Association between serum interleukin 8 levels and cardiovascular events in patients with chronic kidney disease

ABSTRACT

Introduction

Chronic kidney disease (CKD) is associated with systemic inflammation and elevated pro-inflammatory cytokines. While interleukin (IL)-8 has shown harmful cardiovascular effects in preclinical studies, its role in CKD remains underexplored. The study aimed to (i) determine serum IL-8 concentrations across CKD stages, (ii) identify factors associated with IL-8 concentrations, and (iii) evaluate its association with major adverse cardiovascular events (MACEs) and all-cause mortality.

Methods

The Chronic Kidney Disease–Renal Epidemiology and Information Network (CKD-REIN) prospective cohort includes CKD patients with an estimated glomerular filtration rate (eGFR) below 60 mL/min/1.73 m² not on kidney replacement therapy. Baseline serum IL-8 concentrations were centrally measured. MACE was defined as any cardiovascular death, myocardial infarction, stroke and hospital admission for heart failure. Multivariable linear regression was used to identify factors associated with IL-8 concentrations. Adjusted cause-specific Cox proportional hazard models were used to estimate hazard ratios [hazard ratio (HR) (95% confidence interval)] for the first MACE and for mortality.

Results

Among 2389 included patients (66% men; median age 68 years; mean eGFR 34.8 mL/min/1.73 m²), median serum IL-8 concentration was 12.2 pg/mL. Higher IL-8 levels correlated with more advanced CKD (P < .001), and were independently associated with lower eGFR, diabetes, prior cardiovascular disease, anemia, elevated C-reactive protein, more medications and lower serum albumin. Elevated baseline IL-8 was associated with a greater adjusted hazard of MACEs in women [HR for 1-unit change in log(IL-8): 1.75 (1.26; 2.43)] but not in men [HR 1.16 (0.93; 1.45)]. The adjusted HR for all-cause mortality was 1.70 (1.40; 2.06), with no difference between men and women.

Conclusion

In a large cohort of patients with moderate-to-advanced CKD, higher IL-8 levels were associated with a greater risk of MACEs in women (but not in men) and higher mortality in both sexes. Further research is needed to assess the potential of IL-8 as a cardiovascular risk biomarker, clarify the clinical significance of the sex difference observed here and determine whether targeting IL-8 could reduce cardiovascular risk in CKD.

GRAPHICAL ABSTRACT

GRAPHICAL ABSTRACT.

KEY LEARNING POINTS.

What was known:

• Chronic kidney disease (CKD) is associated with systemic inflammation and elevated levels of pro-inflammatory cytokines.

• Interleukin (IL)-8 has not been extensively investigated in CKD, even though it has potentially harmful effects [particularly in cardiovascular disease (CVD)].

This study adds:

• IL-8 levels increased with CKD severity and were independently associated with a lower estimated glomerular filtration rate, diabetes, a history of CVD, anemia, a higher serum C-reactive protein level and a lower serum albumin level.

• Higher IL-8 levels were associated with a greater risk of major adverse cardiovascular events (MACEs) in women but not in men.

• Higher IL-8 concentrations were associated with higher all-cause mortality in both sexes.

Potential impact:

• Our findings suggest that IL-8 is a potential biomarker of the CVD risk in CKD.

• The observation of a sex difference in the association between IL-8 and the MACE risk calls for further research on the underlying mechanisms.

• Targeting IL-8 might emerge as a novel therapeutic strategy for reducing the cardiovascular risk in patients with CKD.

INTRODUCTION

Chronic kidney disease (CKD) is considered to be a major public health problem worldwide [1]. All patients with CKD should be considered to have an elevated risk of cardiovascular disease (CVD), although the underlying pathophysiological mechanisms are not fully understood [2–5]. Indeed, conventional cardiovascular risk factors do not appear to fully explain the elevated risk of CVD in the context of CKD.

CKD is associated with a persistent, low-grade inflammation and elevated levels of pro-inflammatory cytokines, particularly interleukins (ILs), which play a pivotal role in cardiovascular complications [6]. These cytokines promote endothelial dysfunction, vascular inflammation, monocyte recruitment and lipid oxidation, notably contributing to the development of unstable atherosclerotic plaques and vascular calcification [7]. Consequently, systemic inflammation in CKD represents a major non-traditional risk factor for CVD, significantly increasing the already high cardiovascular morbidity and mortality observed in this population. Among these cytokines, IL-6 has been most extensively studied in the context of CKD and is known to be associated with overall and cardiovascular mortality in both pre-dialysis and dialyzed patients [8–11]. Other ILs (such as IL-8) warrant an in-depth evaluation.

