The Proportion of Circulating CD45RO+CD8+ T Cells is Associated with the Coronary Slow Flow

Abstract

Background

Circulating memory CD8+ T cells have been shown to be a crucial mediator of chronic inflammation. This study investigated whether the baseline proportion of circulating CD45RO+CD8+ T cells was associated with the coronary slow flow (CSF) phenomenon.

Methods

A total of 160 consecutive patients [mean (standard deviation (SD)) age, 67.86 (9.55) years; 51.25% male] who were admitted to our hospital between August 2020 and October 2020 for chest pain and underwent coronary angiography with the absence of coronary stenosis were enrolled in this cross-sectional analysis. The patients’ admission CD45RO+ CD8+ T cell plasma levels were measured using flow cytometry. Angiographic CSF was defined as thrombolysis in myocardial infarction (TIMI) flow of ≤ 2 without coronary stenosis, and non-CSF was defined as coronary arteries (< 50% stenosis) with TIMI 3 flow.

Results

The incidence of angiographic CSF was 22.5%. Patients with angiographic CSF had higher levels of CD45RO+CD8+ T cells than those without CSF [56.18 (13.93) vs. 45.26 (16.45); p < 0.001]. After multivariable adjustment, the risk of incident CSF was 2.41 [95% confidence interval (CI) 1.46-3.97] per SD change in CD45RO+ CD8+ T cells. Further, coronary microvascular resistance was significantly higher in patients with CSF than in those without CSF. A positive linear relationship between CD45RO+CD8+ T cells and coronary microvascular resistance was observed.

Conclusions

The proportion of circulating CD45RO+CD8+ T cells is an independent indicator of CSF. This observation may provide insights into the pathophysiological mechanism of CSF.

Abbreviations

ACEI, Angiotensin-converting enzyme inhibitor

ARB, Angiotensin receptor blocker

CAG, Coronary angiography

CFR, Coronary flow reserve

CMD, Coronary microvascular dysfunction

CSF, Coronary slow flow

GAM, Generalized additive model

IFN-γ, Interferon-γ

IL-17A, Interleukin-17A

LAD, Left anterior descending

MRs, Mineralocorticoid receptors

OR, Odds ratio

QFR, Quantitative flow ratio

SD, Standard deviation

TIMI, Thrombolysis in myocardial infarction

INTRODUCTION

Coronary slow flow (CSF) is angiographically defined as the delayed arrival of contrast matter in coronary arteries to the distal vascular bed in the absence of obstructive epicardial coronary disease. Distinguished from the delay in contrast progression in the context of coronary reperfusion therapy for acute myocardial infarction or other ‘secondary’ causes of coronary slow flow, ‘primary’ CSF occurs in up to 5.5-34% of patients with coronary angiography (CAG),¹ and accounts for 4% of unstable angina admissions.² CSF has a debilitating clinical course, and over 80% of CSF patients experience recurrent chest pain, and readmission to the coronary care unit is necessary in almost 20% of cases.³

The primary cause of CSF remains unknown; however, it is associated with coronary microvascular dysfunction (CMD).⁴ Traditional cardiovascular disease risk factors, including smoking, diabetes mellitus, aging, hypertension, and hyperlipidemia, account for less than 20% of the observed data variability,⁵ suggesting other unidentified risk factors for CMD. Interestingly, in various autoimmune diseases, the presence of CMD, as measured by coronary flow reserve (CFR), can be observed. Thus, systemic inflammation also appears to be an essential risk factor. Due to the potential role of inflammatory factors in the induction and perpetuation of CMD, mediators affecting inflammation, such as circulating CD45RO+CD8+ T cells, might be involved in CSF. A significant increase in CD45RO+CD8+ T cells in peripheral blood has been observed in patients with several autoimmune diseases.^6,7 In vitro experiments have confirmed that CD45RO+CD8+ T cells play a harmful role mainly by releasing inflammatory cytokines. In the kidney, memory T cells are predominant sources of interferon-γ (IFN-γ) and interleukin-17A (IL-17A).⁸ These cytokines can promote vascular dysfunction and damage, leading to increased systemic vascular resistance.⁹

As a crucial mediator of inflammation, the role of CD45RO+CD8+ T cells in CSF remains unclear. Therefore, this study sought to investigate whether pre-CAG plasma levels of circulating CD45RO+CD8+ T cells are associated with CSF.

