Interleukin-6 and YKL-40 predicted recurrent stroke after ischemic stroke or TIA: analysis of 6 inflammation biomarkers in a prospective cohort study

Despite widespread use of existing secondary prevention therapy, substantial residual risk of recurrent ischemic events after stroke persisted, varying from approximately 25–30% at 5 years [1]. Traditional risk factors do not explain all epidemiologic features of ischemic vascular event, and there is an urgent need for unraveling new therapeutic targets to reduce the residual risk [2].

Inflammation has been pinpointed as a key regulatory process that links multiple risk factors for ischemic stroke. Some previous epidemiology studies showed that the markers for systemic inflammation, such as interleukin-6 (IL-6), [3] IL-1 receptor antagonist (IL-1Ra) [4, 5] and high sensitive C-reactive protein (hsCRP), [6] and local vascular inflammation produced in atherosclerotic lesions, such as lipoprotein-associated phospholipase A₂ mass (Lp-PLA₂) and activity (Lp-PLA₂-A) [7] and YKL-40 [8, 9] were associated with first or subsequent stroke or functional outcome. This revolution in the thinking about the pathophysiology of atherosclerosis has begun to provide clinical insight and aid patient management, which has been verified in recent randomized trials showing the cardiovascular benefits with anti-inflammatory treatment targeting IL-1 to IL-6 to hsCRP pathways in the patients with coronary disease [10‐12].

However, questions remained, such as the role of IL-1Ra and YKL-40 in recurrent stroke was unclear and controversy still existed regarding the prognostic role of hsCRP, Lp-PLA₂ and Lp-PLA₂-A in recurrent stroke [13‐16]. Moreover, current guidelines do not yet recommend routine measurement of inflammatory markers for risk prediction after stroke, and there is limited data regarding the contribution of combined atherosclerotic inflammatory markers in recurrence after stroke [16, 17].

The Third China National Stroke Registry (CNSR-III) included patients with acute ischemic stroke and transient ischemic attack (TIA) aiming to identify the biological markers for the prognosis and facilitate early evaluation and identification of patients at high risk [18]. In particular, we sought to investigate the contribution of individual and combined atherosclerotic inflammatory markers in recurrent stroke and functional outcome in this biomarker substudy of the CNSR-III.

Instead of focusing on a single inflammatory molecule, we analyzed multiple atherosclerotic inflammatory biomarkers in the current large-scaled multicenter and prospective study, thus providing a comprehensive picture of systemic and vascular inflammation after stroke. We found that the predictive value of IL-6 and YKL-40 in recurrent stroke, composite vascular event and poor functional outcome among the patients with acute ischemic stroke or TIA was more apparent than that of hsCRP and Lp-PLA2 mass and activity.

Different mechanisms might exist for these associations. As an acute-phase reactant produced by hepatocytes, IL-6 has a variety of functions, including propagating the downstream inflammatory response through initiating the Janus kinase and signal transducer and activator of transcription signaling pathways, which contributes to cerebral damage and exerts compact on neurological function [24], and activation of endothelial cells, increased coagulation, and promotion of lymphocyte proliferation and differentiation, and IL-6 has been implicated in progression of atherosclerosis and plaque instability, which may explain how it impacts the development of subsequent stroke [25, 26]. From the point of view of vascular biology, YKL-40 has been shown to activate both mitogen-activated protein kinase and phosphatidylinositol 3′-kinase, which modulates cell proliferation, survival, migration and adhesion, and reflect local vascular bed atherosclerotic inflammation, which consequently might affect the occurrence of vascular event [27‐29]. Among the five stroke subtypes classified according to TOAST criteria, large-artery atherosclerosis and small-vessel occlusion have several risk factors in common that largely contribute to atherosclerosis. Focal inflammation and atherosclerosis have been suggested to be the main pathophysiology of these two subtypes [30]. In our study, we found that the correlations of IL-6 and YKL-40 with recurrent stroke were more apparent in the patients with large-artery atherosclerosis and small-vessel occlusion, implying the association between IL-6 and YKL-40 and atherosclerosis progression, which possibly lead to stroke recurrence. However, the correlations disappeared for IL-6 after adjusting for confounders. Since sample size reduced after TOAST classification, insufficient sample size might be one cause. Further study was needed to verify our results.

On the other hand, though prior Mendelian randomization studies had found that YKL-40 did not play a causative role in the development of cardiovascular disease, [31, 32] these results did not exclude the possibility that YKL-40 could play a role in disease progression, even recurrence. Some previous population-based studies showed that YKL-40 was associated with first stroke, but not myocardial infarction [8, 33]. Moreover, in a European population with stable coronary artery disease, YKL-40 was shown to be associated with composite vascular events, but not cerebrovascular disease after adjusting for other risk factors [34, 35]. These results might suggest the pathophysiological importance of YKL-40 in susceptible vascular bed, and its role as a biomarker reflecting plaque development primarily elicited by risk factor presence in itself, which was consistent with our findings that comorbidity and presence of risk factors were more common in the patients with elevated levels of YKL-40 and YKL-40 predicted recurrence more apparently in the patients with large-artery atherosclerosis. Our results added evidence regarding the positive role of YKL-40 in recurrence after ischemic stroke. On the other hand, it has been shown that YKL-40 could be induced by the cytokine of IL-1β [36], which might explain the positive though relatively weak correlation between YKL-40 and IL-6 in our study. We therefore applied multivariate stepwise regression analyses to reveal the relative usefulness of each marker. When simultaneously considering all markers, IL-6 and YKL-40 remained predictive for recurrent stroke and poor functional outcome. The addition of IL-6 and YKL-40 modestly increased the ROCs of ESSEN and THRIVE models and led to a highly significant NRIs. However, comparing with IL-6 alone, the improvement of predictive value of combining IL-6 and YKL-40 was mild, suggesting the domain role of IL-6.

