Abstract
Objective
We aimed to evaluate the correlation of procalcitonin level with neutrophil to lymphocyte ratio (NLR) and to establish a cut-off value that can be used as an indicator parameter for oncological inpatient.
Methods
Study design is a cross sectional. The study was executed retrospectively on 640 patients with at least one oncological diagnosis. Demographic data, serum albumin, C-reactive protein (CRP), procalcitonin levels, and complete blood count data on the day of admission were attained.
Results
The average procalcitonin level was found to be 63.47% between 01/11/2021 and 30/01/2022 in the Clinic of Oncological Palliative Care, University of Health Sciences Türkiye, İzmir Tepecik Education and Research Hospital. The NLR cut-off value is 6.97 with sensitivity and specificity of 68% and 64%, respectively. There was a medium-level positive correlation between procalcitonin with leukocyte and neutrophil number (r=0.419).
Conclusion
The NLR parameter, which can be determined through complete blood count, can be used nearly as effectively as albumin or CRP for the detection of infection. This approach might prevent unnecessary antibiotic usage, thereby affecting the cost-effectiveness issue positively.
Introduction
Procalcitonin is an acute-phase protein that has been frequently used in recent years for the early diagnosis of bacterial infections, especially septicemia(1). Procalcitonin comprises mainly the metabolization of thyroid hormone and secretion from adipose tissues and the lymphocytes.
Calcitonin is a peptide that is derived from a precursor protein. Serum levels start to rise within 24 hours during bacterial infections due to endotoxin production from parenchymal tissues. The level also rises post-operatively, after burns, heat stroke, and traumas(2). Suggestive levels of septicemia are 2 ng/mL, and those of septic shock are 11.6 ng/mL. If the measurement is >5 ng/mL for patients who are known to be in shock status, the reason might be considered to result from an infection(3).
Procalcitonin might be accepted as an indication for the use of antibiotics in cases of bacteremia while waiting for the result of culture(4). Laboratory parameters such as C-reactive protein (CRP), leucocyte, and neutrophil levels, can be evaluated jointly(5, 6).
CRP is the prototype acute phase reactant of the pentraxin group. The serum levels might increase in 24 to 48 hours during inflammation or infection, and also decrease rapidly(7). Normal levels of CRP are between 0.3 and 1.7 mg/L(8). The increase in the serum level to 10 mg/dL is a sign of bacterial infection. The levels can also be increased in situations such as dental illnesses, hypertension, obesity, smoking, diabetes mellitus, exercise, cancer, chronic fatigue, senility, alcohol consumption, and depression(9). The serum levels can be measured turbidometrically and nephelometrically(10).
Albumin, which constitutes 60% of plasma proteins, is a negative acute-phase protein. The serum and plasma levels decrease during infection, contrary to procalcitonin and CRP(11). Its levels also decrease in situations like lasting starvation, liver dysfunction, and intestinal malabsorption. Albumin is responsible for providing plasma oncotic pressure. It can also be used to evaluate the nutritional status of individuals. Its normal serum levels are 3.5 to 5.0 g/dL(12, 13). Neutrophil to lymphocyte ratio (NLR) is one of the recently proposed parameters for early diagnosis of septicemia (5, 14).
A complete blood count is a test that can be performed at almost any health facility, whereas a procalcitonin test cannot be implemented at every health institution due to its high cost.
The aim in this study was to evaluate the correlation of procalcitonin with NLR and also to establish a threshold level that can be used at early diagnosis during the courses of bacteremia and early phase septicemia. Additionally, the study aims to promote the NLR as an indicator parameter for inpatient clinics, especially for oncological ones where the risk for infection is increased.
Materials and Methods
Ethical approval was obtained from the Non-interventional Ethics Committee of University of Health Sciences of Türkiye, İzmir Tepecik Education and Research Hospital (decision no: 2021/10-28, date: 15.10.2021). The study retrospectively collected data on 640 patients treated at the care clinic in the hospital between November 2021 and January 2022. All the patients had at least one oncological diagnosis.
Demographic data, serum albumin, CRP, procalcitonin levels, and complete blood count data were obtained on the day of hospitalization. The cut-off levels of the biochemical and microbiological laboratories of the study center were procalcitonin: 0.06 ng/dL, albumin: 3.5 g/dL, white blood cell: 10.600/mm3, neutrophil: 7.000/mm3, CRP: 6.0 mg/L, platelet: 140.000/mm3.
Statistical Analysis
Data were analyzed with SPSS 26. Continuous variables were presented as mean and standard deviation, and categorical variables were presented as numbers and percentages. Continuous variables were tested for suitability for normal distribution. The Q-Q plot method and sample size were taken into consideration, and it was observed that they did not comply with the normal distribution assumption. Correlation between variables was investigated using the Spearman’s rho method. Independent group comparisons were made using Kruskal-Wallis and Mann-Whitney U tests, with a Bonferroni correction applied for multiple comparisons. Receiver operating characteristic (ROC) analysis was performed on variables that showed significant differences, and the most appropriate cut-off values were determined according to the Youden index. Binary variables were created for these values. Distribution differences of categorical independent variables were compared with the chi-square test, while the univariate odds ratio was calculated. Type 1 error was evaluated at an a level of 0.05 with a two-tailed test.
