Objectives
Evaluation of the predictive value of serial estimations of serum presepsin (PSP), Creactive
protein (CRP), and procalcitonin (PCT) for the development of infective
complications and/or anastomotic leakage (AL) after elective colorectal resection
with anastomosis for N0M0 colorectal cancers.
Patients and methods
During 7-year study period, 113 patients underwent colorectal resection;
postoperative (PO) morbidities were graded according to the comprehensive
complication index (CCI). Five venous blood samples (T1–5) were collected at
the time of induction of anesthesia, immediate, 1 day, 3 days, and 5 days after
surgery for blood leukocytic count, and calculation of neutrophil-to-lymphocyte ratio
and estimation of serum levels of the studied cytokines. Study outcome is the ability
of serum cytokines’ levels estimated in T2 sample to predict the possibility for the
development of AL.
Results
Incidence of PO infective morbidities was 31%, incidence of AL was 8.85%, and PO
mortality ratewas 2.65%.Operative timewas significantly longer; the total score of the
CCI was significantly higher andPOhospital stay was significantly longer for patients
who had AL. Patients who developed AL had significantly higher neutrophil-tolymphocyte
ratio in T3–5 samples than in the T1 sample. Mean serum levels of
the studied cytokineswere significantly higher inT2–5 samples than inT1sample with
significantly higher levels in patients who developed AL. Regression analysis defined
high serum levels of PSP and PCT in the T3 sample and CRP levels in T5 sample as
predictors ofAL and high levels of the three biomarkers in the T3 sample could predict
mortality. Automatic linear regression defined high serumPSPlevels in theT3sample
and long operative time as significant predictors for high CCI and bad outcome with
accuracy rates of 61 and 39%, respectively.
Conclusion
The level of serum biomarkers could predict AL, but PSP was superior to CRP and
PCT. High serum PSP levels in T2 sample can accurately predict high CCI for PO
complications