Predictive value of inflammatory indexes in in-hospital mortality for patients with acute aortic dissection

Abstract

Background

The purpose of this study was to assess the relationship between admission inflammatory indexes neutrophil-to-lymphocyte ratio (NLR), monocyte-to-lymphocyte ratio (MLR), and systemic immune-inflammation index (SII), and the risk of in-hospital all-cause mortality in acute aortic dissection (AAD) patients.

Methods

A retrospective analysis was conducted on 597 AAD patients (Stanford classification: Stanford type A 365 patients, Stanford type B 232 patients) at a single center. Outcomes were the incidence of in-hospital all-cause mortality. The risk of all-cause death was compared between the groups with low and high inflammatory indexes using the Kaplan-Meier curve. The association between admission inflammatory indexes and outcomes was evaluated using the Cox regression model and restricted cubic splines (RCS). Stratified analysis was performed based on AAD type, age (< 50 years or ≥ 50 years), and gender.

Results

The Kaplan-Meier curves revealed statistically significant differences in outcomes among the low and high inflammatory indexes groups. Cox regression analysis revealed that the in-hospital mortality risk was significantly high in the high inflammatory index groups. MLR was the strongest associated with in-hospital mortality risk. The RCS curve revealed that NLR was non-linearly and J-shaped correlated with in-hospital mortality, and MLR and SII were linearly correlated with in-hospital mortality. Stratified analysis showed interactions between NLR, MLR, and SII and AAD type and age for the risk of in-hospital mortality.

Conclusion

Admission high inflammatory indexes were independently associated with an increased risk of in-hospital all-cause mortality in AAD patients. The inflammatory indexes NLR, MLR, and SII may be useful indicators for predicting in-hospital all-cause mortality in AAD patients.

Introduction

It is still difficult to diagnose and treat acute aortic dissection (AAD), a potentially fatal illness with significant rates of morbidity and mortality [1]. Medial degeneration with intima layer tearing and blood leaking into the arterial wall, which results in the creation of a false lumen inside the middle tunic, are characteristics of AAD [2]. The Stanford classification system categorizes aortic dissection into Type A and Type B. Type A involves dissections that affect the ascending aorta, regardless of the original site of the intimal tear, while all other cases are classified as Type B. In patients with acute aortic dissection, the mortality rate is 33% within the first 24 h if untreated, rising to 50% within 48 h, and approaching 75% if undiagnosed after two weeks. The fatality rate progressively increases by 1% per hour [3]. Predictive techniques for identifying patients with AAD who are at a higher risk of mortality are useful for risk stratification as well as guiding acute treatment.

Numerous in vitro and in vivo studies conducted in recent years have demonstrated that immune inflammatory responses may be important in AD development and are linked to poor patient outcomes [4]. Around the vessel and at the edge of the ruptured media, inflammatory cells may penetrate the aortic wall [5, 6]. Systemic immune-inflammatory biomarkers include the neutrophil-tolymphocyte ratio (NLR), and monocyte-to-lymphocyte ratio (MLR), which reflect the balance of immune response and the overall inflammatory environment [7, 8]. Furthermore, a thorough novel biomarker of inflammation that captures both the body’s general degree of inflammation and localized immunological responses is the systemic immune inflammation index (SII) [9]. These immune-inflammatory markers have been linked in the past to the prognosis of acute and chronic illnesses [10,11,12,13].

The association between inflammatory indexes and prognosis during hospitalization in patients with AAD remains unclear. Therefore, this study aimed to investigate the predictive usefulness of inflammatory indexes and assess the relationship between admission inflammatory indexes and the risk of in-hospital all-cause mortality in AAD patients.

Methods

Study design and population

659 AAD patients from Fuwai Hospital participated in this single-central, retrospective, observational cohort study between March 2017 and March 2018. 15 patients without platelet and white blood cell data and 12 patients without acute AAD type data were eliminated. In the end, the final statistical analysis included 597 participants (Stanford classification: 365 Stanford type A patients and 232 Stanford type B patients) (Fig. 1). In accordance with the Helsinki Declaration, the Ethics Committee of Fuwai Hospital gave its approval for this study. Subjects were not required to sign an informed consent form.

Fig. 1

Flowchart of the selection of patients. AAD, acute aortic dissection; WBC, white blood cell; PLT, platelets

Criteria for inclusion were: (1) Patients who are 18 years of age or older; (2) Magnetic resonance imaging or computed tomography thoracic aortography were used to diagnose two patients with AAD. Patients who refused to participate were excluded.

Data extraction

Patients were interviewed, their doctors were consulted, and medical records were reviewed in order to gather baseline data, which included demographics, history, vital signs, laboratory results, and treatment plans. Upon admission, all patients had venous blood samples taken for laboratory testing. The Data Monitoring Committee would routinely check that all of the data was accurate.

Assessment and definition of inflammatory indexes

Inflammatory markers were calculated as follows: NLR = neutrophil/lymphocyte; MLR = monocyte/lymphocyte; SII = platelet count × neutrophil / lymphocyte [14]. Participants were divided into high and low groups according to the quartiles of these inflammatory markers; those who scored more than the 3/4 quartile were labeled as high, and those who scored lower than the 3/4 quartile were categorized as low.

