El Zaher et al.

World Journal of Surgical Oncology (2022) 20:33

https://doi.org/10.1186/s12957-022-02506-4

RESEARCH Open Access

Role of the triad of procalcitonin,

C-reactive protein, and white blood cell
count in the prediction of anastomotic leak
following colorectal resections
Haidi Abd El Zaher1†, Waleed M. Ghareeb2,3*† , Ahmed M. Fouad4, Khaled Madbouly5*, Hamada Fathy1,
Tomas Vedin6, Marcus Edelhamre6, Sameh H. Emile7* and Mohammed Faisal1,8

Abstract
Purpose: The enhanced recovery after surgery (ERAS) program expedites patient recovery after major surgery. This
study aimed to investigate the role of the triad of procalcitonin (PCT), C-reactive protein (CRP), and white blood cells
(WBC) trajectories as a predictive biomarker for the anastomotic leak (AL) after colorectal surgery.
Method: Patients who had colorectal anastomosis were prospectively included. Postoperative clinical and laboratory
parameters and outcomes were collected and analyzed. The 5-day trajectories of PCT, CRP, and WBC were evaluated.
Based on the trajectory of the three biomarkers, we compared patients with and without AL as detected during the
first 30 days after surgery using the area under receiver operator characteristic curves (AUC) for logistic estimation.
Results: This study included 205 patients, of whom 56% were men and 43.9% were women with a mean age of 56.4
± 13.1 years. Twenty-two patients (10.7%) had AL; 77.3% underwent surgery, and 22.7% were treated with drainage
and antibiotics. Procalcitonin was the best predictor for AL compared to CRP and WBC at three days postoperatively
(AUC: 0.84, 0.76, 0.66, respectively). On day 5, a cutoff value of 4.93 ng/mL for PCT had the highest sensitivity, speci-
ficity, and negative predictive value. The predictive power of PCT was substantially improved when combined with
either CRP or WBC, or both (AUC: 0.92, 0.92, 0.93, respectively).
Conclusion: The 5-day trajectories of combined CRP, PCT, and WBC had a better predictive power for AL than the
isolated daily measurements. Combining the three parameters may be a reliable predictor of early patient discharge,
which would be highly beneficial to ERAS programs.
Keywords: Procalcitonin, C-reactive protein, Anastomotic leakage, Colorectal surgery, Biomarker

*Correspondence: [email protected]; khaled. Introduction

[email protected]; [email protected] Enhanced recovery after surgery (ERAS) programs incor-
†
Haidi Abd El Zaher and Waleed M. Ghareeb contributed equally to this
porates a panel of perioperative protocols and medica-
work.
2
Gastrointestinal Surgery Unit, Department of Surgery, Faculty tions. The use of a minimal access approach, pain killers,
of Medicine, Suez Canal University Hospital, Ismailia, Egypt antiemetic medications, and rehabilitation are commonly
5
Colorectal Surgery Unit, Alexandria University, School of Medicine,
used measures in ERAS programs. Overall, the main aim
Alexandria, Egypt
7
Colorectal Surgery Unit, General Surgery Department, Mansoura of ERAS is to diminish physiological stress, promote the
University Hospital, Mansoura, Egypt early return of total capacity, and decrease healthcare
Full list of author information is available at the end of the article
costs by shortening the length of hospital stay [1].

© The Author(s) 2022, corrected publication 2022. Open Access This article is licensed under a Creative Commons Attribution 4.0
International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you
give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes
were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated
otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not
permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To
view a copy of this licence, visit http://​creat​iveco​mmons.​org/​licen​ses/​by/4.​0/. The Creative Commons Public Domain Dedication waiver
(http://​creat​iveco​mmons.​org/​publi​cdoma​in/​zero/1.​0/) applies to the data made available in this article, unless otherwise stated in a
credit line to the data.
El Zaher et al. World Journal of Surgical Oncology (2022) 20:33 Page 2 of 10

