Novartis' New "Game Changer"
The New York Times article was fairly restrained, simply saying it
has shown a striking efficacy in prolonging the lives of people with heart failure
The Reuters article's headline said
New Novartis drug may upend heart failure treatment
In Forbes, Matthew Herper called it a
The accompanying editorial in the New England Journal of Medicine (1) suggested the drug
may well represent a new threshold of hope for patients with heart failure
On the heels of the debate about the high price of Sovaldi, Gilead's drug similarly touted as a game changer for the also common disease, hepatitis C (look here), almost immediately the concerns about cost followed. Another Reuters article quoted the past president of the European Society of Cardiology,
We are entering a new era of treatment and, of course, it will cost a lot of money, which is a problem
This blog was late to the party for Sovaldi, and did not get to examine the evidence used to support its ostensibly wondrous properties until a while after it was published, and after the drug was approved. Therefore, I thought we should examine the new evidence about Novartis' new drug now, before the drug is reviewed for marketing approval.
Congestive Heart Failure
First, let me give a very informal review of congestive heart failure (CHF). CHF is a syndrome that is the final common pathway for many patients with heart disease. It can be commonly caused by prolonged hypertension, coronary artery disease (often after one or more myocardial infarctions, or heart attacks), cardiac valve disease, or various problems with the heart muscle (cardiomyopathies). In CHF, the heart cannot pump sufficiently to meet the body's needs. The body responds in various ways, not all advantageous. There are two main types of heart failure. In one, systolic dysfunction, the heart's pump function, measured by its ejection fraction, is directly reduced. In the other, sometimes called diastolic dysfunction, the heart becomes stiff, and hence it may not fill with blood adequately. In either case, a common response is accumulation of fluid. Fluid can accumulate in the lungs, causing breathlessness. CHF patients often are fatigued, breathless, find it hard to do physical activity, and may have swelling of the legs or abdomen (edema).
Starting more than 20 years ago, a series of landmark clinical trials found that various types of medication could be used to modify the often disadvantageous compensatory mechanisms evoked by CHF with systolic dysfunction. Controlled trials showed these medications could reduce symptoms, reduce hospitalizations, and prolong life. These medications included those that affected the renin-angiotensin system, used by the body to control blood flow and pressure, including angiotensin converting enzyme inhibitors (ACEIs), and angiotensin receptor blockers (ARBs). Also found to be efficacious were beta-blockers and mineralocorticoid blockers. At this point, all these medications are available in cheap, generic form in the US.
Review of the New Study of Sacubitril and Neprilysin Inhibtion
So let us examine the article just published online that reports a trial of Novartis' new drug.(2)
The new Novartis medication is sacubitril, which inhibits an enzyme called neprilysin, which degrades various active compounds in the body. Inhibition of the enzyme and the resultant increase in the level of these compounds seems to favorably affect the disadvantageous compensation mechanisms found in CHF.
A previous effort to develop a neprilysin inhibitor found that it produced an unacceptable risk of a serious side-effect, angioedema, when combined with an ACE inhibitor. So the current study involved another neprilysin inhibitor, sacubitril, combined with a angiotensin receptor blocker, valsartan, sold by Novartis under the trade name Diovan.
The trial had many good features. It was large, including 8442 patients randomized. It followed patients for longer than the short-term, a median of 27 months. Most importantly, it assessed real clinical outcomes that might be important to patients (that is, "patient-centered outcomes,") including death, acute hospitalization, and a measure of health status and physical function, the Kansas City Cardiomyopathy Questionnaire (KCCQ).
So far, so good. However, a close read revealed some major and some minor problems.
Low Dose of Comparator
This is a bit complex so bear with me.
The study assessed a combination of sacubitril and valsartan versus a well-known ACE inhibitor, enalapril. The target dose of the sacubitril and valsartan, rather confusingly called LCZ696, was 160 mg of valsartan twice daily, for a total of 320 mg/ day of valsartan. The target dose of enalapril was 10 mg twice daily, for a total of 20 mg.
Note that both ACEIs like enalapril and ARBs like valsartan are considered appropriate first line treatment of CHF with systolic dysfunction. In the most recent ACCF/ AHA guideline for heart failure, the maximum suggested dose of enalapril is 10-20 mg twice a day, and of valsartan, 160 mg twice a day. Previous studies have suggested that pushing these doses as high as patients can tolerate improves outcomes.
So the current study compared a targeted maximum dose of valsartan (plus sacubitril) to a less than maximum dose of enalapril. The article did not explain why the study did not call for a maximum dose of enalapril, or indeed why valsartan plus sacubitril was not compared to valsartan alone.
