Abstract
Objectives
We aimed to investigate the value of leukocyte biomarkers and disease scores for the early detection of infection in patients who have undergone elective colorectal surgery for malignancy.
Methods
We conducted a prospective study at a training and research hospital. Patients who developed infection were considered to be an Infection group, and the others were regarded as a Control group. For individuals in both groups, the Sequential Organ Failure Assessment Score (SOFA), quick SOFA, and National Early Warning Score (NEWS) were calculated and blood samples were collected for flow cytometry analysis. A model was developed using logistic regression analysis to identify parameters that were predictive of mortality.
Results
One hundred thirty-two patients were included in the study. Infections developed in 36 (27.3%) of the participants, of which 14 (38.9%) were intra-abdominal, 10 (27.8%) were pneumonia, 8 (22.2%) were superficial incisional infections, and 4 (11.1%) were urinary tract infections. The NEWS was the most effective parameter for the detection of early infection in patients undergoing surgery for colorectal malignancy.
Conclusion
The NEWS score can be easily used to predict infection soon after surgery for colorectal malignancy.
Keywords
Introduction
Infectious complications that arise in patients who undergo surgery for the treatment of gastrointestinal malignancy worsen their clinical condition and complicate their care. Furthermore, they cause an increase in the cost of treatment, and lengthen the postoperative recovery period and the hospital stay of the patients. 1 Various healthcare-associated infections, including surgical site infections and pneumonia, urinary tract infections (UTIs), gastroenteritis and bloodstream infections, have been reported during the postoperative period. 2 Factors such as malignancy-related immunosuppression, age and nutritional status, in addition to the intervention and colonisation by resistant bacteria owing to frequent hospital admissions, facilitate the development of healthcare-associated infections during the postoperative period.3,4
Infections that develop following surgery for gastrointestinal malignancy increase the risk of perioperative morbidity and may worsen survival. Therefore, the early diagnosis and initiation of treatment are essential. Postoperatively, there is typically an increase in the synthesis of inflammatory cytokines in patients as part of a systemic inflammatory response syndrome, which makes it challenging to identify infection when the clinical signs and symptoms are subtle. Although the sensitivities and specificities of the most commonly used laboratory parameters, the leukocyte count and C-reactive protein (CRP) concentration, are not high, they are still used in routine patient monitoring, owing to their rapidity, accessibility and cost-effectiveness. In contrast, the measurement of procalcitonin and interleukin-6 concentrations is costly and not routinely performed in every centre, and especially in resource-limited countries. However, previous studies have demonstrated their superior performance to CRP.5–8 In recent years, several studies have shown the utility of neutrophil CD4+ counting by flow cytometry and the neutrophil CD64 value for the identification of infection after cancer surgery.1,8,9 However, studies of CD64+ monocytes and CD15+ neutrophils have been rare. Disease scores such as the Sequential Organ Failure Assessment (SOFA) sand Quick (q)SOFA scores have been previously evaluated for their utility in determining the need for intensive care and the risk of mortality during the postoperative period. In addition, the National Early Warning Score (NEWS) was developed to predict the acute deterioration of patients in the emergency room, during hospitalisation, or in the intensive care unit.10–12 However, its value as a marker of early infection after colorectal malignancy surgery has not been investigated. In clinical environments with a heavy workload, such as surgery departments, there is a need for rapid, easily applicable, and accessible biomarkers and disease scores. Therefore, in this study, we aimed to investigate the value of leukocyte-related biomarkers and disease scores for the early detection of infection in patients who have undergone elective colorectal surgery for malignancy.
Methods
We performed a prospective observational single-centre study on patients who underwent elective colorectal malignancy surgery at a training and research hospital between August 2022 and October 2023. The study was approved by the local ethics committee, with reference number 2022/126. A diagnosis of colorectal malignancy was made histopathologically. The inclusion criteria were elective gastrointestinal surgery for the treatment of malignancy and agreement to participate in the study, through the provision of verbal informed consent. The exclusion criteria were the necessity for emergency surgery owing to the development of ileus and the performance of palliative surgical procedures.