The rate of anastomotic failure varies according to The study is registered to www.​clini​caltr​ials.​gov under
the site of anastomosis. The overall incidence of colo- number: NCT0515902
rectal anastomotic leakage (AL) ranges between 2% and
14% when surgery is performed by an experienced sur- Selection criteria and sample size calculation
geon [2–6]. Early AL usually becomes clinically evident We included adult patients of either sex who underwent
between 5 and 7 days postoperatively [7]. Dehiscence of colorectal surgery entailing an anastomosis. The exclu-
colorectal anastomosis may increase the local recurrence sion criteria included patients younger than 18 years,
rate of malignant tumors and postoperative mortality those with an active infection at the time of surgery,
secondary to peritonitis and septicemia [8]. The interest those who had received chemotherapy or radiotherapy,
in identifying a biological marker for the early detection and those on long-term corticosteroid therapy. The sam-
of AL is growing [9]. Such a marker could play a vital role ple size was calculated using online software for sample
in modern fast-track multimodal protocols, allowing safe size calculation for observational studies (http://​www.​
and early discharge of patients after colorectal surgery raoso​ft.​com/​sampl​esize.​html). In light of the popula-
with a low readmission rate. C-reactive protein (CRP) has tion size, referred to as the number of patients having a
been identified as a valid parameter for detecting post- colorectal anastomosis in our hospital per year (400–420
operative infectious complications after rectal resection patients per year), and with a margin of error set at 5%
[10]. A serum CRP level greater than 12.4 mg/dL on post- and confidence level set at 95%, a minimum sample size
operative day (POD) 4 is considered predictive of septic of 201 was required to be included
complications [11]. According to a recent analysis, the
changes in the trajectory of CRP levels might be more Preoperative assessment
beneficial than a snipped point. All patients were subjected to a thorough review of medi-
Moreover, the trajectory has negative predictability of cal history, physical examination, and routine preop-
up to 99.3% [12]. Another interesting biomarker is proc- erative investigations, including complete blood count,
alcitonin (PCT), the prohormone of calcitonin produced serum CRP, and PCT. Pelvic-abdominal ultrasonography,
by parafollicular C cells in the thyroid. Typically, it has a magnetic resonance imaging, and pelvic-abdominal com-
low plasma concentration in healthy individuals (0.01– puted tomography (CT) with double contrast were per-
0.05 ng/mL), and it increases during severe generalized formed with colonoscopy in the elective cases. Previously
bacterial, parasitic, or fungal infections, but not in non- published ERAS protocols were followed in the present
infectious inflammatory reactions [13]. Procalcitonin has study [19–22].
been described as an early, sensitive, and specific marker
of sepsis [14]. Moreover, the plasma concentration of Data collected
PCT has been used as an early predictor of infection in For each intervention, data on the surgical approach
acute pancreatitis [15], secondary peritonitis, and infec- (laparotomy or laparoscopy), underlying pathology, type
tious complications after thoracic, esophageal, and car- of resection (right/left hemicolectomy, rectal resection,
diac surgeries [16]. In addition, elevated white blood cell Hartman’s reversal, or closure of colostomy), and type
(WBC) count is associated with AL after gastrointestinal of the anastomosis (stapled or hand-sewn, end-to-end,
surgeries [17, 18]. Therefore, this study was conducted to side-to-side, or end-to-side) were recorded. The choice
evaluate the utility of CRP, PCT, and WBC count trajec- between open and laparoscopic surgery was driven by
tories, as separate and combined biomarkers for predict- the presence of contraindications for laparoscopy and,
ing AL after colorectal surgery. according to the patients’ desire, after surgeon coun-
seling. Choosing hand-sewn or stapled anastomosis was
decided according to resources available in our center.
Patients and methods
Study design and setting Postoperative assessment and outcomes
The present study is a prospective cross-sectional study Patients were examined daily to assess their clinical con-
on consecutive patients who underwent elective or emer- dition in terms of the presence of pain, fever, hemody-
gency surgery with a colorectal anastomosis between namic status, abdominal examination, return of bowel
March 2018 and March 2020 at the Surgical Oncol- function, wound condition, and hemoglobin level (if
ogy Unit of Suez Canal University Hospitals. All eligi- required).
ble patients provided written informed consent before The primary outcome of the study was AL which was
inclusion in the study. The research ethics committee has defined as a disruption in the integrity of the intestinal
approved the current study in the Faculty of Medicine wall at the anastomotic site that required surgical or
Suez Canal University with reference number (#8037). radiological intervention. Upon clinical suspicion, the
El Zaher et al. World Journal of Surgical Oncology (2022) 20:33 Page 3 of 10

diagnosis of AL was confirmed with pelvic-abdominal CT using ROC analysis for continuous variables or 2 × 2
with double contrast. Both clinical and radiological ALs, tables for binary variables (e.g., a trajectory of more than
including those managed conservatively, were included 50 mg/L increase in CRP, 0.5 ng/mL increase for PCT, or
in the present study. Clinical AL is defined as patients 1000/mm3 for WBC). P values of less than 0.05 were used
with evidence of a leak that needs active management, to denote statistical significance at a 95% level of confi-
whether therapeutic or surgical intervention. A discov- dence. Intra-individual variability was further evaluated
ery at reoperation, feculent drainage , fecal debris from with three indices:
the incision, extravasation of contrast on enema, or the
existence of air or fluid in the anastomotic area observed 1. Within-individual standard deviation (SD)
by computed tomographic (CT) scan were all considered 2. Coefficient of variation (CV) which was calculated as
anastomotic leaks [23]. Patients were then divided into the ratio of SD to the mean)
two groups according to the presence or absence of AL. 3. Variability independent of the mean (VIM), which
The two groups were compared concerning the following was calculated as the SD divided by the within-indi-
parameters: sex, age, underlying pathology, the urgency vidual mean to the power x and multiplied by the
of intervention (elective or urgent), surgical approach mean value of the laboratory parameter in the cohort
(laparotomy or laparoscopy), type of resection, length to the power x
of hospital stay, and postoperative morbidity and mor-
tality. C-reactive protein and PCT levels were measured The power x was obtained by fitting a curve through a
before surgery and on a daily basis until POD 5 or dis- plot of SD against mean laboratory parameter, using the
charge. Normal serum level of PCT in adults is < 0.1 ng/ model SD=a*mean x, where x is derived by nonlinear
ml and normal serum CRP is < 1 mg/dl. The assessment regression [26, 27].
of the CRP, PCT, and WBCs trajectories in predicting
AL was investigated prospectively. CRP was assessed by Results
immunonephelometry on an automated dimension Vista Patients’ characteristics
analyzer [24] (Siemens, Erlangen, Germany). Procalci- Overall, 217 patients who had colorectal surgery to diag-
tonin was assessed with homogeneous phase sandwich nose new or previous colorectal cancer were initially
enzyme-linked immunosorbent assay analysis (Brahms, screened. Twelve patients did not meet the study inclu-
Hennigsdorf, Germany) [25]. Patients were followed up sion criteria and were excluded, and thus 205 patients
for 30 days postoperatively to detect late AL. were ultimately included. Males represented 56.1% of
patients, while 43.9% were females. Patient age ranged
Statistical analysis from 25 to 78 years, with a mean of 56.4 years. The mean
All statistical analyses were performed using Statisti- body mass index was 29.6 ± 2.9 kg/m2. Approximately
cal Package for the Social Sciences, version 25.0 (IBM half of the patients had chronic medical conditions, par-
Corp., Armonk, NY, USA). Continuous variables were ticularly diabetes mellitus (20.6%) and cardiovascular
presented as mean and standard deviation (SD) if nor- conditions (14.2%), but none had liver cirrhosis or other
mally distributed or median and range if abnormally pathology. Approximately 84% of the patients had locally
distributed. Categorical variables were presented as fre- advanced tumor stages (T3/T4). A hemicolectomy was
quencies and percentages. All comparisons of continuous performed on 56.1% of the patients, whereas rectal resec-
variables were performed using the Mann–Whitney U tion was performed on 26.8%. Other procedures included
test. Biomarker trajectory was calculated as the average colostomy closure (10.2%) and Hartman’s reversal (6.8%).
linear trend from day 0 to day 5 (5-day trajectory) and Open surgery was performed in 86.3% of the patients
between every two consecutive days using linear regres- and laparoscopic procedures in 13.7%. Side-to-end and
sion. A logistic regression model for each biomarker tra- end-to-end anastomoses were the most common anas-
jectory and their combinations were used with AL as the tomoses performed (42% and 29.3%, respectively). The
outcome. The predicted values of the logistic regression hand-sewn anastomosis was performed on 6.5%, while
models were used to examine the predictive performance 38.5% were stapled. The mean operation time was 166
of biomarkers trajectories using receiver operating char- minutes (Fig. 1, Table 1).
acteristic (ROC) analysis. The respective areas under the
curve (AUC) and its 95% confidence interval were cal- Outcomes
culated to evaluate the predictive performance of CRP, One hundred and seventy-four patients (84.9%) were
PCT, and WBC for AL. The sensitivity (SN), specificity discharged without complications after a mean hospital
(SP), negative predictive value (NPV), and positive pre- stay of 10.7 ± 3.8 days. In contrast, thirty-one patients
dictive value (PPV) of these parameters were calculated (15.1%) experienced complications. Wound infection
El Zaher et al. World Journal of Surgical Oncology (2022) 20:33 Page 4 of 10