METHODS

Study design and subjects

This was a cross-sectional analysis. Consecutive inpatients aged > 18 years with suspected angina who were admitted to Shanghai General Hospital from August 2020 to October 2020 were eligible to participate in this study. Patients were only included if angina was considered in the differential diagnosis by the attending physician and CAG was planned. Patients were excluded if they met any of the following exclusion criteria: 1) CAG showed the presence of epicardial coronary stenosis of ≥ 50%; 2) acute coronary syndrome; 3) inflammatory or infectious conditions; 4) treatment with anti-inflammatory drugs, such as antibiotics and non-steroidal anti-inflammatory drugs; 5) autoimmune disease or treatment with glucocorticoids; 6) allergic disease; 7) malignancy; 8) treatment with immunosuppressive agents; and/or 9) those who were unwilling or unable to participate in the study. None of the patients included in the study had been diagnosed with advanced liver disease, renal failure, or valvular heart disease. The Ethics Committee of the Shanghai General Hospital approved the study, and all patients consented to some of their blood being used for scientific purposes. All patients underwent the laboratory assays planned for the study during their hospital stay.

Flow cytometry analysis

Two ml of peripheral blood samples were collected in an ethylenediaminetetraacetic acid-coated tube (No. GD020EK Zhejiang Guangdong Medical Technology Co. Ltd., Zhejiang, China). The flow cytometry analyses were carried out using a 4-color Beckman Fc500 (Beckman Coulter, Fullerton, CA, USA) flow cytometer and analyzed using WinMDI 2.9 software (Joseph Trotter, Scripps Research Institute, La Jolla, CA, USA). To examine the expression of T cells, the peripheral blood cells were stained with fluorescein isothiocyanate-conjugated anti-human CD3, ECD, anti-human CD4, PC5 anti-human CD8, and anti-human CD45RO. All antibodies were acquired from Beckman (Beckman Coulter, Fullerton, CA, USA). Non-specific staining using isotype-matched control monoclonal antibodies was < 1%. The intra- and inter-assay variability were < 10%. All specimens were tested within 24 hours of collection.

Coronary angiography and CSF diagnosis

A cardiologist with ≥ 1 year of experience in interventional therapies for coronary heart disease performed the CAGs. One day before CAG, each patient was treated with 300 mg of aspirin and 300 mg of clopidogrel. During the procedure, a bolus of 3000 IU of heparin was administered. All CAG procedures were performed using a radical approach with a 5F catheter. All angiographic endpoints were evaluated after CAG, and the angiographer completed a standardized form. Patients with CAG results of epicardial coronary artery stenosis exceeding 50% were excluded. The thrombolysis in myocardial infarction (TIMI) flow grade was assessed at the angiographic core laboratory. Two interventional cardiologists who were blind to the immunocyte assay results performed case adjudication by independently reviewing the CAG imaging. Case adjudication was dichotomic: CSF was present or absent. CSF was defined as angiographically normal or near-normal (< 50% stenosis) coronary arteries with TIMI ≤ 2 flow (i.e., requiring three or more beats to opacify the distal vasculature) in at least one major epicardial artery in the absence of specific ischemic provocative maneuvers (e.g., coronary angioplasty). Non-CSF was defined as coronary arteries (< 50% stenosis) with TIMI 3 flow. In cases of disagreement, the cases were adjudicated after discussion.

Parameter measurements associated with coronary microvascular dysfunction

Microcirculation resistance was measured according to the Quantitative Flow Fraction Measurement System of Bodong Medical Imaging Technology (Shanghai) Co., Ltd. Two cardiologists who were qualified to take measurements at our hospital made all measurements together. Two angiographic projections, at least 25° apart, were transferred to the system for the measurements. A contrast flow model, incorporating contrast flow velocity based on the frame count method, was used to compute the quantitative flow ratio (QFR).

