Here at the European Society of Cardiology (ESC) 2017 Congress, only hours after the presentation of the CANTOS trial, I see two ways to think about this landmark trial. One is scientific, the other clinical.
The Scientific: (Almost) Linking Inflammation to Atherosclerosis
For years, clinicians have felt inflammation played a causative role in atherosclerosis. But the exact link has remained uncertain. No blocker of inflammation has yet proven successful in improving outcomes.
IL-1β sits upstream in the inflammatory cascade. Blocking it with canakinumab (Ilaris, Novartis) should have significant anti-inflammatory effects. And it does. The drug works in rare rheumatologic conditions, and it’s been shown to reduce levels of the inflammatory biomarkers CRP and IL-6 in patients with diabetes.
The question in CANTOS was whether would it work in patients with established heart disease who have a high inflammatory burden—as measured by elevated levels of CRP. It did.
This is a remarkable discovery. Millions of people in this world suffer and die from atherosclerosis. A completely new target—independent of LDL-lowering—that may provide a spark to study other targets in this area is huge.
Will future generation of clinicians look back on the work of Dr Paul Ridker’s team as the beginning of a new era? It’s possible.
One caveat: I used the modifier almost to describe the claim that CANTOS validates the inflammation hypothesis of atherosclerosis.
In an email, Dr Sanjay Kaul (University of California, Los Angeles) noted that without a low–hs-CRP arm in CANTOS, it’s hard to make that claim. “If treatment benefits are evident in a low–hs-CRP arm as well as in high–hs-CRP arm (as was observed with rosuvastatin in the HOPE-3 trial and not possible to evaluate in JUPITER because patients with hs-CRP <2 mg/L were not enrolled), can you really claim inflammatory hypothesis of atherosclerosis is validated?”
The Clinical: Not Likely to Make Prime Time
I doubt canakinumab will become a new cardiovascular drug. The benefits seen in CANTOS may have been statistically significant, but they are not clinically significant.
A 15% relative risk reduction of the primary end point (a composite of nonfatal MI, nonfatal stroke, or CV death) with the 150-mg dose of canakinumab translates to an absolute risk reduction of 0.64% or an NNT of 156. Nearly all of that reduction came in nonfatal MI. There was no significant difference in stroke and cardiovascular death. Overall mortality, too, was not lowered.
And you would have to balance these paltry reductions against the safety signal: a small rise in serious infection risk.
On the matter of cost, the drug is now priced at $200,000 per year. Ridker points out that one cannot use that number because it reflects an orphan drug cost. If Novartis went for a secondary-prevention indication, it would no longer be an orphan drug—and it would cost less. I would argue, though, based on recent developments in US drug pricing, the cost of this novel drug would still be quite high—much higher than its net clinical benefit.
Perhaps the more interesting clinical aspect of canakinumab was its marked reduction in cancer mortality. CANTOS authors published this exploratory finding in a separate paper in the Lancet. The relative risk reduction with the 300-mg dose reached 51%, mostly driven by a 77% reduction in death from lung cancer.
Ridker told me this was not a surprise. They had a cancer adjudication committee on day 1 of trial development. That’s because IL-1β and inflammation have been implicated in cancer development and progression.
This seemingly dramatic finding deserves great caution in its interpretation. CANTOS enrolled 10,061 patients, and 196 (1.9%) of them died of cancer. So the analysis of cancer deaths stems from a very small subgroup of a trial not designed to answer that question. Also, the 51% relative reduction of cancer death translates to a tiny 0.33% in absolute terms.
A passionate clinical scientist recently described his early discoveries as “mere snowflakes on a mountain.”
Is CANTOS the beginning of a major blizzard or a passing snow flurry? Only time will tell.