Is Novartis' LCZ696 "revolutionary" or just a marginal improvement?

Over the Labor Day weekend, Novartis presented the results of a trial for its cardiovascular compound, LSC696, at the European Society of Cardiology (ESC) meeting in Barcelona, Spain.  Reactions to those results were the number one topic on the pharma blogosphere for several days afterward.

Prior to the presentation, Wall Street's sales projections for the drug were in the vicinity of $1 billion per year by 2020.  Immediately afterward the forecasts skyrocketed, especially among the most bullish analysts. Morningstar analyst Damien Conover raised his forecast from an annual peak of just over $1 billion to approximately $6 billion.  Leerink Swann's Seamus Fernandez doubled his estimate from $3.2 billion a year to $6.4 billion.  Sanford C. Bernstein's Timothy Anderson was especially exuberant, predicting annual sales as high as $8 billion.

So do these enormous sales projections for LCZ696, following its Labor Day splash, indicate that the product is a "revolutionary breakthrough" for treating heart failure, as its lead investigator calls it?  Or is this just one more instance where a pharma company promotes a small, tweaked improvement as a breakthrough to justify squeezing payers for premium dollars?

First, a bit of background.

LSC696 actually combines two compounds: valsartan (which is currently sold by Novartis under the name Diovan) and the neprilysin inhibitor, sacubitril.

The valsartan works within a physiological pathway known as the RAAS cascade.  Researchers classify it as an angiotensin II receptor blocker (ARB) because it opposes the action of angiotensin II, a naturally occurring substance in the body that raises blood pressure and forces the heart to work harder.  ARBs help relax and widen blood vessels, thereby lowering blood pressure and making it easier for the heart to pump blood.

The sacubitril is a neprilisyn inhibitor, which means it decreases the body's neprilysin enzyme that, in turn, helps control blood volume and lower blood pressure.  Given this combined mode of action, researchers call LSC696 a RAAS-NEP inhibitor.

Amid all the Wall Street drumbeating, Novartis' design for the study presented in Barcelona was the first thing that aroused skepticism.  Novartis compared the new regimen to enalapril.  Enalapril acts on the RAAS cascade differently than ARBs.  It prevents the conversion of angiotensin I to angiotensin II.  As such it is classified as an ACE inhibitor.  Within this treatment group, enalapril is actually one of the oldest and weakest ACE inhibitors, having been approved in 1985 as branded Vasotec.  Although matching the LCZ696 test drug/regimen against another compound is preferable to pharma's usual approach of comparing something new to placebo, a comparison to something that is less than state of the art seems to stack the deck in favor of the new medication.

Just as importantly, contributing authors in the New England Journal of Medicine pointed out that the enalapril dose used in the trials was lower than the one generally recommended.  At the same time, investigators had the liberty to increase dosing on the valsartan portion of LCZ696 to its maximum level.  That means Novartis' researchers may have compared a full dose of valsartan to a moderate dose of enalapril, in which case something other than LCZ696's use of a NEP inhibitor is what produced a better result.  This observation is part of what led several cardiologists who attended the meeting in Barcelona to ask why LCZ696, if it represents a truly improved level of care, failed to reduce atrial fibrillation, a key characteristic of worsening heart failure.

Some observers question yet another aspect of the LSC696 study design.  They point out that patients who received the Novartis test regimen had significantly lower systolic blood pressure before the study than those who received enalapril.  The higher ingoing blood pressure among enalapril users may have created a more advanced cardiovascular condition that made their hospitalizations or deaths more likely.

Other cardiologists at ESC questioned the part of Novartis' study plan that included a "washout" period for patients enrolled in the trial.  Investigators switched patients from the heart medications they were taking prior to the study and gave them either enalapril or the LSC696 duo before starting the trial.  Those patients that could not tolerate the therapies were not enrolled.  That means the incidence of side effects and other adverse events in the study was likely lower than what cardiologists would see in real world practice because the enrollment was skewed to favor patients who can tolerate valsartan+sacubitril.

