Hopes Fade for a CV Indication for Canakinumab: What’s Next for the Inflammatory Hypothesis?
The CANTOS sponsor has given up on a CV indication for its monoclonal antibody, leaving some to ask if marketing trumped medicine.
The prospect of targeting inflammation to treat residual cardiovascular risk using the monoclonal antibody canakinumab (Ilaris) appears to be slim to none in the wake of recent decisions from regulators in the United States and Europe. While that news might be frustrating for some clinicians and researchers, it’s quite possible that Novartis, the drug’s manufacturer, isn’t overly disappointed that its agent won’t be used any time soon for patients with cardiovascular disease.
Crunching the numbers, the cardiovascular indication always posed a sticky business proposition for Novartis. Canakinumab is currently approved—and priced—as an orphan drug for several conditions, including active systematic juvenile idiopathic arthritis. If the drug had gained an indication for cardiovascular disease prevention, the company would have had to drastically cut the price of the agent by as much as 98% for it to make financial sense for the large number of patients who would be eligible for treatment.
Such pricing concerns, however, are likely no longer relevant.
Late last year, the US Food and Drug Administration issued a complete response letter to the sponsor, essentially rejecting the application on the grounds that the submitted data were insufficient to support the company’s proposed indication. In Europe, Novartis opted to close the file on the cardiovascular indication when it withdrew from the approval process without addressing questions from regulators there.
The FDA rejection and European withdrawal seem in stark contrast to the enthusiastic reception CANTOS received in 2017. At the time, financial analysts pegged the anti-inflammatory agent canakinumab as a potential blockbuster, with future sales estimates in the billions.
For Sanjay Kaul, MD (Cedars-Sinai Medical Center, Los Angeles, CA), the decision to walk away from the European process—Novartis withdrew its application for a cardiovascular indication in December 2018—appears to be a case of marketing considerations trumping science.
“It seems to me, absent any cogent reasons for not pursuing the claim, marketing decisions might have driven the decision to withdraw,” said Kaul. “Half a billion dollars in revenue was at risk.”
While a cardiovascular indication could have earned Novartis multiples of what it makes selling canakinumab for rare conditions, there was always some uncertainty involved. The company would have had to forgo orphan-drug pricing—thereby giving up a dependable revenue stream—at a time when the financial prospects from the cardiovascular indication weren’t guaranteed, say experts.
No Guarantee of Success
In 2018, canakinumab earned Novartis roughly $400 million in sales, according to estimates. As an orphan drug, the monoclonal antibody costs roughly $200,000 per year when injected monthly, or $16,000 per injection. If delivered every 3 months, as it was administered in the CANTOS trial, the annual costs are substantially lower but still unreasonably high for a therapy expected to treat a common condition like cardiovascular disease.
In fact, according to a recent economic analysis published online January 16, 2019, ahead of print in JAMA Cardiology, the incremental cost-effectiveness ratio for canakinumab at its current market price is a staggering $6.4 million per quality-adjusted life-year (QALY) gained, a number that far exceeds the typical willingness-to-pay threshold of $100,000 per QALY gained.
Jenny Bjerre, MD, who led the cost-effectiveness analysis along with Thomas Sehested, MD (both Copenhagen University Hospital, Denmark), told TCTMD that they assumed if canakinumab did get an approval from FDA for this indication, the price would be reduced. At the current orphan price, though, it widely misses the mark.
Robert Harrington, MD (Stanford University School of Medicine, CA), told TCTMD that if canakinumab was approved broadly for a common condition like secondary prevention of cardiovascular disease, Novartis would certainly have had to restructure its pricing. Kaul pointed out that had Novartis gained the cardiovascular indication, the number of patients eligible to take the drug would have expanded dramatically. But, he added, there was no guarantee of financial success given potential restrictions and red tape from payers.
Witness, for example, the relatively slow uptake of costly drugs like the PCSK9 inhibitors alirocumab (Praluent; Regeneron/Sanofi) and evolocumab (Repatha; Amgen) and Novartis’ own heart-failure drug sacubitril/valsartan (Entresto). The PCSK9 inhibitors—both of which were criticized for their initial market price of $14,000 per year—have had a rough ride from payers, and patients have faced an equally hard time getting their prescribed medications, leading the drug makers to cut prices or work out deals to make the drugs more palatable.
Questioning the Emphasis on ‘Responders’
In the CANTOS trial, which included 10,061 patients with a previous MI and elevated high-sensitivity C-reactive protein (hsCRP), treatment with 150 mg of canakinumab reduced the relative risk of nonfatal MI, nonfatal stroke, or cardiovascular death—the study’s primary composite endpoint—by 15% compared with placebo (P = 0.021). Treatment with canakinumab was also associated with a reduction in the risk of cancer, particularly lung cancer, although the risk of fatal and nonfatal infection was higher among patients who received the monoclonal antibody.