Study size calculation

Consistent with the data from previous studies, the results of our preliminary experiment showed that the incidence of CSF was 10%. The research hypothesis was that CD45RO+CD8+ T cells would have good diagnostic value, with an area under the curve of 0.75, an α of 0.05, and power (1-β err prob) of 0.90. The ratio of non-CSF (negative) to CSF (positive) was 9:1, and a bilateral superiority test was performed. PASS11 calculated the sample size. The final results indicated that at least 150 patients were needed for the study.

Statistical analysis

The baseline characteristics of the participants are indicated as mean [standard deviation (SD)] (Gaussian distribution) or median (Q1-Q3) (skewed distribution) for the continuous variables, and as percentages for categorical variables. We used the chi-square test (categorical variables), Student’s T-test (normal distribution), or Mann-Whitney U-test (skewed distribution) to test for differences among groups. We used univariate and multivariable binary logistic regression models to test the association between CD45RO+CD8+ T cells and CSF with three distinct models. Model 1 was non-adjusted (no covariates adjusted). Model 2 was a minimally adjusted model (only sociodemographic variables adjusted). Model 3 was fully adjusted; Table 1 presents the covariates with marginal associations with CSF (p ≤ 0.1). Subgroup analyses were performed using a stratified binary logistic regression model. A continuous variable was first converted to a categorical variable according to the clinical cut-off point, and an interaction test was then performed. A likelihood ratio test followed the effect modification tests for the subgroup indicators. To test the robustness of our results, a sensitivity analysis was performed. We converted CD45RO+CD8+ T cells into a categorical variable by quartile to calculate the p for the trend to verify the results of CD45RO+CD8+ T cells as the continuous variable and examined the possibility of non-linearity.

Table 1. Patient characteristics.

	Non CSF (n = 124)	CSF (n = 36)	p value
Age, years, mean (SD)	68.11 (9.45)	66.97 (9.96)	0.53
Male, n (%)	58 (46.77)	24 (66.67)	0.04
Body mass index, kg/m2, mean (SD)	24.69 (4.00)	24.77 (3.00)	0.91
Current tobacco user, n (%)	20 (16.13)	9 (25.00)	0.22
History of hypertension, n (%)	95 (76.61)	23 (63.89)	0.13
History of ischemic stroke, n (%)	23 (18.55%)	3 (8.33%)	0.14
History of atrial fibrillation, n (%)	5 (4.03%)	4 (11.11%)	0.11
History of diabetes, n (%)	23 (18.55%)	8 (22.22%)	0.62
History of chronic obstructive pulmonary disease, n (%)	11 (8.87%)	3 (8.33%)	0.92
History of pulmonary nodule, n (%)	53 (42.74%)	18 (50.00%)	0.44
Systolic blood pressure, mmHg, mean (SD)	138.92 (19.43)	136.03 (17.47)	0.42
Diastolic blood pressure, mmHg, mean (SD)	76.87 (11.40)	76.53 (10.93)	0.87
Hemoglobin, g/L, mean (SD)	131.91 (15.69)	140.25 (18.60)	0.01
White blood cell count, ×109/L, mean (SD)	6.04 (2.23)	6.02 (1.60)	0.97
Neutrophilic granulocyte percentage, %, mean (SD)	58.34 (12.21)	55.18 (14.64)	0.19
Monocytes percentage, %, median (Q1-Q3)	6.55 (5.50-7.90)	7.00 (6.10-7.90)	0.46
Lymphocyte percentage, %, mean (SD)	30.41 (9.76)	31.19 (8.34)	0.66
Platelet, ×109/L, mean (SD)	197.73 (70.43)	190.19 (46.04)	0.55
C-reactive protein, μg/ml, median (Q1-Q3)	1.15 (0.50-2.58)	1.45 (0.67-2.80)	0.46
B-type natriuretic peptide, median (Q1-Q3)	47.50 (29.00-85.25)	55.50 (24.50-117.50)	0.52
Triglyceride, mmol/L, median (Q1-Q3)	1.30 (0.90-1.83)	1.35 (1.10-1.90)	0.28
Total cholesterol, mmol/L, mean (SD)	4.16 (1.14)	4.22 (1.16)	0.78
High-density lipoprotein cholesterol, mmol/L, mean (SD)	1.19 (0.31)	1.14 (0.30)	0.38
Low-density lipoprotein cholesterol, mmol/L, mean (SD)	2.64 (0.86)	2.71 (0.84)	0.67
Serum creatinine, μmol/L, mean (SD)	72.51 (31.76)	70.98 (11.20)	0.78
Uric acid, μmol/L, mean (SD)	341.43 (90.20)	376.91 (81.93)	0.04
Glycosylated hemoglobin, %, mean (SD)	6.47 (0.94)	6.57 (1.13)	0.57
Aspirin use, n (%)	39 (31.45%)	5 (13.89%)	0.04
Platelet ADP receptor antagonist use, n (%)	19 (15.32%)	4 (11.11%)	0.53
Statin use, n (%)	47 (37.90%)	7 (19.44%)	0.04
Angiotensin-converting enzyme inhibitors/angiotensin receptor blockers use, n (%)	52 (41.94%)	9 (25.00%)	0.07
Diuretic use, n (%)	9 (7.26%)	4 (11.11%)	0.46
β-blocker use, n (%)	30 (24.19%)	5 (13.89%)	0.19
Calcium channel blocker use, n (%)	40 (32.26%)	11 (30.56%)	0.85
Oral hypoglycemic drugs use, n (%)	32 (25.81%)	7 (19.44%)	0.43
Insulin use, n (%)	6 (4.84%)	1 (2.78%)	0.60
Proportion CD45RO+CD8+ T cells, %, mean (SD)	45.26 (16.45)	56.18 (13.93)	< 0.001
Proportion CD45RO+CD8+ T cells group*, n(%)			0.03
< 22.5%	12 (9.7%)0	0 (0%)
22.5-43.1%	43 (34.7%)	8 (22.2%)
> 43.1%	69 (55.6%)	28 (77.8%)