IL-8 is produced by a variety of cell types early in the inflammation response but can persist for days or even weeks. While primarily produced by monocytes and macrophages, it is also secreted by other cell types, notably smooth muscle cells and endothelial cells within atherosclerotic plaques, suggesting its important role in the pathogenesis of atherosclerosis [12]. IL-8 has been poorly studied in patients with CKD, although the results of preclinical studies suggest that it has a harmful effect. In a recent study of the disease mechanisms underlying vascular calcification in CKD, IL-8 secretion by endothelial cells appeared to have a predominant role [13]. IL-8 stimulated the calcification of human aortic smooth muscle cells (induced by phosphate and indoxyl sulphate) in a concentration-dependent manner. In the few small clinical studies that assayed IL-8 in patients with CKD, the serum concentrations appear to be elevated [14]. One study found significantly higher serum IL-8 concentrations in children with CKD (n = 50) than in controls [15]. In non-dialyzed patients with CKD, IL-8 was associated with coronary artery calcification [16]. In addition, in a study of 76 patients on hemodialysis, elevated IL-8 concentrations were reportedly associated with mortality and cardiovascular events [17].

Our starting hypothesis was that a high serum IL-8 concentration is independently associated with the occurrence of major adverse cardiovascular events (MACEs) in patients with CKD. Thus, the objectives of the present study were to (i) describe the serum IL-8 concentration in patients at various CKD stages, (ii) identify factors associated with the serum IL-8 concentration and (iii) determine whether serum IL-8 concentrations are associated with MACEs and all-cause mortality before kidney replacement therapy (KRT) in patients with CKD, after adjustment for the estimated glomerular filtration rate (eGFR) and other risk factors.

MATERIALS AND METHODS

The study’s results are reported in accordance with the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines [18].

Study design and participants

The Chronic Kidney Disease–Renal Epidemiology and Information Network (CKD-REIN) is a prospective cohort of adult patients (over 18 years of age) with a confirmed diagnosis of CKD and an eGFR <60 mL/min/1.73 m² (stage 3–5) and who were not on maintenance dialysis and had not undergone kidney transplantation by the time of study entry. Extensive details of the study protocol have been published elsewhere [19]. Patients were recruited from 40 nephrology outpatient facilities in France. The facilities were nationally representative with respect to geography and legal status (public or private sector). A total of 3033 patients were included during a routine nephrology outpatient appointment between 2013 and 2016 and were followed up by clinical research associates (CRAs) for 5 years or until 31 December 2020. The study protocol was approved by the institutional review board at the French National Institute of Health and Medical Research (INSERM; reference: IRB00003888) and was registered at ClinicalTrials.gov (NCT03381950).

For the purposes of the present analysis, we excluded patients who did not have a serum IL-8 measurement at baseline (n = 475) and patients for whom serum samples were collected >90 days after inclusion in the study (n = 169). Hence, a total of 2389 patients were included in the present analysis.

Study data

The CRAs collected data from patient interviews and medical records at baseline and then annually. Patients’ characteristics, including the history of hypertension, diabetes and CVD, were recorded. Height and weight were measured to calculate body mass index (BMI). A specific electronic case report form [linked to the international Anatomical Therapeutic and Chemical (ATC) thesaurus] was used by the CRAs to record the drugs prescribed to patients in the 3 months prior to the enrolment visit. Standard blood and urine tests (those recommended by the French health authorities for the routine management of CKD) were carried out for all patients in their usual medical laboratory. The GFR was estimated using the Chronic Kidney Disease Epidemiology Collaboration creatinine equation [20]. The urinary albumin-to-creatinine ratio (ACR) was determined by either measurement or estimation using an equation based on proteinuria measurements [21]. We therefore classified patients according to the Kidney Disease: Improving Global Outcomes 2012 guideline stages as follows: A1 (normal or minimal increase), ACR <3 mg/mmol; A2 (moderate increase), ACR 3–30 mg/mmol; A3 (severe increase), ACR >30 mg/mmol [22].

IL-8 measurements

Serum samples were collected at the time of patient’s enrolment, stored at 4°C, and aliquoted within 6 h without further processing. All samples were stored at –80°C in a biological resource center (Biobanque de Picardie, Amiens-Picardie University Hospital, Amiens, France, BRIF number: BB-0033–00 017). The IL-8 concentration was measured using an enzyme-linked immunosorbent assay ELISA (BD OptEIA™ Human IL-8 ELISA Set; BD Biosciences, Franklin Lakes, NJ, USA). As the data on the IL-8 concentration were not normally distributed, the values were log-transformed for statistical analyses as a continuous variable.

Study outcomes

Hospital admissions occurring during study follow-up were identified from medical reports, hospital records and/or patient interview. Deaths were ascertained from death certificates, hospital records, reports by family members and by linkage with the national vital status registry. A physician reviewed and coded all events using the International Classification of Diseases, 10th Revision. Based on these codes, this physician then identified and assessed cardiovascular events according to the Cardiovascular and Stroke Endpoint Definitions for Clinical Trials [23]. A senior cardiologist (N.M.) further adjudicated all cardiovascular deaths.

The present study’s primary endpoint was the occurrence of the fatal or non-fatal MACE, defined as any cardiovascular death, myocardial infarction, stroke and hospital admission for heart failure. The secondary endpoints were the first instance of fatal or nonfatal CVD (whether atheromatous or non-atheromatous), the first instance of fatal or nonfatal atheromatous CVD, the first instance of fatal or nonfatal non-atheromatous CVD and all-cause deaths (Supplementary data, Table S1).

