Evaluation of the inflammatory markers CCL8, CXCL5, and LIF in patients with anastomotic leakage after colorectal cancer surgery

Anastomotic leakage (AL) after colorectal surgery constitutes a dreaded complication after colorectal surgery [1]. Depending on the anatomical localization, insufficiency rates of about 3% after colon surgery and between 3 and 19% after colorectal surgery are described [2‐4]. Even in high-volume centers, a portion of patients develops an anastomotic leakage. However, high-volume centers reveal less in-hospital mortality rates of patients with anastomotic leakage [5] . Consensus definition states anastomotic leakage as a communication between the intra- and extraluminal compartments resulting from a defect in the integrity of the intestinal wall at the anastomosis. Leakages originating from the suture or staple line of a neorectal reservoir, as well as a pelvic abscess in the proximity of the anastomosis, are also considered an anastomotic leakage [6, 7]. Defined by the International Study Group of Rectal Cancer, three grades of AL exist ranking the AL due to its clinical severity. Grade A is called a radiologic leakage meaning the patient has no clinical symptoms or increased infectious values in the blood test. This kind of anastomotic leakage entails no active therapeutic intervention. Grade B patients present with leukocytosis, an increase of CRP, abdominal pain, and distension and require an active therapeutic intervention in terms of antibiotics or an interventionally placed pelvic drain. But there is no need for relaparotomy. Grade C AL includes the symptoms as Grade B together with an ensuing peritonitis or sepsis. Patients with Grade C AL require a relaparotomy which is often associated with Hartmann’s procedure [6, 7]. The mean occurrence of colorectal anastomotic leakage (CAL) has been described for the eighth postoperative day (POD), with an interval between the sixth and thirteenth POD [8, 9]. Preoperative and intraoperative risk factors for CAL are male sex, distal anastomosis, advanced tumor stage, emergency surgery, duration of surgery, or amount of blood loss [1, 10‐16]. Nevertheless, CAL rates remain stable over the past years [17]. AL after colon or colorectal resection is associated with a prolonged hospital stay, substantial negative impact on morbidity and mortality rates, as well as higher cancer recurrence frequency [1, 3, 18, 19]. Therefore, an objective of utmost importance is the early detection of AL to limit the negative postoperative outcome to a minimum. Occurrence of an anastomotic leakage is associated with a local inflammation at this site. Moreover, an upregulation of inflammatory cytokines and chemokines in case of inflammation is commonly known [20]. As acute-phase proteins, cytokines or chemokines, respectively, were synthesized in the liver and at the site of inflammation, protein levels taken from pelvic drain fluid represent the local milieu and could be more specific for the detection of an anastomotic leakage [21]. To investigate markers for AL, we analyzed a panel of inflammatory markers in sera and peritoneal fluid from the abdominal drain on the third postoperative day from patients with and without CAL after colorectal surgery due to colorectal cancer. CCL-1/I-309 (C-C motif ligand 1), CCL8/MCP-2 (monocyte chemotactic protein-2), CCL13/MCP-4 (monocyte chemotactic protein-4), CXCL5/ENA-78 (epithelial neutrophil-activating peptide), LIF (leukemia inhibitory factor), IL-16 (interleukin-16), and IL-21 (interleukin-21) were chosen for analysis. Selected markers were chosen based on a literature research because of their known role in inflammatory processes. CCL-1/I-309 is produced mostly by T_regs at the site of inflammation [22]. CCL8/MCP-2 activates immune cells like natural killer cells as a proinflammatory mediator [23]. CCL13/MCP-4 carries out proinflammatory actions through chemotaxis of monocyte-derived macrophages, lymphocytes, and basophils [24]. CXCL5/ENA-78 is detected in inflamed intestinal mucosa. Il-16 is produced by T lymphocytes, eosinophils, mast cells, and macrophages during inflammatory responses and is recruited if cell necrosis occurs [25, 26]. Il-21 expression is induced by other cytokines, e.g., Il-6, and it regulates the proliferation and function of numerous immune cells like natural killer cells [27]. And LIF promotes recruitment of inflammatory cells to the area of damage [28].Moreover, until now no study investigated the influence of these inflammatory markers in anastomotic leakage after colorectal surgery. We hypothesize that the above named inflammatory markers - measured in the peritoneal fluid - could predict the occurrence of an anastomotic leakage after colorectal surgery already on the third postoperative day, prior to the mean occurrence at the eight postoperative day.

The present study reveals altered expression of CCL8/MCP-2, CXCL5/ENA-78, and LIF in peritoneal fluid samples of patients with an anastomotic leakage after colorectal surgery.

