Dangerous Liaisons

Alliances are a favorite of corporate strategists everywhere. More than 10,000 interfirm collaborations were formed worldwide in 2000, double the number of five years before. Alliances now generate 25 percent of the top 1,000 public US companies' revenues, up from 7 percent in 1990.

Pharma and biotech companies are particularly prone to partnering. The 20 biggest pharmaceutical firms formed nearly 1,500 alliances with biotech companies between 1997 and 2002. According to Strategic Decisions Group, a consulting firm, "40 50 percent of products in development by global pharmaceutical firms are externally sourced," and "more than half of the current 20 best-selling prescriptions drugs are co-developed, co-marketed, or in-licensed." By 2007, in-licensing alone is expected to account for 40 percent of the revenues of pharma's top 20 companies.

What's driving biotech and pharma into each other's arms is the need to find complementary resources pharma's innovation gap meets biotech's funding crisis plus the desire to share risk and access new markets and information. Given these trends, it seems there are two kinds of pharma executives today those participating in biotech alliances and those who will be soon.

What is the reality of alliance life? Two of the most celebrated biotech pharma alliances in recent times can serve as an introduction. One, a phenomenally productive venture, scientifically and commercially, has degenerated into a bitter squabble over the spoils. The other, just formed, bears a striking resemblance to the most successful biotech pharma hookup of all time.

Let's Make an Ordeal

In 1993, Knoll Pharmaceuticals, a divison of BASF, the German chemical company, called on UK-based Cambridge Antibody Technology (CAT). Knoll had six autoimmune disease targets. CAT had an advanced phage display technology ideal for identifying high-quality, lead-optimized candidates for human antibody drugs. CAT also had a vast library of antibodies (now more than 100 billion, according to the company). They were a perfect match.

The two signed an agreement (updated in 1995) that committed Knoll to pay up-front research fees to CAT and milestones along the way. Knoll would license promising products, assume their development and commercialization, and pay CAT royalties on any resulting sales.

Amazingly, the first target they selected, tumor necrosis factor-alpha (TNFa), was successfully neutralized. The partnership was off to a promsing start, then, in 2001, BASF sold Knoll to Abbott Laboratories for $7.1 billion. The key asset was the TNFa candidate developed by CAT called D2E7.

Royalty or Loyalty

In 2003, D2E7 was approved for the treatment of rheumatoid arthritis in the United States and Europe as Humira (adalimumab). That year, its first on the market, sales hit $280 milllion. This year they are expected to exceed $700 million, then rise to nearly $1.2 billion in 2005.

It took 10 years but the alliance hit a home run in its first at bat. Humira, Abbott claims, is "one of a few new potential blockbuster drugs to emerge from a collaboration between biotech and Big Pharma." CAT became not only the first British biotech to produce a potential blockbuster but also one of the few European biotechs to get anything to market.

In March 2003, with royalties due but not yet paid, Abbott told CAT it intended to cut the royalty rate from what analysts estimate to be between 5 and 6 percent to the contractual minimum, perhaps 2 percent. Multiply the difference by strong sales over a long time and it amounts to hundreds of millions of dollars at net present value. (CAT is not the only one affected. It owes a portion of its royalties to the Medical Research Council, the UK equivalent of NIH, the originators of phage display technology. MRC, in turn, owes something to the Scripps Institute.)

The timing of Abbott's pronouncement couldn't have been worse for CAT, which had made an all-stock bid to acquire Oxford Glycosciences (OGS). Investors responded to the news by knocking down CAT's shares, and OGS went to Celltech instead.

CAT, of course, disputed Abbott's decision. The parties first held private talks then engaged in four mandated mediation sessions, all of which proved fruitless. In November 2003, CAT filed legal proceedings in the UK High Court. A trial before a specialist judge in the patent division is likely unless a settlement is reached within 7 13 months.

Contract Killers

The source of the dispute in this uncommonly successful collaboration is two overlapping and possibly conflicting contractual clauses, Articles 5 and 12. Both relate to royalty offsets, a common element of biotech pharma licenses that offers protection to licensors of intellectual property (IP).

To get a drug like Humira to market, you need three pieces of IP: one for the target, one for the product, and one for the downstream manufacturing process. Knoll owned the target IP, they came to CAT for the product IP, and they licensed the manufacturing IP from Genentech.

A company's claim to own intellectual property may not free and clear. Sign a license with one party and you may have to deal with others claiming their IP rights have been infringed. If that happens, the offset clause kicks in, letting you cut payments due the initial claimant in order to pay the rest.

Knoll had more than the usual reasons to request an offset (or anti-stacking) provision because CAT's patents were still pending when the contract was signed. Even Peter Chambre, CAT's CEO since 2002, admits: "It was quite natural for Knoll to seek some protection in the event that it had to take multiple licenses to operate the phage display technology described in our patents."

This provision, Article 5, is fairly standard stuff. Translated from the legalese, it says that if Knoll needs to pay anyone else (other than MRC) to use CAT's technology, those payments will come out of CAT's end.

Back to the Future

CAT, in any event, got its patents. Knoll and, for that matter, Abbott never had reason to pay anyone else for CAT's IP. So CAT had every reason to expect a full royalty. But it seems that Abbott is reading another clause, Article 12, to mean any patent royalties paid for any purpose even those on the target (owed to Serono and Peptech) or the manufacturing process and not just the IP licensed from CAT, could be used to offset payments to CAT.