Only events that occurred before KRT were analyzed. KRT events (defined as initiation of maintenance dialysis or pre-emptive kidney transplantation) were identified from medical records, patient interviews and/or by linkage with the national REIN (French Renal Epidemiology and Information Network) registry. KRT events and non-cardiovascular deaths before KRT were considered to be competing events for cardiovascular events and cardiovascular mortality. Patients were censored at the date of the competing event, the end of their 5-year follow-up or at study loss to follow-up.

Statistical analysis

The baseline characteristics were described first for the study population as a whole (n = 2389) and then for serum IL-8 tertile (T) subgroups (T1, ≤9.33 pg/mL; T2, 9.34–15.49 pg/mL; T3, >15.49 pg/mL). Continuous variables were reported as the mean [standard deviation (SD)] or the median [interquartile range (IQR)], depending on the distribution. Categorical variables were reported as the frequency (percentage). Depending on the distribution, we used a chi-squared test or Fisher’s exact test to compare values of categorical variables, and an analysis of variance or the Kruskal–Wallis test to compare values of continuous variables.

Multivariable linear regressions (based on sociodemographic variables, clinical variables and prescription drugs) were used to estimate independent associations with the log-transformed serum IL-8 concentration (expressed as a non-standardized coefficient [95% confidence interval (CI)] and a standardized coefficient). Our choice of the variables included in multivariable models was based on a literature review and a P-value <.2 in a univariable analysis.

We used cause-specific Cox models to test the crude and adjusted associations between the risk of a first MACE and the serum IL-8 concentration at baseline. Cox models were adjusted for a set of confounding factors selected from a directed acyclic graph [24]. This approach enables the selection of an optimal set of adjustment factors, by closing non-causal pathways between the exposure (i.e. the IL-8 concentration) and the outcome (the MACE). The selected covariates were age, sex, smoking status, diabetes, a history of CVD, anemia, eGFR, pulse pressure, BMI, the log-transformed serum C-reactive protein (CRP), albumin and urea concentrations, the log-transformed ACR, and the use of diuretics, statins and antithrombotics. We also tested the association between the baseline serum IL-8 concentration and all-cause mortality, and the risk of an atheromatous or non-atheromatous cardiovascular event. The adjustment factors were the same as those used to assess the risk of MACEs. We looked for interactions between serum IL-8 concentration and other covariates, and we tested the proportional hazards hypothesis by analyzing the Schoenfeld residuals for each model. Using restricted cubic splines in the adjusted Cox models [with four knots placed at the 5th, 35th, 65th and 95th percentiles of log(IL-8) values] [25], we investigated the functional relationship between the serum IL-8 concentration and the risk of MACE, the risk of atheromatous or non-atheromatous cardiovascular event, and all-cause mortality.

After assuming that data were missing at random, we managed missing covariate data with multivariate imputation by chained equations with predictive mean matching in the mice package in R [26, 27]. A total of 25 datasets were created, and the number of iterations was set to 20. All covariates present in the final Cox models were included in the imputation model. Fitted Cox models were generated for each dataset, and pooled regression coefficients were obtained using Rubin’s rules.

All tests were two-tailed, and the threshold for statistical significance was set to P < .05. All statistical analyses were performed with R software (version 4.4.1) [27].

RESULTS

Characteristics of the patients at baseline

Of the 3033 patients enrolled in the CKD-REIN cohort, 2389 were analyzed. The median (IQR) age was 68 (60–76) years, 66% of the patients were men, the mean (SD) eGFR was 34.8 (13.4) mL/min/1.73 m², 52% of the patients had a history of CVD and 41% had diabetes (Table 1). For the study population as a whole, the median (IQR) serum IL-8 concentration was 12.2 (8.0–17.7) pg/mL. Compared with patients in T1 and T2, patients in T3 had a higher prevalence of diabetes and a history of CVD. Furthermore, the patients in T3 had a higher ACR, higher serum CRP and urea levels, lower serum albumin and hemoglobin levels, and more drug prescriptions. The male-to-female ratio was essentially the same in all tertiles.

Table 1:

Baseline characteristics of the study population.

