Prognostic value of systemic immune-inflammation index, neutrophil-lymphocyte ratio, and thrombocyte-lymphocyte ratio in critically ill patients with moderate to severe traumatic brain injury

Abstract

Traumatic brain injury (TBI) is a significant health problem with a high mortality rate. Inflammatory markers can predict the prognosis of TBI where neuroinflammation is essential. In this study, the prognostic value of the systemic immune-inflammation index (SII), neutrophil-lymphocyte ratio (NLR), and platelet-lymphocyte ratio (PLR) at admission in patients with critical TBI was investigated. Patients with moderately severe TBI in the intensive care unit (ICU) of a tertiary center between June 2020 and June 2022 were retrospectively reviewed. Patients were classified into survivor and mortality groups. The predictive performance of SII, PLR, and NLR levels calculated from blood results at admission and 28-day mortality and patient outcomes were analyzed. One hundred sixty-one patients were included in this study. The median age of the entire population was 41 (18–90) years, and 80.7% (n = 130) of the patients were male. Falls (42.2%) and traffic accidents (40.4%) were the most common causes of TBI. The most common primary diagnoses in patients with TBI were acute subdural hematoma (30.4%) and subarachnoid hemorrhage (26.1%). The SII and NLR levels were significantly higher in the mortality group, and PLR levels were significantly lower (P = .004, P < .001, P < .001, respectively). In multivariate regression analysis, SII and PLR were independent predictors of mortality (P = .031 and P < .001, respectively). In the receiver operating characteristics (ROC) curve analysis, the cutoff value for SII was ≥ 2951, and the area under the curve (AUC) was 0.662 (95% CI, 0.540–0.784). The cutoff value for NLR was ≥ 9.85, AUC was 0.717 (95% CI, 0.600–0.834), and the cutoff value for PLR was ≤ 130.4, AUC was 0.871 (95% CI, 0.796–0.947). 28-day mortality was 21.1%. Neuroinflammation is essential in patients with critical TBI, and inflammatory markers SII, NLR, and PLR have prognostic importance. SII and PLR are independent predictors of mortality. Early detection of those with a poor prognosis in critically ill TBI patients and planning aggressive treatments may contribute to reducing mortality.

1. Introduction

Traumatic brain injury (TBI) is defined as the pathological changes in the brain that occur due to physical trauma.^[1] Although its incidence has increased in recent years, it is estimated that approximately 70 million people worldwide are affected by TBI annually.^[2] TBI is a significant cause of disability and mortality and has become a critical health problem that causes heavy economic loss.^[3] Intracranial hemorrhages, epidural and subdural hematomas, cerebral contusion, and diffuse axonal injuries are the most common causes of acute primary TBI. The inflammatory response after primary brain injury and the release of various mediators can trigger secondary brain damage.^[4] According to the Glasgow coma score (GCS), TBI is classified as mild, moderate, and severe. Mortality rates of 30–40% have been reported in critically ill patients with severe TBI.^[5,6] Identifying patients with poor prognosis among TBI patients admitted to intensive care units (ICU) is essential in applying aggressive treatment and preventing mortality.

Inflammation plays a vital role in TBI. Inflammatory response and cell activation, migration and recruitment of neutrophils, and release of inflammatory mediators can lead to secondary damage.^[7] Neutrophils are the most abundant cells in the circulation after TBI, both in the acute and post-injury phases. The number of circulating neutrophils increases significantly in patients with TBI compared to healthy controls and doubles within 3 to 4 hours.^[8] The systemic immune-inflammation index (SII), neutrophil-to-lymphocyte ratio (NLR), and platelet-to-lymphocyte ratio (PLR), which can be calculated from neutrophil, platelet, and lymphocyte counts, are known to reflect the balance between the immune status and the host inflammatory process. It has been reported that SII is essential in evaluating inflammation and helps predict prognosis in many clinical conditions, especially cancer.^[9,10] Although there are studies investigating the prognostic value of NLR and PLR levels in patients with TBI, there is no consensus regarding the prognostic value of SII.

This study aims to evaluate patients with critical TBI in the ICU of a tertiary center and to investigate the predictive performance of SII, NLR, and PLR levels at admission for 28-day mortality.