Results
The study was carried out with 640 patients; of these, 44.53% (n=285) were male. The percentage of individuals who were 65 years of age and over was 50.78% (n=325). The most prevalent oncological disease of the patients was breast cancer with 15.75% (n=97); the second most prevalent was stomach cancer with 7.03% (n=45).
The average procalcitonin level of the patients was 63.47 ng/mL. The averages for white blood cell number were 9.780/mm3; for neutrophil number, they were 7.830/mm3; and lymphocyte numbers, they were 1.130/mm3 (Table 1).
There was a medium-level positive correlation between procalcitonin and white blood cells and neutrophil number (r=0.419), a medium-level positive correlation between procalcitonin and NLR (r=0.365) (Table 2).
According to the ROC analysis, the area under the curve used to determine a cut-off value for NLR was calculated to be 0.697. Therefore, the cut-off was determined as 6.00. This condition was recorded with an ideal sensitivity of 68% and a specificity of 64% (Figure 1).
Discussion
Increases in procalcitonin levels, may be mistakenly interpreted as a positive sign of infection. Sometimes, overuse of antibiotics can mask an infection despite low procalcitonin levels. Therefore, it may be wise to set a cut-off value. Eraldemir(1) has intercrossed procalcitonin levels at cut-off values 0.1, 1, 2, and 5 ng/mL in separate groups with NLR, and he identified cut-off values for NLR to be 4.71, 6.44, 7.65, and 8.24 in the respective order. When 4.71 was used as the cut-off value for NLR, he determined the sensitivity to be 58.16% and the specificity to be 78.76%. Önmez et al.(15) have studied patients with pancreatitis and specified that the highest correlation of NLR parameter with the severity of pancreatitis was 68% in the ROC analysis, with an Area under the ROC curve of 0.687 (confidence interval: 0.570-0.804). Furthermore, the sensitivity of this value was determined to be 61% and the cut-off value has been confirmed to be 6.0451(15). Another study reported that NLR with a cut-off value of 4.68 can significantly differentiate between normal and inflamed appendices(16). Sayah et al.(17) have analyzed the potential fatality of coronavirus disease 2019. According to this, interleukin-6 has been shown to have a positive correlation with NLR where the cut-off value was 7.4, sensitivity was 96.3%, and specificity was 70.5%. Çil et al.(18) have specified NLR to be 7.21 in patients with pneumonia who had high procalcitonin levels. Our study revealed the NLR cut-off value to be 6.97 using ROC analysis, with sensitivity and specificity at 68% and 64% respectively. Therefore, we consider these values to be self-consistent and to have high predictive value for septicemia.
The lymphocyte number decreases while neutrophils increase during systemic inflammatory response and septicemia, thereby providing credibility for NLR increase in its medical implementation(19). de Jager et al.(20) revealed that NLR increase is more valuable in the early diagnosis of bacteremia at the emergency service compared to neutrophil and leucocyte increase. Beyazit et al.(21) revealed that NLR is superior to leucocytes in conditions of bacteremia and septicemia. We identified a medium-level positive correlation between procalcitonin and NLR. Therefore, we can consider that the implementation of NLR instead of procalcitonin, at least for oncological patients, might prevent unnecessary antibiotic usage and positively affect cost-effectiveness. One must bear in mind that a complete blood count provides an advantage as it is both fast and low-cost.
Tekin et al.(22) revealed that in patients presenting with high fever, CRP levels increase significantly with bacterial infection compared to those with viral. Yıldız Celtek et al.(23) found CRP levels to be significantly higher in patients. Ljungström et al.(5) determined that single use of NLR or procalcitonin has similar diagnostic power; nevertheless, using NLR with CRP acts as a confirmatory factor.
Arikan et al.(24) reported that there was a significant difference between all groups in NLR and platelet to lymphocyte ratio values(25). In a study conducted in Wuhan, it was determined that high values of NLR had high sensitivity and specificity in predicting mortality(26). There are studies in the literature in which NLR is successful in predicting sepsis and mortality. In this sense, this finding supports our study by corroborating our hypothesis.
Study Limitation
The most important limitation of our study is in investigating the presence of clinical variability changes, and it being conducted in a single center. Therefore, studies with larger samples are needed.
Conclusion
Both the cost and unattainability of procalcitonin at some health institutions directs clinicians to use alternative infection markers. We consider that the NLR parameter, which can be determined through complete blood count, can be used nearly as effectively as procalcitonin for the detection of infection. Further studies are needed to conduct more multicenter research, especially those without infection variability.