		IL-8 tertiles (pg/mL)
	Total	T1: ≤9.33	T2: 9.34–15.49	T3: >15.49		Imputed data
	(N = 2389)	(N = 797)	(N = 796)	(N = 796)	P-value	(N = 2389) (%)
IL-8 (pg/mL)	12.2 (8.0–17.7)	6.5 (5.0–8.0)	12.2 (10.8–13.7)	21.2 (17.7–28.3)	<.001	0
Age (years)	68 (60–76)	67 (57–75)	68 (60–76)	69 (63–77)	<.001	0
Men (%)	66	67	66	66	.92	0
Caucasian (%)	97	97	97	98	.57	0
Level of studies (%)					<.001	0
<9 years	15	12	16	17
9–11 years	49	45	47	54
≥12 years	36	43	37	29
Hypertension (%)	96	95	97	97	.06	0
Diabetes (%)	41	30	41	51	<.001	0
Dyslipidemia (%)	73	69	75	75	.003	0
History of CVD (%)	52	40	54	60	<.001	0
Gastrointestinal bleeding (%)	4	4	4	5	.40	4
Alcohol abuse (%)	1.0	0.4	1.1	1.5	.07	0
Smoking status (%)					.66	0.5
Non-smoker	40	41	41	39
Past	47	48	46	48
Current	12	11	12	13
Active cancer or history of cancer (%)	21	23	19	23	.03	2
BMI (kg/m²)	28.7 (5.8)	28.0 (5.5)	28.9 (5.8)	29.1 (6.0)	<.001	2
Pulse pressure (mmHg)	63.3 (18.0)	61.7 (17.3)	62.6 (17.8)	65.6 (18.5)	<.001	0.6
eGFR (mL/min/1.73 m²)	34.8 (13.4)	37.8 (13.4)	34.8 (13.7)	31.8 (12.5)	<.001	0
Albumin- or protein-to-creatinine ratio (mg/g)	121 (23–626)	85 (16–543)	113 (21–550)	178 (37–857)	<.001	10
Albumin- or protein-to-creatinine ratio classes (%)					<.001	10
Normal or minimal increase (A1)	28	34	29	22
Moderate increase (A2)	35	33	36	36
Severe increase (A3)	37	33	35	42
CRP (mg/L)	2.5 (1.1–5.9)	1.8 (0.8–3.8)	2.7 (1.2–6.0)	3.6 (1.6–8.4)	<.001	0.7
Sodium (mmol/L)	140 (2.7)	140 (2.6)	140 (2.6)	140 (3.0)	.91	1.1
Potassium (mmol/L)	4.5 (0.5)	4.5 (0.5)	4.5 (0.5)	4.6 (0.5)	.24	0.8
Chloride (mmol/L)	104 (3.8)	104 (3.4)	104 (3.8)	104 (4.2)	.02	7
Bicarbonates (mmol/L)	24.9 (3.4)	25.0 (3.3)	25.0 (3.3)	24.8 (3.7)	.22	12
Calcium (mmol/L)	2.35 (0.13)	2.35 (0.12)	2.35 (0.13)	2.35 (0.14)	.45	4
Phosphates (mmol/L)	1.15 (0.23)	1.13 (0.22)	1.15 (0.21)	1.18 (0.24)	<.001	6
Serum albumin (g/L)	40.6 (4.1)	41.4 (3.8)	40.7 (3.6)	39.5 (4.5)	<.001	0
Urea (mmol/L)	13.3 (10.2–18.0)	12.2 (9.4–16.0)	13.4 (10.1–18.3)	14.7 (11.3–19.7)	<.001	0
Hemoglobin (g/dL)	13.0 (1.7)	13.4 (1.6)	13.1 (1.7)	12.6 (1.7)	<.001	2
Uric acid (µmol/L)	429 (120)	418 (108)	434 (120)	434 (131)	<.001	10
Number of prescription drugs per patient	8 (4)	7 (4)	8 (4)	9 (4)	<.001	0
Proton pump inhibitors (%)	32	24	33	40	<.001	0
Diuretics (%)	54	45	55	61	<.001	0
Beta-blockers (%)	42	35	44	47	<.001	0
Calcium channel blockers (%)	47	41	47	53	<.001	0
Renin–angiotensin inhibitors (%)	76	76	79	74	.09	0
Statins (%)	58	55	59	60	.10	0
Antithrombotics (%)	52	42	52	61	<.001	0
Anticoagulants (%)	14	10	15	18	<.001	0
Corticosteroids (%)	7	7	7	8	.54	0
Immunosuppressants (%)	6	6	5	6	.93	0
Antigout preparations (%)	35	31	36	37	.02	0
Phosphate binders (%)	4	4	4	3	.27	0

Data are presented as the mean (SD) or the median (first-third quartiles), or frequency (%). P-values in bold indicate p-values <0.05.

Concentrations of IL-8 were similar between women and men [median (IQR) 12.1 (8.0–17.7) pg/mL vs 12.3 (7.9–17.7) pg/mL; P = .88; Supplementary data, Table S2]. Men were slightly older and more frequently had a history of CVD, hypertension, diabetes and dyslipidemia. Women had a lower mean eGFR, whereas men had higher median ACR. Regarding medications, men were more often treated with antihypertensive agents, statins, antithrombotics and antigout drugs, whereas proton pump inhibitors were more frequently prescribed to women.

A small but significant negative correlation was found between log-transformed serum IL-8 and the eGFR (r = –0.21; P < .001). The more advanced CKD stages were associated with higher serum IL-8 concentrations (P < .001; Fig. 1).

Figure 1:

Distribution of serum IL-8 concentrations by CKD stage. ND, not on dialysis.

Factors associated with the serum IL-8 concentrations

In linear regression models, log-transformed serum IL-8 levels were independently and positively associated with diabetes, a history of CVD, anemia, serum CRP and the number of prescription drugs, and negatively associated with eGFR and serum albumin (Table 2). Log-transformed CRP, eGFR, albumin and diabetes were the variables that best predicted IL-8 levels. Sex was not associated with the IL-8 level.

