Neutrophil-to-lymphocyte Ratio vs C-reactive Protein as Early Predictors of Anastomotic Leakage After Colorectal Surgery

Introduction Colorectal surgery (CRS) has improved immensely over recent decades as a consequence of better presurgical preparation, antibiotic prophylaxis, surgical technique, and postoperative management. However, it is may still be associated with somehow high morbidity and mortality rates. Prospective multicenter studies have demonstrated that 30-day morbidity and mortality may go up to 35% and 9% respectively. The most frequent postoperative complications include surgical site infection, intraabdominal abscess, ileus, hemorrhage, and anastomotic leakage (AL). Of these, AL is considered the most serious, with an incidence that varies from 3% to 15% and is responsible for up to a third of mortality in these patients. It is worth noting that most of them become evident between postoperative days 5 and 7, and the highest rates occur in coloanal anastomosis (10 to 20%). Early detection of serious postoperative complications such as AL in patients who undergo CRS is crucial for effective decision-making that anticipates septic complications. Although few markers have been studied as predictors of postoperative complications, in an era of enhanced recovery after surgery (ERAS) or Fast Track Surgery in which patients are discharged rapidly, the identification and use of pre-clinical inflammatory biomarkers may become of utmost importance to allow early diagnosis of serious complications. C-reactive protein (CRP) is an acute-phase reactant of hepatic synthesis, which raises in most of the inflammatory processes in response to proinflammatory cytokines as interleukin (IL) 6, IL-1 beta, tumor necrosis factor-alpha, and interferon-gamma. It has been noted that high concentrations of CRP at the third and fourth postoperative day may be associated with the development of septic intra-abdominal complications such as AL after CRS. On the other hand, pre and postoperative neutrophil-to-lymphocyte ratio (NLR), a simple and costless marker of a subclinical inflammatory response, has recently been identified as a useful predictor of major complications after surgery. Moreover, it has been recently suggested that preoperative NLR might be better than CRP as a predictor of 30-day morbidity after major abdominal surgery. With regards to CRS, it has been recently suggested that an elevated preoperative NLR may be a risk factor for major surgical complications following colorectal resection, with a trend towards the occurrence of AL. Although NLR may be advantageous in the clinical setting given that it has shown a faster kinetic pattern than CRP in response to surgical trauma, whether postoperative NLR can early predict the occurrence of AL in after CRS remains unknown.

The aim of the present study was to determine the value of postoperative NLR as an early predictor of AL in patients who undergo CRS and compare its diagnostic performance with CRP.

Material and methods Study design and population: The current report represents a single-institution retrospective cohort analysis of a prospectively maintained database. Data for consecutive patients submitted to colorectal surgery from June 2015 to April 2019 at a General Surgery Service of a tertiary referral University Hospital was extracted. All colorectal resections were included in the study analysis regardless on the approach used (open or laparoscopic), the confection of a diverting loop ostomy or the timing of surgery (urgent or scheduled). From a total 152 patients, 36 were excluded due to incomplete medical records, hematological diseases, extra-abdominal infections, definitive colostomy (i.e. Miles operation), or those who did not have a primary anastomosis (e.g. Hartmann's operation). Informed consent was obtained for all patients before surgery.

Surgical Technique and follow-up: All the procedures were performed by staff surgeons, including a colorectal surgery specialist. Regarding the techniques used for the anastomosis, all colorectal or ileorectal anastomosis were performed used staplers and ileocolic anastomosis were either performed using staplers or hand-sewn technique. After discharge, all patients were controlled through clinical evaluation and routine laboratory tests in an outpatient office at 2 weeks and 1 month after surgery, and every 4 months thereafter in oncological patients. Data on each patient was recorded prospectively from the date of index operation up to a minimum of 12 months after surgery or until death whichever occurred first.

Study variables and definitions: Patients demographics, indication for surgery, timing of the operation (elective or emergency), surgical risk according to ASA (American Society of Anesthesiologists), resection type, surgical approach (open or laparoscopic), need of diverting ileostomy, hospital stay and postoperative complications were analyzed. The Comprehensive Complication Index (CCI) was used to grade postoperative morbidity, which integrates in a single formula all the complications suffered by the patient according to the Clavien-Dindo score, summarizing postoperative (POP) morbidity in a grading score that goes from 1 to 100. Serum concentration of CRP and NLR during the first 5 postoperative days was recorded. All patients had complete cytology every day up to day 5 in order to calculate the ratio between the absolute value of neutrophils and the absolute value of lymphocytes. A CRP level below 5 mg/dL was considered normal. We analyzed the relationship between the variables and the development of AL. According to laboratory findings and clinical judgement, additional imaging studies were employed to rule out AL. AL was defined as suture line disruption with intestinal content leakage or abscess formation, associated to fever or abdominal pain, and confirmed by a CT-scan or re-operation up to 3 months after CRS.

Statistical Analysis: Continuous variables are expressed as means and standard deviation (SD) for symmetrically distributed, and median (range) for non-symmetrically distributed data. Categorical variables were expressed as frequencies (percentages). The Mann-Whitney test was used for comparison of continuous variables and Chi-squared test or Fisher's exact test was used for comparisons of categorical variables between patients with and without AL. The differences between variables were considered significant at a value of p≤0.05. Receiver operating characteristic (ROC) curves were used to evaluate the diagnostic accuracy of CRP and NLR as predictors of AL, determining sensitivity (Se), specificity (Sp), positive predictive value (PPV) and negative predictive value (NPV). The diagnostic accuracy of the plasmatic markers studied was based on the area under the ROC curve (AUC). We used the Youden Index to find the cutoff point with the best diagnostic performance of CRP and NLR on the first five postoperative days. SPSS versión 24 and GraphPad version 7 were used for statistical analysis.