–Modest treatment effect of canakinumab may limit clinical role in heart disease, but strong anti-cancer signal sparks interest.
Despite questions about its clinical importance, the CANTOS trial is being hailed by experts for finally validating the role of inflammation in heart disease. The cardiovascular benefits in the trial were real but modest. Even more exciting, exploratory findings may lead to an important new line of cancer research.
The main results of the trial were presented by Paul Ridker (Brigham & Women’s Hospital) at the European Society of Cardiology meeting in Barcelona and published simultaneously in the New England Journal of Medicine. The cancer findings were published separately in the Lancet.
CANTOS (The Canakinumab Anti-Inflammatory Thrombosis Outcomes Study) randomized 10,061 patients with a previous myocardial infarction (MI) and elevated levels of hs-CRP (2 mg or more per liter) to placebo or 50, 150, or 300 mg of canakinumab, a monoclonal antibody that targets interleukin-1β. The drug was administered subcutaneously every three months. In the combined 150/300 treatment group hs-CRP was reduced by 39%. There were no differences between the groups in lipid levels.
After a median followup of 3.7 years all three canakinumab doses led to numerical reductions in the primary efficacy endpoint (nonfatal MI, nonfatal stroke, or CV death), but the 150 mg group was the only one that achieved statistical significance. Most of the difference was attributed to a reduction in the nonfatal MI endpoint.
In the combined 150/300 mg treatment group there was a 15% reduction in the primary endpoint and a 30% reduction in the need for revascularization procedures.
Here are the incidence rate (events per 100 person-years) for the primary endpoint in each group:
- Placebo: 4.50
- 50 mg: 4.11
- 150 mg: 3.86
- 300 mg: 3.90
There were no significant differences in all cause or cardiovascular mortality, though the direction of the numbers favored canakinumab. However, canakinumab was associated with a significantly higher rate of fatal infection compared to placebo. (Fatal infections per 100 person years were 0.18 in the placebo group and 0.31, 0.28, and 0.34 in the 50, 150, and 300 mg canakinumab groups, respectively.) On the positive side, cancer mortality was lower in the canakinumab groups (see below for more on this).
“CANTOS has helped move the inflammatory hypothesis of coronary artery disease forward scientifically,” wrote Bob Harrington (Stanford University), in an accompanying editorial in NEJM. “However, the modest absolute clinical benefit of canakinumab cannot justify its routine use in patients with previous myocardial infarction until we understand more about the efficacy and safety trade-offs and unless a price restructuring and formal cost-effectiveness evaluation supports it.”
Harrington noted that canakinumab, which is currently approved for use in rare auto-inflammatory disorders, costs about $200,000 per year in the US. “Such pricing may be suitable for rare diseases, but not for a common indication such as coronary artery disease, even if given every 3 months.”
Less expensive antiinflammatory drugs may one day be available for a cardiovascular inflammation indication. Ridker is currently heading up the NHLBI CIRT trial, which is testing methotrexate in 7,000 patients. Another trial underway is the Lodoco trial from Australia and the Netherlands, in which 4,300 patients are being randomized to low dose colchicine or placebo.
Strong Praise For Trial Despite Questions About Clinical Utility
A broad array of cardiologists praised the trial and its significance but cast doubt on the likelihood of it having an immediate effect on clinical practice.
“CANTOS is a remarkable scientific success, validating the concept that treatment of inflammation will reduce atherosclerosis,” said Sek Kathiresan (Broad Institute). “Now, the discussion will shift to risk-benefit and cost-benefit for this specific medicine – canakinumab.”
Harlan Krumholz (Yale University) said CANTOS “may have little effect on patient care (given the drug’s price and effect) but will have an outsized effect on the field. We now have evidence that manipulating the immune system can reduce cardiovascular risk. And targets for such drugs are ample. And there is the prospect that older drugs, such as methotrexate (being tested now in an NIH trial by Dr. Ridker) may produce marked benefit… this study now indicates that inflammation as a modifiable risk factor may be moving mainstream.”
Ethan Weiss (UCSF) said the “authors are to be commended for a well-designed trial that really does address the impact of modulating inflammatory cytokines in high-risk CV patients with elevated CRP…. There is what appears to be a real but small reduction in the primary endpoint (2% absolute risk reduction) over the course of the trial. This result is believable and the trends in the other doses support that. It is biologically very interesting especially with the lack of effect on lipids or other risk markers. However, it is a small effect and it comes with some real safety concerns.”
Scott Ratner, a cardiologist in Long Island, NY, expressed caution. He said that the trial “seems like an interesting idea but the concept that the first biologic you try for a complex disease like atherosclerosis would magically work seems like a long shot.”