Diagnosis of infection
Patients who developed infection were considered to constitute an Infection group, and those who did not were considered to constitute a Control group. A diagnosis of infection was made in the presence of symptoms such as purulent sputum production, high fever, redness, suppuration, watery and frequent bowel movements, dysuria and pollakiuria; and in the presence of signs such as meningeal irritation and auscultatory findings of rales and rhonchi; supplemented by bacterial culture for pathogen isolation. The occurrence of surgical site infections, pneumonia, central venous catheter-related bloodstream infections, UTIs and enterocolitis was documented. The criteria for nosocomial infections of the Centers for Disease Control and Prevention were used. 2
Data collection
Demographic information, including age and sex; the underlying disease (hypertension, chronic obstructive pulmonary disease, diabetes mellitus, congestive heart failure, chronic kidney failure or cerebrovascular disease) and the Charlson Comorbidity Index (CCI) values for the participants were recorded in the case follow-up form for each. Each of the participants received intravenous systemic antibiotic prophylaxis during the induction of anaesthesia.
On the day of the surgery (D0), blood samples were collected for the analysis of haematological parameters and the CRP (reference range 0–47.6 nmol/L) concentration. The haemogram and CRP concentration measurements were repeated on postoperative days 3 (D3) and 5 (D5). On D5, EDTA-anticoagulated blood samples were collected for the evaluation of neutrophil CD15 (nCD15), neutrophil CD64 (nCD64), and monocyte CD64 (mCD64) levels using flow cytometry.
Flow cytometry method
A three-laser, eight-colour, and 10-parameter flow cytometry device (DxFLEX, Beckman Coulter, Brea, CA, USA) was used to measure the nCD15, nCD64 and mCD64 levels in peripheral blood samples. A 2-ml EDTA-anticoagulated peripheral blood sample was obtained for each analysis, and all the samples were processed within 6 hours of collection. Fluorescent dyes (antiCD45–APC, antiCD15–PC5 and antiCD4–FITC; Beckman Coulter Life Sciences, Indianapolis, IN, USA) were employed. The nCD15, nCD64 and mCD64 data are expressed as the mean fluorescence intensity.
Disease scores
The qSOFA and SOFA scores and the NEWS were calculated for each participant on D5. The qSOFA score was calculated by awarding 1 point for low blood pressure [systolic blood pressure (SBP) ≤100 mmHg], 1 point for a high respiratory rate (≥22 breaths per minute) and 1 point for an altered level of consciousness (Glasgow Coma Scale score <15). Thus, the qSOFA score could range from 0 to 3 points. The SOFA score was calculated using seven parameters [the partial pressure of oxygen/fraction of inspired oxygen, platelet count, bilirubin concentration, creatinine concentration, Glasgow Coma Scale score, the mean blood pressure/administration of vasoactive agents/oxygen delivery, and the use of continuous positive airway pressure or invasive ventilation. The NEWS was based on six physiological parameters (respiration rate, oxygen saturation, SBP, pulse rate, level of consciousness or new confusion, and body temperature). The NEWS had a range of 0 to 19, with scores ≤4 corresponding to a low level of severity, 5 to 6 corresponding to an intermediate level of severity and ≥7 corresponding to a high level of severity.10–12 Comorbidities included in the CCI were identified using patient interviews and International Classification of Diseases 10th Revision codes. Patients were followed over a 28-day observation period, during which they were admitted to hospital or attended as outpatients, to monitor them for the development of healthcare-associated infections.
Statistical analysis
Statistical analyses were performed using IBM SPSS v.26 (IBM Corp., Armonk, NY, USA). The conformity of the data with a normal distribution was evaluated using the Kolmogorov–Smirnov and Shapiro–Wilk tests. Continuous datasets are described as mean ± standard deviation and categorical datasets are described using frequency and percentage. Student’s t-test and the Mann–Whitney U-test were used to compare datasets that were normally and non-normally distributed, respectively; and the chi-square or Fisher’s exact test was used to compare categorical datasets. A model was developed using logistic regression (the enter and forward methods) analysis to identify parameters that predicted mortality. P < 0.05 was considered to indicate statistical significance.