Fig. 1 Flow chart depicting the indications for and types of colorectal anastomoses performed in the study

was recorded in seven patients (3.4%), whereas respira- cutoff value of 4.93 ng/mL for PCT had the highest SN,
tory and urinary tract infections were recorded in 1.96% SP, and NPV (77.3%, 96.7%, and 97.3%, respectively).
and 1.46% of the patients, respectively. Anastomotic leak However, the cutoff values for CRP and WBC had their
was detected in 22 (10.7%) patients between the 6th and highest SN on PODs 1–3 and their highest SP on PODs
14th postoperative day (POD); 14 had clinical AL mani- 3–5. The highest NPV was achieved for CRP ≥ 114.1
fested by fever, pain, tachycardia, and peritonitis, and 8 mg/L and WBC ≥ 9.02 × 1000/mm3 on PODs 3 and 5,
had radiologically evident (subclinical) AL. Anastomotic respectively (97.5% and 97.7%, respectively) (Table 3).
leak was surgically treated in 77.3% of patients and was Trajectory analysis of PCT, CRP, and WBC count and
treated with percutaneous drainage and antibiotics in their combinations over the 5-day observation period
22.7% (Table 1). The anastomotic leak occurred mainly revealed that the PCT trajectory had the highest AUC
after left hemicolectomy, rectal resection, and Hart- compared with CRP and WBC (0.88 vs. 0.81 and 0.68,
man’s reversal (16.7%, 16.7%, and 14.3%, respectively). Six respectively). The predictive assessment of the PCT tra-
(2.9%) patients died after a mean of 16 PODs, 4 of whom jectory showed a substantial improvement when com-
had a rectal resection, and 2 had left hemicolectomy, but bined with the trajectories of either CRP or WBC, or
none of the deaths was related to AL. both (AUC: 0.92, 0.92, or 0.93, respectively) (Fig. 2). Fol-
low-up of the patients revealed no AL beyond 30-days
CRP, PCT, and WBC count measurements after surgery.
The mean preoperative CRP, PCT, and WBC levels of Furthermore, an increase of CRP of more than 50 mg/L
patients who developed AL were comparable to those between any two consecutive days had an AUC of 0.84
without AL. The mean postoperative CRP, PCT, and (SN: 90.9%, SP: 77.6%, PPV: 32.7%, NPV: 98.6%) with
WBC levels in patients with AL were significantly higher the highest predictive performance between POD 2 and
than in patients without AL starting from POD2 of the 3 (AUC: 0.85, SN: 81.8%, SP: 88.5%, PPV: 46.1, NPV:
5-day observation period following surgery (Table 2). 97.6%). Likewise, an increase of PCT more than 0.5 ng/
The predictive power of CRP, PCT, and WBC levels on ml between any two consecutive days had an AUC of 0.93
individual time points from POD 1 to POD 5 showed that (SN: 95.5%, SP: 89.6%, PPV: 52.4%, NPV: 99.4%) with the
the AUC of PCT was higher than that of CRP and WBC highest predictive value between POD 4 and 5 (AUC:
at each corresponding time point, particularly on PODs 0.92, SN: 86.4%, SP: 97.3%, PPV: 79.1, NPV: 98.3%). In
3, 4, and 5 (AUC = 0.84, 0.85, and 0.89, respectively). contrast, WBC trajectory had less value in predicting AL.
On other PODs, the AUC for PCT ranged from 0.73 to An increase in WBC of more than 1000/mm3 between
0.89, whereas those of CRP and WBC ranged from 0.51 any two consecutive days had an AUC of 0.75 (SN:
to 0.68 and from 0.51 to 0.89, respectively. On POD 5, a 100%, SP: 49.2%, PPV: 19.1, NPV: 100%) with the highest
El Zaher et al. World Journal of Surgical Oncology (2022) 20:33 Page 5 of 10

