Is There Another Roche-Genentech Out There? Signposts for the Rare Beast

The emotional dust raised by Roche’s takeover bid for Genentech has largely settled (you can download a free collection of our coverage of the event and some of its implications, here).

But there’s still plenty of strategic dust obscuring the view. The Genentech/Roche model – the most admired relationship in industry history -- seemed almost uniquely able to solve two enormous challenges: biotech’s ability to access capital and Pharma’s to feed its pipeline.

We don’t believe that Roche's move to acquire its junior partner means the structure can't work elsewhere. Roche simply figured Genentech’s run of innovation was close to the an end, and that the deal’s cost of accessing further innovation (royalties, milestones, geographic limiations) was simply not worth the candle.

But the fact is that the model has been rarely attempted. Here’s why. Thanks to a unique confluence of partnering requirements, Roche and Genentech both got something they needed from this deal without insisting on getting more (until Roche did by announcing its bid for Genentech). That set of circumstances has rarely occurred in the past -- and will rarely do so in the future.

To review these circumstances: Genentech wanted to fund R&D at what it believed was the requisite level – but couldn’t without killing its share price and thus closing off its access to capital. Not so rare a situation, by any means.

So it convinced Roche to offer Genentech stockholders a generous put on their shares: if the stock began trading too low, Genentech shareholders could force Roche to buy them out. That meant that Genentech’s shares could only fall so far while its competitors in the game of attracting capital could make no such reassuring promise. Many had to – still have to -- overdose on dilution.

But that put cost Genentech something few biotechs would be willing to pay today: a fixed-price option for ex-US rights to its pipeline. The option, which secured Roche some of the most successful biotech drugs, was based on deal prices from the mid-90s. Now, no one could really forecast the extraordinary inflation in deal values consequent upon the equally extraordinary lack of R&D productivity in Big Pharma (not excepting Roche). But it’s hard to believe that anyone today, knowing now what we’ve learned about biotech dealmaking over the past decade, would agree to that kind of fixed-priced option.

The option didn’t turn out to be all that bad for Genentech either. As we could not forecast inflation in deal prices, we could not have forecast the ability to regularly charge $50,000 for a course of therapy applicable to hundreds of thousands of people. Genentech, in short, limited itself to the US market at precisely the moment the US market was at its most lucrative for the kinds of products Genentech was selling.

And one other thing: Roche was getting something it could not develop on its own or easily find on the outside: large-molecule discovery, development, production and marketing capabilities. Roche couldn’t duplicate Genentech’s pipeline. That means that there were relatively few entrenched interests at Roche who would see Genentech as duplicating their own efforts -- and therefore competition. Moreover, because Genentech had actually created three approved products (and developed and sold two of them – TPA and human growth hormone; Lilly developed and sold the third, human insulin), Genentech’s pipeline didn’t look like a pipe dream.

So: a rare set of circumstances. There are several companies, like Genentech, with a platform for producing a new kind of drug. But few of them would be willing to sell a fixed-price option to their entire pipeline. The pressures on US pricing (compounded by the likely advent of biosimilars) will force companies to be stingier in parceling out ex-US rights all at one go. Moreover, there aren’t many companies like the Genentech of 1995, which had successfully developed and commercialized products -- but which still needed help in accessing capital. Celgene, for example, certainly doesn’t need Big Pharma’s guarantee to wring money out of Wall Street; nor does Genzyme; nor does Gilead.

The two closest recent examples of a Genentech/Roche-like model – the Idenix/Novartis transaction from 2003 or the Theravance/GlaxoSmithKline deal of 2004 – don’t exactly shine as examples of successful development organizations. Moreover, neither of those companies provide their would-be acquirors any special new platform. Theravance, in fact, is all about me-better small-molecule drugs. And the fact that GSK chose its own long-acting beta agonist to work on, rather than Theravance’s, does at least call into question the acquisition value of Theravance’s platform – indeed, last year GSK formally decided not to acquire the additional Theravance shares it could have.

So, granted the rarity of candidates, who might fit? Which companies with unique or at least important R&D platforms have produced an approved but only marginally successful drug … and who thus would be willing to pay a price similar to the one Genentech paid in order to access capital at a reasonable rate?

RNAi platforms are certainly hot. But none of the independent RNAi players (like Alnylam or Silence) has actually developed a drug let alone proven that the platform actually can produce them. Too early for the 60% solution.

Aptamers haven’t generated huge partnering buzz but there’s at least one on the market (Macugen, from Eyetech/Pfizer) so Archemix, granted it can push its pipeline along, might be a good candidate for a 60% deal (the VCs in that company which have been unable to get it public would be happy with such an outcome).

Isis Pharmaceuticals’ antisense platform has attracted various partners; it’s got a late-stage program in development, mipomersen, that it pushed through on its own. But the market likes Isis now; the stock has dramatically outperformed the biotech index. So if we were CEO Stan Crooke we might insist on a bit more valuation than most Big Pharmas would be willing to tolerate – particularly since the company already sold Genzyme the rights to mipomersen. So: better candidate than Alnylam (would Genzyme’s Termeer do a 60% deal for Isis? We think he’d consider it) but not so good, or as inexpensive, as Archemix.

Exelixis has managed to put together a pretty fair set of targets and chemistries; has gotten a few products into pivotal trials, albeit none further than that. But it sure hasn’t secured the unalloyed confidence of investors. That combination of scientific quality and market skepticism has made Exelixis attractive to, and attracted by, innovative financiers like Symphony Capital and Deerfield – and which probably would attract them to a 60% acquirer, were one to come along. But it’s not clear to us that an acquirer would be as interested: small molecules are Pharma's bread and butter, too close to what they think they already know.

And a few ideas a bit further afield.

Big Pharma is increasingly interested in generics, particularly of the large-molecule variety. Might make sense for investor-poor, technology-rich Momenta to stay quasi-independent but let a Big Pharma sell its biosimilars as part of a push into emerging markets.

And then there's China; could Pharma tap into that increasingly attractive and rapidly transforming market by teaming with a service play like WuXi Pharmatech or a still-small and home-grown biotech play like Hutchison China-Meditech's recently emancipated MediPharma subsidiary?

So another Roche/Genentech is possible. Likely? Not so sure. Because there was one other element to that we haven't discussed: strategic courage. And that may be the rarest element of all.

Image from flickr user Steffe used under a creative commons license.