James Stein (University of Wisconsin) said: “This is an important paper that does what the authors proclaim: it shows for the first time that a drug that specifically targets inflammation and does not alter major serum lipids can reduce CVD events. Although we have pathological, physiological, and animal studies supporting the role of inflammation in atherosclerosis, this type of direct intervention shows we can manipulate the process and that specific targeting of inflammation may be useful treatment. I doubt that canakinumab will become a major treatment since the effect size was small and there are several safety concerns. Inflammation is an integral part of many salutary biological processes, especially response to injury including infection, so some side effects related to infection are expected, but here the risk was serious relative to the benefits. This study opens up many promising pathways and was a hard study to do, so kudos to the study team.”
Sanjay Kaul (Cedars Sinai) praised the trial but expressed doubts about the drug’s prospects. “From a regulatory perspective, the treatment effect is not statistically persuasive (for a single trial); in fact the p value for the 150 mg dose barely made it, which means that only a few extra events in the treatment arm would overturn statistical significance. It would not surprise me if the treatment is not allowed a claim for cardiovascular protection provided the sponsor submits an NDA for this indication.”
Kaul said also that “the fatal infection risk is going to complicate the benefit-risk profile of this drug.”
In an interview, Ridker said that “any discussion of cost is phenomenally naive,” since the current high price of the drug is based on its orphan status. “Obviously if they decide to get a label they lose orphan drug status,” he explained. Further, he explained that “no one is going to treat everyone with this.” He said that those patients who had the largest CRP reduction had a large 25-30% reduction in MI. He proposed a strategy in which patients with high levels of CRP might receive the first dose for free. Only those who experienced a large reduction in CRP would then go on to receive subsequent doses (and, of course, then pay the cost of the drug).
According to Kathiresan we are now entering “an era with a plethora of treatment options for the secondary prevention of atherosclerotic cardiovascular disease. Human genetics has highlighted that atherosclerotic cardiovascular disease is multi-factorial, with several underlying causal pathways. We now have a treatment for several of these pathways.”
“After a heart attack,” he continued, “the average patient right now will be treated with several (generic) standard-of-care meds – aspirin, statin, beta-blocker, ACE-i, and metformin. Which additional meds should be added on? The options now include treatment for several distinct mechanisms: 1) LDL (PCSK9i, ezetimibe, anacetrapib); 2) inflammation (canakinumab); 3) thrombosis/platelets (ticagrelor, clopidogrel); 4) thrombosis/coagulation (low-dose rivaroxaban); and 5) metabolic (SGLT2 inhibitors). And in future, we will likely have therapies for hypertriglyceridemia (APOC3 inhibition, ANGPTL3 inhibition) and lipoprotein(a) (Lp(a) inhibitors). How is the practicing physician to decide which therapies to add for any given patient?”
Kathiresan concluded: “The future of secondary prevention is identifying which mechanistic pathway would be most beneficial for a given patient.”
Kathiresan’s remarks dovetailed with a statement by Ridker in a press release from Brigham & Women’s Hospital. “Cardiologists will need to learn about inflammation today, the same way we learned about cholesterol 30 years ago,” said Ridker. “CANTOS is a demonstration of how personalized medicine will occur in the future, as we now need to distinguish those heart disease patients who have ‘residual cholesterol risk’ from those who have ‘residual inflammatory risk.’ These two groups will require different interventions.”
Cancer Results Might Overshadow The Main Trial
It is possible that the positive findings of the exploratory cancer analysis will gain more attention and excitement than the more modest cardiovascular findings. The trial was not designed to explore the benefits of cancer, but the fact that a mortality benefit from cancer was observed may overshadow the more modest cardiovascular results.
“CANTOS provides the first evidence from a randomised trial in human beings that inhibition of interleukin 1β with the monoclonal antibody canakinumab is associated with reduced incidences of fatal cancer, lung cancer, and fatal lung cancer,” the authors wrote in the Lancet. But, they cautioned, their “data should be interpreted in the context that the primary aim of the trial was to investigate cardiac events rather than cancer events…. Our exploratory data should be replicated and extended in settings directly related to early cancer screening and initial treatment of cancers, particularly lung cancer.”
Over the course of the study there were 196 cancer deaths. Compared to placebo there was a 14% reduction in the 50 mg group, a 22% reduction in the 150 mg, and a 51% reduction in the 300 mg group. The reduction in the 300 mg group achieved statistical significance, though the authors cautioned about the use of statistical tests in an exploratory analysis.
Nearly all the benefit occurred in lung cancer. Incident lung cancer was reduced by 23% in the 50 mg group, 39% in the 150 mg group, and 67% in the 300 mg group. Fatal lung cancer was reduced by 77% in the 300 mg group.
The authors speculated that canakinumab didn’t prevent cancer from starting but said that “a more biologically plausible explanation is that canakinumab reduced the rate of progression, invasiveness, and metastatic spread of lung cancers that were prevalent but undiagnosed at trial entry.”
Sanjay Kaul warned that “the cancer benefit is interesting, but hypothesis-generating.”
Ethan Weiss said the cancer results “diminish the impact of the CV outcome which were paltry to begin with.” He said he was struck by the fact that there was “a clear dose response in the cancer signal and none in the CV endpoints. That is troublesome.”