Table 2:

Factors associated with the serum IL-8 level.

	Univariable model		Multivariable model
Variables	Unstandardized unadjusted coefficient	P-value	Unstandardized coefficients	P-value	Standardized coefficients
Age (years)	0.01 (0.01; 0.01)	<.001	0.001 (–0.002; 0.003)	.69	0.001
Men (ref: women)	–0.01 (–0.07; 0.05)	.79	–0.001 (–0.07; 0.07)	.98	0.0005
Level of studies
<9 years	Ref		Ref
9–11 years	–0.03 (–0.11; 0.05)	.42	0.00005 (–0.09; 0.09)	.99	0.00 004
≥12 years	–0.21 (–0.29; –0.13)	<.001	–0.08 (–0.18; 0.02)	.10	0.06
Diabetes	0.25 (0.20; 0.31)	<.001	0.14 (0.06; 0.21)	<.001	0.10
Dyslipidemia	0.10 (0.03; 0.16)	.003	–0.06 (–0.14; 0.01)	.11	0.04
History of CVD	0.22 (0.17; 0.28)	<.001	0.11 (0.04; 0.18)	.002	0.08
Obesity	0.12 (0.06; 0.18)	<.001	–0.06 (–0.13; 0.01)	.11	0.04
Anemia	0.28 (0.23; 0.34)	<.001	0.10 (0.03; 0.17)	.005	0.07
Systolic blood pressure (mmHg)	0.002 (0.001; 0.003)	.006	0.0002 (–0.001; 0.002)	.77	0.01
eGFR (mL/min/1.73 m²)	–0.01 (–0.01; –0.01)	<.001	–0.01 (–0.01; –0.004)	<.001	0.13
Log(ACR) (mg/g)	0.04 (0.03; 0.06)	<.001	0.01 (–0.01; 0.02)	.56	0.02
Log(CRP) (mg/L)	0.16 (0.13; 0.18)	<.001	0.10 (0.07; 0.13)	<.001	0.17
Chloride (mmol/L)	–0.01 (–0.02; 0.0001)	.052	–0.01 (–0.02; 0.002)	.10	0.04
Bicarbonates (mmol/L)	–0.01 (–0.02; 0.00004)	.051	–0.01 (–0.02; 0.002)	.14	0.04
Phosphates (mmol/L)	0.34 (0.22; 0.47)	<.001	–0.04 (–0.19; 0.12)	.66	0.01
Serum albumin (g/L)	–0.04 (–0.04; –0.03)	<.001	–0.02 (–0.03; –0.01)	<.001	0.11
Urea (mmol/L)	0.02 (0.02; 0.02)	<.001	0.001 (–0.01; 0.01)	.79	0.01
Uric acid (µmol/L)	0.001 (0.0003; 0.001)	<.001	0.0002 (–0.0001; 0.0005)	.14	0.03
Number of prescription drugs per patient	0.04 (0.04; 0.05)	<.001	0.01 (0.001; 0.02)	.02	0.07

Data are presented as the coefficient (95% CI). P-values in bold indicate p-values <0.05.

Major adverse cardiovascular events

In the adjusted Cox model, there was a significant interaction between log-transformed serum IL-8 and sex. Therefore, a subgroup analysis by sex was performed, revealing differential effects of IL-8 according to sex, despite no statistically significant differences in IL-8 concentrations. In women, the adjusted hazard ratio (HR) (95% CI) for MACEs associated with the baseline serum log-transformed serum IL-8 concentration was 1.75 (1.26; 2.43). A restricted cubic spline analysis showed that the relationship between the serum IL-8 concentration and the hazard of MACEs was linear [P_{(non-linearity)} = .09], with a greater hazard at high concentrations (Fig. 2). In men, the adjusted HR (95% CI) for MACEs associated with the baseline log-transformed serum IL-8 concentration was 1.16 (0.93; 1.45). A restricted cubic spline analysis confirmed this (non-significant) trend (Fig. 3). Sensitivity analyses using IL-8 tertiles yielded the same results (Supplementary data, Fig. S1).

Figure 2:

Adjusted HR for the occurrence of the first MACE, as a function of the baseline serum IL-8 concentration, using restricted cubic spline terms, in women. Note: HRs are adjusted for age, smoking status, diabetes, history of CVD, anemia, eGFR, pulse pressure, BMI, log-transformed ACR, log-transformed serum CRP, albumin, urea, and the use of diuretics, statins and antithrombotics. The continuous line represents predictions with restricted cubic splines in Cox models (dashed lines represent the 95% CIs). The red line represents predictions made without applying a spline. Ticks on the x-axis represent the distribution of the baseline serum IL-8 level.