Table 1 Characteristics of the patients of the study (n = 205) Table 2 Distribution of CRP, PCT, and WBC among patients with
and without anastomotic leakage at different time points (n =
Variables Frequency (%)
205)
Sex
Measurements Without leakage With leakage P value
Male 115 (56.1%)
Female 90 (43.9%) CRP (mg/L)
Age (years), mean (SD), range 56.4 (13.1), 25–78 Preoperative 3.8 (1.8) 3.69 (2.09) 0.701
Comorbidities POD 1 19.80 (8.25) 19.91 (8.16) 0.891
None 101 (49.5%) POD 2 44.96 (28.45) 66.03 (24.77) 0.002*
Diabetes mellitus 42 (20.6%) POD 3 56.40 (45.02) 111.13 (51.00) < 0.001*
Cardiovascular 29 (14.2%) POD 4 61.02 (56.90) 142.14 (81.60) < 0.001*
COPD 22 (10.8%) POD 5 58.61 (65.57) 150.90 (120.17) 0.006*
Renal failure 10 (4.9%) PCT (ng/mL)
Tumor stage Preoperative 0.57 (0.34) 0.61 (0.30) 0.478
T1 3 (1.5%) POD 1 0.84 (0.45) 1.28 (0.55) < 0.001*
T2 29 (14.1%) POD 2 1.32 (0.58) 2.22 (1.15) < 0.001*
T3 110 (53.7%) POD 3 1.68 (0.92) 3.39 (1.49) < 0.001*
T4 63 (30.7%) POD 4 1.99 (0.95) 4.75 (2.21) < 0.001*
Type of operation POD 5 2.21 (1.08) 6.31 (2.65) < 0.001*
3
Right hemicolectomy 61 (29.8%) WBC (× 1000/mm )
Left hemicolectomy 54 (26.3%) Preoperative 6.41 (1.60) 5.85 (0.88) 0.187
Rectal resection 55 (26.8%) POD 1 6.96 (1.44) 6.98 (0.84) 0.983
Closure of colostomy 21 (10.2%) POD 3 7.99 (1.57) 8.79 (1.13) 0.014*
Hartman’s reversal 14 (6.8%) POD 5 8.34 (3.34) 10.01 (1.51) < 0.001*
Surgical approach CRP C-reactive protein, PCT procalcitonin, WBC white blood cell count, POD
Open 177 (86.3%) postoperative day
*
Statistically significant difference at p values of less than 0.05
Laparoscopic 28 (13.7%)
Values are presented as mean (standard deviation)
Type of anastomosis
End-to-end 60 (29.3%)
End-to-side 26 (12.7%) predictive value between POD 3 and 5 (AUC: 0.84, SN:
Side-to-end 86 (42.0%) 90.9%, SP: 77.0%, PPV: 32.2, NPV: 98.6%) (Table 4).
Side-to-side 33 (16.1%) The SD and CV indices of intra-individual variabil-
Anastomotic technique ity showed that intra-individual variability of CRP, PCT
Handsewn 126 (61.5%) and WBC was significantly different between patients
Stapled 79 (38.5%) with AL, and those without AL; patients with AL showed
Operation time (min), mean (SD), range 166.2 (19.5), 120–200 higher variability. Although VIM of CRP in AL was sig-
Hospital Stay (days), mean (SD), range 10.7 (3.8), 4–21 nificantly higher than non-AL, the difference for PCT or
Postoperative complications WBC was not significant (Table 5).
No complications 168 (81.9%)
Total complications: 37 (18.1%) Discussion
Anastomotic leakage 22 (10.7%) Anastomotic leak is considered a common and serious
Wound Infection 7 (3.4%) complication of colorectal surgery. Early recognition
Respiratory Infection 4 (1.96%) of AL is imperative to reduce mortality and morbid-
Urinary tract infection 3 (1.46%) ity. However, early detection of AL can be challenging
Mortality 6(2.9%) because early clinical and radiologic signs are rather
Management of anastomotic leakage (n = 22) non-specific. Therefore, accurate biomarkers for early
Reoperation 17 (77.3%) AL detection are highly required [28, 29]. This study
PCT drainage and antibiotics 5 (22.7%) found a strong association between AL and the trajec-
COPD chronic obstructive pulmonary disease, PCT percutaneous drainage tory of CRP, PCT, and WBC over 5 PODs. The area
under the curve (AUROC) is a commonly used metric
to determine the discriminatory power of a diagnos-
tic method. The AUROC has the highest value of 1.0,
signifying a (theoretically) flawless test. AUROC of
El Zaher et al. World Journal of Surgical Oncology (2022) 20:33 Page 6 of 10

Table 3 Predictive performance of CRP, PCT, and WBC for anastomotic leakage on isolated time points
Isolated time points AUC​ 95% CI Cutoff values SN SP PPV NPV

CRP (mg/L)
POD 1 0.51 0.44–0.58 12.5 90.9 24.0 12.6 95.7
POD 2 0.70* 0.63–0.76 62.6 81.8 67.2 23.1 96.9
POD 3 0.76* 0.69–0.81 114.1 81.1 85.2 40.0 97.5
POD 4 0.75* 0.68–0.81 168.4 72.7 94.5 61.5 96.6
POD 5 0.68* 0.61–0.74 210.5 59.1 97.8 76.5 95.2
PCT (ng/mL)
POD 1 0.73* 0.66–0.79 1.30 50.0 88.0 33.3 93.6
POD 2 0.73* 0.67–0.79 1.89 63.6 86.3 35.9 95.2
POD 3 0.84* 0.79–0.89 2.60 77.3 90.7 50.0 97.1
POD 4 0.85* 0.80–0.90 3.65 72.7 94.5 61.5 96.6
POD 5 0.89* 0.84–0.93 4.93 77.3 96.7 73.9 97.3
WBC (× 1000/mm3)
POD 1 0.50 0.45–0.59 6.11 90.9 30.6 13.6 96.6
POD 3 0.66* 0.59–0.73 7.50 91.0 41.5 15.7 97.4
POD 5 0.79* 0.73–0.85 9.02 86.4 71.0 26.4 97.7
CRP C-reactive protein, PCT procalcitonin, WBC white blood cell count, SN sensitivity, SP specificity, PPV positive predictive value, NPV negative predictive value, POD
postoperative day
*
Statistically significant different AUC from the reference diagonal line at p values of less than 0.05