After considering the study's design, a second reason for skepticism relates to interpreting the results. The press releases tout the fact that LCZ696 confers a 20% relative risk reduction versus enalapril for developing the study endpoints of heart failure hospitalization or death. What that actually means is there was a 26.5% chance for hospitalization/death with enalapril versus a 22% chance with LCZ696.  Now a 4.5% smaller risk can make a huge commercial difference, because that was the approximate extent to which Plavix reduced risk versus aspirin.  Plavix went on to become the world's second highest selling product behind Lipitor.  Clinically and epidemiologically, however, that represents a good but not a great difference. Thirty-two people need to be treated (NNT) with LCZ696 to prevent one death and the rule of thumb is that an NNT of 50 or higher means something is fairly useless.  In other words, LCZ696 may represent yet another marginal, incremental improvement rather than any sort of revolutionary breakthrough.

In addition to issues related to the study's design and interpretation, a number of cardiologists and other researchers raise an issue about sacubitril's mode of action:  inhibiting the neprilysin enzyme.  They fear the process could affect other metabolic pathways and produce harmful consequences.  For example, some researchers argue that neprilysin seems to play a role in reducing beta-amyloid in the brain.  Beta-amyloid forms the clumps and tangles characteristic of Alzheimer's disease and reducing the neprilysin enzyme may raise the risk of patients developing Alzheimer's.  Novartis ended the study reported in Barcelona too early to assess such side effects.

LSC696 is not the first RAAS+NEP inhibitor promoted by its developer as something that will revolutionize the way cardiologists treat the hypertension-to-heart failure process.  The same earth-shaking predictions for Bristol-Myers Squibb's omapatrilat (Vanlev) were all over the media a decade ago.  Novartis's lead investigator on LSC696 is Dr. Milton Packer from the University of Texas Southwestern in Dallas.  Ten years ago the same Dr. Packer, then at Columbia University in New York, was BMS's lead investigator on Vanlev and he was equally effusive in touting that product as a revolutionary breakthrough.  Alas, Vanlev never even made it to the market.  The FDA decided against approving it because a number of patients in the studies developed potentially serious angioedema, a swelling of the lips and throat.

Around the same 2004-2005 timeframe, Pharmacia was developing another RAAS-NEP inhibitor, eplerenone (Inspra), one that did receive FDA approval.  Nonetheless, its modest benefits led Pfizer to promptly bury Inspra as soon as they bought Pharmacia.

Is LCZ696 sufficiently different from these other RAAS-NEPs to be more effective in controlling heart failure while avoiding their side effects and safety-tolerability problems?  The Novartis therapy replaces an ACE inhibitor with an ARB and it uses a different neprilisyn inhibitor.  But as often happens in pharmacology, the several different compounds within a chemical class produce the same clinical effect.

Whether LCZ696 lives up to its projections for pulling in $6-8 billion a year will also depend on how Novartis prices the product.  In the past, the company has been remarkably tone deaf on that factor, as witness the launch of its MS product Gilenya.

Novartis has about six or seven years all to itself for using this RAAS-NEP inhibitor before a competitor potentially can bring another one to market.  That's both an advantage and a disadvantage.  The advantage of working without a competitor is obvious, but the disadvantage is that Novartis will have to generate the primary demand for a RAAS-NEP all alone.  Acting by themselves, they will have to provide data that satisfactorily answer all the questions about the study.  Without competing RAAS-NEP brands on the market, Novartis will be the only company looking to demonstrate and publicize better long-term outcomes that can justify premium pricing.  It remains to be seen how well they will do all of this before a competitor or two come along.

Finally, it seems questionable whether any branded pharma company would pair either omapatrilat or eplerenone with an off-patent ARB and sponsor the studies needed to obtain regulatory approval.  Eplerenone is already a generic and omapatrilat will likely lose patent protection this year.  Nonetheless, it is worthwhile speculating whether a Teva, an Actavis or a Dr. Reddy's – all of which know how to make good margins on generics – might be willing to sponsor and conduct the studies that test how well one of them, combined with a generic ARB, treats heart failure.

If one or more generics companies did make that play, the sales projections for LCZ696 would likely fall back to what they were before Labor Day or possibly even lower.

When the FDA shot down Vanlev, some observers predicted that it was unlikely another RAAS-NEP would ever see the light of day.  Perhaps the public relations Novartis manufactured for LCZ696 might provide its most important benefit if all the talk inspires the development of an all-generic medication for heart failure, one where its slightly better effect is matched by a commensurately modest price.

Read more from the Check Up blog »