**4. Discussion**

Accumulating evidence indicates that oxidative stress is not only merely a consequence of chronic inflammation, but may have an essential role in the development and maintenance of inflammation and aberrant immune response in CD. In this regard, our results demonstrate an overall increase in oxidative stress biomarkers in CD patients at surgery when compared to the controls, highlighting that severe clinical activity is reflected by systemic oxidative stress. Among the markers analyzed, AOPP demonstrated the greatest diagnostic ability in differentiating CD patients from the controls. AOPPs are di-tyrosine-containing and cross-linking products, formed by the reaction of plasma proteins, mainly albumin, with chlorinated compounds resulting from the activity of myeloperoxidase (MPO) [14,18]. For this reason, AOPP are recognized as both oxidative protein damage markers and mediators of inflammation.

Increased plasma AOPP levels in patients with chronic diseases including active CD patients have been reported by others [19–21]. In our study, CD patients had very high levels of AOPP in their serum, much higher when compared to the study by Krzystek-Korpacka and co-workers [20], who measured a mean level of 1.87 μmol/g of albumin in the plasma of active CD patients. We measured AOPP levels in the serum of CD patients at surgery, therefore with severe clinical conditions, and these high levels can be a consequence of their clinical status.

Interestingly, some mechanistic studies have demonstrated the role of AOPP in the pathogenesis and progression of CD. There is in fact in vitro and in vivo evidence that AOPPs induced depletion of intestinal epithelial cells and inflammatory changes that alter the structural integrity of the intestinal mucosa [22–24]. These compounds are also able to modulate cell cycle arrest [25]; Shi and co-worker recently demonstrated that AOPPs exhibit their negative regulatory function on intestinal epithelial cell cycle progression by activating the RAGE/CD36-c-jun N-terminal kinase (JNK)-p27kip1 signaling pathway [21]. By interacting with RAGE and CD36 receptors, AOPP activate protein kinase C and nicotinamide adenine dinucleotide phosphate (NADPH) oxidase as well as the NF-κB-dependent inflammatory pathway [26]. Xu et al. [27] also found that the deposition of AOPPs in fibrotic lesions from CD patients promoted epithelial–mesenchymal transition, a fundamental mechanism in complications of CD such as intestinal fibrosis through the oxidative and inflammatory pathway.

Other than AOPP, the first identified ligands of RAGE were AGEs. Although RAGE/AGE signaling has mainly been studied in diabetes, there is also evidence of its activation in neurodegenerative diseases, cancer, and in various inflammatory diseases including CD [28–30]. Mice lacking RAGE receptors are in fact protected from chemically-induced intestinal inflammation and treatment with a RAGE-specific inhibitor protects mice from indomethacin-induced enteritis and dextran sulfate sodium-induced colitis, suggesting that the RAGE signaling pathway could be a promising therapeutic target for IBD patients [30]. On the contrary, there are no reports available in the literature on the serum level of AGEs in patients with CD. AGEs are heterogeneous compounds mainly generated through the non-enzymatic glycation of protein, lipids, and nucleic acid driven by hyperglycemia and oxidative stress, but increased serum levels were associated with hypercholesterolemia [31] and cigarette smoking [32]. In addition, humans are exposed to dietary sources of AGEs through animal-derived foods and cooking processes that result in the formation of new AGEs [33]. Moreover, AGEs such as glycol aldehyde and 2-hydroxy-propanal may be generated by activated neutrophils, even in the absence of sugars [34].

Lipid peroxidation products are also implicated in the pathogenesis of IBD. For instance, it has been reported that 4-hydroxynonenal treatment suppressed colonic expression of tight junction proteins, enhanced bacterial translocation from the gu<sup>t</sup> into the systemic circulation, and increased activation of Toll-like receptor 4 signaling [35].

In our cases, circulating TBARS were significantly higher when compared to the controls and these results are in line with those obtained by others who measured elevated lipid peroxidation markers malondialdehyde (MDA)/TBARS in CD patients [35–38]. In the study by Sampietro and co-workers, CD patients at surgery showed a significantly higher basal peroxidative state when compared to the controls, but while the inflammatory and oxidative indices were significantly reduced, two months later, and maintained low one year after surgery, TBARS did not reach levels comparable to those in the control subjects, indicating that in quiescent CD, there is an upregulated level of plasma peroxidation [39].

Szczeklik and co-workers recently described the presence of an upward trend in the serum (and saliva) MDA levels, depending on the severity of CD and a correlation between the MDA levels and the visible symptoms of inflammation [40]. Our results showed a significant correlation among lipid peroxidation, disease location, and behavior and with the severity of the first clinical presentation of CD. All these data sugges<sup>t</sup> the role of TBARS as a potential marker of the severity of the disease.

Serum TBARS levels were also correlated with smoking habit being higher in smokers and to a lesser extent, in former smokers when compared to non-smokers. Cigarette smoke is a well-known source of ROS and one of the most powerful oxidative stress inducers. It is considered both an etiological risk factor for CD as well as for its recurrence; smokers have a higher risk when compared to non-smokers of developing a postoperative recurrence and the risk increases in relation to the number of cigarettes smoked [41,42].

We also defined an oxidative score to take into account the overall oxidative–antioxidant status of each patient and identified positive correlations with smoking habit and the presence of cutaneous manifestation of the disease, a well-recognized complication of IBD that frequently occurs in CD and is associated with a worse prognosis [43,44]. The pathogenic mechanism underlying the development of cutaneous manifestations in CD patients is still not known, but our results sugges<sup>t</sup> that oxidative stress may have a role.

Despite the high degree of oxidative stress observed in CD patients, witnessed by the increase in all oxidative markers measured, all correlated to each other, we did not observe differences in their antioxidant status, measured as FRAP, when compared to controls. On the contrary, several studies reported a reduced antioxidant capacity in CD patients, measured both as plasma carotenoids and other vitamin content [45,46], or as total antioxidant capacity [12,47,48], or as plasma free thiols [10]. In particular, Bourgonje and coworkers reported strong and negative correlations among albumin-adjusted plasma free thiols and a number of pro-inflammatory markers of disease activity [10]. In this study, a more favorable redox status was also observed in CD patients with ileal disease compared to patients with colonic localization; in contrast, in our study, the ileal localization was significantly associated with increased serum TBARS levels. We previously reported that disease located in the upper part of the intestine is a risk factor for recurrence when compared to diseases located in the distal ileum and colon [49].

The serum/plasma total antioxidant activity is the sum of the contribution of endogenous (uric acid, bilirubin, albumin) and exogenous (medications and food-derived) antioxidants, thus this disagreement may be due, not only to the complexity of the disease, but also to the variability in patient medications and dietary habits or supplement use. In this context, we noted differences in AOPP values associated with the treatment with mesalazine, known to exert anti-inflammatory and antioxidant effects, and in the FRAP levels that were significantly higher in patients under azathioprine treatment, which may be related to its contribution to oxidative stress [50].

Moreover, the lack of an association between circulating markers of oxidative damage and total antioxidant capacity suggests that in CD, oxidative stress is not just the result of an imbalance between oxidants and antioxidants, but may prime, at least in the acute phase of the disease, pro-inflammatory mechanisms through RAGE activation.

The main limitations of this study are the relatively small sample size, the cross-sectional design, the lack of follow-up, and of recurrent measurements during the course of the disease, which are necessary to validate the relevance of using oxidative stress markers in the clinical setting.