Figure 3:

Adjusted HR for the occurrence of the first MACE, as a function of the baseline serum IL-8 concentration, using restricted cubic spline terms, in men. Note: HRs are adjusted for age, smoking status, diabetes, history of CVD, anemia, eGFR, pulse pressure, BMI, log-transformed ACR, log-transformed serum CRP, albumin, urea, and the use of diuretics, statins and antithrombotics. The continuous line represents predictions with restricted cubic splines in Cox models (dashed lines represent the 95% CIs). The red line represents predictions made without applying a spline. Ticks on the x-axis represent the distribution of the baseline serum IL-8 level.

Atheromatous or non-atheromatous cardiovascular events

In the adjusted Cox model, there was a significant interaction between log-transformed serum IL-8 and sex. Again, a subgroup analysis by sex was performed. In women, the adjusted HR (95% CI) for atheromatous or non-atheromatous cardiovascular events associated with the baseline log-transformed serum IL-8 concentration was 1.37 (1.04; 1.80). A restricted cubic spline analysis showed that the relationship between the serum IL-8 level and the hazard of an atheromatous or non-atheromatous cardiovascular event was linear in women [P_{(non-linearity)} = .51], with a greater risk at high concentrations (Supplementary data, Fig. S2). In men, the adjusted HR (95% CI) for atheromatous or non-atheromatous cardiovascular events associated with the baseline log-transformed serum IL-8 concentration was 1.03 (0.87; 1.21). The tertile analysis and the restricted cubic spline analysis confirmed this (non-significant) trend (Supplementary data, Figs S1 and S3).

Subtype analyses revealed a consistent trend for non-atheromatous cardiovascular events (Supplementary data, Figs S1, S4 and S5). In contrast, the effect of the interaction between log-transformed serum IL-8 and sex on atheromatous cardiovascular events was not significant, and the association between log-transformed serum IL-8 and the occurrence of atheromatous cardiovascular events was not significant (Supplementary data, Figs S6 and S7).

All-cause mortality

In the adjusted Cox model for all-cause mortality, the interaction between log-transformed serum IL-8 and sex was not significant (P = .82). After multiple adjustments, the HR (95% CI) for death before KRT as a function of the log-transformed serum IL-8 concentration was 1.70 (1.40; 2.06). The restricted cubic spline analysis showed that the relationship between serum IL-8 levels and the hazard of death was linear [P_{(non-linearity)} = .76], with a higher hazard at elevated concentrations (Fig. 4). Sensitivity analyses using IL-8 tertiles yielded the same results (Supplementary data, Fig. S7).

Figure 4:

Adjusted HR for overall mortality, as a function of the baseline serum IL-8 concentration, using restricted cubic spline terms. Note: HRs are adjusted for age, smoking status, diabetes, history of CVD, anemia, eGFR, pulse pressure, BMI, log-transformed ACR, log-transformed serum CRP, albumin, urea, and the use of diuretics, statins and antithrombotics. The continuous line represents predictions with restricted cubic splines in Cox models (dashed lines represent the 95% CIs). The red line represents predictions made without applying a spline. Ticks on the x-axis represent the distribution of the baseline serum IL-8 level.

DISCUSSION

In our analysis of a large, prospective cohort of non-dialyzed patients with CKD, the serum IL-8 concentration increased with CKD severity. Our results highlighted an elevated hazard of MACEs with higher serum IL-8 concentrations in women but not in men. Although serum IL-8 levels were similar in men and women, there was a clear sex difference in the association between IL-8 and cardiovascular outcomes. Furthermore, IL-8 concentrations were associated with all-cause mortality in the cohort as a whole.

CKD is associated with an increase in inflammation, which can progress to a chronic state. As a result, levels of pro-inflammatory cytokines (and particularly ILs) tend to increase and potentially contribute to the higher cardiovascular risk observed in this population. IL-6 has been extensively studied in the context of CKD and is known to be associated with overall and cardiovascular mortality in both pre-dialysis and dialyzed patients [8–11]. Moreover, the IL-6 inhibitors ziltivekimab and clazakizumab are being trialed in the clinic, with a view to reducing persistent inflammation and the associated cardiovascular risk in patients with CKD [28, 29], highlighting that, in the near future, these biomarkers may not only serve prognostic purposes but also guide therapeutic decisions. However, the roles of other ILs (such as IL-8) warrants further investigation. IL-8 has been poorly studied in CKD, and only few small clinical studies have evaluated serum IL-8 levels in patients with CKD and the potential association with cardiovascular outcomes or mortality.

Firstly, we conducted an in-depth investigation of the factors associated with serum IL-8. Serum IL-8 concentrations increased with the CKD stage; this is in line with the results of previous studies of both non-dialyzed and dialyzed patients with CKD [14, 15, 30, 31]. Our present results confirmed this trend in a large cohort of over 2000 non-dialyzed patients with CKD. Furthermore, a multivariable linear regression analysis revealed a negative association between the serum IL-8 level and eGFR. Serum IL-8 was positively associated with diabetes, a history of CVD, anemia, serum CRP and the number of prescription drugs taken, and negatively associated with the serum albumin level. All these parameters are typically associated with inflammation [32–35]. We did not observe a significant difference in the IL-8 level between men and women. To date, few data on sex-specific variations in IL-8 levels in the context of CKD and/or CVD have been reported, and most studies have not addressed this aspect.