2. Materials and methods

This retrospective observational study was started by the principles of the Declaration of Helsinki after receiving approval from the local Clinical Research Ethics Committee (date: 02.08.2023 KAEK/2023.02.16). Patients with TBI who were followed up and treated in the ICU of the University of Health Sciences Turkey, Istanbul Kanuni Sultan Süleyman Training and Research Hospital, during the 2 years between 01.06.2020 and 01.06.2022, were retrospectively scanned.

Inclusion criteria are as follows: age ≥ 18 years; moderate to severe TBI (GCS ≤ 13); isolated head trauma; admitted within 24 hours of initial symptom onset, confirmed by magnetic resonance imaging or computed tomography. Exclusion criteria included: ICU stay of <24 hours; pregnancy; malignancy, autoimmune and hematological disease; use of medications that may affect platelet and lymphocyte counts (such as antiplatelet, glucocorticoids); viral or bacterial infection, including COVID-19, within 2 weeks before the trauma; currently active infection; missing data.

Demographic data, primary diagnoses upon admission to the ICU, trauma etiologies, comorbid diseases, length of stay in the ICU and mechanical ventilation, GCS and APACHE II (Acute Physiology and Chronic Health Evaluation II) scores, and 28-day mortality were recorded. Inflammatory markers SII, PLR, and NLR values were obtained from the patient blood results routinely checked during admission to the ICU. In patients admitted to the ICU after being operated on by a neurosurgeon, the blood test results were obtained from the preoperative emergency department. SII was obtained by multiplying the neutrophil and platelet counts and dividing by the lymphocyte count. PLR and NLR rates were calculated by dividing the platelet and neutrophil numbers by the lymphocyte number, respectively. Patients were classified into the survivor and mortality groups according to 28-day mortality. Inflammatory markers (SII, PLR, and NLR) and clinical data were compared between groups.

The sample size was calculated using the G* Power 3.1 program. The study primary aim is to investigate the relationship between SII, PLR, and NLR rates and mortality in patients with TBI during ICU admission. For T-tests, when P < .05, the effect size is 0.5, and the power of the study is determined as 80%, 140 patients must be included in the study. All patients who were followed up in the ICU due to moderate-severe TBI between the relevant dates and met the inclusion criteria were included in the study.

2.1. Statistical analysis

SPSS 26.0 (SPSS Inc., Chicago) program was used to analyze the data. Descriptive data were expressed as number of patients, percentage, median, and range. The conformity of the variables to the normal distribution was evaluated with the Shapiro-Wilks test and histogram. Quantitative variables that were not normally distributed were analyzed with the Mann–Whitney U test. The Chi-square and Fisher exact tests were used to evaluate qualitative data. Multivariate regression analysis was used to determine whether SII, PLR, and NLR differed significantly between the groups and were independent predictors of mortality. Logistic regression analysis results were presented as odds ratio and 95% confidence interval (CI). Receiver operating characteristics (ROC) curve analysis was performed to determine SII, PLR, and NLR prognostic value. The significance level was accepted as P < .05.

3. Results

One hundred sixty-one patients with moderate to severe TBI who were followed up in the ICU were included in the study (Fig. 1). The median age in the entire population was 41 (15–90) years, and 80.7% of the entire population was male (n = 130). The mean age of the patients in the survival group was significantly lower than the mortality group (P = .027). Blunt trauma was detected in 96.7% (n = 156) of head traumas. In the mortality group, GCS scores were significantly lower than in the survival group, and APACHE II scores were significantly higher (P < .001). Surgery was performed in 47.2% (n = 85) of patients with TBI. Operative status did not affect mortality (P = .117). Comorbid diseases were significantly higher in the mortality group (P = .038). Brain death was detected in 1.2% of the entire population (n = 2), and 1 patient became a donor for organ transplantation. The 28-day mortality in the entire population was 21.1%.

Figure 1.

Flow chart of the study.

When hematological data were compared between the groups, absolute neutrophil (median 27.1 vs 8.9) and lymphocyte (median 2.5 vs 1.2) counts were found to be significantly higher in the mortality group (P < .001). SII was significantly higher in the mortality group (median 2536 vs 1939, P = .004). In addition, NLR was found to be significantly higher (median 10 vs 7.8, P < .001), and PLR was significantly lower (median 97.1 vs 215, P < .001) in the mortality group (Table 1).