CSF, coronary slow flow; Q, quartile; SD, standard deviation.

* Grouped based on clinical cut-off value.

Further, to address the non-linearity of continuous variables in Model 3 and CSF risk, a weighted generalized additive model (GAM) and smooth curve fitting (using the penalized spline method) were conducted. All data were analyzed using the statistical software packages R (http://www.R-project.org, The R Foundation) and Empower Stats (http://www.empowerstats.com, X&Y Solutions, Inc, Boston, MA). All statistical tests were 2-sided, and a p value < 0.05 was considered statistically significant.

RESULTS

Clinical characteristics of the study participants

Overall, of the initial 761 patients who presented with angina for planned CAG, 601 were excluded because they were unwilling or unable to participate, had CAG with epicardial coronary artery stenosis ≥ 50%, had inflammatory or infectious conditions, or had been treated with anti-inflammatory drugs. Thus, a total of 160 patients [mean (SD) age, 67.86 (9.55) years; 51.25% male] were included in this study (see Figure 1). The baseline features of the patients are summarized in Table 1. The incidence of angiographic CSF, defined as an angiographic coronary TIMI flow ≤ 2, was 22.5%. CSF patients had higher levels of hemoglobin and uric acid than non-CSF patients. Compared to the non-CSF patients, very few CSF patients had a recent history of aspirin, statin, or angiotensin-converting enzyme inhibitor (ACEI)/angiotensin receptor blocker (ARB) use. There were more CD45RO+CD8+ T cells in the CSF patients than in those with TIMI 3 flow between groups. Accordingly, the patients with angiographic CSF had higher levels of CD45RO+ CD8+ T cells than those without CSF [56.18 (13.93) vs. 45.26 (16.45); p < 0.001; see Table 1, and Figure 2A).

Figure 1.

Study populations. CAG, coronary angiography; CSF, coronary slow flow; TIMI, thrombolysis in myocardial infarction.

Figure 2.

CD45RO+CD8+ T cells and coronary slow flow. (A) Box-plot of the CD45RO+CD8+ T cells distribution in patients with or without coronary slow flow. (B) Coronary slow flow risk increased across quartiles of CD45RO+CD8+ T cells. CSF, coronary slow flow.