Secondly, we investigated the association between serum IL-8 and CVD in a CKD population. The link between serum IL-8 and CVD has been explored in various conditions other than CKD. In patients with coronary artery disease and acute myocardial infarction, IL-8 independently predicted cardiovascular events, including MACEs [36, 37]. In the Controlled Rosuvastatin Multinational Trial in Heart Failure (CORONA) study of 1464 patients with chronic heart failure, baseline serum IL-8 was independently associated with a composite outcome (cardiovascular mortality, non-fatal myocardial infarction or non-fatal stroke) and all-cause mortality [38]. In a study of a large, community-based cohort of elderly individuals, however, IL-8 was associated with all-cause mortality but not with cardiovascular events [39]. Given the high prevalence of CVD in patients with CKD, investigating the association between IL-8 and CVD appears to be particularly relevant. Indeed, the results of several preclinical studies have suggested a potential role for IL-8 in the context of uremia. A study of the mechanisms of vascular calcification in CKD found that the IL-8 secreted by endothelial cells exposed to uremic toxins (such as phosphate and indoxyl sulphate) promoted vascular calcification [13]. Another study demonstrated that the in vitro exposure of primary human aortic valvular interstitial cells to IL-8 enhanced phosphate-induced calcification [40]. Furthermore, in a rat model of CKD-associated calcific aortic valve disease, treatment with an antagonist of CXCR2 (one of the main IL-8 receptors, along with CXCR1) limited disease progression and thus highlighted the potential therapeutic value of IL-8 inhibition [40].

Our present results highlighted an elevated risk of MACEs in women (but not men) with higher serum IL-8 concentrations and an elevated risk of all-cause mortality in both sexes. Only a few small studies have investigated the association between IL-8 levels and CVD or mortality in patients with CKD, and none specifically evaluated sex interactions. In a study of 76 dialysis patients followed up for 18 months, the baseline IL-8 concentration was found to be an independent predictor of both all-cause mortality and cardiovascular events. However, the small number of observed events (15 deaths and 17 cardiovascular events) limited the adjustment for key confounders in the statistical analysis [17]. Furthermore, in an exploratory, cross-sectional study of 148 patients with stage 3–4 CKD, the plasma IL-8 level was associated with coronary artery calcification [16]. It is noteworthy that none of these studies investigated or reported sex differences in outcomes. In patients with coronary heart disease and unknown CKD status, cytokine concentrations were associated with the risk of MACE in women but not in men. Notably, plasma cytokine concentrations were significantly higher in women than in men in this cohort; however, IL-8 concentrations were not measured [41]. In the CKD-REIN cohort, a study by Faucon et al. investigating sex differences in CVD incidence showed that women had a lower risk of atheromatous events than men, while rates of non-atheromatous events were similar [42]. In our analyses, serum IL-8 was significantly and linearly associated with non-atheromatous events in women, with no significant association with atheromatous events. While addressing a different question—the prognostic value of IL-8 levels for cardiovascular outcomes stratified by sex—our findings complement rather than contradict those of Faucon et al., and suggest that IL-8 may contribute to explaining part of the residual cardiovascular risk in women, particularly for non-atheromatous events. These findings highlight IL-8 as a potential contributor to the pathophysiology of cardiovascular risk in women with CKD.

As mentioned above, and despite the absence of a significant sex difference in the serum IL-8 level, the association between IL-8 and cardiovascular outcomes differed in men vs women. To the best of our knowledge, this difference has not been reported previously. Few of the published studies evaluating the association between the cardiovascular risk or mortality and the IL-8 concentration tested for an interaction with sex. In the EPIC-Norfolk study, a sex interaction with IL-8 was evaluated in a model predicting coronary artery disease but was found to be non-significant [43]. It is also noteworthy that the great majority of animal studies (especially in cardiovascular research) are conducted on males [44], which might limit our understanding of sex-specific inflammatory responses. Further research is needed to clarify how sex influences the relationship between inflammation and the cardiovascular system. In our analysis, sex modified the effect of IL-8 on cardiovascular outcomes. This suggests that the same increment in IL-8 levels confers a differential risk in men vs women and might ultimately lead to sex-specific disease manifestations. The clinical implications of sex-based differences in blood biomarkers remain incompletely understood. However, this knowledge might have important applications, including the establishment of sex-specific biomarker cut-offs and the development of targeted therapies to enhance treatment effectiveness in men and in women.

In the present study, IL-8 was associated with a higher risk of all-cause mortality. This risk was similar in men and women. However, evidence from a prospective cohort of approximately 1000 elderly individuals suggested that the association was sex-specific; elevated IL-8 levels independently predicting a higher mortality risk in women but not in men [45]. The baseline IL-8 concentrations in that cohort were slightly but significantly higher in women than in men [median IL-8 (IQR) (pg/mL): 6.8 (6.4; 7.2) and 6.3 (6.0; 6.5), respectively; P = .03]. In contrast, we did not observe that type of difference; the median (IQR) IL-8 level was similar in women and men [12.1 (8.0; 17.7) and 12.3 (7.9; 17.7) pg/mL, respectively; P = .88].

