C-reactive protein in drainage fluid as a predictor of anastomotic leakage after elective colorectal resection C-reaktivni protein u drenažnoj

Background/Aim. C-reactive protein (CRP) is considered to be an indicator of postoperative complications in abdominal surgery. The aim of this study was to determine the significance of serial measurement of CRP in drainage fluid in the detection of anastomotic leakage (AL) in patients with colorectal resection. Methods. CRP values in serum and drainage fluid, respectively, were measured on the first, third, fifth, and seventh postoperative day (POD) in 150 patients with colorectal resection and primary anastomosis. The values obtained were compared between the groups of patient without complications of surgical treatment and those with AL. Results. Clinically evident AL was observed in 15 patients – in two (4.2%) patients with left colonic surgery, and 13 (12.6%) patients with colorectal anastomosis. Mean values of CRP were higher in the patients with AL than in the patients without complications, both in serum and drainage fluid, with the most significant differences recorded on the PODs 5 and 7 (p < 0.001). Correlation analysis showed a positive correlation between serum and drainage fluid CRP levels in both groups of patients. Serum and drainage fluid CRP values on the PODs 5 and 7 are most important in the detection of AL. In 80% of patients with CRP values in the drainage fluid of 53 mg/L for the POD 5 and 42 mg/L for the POD 7 AL was observed. The method specificity was 77% for the POD 5, and 83% for the POD 7. All the patients with CRP values in drainage fluid above 108 mg/L on the POD 5 and 93 mg/L on the POD 7 had AL. Conclusion. Serial measurement of CRP in drainage fluid can reliably be used in the detection of AL in patients with colorectal resection. The most significant values obtained on the PODs 5 and 7 were positively correlated with the values registered in serum.

Some studies including our one 12 have shown that elevated serum C-reactive protein (CRP) concentrations in the postoperative period may predict an increased chance of postoperative infection 13 and AL 14 .The fact that CRP begins to increase before the occurrence of postoperative infectious complications, in contrast to clinical signs such as fever, tachycardia, and pain 13 , makes it an ideal predictor of postoperative infectious complications.Since this is a nonselective marker of inflammation, before searching for specific, it is necessary to exclude the other infectious complications 13,14 .
According to our knowledge there are no data on the determination of CRP in drainage fluid in the detection of AL in patients with colorectal resection.Peritoneal fluid in its composition is a filtrate of plasma and it is in equilibrium with serum.Our assumption is that the values of CRP in drainage fluid reflect values obtained in serum.
The aim of this study was to determine the sensitivity and specificity of CRP in drainage fluid in the detection of AL in patients with elective colorectal resection.

Methods
Our prospective study enrolled 150 patients with cancer of the left colon and rectum surgically treated at the Clinic for General Surgery, Military Medical Academy, Belgrade in the period from April, 2011 to November, 2012.Characteristics of these patients were given in our previous article 12 .
The patients with clinical signs of infection or some other inflammatory condition present preoperatively were excluded from the study.The analysis involved only the patients in whom conventional elective, radical, or palliative surgical intervention was done, with colocolonic or colorectal anasomosis, handsewn or stapled.All the operations were doune by the surgeouns performing at least 30 similar surgical procedures per year.Mechanical preoperative large bowel preparation was done only in those with rectal cancers.The patients surgically treated for tumor recurrence were excluded from the study.The creation of diverting ileostomy or transversocolostomy depended on the individual assessment of surgeons.Before the closure of laparotomy incision, abdominal cavity of the patients was routinely drained with at least one drain placed in the area of the pouch of Douglas or in the presacral area, in the region of colorectal anastomosis.
In the immediate postoperative course, within a month of surgery, all remote (pneumonia, urinary infection, infections caused by central venous line) and surgical site (wound infection, AL, intraabdominal abscess collections) infectious complications were registered.Redness, edema, and purulent secretion at the site of laparotomy wound were the clinical criteria establishing the presence of infection in the surgical incision site 16 .Clinical parameters of AL were defined by the presence of purulent or fecal content at the drain site, pelvic abscess, peritonitis, rectovaginal fistula, or the appearance of purulent content from the rectum (per recti) 17 .Routine, contrast-enhanced x-ray control of the anastomosis was not implemented, since the patients with asymptomatic leakage, were not relevant for the study.In patients with low colorectal anastomosis, digital rectal examination was an integral part of the examination to detect possible AL.Intraabdominal abscesses were detected by way of the presence of purulent secretion after surgical or percutaneous ultrasoundguided drainage of these collections 16 .Appropriate clinical presentation with a positive x-ray finding in pneumonia, urinary sediment and urine culture in urinary infection, and positive blood culture in infections caused by central venous line, defined the presence of individual remote infections.
On the first, third, fifth, and seventh postoperative day (POD), serum CRP levels were measured, utilizing the method immunonephelometry on a SIEMENS autoanalyzer (DADE Behring BN II).Drain fluid was collected from the intraperitoneal drains placed in the pelvis and CRP levels were measured by the same method at the Institute of Medical Biochemistry, Military Medical Academy, Belgrade.
The usual descriptive statistic parameters were used in statistical analysis of the obtained results (mean value, standard deviation, range, 95% confidence interval, frequency of individual characteristics).Depending on the normality of distribution of the observed parameters and the number of groups among which the statistical significance was sought for, non-parametric tests, the Mann-Whitney U-test and Pe-arson´s correlation test were used.The sensitivity and specificity of relevant biochemical markers were analyzed using the receiver operating characteristic curve (ROC).
Commercially available statistical software package SPSS version 17 (USA) was used for statistical analysis.