Thus the advantages of valsartan plus sacubitril found in the study could be entirely due to the superiority of maximum doses of valsartan versus moderate doses of enalapril. I do not believe the study design can allow conclusions to be drawn about the efficacy of sacubitril.
Note that this problem was noted in an online comment on the article by Dr Martin Nitschke.
Choosing to compare the new drug to a less than maximally efficacious dose of the comparator medicine could make the new drug appear to be more efficacious than it actually is. This apparent major design problem has been listed as one of the ways commercially sponsored studies may be manipulated to make the sponsors' products appear better. The particular tactic of conducting a trial "against a treatment known to be inferior" was listed in a 2005 article on clinical research manipulation by Richard Smith.(3)
Active Run-In Period
Before patients were enrolled in the trial, they went through two "run-in" periods. In the first, they were switched from their previous doses of ACEIs or ARBs to enalapril, up to 10 mg twice daily. Patients unable to tolerate this were not randomized.
Then the patients were switched from enalapril to valsartan 160 mg plus sacubitril up to . Again, if they failed to tolerate this, they were also not randomized.
The article noted that substantial patients were disqualified during these active run-in periods. During the enalapril period, of 10,513 patients, 1102 were dropped from the study, including 591 (5.6%) due to an adverse effect, 66 (0.6%) due to an abnormal test results, and 49 (0.5%) who died. During the valsartan plus sacubitril period, of 9419 patients, 977 were dropped, including 547 (5.8%) for adverse effects, 58 (0.6%) for abnormal test results, and 47 (0.5%) who died.
There are major problems with such active run-in periods. First, by eliminating patients who had adverse effects from the main results, they make the adverse effect rate appear smaller.
Second, they raise questions about the patient population to whom the study results apply. The study results only clearly apply to patients who are known to be able to tolerate valsartan and sacubitril. In practice, should the drug be approved, at that time the only such patients would be those who already participated in trials. (Of course, the results only apply to patients known to be able to tolerate enalapril. This drug, however, has been on the market for years, and is widely used for problems, like hypertension, other than CHF. So there are already quite a few patients known to tolerate it.)
The problems with interpreting studies that included active run-in periods were clearly described in an article by Pablos-Mendez et al in 1998.(4)
Failure to Consider Adverse Effects Specific to Sacubitril
Valsartan and enalapril are now old drugs. Their adverse effects are well known.
Sacubitril is a new drug in a new class. It inhibits an enzyme that in turn affects the metabolism of many substances. Thus, it is possible it has widespread effects, and may have multiple, and possibly unusual side-effects.
Yet the current study did not apparently include any efforts to anticipate and attempt to assess such side effects. This problem was pointed out in a comment by Dr Ivan Boyadzhiev.
Failure to consider specific adverse effects may mean these effects are overlooked. Thus, it is possible that the new drug has adverse effects that go beyond what are addressed in the article.
Large Numbers of Sites and Quality Control
The study was done in 1043 centers in 47 countries. The list of investigators took up six pages of the supplementary documentation. Thus the average number of patients recruited per center was less than nine.
It seems improbable that a study involving so many investigators and centers, most of whom must have devoted little of their time and effort to this particular study, would have adequate quality control. One indicator that quality control may have been a problem was that although the study inclusion criteria included patients with class II-IV heart failure, about 5% of study patients had class I heart failure.
The article and the supplementary material did not discuss study coordination or quality control. It is not possible to tell whether any such problems with quality control may have distorted the study's results.
Lack of Generalizability in the Patient Population
The complete list of exclusion criteria, only appearing in the supplementary material, was extensive. Patients with many common problems were supposed to be excluded, and the definition of the some exclusion criteria were vague and subjective. These problems included coronary or carotid artery disease "likely to require" surgical or percutaneous intervention within 6 months, "history of severe pulmonary disease," "any surgical or medical condition which might significantly alter the absorption, distribution, metabolism or excretion of study drugs," including inflammatory bowel disease, duodenal or gastric ulcers, evidence of liver disease, and "presence of any other disease with a life expectancy of less than 5 years."
CHF patients are commonly elderly and have other medical problems. Thus the study's results may not generalize to many patients with CHF in real-life.
It would certainly be nice to have a new medicine to improve the management of chronic CHF. Sacubitril may have promise to do so. Unfortunately, the clinical trial that just generated much favorable publicity ahead of regulatory consideration of this drug had major, possibly fatal flaws. Use of a probably non-comparable and inferior "comparator" medication made it impossible to tell whether sacubitril actually is efficacious. Use of an active run-in period, and failure to consider whether this new drug may have unexpected side-effects made it likely that the study underestimated its adverse effects. Study results may not generalize to many real-life patients. There are reasons to be concerned about whether the quality of study implementation was adequate.
