Abstract
Background. Carotid artery stenosis is often considered a stable clinical condition, and the underlying atherosclerosis is thought to have an inflammatory background.
Objectives. The aim of the study was to assess the value of different parameters obtained from whole blood counts for the prediction of advanced carotid artery atherosclerosis, including vessel occlusion, irrespective of symptom occurrence.
Materials and methods. The study group comprised 290 patients (84 (29%) females and 206 (71%) males) with a mean age of 68 ±8 years, who were admitted to the Vascular Surgery Department due to significant carotid artery disease. Patients were retrospectively divided into 2 subgroups regarding the presence or absence of artery occlusion. The demographic, clinical and laboratory preoperative data were compared between both groups.
Results. We found significant differences in preoperative large unstained cell (LUC) counts between patients with and without carotid artery occlusion (p = 0.003), when analyzed with the Mann–Whitney test for independent samples. The receiver operating characteristic (ROC) curve showed that LUC count has prognostic properties for carotid artery occlusion, with an area under the curve (AUC) of 0.637 (p = 0.033), yielding a 69.70% sensitivity and a 51.75% specificity.
Conclusions. Large unstained cells represent an acute inflammatory state related to artery occlusion. An LUC count below the cutoff value of 0.16×109/L may be a predictor of carotid artery occlusion. Therefore, carotid artery occlusion should not be regarded as a chronic state, but as a clinical challenge being promoted by active inflammatory processes.
Key words: inflammation, atherosclerosis, occlusion, carotid stenosis, large unstained cells
Background
Carotid artery disease is a challenging clinical problem that has recently been recognized as one of the most common causes of stroke.1 Instead of viewing age as the traditional risk factor, clinical considerations have assumed a greater significance for carotid plaques and stroke prediction.2 Asymptomatic carotid artery disease is diagnosed in less than 1% of those aged below 50 years, and in over 3% of patients over 80 years.3, 4 The monitoring of atherosclerotic plaques plays an important role in stroke prevention,5 with the prevalence of symptomatic intracranial stenosis being higher in elderly patients than in those younger than 70 years.6
Numerous clinical factors have been associated with the formation of atherosclerosis, including components of the metabolic syndrome, such as diabetes, obesity and hyperlipidemia, all of which show an elevated inflammatory response. In obesity, perivascular adipose tissue which surrounds blood vessels, where it becomes dysfunctional and secretes pro-inflammatory molecules, promotes the infiltration of inflammatory cells, and furthers the development of atherosclerosis.7, 8, 9
There is an increased body of evidence suggesting that the size of carotid artery atherosclerotic plaques, more so than their composition, plays a significant role in the clinical presentation of carotid artery disease.10 However, studies examining the makeup of plaques and subsequent remodeling and influence on mechanical forces should be taken into consideration with the use of ultrasound Doppler, computed tomography (CT) and positron emission tomography (PET), or magnetic resonance imaging (MRI) for predicting possible complications.11, 12, 13
Atherosclerosis is considered a lipid-derived disease with an inflammatory background.14 The inflammatory reactions initiate the formation of plaques if the endothelium becomes dysfunctional, while also facilitating disease progression.15, 16, 17 The infiltration of inflammatory cells has previously been presented as a hallmark feature of plaque instability,18 while a reduction of the inflammatory response was associated with plaque reduction in animal models.19 In our previous reports, we found a significant relationship between simple inflammatory indices obtained from the whole blood counts and overall mortality.20, 21, 22
Objectives
The current study aimed to assess the value of different morphological parameters obtained from whole blood counts for the prediction of advanced carotid artery atherosclerosis, including vessel occlusion and irrespective of symptom occurrence.
Materials and methods
Study patients
Three hundred ninety-one patients were admitted to the Department of Vascular Surgery at the Poznan University of Medical Sciences (Poland) between January 2018 and December 2020 due to significant carotid artery disease. Of this group, 290 patients (84 (29%) females and 206 (71%) males) with a mean age of 68 ±8 years underwent detailed laboratory evaluation and were enrolled in the final retrospective single-center analysis. The laboratory tests were performed upon hospital admission. The study group received carotid artery treatment including percutaneous (50 (17%)) and surgical (239 (82%)) interventions (Figure 1). One patient was disqualified due to a high perioperative mortality risk. Patients requiring unplanned intervention or concomitant surgery were also excluded from the study. Additional exclusion criteria encompassed inflammatory, autoimmune, oncological, or hematological proliferative diseases.