Results
We analyzed 150 patients in total, 94 (66.7%) men and 56 (33.3%) women (male-to-female ratio, 1.7 : 1).The youngest operated patient was 33, and the oldest 87 years of age.The average age of the patients was 65 ± 11 years.
The overall morbidity rate associated with surgical treatment was 34%, and mortality rate 4%.Surgical site infections were observed in 41 (27.3%) and remote infections in 10 (6.7%) patients.In 99 surgically treated patients postoperative course was without any complications.
Clinically evident AL was observed in 15 patients -in 2 (4.2%) patients with left colonic surgery, and 13 (12.6%)patients with colorectal anastomosis.The postoperative mortality rate associated with AL was 13%, and out of 6 fatal outcomes, leakage was the immediate cause of death in two of the patients.
Table 1 shows a comparison of the mean values of CRP in serum and drainage fluid for the observed PODs between the group of patients without complications and those with AL.In all the patients, there was an increase of CRP, with maximum values reported for the POD 3, except for the CRP values in drainage fluid in the group of patients with AL.In those without complications, there was a descending tendency in serum and drainage fluid CRP values, significantly more evident compared to descending CRP values in those with AL.On the POD 1, there was no statistically significant differences in CRP values in the group of patients without complications and those with AL.On the PODs 5 and 7, there was a statistically most significant difference in the values of CRP in serum and drainage fluid between the group of patients without complications and with AL (p < 0.001).
Table 2 shows the results of correlation analysis for CRP values in serum and drainage fluid in the group of patients without complications and those with AL for the observed days.Statistically highly significant positive correlation (p < 0.001) between the value of CRP in serum and drainage fluid is present for all the observed days in patients without complications, while the highest correlation in the group with AL was observed on the POD 7 (the highest value of the correlation coefficient).
Sensitivity and specificity of serum and drainage fluid CRP measurement in the detection of AL were analyzed using the ROC curve and were shown in Table 3 and Figures 1 and 2.