Fig. 2 Area under the receiver operator curve estimates for the models predicting anastomotic leakage, including CRP, PCT, and WBC trajectories
and their combinations, over the 5-day postoperative observation period. CRP, C-reactive protein; PCT, procalcitonin; WBC, white blood cell count;
AUC, area under the curve
El Zaher et al. World Journal of Surgical Oncology (2022) 20:33 Page 7 of 10

Table 4 Predictive performance of CRP, PCT, and WBC Trajectories for the anastomotic leak
Trajectories AUC​ 95% CI SN SP PPV NPV

CRP > 50 mg/l

Between any 2 days 0.84* 0.79–0.89 90.9 77.6 32.7 98.6
   From POD 1 to POD 2 0.83* 0.76–0.88 86.4 80.3 34.5 98.0
   From POD 2 to POD 3 0.85* 0.80–0.90 81.8 88.5 46.1 97.6
   From POD 3 to POD 4 0.81* 0.75–0.86 72.7 89.6 45.6 96.5
   From POD 4 to POD 5 0.78* 0.71–0.83 59.1 96.2 64.9 95.2
PCT > 0.5 ng/ml
Between any 2 days 0.93* 0.88–0.96 95.5 89.6 52.4 99.4
   From POD 1 to POD 2 0.79* 0.73–0.84 68.2 89.6 44.0 95.9
   From POD 2 to POD 3 0.86* 0.80–0.90 77.3 94.5 62.9 97.2
   From POD 3 to POD 4 0.87* 0.81–0.91 77.3 96.2 70.8 97.2
   From POD 4 to POD 5 0.92* 0.87–0.95 86.4 97.3 79.1 98.3
WBC > 1.0 (× 1000/mm3)
Between any 2 days 0.75* 0.68–0.80 100 49.2 19.1 100
   From POD 1 to POD 3 0.72* 0.65–0.78 86.4 57.4 19.5 97.2
   From POD 3 to POD 5 0.84* 0.78–0.89 90.9 77.0 32.2 98.6
CRP C-reactive protein, PCT procalcitonin, WBC white blood cell count, SN sensitivity, SP specificity, PPV positive predictive value, NPV negative predictive value, POD
post-operative day
*
Statistically significant different AUC from the reference diagonal line at p value < 0.05

Table 5 Day-to-day intra-individual variability Indices of CRP, Being a serious complication [31], there were several
PCT, and WBC for anastomotic leakage attempts to investigate the risk factors for developing
Within No AL AL Mean difference p value AL [32]. Paliogiannis et al., as repoted the neutrophil to
table- (95% CI) lymphocyte (NLR), derived neutrophil to lymphocyte
individual (dNLR), lymphocyte to monocyte (LMR), and platelet
variability
to lymphocyte (PLR) ratios, however, the AUC as not
CRP SD 30.72 (24.9) 74.67 (33.8) 43.9 (32.4–55.5) < 0.001* exceed 0.744; 95% CI 0.719–0.768 in predicting AL [33].
CV 0.68 (0.22) 0.93 (0.14) 0.25 (0.16–0.35) < 0.001* Recently, PCT has been studied as a marker of early
VIM 31.90 (9.78) 41.45 (8.25) 9.56 (5.27–13.8) < 0.001* inflammatory changes earlier than CRP. PCT tends to
PCT SD 0.66 (0.45) 2.27 (1.09) 1.61 (1.12–2.10) < 0.001* reflect the magnitude of the systemic inflammatory
CV 0.47 (0.14) 0.71 (0.21) 0.24 (0.14–0.33) < 0.001* response in the first 12 h postoperatively, particularly
VIM 0.87 (0.37) 0.85 (0.27) − 0.02 (− 0.18–0.15) 0.673 in bacterial infection with a systemic response facilitat-
WBC SD 1.25 (1.55) 1.92 (0.60) 0.67 (0.01–1.33) < 0.001* ing early diagnosis of AL, allowing for early therapeu-
CV 0.17 (0.13) 0.24 (0.06) 0.08 (0.02–0.13) < 0.001* tic interventions, and conferring better outcomes [34].
VIM 1.67 (2.19) 1.63 (0.59) − 0.03 (− 0.96–0.89) 0.062 Nonetheless, PCT remains a questionable biomarker
OR odds ratio, CI confidence interval, SD standard deviation, CV coefficient of
for AL. Smith et al. reported lower accuracy of the PCT
variation, VIM variability independent of the mean trajectory than CRP (AUC: 0.763 vs. 0.961, respectively)
*Statistically significant p value (< 0.05); Mann-Whitney test [12]. In contrast, Garcia-Granero et al. found PCT to be
the most accurate biomarker with an AUC of 0.86 [35],
which is in line with the findings of our study.
0.5 implies no discriminative value and is depicted as Several studies have emphasized the utility of CRP as
a straight, diagonal line running from the bottom left a diagnostic indicator for AL. CRP has an adequate dis-
corner to the top right corner. Any measure with an criminatory capacity for AL with an AUC varying from
AUROC value > 0.9 has high preferential strength [30]. 0.69 to 0.87 [36–39]. The PREDICT study assessed the
The solo trajectory of PCT was the most reliable bio- trajectory of CRP levels along with the PODs on 833
marker compared to CRP and WBC for detecting AL, patients. The study concluded that the CRP trajectory
and its combination with the CRP and WBC trajecto- could accurately exclude postoperative AL [40]. How-
ries provided the maximum AL diagnostic accuracy ever, the PREDICT study did not assess the accuracy of
with an AUC of 0.93. combined trajectories of other potential biomarkers. Our
El Zaher et al. World Journal of Surgical Oncology (2022) 20:33 Page 8 of 10

