-parametric data: Mann–Whitney U and Wilcoxon tests. P values <0.05 were considered statistically significant. (A and C), PaO₂ values in the 0.80 FiO₂ group were significantly higher (P < 0.001) than in the 0.30 FiO₂ group's values, before and after surgery. (B and D), in PaCO₂ values, no differences were found between the study groups, before and after surgery.

Table 3. Effects of FiO₂ on arterial blood gas analysis in the two groups of cancer patients

	0.30 FiO2 group	0.80 FiO2 group	P
Before surgery
pH	7.43 (7.36–7.56)	7.42 (7.38–7.53)	0.953
PaO2 (mmHg)	95 (72–131)	291(181–491)	<0.001
PaCO2 (mmHg)	33 (24–37)	33 (23–40)	0.854
PaO2/FiO2	338 (239–443)	386 (226–614)	0.169
SpO2 (%)	98 (95–100)	100 (99–100)	< 0.001
Haemoglobin (g/dl)	12.5 (10.30–15.8)	12.3 (10.4–12.8)	0.678
After surgery
pH	7.37 (7.33–7.40)	7.36 (7.26–7.42)	0.726
PaO2 (mmHg)	103 (74–131)	311 (246–383)	<0.001
PaCO2 (mmHg)	31 (23–36)	32 (24–38)	0.931
PaO2/FiO2	350 (285–522)	393 (286–479)	0.073
SpO2 (%)	97 (90.0–100)	100 (99–100)	<0.001
Haemoglobin (g/dl)	11.7 (9.4–14.8)	11.9 (10.0–14.4)	0.8010

Values expressed as median (range). Differences were analysed with two-tailed tests for non-parametric data: Mann–Whitney U and Wilcoxon tests. P values <0.05 were considered statistically significant. SpO₂ (haemoglobin oxygen saturation) and PaO₂ (oxygen partial pressure) values in the 0.80 FiO₂ group were significantly higher (P < 0.001) than in the 0.30 FiO₂ group's values, before and after surgery.

Fig. 2 summarizes the results of oxidative stress markers measured in blood plasma. Blood plasma levels of MDA after surgery were lower in the 0.80 FiO₂ group than in the 0.30 FiO₂ group (P < 0.001). In addition, in the 0.30 FiO₂ group, blood plasma levels of MDA after surgery were higher than before surgery (P < 0.029). However, in the 0.80 FiO₂ group, blood plasma levels of MDA after surgery were lower than before surgery (P < 0.010). Blood plasma levels of GSSG after surgery were lower in the 0.80 FiO₂ group than in the 0.30 FiO₂ group (P < 0.048). In addition, in the 0.30 FiO₂ group, blood plasma levels of GSSG after surgery were higher than before surgery (P < 0.039).

Figure 2. Effects of FiO₂ on levels of oxidative stress markers like MDA and GSSG in arterial blood of the two groups (0.30 FiO₂ and 0.80 FiO₂) of cancer patients, before surgery and after surgery. Values expressed as scatter plot and median (horizontal bars). Differences were analysed with two-tailed tests for non-parametric data: Mann–Whitney U and Wilcoxon tests. P values <0.05 were considered statistically significant. (A and B), MDA and GSSG values respectively – for these markers no differences were found between the study groups before surgery. (C and D), MDA and GSSG, respectively, in the 0.30 FiO₂ group were significantly higher than in the 0.80 FiO₂ group after surgery.

Fig. 3 summarizes the results measured in colonic mucosa from removed piece of colon in the two groups of cancer patients. Tissue values of MDA after surgery were lower in the 0.80 FiO₂ group than in the 0.30 FiO₂ group (P = 0.005). In colonic mucosa, tissue values of XO activity were higher in the 0.30 FiO₂ group than in the 0.80 FiO₂ group (P = 0.044), and tissue values of XO/(XO + XDH) ratio, to express XDH-to-XO conversion, were higher in the 0.30 FiO₂ group than in the 0.80 FiO₂ group (P = 0.047).