Associations between CD45RO+CD8+ T cells and CSF

The results of univariate analysis (Model 1) showed that for each additional unit of plasma level of CD45RO+ CD8+ T cells, the risk of incident CSF was 1.40 [95% confidence interval (CI): 1.18-1.74]. A minimally adjusted model and fully adjusted model were also used. Compared to the unadjusted model [odds ratio (OR): 1.04, 95% CI: 1.02-1.07], the CD45RO+CD8+ T cells in the minimally adjusted model (adjusted for sex and age only) was still positively correlated with the risk of CSF, and there was no significant change in the effect size (OR: 1.05, 95% CI: 1.02-1.07). In the fully adjusted model (adjusted for age, sex, uric acid, hemoglobin, atrial fibrillation, aspirin use, statin use, and ACEI/ARB use), the effect size was 1.05 (95% CI: 1.02-1.09) (see Table 2). In all models, the ORs for CSF risk progressively increased across CD45RO+CD8+ T cell quartiles (see Table 2). Accordingly, the angiographic CSF incidence significantly increased with a higher quartile of CD45RO+CD8+ T cells (p value for trend = 0.014; see Figure 2B). In the final fully adjusted multivariable model, for CD45RO+CD8+ T cells exceeding the clinical normal range (> 43.1%), the OR (95% CI) for CSF was 2.87 (1.06-7.79), and for a per SD increase in CD45RO+CD8+ T cells, the OR (95% CI) for CSF was 2.41 (1.46-3.97). To address non-linearity, we also used a GAM to adjust the continuous variables. Despite these transformations (in which continuous variables were smoothed), the results did not change significantly (see Table 2). The smooth curve and results of the GAM showed that the relationship between CD45RO+ CD8+ T cells and CSF was approximately linear after adjusting for age, male sex, hemoglobin, uric acid, atrial fibrillation, aspirin use, statin use, and ACEI/ARB use (see Figure 3).

Table 2. Association of proportion CD45RO+CD8+ T cells with incident coronary slow flow events.

	Model 1	Model 2	Model 3	GAM
Proportion CD45RO+CD8+ T cells, %	1.04 (1.02-1.07)	1.05 (1.02-1.07)	1.05 (1.02-1.09)	1.07 (1.03-1.12)
Proportion CD45RO+CD8+ T cells, %
Q1	Ref	Ref	Ref	Ref
Q2	1.69 (0.44-6.52)	1.88 (0.48-7.40)	2.03 (0.45-9.22)	4.01 (0.73-22.19)
Q3	3.19 (0.92-11.05)	3.20 (0.91-11.24)	5.48 (1.32-22.72)	8.24 (1.46-46.45)
Q4*	5.40 (1.60-18.20)	5.90 (1.70-20.49)	8.37 (1.98-35.44)	14.54 (2.57-82.39)
Proportion CD45RO+CD8+ T cells group†
22.5-43.1%	Ref	Ref	Ref	Ref
< 22.5%	_#	_#	_#	_#
> 43.1%	2.18 (0.91-5.22)	2.19 (0.90-5.30)	2.87 (1.06-7.79)	3.44 (0.96-12.38)
Proportion CD45RO+CD8+ T cells, % (per SD increase)	2.04 (1.33-3.13)	2.12 (1.36-3.30)	2.41 (1.46-3.97)	3.18 (1.67-6.07)

Data are odds ratio (95% CI). Mean (SD) proportion CD45RO+CD8+ T cells was 26.25 (7.50) in the Q1, 41.70 (4.47) in the Q2, 54.83 (3.61) in the Q3, and 68.10 (5.97) in the Q4.

* p for trend, p < 0.05. ^# The model failed because of the small sample size. ^† Grouped based on clinical cut-off value.

Model 1: The non-adjusted model with no covariates adjusted. Model 2: The minimally adjusted model with only age and sex-adjusted. Model 3: The fully adjusted model with hemoglobin, uric acid, atrial fibrillation, age, sex, aspirin, statin, and ACEI/ARB use. GAM: all continuous variables in the covariates were adjusted as smooth.

ACEI, angiotensin-converting enzyme inhibitor; ARB, angiotensin receptor blocker; CI, confidence interval; GAM, generalized additive model; Q, quartile; SD, standard deviation.

Figure 3.

The relationship between CD45RO+CD8+ T cells and coronary slow flow (CSF). The solid red and dotted blue lines indicate the associated probability and its corresponding 95% confidence interval. A positive relationship between CD45RO+CD8+ T cells and CSF risk was detected after adjusting for age, the male sex, hemoglobin, uric acid, atrial fibrillation, aspirin use, statin use, and angiotensin-converting enzyme inhibitor (ACEI)/angiotensin receptor blocker (ARB) use.

