Abstract
Atherosclerotic cardiovascular disease (ASCVD) is the leading cause of mortality in the Western world. Atherogenesis is an inflammatory process originating from a nascent fatty streak lesion culminating in plaque rupture.1–4 Various biomediators of inflammation, including adhesion molecules, cytokines, chemokines, etc., are known to play a role in atherosclerosis in both animal models and humans. The prototypic marker of inflammation, high-sensitivity C-reactive protein (hsCRP) for example, has previously been shown to independently predict cardiovascular events (CVE).1–4
Both low-density lipoprotein cholesterol (LDL-C)-lowering therapy and antihypertensives have been shown to reduce CVE. An anti-inflammatory effect in addition to reduction of blood pressure, especially with blockade of the angiotensin–aldosterone system and LDL-C lowering, cannot be excluded by the above therapies. The benefit of statin therapy in reducing CVE can to some extent be ascribed to pleiotropic anti-inflammatory effects since the best benefit is accrued by concomitant reduction of both hsCRP and LDL-C, and statins have been shown to be anti-inflammatory in human studies.3–5 A therapy that is anti-inflammatory but lipid-neutral has not been studied however. In a recent prospective study, Ridker and colleagues report on a monoclonal antibody targeting anti-interleukin 1 beta (anti-IL-1 beta) on CVE. 6
The Canakinumab Anti-inflammatory Thrombosis Outcomes Study (CANTOS), a randomized, double-blind, placebo-controlled trial evaluated the effect of canakinumab on the prevention of recurrent vascular events in individuals with existing ASCVD. The study included 10,061 individuals, 25.7% of which were women, 40% were diabetic, and around 80% were hypertensive, with a mean age of 61 years, who had had a previous myocardial infarction, and had a persistent pro-inflammatory state as indicated by levels of hsCRP protein at or greater than 2 mg/L. Importantly, 91% of patients were on statins with a median LDL-C of 2.13 mmol/L and a median hsCRP at baseline of 4.2 mg/L. Crucial exclusion criteria included chronic or recurrent infection, previous cancer except basal cell carcinoma, an immunocompromised state, history or risk for tuberculosis, disease related to HIV, or ongoing use of other anti-inflammatory therapies. The patients were randomly assigned to placebo, and three doses of canakinumab, that is, 50 mg, 150 mg, and 300 mg administered subcutaneously once every 3 months in a ratio of 1.5:1:1:1. The primary efficacy endpoint was a composite of nonfatal myocardial infarction, nonfatal stroke, or cardiovascular death. The secondary endpoints included components of primary endpoints and hospitalization for unstable angina requiring urgent revascularization and death from any cause. Comparisons of changes in levels of hsCRP and IL-6 between placebo and canakinumab groups were assessed up to 48 months and 12 months, respectively. A large majority of participants had undergone previous revascularization procedures and were on antithrombotic, lipid-lowering, and anti-ischemic agents. At the end of 4 years, reductions in hsCRP levels compared with placebo group were 26%, 37%, and 41% in individuals receiving 50 mg, 150 mg, and 300 mg canakinumab (p < 0.001), respectively. Similar reductions in levels of IL-6 were also observed at 1 year, but notably there was no significant change in levels of LDL-C from baseline. At the end of follow up, the incidence rate for primary endpoint was observed to be 4.50, 4.11, 3.86, and 3.90 events per 100 person-years in the placebo, 50 mg, 150 mg, and 300 mg dose canakinumab group, respectively. No significant effect was seen with the 50 mg dose compared with placebo. A significant effect was observed in the 150 mg dose group, however, the hazard ratio (HR) versus placebo was 0.85 (p = 0.02075), which was lower than the threshold p value of 0.02115. Although the HR in the 300 mg dose group was 0.86 (p = 0.03), the threshold was higher than the threshold p value of 0.01058. Comparison of all doses of canakinumab combined with placebo revealed a p value of 0.02 for trend. Similarly, for secondary endpoints, the incidence rates were lower with all the canakinumab doses compared with placebo. In the 150 mg dose group, the HR for secondary endpoint was 0.83 with a p value of p = 0.00525, which was lower than the threshold p value of 0.00529. No significant difference was observed in all-cause mortality between the different dose groups of canakinumab compared with placebo. A significantly higher incidence of leukopenia was observed in patients who received canakinumab and significantly more deaths attributed to infection or sepsis were observed in the canakinumab groups (incidence rates versus placebo of 0.31 versus 0.18 per 100 person years, p = 0.02). In addition, thrombocytopenia was reported more frequently in individuals receiving canakinumab with no increased hemorrhage and there were fewer reports of gout and osteoarthritis in the canakinumab groups compared with placebo. Although there were no significant changes in cancer risk and occurrence, there was a significantly lower cancer-related mortality in participants in combined canakinumab groups than in the placebo group (p = 0.02 for trend across active dose groups).
The most promising finding in this study is that an anti-inflammatory strategy targeting IL-1, resulting in a decrease in hsCRP with no decrease in LDL-C, reduced CVE significantly in patients receiving the 150 mg dose. This was also true for the secondary endpoints with the 150 mg dose. It is unclear why data on IL-6 were not reported at trial end like hsCRP. We can also conclude that patients receiving the 50 mg dose accrued no benefits and it appears that the 300 mg dose resulted in a reduction in CVE based on the trend analyses. Hence we can argue that this study fulfills Koch’s postulates for ASCVD since inflammation, including the pro-inflammatory cytokine, IL-1 beta, has been shown in animal models to contribute to atherosclerosis, and now this study shows that blocking IL-1 with the 150 mg dose results in a significant decrease in CVE. Furthermore, we need to keep in mind that significance was attained at the third decimal point and a confirmatory study with a more significant p value with an anti-inflammatory agent will be a more promising therapy for clinical use. There appears to be additional benefits on gout and osteoarthritis with the puzzling finding of no decrease in cancer prevalence but a significant decrease in cancer mortality, possibly due to an anti-inflammatory effect. For both gout and osteoarthritis, the rheumatologist might now consider this to be a more favorable option. The data support a role for hsCRP in atherothrombosis and the authors cannot exclude the possibility that reducing hsCRP was also a potential mechanism. 7 A serious concern is the increase in deaths due to sepsis and infections, especially in diabetic patients, despite stringent exclusion criteria to avoid this serious adverse event. Ideally, an anti-inflammatory strategy that reduces CVE without increasing infectious mortality will be more acceptable for clinical practice.
There are other such strategies which are presently being tested in human clinical trials.5,8 Ridker and colleagues need to be lauded for conducting such a large trial to fruition, and in our view, fulfilling Koch’s postulates with respect to the crucial role of inflammation in atherosclerosis. Yet, further clinical trials are needed to consider it part of our mainstay therapy to further reduce CVE. Furthermore, since this study was conducted in patients with ASCVD, the findings cannot currently be translated to patients without ASCVD. Nonetheless, it remains a significant therapeutic advance in cardiovascular disease.