study found biomarker trajectories to be more predictive values rather than trajectory [47, 48]. PCT has also been
of AL than their individual values at each POD. Moreo- investigated recently as an early marker for septic compli-
ver, the combination of various biomarker trajectories cations after surgery. The iCral study [49] and Spoto et al.
can maximize the predictive power. [50] have shown that high postoperative PCT levels are
Previous studies have demonstrated the possibility of associated with significant complications and suggested
using CRP as a marker for infection-related complica- that patients with high postoperative PCT levels should
tions after gastrointestinal operations [41]. It has been undergo imaging studies to search for surgical compli-
shown that the serum concentration of CRP signifi- cations. A meta-analysis by Cousin et al. confirmed that
cantly increases immediately after surgery and returns PCT, measured on POD 5, is a helpful biomarker for the
to normal on POD 3 in patients without complications. early diagnosis of intra-abdominal infection, including
According to Yeung et al., the increase in CRP after rectal AL, after colorectal surgery [51]. In another study, Giac-
resection suggests AL, and its evaluation in the postoper- caglia et al. showed that low levels of these two biomark-
ative period can be useful for early detection of AL. With ers on PODs 3 and 5 were associated with a low risk of
a cutoff value of 148 mg/dL on POD 3, the SN and SP of AL [52].
CRP was 95% [42]. In another study, Ortega-Deballon The strengths of the present study include the assess-
et al. reported that CRP is a good predictor for AL, with ment of the predictive power of three different biomark-
an AUC of 0.8 on POD 4 [37]. ers separately and combined, which was not reported on
Similarly, in a study by Zawadzki et al., serum CRP was previously. Limitations of the study include the single-
markedly elevated at the third POD among patients with center nature and relatively small numbers of patients
AL [28]. Messias et al. [43] did not find any statistically included. The predictive utility of the combined trajec-
significant alterations in serum CRP levels in the first 3 tory of the biomarkers assessed in the current study
PODs. However, starting from POD 4, serum CRP val- needs to be investigated in more extensive multicenter
ues in patients with AL were significantly higher than studies. Furthermore, data on other confounding vari-
those in patients without AL The highest CRP levels were ables such as diabetes, immunosuppressive medical con-
observed in patients with AL on POD 5. Su’a et al. [44] ditions, smoking were unavailable. In addition to AL,
analyzed 11 studies on AL and found a broad range of many other infectious diseases (for example: postopera-
CRP cutoff values on the third and fourth PODs, varying tive pneumonia) cause PCT, CRP and WBC to rise. Thus,
from 94 to 190 mg/L. Medications, such as corticoster- we have considered using “propensity score matching”;
oids and statins may change this reaction, reduce serum however, the sample size of our study, particularly the
CRP thresholds, and alter the perception of cutoff values number of patients with AL, was too small to perform
by 22%. Singh et al. [36] conducted a systematic review this type of analysis which may render it underpowered.
of seven studies, including more than 2400 patients [41], Therefore, we preferred not to lose the power of the study
and found that CRP levels were comparable on PODs 3, and proceed with the current approach of data analysis.
4, and 5.
Smith et al. reported CRP as the most accurate bio- Conclusion
marker for the anastomotic leak. However, the con- This study provides evidence on the usefulness of the
sistency of the CRP trajectory in the present study was combined triad of PCT, CRP, and WBC trajectories as
not observed [12]. Although the AUC values did not accurate biomarkers for AL after colorectal surgery, par-
exceed those known to be highly informative (0.81), CRP ticularly on PODs 3 and 5. Furthermore, this combina-
monitoring was proper as indicated by its NPV (97.5%) tion can be a reliable predictor for early patient discharge,
along with its high SP starting at POD 3. These findings which would be highly beneficial to ERAS programs.
were supported by the PREDICT study [40], which has
Acknowledgements
reported an AUC of 0.85 and NPV of 0.95 for CRP. Not applicable.
According to our study, the PCT trajectory is the best
solo predictor of major AL at PODs 1, 3, and 5 with a Authors’ contributions
WMG, HA, and MF conceived and designed the study. HA, HF, AF, MF and
maximum AUC of 0.89 on POD 5 with a cutoff value of WMG carried out the study and analyzed the data. HA and HF collected data.
4.93 mg/L and an NPV of 97.3%. Moreover, it can give KM, TV, and ME helped design the experiments. WMG and HA wrote the
a maximum discriminatory value for AL diagnosis when paper. KM, ME, and SE checked the paper. SE and TV performed language cor-
rection. All authors read and approved the final manuscript.
combined with the 5-day trajectories of CRP and WBC.
In terms of WBC as AL marker, it was not a point of Funding
investigation by some researchers [45, 46] or it has been Open access funding provided by The Science, Technology & Innovation
Funding Authority (STDF) in cooperation with The Egyptian Knowledge Bank
reported that there was no relation to AL, however, they (EKB).
have included only symptomatic AL patients and isolated
El Zaher et al. World Journal of Surgical Oncology (2022) 20:33 Page 9 of 10