Figure 3. Effects of FiO₂ on MDA levels, XO activity, XDH activity, XO + XDH activities, and XO/(XO + XDH) ratio in colonic mucosa from excised colon in the two groups of cancer patients. MDA, malondialdehyde; XO, xanthine oxidase; XDH, xanthine dehydrogenase. Values expressed as scatter plot and median (horizontal bars). Differences were analysed with two-tailed tests for non-parametric data: Mann–Whitney U and Wilcoxon tests. P values <0.05 were considered statistically significant. (A) MDA values were significantly lower in the 0.80 group than in the 0.30 group. (B) XO activity was significantly lower in the 0.80 group than in 0.30 group. (C) On XDH activity no differences were found between the study groups. (D) On XO + XDH activities no differences were found between the study groups. (E) On XO/(XO + XDH) ratio, values were significantly lower in 0.80 group than 0.30 group.

Significant and positive correlations were found between MDA tissue levels (as a dependent variable) and XO enzymatic activity and XO/(XO + XDH) ratio (as independent variables) in both groups. The Pearson's correlation coefficient (r) values were higher in the 0.30 FiO₂ group than in the 0.80 FiO₂ group (see Fig. 4).

Figure 4. Effects of FiO₂ on the correlations between MDA levels and XO activity and XO/(XO + XDH) ratio in colonic mucosa from removed piece of colon in the two groups of patients. MDA, malondialdehyde; XO, xanthine oxidase; XDH, xanthine dehydrogenase. r, Pearson's correlation coefficient. XO enzymatic activity and XO/(XO + XDH) ratio as independent variables and MDA as a dependent variable. P values <0.05 were considered statistically significant. (A) correlation between MDA levels and XO activity was very significant in 0.30 FiO₂ group. (B) correlation between MDA levels and XO activity was significant in 0.80 FiO₂ group and the r value was lower than 0.30 FiO₂ group. (C) correlation between MDA levels and XO/(XO + XDH) ratio on 0.30 FiO₂ group was statistically significant. (D) correlation between MDA levels and XO/(XO + XDH) ratio in 0.80 FiO₂ group was not statistically significant and the r value was lower than 0.30 FiO₂ group.

Discussion

This study shows that inhalation of 0.80 oxygen during colon surgery results in an increase in arterial oxygen partial pressures, which results in a decrease of XO enzymatic activity and XO/(XO + XDH) ratio and MDA tissue levels in the colonic mucosa. These results may explain, at least in part, a possible mechanism of oxidative stress induced by ischaemia–reperfusion in these patients.

MDA is a final product of fatty acid peroxidation in cell membranes and a classical marker of oxidative stress in tissues and blood.¹⁶ We recently reported a reduction of MDA blood plasma levels and other blood oxidative stress marker levels, such as GSSG, in patients undergoing colon surgery with supplemental oxygen (0.80 FiO₂).¹⁵ In the present study, we again observed an increase of MDA and GSSG levels in the blood plasma, as well as in the colonic mucosa, of patients who inspired 0.30 FiO₂ during colon surgery. However, patients who inspired 0.80 oxygen during the same study period did not show these increased oxidative stress markers. Therefore, administration of 0.80 FiO₂ in ventilated patients undergoing colon surgery may protect against the peroxidation of fatty acids in the colonic mucosa by oxidative stress.

Splanchnic vasoconstriction induced by surgical stress is known to influence intestinal hypoxia during the operative period. Tissue hypoxia triggers a sequence of events that may ultimately lead to cellular dysfunction and XO activation with generation of superoxide anions, which are very reactive oxygen species.^{5–8,20–22} XO is derived from XDH in those hypoxia situations, and XO is a major source of superoxide anions during reperfusion.^20,22 In this study we found that XO enzymatic activity and XO/(XO + XDH) ratio, to express XDH-to-XO conversion, in colonic mucosa during colon surgery were lower in patients who inspired 0.80 FiO₂ than in patients who inspired 0.30 FiO₂.