Subgroup analysis

We further explored the role of other covariables on the association between CD45RO+CD8+ T cells and CSF risk in subgroup analysis. As Figure 4 shows, the associations between elevated CD45RO+CD8+ T cells and CSF risk were consistent across all covariate subgroups (all p values for interaction > 0.05).

Figure 4.

Subgroup analysis and interaction analysis results for the relationship between CD45RO+CD8+ T cells and CSF. ACEI, angiotensin-converting enzyme inhibitors; ARB, angiotensin receptor blockers; CI, confidence interval; CSF, coronary slow flow; OR, odds ratio.

Increased CD45RO+CD8+ T cells in CSF and association with coronary microvascular resistance

The CAG results of 86 patients met the Quantitative Flow Fraction Measurement System’s requirements among all analyzed patients. Finally, parameters of the left anterior descending (LAD) were analyzed. As Table 3 shows, the CSF patients had slower flow velocity and more frame counts than the non-CSF patients. Accordingly, microvascular resistance was significantly higher in the patients with CSF than in those without CSF. Contrast QFR, which reflects whether the epicardial coronary artery can cause ischemia, did not differ between the two groups. A positive linear relationship was observed between CD45RO+CD8+ T cells and coronary microvascular resistance (see Figure 5).

Table 3. Parameter measurement associated with coronary microvascular dysfunction.

	Non-coronary slow flow (n = 64)	Coronary slow flow (n = 22)	p value
LAD frame count, frame, median (Q1-Q3)	8.00 (5.00-12.00)	20.50 (19.00-26.00)	< 0.001
LAD contrast QFR, mean (SD)	0.89 (0.06)	0.86 (0.07)	0.13
LAD MR, mmHg*s/m, median (Q1-Q3)	71.48 (38.34-103.25)	152.48 (90.51-218.58)	< 0.001
LAD flow velocity (m/s), median (Q1-Q3)	0.12 (0.10-0.18)	0.08 (0.06-0.10)	0.001

LAD, left anterior descending; MR, microvascular resistance; Q, quartile; QFR, quantitative flow ratio; SD, standard deviation.

Figure 5.

The relationship between CD45RO+CD8+ T cells and coronary microvascular resistance. (A) A positive linear relationship between CD45RO+CD8+ T cells and LAD coronary MR was detected. (B) A positive relationship between CD45RO+CD8+ T cells and LAD coronary frame count was detected. (C) A positive relationship between CD45RO+CD8+ T cells and LAD coronary flow velocity was detected. (D) CD45RO+CD8+ T cells was not related to LAD QFR. LAD, left anterior descending; MR, microvascular resistance; QFR, quantitative flow ratio.

DISCUSSION

Our results showed that higher plasma levels of CD45RO+CD8+ T cells were independently associated with CSF as assessed by angiographic parameters. Thus, CD45RO+CD8+ T cells appear to play an important pathophysiological role in CSF. This positive effect was evident in all subgroups considered and after careful adjustments were made. Further, the pre-CAG level of CD45RO+CD8+ T cells was positively associated with microvascular resistance, implying that CD45RO+CD8+ T cells may be involved in CMD.

CMD plays a pivotal role in the pathogenesis of CSF due to endothelial dysfunction, smooth muscle cell dysfunction, and vascular remodeling.¹⁰ Recent studies have found evidence of reduced CFR in CSF patients.¹¹ Similarly, this study also found that the microcirculation resistance parameters of CSF patients were significantly higher than those of non-CSF patients, which further suggests that the CSF phenomenon is a kind of angiographic manifestation of CMD. Indeed, we observed that the CSF patients had slower flow velocity and higher microvascular resistance compared with the non-CSF patients based on Quantitative Flow Fraction Measurement. Reduced plasma levels of nitric oxide have also been found in some studies involving CSF.¹² Notably, activated immune cells have been shown to infiltrate the adventitia and adhere to the endothelium, which likely contributes to CMD development.