Availability of data and materials 13. Vijayan AL, et al. Procalcitonin: a promising diagnostic marker for sepsis
All data are available on reasonable request and antibiotic therapy. J Intensive Care. 2017;5:51.
14. Urrechaga E. Reviewing the value of leukocytes cell population data
(CPD) in the management of sepsis. Ann Transl Med. 2020;8(15):953.
Declarations 15. Meher S, et al. Role of biomarkers in diagnosis and prognostic evaluation
of acute pancreatitis. Journal of Biomarkers. 2015;2015:519534.
Ethics approval and consent to participate 16. Slieker JC, et al. Procalcitonin-guided antibiotics after surgery for
This study was approved by the research ethics committee in the Faculty of peritonitis: a randomized controlled study. Gastroenterol Res Pract.
Medicine-Suez Canal University with reference number (#8037/2020). 2017;2017:3457614.
17. Liesenfeld LF, et al. Prognostic value of inflammatory markers for
Consent for publication detecting anastomotic leakage after esophageal resection. BMC Surg.
Written informed consent for publication was obtained from all participants. 2020;20(1):324.
18. Stearns AT, et al. Physiological changes after colorectal surgery suggest
Competing interests that anastomotic leakage is an early event: a retrospective cohort study.
All authors declare that they have no competing interests. Colorectal Dis. 2019;21(3):297–306.
19. Ramírez JM, Blasco JA, Roig JV, et al. Enhanced recovery in colorectal sur-
Author details gery: a multicentre study. BMC Surg. 2011;11:9. https://​doi.​org/​10.​1186/​
1
Surgical Oncology Unit, Department of Surgery, Faculty of Medicine, Suez 1471-​2482-​11-9.
Canal University Hospital, Ismailia, Egypt. 2 Gastrointestinal Surgery Unit, 20. Arroyo A, et al. Influence of size and complexity of the hospitals in an
Department of Surgery, Faculty of Medicine, Suez Canal University Hospital, enhanced recovery programme for colorectal resection. Int J Colorectal
Ismailia, Egypt. 3 Faculty of Medicine, Suez Canal University, Ismailia, Egypt. Dis. 2012;27(12):1637–44.
4
Department of Public Health, Community Medicine, Occupational & Environ- 21. Esteban F, et al. A multicentre comparison of a fast track or conventional
mental Medicine, Faculty of Medicine, Suez Canal University, Ismailia, Egypt. postoperative protocol following laparoscopic or open elective surgery
5
Colorectal Surgery Unit, Alexandria University, School of Medicine, Alexandria, for colorectal cancer surgery. Colorectal Dis. 2014;16(2):134–40.
Egypt. 6 Department of Surgery, Helsingborg Hospital, University of Lund 251 22. Moya P, et al. Perioperative immunonutrition in normo-nourished
87, Helsingborg, Sweden. 7 Colorectal Surgery Unit, General Surgery Depart- patients undergoing laparoscopic colorectal resection. Surg Endosc.
ment, Mansoura University Hospital, Mansoura, Egypt. 8 General Surgery 2016;30(11):4946–53.
Department, Sahlgrenska University Hospital, Gothenburg, Sweden. 23. Adams K, Papagrigoriadis S. Little consensus in either definition or diag-
nosis of a lower gastro-intestinal anastomotic leak amongst colorectal
Received: 26 September 2021 Accepted: 5 February 2022 surgeons. Int J Colorectal Dis. 2013;28(7):967–71.
24. Lagoutte N, et al. C-reactive protein and procalcitonin for the early detec-
tion of anastomotic leakage after elective colorectal surgery: Pilot study
in 100 patients. J Visceral Surg. 2012;149(5):e345–9.

tion: LIAISON® BRAHMS PCT® II GEN. Pract Lab Med. 2016;6:1–7.

25. Fortunato A. A new sensitive automated assay for procalcitonin detec-
References

surgery: Enhanced Recovery After Surgery (ERAS(®)) Society recommen-

1. Gustafsson UO, et al. Guidelines for perioperative care in elective colonic 26. Rothwell PM, et al. Prognostic significance of visit-to-visit variability,
maximum systolic blood pressure, and episodic hypertension. Lancet.
dations. World J Surg. 2013;37(2):259–84. 2010;375(9718):895–905.
2. Hyman N, et al. Anastomotic leaks after intestinal anastomosis: it’s later 27. Asayama K, et al. Prognosis in relation to blood pressure variability: con
than you think. Ann Surg. 2007;245(2):254–8. side of the argument. Hypertension. 2015;65(6):1170–9 discussion 1179.
3. Alves A, et al. Factors associated with clinically significant anastomotic 28. Zawadzki M, et al. Serum cytokines in early prediction of anastomotic
leakage after large bowel resection: multivariate analysis of 707 patients. leakage following low anterior resection. Wideochirurgia i inne techniki
World J Surg. 2002;26(4):499–502. maloinwazyjne. 2018;13(1):33–43.
4. Phitayakorn R, et al. Standardized algorithms for management of anasto- 29. Reynolds IS, et al. C-reactive protein as a predictor of anastomotic leak in
motic leaks and related abdominal and pelvic abscesses after colorectal the first week after anterior resection for rectal cancer. Colorectal Disease.
surgery. World J Surg. 2008;32(6):1147–56. 2017;19(9):812–8.
5. Frye J, et al. Anastomotic leakage after resection of colorectal can- 30. Fan J, Upadhye S, Worster A. Understanding receiver operating character-
cer generates prodigious use of hospital resources. Colorectal Dis. istic (ROC) curves. Cjem. 2006;8(1):19–20.
2009;11(9):917–20. 31. Shetty GS, et al. Complications as indicators of quality assurance after 401
6. Bruce J, et al. Systematic review of the definition and measure- consecutive colorectal cancer resections: the importance of surgeon vol-
ment of anastomotic leak after gastrointestinal surgery. Br J Surg. ume in developing colorectal cancer units in India. World J Surg Oncol.
2001;88(9):1157–68. 2012;10(1):15.
7. Li Y-W, et al. Very early colorectal anastomotic leakage within 5 post- 32. Kryzauskas M, et al. Risk factors for anastomotic leakage and its impact
operative days: a more severe subtype needs relaparatomy. Scientific on long-term survival in left-sided colorectal cancer surgery. World J Surg
Reports. 2017;7(1):39936. Oncol. 2020;18(1):205.
8. Cong ZJ, et al. Incidence and mortality of anastomotic dehiscence 33. Paliogiannis P, et al. Blood cell count indexes as predictors of anastomotic
requiring reoperation after rectal carcinoma resection. Int Surg. leakage in elective colorectal surgery: a multicenter study on 1432
2014;99(2):112–9. patients. World J Surg Oncol. 2020;18(1):89.
9. Scepanovic MS, et al. C-reactive protein as an early predictor for 34. Adamina M, et al. Meta-analysis of the predictive value of C-reactive
anastomotic leakage in elective abdominal surgery. Techn Coloproctol. protein for infectious complications in abdominal surgery. Br J Surg.
2013;17(5):541–7. 2015;102(6):590–8.
10. Straatman J, et al. C-reactive protein in predicting major postoperative 35. Garcia-Granero A, et al. Procalcitonin and C-reactive protein as early
complications are there differences in open and minimally invasive colo- predictors of anastomotic leak in colorectal surgery: a prospective obser-
rectal surgery? Substudy from a randomized clinical trial. Surg Endosc. vational study. Dis Colon Rectum. 2013;56(4):475–83.
2018;32(6):2877–85. 36. Singh PP, et al. Systematic review and meta-analysis of use of serum
11. Welsch T, et al. C-reactive protein as early predictor for infectious C-reactive protein levels to predict anastomotic leak after colorectal
postoperative complications in rectal surgery. Int J Colorectal Dis. surgery. Br J Surg. 2014;101(4):339–46.
2007;22(12):1499–507. 37. Ortega-Deballon P, et al. C-reactive protein is an early predictor of
12. Smith SR, et al. Biomarkers and anastomotic leakage in colorectal septic complications after elective colorectal surgery. World J Surg.
surgery: C-reactive protein trajectory is the gold standard. ANZ J Surg. 2010;34(4):808–14.
2018;88(5):440–4.
El Zaher et al. World Journal of Surgical Oncology (2022) 20:33 Page 10 of 10