Results
A total of 141 patients underwent surgery for the treatment of colorectal malignancy. Six patients who underwent emergency surgery, three who underwent palliative surgery and two who could not be followed up after surgery were excluded from the study (Figure 1). Therefore, 132 patients were included in the study, of whom 50 (37.9%) were female and 82 (62.1%) were male. The ages of the participants ranged from 37 to 87 years and their mean age was 64.3 ± 14.2 years. Eighteen of the patients had diabetes mellitus, 10 had chronic lung disease, 6 had chronic heart failure, 6 had dementia, 2 had cerebrovascular disease and 2 had connective tissue disease. However, none of them had chronic renal failure that necessitated haemodialysis or cirrhosis. Approximately 79% of the participants had locally advanced tumours (stages T3/T4). Ninety-two (70%) participants had colonic cancer and 40 (30%) had rectal cancer. Most of the participants underwent a right colectomy (n = 41, 31.1%), and the others underwent a left colectomy (n = 19, 14.4%), transverse colectomy (n = 8, 6.1%), sigmoid resection (n = 26, 19.7%), low anterior resection (n = 27, 20.5%) or abdominoperineal resection (n = 11, 8.2%).
Flowchart of the inclusion and exclusion of participants in the study.
Infection developed in 36 (27.3%) of the participants. Of these infections, 14 (38.9%) were intra-abdominal, 10 (27.8%) were pneumonia, 8 (22.2%) were superficial incisional infections and 4 (11.1%) were UTIs. The mean time to the development of infectious disease was 7.6 ± 4.7 days. The demographic and clinical characteristics of the two groups are shown in Table 1. In the Infection group, the leukocyte counts on D3 (P = 0.018) and D7 (P = 0.004), the neutrophil counts on D3 (P = 0.001) and D7 (P = 0.005), the CRP concentrations on D3 (P = 0.001) and D7 (P = 0.001), and the nCD64 (P = 0.001), mCD64 (P = 0.001), qSOFA score (P = 0.001), SOFA score (P = 0.001) and NEWS (P < 0.001) were significantly higher than those of the Control group. The laboratory data and disease scores of the two groups are shown in Table 2. Forward logistic regression analysis revealed that the NEWS (P < 0.001, standard error: 0.171, Wald statistic: 31.393, Exp(B): 2.606, 95% confidence interval: 1.864–3.644) was the best early identifier of infection in patients who had undergone surgery for colorectal malignancy.
Demographics and clinical characteristics of the participants.
Data are mean ± SD or n.
Alcohol abuse was defined as a consumption of >2 units/day (1 unit = 12 g), and smoking: was defined as the smoking of one pack or more per day.
The data were analysed using Student’s t-test or the chi-square test, as appropriate.
Laboratory data and disease scores for the Infection and Control groups.
Data are mean ± SD.
Student’s t-test or *the Mann–Whitney U-test were used for the comparisons.
D, day following surgery; CRP, C-reactive protein; n, neutrophil; m, monocyte; SOFA, Sequential Organ Failure Assessment Score; qSOFA, Quick SOFA; NEWS, National Early Warning Score; MFI, mean fluorescence intensity.
