Clinical outcome following acute ischaemic stroke relates to both activation and autoregulatory inhibition of cytokine production

Abstract

Background: As critical mediators of local and systemic inflammatory responses, cytokines are produced in the brain following ischaemic stroke. Some have been detected in the circulation of stroke patients, but their role and source is unclear. Focusing primarily on interleukin(IL)-1-related mechanisms, we serially measured plasma inflammatory markers, and the production of cytokines by whole blood, from 36 patients recruited within 12 h and followed up to 1 year after acute ischaemic stroke (AIS).

Results: Admission plasma IL-1 receptor antagonist (IL-1ra) concentration was elevated, relative to age-, sex-, and atherosclerosis-matched controls. IL-1beta, soluble IL-1 receptor type II, tumour necrosis factor (TNF)-alpha, TNF-RII, IL-10 and leptin concentrations did not significantly differ from controls, but peak soluble TNF receptor type I (sTNF-RI) in the first week correlated strongly with computed tomography infarct volume at 5-7 days, mRS and BI at 3 and 12 months. Neopterin was raised in patients at 5-7 d, relative to controls, and in subjects with significant atherosclerosis. Spontaneous IL-1beta, TNF-alpha and IL-6 gene and protein expression by blood cells was minimal, and induction of these cytokines by lipopolysaccharide (LPS) was significantly lower in patients than in controls during the first week. Minimum LPS-induced cytokine production correlated strongly with mRS and BI, and also with plasma cortisol.

Conclusion: Absence of spontaneous whole blood gene activation or cytokine production suggests that peripheral blood cells are not the source of cytokines measured in plasma after AIS. Increased plasma IL-1ra within 12 h of AIS onset, the relationship between sTNF-RI and stroke severity, and suppressed cytokine induction suggests early activation of endogenous immunosuppressive mechanisms after AIS.

Figure 1

Plasma cytokine kinetics in all patients and patients without infection (in the 6 weeks preceding assessment). Patient values are expressed as multiples (ratios) of their corresponding controls (with 95% CIs) to account for skewed distributions. The 24 h time point includes "next 9 am" data where admission was between 07:00 and 11:00. *p ≤ 0.01 **p ≤ 0.001 (relative to controls).

Figure 2

Lipopolysaccharide-stimulated cytokine production kinetics in all patients and patients without infection (in the 6 weeks preceding assessment). Box plots show medians, 10^th and 90^th percentiles and outliers. The 24 hour timepoint includes "next 9 am" data where admission was between 07:00 and 11:00. *p ≤ 0.01 **p ≤ 0.001 (relative to controls).

Figure 3

Minimum LPS-stimulated cytokine production in first week in survivors and non-survivors to 12 months. P values generated from analysis of log-transformed data (logistic regression for comparison of survivors and non-survivors, paired t-test for comparisons with controls).