With the increasing development of medications intended to reduce inflammation in CKD, the inhibition of pro-inflammatory ILs is now of major interest. However, translating these findings into clinical practice requires additional steps, including standardization of biomarker assays and evidence that acting upon these markers improves patient outcomes. Despite growing awareness of the need for adequate female representation in clinical trials, pharmacological research has historically neglected sex differences. Our results highlight the potential importance of considering sex-specific effects in the clinical evaluation of IL-8-targeting therapeutics. IL-8 receptor antagonists (such as the CXCR1/2 inhibitor reparixin) are currently being investigated for the treatment of various conditions, including inflammatory lung diseases and cancer [46, 47]. However, no IL-8-targeting therapy is currently in clinical development in CKD or CVD. Given IL-8’s role in inflammation, the use of these therapeutics in patients with CKD might mitigate chronic inflammation and reduce the cardiovascular risk. In view of our results, a potential sex-specific association of IL-8 with cardiovascular outcomes should be carefully considered in trials of IL-8 receptor antagonists.

The present study had several strengths. Firstly, to the best of our knowledge, this study was the first to have investigated the association between serum IL-8 concentrations and the occurrence of cardiovascular events and mortality in a large, nationally representative, multicenter, prospective cohort of patients with CKD not undergoing KRT. This design enabled comprehensive adjustments for potential confounding factors and exploration of the interaction between sex and IL-8. Secondly, IL-8 concentrations were measured in the same central laboratory, which probably minimized inter-assay variability and increased reliability. Lastly, all cardiovascular events were rigorously assessed using standardized definitions, which enhanced the accuracy and consistency of outcome classification.

Our study also had some limitations. Firstly, IL-8 concentrations were measured only once at baseline, which prevented us from examining potential associations between clinical outcomes and changes over time in IL-8 levels. Secondly, the study’s observational design precluded any causal interpretations. Finally, echocardiographic data and vascular calcification scores were not available in the CKD-REIN cohort, preventing us from examining the association between IL-8 and these parameters.

In conclusion, our analysis of a large cohort of patients with moderate-to-advanced CKD demonstrated that elevated IL-8 concentrations were associated with (i) an elevated risk of MACEs in women but not in men, and (ii) an elevated risk of mortality in both sexes. Further investigation is needed to assess the relevance of IL-8 as a biomarker for cardiovascular risk and the clinical implications of the observed sex differences. Lastly, the potential use of IL-8 antagonists to prevent CVD in patients with CKD warrants further exploration.

FUNDING

CKD-REIN is funded by the French Agence Nationale de la Recherche through the 2010 ‘Cohortes-Investissements d’Avenir’ program (ANR-IA-COH-2012/3731) and by the 2010 national Programme Hospitalier de Recherche Clinique. CKD-REIN is also supported through a public–private partnership with GlaxoSmithKline (GSK) since 2012, Boehringer Ingelheim France since 2022, Novo Nordisk since 2024, Fresenius Medical Care from 2012 to 2024, Vifor France from 2018 to 2023, Sanofi-Genzyme from 2012 to 2015, Baxter and Merck Sharp & Dohme-Chibret (MSD France) from 2012 to 2017, Amgen from 2012 to 2020, Lilly France from 2013 to 2018, Otsuka Pharmaceutical from 2015 to 2020 and AstraZeneca from 2018 to 2021. Inserm Transfert set up and has managed this partnership since 2011. This research was funded by the Fondation du Rein, under the aegis of the Fondation pour la Recherche Médicale (grant reference: FDR202212016813). The funding sources had no roles in study design, conduct, reporting or the decision to submit for publication.

AUTHORS’ CONTRIBUTIONS

Each author contributed important intellectual content during manuscript drafting or revision and agrees to be personally accountable for the individual’s own contributions and to ensure that questions pertaining to the accuracy or integrity of any portion of the work—even one in which the author was not directly involved—are appropriately investigated and resolved, including with documentation in the literature if appropriate.

DATA AVAILABILITY STATEMENT

The data that support the findings of this study are available upon reasonable request by contacting the CKD-REIN study coordination staff at ckdrein@inserm.fr.

CONFLICT OF INTEREST STATEMENT

M.P., S.M.L., S.L., Y.B., S.K. and C.C. have nothing to declare. Z.A.M. reports having received grants for CKD-REIN and other research projects from Amgen, Baxter, Fresenius Medical Care, GlaxoSmithKline, Merck Sharp & Dohme-Chibret, Sanofi-Genzyme, Lilly, Otsuka, AstraZeneca, Vifor and the French government, as well as fees and grants to charities from AstraZeneca, Boehringer Ingelheim and GlaxoSmithKline. N.M. declares honoraria for lectures given for Bristol-Myers Squibb. N.A.P. declares financial support from pharmaceutical companies involved in the CKD-REIN study’s public–private partnership: GlaxoSmithKline (GSK), Boeringher Ingelheim and Novo Nordisk; all the grants were made to Paris Saclay University.