It has been noted that the prognosis of recurrent stroke is unfavorable, and in China, stroke has ranked third among the leading causes of death behind malignant tumors and heart disease [37, 38]. We therefore adjusted for stroke recurrence to exclude the possibility that poor functional outcome was resulted in by recurrence. The predictive accuracy of Essen Stroke Risk Score and the Totaled Health Risks in Vascular Events Score was lower than that in the previous study [22, 23], and the improvement by adding inflammatory markers was significantly moderate, which might be mainly due to improvement of secondary prevention management of ischemic cerebrovascular disease, proved by 38% decrease (9.9% vs. 16%) in the rate of stroke recurrence in the current study [22]. We believe our data at least have potential importance for risk stratification of patients as well as the design of future trials when choosing systemic or local inflammation markers or treatment targets.

On the other hand, the residual risk of recurrent stroke and death is substantial despite of early management and administrating secondary prevention therapy, calling for measures from other perspective. Atherosclerotic plaque inflammation has been suggested to be an important contributor to plaque destabilization and thromboembolic events. In our study, the relationship between inflammatory markers and recurrent stroke persisted even after adjusting for currently available secondary preventions, including statin and antiplatelet, antihypertensive and hypoglycemic agents. The CANTOS and LoDoCo2 (Low-Dose Colchicine for Secondary Prevention of Cardiovascular Disease 2) studies have demonstrated the protective effect of anti-inflammatory treatment in the patients with coronary heart diseases [10, 39]. However, little was known for stroke prevention. The ongoing CONVINCE (Colchicine for Prevention of Vascular Inflammation in Non-cardio Embolic Stroke) study evaluating the effect of low-dose of colchicine in reducing the rate of recurrent stroke in patients with stroke would shed some light on it [40]. Moreover, our findings also have implications for patients care. Beyond pharmacologic interventions, lifestyle such as diet, exercise and smoking cessation all lowered vascular inflammation. Therefore, lifestyle modification should be emphasized as well, and measurements of inflammation marker levels could be used to motivate lifestyle choices.

IL-1β induces the production of IL-1Ra, which is an endogenous inhibitor of IL-1β [41]. Since a direct measurement of IL-1β was not applicable due to extremely low circulation levels, IL-1Ra could serve as a detectable surrogate parameter for high IL-1β activity [42]. Our data regarding the positive association between IL-1Ra and poor functional outcome merited careful consideration. As an antagonist of IL-1β, IL-1Ra treatment has been highly neuroprotective experimentally [43], while endogenous levels of IL-1Ra have been elucidated to be upregulated and related to adverse outcome after cardiovascular disease and stroke [4, 5]. The functional role of IL-1Ra in the context of atherogenesis has yet to be fully determined [44]. In this regard, two major assumptions might be postulated. First, elevated IL-Ra levels might be atheroprotective, as an insufficient attempt by the body to counter-regulate the concomitant increases of IL-1β activity [45]. Therefore, IL-1Ra might mainly serve as an indirect indicator for IL-1β. Second, IL-1Ra itself might induce atherosclerosis. A large genetic study has found that IL-1Ra-rising alleles lead to an increased cardiovascular risk [46]. However, the hypothesis that binding of IL-1Ra to the IL-1 receptor triggered harmful downstream signaling still needed to be verified. Moreover, given the result of the CANTOS trial, which lowered rate of recurrent cardiovascular event by using anti-inflammatory therapy targeting interleukin-1β, [10] the latter assumption became less likely, but, on the other hand, cannot be totally excluded.

Our study also had several limitations. First, only one point measurements of markers were available, incapacitating us to evaluate the effect of changes in the levels of these markers over time. Second, we did not collect information about how the traditional risk factors were controlled during follow-up under the secondary prevention treatment. Third, though centralized etiology classification of index event was performed, acquirement of stroke subtypes of recurrent stroke would be informative. Fourth, all participants were Chinese, thus, our findings may not be generalizable to other races and ethnicities.

In the patients with ischemic stroke or TIA, we found independent associations of IL-6 or YKL-40 with recurrent stroke, composite vascular events and poor functional outcome, which were more apparent than that of hsCRP and Lp-PLA2 mass and activity. The addition of IL-6 and YKL-40 improved risk classification of clinical risk algorithms. Our findings provided a comprehensive picture of systemic and vascular inflammation after stroke, helped optimizing risk stratification and might shed some light on choosing inflammation target when designing the clinical trials of stroke in the future.