A transmural gradient of PaO₂ exists across the colon with mucosal PaO₂ far lower than serosal PaO₂. Both serosal and mucosal PaO₂ decrease during hypovolemia in shocked rats.²³ However, our patients were well hydrated and they did not have hypovolemia. Therefore, it is reasonable that a hyperoxic status and an increased PaO₂ improve oxygen delivery to the places most in need of oxygen, such as the colon mucosa.

Arterial blood values (SpO₂ and PaO₂) were significantly higher in patients who received 0.80 FiO₂ compared to patients receiving 0.30 FiO₂. Although we have not used a Clark type polarographic oxygen electrode for measurement of tissue oxygen tension, we assumed that oxygen tension was also higher in the intestinal mucosa in patients receiving 0.80 FiO₂.^8–10 In rats it has been reported that a marked degrees of tissue hypoxia can be tolerated before major histological damage. Non-viable bowel had significantly impaired tissue oxygenation when compared with viable bowel.²⁴

It is possible that the high PaO₂ may also have had an effect on the blood cells, because recently it has been shown that 80% oxygen may decrease the production of TNF alpha and increase production of reactive oxygen species in human leukocytes, and exerts significant effects on multiple cellular and immunological parameters, providing a potential mechanism for benefits from the use of supplemental oxygen.²⁵

This increase in oxygen tissue could explain a decrease in the activity of XO and XO/(XO + XDH) ratio in the group of patients who breathed 0.80 FiO₂ since XDH is activated by hypoxia and XO is activated by reperfusion. This mechanism might explain, at least in part, the increase of oxidative stress in patients who inspired 0.30 FiO₂.¹⁵ Furthermore, the degree of oxidative stress, expressed as tissue levels of MDA in colonic mucosa, correlated positively and very significantly with the activity of XO and XO/(XO + XDH) ratio. These correlation coefficients were higher in the group of patients who breathed 0.30 FiO₂ than in those who breathed 0.80 FiO₂.

We could not find respiratory complications in our patients during the study period. Administration of 0.80 FiO₂ during surgery and 2 hours after surgery did not worsen pulmonary function or cause atelectasis, as confirmed by chest computed tomographic scans.^11,26 Furthermore, no changes were observed in the PaO₂/FiO₂ ratio, which rules out respiratory distress.

In relation to the clinical implications of supplemental oxygen and hyperoxia, recently it has been shown that hyperoxia has a suppressive effect on radical superoxide generation in blood, with improvement of oxidative stress, inflammatory response, and early endothelial damage in rats subjected to cerebral ischaemia–reperfusion.²⁷ In this context, it also has been recently demonstrated that hyperoxic preconditioning of rats in vivo protects against ischaemia–reperfusion myocardial damage by inhibition of mitochondrial transition pore opening and permeability cytochrome c release.²⁸

In summary, the results of our study showed an increase in oxidative stress marker levels in blood and colonic mucosa occur when 0.30 FiO₂ is used during colon surgery, possibly through an increase in XO enzymatic activity in the colonic mucosa. The administration of 0.80 FiO₂ prevented oxidative stress, with a reduction of lipid peroxidation and glutathione oxidation; this may be due to decreases in XO enzymatic activity and XO/(XO + XDH) ratio in the colonic mucosa. No signs of lung alteration were observed when the high concentration of oxygen was used during surgery. Our results support the administration of 0.80 FiO₂ during colon surgery, because it prevented oxidative stress¹⁵ and can be effective in the reduction of surgical wound infections, return of bowel function and duration of hospitalization. The primary defence against surgical pathogens is oxidative killing by neutrophils and oxidative killing is a function of tissue oxygen partial pressure.^11,12,29,30 Moreover, supplemental oxygen appears to confer few risks to the patients and little associated cost.

Acknowledgement

The authors thank Maria Dolores Royo for her generous and skilful technical assistance and Koen J. Mateboer for proofreading the English manuscript.