Thirty-one (11%) patients admitted for vascular intervention presented with carotid artery occlusion, while 259 (89%) had significant stenosis. Comorbidities included hypercholesterolemia 196 (68%)), arterial hypertension (178 (61%)), a history of stroke (146 (50%)), tobacco use (96 (33%)), diabetes mellitus (73 (25%)), and permanent atrial fibrillation (AF) (19 (7%)) (Table 1).
Study design – laboratory analysis
The blood samples for whole blood analysis were collected upon patient admission. The study group was divided retrospectively into 2 subgroups, namely those with carotid artery stenosis and those with carotid artery occlusion. The blood morphological results were compared between both subgroups, and the ability of various blood markers to predict carotid artery occlusion was analyzed.
Inflammatory indices were calculated, including neutrophil-to-lymphocyte ratio (NLR), monocyte-to-lymphocyte ratio (MLR), platelet-to-lymphocyte ratio (PLR), the systemic inflammatory index (SII) (the quotient of neutrophils and platelets divided by the lymphocyte count), the systemic inflammatory response index (SIRI) (the quotient of neutrophils and monocytes divided by the lymphocyte count), and the aggregate index of systemic inflammation (AISI) (the proportion of neutrophils, monocytes and platelets divided by the lymphocyte count).
Analysis
We analyzed demographic, clinical and laboratory data comprising whole blood count parameters with the use of a routine hematology analyzer (Sysmex Europe, Norderstedt, Germany).
The researchers adhered to the principles of good clinical practice and the Declaration of Helsinki, and the study was approved by the Local Ethics Committee of the Poznan University of Medical Sciences (approval No. 784/21, October 13, 2021).
Statistical analyses
Numerical data were presented as mean ± standard deviation (M ±SD) when they followed normal distribution (Shapiro–Wilk test). Otherwise, data were reported as medians and interquartile range (Q1–Q3), where Q1 is the lower quartile and Q3 is the upper quartile. Categorical variables were presented as counts and percentages. The comparison between carotid stenosis patients and occlusion patients was performed with the Student’s t-test when data followed normal distribution (Table 2) and variation between groups was homogenous (Levene’s test). When data did not follow normal distribution, the Mann–Whitey test was used. Categorical data were compared using the χ2 test. The receiver operating characteristic (ROC) curve analysis was used to find the parameters that have prognostic properties for the presence of occlusion. The cutoff point was estimated with the Youden’s index. The uni- and multivariable logistic regression with backward stepwise selection was performed to find factors which increased the occlusion risk. The results are presented as odds ratio (OR) and 95% confidence interval (95% CI). The inclusion criteria for the study was whole blood count results in the normal range to prevent outliers. The lack of multicollinearity of explanatory variables was checked by assessing the correlation and by estimating variance inflation factor (VIF). For all analyzed predictors, the VIF < 1.7. The assumption regarding the linear relationship between explanatory variables and the logic of the response variable was checked using the Box–Tidwell test. The statistical analysis was performed using MedCalc® Statistical Software v. 20.027 (MedCalc Software Ltd., Ostend, Belgium). The assumptions regarding logistic regression were performed using Stata 17 software (StataCorp, College Station, USA). A p-value of 0.05 was considered statistically significant for all tests.
Results
A total of 290 patients were analyzed, including 155 (53%) who presented with clinical symptoms (visual disturbances in 29 (10%) patients and vertigo in 31 (11%) patients). Stroke was reported in 146 (50%) patients, and there were 178 (61%) and 112 (39%) patients diagnosed with significant (more than 70% of lumen narrowing) right and left carotid artery disease, respectively. Collateral atherosclerosis of carotid arteries was present in 108 (37%) patients.
A total of 169 (58%) and 16 (6%) patients underwent surgery, while 29 (10%) and 8 (4%) underwent angioplasty procedures, in the stenosis and occlusion groups, respectively. It was found that 13 (5%) episodes of perioperative neurological complications were reported, including 11 (4%) transient ischemic attacks (TIA) and 2 (1%) strokes.
The Mann–Whitney test for independent samples revealed significant differences in preoperative large unstained cells (LUCs) between patients with carotid artery occlusion and stenosis (p = 0.003). The ROC curve analysis showed that LUC count has prognostic properties for carotid artery occlusion featuring an area under the curve (AUC) = 0.637, 95% CI: 0.58–0.69, and a p-value of 0.033 with a 69.70% sensitivity and a 51.75% specificity (Figure 2).