Therefore, all the enthusiasm about this drug may be premature, and does not appear to be evidence-based. That clinical research sponsored by organizations that sell health care goods and services may be manipulated to make the sponsors' products look better than they really are is now an old story. We have seen multiple instances in which drugs and devices turned out to be less efficacious and/or more dangerous than originally advertised. Excess enthusiasm about such new innovations may drive up costs, and worse, hurt patients. Physicians, other health care professionals, and those concerned about health policy ought to be much more skeptical about every new instance of a purportedly wondrous innovation.
Evidence-based medicine rigorously applied suggests that individual health care and health policy decisions should be driven by the best available evidence, mostly from clinical research, about the benefits and harms of tests, treatments, programs, and so on, in the context of what outcomes matter to patients. The skepticism EBM should engender lead to health care that is more about patients and their outcomes, and less about ideology, hype, and hucksterism.
ADDENDUM (3 September, 2014) - See also post by Dr Vinay Prasad on the CardioExchange site. He drew similar conclusions about the choice of comparator, and the active run-in period, and added a new concern about the early stopping of the trial. The comments include a - ahem - vigorous exchange between Dr Prasad and Dr Milton Packer, the Principal Investigator of the trial. IMHO, Dr Packer did not completely understand or appreciate the points made by Dr Prasad. Also, see our later blog post on this exchange here.
ADDENDUM (9 September, 2014) - See also a post by Gary Schwitzer on the HealthNewsReview.org blog that summarizes other critical online comments about the PARADIGM - HF study. In particular, Dr Richard Lehman on his Journal Review blog on the British Medical Journal website wrote this scathing paragraph:
And now, alas, to the much-vaunted PARADIGM-HF trial. This has been hyped as the biggest breakthrough in heart failure for at least 20 years. A fixed dose of enalapril was compared with a substance called LCZ696. This turns out to have been a mixture of valsartan and sacubitril, a neutral endopeptidase inhibitor, and most participants ended up receiving a daily dose of 320mg of valsartan, versus 20mg of enalapril. A total of 711 patients (17.0%) receiving LCZ696 and 835 patients (19.8%) receiving enalapril died after a median follow-up of 27 months, at which point the trial was stopped prematurely. So just what are we looking at here? How can we distinguish the effect of the sacubitril from the effect of the high dose of valsartan? Well, we can’t. And this trial is a perfect example of everything that is wrong with heart failure trials. The mean age of the 8442 patients was 63.8, nearly 80% were male, and they were selected by reduced ejection fraction in 1043 centres across 47 countries. A logistic nightmare, but a great way for Novartis to spread influence. And Novartis then collected, managed and analyzed all the data itself. As I have said, a submaximal dose of one drug was compared with a maximal dose of another plus an extra ingredient. There was a run-in period, in which patients who were intolerant to the new treatment (12%) dropped out. Adding bias to bias, the trial was terminated prematurely. One of the primary end-points was hospitalization, which was ably demolished in a JAMA article I pointed out last week. And yes, there was a mortality benefit, but the number-needed-to-treat was about 35 to prevent one death in 2 and half years, in a population far removed from the elderly co-morbid patients we see in real life. As I draw to the end of a clinical career in which I’ve tried to help people with heart failure, I stand bemused. This is just how things have been done for the last 30 years, and it’s not good enough. At the very least, Novartis must make its full data set available for independent analysis. And before sacubitril is licensed, we need a properly designed trial, say between valsartan 160mg b.d. alone and valsartan plus sacubitril, in a typical population with heart failure. It will be very interesting to see what the FDA, the EMA and NICE decide.
1. Jessup M. Neprilysin inhibition - a novel therapy for heart failure. N Engl J Med 2014; DOI: 10.1056/NEJMe1409898. Link here.
2. McMurray JJV, Packer M, Desai AS et al. Angiotensin - neprilysin inhibition versus enalapril in heart failure. N Engl J Med 2014; DOI: 10.1056/NEJMoa1409077 Link here.
3. Smith R. Medical journals are an extension of them marketing arms of pharmaceutical companies. PLoS Medicine 2005; DOI: 10.1371/journal.pmed.0020138. Link here.
4. Pablos-Mendez A, Barr G, Shea S. Run-in periods in randomized trials: implications for the application of results in clinical practice. JAMA 1998; 279(3): 222-225. doi:10.1001/jama.279.3.222. Link here.