Tuesday, September 15, 2009

Vivus Can Teach Old Drugs New Tricks, But Does It Have a Bidil Problem?

A lot has been said and written over the past week about Vivus' obesity drug candidate Qnexa, which is a combination of two older, approved products. Its Phase III data was surely quite impressive, as impressive perhaps as the obesity market has been intractable.

But setting aside questions about safety and efficacy--lets assume the company is on to a winner in those respects--and likewise ignoring the not-so-successful track record of existing drugs approved to treat obesity (two recent posts by our friends at Bnet make each of those points well, here and here), there are some complicating and inconvenient factors for Qnexa that aren't getting much play.

First some background. On September 9 the company announced spectacular weight-loss data from two Phase III trials; one study showed an average of 37lbs (14.7%) weight loss for patients on the drug for 56 weeks. The news, which sent Vivus shares up a whopping 70%, also put pressure on late-stage rivals Arena and Orexigen.

(As an aside, the fact that there are three Phase III obesity drug candidates out there, each showing a little extra leg for eligible pharma partners and each surpassing FDA's 5%-weight-loss hurdle guideline for obesity drugs, is a drug-journo's dream, so expect lots more stories on these candidates from your favorite news outlets.)

Arena is developing the serotonin activator lorcaserin (a first Phase III was successful, showing average 5.8% weight loss; a second Phase III trial is expected to report out any day now) and Orexigen is developing Contrave, a combination of the antidepressant bupropion (aka Wellbutrin) and a sustained release form of naltrexone, a an opioid blocker marketed to treat various addictions. Contrave's Phase III data in July was also solid, showing an average 6.1% and 6.4% weight loss in two trials.

Each company even managed to take advantage of their good data and investors' returning appetite for biotech: Arena raised more than $52 million and Orexigen netted $81.6mm in a FOPO on the backs of their respective news. Vivus has yet to pull the trigger on a stock offering (it had just over $144mm in cash and equivalents at mid-2009).

OK. So far, so good for Vivus: it has shown the most impressive results and could therefore ink a successful partnership and have a leg up on its rivals for share in tomorrow's obesity-drug-marketplace. Perhaps! But Qnexa is, like Contrave, a combination of two generic drugs: the stimulant phentermine and the epilepsy and migraine treatment topiramate. Vivus' explanation for how the drugs work together is here.

Now we are aware that this isn't exactly the same scenario, but Qnexa reminds us of another generic-generic combination that was destined for blockbuster-status (only to fall very flat): Nitromed's Bidil. And why did Bidil hit the skids so spectacularly? One word: Pricing.

It's by no means a perfect comparator, but the similarities are there. Bidil is a fixed-dose combination of the generics isosorbide dinitrate and hydralazine hydrochloride used to treat heart failure in black patients. The drug showed stunning results in this population, FDA approved it in June 2005, and then Nitromed priced it at $1.80 per pill--or about $5.40 per day (up to as much as $10.80/day depending on a patient's dose).

That price was four times the retail cost of the combined generics and much higher than even industry analysts thought Nitromed would go. FDA said yes on safety and efficacy, but would doctors prescribe it at that price? Would payors foot the bill? See this IN VIVO feature from a few months post-launch for all the details.

Suffice it to say, although the company had blockbuster dreams, Bidil never took off, at least in part because of its price relative to generics (though the company surely tried to nip generic prescribing in the bud, going as far to file a Citizen's Petition to get FDA to reiterate that there was no generic substitute for Bidil; FDA complied). Did Nitromed strike out while swinging for the fences, when it could have settled for a nice double?

Now, again, Qnexa isn't Bidil, and the circumstances aren't identical. The components of Qnexa aren't available commercially right now in the exact doses that Vivus has tested in combo. And Qnexa's formulation involves controlled release, minimizing certain side effects like tingling, which is related to topiramate. Doctors also "won't take on liability of prescribing off-label when there is a drug approved by FDA," Vivus COO Peter Tam told our colleagues at 'The Pink Sheet' DAILY last week, adding "we are not concerned about generic substitution." (Orexigen execs are similarly sanguine about generics.)

But a highly priced Qnexa may still face some of the same hurdles as Nitromed's Bidil. Off-label prescribing may become a factor. Payors that universally prefer behavioral changes--diet and exercise--in this kind of market may balk at reimbursement for all but the most intractable of patients, limiting Qnexa's market.

Orexigen has in fact said that they planned to price Contrave roughly in line with or even lower than the total cost of its component generics, though at about $8 per day for bupropion and naltrexone generics (a price quoted by Orexigen execs during a presentation in January) that isn't exactly a low ceiling. Topiramate and phentermine cost considerably less, under $1/pill for 100mg topiramate and just over $1/pill for 15 mg phentermine at, for example. (Qnexa's highest dose in Phase III--the one that achieved the best results--was 92mg of topiramate plus 15 mg of phentermine, once a day.)

Now, another potential difference between Bidil and--should either drug make it to market--Qnexa and Contrave: Nitromed tried to sell Bidil itself, without a partner. Both Orexigen and Vivus are out pounding the pavement, in search of pharmaceutical partnerships, or perhaps acquirers, to provide the necessary commercial infrastructure for selling a mass market obesity drug.

We won't be surprised to see either company land a deal, though we have our doubts--because of the pricing problem described above--that either drug will land the kind of blockbuster alliance or acquisition that each company is surely hoping for. What are we expecting: lots of risk-sharing stuff like sales milestones, lower-than-typical-Phase III upfronts.

In short, maybe a nice double.


Anonymous said...

Check the weight of the average participant in the Vivus trial. Then check the diet and exercise progran for those participants.

Check the same for the Arena trial.

The differnces in the weight of the average participant (the Vivus trial participants were substantially more overweight that the participants in the Arena trial) and the diet and exercise regimens (the Vivus regime was far more exacting) may explain why Vivus did so well.

Hard to make a head-to-head comparison of the two products until adjustments are made for these differences.

Anonymous said...

Isn't that factor taken into account by the placebo group?

Anonymous said...

In addition, the author compared completer mean weight loss for Qnexa with top-line ITT mean weight loss from Contrave. If you are going to compare two drugs across trials, at least present equivalent results. The completer data for Contrave was 8.4% (17.6 lbs); the ITT data from Qnexa was 10.4%.

Chris Morrison said...

Thanks for your comments. the purpose of this post is NOT to compare the three candidates' Phase III results, but if you want full results i've linked to all the company press releases in the text. but that said, sorry if i confused any ITT with completer responses etc.


Trista Morrison, BioWorld/BNET said...

Chris - interesting point. I wondered the same thing, and the Vivus folks told me the same thing they told your pals at Pink: docs won't risk combining off label drugs for obesity - not after the Phen-Fen combo disaster.

A couple other points they mentioned: low-dose Qnexa is as little as 1/16 and 1/8 of the available generic doses - so would be tough to split the generic pills. Also, Qnexa requires titration that would be tough to duplicate w/ generics.

I'm not sure I'm convinced, but we shall see...