Univariable analysis
According to the univariate logistic regression analysis (Table 3), significant preoperative factors obtained from the whole blood analysis included LUC count below the cutoff value of 0.16 (OR = 2.47, 95% CI: 1.13–5.39, p = 0.024).
Moreover, the univariable analysis of demographical data including gender (OR = 1.58, 95% CI: 0.66–3.81, p = 0.300) or obesity (OR = 0.63, 95% CI: 0.20–2.01, p = 0.438) did not reveal significance of these factors towards an increased risk of coronary artery disease (CAD). Neither preoperative symptoms (OR = 1.06, 95% CI: 0.38–2.91, p = 0.915) nor standard risk factors including smoking (OR = 1.06, 95% CI: 0.38–2.91, p = 0.915), diabetes mellitus (OR = 1.33, 95% CI: 0.60–2.96, p = 0.472), hypercholesterolemia (OR = 1.27, 95% CI: 0.78–2.11, p = 0.611), hypertension (OR = 1.29, 95% CI: 0.60–2.78, p = 0.508), or body mass index (BMI; OR = 0.99, 95% CI: 0.90–1.11, p = 0.992) showed statistical significance for either coexisting CAD (OR = 2.08, 95% CI: 0.98–4.42, p = 0.058) or peripheral artery disease (OR = 1.95, 95% CI: 0.78–4.88, p = 0.151) (Table 3). The only significant factor discovered in our analysis was AF (OR = 3.09, 95% CI: 1.03–9.25, p = 0.043). Similarly, laboratory results were not predictive, including serum cholesterol (OR = 0.86, 95% CI: 0.62–1.19, p = 0.363), low-density lipoprotein (LDL; OR = 0.93, 95% CI: 0.63–1.36, p = 0.698), or high-density lipoprotein (HDL; OR = 0.67, 95% CI: 0.28–1.63, p = 0.383).
Multivariable analysis
According to the multivariate logistic regression analysis (Table 3), the only significant preoperative factor was LUC count below the cutoff value of 0.16 × 109/L (OR = 2.70, 95% CI: 1.22–6.03, p = 0.015) and AF (OR = 3.75, 95% CI: 1.22–11.65, p = 0.022).
Moreover, the logistic regression with the Hosmer–Lemeshow test for goodness-of-fit, log-likelihood ratio test p-values, and Nagelkerke pseudo R2 test of analyzed parameters were performed (Table 4). The tests revealed significance of LUC count and CAD co-existance for carotid artery occulsion.
Discussion
To the best of our knowledge, our study is the first to reveal the predictive value of LUC count for carotid artery occlusion. We present the results of a multivariable analysis of preoperative whole blood count, with a cutoff value of 0.16 × 109/L as a predictive factor, regardless of symptoms.
Contrary to popular belief, our analysis revealed neither predictive value of potential comorbidities of atherosclerosis, nor predictive values of laboratory results of serum cholesterol fractions for carotid artery obstruction. Moreover, while some studies23, 24 have shown gender dependence of peripheral atherosclerotic disease, we did not find such correlation with carotid artery occlusion. While our study examined the whole blood count analysis and compared its results to diagnostic tools such as ultrasound imaging in patients with defined carotid disease, our rationale was to find predictive indicators for carotid artery disease progression that could be performed during routine check-ups. Based on our results, we believe that LUCs can be regarded as a simple marker within whole blood that can help distinguish patients with more advanced stages of carotid artery disease.
The LUC population reflects activated lymphocytes and peroxidase-negative large cells that do not contain morphological features of lymphocytes, eosinophils, basophils, or neutrophils.25, 26 This population can include virally activated lymphocytes, plasma cells, pediatric lymphocytes, hairy cells, and peroxidase-negative blasts. Those cells are beyond clear classification but have been postulated to be clinically relevant during inflammatory states, viral infections and hematological malignancies.27, 28, 29. Their increased amount in whole blood analysis was found to be correlated with immunological activation.27 Vanker and Ipp have indicated LUCs as a valuable marker of both innate immunity and CD8+ lymphocyte activation.30 Previously, LUCs have only been analyzed in a small number of studies related to leukemia, myelodysplastic syndromes and viral infection.31 Though LUCs reflect organism activation to variable factors, we are the first to present their association with atherosclerosis.