Table 1.

Demographic data of the groups and some clinical features.

	All population (n = 161)	Survivor Group (n = 127)	Mortality group (n = 34)	P value
Age (yr)	41 (18–90)	39 (18–90)	54.5 (18–89)	.028
Gender, n (%)				.824
Female	31 (19.3)	24 (18.9)	7 (20.6)
Male	130 (80.7)	103 (81.1)	27 (79.4)
Comorbidity, n (%)	52 (32.3)	36 (28.3)	16 (47.1)	.038
Trauma type, n (%)				.285
Blunt	156 (96.9)	124 (97.6)	32 (94.1)
Penetrating	5 (3.1)	3 (2.4)	2 (5.9)
Trauma severity, n (%)				<.001
Moderate	75 (46.6)	73 (57.5)	2 (5.9)
Severe	86 (53.4)	54 (42.5)	32 (94.1)
Neurosurgery, n (%)	85 (47.2)	63 (49.6)	22 (64.7)	.117
GCS	8 (3–13)	10 (3–13)	3 (3–10)	<.001
APACHE II	17 (4–52)	14 (4–45)	31 (20–52)	<.001
Duration of ICU (d)	7 (2–178)	6 (2–178)	15 (2–95)	<.001
Duration of Mv (d)	2 (0–75)	1 (0–50)	14.5 (2–75)	<.001
Platelet, ×109/L	254 (95–532)	257 (98–467)	230 (95–532)	.062
Neutrophil, ×109/L	10.1 (6.1–39.9)	8.9 (6.1–18.3)	27.1 (10.7–39.9)	<.001
Lymphocyte, ×109/L	1.3 (0.8–5.8)	1.2 (0.8–3.8)	2.5 (0.8–5.8)	<.001
SII	1996 (6.1–39.9)	1939 (742–3903)	2536 (1030–7420)	.004
PLR	197.7 (6.1–39.9)	215 (51.8–476.6)	97.1 (40.6–265)	<.001
NLR	8 (6.1–39.9)	7.8 (4.2–12.8)	10 (4.8–35)	<.001

Data are given as number of patients (n) and percentage, median, and range.

APACHE II = Acute Physiology and Chronic Health Assessment-2, GCS = Glasgow Coma Scale, ICU = Intensive care unit, Mv = Mechanical ventilation, NLR = Neutrophil-to-lymphocyte ratio, PLR = Platelet-to-lymphocyte ratio, SII = Systemic immune- inflammation index.

Falls (42.2%, n = 689), traffic accidents (40.4%, n = 65), and assaults (10.6%, n = 17) were the most common causes of TBI (Table 2). The most common primary diagnoses in patients with TBI were acute subdural hematoma (30.4%), subarachnoid hemorrhage (26.1%) and epidural hematoma (18.6%) (Table 3).

Table 2.

Trauma etiologies of patients with traumatic brain injury.

	All population (n = 161)	Survivor group (n = 127)	Mortality group (n = 34)
Falls	68 (42.2)	44 (34.6)	24 (70.6)
Traffic accidents	65 (40.4)	58 (45.7)	7 (20.6)
Assault	17 (10.6)	16 (12.6)	1 (2.9)
Work accidents, crush	6 (3.7)	6 (4.7)	0
Gunshot wounds/explosions	4 (2.5)	2 (1.6)	2 (5.9)
Stab wounds	1 (0.6)	1 (0.8)	0

Values are expressed as the number of patients (n) and percentage.

Table 3.

Primary diagnosis of patients with traumatic brain injury.

	All population (n = 161)	Survivor group (n = 127)	Mortality group (n = 34)
Acute subdural hematoma	49 (30.4)	35 (27.6)	14 (41.2)
Subarachnoid hemorrhage	42 (26.1)	32 (25.2)	10 (29.4)
Epidural hematoma	30 (18.6)	26 (20.5)	4 (11.8)
İntraserebral hematoma	20 (12.4)	15 (11.8)	5 (14.7)
Contusio cerebri	12 (7.5)	11 (8.7)	1 (2.9)
Brain edema	6 (3.7)	6 (4.7)	0
Pneumocephalus	2 (1.2)	2 (1.6)	0

Values are expressed as the number of patients (n) and percentage.