Discussion
Despite many advances in surgery, the quest for uneventful healing of the intestinal anastomosis remains a challenge after colon and rectal resections.AL is the most serious complication of surgical treatment and a significant obstacle to the successful treatment of patients with colorectal resection 18 .It is more common after rectal surgery, between 8% and 14% 2,10,19 , compared to the colon, ranging from 3% to 7% 20,21 .
Differences in the frequency of, among other things, are due to different criteria used for the diagnosis of leakage.In order to better define, a new grading system was proposed by the International Study Group of Rectal Cancer (ISGRC) 22 .In analysis of our patients, we used exclusively the clinical criteria to establish AL 22,23 , and it was reported in 15 patients, in two (4.2%) patients with left colonic surgery, and 13 (12.6%)patients with colorectal anastomosis.The rates of morbidity and mortality significantly increase after AL, with mortality reported between 12% and 27% 1,24,25 .Postoperative mortality rate with AL in our study was 13%, and out of six deaths, leakage was the immediate cause of death in two.Early detection of this, potentially most dangerous complication, in the absence of clear clinical manifestations, would make possible an early introduction of appropriate therapeutic measures intended to alleviate or eliminate adverse effects.
CRP is the most popular and most widely available marker of the acute inflammatory response 26 .The production of CRP occurs almost exclusively in the liver by hepatocytes as part of the acute-phase response upon stimulation by interleukin (IL)-6, tumor necrosis factor-α, and IL-1-β originating at the site of inflammation.It is a pentameric protein with various molecular functions including complement activation and opsonization 27 .Within 6 h after stimulation, CRP serum levels exceed normal values and peak after about 48 h.CRP has a nearly constant serum half-life of about 19 hours.Therefore, the CRP serum concentration is determined by its synthesis rate and reflects the intensity of the stimulus for acute inflammatory responses 28 .In numerous studies, the significance of the serial measurement of CRP in serum in detecting infectious complications of surgical treatment and/or AL was shown [29][30][31][32] .
According to the available literature, we found no data on whether someone is determined CRP levels in drainage fluid.While some advocates 8 , and others do not find justification 33 , a routine approach in our institution involves mandatory drainage of the abdominal cavity after colorectal resection by placing at least one drain that runs through a separate incision in the skin.A few years ago another option proved to be interesting in patients with drainage done and refers to the ability to analyze the drainage of fluid in the purpose of early detection of AL 34 .Some biomarkers have been proposed as objective diagnostic parameters 35 and are collected from the fluid through the drain located near the site of interest.Potential biomarkers include immune (IL-1, IL-6 and IL-10, tumor necrosis factor alpha) parameters, tissue repair (matrix metalloproteinase-1, 2, 9 and 13) parameters, parameters of ischemia (glucose, lactate, glycerol), and microbiological (lipopolysaccharides, pH, pCO2, and pO2) parameters 35 .
In aim to verify the results obtained in drainage fluid in our patients, we compared them with the serum CRP values, between the group of patients without complications of surgical treatment and those with AL.The mean values of CRP in drainage fluid were lower compared with the values the serum for both groups of observed patients (without complications and with AL).The highest CRP levels were registered on the POD 3, except for the values in drainage fluid in the group of patients with AL with the highest CRP registered on the POD 5.In patients without complications, after an initial rise of CRP values, there was a gradual decline of CRP on days to follow.In those with AL, high CRP values persisted on the PODs 5 and 7 (Table 1).Similar trends in changes of CRP in serum is found in the papers of other authors 13,29,30  fluid between patients without complications and with AL.Statistically, the most significant difference occurred on the fifth and seventh POD (Table 1).
Correlation analysis showed a highly statistically significant positive correlation between the measured values of CRP in serum between the groups of patients without complications and those with AL for all the observed days.In the group of patients with AL statistically significant positive correlation was found on the PODs 1, 3 and 7, while on the POD 5 positive correlation was at the border of statistical significance (p = 0.055) (Table 2).We can conclude that changes in serum CRP values were accompanied by appropriate changes in the values of CRP in drainage fluid.
Serum and drainage fluid CRP values on the PODs 5 and 7 are most important in the detection of AL.The area under the ROC curve on the POD 5 was 0.920 and 0.960 on the POD 7 for serum CRP, and 0.879 and 0.824 for CRP values in drainage fluid.AL was observed in 85% of the patients on the POD 5 and in 92% on the POD 7, with serum CRP values of 77 mg/L and 90 mg/L.The method specificity was 77% for the POD 5, and 88% for the POD 7. All the patients with AL had serum CRP values above 139 mg/L on the POD 5 and 150 mg/L on the POD 7. In 80% of patients with CRP values in the drainage fluid of 53 mg/L for the POD 5 and 42 mg/L for the POD 7 AL was observed.The method specificity was 77% for the POD 5, and 83% for the POD 7. All the patients with CRP values in drainage fluid above 108 mg/L on the POD 5 and 93 mg/L on the POD 7 had AL.
However, as a nonselective marker of inflammation and is not a completely reliable indicator of infection, CRP could be taken into consideration only within the clinical presentation context 36 .

Conclusion
Our results indicate that measurement of CRP in drainage fluid can be reliably used in early detection of anastomotic leakage as the most serious infectious complication after colorectal resection.These values were positively correlated with CRP in serum, especially on the fifth and seventh day, when the risk from the appearance of this complication is greatest.Whether measurement of C-reactive protein in common with other biomarkers, in drainage fluid, contributes to more accurate early detection of patients at risk of developing anastomotic leakage after colorectal resection remains to be confirmed in future studies.Anastomotic leakage is inevitable companion of colorectal resection in a certain number of patients, and hence, its detection is of crucial importance because it allows early implementation of therapeutic measures to reduce the adverse effects.

Fig. 1 -Fig. 2 -
Fig. 1 -Diagnostic accuracy of serum C-reactive protein values in the detection of anastomotic leakage, expressed as a receiver operating characteristic curve; POD -postoperative day

Table 2 Correlation of C-reactive protein (CRP) values in serum and drainage fluid in the patients without complications and those
with anastomotic leakage (AL), by postoperative days (POD 1, 3, 5, 7) *Pearson'

Table 3 Sensitivity and specificity of serum and drainage fluid C-reactive protein (CRP) values in the patients with anastomotic leakage by postoperative days (POD 1, 3, 5, 7), expressed as an area under the ROC curve
ROC -receiver operating characteristic; AUC -area under the curve; CI -confidence interval.
. On the POD 1, there was no statistically significant difference in the values of CRP in serum and drainage Kostić Z, et al.Vojnosanit Pregl 2016; 73(3): 228-233.