Given the positive primary endpoint, Novartis submitted the data to the FDA in hopes of gaining a cardiovascular indication. However, the company was not looking for a broad indication based on the overall CANTOS trial, but rather for a specific approval in patients considered “responders” to treatment. That proposed indication was based on a subgroup analysis that showed patients who achieved a substantial reduction in hsCRP (to < 2.0 mg/L) had a bigger benefit from treatment, including a significant reduction in cardiovascular and all-cause mortality.
Several experts who spoke with TCTMD suggested there was an extremely limited possibility—one doctor called it a “vanishingly small chance”—that the FDA would approve canakinumab among responders given that the subgroup analysis was largely viewed as hypothesis-generating. Harrington, who has served on and chaired numerous FDA advisory committees, said the responder analysis likely “wouldn’t fly” if it were presented before the agency’s Cardiovascular and Renal Drugs Advisory Committee members.
The FDA rejected the proposed indication, stating that the data did not support use of canakinumab as a targeted therapy for patients with cardiovascular disease who achieved a reduction of hsCRP below the 2-mg/L mark in response to the initial dose.
“Overall, [canakinumab] had a fairly modest effect in a single trial that had some issues attached to it,” said Harrington. “So I’m not surprised they received a complete response letter. That doesn’t mean they couldn’t have tried to recharacterize things and potentially lobby for an advisory committee meeting. I’m not privy to the details at Novartis, but they obviously chose not to do that.”
To TCTMD, Kaul said there was zero chance the FDA would have approved canakinumab based on the subgroup analysis. The agency will take advantage of subgroup analyses to refine the benefit-risk trade-off, but only if the results are positive in the overall cohort. With CANTOS, though, the responder analysis was prone to statistical confounding because it was based on a postrandomization variable, he said. Moreover, the results of the overall trial weren’t overwhelming.
So why focus the potential indication on canakinumab responders based on a subgroup analysis, which had a limited chance to be approved by the FDA?
A Novartis spokesperson said the company did so because the responder analysis, as well as other internal analyses using different methodologies, identified these patients as having the largest benefit and strengthened their belief in response-guided therapy. “This would have enabled physicians to take a personalized medicine approach with canakinumab by enabling assessment of a patient’s response early in the course of therapy and [to] make treatment decisions accordingly,” according to Novartis.
Even in the unlikely event it was approved for responders only, Novartis was still looking at a drastic price cut for canakinumab. Despite a significant 31% decrease in all-cause mortality in the responder group, which improved the drug’s cost-effectiveness, results showed it was still far from making financial sense at $819,000 per QALY gained when priced at $73,000 annually.
Withdrawal of European Application
In Europe, the regulatory details differed, but the process still led some to question just how badly Novartis wanted canakinumab to be approved for cardiovascular risk prevention.
In response to Novartis’ marketing request, the Committee for Medicinal Products for Human Use (CHMP), the group responsible for assessing products for the European Medicines Agency, came back to the company with a number of outstanding issues that needed to be addressed. At that time, CHMP felt the data from CANTOS were “not robust enough to clearly demonstrate that canakinumab is effective in all patients who have had a heart attack.”
Rather than answer the questions raised by CHMP, Novartis withdrew its application, stating it would be unable to address the major objections to support a cardiovascular indication within the required time frame. The company did reserve “the right to make further submissions at a future date in this or other therapeutic indications.”
Asked about the ethics of walking away from the regulatory process, Kaul was careful in his answer. He said it’s uncertain whether such a move is unethical, but added that it’s highly unusual, if not unprecedented, for a company to withdraw a claim because it didn’t have enough time to address objections raised by regulatory authorities, particularly if the clinical trial that supports the indication reached its primary endpoint.
Harrington also commented on whether Novartis had an ethical obligation to fully pursue a broad clinical indication and took a pragmatic view by framing his response around the trial results.
If CANTOS were “wildly positive,” with significant reductions in mortality and MI, then there would be a large push from the clinical community to demand the therapy be made available to physicians treating patients with coronary artery disease. On the flip side, if the drug didn’t work, but instead caused harm, it’s an easy call not to ask for a cardiovascular indication.
“Instead, here, we end up in the middle where canakinumab has a modest effect, that effect is balanced by some harm, and there is no effect on mortality,” said Harrington. “So I don’t think Novartis has a moral imperative here—the data are not so clean-cut.”
Commenting on the decision to withdraw the application for European approval, Novartis acknowledged that committee members had “some major objections.” In looking at their concerns, “our conclusion was that the data from the trial was not adequate to address those objections.” Asked if the decision to walk away from Europe was based on market factors, the Novartis spokesperson reiterated that the major driver was the inability to address CHMP’s questions with the available data.
TCTMD also asked Paul Ridker, MD (Brigham and Women’s Hospital, Boston)—who, along with Peter Libby, MD (Brigham and Women’s Hospital), is credited as the driving force behind this field of research—about the approval strategy pursued by Novartis and the company’s apparent decision to set aside plans for a cardiovascular indication; he declined to comment. Instead, Ridker praised Novartis for the leap of faith it took in supporting an academic-industry collaboration for an unproven hypothesis in cardiovascular disease progression. Without the company’s support, he said, neither the cardiovascular community nor the oncology community would be where they are today.