Previous research has suggested that CD8+ T cells act as intermediaries in endothelial dysfunction and vascular rarefaction. Increased accumulation of CD8+ T cells has been observed in the kidneys and aortas of hypertensive mice to promote vascular resistance. Mice lacking CD8+ T cells have been shown to develop blunted vascular rarefaction to Ang II in the kidneys and exaggerated microvascular injury measured by examining microvascular remodeling and stiffness. In addition, the adoptive transfer of T cells has been shown to restore vascular dysfunction. Sun et al. recently showed that CD8+ T cells express mineralocorticoid receptors (MRs), and that these T cell receptors promote IFN-γ production by CD8+ T cells.¹³ They further showed that the specific deletion of MRs in T cells caused a dramatic decrease in vascular resistance and vascular damage. Conversely, the overexpression of MRs in T cells exacerbated vascular resistance. Importantly, there is growing evidence that MRs have pro-hypertensive effects in endothelial cells and SMCs, suggesting that they play an underlying role in CMD development.¹⁴

Clinical data has also shown that circulating T cells in humans with hypertension exhibit evidence of activation as indicated by the increased percentage of memory T cells and increased production of IL-17A and IFN-γ.¹⁵ An animal experiment also found expansion of the memory cell marker CD45RO in the aorta, kidneys, and lymph nodes under AngII stimulation. As the major source of IFN-γ and IL-17A in mice with increased vascular resistance, memory T cells infiltrate the periadventitial tissues, releasing inflammatory cytokines that alter vascular function and participate in end-organ damage.¹⁶ The genetic deletion of these cytokines or treatment with their specific antagonists has been shown to blunt vascular resistance and its attendant end-organ dysfunction. Thus, the finding of increased circulating memory cells might have a pathophysiological significance in increasing microvascular resistance, which could also be true of CMD-mediated CSF. Consistent with these findings, the present study showed that coronary artery blood flow velocity slowed with the increase in CD45RO+ CD8+ T cells, and that microvascular resistance increased significantly in the patients without apparent angiographic stenosis. In CSF, CD45RO+CD8+ T cells appear to have a pro-inflammatory effect on endothelial cells, causing a reduction in nitric oxide bioavailability and increasing chemokine and cytokine expressions, such as IFN-γ, which further leads to slow coronary microvascular flow.¹⁷ Additionally, this stimulates inflammatory elements and coagulation aggregation by inducing the expression of adhesion molecules on the cellular surface. Through the vascular endothelium-homing receptor CX3CR1, circulating memory CD8+ T cells homing to the endothelium can be activated by platelet-mediated regulation.¹⁸ Complex interactions between inflammatory elements and coagulation at endothelial surfaces may play a role in CSF. Finally, activated CD8+T cells have been associated with carotid intimal thickness and reduced carotid distensibility, indicating their role in vascular remodeling.^19,20 Taken together, these effects may explain the association found in our study between CD45RO+CD8+ T cell levels and CSF.

Some studies have suggested that statins may attenuate the CSF phenomenon by suppressing inflammation and improving endothelial function.²¹ Our study found that, independently of statin treatment, CD45RO+CD8+ T cell levels were higher in the patients with CSF than in those with normal flow. We also found that uric acid was associated with CSF, which is in keeping with previous studies that have suggested that uric acid is associated with CMD, and that there can be adverse outcomes in patients with no evidence of significant coronary artery disease at the initial angiography.²² Notably, in our study, CD45RO+CD8+ T cells predicted CSF independently of uric acid, suggesting that it may be an essential modulator of CSF.

CONCLUSIONS

In summary, this appears to be the first study to show that the baseline level of CD45RO+CD8+ T cells is an independent indicator of CSF in patients with an absence of coronary stenosis, which provides insights into the interpretation of this phenomenon. Further studies on the mechanisms underlying the association between CD45RO+CD8+ T cells with CMD may be useful to prevent and treat CSF.

FUNDING

This work was supported by the National Natural Science Foundation of China (NSFC, No. 81803759, 82000332, and No. 82000312), Shanghai Sailing Program (No. 20YF1439800), and the Clinical Research Innovation Plan of Shanghai General Hospital (No. CTCCR-2019C06)

DECLARATION OF CONFLICT OF INTEREST

There are no conflicts of interest.