Endpoints

The study’s endpoint was the incidence of in-hospital all-cause mortality. Mortality from any cause that occurs while a patient is in the hospital is known as in-hospital all-cause mortality.

Statistical analysis

First, the normality of continuous variables was tested; otherwise, non-normally distributed data were expressed in the median with interquartile range (IQR), and regularly distributed data were expressed as mean ± standard deviation (SD). Absolute numbers with percentages were used to represent categorical variables. Continuous variables were compared using a t-test or Wilcoxon rank-sum test, and categorical variables were compared using chi-square or Fisher exact tests to identify differences in baseline characteristics between the survival group and death group.

The risk of all-cause death was compared between the groups with low and high inflammatory indexes using the Kaplan-Meier curve. Cox analysis, both univariate and multivariate, was used to evaluate the associations between inflammatory indexes and all-cause death risk. Model 2 adjusted for gender, age, AAD type, SBP, HR, smoke, and ascending aorta diameter, Model 3 further adjusted for surgery and TEVAR. For Cox regression, NLR, MLR, and SII were employed as continuous and categorical variables, respectively. To elucidate the non-linear correlation with the probability of outcome events, NLR, MLR, and SII were further examined as continuous variables using restricted cubic splines (RCS).

Additionally, stratified analysis was performed based on AAD type, age (< 50 years or ≥ 50 years), and gender. The capacity of NLR, MLR, and SII to predict the risk of all-cause death in AAD was assessed using the area under the receiver operating characteristic (ROC) curve (AUC). R software (version 4.3.2) and SPSS software (version 21.0) were used to run all statistical analyses, and a two-tailed P-value < 0.05 was deemed statistically significant.

Results

Baseline characteristics of study individuals

597 AAD patients (Stanford type A 365 patients, Stanford type B 232 patients) were included in this study. Males made up 74.71% of the sample, and the median (interquartile range) age was 53 (44, 62). There were 92 recorded fatalities over a median (95%CI) in-hospital duration of 12.00 (11.35, 12.65) days. In the death group, patients were older, more had a history of smoking, more were Stanford type A AAD, had higher SBP, HR, WBC, hemoglobin, neutrophil, ascending aorta diameter, NLR, MLR, and SII, had lower PLT and lymphocyte, more had rupture of AD, stroke, and massive pericardial effusion (all P < 0.05) (Table 1).

Inflammatory indexes associated with in-hospital all-cause mortality

Results for the groups with low and high inflammatory indexes differed statistically significantly (log-rank P < 0.05), according to the Kaplan-Meier curves (Fig. 2). AAD patients with high NLR, MLR, and SII had a higher risk of in-hospital all-cause mortality. As shown in Table 2, the results of Cox proportional-hazard regression analysis demonstrated that when considering inflammatory indexes as continuous variables, all the 3 models exhibited significant associations between NLR [Model 3 h, 1.030 (95%CI, 1.006–1.056) P = 0.016], MLR [Model 3 h, 1.809 (95%CI, 1.108–2.952) P = 0.018], and SII [Model 3 h, 1.000146 (95%CI, 1.000025–1.000266) P = 0.018] and the occurrence of in-hospital all-cause mortality. When employing NLR, MLR, and SII as categorical variables, Model 1 and Model 2 demonstrated an association between high inflammatory indexes and in-hospital all-cause mortality. The RCS curve revealed that NLR was non-linearly and J-shaped correlated with in-hospital mortality, MLR and SII were linearly correlated with in-hospital mortality (Fig. 3). The ROC analysis showed that the AUC (Area Under Curve) of NLR was 0.675 (95% CI 0.620–0.730), the AUC of MLR was 0.646 (95% CI 0.584–0.708), and the AUC of SII was 0.600 (95% CI 0.537–0.662) (See Fig. 4).

Fig. 2

Kaplan–Meier curves for unadjusted in-hospital mortality. (a) NLR (b) MLR (c) SII.NLR, neutrophil-to-lymphocyte ratio; MLR, monocyte-to-lymphocyte ratio; SII, systemic immuneinflammation index

Table 1 Baseline characteristics between the survival group and death group

Table 2 Cox proportional hazard models of inflammatory indexes with in-hospital mortality

Table 3 Subgroup analysis

Fig. 3

Restricted cubic spline regression analysis of inflammatory indexes with in-hospital mortality. (a) NLR Value of 8.476 was set as a reference (vertical dashed lines) for NLR to predict the risk of in-hospital mortality. (b) MLR Value of 0.562 was set as a reference (vertical dashed lines) for MLR to predict the risk of in-hospital mortality. (c) SII Value of 1500.819 was set as a reference (vertical dashed lines) for SII to predict the risk of in-hospital mortality. NLR, neutrophil-to-lymphocyte ratio; MLR, monocyte-to-lymphocyte ratio; SII, systemic immuneinflammation index; HR, hazard ratio; CI, confidence interval

Fig. 4

ROC of inflammatory indexes. AUC, Area Under Curve. (a) NLR (b) MLR (c) SII. NLR, neutrophil-to-lymphocyte ratio; MLR, monocyte-to-lymphocyte ratio; SII, systemic immuneinflammation index

Subgroup analysis

We found significant interactions between NLR, MLR, and SII and AAD type and age for the risk of in-hospital all-cause mortality (P for interaction < 0.01) (Table 3).

