Residual inflammatory risk associated with IL-6 and IL-18 after IL-1β inhibition
Residual inflammatory risk associated with interleukin-18 and interleukin-6 after successful interleukin-1β inhibition with canakinumab: further rationale for the development of targeted anti-cytokine therapies for the treatment of atherothrombosis
Introduction and methods
The Canakinumab Antiinflammatory Thrombosis Outcomes Study (CANTOS) showed that targeting inflammation with interleukin-1β (IL-1β) inhibition with canakinumab lowered CV event rates, without affecting blood pressure and cholesterol . The CANTOS trial thereby established the importance of inflammation in human atherosclerosis. It also provided proof-of-principle of targeting pro-inflammatory cytokine pathways, with the magnitude of inflammation inhibition being related to the magnitude of clinical benefit [2,3]. CANTOS participants treated with high-intensity statins and canakinumab remain, however, at considerable risk for recurrent CV events.
It has been suggested that IL-18 may play a role in the residual inflammatory risk [4-6]. No data is available, however, to date that addresses whether IL-18 levels remain a significant predictor of residual risk after therapy with aggressive IL-1β inhibition. This would be relevant in light of development of NLRP-3 inflammasome inhibitors, which would inhibit active forms of IL-1β as well as IL-18.
This study therefore tested the hypothesis in CANTOS data that plasma IL-18 levels associate with future vascular risk both before and after treatment with canakinumab. The relation of IL-6 with future risk was also assessed, as IL-6 signaling has been linked to plaque progression and rupture [7-9]. Plasma levels of IL-18 and IL-6 were measured at randomization and after three months of therapy in 4848 CANTOS participants. Individuals in this analysis were mostly enrolled in Western Europe, the USA and Canada. Any effect of canakinumab on the median percent change in IL-18 and IL-6 between the two measurements was assessed. Assuming regression to the mean for the change in IL-18 and Il-6 in the placebo group over the 3 months period, analyses for the canakinumab therapy were placebo-subtracted. Relative hazards for major adverse CV events (MACE), consisting of myocardial infarction, stroke, or CV death, and other composite CV endpoints were calculated.
- Effects of IL-1β inhibition on IL-6 concentrations were similar to those previously published for hsCRP, such that canakinumab lowered IL-6 levels in a dose-dependent manner, with a placebo-subtracted median % reductions at 3 months were 24.8%, 36.3% and 43.2% for the 50, 150 and 300 mg doses. Canakinumab had no significant effect on IL-18 levels at 3 months.
- Baseline plasma levels of IL-18 and IL-6 were associated with future CV risk. Each tertile increase in IL-18 was associated with 15% higher risk of MACE (95%CI: 4-28%, P=0.0085), and each tertile increase in IL-6 with 50% higher risk (95%CI: 34-67%), P<0.0001). Other endpoints gave similar results.
- After adjustment for covariates, each tertile increase in IL-6 prior to canakinumab treatment associated with a 42% increase in risk of MACE (95%CI: 27-59%, P<0.0001). IL-18 was no longer significantly associated with MACE after this adjustment.
- After initiation of canakinumab treatment, plasma levels of both IL-18 and IL-6 at 3 months were associated with future CV risk. Each tertile increase in IL-18 levels was associated with 15% higher MACE risk (95%CI: 3-29%, P=0.016), and each tertile increase in IL-6 with 42% higher MACE risk (95%CI: 26-59%, P<0.0001). Similar effects were observed for other endpoints, which were still significant after adjustment for age and gender. IL-6 remained significantly associated with CV risk after further adjustment for covariates (27% higher MACE risk per higher tertile IL-6).
- Individuals in the highest baseline tertiles of both cytokines had the highest risk for future MACE (HR: 2.75, 95%CI: 1.92-3.95, P<0.0001 vs. those in lowest tertiles). After canakinumab therapy, the risk in individuals with IL-18 and IL-6 in the highest tertiles was similarly elevated (HR: 2.67, 95%CI: 1.83-3.89, P<0.0001). The additive effects were not significant in tests of interaction on a multiplicative scale.
These randomized data show that IL-1β inhibition with canakinumab lowers plasma IL-6 but not IL-18 levels, and baseline and on-treatment levels of IL-6 or IL-18 predict risk of recurrent CV events. IL-6 showed a greater association with residual inflammatory risk after IL-1β inhibition than IL-18. These data are informative for the development of therapies targeting cytokines other than IL-1β.