38. Platt JJ, et al. C-reactive protein as a predictor of postoperative infective

complications after curative resection in patients with colorectal cancer.
Ann Surg Oncol. 2012;19(13):4168–77.
39. Kørner H, et al. Diagnostic accuracy of C-reactive protein for intraab-
dominal infections after colorectal resections. J Gastrointest Surg.
2009;13(9):1599–606.
40. Stephensen BD, et al. C-reactive protein trajectory to predict colorectal
anastomotic leak: PREDICT Study. Br J Surg. 2020;107(13):1832–7.
41. Neary C, et al. C-reactive protein-lymphocyte ratio identifies patients at
low risk for major morbidity after oesophagogastric resection for cancer.
Dig Surg. 2020;37(6):515–23.
42. Yeung DE, Peterknecht E, Hajibandeh S, et al. C-reactive protein can
predict anastomotic leak in colorectal surgery: a systematic review and
meta-analysis. Int J Colorectal Dis. 2021;36:1147–62. https://​doi.​org/​10.​
1007/​s00384-​021-​03854-5.
43. Messias BA, et al. Serum C-reactive protein is a useful marker to exclude
anastomotic leakage after colorectal surgery. Sci Rep. 2020;10(1):1687.
44. Su’a BU, et al. Systematic review of the role of biomarkers in diagnosing
anastomotic leakage following colorectal surgery. BJS (British Journal of
Surgery). 2017;104(5):503–12.
45. Wen J, et al. Nomogram to predict postoperative infectious complica-
tions after surgery for colorectal cancer: a retrospective cohort study in
China. World J Surg Oncol. 2021;19(1):204.
46. Fukada M, et al. Risk and early predictive factors of anastomotic leakage
in laparoscopic low anterior resection for rectal cancer. World J Surg
Oncol. 2019;17(1):178.
47. Qi X, et al. Risk factors of symptomatic anastomotic leakage and its
impacts on a long-term survival after laparoscopic low anterior resection
for rectal cancer: a retrospective single-center study. World J Surg Oncol.
2021;19(1):187.
48. Zhou S, et al. Predictive risk factors for anastomotic leakage after anterior
resection of rectal cancer in elderly patients over 80 years old: an analysis
of 288 consecutive patients. World J Surg Oncol. 2019;17(1):112.
49. Anastomotic leakage after elective colorectal surgery. A prospective mul-
ticentre observational study on use of the Dutch leakage score, serum
procalcitonin and serum C-reactive protein for diagnosis. BJS Open.
2020;4(3):499–507.
50. Spoto S, et al. The role of procalcitonin in the diagnosis of bacterial infec-
tion after major abdominal surgery: Advantage from daily measurement.
Medicine (Baltimore). 2018;97(3):e9496.
51. Cousin F, Ortega-Deballon P, Bourredjem A, Doussot A, Giaccaglia V,
Fournel I. Diagnostic Accuracy of Procalcitonin and C-reactive Protein for
the Early Diagnosis of Intra-abdominal Infection After Elective Colorectal
Surgery: A Meta-analysis. Ann Surg. 2016;264(2):252–6. https://​doi.​org/​10.​
1097/​SLA.​00000​00000​001545.
52. Giaccaglia V, et al. Procalcitonin, as an early biomarker of colorectal
anastomotic leak, facilitates enhanced recovery after surgery. J Crit Care.
2014;29(4):528–32.

Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in pub-
lished maps and institutional affiliations.

Ready to submit your research ? Choose BMC and benefit from:

• fast, convenient online submission

• thorough peer review by experienced researchers in your field
• rapid publication on acceptance
• support for research data, including large and complex data types
• gold Open Access which fosters wider collaboration and increased citations
• maximum visibility for your research: over 100M website views per year

At BMC, research is always in progress.

Learn more biomedcentral.com/submissions