Discussion
Infections following surgery for colorectal malignancy affect the prognosis of the disease, and therefore early diagnosis and treatment is important. In the present study, the leukocyte counts, neutrophil counts, CRP concentration, nCD64, mCD64, qSOFA score, SOFA score, and NEWS were found to be significantly higher in patients that developed infection after elective surgery for gastrointestinal malignancy than in those who did not. Kostic et al. 6 reported that 27.3% of 150 patients undergoing resection and primary anastomosis for colorectal cancer subsequently developed surgical site infection. In this study, CRP (cutoff value 695 nmol/L) had the highest sensitivity (79.6%) and specificity (88.8%) on postoperative day 6, with an area under the curve (AUC) of 0.88. Comesaña et al. 7 reported that 5.8% of 120 patients undergoing surgery for colorectal cancer developed intra-abdominal infection, and that their procalcitonin concentrations were high on D1 and D3 and their CRP concentrations were high on D3. In this study, when the threshold value for the D3/D1 ratio of the concentrations of CRP was set at 1.8, a higher sensitivity for the detection of infection was obtained (90.9%, with a specificity of 27.3%). In addition, procalcitonin had a sensitivity of 100% and specificity of 73.8% when a threshold concentration of 0.45 ng/mL on D3 was used. Facy et al. 13 reported an incidence of intra-abdominal infection of 11.8% in 501 patients undergoing colorectal surgery. In this study, CRP was found to be more discriminative than procalcitonin on D4 for the detection of intra-abdominal infection (P = 0.03). Xiao et al. 14 reported that procalcitonin concentration was a more reliable means of detecting infection after radical gastrectomy than the leukocyte or neutrophil count. Postoperative procalcitonin concentrations of <0.695 ng/mL on D3 and <0.515 ng/mL on D5 were found to be associated with a lower probability of postoperative infection. However, it is more expensive to measure procalcitonin concentration than to perform a complete blood count or CRP assay and is not used in many healthcare facilities with limited resources.
The measurement of nCD64 using flow cytometry has been considered in patients undergoing malignancy surgery.1,9 CD64 is an integrated membrane glycoprotein that binds monomeric IgG-type antibodies with high affinity and is present in macrophages and monocytes. High levels of CD64 in these cells indicates their activation and phagocytosis as part of the immune response. Therefore, the level of CD64 expression in neutrophils and monocytes can be used to provide evidence of infection. 15 The data obtained in the present study are consistent with this. Povsic et al. 1 reported that 15% of 200 patients undergoing surgery for colorectal cancer developed intra-abdominal infection, and that the best indicators of this were nCD64 on D1 and nCD64 and CRP concentration on D3. Achkasov et al. 16 found that of 73 patients undergoing colorectal resection, 13.7% developed infection, and that the nCD64 level was higher on D3 (P = 0.0017) and D6 (P = 0.018) in those who developed infection. Tomislav et al. 17 found that CD64 was the only biomarker that could predict postoperative infection in 229 patients who underwent colorectal, face or open-heart surgery (P = 0.001). Furthermore, nCD64 expression was found to be useful for the identification of postoperative infections and to predict early discharge from healthcare facilities. 18
Other studies have investigated the utility of further markers of postoperative infections. Butyrylcholinesterase (BChE) is an alpha-glycoprotein that is expressed in most tissues, and especially in the liver. Low BChE expression has been shown to be associated with higher risks of mortality, infectious diseases, and septic shock in patients undergoing liver transplantation surgery.19–22 Verras et al. 23 evaluated BChE expression in 402 patients undergoing colorectal surgery as a means of identifying surgical site infection. Multivariate analysis revealed that low BChE expression on D1 was associated with a 2.6-fold higher risk of developing surgical site infection [odds ratio (OR): 2.6, 95% confidence interval (CI): 1.3–3.9, P < 0.05]. Similar results were obtained for low BChE expression on D3 (OR: 2.5, 95% CI: 1.27–3.87, P < 0.05). Therefore, the monitoring of BChE expression may assist physicians with the routine follow-up of patients in the near future.
