(Updated with an additional comment from Milton Packer)
More safety concerns are being voiced about Entresto, the new, ballyhooed heart failure drug from Novartis. Fueling these concerns is a mix of speculation, worry about the FDA’s rapid approval of the drug, and a long-running feud between two leading heart failure researchers.
The concerns, raised in a JAMA Viewpoint by heart failure expert, Arthur Feldman, and co-authors Julia Haller and Steven DeKosky, focus on the possibility that a key mechanism of the drug, neprilysin inhibition, may have troubling repercussions in the central nervous system and the eye. Entresto is a combination of the novel neprilysin inhibitor sacubitril and the older, well-established angiotensin receptor antagonist valsartan.
The core concern is that neprilysin plays a central role in the degradation of beta amyloid in the brain and the eye. Inhibition of neprilysin might then lead to an increased risk for Alzheimer’s disease and age-related macular degeneration.
The macular degeneration issue is new and has not been raised in the cardiology community until now, since a possible role for beta amyloid in macular degeneration was only recently discovered.
The concern about Alzheimer’s disease is not entirely new, having been aired earlier this year in the European Heart Journal, as I reported at the time. But although the drug has now gained approval the concerns about beta amyloid has not been eliminated, and they have been “largely absent from public discussions about the beneficial effects of” Entresto, write Feldman and co-authors.
An Observational Study?
At the request of the FDA the neurocognitive question is being rigorously studied in PARAGON, the next big Entresto trial in heart failure patients with preserved ejection fraction. But the viewpoint authors call for “substantive steps… to acquire data in a more timely manner” and to assess the issue in the heart failure patients with reduced ejection fraction who are now currently taking the drug. They recommend a “public-private partnership model” for a postmarked surveillance system in the current patient population.
The viewpoint authors write that “there is equipoise regarding the utility of valsartan/sacubitril in patients with heart failure that will likely result in a cohort of patients receiving valsartan/sacubitril and a second cohort treated with traditional therapy of an angiotensin-converting inhibitor or an angiotensin receptor antagonist.”
But there is no such equipoise, responds Milton Packer, the co-principal investigator of the pivotal PARADIGM-HF trial, in an interview. Given the highly significant clinical benefits in PARADIGM-HF he said it would be unethical to ask patients to remain on ACE inhibitors. An observational study such as they propose “would require physicians to maintain patients on an ACE inhibitor for years and years even though there is persuasive evidence” that Entresto saves lives. “What,” Packer asks, “would the informed consent look like?”
Packer also pointed to the inevitable limitations of observational studies because it it is impossible for an observational study to completely correct for confounding effects. “Randomization controls for confounding in an observational study.”
In an interview Feldman asserted that equipoise does exist because there is significant residual doubt in the heart failure community about the drug, thereby creating a “natural experiment” that can help resolve troubling questions about the drug.
Packer also said that the results of PARAGON will be known before the observational study could even be completed. (In their viewpoint the authors write that “the timetable for the final report” of PARAGON is in 2022, but Packer said the trial will be finished much earlier than that.)
The viewpoint authors acknowledge that no suggestion of neurocognitive or ocular adverse effects were found in PARADIGM-HF, the key trial supporting the drug’s use, or in other studies. But, they write,
“These studies do not fully dispel concern, however, because patient follow-up in PARADIGM-HF was relatively short; Alzheimer disease–specific, dementia-related adverse events were not prespecified; and executive dysfunction, a pathognomonic finding in Alzheimer disease, was not measured. Furthermore, young monkeys and normal human volunteers do not have presenile plaque or blood-brain barrier dysfunction, both of which have the potential to increase the leakage of drugs into the central nervous system. By contrast, patients with heart failure have multiple risk factors for both Alzheimer disease and blood-brain barrier disruption including older age, hypertension, elevated cholesterol levels, cerebrovascular disease, and diabetes. In addition, the duration of time necessary for patients to manifest cognitive or behavioral symptoms is not immediate and is likely dependent on whether they had preclinical amyloid plaque already present in the brain or in the eye.”
For now they write, “the risks of neprilysin inhibition in the brain and in the eye remain speculative.” But since “the emergence or worsening of cognitive or visual impairments would be devastating for patients with heart failure” they recommend “a prudent approach” that would “follow high-risk patients closely with cognitive assessments, amyloid positron emission tomography, and retinal imaging until definitive answers emerge.”
Packer said that it is impossible to dismiss the theoretical dangers of neprilysin inhibition but that the unequivocal benefits of valsartan-sacubitril in PARADIGM-HF should not be dismissed. He noted that the “relatively short” followup period “was because the trial was terminated early by the overseeing ethical committee when it observed that fewer patients were dying when they were treated with sacubitril/valsartan.” He asked: “How many excess deaths can Dr. Feldman tolerate to justify his desire for clarity? I can assure him that — with greater follow-up — his fears might be justified, but only because it is not possible for dead patients to develop dementia.”
The Middle of the Journey
It is worth noting here that the JAMA viewpoint will almost certainly not be the last attempt to raise questions about the role of Entresto in clinical practice. I have heard a rumor that a major journal will be soon publishing a paper raising more questions about PARADIGM-HF.
It is also worth noting that this is not the first dance for Feldman and Packer. They have been at odds for more than 20 years. Back in 1993 Packer wrote a NEJM editorial critical of a trial run by Feldman studying vesnarinone, a novel heart failure drug that ultimately failed.
In response to this story Milton Packer sent the following statement:
“It is really important to understand the difference between evidence and speculation. The PARADIGM-HF trial provides conclusive evidence that sacubitril/valsartan reduces cardiovascular mortality more effectively than guideline-recommended doses of enalapril. None of the clinical trials to date with neprilysin inhibition have noted any concerns about cognition or vision. Therefore, Dr. Feldman’s opinion piece should be regarded for what it is: pure speculation.
“The question is: what purpose does it serve? If Dr. Feldman wants to raise questions about what might happen to patients whose lives are prolonged by sacubitril/valsartan, I have no problem with that. But it would be irresponsible for him to rely on such speculation to suggest that sacubitril/valsartan not be used instead of ACE inhibitors to prolong life in the first place.”