Anastomotic leakage is still a detrimental complication after colorectal surgery with adverse prognostic effects for the patients including prolonged hospital stay as well as increased morbidity and mortality rates [1, 18]. Moreover, anastomotic leakage displays a risk factor for local recurrence and shortens long-term cancer-specific survival [29]. Clinical signs of an anastomotic leakage before the fifth postoperative day are unusual, and as mean day of diagnosis, the eighth postoperative day is stated commonly [8]. Notwithstanding the advancements in operative techniques and definition of risk factors, the incidence of anastomotic leakages after colorectal surgery has not dwindled over the last decade [1, 4, 30]. Additionally, a delay in diagnosis of an anastomotic leakage is common due to diagnostic methods lacking sensitivity and specificity [21, 31]. Doeksen et al. described a sensitivity of 54% and a specificity of 66% for the detection of a colorectal anastomotic leakage using CT scans. However, it is not described if a rectal filling with contrast agent was performed [32]. Hence, additional indicators for the detection of an anastomotic leakage are of utmost importance. A variety of studies evaluated several inflammatory markers like Il-6, TNF-α, MMP8, MMP9, or procalcitonin in peritoneal fluid and/or serum samples of patients suffering anastomotic leakage after colorectal resection in order to determine diagnostic biomarker [17, 33‐38]. One of the most assessed markers constitutes CRP value. C-reactive protein is an acute-phase reactant protein which is synthesized mainly by hepatocytes but also by blood monocytes in response to infection or inflammation upon stimulation by proinflammatory cytokines as interleukin-6 (Il-6) and tumor necrosis factor-α (TNF-α) [39]. CRP activates complement pathway and boosts phagocytosis of damaged cells [40], while it is a reliable inflammatory marker, but non-specific [41]. Singh et al. defined in their systematic review and meta-analysis serum CRP on PODs (postoperative days) 3, 4, and 5 to be a diagnostic tool for the detection of an anastomotic leakage after colorectal surgery with an area under the curve of 0.81 and a CRP cutoff value of 172 mg/l on POD 3 [41]. Komen et al. assessed significantly elevated CRP level in peritoneal fluid samples on POD 3 and POD 5 in patients suffering an anastomotic leakage after colorectal surgery [9]. No area under the curve or cutoff values were defined in this study. Congruently with the abovementioned studies, we observed significantly elevated CRP levels on POD 3 in both serum and peritoneal fluid—with an area under the curve of 0.811 likewise and a CRP cutoff value of 153 mg/l and an area under the curve of 0.703 and a CRP cutoff value of 62.92 mg/l, respectively—of patients developing an anastomotic leakage compared with patients without an anastomotic leakage, indicating that our study collective constitutes a representative study cohort.

We found a differential expression of CCL8/MCP-2, CXCL5/ENA78, and LIF in peritoneal fluid samples of patients developing an anastomotic leakage on POD 3 compared with patients without an anastomotic leakage. As acute-phase proteins were synthesized in the liver and at the site of inflammation, protein levels taken from pelvic drain fluid represent the local milieu and could be more specific for the detection of an anastomotic leakage than serum markers [21]. Upregulation of cytokines and chemokines in case of inflammation is commonly known [20]. Induced by fibroblasts, macrophages, and endothelial cells, CCL8/MCP-2 acts through CCR1–3, CCR2, CCR3, and CCR5, activating immune cells including neutrophils, eosinophils, basophils, monocytes, T cells, and natural killer cells as a proinflammatory mediator [23, 42, 43]. Due to local inflammation at the site of the anastomotic leak, it is comprehensible that CCL8/MCP-2 is higher expressed in peritoneal fluid samples of patients developing anastomotic leakage. Moreover, the chemokine CXCL5/ENA-78 is higher expressed in patients with an anastomotic leakage, too. CXCL5/ENA-78 is an epithelial-derived chemokine appearing in the inflamed intestinal mucosa recruiting neutrophils from the lamina propria to the epithelial layer [44‐46]. Further, it is produced by enterocytes, monocytes, and endothelial cells and constitutes a powerful neutrophil chemoattractant interacting with CXCR2 [44, 47‐49]. And it drives through cancer-related angiogenesis [50]. LIF—a member if the IL-6 cytokine family—acts through LIF receptor ß which is located on macrophages and monocytes [51]. LIF activates STAT3/JAK1-kinase, PI3-kinase, and MAPK pathway and epitomizes proinflammatory characteristics like induction of acute-phase proteins and promotes recruitment of inflammatory cells to the area of damage [28, 52]. As for CCL8 and CXCL5, we found a higher expression of LIF in the peritoneal fluid samples of insufficient patients.

Taken together we found peritoneal fluid level of CCL-8/MCP-2, CXCL5/ENA-78, and LIF on the third postoperative day to be higher in patients with than without an anastomotic leakage. However, there are some limitations of the current study. First, the limited sample size of the small patient cohort due to prospectively collected probes could be an issue of confounding. Second, male sex, higher T status, and rectal anastomosis are known to be associated with higher AL rates and could depict confounder variables [1, 2, 4]. Therefore, we compared the clinical patient’s data with the anastomotic leakage rate. As described in the literature, in this study, male sex and rectal anastomosis were associated with an occurrence of an anastomotic leak. For the other tested clinical parameters like age, BMI, smoker, TNM category, R, grading, and neoadjuvant therapy, no association with the occurrence of an anastomotic leakage was assessed, resulting in a low risk of bias.

Third, only seven out of a high number of cytokines and chemokines were measured due to precise method of the Luminex® bead-based multiplex assay. Fourth, patients with other postoperative complications—like wound infection, urinary tract infection, pneumonia, and secondary carcinomas—were excluded in order to recruit a consistent homogenous patient collective with no other inflammatory postoperative comorbidities than AL. Hence, a difficulty could be to detect an anastomotic leakage with the assessed markers, if the patient has more than one postoperative complication. Therefore, further studies with a larger patient cohort are needed.

This study demonstrates an elevated level of the inflammatory markers CCL8/MCP2, LIF, and CXCL5/ENA-78 on the third postoperative day in peritoneal fluid samples from patients developing anastomotic leakage after colorectal surgery compared with patients without anastomotic leakage. ROC analysis showed that CCL8/MCP2, LIF, and CXCL5/ENA-78 might be potential markers for an early detection of anastomotic leaks after colorectal surgery. Further studies with a larger patient cohort with inclusion of different variables are needed to evaluate their potential as predictive biomarkers for anastomotic leakage.