“The research community’s task now is to find alternative approaches to inflammation inhibition that are as effective but hopefully less expensive so that payers will be comfortable with broad usage,” Ridker said.
What About the Inflammatory Hypothesis?
But for some, the regulatory to-and-fro with canakinumab can’t help but cast a shadow over the larger discussion about the “inflammatory hypothesis” of cardiovascular disease progression. While it could be argued CANTOS was an academic exercise to prove that inflammation does contribute to atherosclerotic risk, and that regulatory approval required a larger hurdle to be cleared than could be addressed in the trial, that argument only leads to confusion.
“If the data are not robust enough for the company to pursue a label claim, are they robust enough to claim CANTOS validated the inflammatory hypothesis?” asked Kaul. “I don’t think you can separate the two. I have a hard time reconciling that. That’s purely from an academic perspective. When you say the trial validates [the inflammatory hypothesis], the question you need to ask is: what strength of evidence is needed to validate a hypothesis that investigators have been working on for 30 years? I would say it requires ‘beyond a reasonable doubt,’ not just simply a preponderance of evidence. That’s the biggest dilemma with the regulatory claim.”
As the CHMP’s list of objections makes clear, CANTOS didn’t prove beyond a reasonable doubt that the anti-inflammatory agent canakinumab reduced atherosclerotic cardiovascular events, said Kaul. While CANTOS showed a modest reduction in the risk of nonfatal MI, nonfatal stroke, or cardiovascular mortality with the 150-mg dose of canakinumab, the benefit was mainly driven by a 24% reduction in MI.
Harrington agreed, stressing that the CANTOS trial wasn’t a slam dunk for canakinumab. For example, the CANTOS investigators originally intended to enroll 17,200 patients but Novartis reduced the size of the trial for financial considerations (done without access to data). Additionally, the trial originally began with the 150-mg and 300-mg doses of canakinumab, then a 50-mg dose was later added at the request of regulatory bodies. Finally, Harrington, like Kaul, noted there was no effect on cardiovascular mortality and no details about the type of MIs prevented.
Unlike Kaul, however, Harrington said he believes the inflammatory hypothesis was largely confirmed with the trial. “You have a therapy that only works through an anti-inflammatory mechanism and you had a demonstrable effect on clinical outcomes,” he said. “That’s positive for the field.”
Ridker said CANTOS provided “definitive proof of principle” that inhibiting inflammation—at least with the interleukin-1β inhibitor canakinumab—reduces cardiovascular events independent of blood pressure and cholesterol.
Recently, the Cardiovascular Inflammation Reduction Trial (CIRT), testing a different anti-inflammatory agent and also led by Ridker, came up empty-handed. In that study, low-dose (and inexpensive) methotrexate, used for the treatment of rheumatoid arthritis and other autoimmune disorders, failed to prevent cardiovascular events among patients with a prior MI or multivessel coronary artery disease. CIRT, however, did not include patients with extensive inflammation, and treatment with methotrexate had no effect on pro-inflammatory cytokines interleukin-1β and interleukin-6 or hsCRP.
The door has not closed altogether on the inflammation hypothesis. Several other anti-inflammatory agents are currently being tested, including colchicine in the OASIS-9 CLEAR-SYNERGY trial and COLCOT.
As for what’s next with canakinumab, the future is uncertain. Novartis is exploring its use in other disease states, such as non-small-cell lung cancer and diseases that involve the interleukin-1β pathway.
Sehested TSG, Bjerre J, Ku S, et al. Cost-effectiveness of canakinumab for prevention of recurrent cardiovascular events. JAMA Cardiol. 2019;Epub ahead of print.
- Sehested reports receiving grant support from the Lundbeck Foundation.
- Bjerre reports receiving grants from the Danish Society of Cardiology and the Danish Heart Foundation.
- Kaul reports no relevant conflicts of interest.
- Ridker reports receiving research grant support from Novartis Pharmaceuticals to conduct the CANTOS trial. He reports serving as a consultant to Novartis and is listed as a co-inventor on patents held by Brigham and Women’s Hospital that relate to the use of inflammatory biomarkers in cardiovascular disease and diabetes that have been licensed to AstraZeneca and Siemens.
- Harrington reports receiving grants from Novartis, GlaxoSmithKline, Regado, Sanofi Aventis, AstraZeneca, Janssen, and Bristol-Myers Squibb; grants and personal fees from Merck and The Medicines Company; personal fees from Amgen, Gilead, and WebMD; other from Element Science; and personal fees and other from MyoKardia. He serves on the board of directors for SignalPath (software), Scanadu (mobile health), the American Heart Association, and Stanford HealthCare. He also chaired the Data Safety and Monitoring Board for the CIRT.