Since LUCs represent activated lymphocytes, they can also be regarded as a carotid stenosis progression indicator and the underlying inflammatory activation indicator.32 The inflammatory origin of atherosclerosis has been previously presented and suggests this origin is involved in both disease initiation and progression.33, 34, 35 Simple parameters from whole blood count, including NLR, MLR and SIRI, were postulated to be related to atherosclerosis progression.36, 37, 38, 39, 40, 41, 42 However, in our analysis, none of these indices were related to carotid artery occlusion, although we previously presented the prognostic value of MLR for collateral carotid artery involvement.32 The significant difference between LUCs and lymphocyte, monocyte and neutrophil counts is based on the characteristics of the cells; LUC counts take into account activated cells, while other cell counts measure the concentration of the cells.
The standard inflammatory markers such as C-reactive protein (CRP) and erythrocyte sedimentation rate (ESR) were analyzed in patients with carotid disease as possible simple markers. In the study by Liu et al., CRP combined with lipoprotein-associated phospholipase A2 was found to be a significant marker for carotid atherosclerosis.43 Moreover, Boaz et al. showed the relationship between intimal thickening and CRP.44 However, Schmidt et al. did not reveal any interactions between vascular risk factors for carotid atherosclerosis and CRP, as CRP was found to be related to the severity of small brain vessel injury.45 Finally, the CRP concentration as a serological determinant of carotid plaques vulnerability was postulated by Fittipaldi et al., who demonstrated the relation between atherosclerotic plaques vulnerability evaluated in histological examination and CRP values >5 mg/L.46
Large unstained cells have been described as peroxidase-negative cells. Myeloperoxidase (MPO) is a protein of 140 kDa molecular weight, composed of a tetramer from 2 α chains (60 kDa) and 2 β chains (14 kDa).47 This protein is commonly used for the classification of acute leukemia during diagnosis.48 The MPO deficiency has been reported as of either primary (genetic) or secondary origin as a disease consequence,49, 50 and has been described in renal failure, cardiovascular disease and diabetes mellitus.51, 52, 53 According to previous reports, MPO deficiency can be related to chronic disease development, including the extent of brain damage during stroke.54 An increased number of LUCs as peroxidase negative cells in peripheral blood test, especially in patients with carotid artery occlusion, may be regarded as a secondary type of deficiency. This cell type is activated by various factors, although not those typically associated with monocyte or lymphocyte activation, and may be associated with vasculitis.55 Moreover, LUCs were related to acute inflammatory reaction.31 According to previous reports, LUCs should be regarded as a mixture of activated lymphocytes, monocytes and lymphoblasts, and therefore we concluded that chronic carotid artery occlusion induces an active inflammatory response.32, 56
The role of monocytes in atherosclerosis has been described as facilitating increased cytokine release and being involved in plaque destabilization.57, 58 The inflammatory indices have been gaining scientific attention in recent years due to their low cost and easy availability as a possible predictive tool in cardiovascular disease.59, 60
The role of lymphoblasts in atherosclerosis has not been previously investigated in a general population, although their presence in premature atherosclerosis in hematological diseases has been postulated.61 The results of our study may shed new light on the role of lymphoblasts, especially in the narrowing of arteries during the atherosclerotic processes.
The second factor which appeared predictive for carotid artery occlusion was AF. Obviously, inflammatory activation, with the co-occurrence of AF, may serve as a trigger and result in carotid artery disease, as AF may enhance the inflammatory response.62 However, the inflammatory activation related to several conditions and diseases, such as carotid artery disease, heart failure and acute coronary syndrome, may trigger AF occurrence.63, 64 It has already been observed in patients after cardiac surgery that AF in the early postoperative period may occur even without the arrhythmia in the patient’s history.65
Limitations
The study was performed as a single-center, retrospective study and involved only patients with advanced stages of carotid artery disease referred for surgical intervention. Future studies including patients with a wide spectrum of carotid artery severity would be beneficial. The results of the study may be relevant to subgroups of patients with carotid disease, irrespectively of clinical symptoms or comorbidities, indicating those who present with artery occlusion. Second, our study lacked a healthy control group. Although the results of logistic regression model are significant, the pseudo R2 is low, possibly due to the relatively small sample size and lack of control group.
Conclusions
Large unstained cells represent an acute inflammatory state related to artery occlusion, and their concentration below a cutoff value of 0.16×109/L may predict carotid artery obstruction. Carotid artery occlusion should not be regarded as a chronic state, but as a clinical challenge promoting an active inflammatory process.