In the multivariate regression analysis of SII, PLR, and NLR, which are markers that showed significant differences between the groups, SII and PLR were found to be independent predictors of mortality (P = .031 and P = .001, respectively) (Table 4).

Table 4.

The multivariate logistic regression analysis results.

	Variables	OR 95% CI (min-max)	P value
SII	0.996	0.993–1.000	.031
NLR	0.484	0.211–1.110	.087
PLR	1.118	1.050–1.192	.001
Constant	1.469	0.848

CI = confidence interval (minimum-maximum), NLR = neutrophil-to-lymphocyte ratio, OR = odds ratio, PLR = platelet-to-lymphocyte ratio, SII = systemic immune-inflammation index.

In the ROC Curve analysis of SII, PLR and NLR, the cutoff value of SII in mortality prediction was ≥ 2951, area under curve (AUC) 0.662 (0.540–0.784), the cutoff value for PLR was ≤ 130.4, AUC 0.871 (0.796–0.947) and the cutoff value for NLR was ≥ 9.85, AUC 0.717 (0.600–0.834) (Table 5).

Table 5.

Prognostic performance of SII, PLR, and NLR for predicting in-hospital mortality.

	Cutoff	Sensitivity	Specificity	AUC (95% CI)
SII	2951	0.641	0.921	0.662 (0.540–0.784)
NLR	9.85	0.529	0.945	0.717 (0.600–0.834)
PLR	130.4	0.794	0.890	0.871 (0.796–0.947)

AUC = area under curve, CI = confidence interval (minimum-maximum), NLR = neutrophil-to-lymphocyte ratio, PLR = platelet-to-lymphocyte ratio, SII = systemic immune-inflammation index.

4. Discussion

In this study investigating the prognostic value of inflammatory markers in critically ill patients with moderate TBI, it was found that SII, NLR, and PLR, which can be easily calculated from blood results on admission to the ICU, help predict mortality. At the same time, multivariate regression analysis determined that SII and PLR were independent predictors of mortality in patients with TBI. When the prognostic values in predicting mortality were compared with ROC analysis, the areas under the curve were determined as PLR (0.87), NLR (0.71), and SII (0.66) from largest to smallest. It was determined that all 3 inflammatory markers at the time of admission to the ICU in patients with moderate to severe TBI, where high mortality rates are observed, may be helpful in the early detection of patients with a poor prognosis and in planning aggressive treatments.

TBI is a significant cause of morbidity and mortality in the population under the age of 40 in developed and developing countries. Although its global annual incidence is reported to be between 27 and 69 million, it is also known as the “silent epidemic” due to the increase in its incidence.^[11] It has been reported that the most common causes of TBI are traffic accidents and falls, and the most common primary diagnoses in patients are subarachnoid hemorrhage and acute subdural hematoma.^[12] In our study, consistent with the literature, the most common causes of TBI were falls and traffic accidents, and the most common primary diagnoses were acute subdural hematoma and subarachnoid hemorrhage.

There are 2 stages to TBI: primary and secondary injury. Primary damage is neural tissue damage that occurs after a mechanical impact. Neuroinflammation is effective in secondary damage. An inflammatory response can disrupt the blood-brain barrier (BBB), neurotoxicity, and neurodegeneration.^[13] Neuroinflammation is a complex response that can occur during the acute phase of TBI. It has been reported that the inflammatory response associated with TBI begins immediately after trauma and peaks within the first 24 hours.^[14] Immediately after trauma, various molecules, such as pro-inflammatory cytokines and chemokines, are released from damaged neuronal cells. These molecules activate neutrophils, resulting in phagocytosis of damaged cells remaining in this area. Neutrophils are essential components of the innate immune system and are the first and most important cells that respond immediately to injury. Neutrophil stimulation also leads to degranulation of granules and the release of bactericidal proteinases such as permeability-increasing proteins, elastase, and metalloproteinases. Inappropriate activation of endothelial cells can further disrupt the integrity of the BBB and lead to the passage of protein fluid into the interstitial space and significant leukocyte infiltration an increase in BBB permeability is observed hours after head trauma. In addition, disruption of the vascular wall leads to more plasma and molecules leakage into the extravascular space, exacerbating brain edema and secondary damage.^[15] Lymphocytes, part of the immune system, are essential for the body immune response, including the generation of antibodies and cell-mediated immunity. Activated platelets have inflammatory roles in several physiological and pathological conditions by affecting the permeability of endothelial cells and the functions of neutrophils and macrophages. NLR and PLR are simple and reliable markers calculated from neutrophil, platelet, and lymphocyte counts obtained from complete blood counts, and their use in clinical practice is increasing. NLR and PLR are generally considered indicators of inflammation before any clinical signs are observed. Studies have shown that high NLR and low PLR levels may be prognostic markers in detecting patients with poor TBI prognoses.^[16,17] Low platelet counts, and acute lymphocytosis explained this situation due to coagulation disorder and disseminated intravascular coagulation that occurred in the early period of trauma. In our study, consistent with the literature, median NLR levels were significantly higher, and PLR levels were significantly lower in the mortality group. Additionally, PLR was found to be an independent predictor of mortality.