The qSOFA and SOFA scores and the NEWS, which were evaluated in the present study, were found to be predictors of infection within the 28 days following surgery for colorectal malignancy, and the NEWS score was the best predictor. The NEWS has been reported to predict clinical deterioration and the need for intensive care in both patients undergoing surgery and those hospitalised for non-surgical treatment.11,12 However, its utility for the prediction of healthcare-associated infection in patients undergoing surgery for malignancy has not been previously evaluated.11,12 Although the qSOFA score is often used by clinicians to diagnose sepsis, it is also predictive of in-hospital mortality for patients who are not in intensive care. 24 The qSOFA score includes the patient’s mental status, respiratory rate and SBP; and the NEWS includes these parameters, but also the routinely measured parameters of oxygen saturation, oxygen support, body temperature and heart rate.11,12 It is not expensive to obtain, because it does not include any laboratory parameters, but requires dynamic patient monitoring. The SOFA score is a scoring system that is based on clinical and laboratory parameters reflecting respiratory, cardiovascular, hepatobiliary, coagulation, neurological and renal function, and it is often used in intensive care units to predict sepsis-related mortality. 25 Studies comparing the qSOFA, SOFA and NEWS for the diagnosis of sepsis can be found in the literature. Oduncu et al. 26 found sepsis in 62% (287) of 463 patients who presented to the emergency department and shock in 13.8% of 64 patients. In this study, the sensitivities of the qSOFA, Systemic Inflammatory Response Syndrome (SIRS), and NEWS for the prediction of sepsis were 23%, 77% and 58%, respectively, with specificities of 99%, 35% and 81%. NEWS and qSOFA were found to be similarly good predictors, and more useful than SIRS (P < 0.001). Zhou et al. 27 compared the NEWS; Confusion, Urea, Respiratory rate, Blood pressure and age ≥65 years (CURB65); pneumonia severity index; qSOFA; SOFA; Emergency Department Sepsis Mortality Score (MEDS); and lactate concentration for the prediction of 28-day mortality, intensive care admission, and the need for mechanical ventilation in 340 patients with community-acquired pneumonia and sepsis who presented to the emergency department. In this study, the NEWS (threshold value 9) was the most useful predictor of 28-day mortality (AUC: 0.861), intensive care admission (AUC: 0.895) and the need for mechanical ventilation (AUC: 0.873). The NEWS is an easily calculable score that does not increase the physician’s workload. It can be easily used in surgical departments where postoperative intensive care is often required.
Managing the factors that facilitate the development of surgical site infection is as important as the early recognition of infection. Panos et al. 28 prospectively evaluated the risk factors for surgical site infection in 133 patients who had been diagnosed with colorectal cancer and underwent elective surgery. They identified age >70 years (P = 0.033), body mass index ≥30 kg/m2 (P = 0.021), American Society of Anaesthesiologists (ASA) score >2 (P = 0.028), diabetes mellitus (P = 0.021) and chronic steroid use (P = 0.049) as risk factors. Mulita et al. 29 prospectively evaluated 141 patients who underwent elective surgery for colorectal cancer with respect to the development of postoperative sepsis. Postoperative sepsis was detected in 18 (12.8%) patients, and the most common cause was anastomotic leak (3.6%). The risk of developing sepsis was highest in patients >65 years of age (P = 0.034), with ASA score >2 (P = 0.008) and with diabetes mellitus or cardiovascular disease (P = 0.013 and P = 0.009, respectively). Thus, age, concomitant disease, body mass index, ASA score, emergency vs. elective surgery, the use of open vs. laparoscopic methods and the duration of surgery also affect the risk of postoperative infection.
Conclusion
We have shown that the leukocyte counts on D3 and D7, the neutrophil counts on D3 and D7, the CRP concentrations on D3 and D7, nCD64, mCD64, qSOFA, SOFA, and NEWS are significant predictors of infection after elective surgery for colorectal malignancy, of which the best predictor was the NEWS score. The principal limitation of the present study was the relatively small number of patients, which was the result of the recruitment of a specific patient group at a single centre. However, to the best of our knowledge, this was the first study to evaluate the use of the NEWS for the prediction of infection after elective colorectal malignancy surgery. The NEWS is a disease score that can be calculated at the bedside, does not increase the workload of physicians, and does not incur additional costs, making it easily usable in surgical departments.
Footnotes
Acknowledgements
We extend our sincere gratitude to all the patients who participated in the study.
Author contributions
B.C. had full access to all of the data in the study and was responsible for conceiving the study. S.C., M.Y. and S.T. drafted the manuscript. S.K., B.C. and E.K. collected data. All the authors read and approved the final version of the manuscript.
Declaration of conflicting interest
The authors declare that there is no conflict of interest.
Funding
This study was supported by Eriskin Asi Dernegi (no. 2).