Discussion

The relationship between the risk of in-hospital all-cause mortality in patients with AAD and the admission inflammatory indexes NLR, MLR, and SII was assessed retrospectively in this study. The results demonstrated that admission NLR, MLR, and SII were independently positively associated with the risk of in-hospital all-cause mortality in AAD patients. Subgroup analysis in this study suggested significant interactions between NLR, MLR, and SII, and AAD type and age for the risk of in-hospital all-cause mortality. The inflammatory indexes NLR, MLR, and SII may be useful indicators for predicting in-hospital all-cause mortality in AAD patients.

Systemic immune-inflammatory biomarkers include NLR, MLR, and SII which reflect the balance of immune response and the overall inflammatory environment [7, 8]. These immune-inflammatory markers have been linked to the prognosis of acute and chronic illnesses, including non-alcoholic fatty liver disease risk, acute myocardial infarction, chronic kidney disease, and acute kidney injury [10,11,12,13].

Numerous studies have demonstrated the critical role inflammation plays in AD [2, 6, 15, 16]. A significant histologic characteristic of AD is medial degeneration, which is accompanied by a significant infiltration of inflammatory cells [17]. By entering the aorta wall through adventitia, wound tissue, or vascular pathways, inflammatory cells encourage the smooth muscle cells in the aortic tissue to undergo apoptosis, which results in AD [18, 19]. Patients with significant clinical symptoms and progression had higher levels of inflammatory cell activity in the aorta wall compared to those who were asymptomatic and clinically stable [5]. All of these suggest that inflammatory markers could be a useful predictor for accurately assessing AD patients’ prognoses. A retrospective study showed that neutrophil count was an independent risk factor for in-hospital mortality in patients with Stanford type A AAD [20]. Previous studies observed U-shaped associations between platelet-lymphocyte ratio (PLR) and in-hospital death in patients with Stanford type A and Stanford type B AAD [21, 22]. Admission NLR and PLR levels were independently associated with in-hospital mortality of Stanford type A AAD patients [23]. In patients with AAD undergoing surgery, a lower lymphocyte-to-monocyte ratio (LMR) was found to be independently linked to a greater in-hospital death rate [17].

In our study, patients in the death group had higher neutrophils, lower PLT, and lymphocytes. Neutrophils are a significant inflammatory agent and comprise 50–70% of circulating WBCs. Mechanical injury induced the expression of neutrophil chemoattractant [20]. The aneurysmal vascular wall is destroyed by the neutrophil excess, which encourages gradual growth and rupture [20]. Recent research indicates that lower PLTs are linked to in-hospital mortality in Stanford type A AAD, although the association between thrombocytopenia at admission and AAD outcomes is still complicated and unclear [24]. Activated platelets attract leukocytes and cause platelet aggregation by releasing inflammatory mediators into the surrounding microenvironment [25]. They may worsen aortic intimal damage, worsen coagulation and inflammation, and raise the risk of unfavorable results. Moreover, excessive platelet consumption might exacerbate organ ischemia and raise the risk of dissection rupture [22]. In patients with cardiovascular illnesses, low lymphocyte counts may indicate the degree of inflammation, which is linked to bad outcomes [26]. The release of catecholamines and inflammatory mediators, neurohumoral activation, and a rise in serum cortisol are all triggered by AAD, which lowers the lymphocyte count [5]. Significant and chronic lymphopenia is frequently seen in AD patients with unfavorable prognoses [17]. Aortic dissection rupture was associated with lymphatic infiltration, and there was a strong correlation between lymphoid infiltration and peripheral blood lymphocyte count [5]. According to a prior study, lymphocytes in these individuals’ aortas trigger death-promoting pathways, which in turn trigger the apoptosis of smooth muscle cells [6]. When considered collectively, inflammation has a significant impact on the pathophysiology and outcome of AAD.

Our study has several limitations. First, the data should be interpreted with caution because this study was retrospective and single-center. We performed subgroup analysis and accounted for a number of variables, but we were unable to totally rule out the impact of possible confounders on the results. Secondly, our study had a moderate sample size, and in order to validate our findings, larger cohort studies are required. The third, dynamic observation was not used in this investigation; only the baseline inflammatory indexes were examined. Lastly, it might not be suitable to extrapolate the results to other ethnic groups because this study only included Chinese patients.

Conclusion

In this study, we revealed the predictive value of inflammatory indexes for the risk of in-hospital all-cause mortality in AAD patients. Admission high inflammatory indexes were independently associated with an increased risk of in-hospital all-cause mortality in AAD patients. The inflammatory indexes NLR, MLR, and SII may be useful indicators for predicting in-hospital all-cause mortality in AAD patients.