It has been reported in the literature that high NLR levels help predict the functional outcomes of patients with spontaneous subarachnoid hemorrhage and are also predictive of the risk of rebleeding within the first 72 hours in ruptured aneurysms.^[18] Spontaneous bleeding, such as chronic subdural hematoma, may occur in surviving TBI patients later in life. It has also been stated that NLR levels are predictive of chronic subdural hematoma recurrence after trauma.^[19] The current study did not investigate the hemorrhagic recurrence profiles of patients with TBI after discharge.

SII was initially identified as a potential marker for tumor and cardiovascular diseases. Subsequent studies have reported that it is associated with disease severity and prognosis in various inflammatory conditions. SII is calculated by multiplying the number of platelets by the number of neutrophils and dividing by the number of lymphocytes.^{[9,10,17,20,21]} Since inflammation has a vital role in the pathogenesis of TBI, SII was thought to be a prognostic marker. However, there needs to be more studies in the literature investigating the prognostic role of SII in patients with TBI. Chen et al^[22] stated that the prognostic performance of SII was higher than PLR and NLR in patients with severe TBI. The authors evaluated the Glasgow outcome score of patients with TBI at 6 months post-discharge. In another study, it was suggested that high SII levels were an independent predictor of poor prognosis at discharge in patients with cerebral hemorrhage.^[23] Our study considered the 28-day mortality of patients with critical TBI. In this study, we contributed to the literature. We determined that SII, which can be easily calculated from blood results on admission to the ICU in patients with moderate TBI, can predict mortality and be used as an independent predictor of mortality. In our study, the cutoff level for SII was determined as 2951. SII values > 2951 can predict 28-day mortality with 64% sensitivity and 92% specificity in patients with critical TBI admitted to the ICU. However, the prognostic performance of SII was found to be lower than PLR and NLR.

GCS score can predict prognosis in trauma patients followed in ICU.^[24,25] It has been found that the prognostic role of NLR levels in predicting adverse outcomes in patients with severe TBI is similar to GCS.^[25] Our study found that GCS scores were significantly lower in the mortality group, consistent with the literature. However, the relationship between GCS and inflammatory markers was not investigated.

Our study has some limitations. The first is retrospective and single-center. Second, patients with mild TBI were not included in the study. Third, although patients with TBI were admitted to the emergency department within 24 hours after the head injury, the exact time of the head injury could not be determined.

5. Conclusion

In conclusion, SII, NLR, and PLR, low-cost inflammatory markers easily calculated from the blood count at admission, help predict mortality in critically ill TBI patients in the ICU. In addition, SII and PLR are independent predictors of mortality in patients with moderate to severe TBI.

Author contributions

Conceptualization: Kadir Arslan, Ayca Sultan Sahin.

Data curation: Kadir Arslan.

Formal analysis: Kadir Arslan.

Funding acquisition: Kadir Arslan.

Investigation: Kadir Arslan.

Methodology: Ayca Sultan Sahin.

Supervision: Kadir Arslan, Ayca Sultan Sahin.

Visualization: Kadir Arslan.

Writing – original draft: Kadir Arslan.

Writing – review & editing: Kadir Arslan, Ayca Sultan Sahin.