Article 12, available from an SEC filing, says: "In the event that Knoll or its Collaborators must pay royalties or license fees to third parties under one or more claims of one or more patents to enable Knoll or its Collaborators to utilize or have utilized the inventions of the Patents, [content omitted] shall be credited to Knoll's royalty obligation to CAT for sales in that country where the patents exist." But it then goes on to state: "This offset shall not include royalties or license fees which are beyond the scope of the technology described in the Patents, for example fees paid to third parties for delivery systems."

The same language applies to the other five targets Knoll brought to CAT as well as the five additional indications for Humira now in late-stage clinical trials. Presumably, any resulting products will be subject to similar dispute. There's a lot to quarrel over. Abbott estimates Humira, all told, has a market potential of $14 billion by 2010.

In a statement, Abbott says, "The dispute is related to the proper interpretation of the royalty offset provisions of the agreement under which Humira was developed," but makes no attempt to square that interpretation with what the contract says. Meanwhile, Chambre says, "We strongly believe that the offset provisions were not designed for the purpose to which Abbott is putting them. We are seeking an outcome consistent with that, and are determined to get it."

Try a Little Tender

The second case history opened just a few weeks ago, when GlaxoSmithKline and privately held Bay Area biotech Theravance announced "an innovative broad-based strategic alliance" hailed by some experts. "A very, very, very smart deal," said Edward Saltzman, president of Defined Health, a biopharmaceutical consulting firm that specializes in business development strategy, including helping biotech clients find the right partner. Fintan Walton, PhD, chief executive of PharmaVentures, which also helps biotechs do deals, described it as "a good example of a next-generation deal."

Why such praise? GSK paid $129 million to up its stake in Theravance from 6 to 19 percent. GSK also obtained an exclusive option to license Theravance's product portfolio through August 2007, with milestone payments worth potentially between $162 million and $240 million per program. If GSK winds up owning more than 50 percent of Theravance, its licensing option will be extended five years.

So far, nothing special. But GSK's investment also paid for a call option permitting it to buy up to 50 percent of the shares outstanding in 2007. What's more, a put option was established giving Theravance shareholders (other than GSK) the right to sell half of their shares to GSK, also in 2007.

In an interview, Theravance CEO Rick Winningham pointed out that "a floor valuation of $525 million for the 50 percent of non-GSK stock means GSK has put $654 million on the table."

He also stressed that "Theravance remains, both in the short and the long term, an independent company, in charge of all decisions related to a program prior to the point at which GSK might opt in."

He would not comment on plans for an IPO but admitted that "the transaction certainly creates an attractive security for current and future shareholders of Theravance," adding that "since it pertains to only 50 percent of the shares," the call option "provides unlimited long-term potential for our shareholders."

Winningham described some possible scenarios: "The price of the call, which has been defined but not disclosed, is at a multiple a premium of the put. In the event that our stock were a public security and trading below the put price, owners may decide to sell to GSK at the put price. If it were trading above the put price, they would probably decide not to sell. If the stock were trading above the call price, then it may make sense for GSK and it is completely their decision to exercise the call and buy some 50 percent of the outstanding stock. If the put or call were exercised, GSK would own roughly 60 percent of Theravance in 2007 own but not control. They would have a minority of board seats."

Theravance's only prior collaboration, also with GSK, was similarly unusual, and may shed light on the nature of their relationship. David Brinkley, Theravance's senior vice-president, commercial development, explains that it was "structured as a pooling of assets." Each company "put two compounds into a pool. Those compounds then moved forward based on their merits not origins. Theravance receives the same royalty rate on sales of any compound or medicine that results, even if GSK discovered and developed it." The same holds for GSK as well. "What's extraordinary about that," Brinkley says, "is GSK's willingness to say great compounds, great ideas, great approaches can come from anywhere. They don't have a monopoly, even in an area where they have a certain amount of expertise, such as beta agonists. This latest deal is just a natural step in the evolution" of this philosophy. He says, "We've heard GSK refer to it as 'increasing the genetic diversity of our discovery effort.'"

Winningham predicts that "transactions like this, that leverage the strengths of small companies, are going to continue to be done with large pharma, probably at a greater rate, because large pharmaceutical companies require an incredible level of innovation to maintain their financial strength and success in the marketplace."

The Heart of Alliance

What is an alliance? Professor Gomes-Casseres, author of Mastering Alliance Strategy, says: "It's a way to combine the capabilities of companies that gives added value, but in a way that is not transactional and also not fully merged." Companies collaborate when a simple exchange I buy, you sell won't answer their needs, and neither will buying an entire company. He adds: "An alliance is a longer relationship where you need to manage joint decision making. You can't set everything down in the beginning and just execute against it. You have to continually come up with answers for problems that arise that you didn't foresee at the beginning. It's a moving target."

Lawyers have a way of talking about this, also used in economics: an incomplete contract. "A complete contract," Gomes-Casseres says, "sets down in explicit terms every contingency, every state of the world like your mortgage. There is nothing you can litigate. An alliance is not like that because you deal with an uncertain project, things that cannot be predicted. So what we do is create a kind of contract which is not complete but which allows us to go with the target and answer new questions as they arise and make joint decisions when new problems arise. You must create a decision making structure for your relatively open-ended relationship, what is called a relational contract. You contract on the fact that you're going to have a relationship to resolve issues. You don't necessarily contract on all the answers to the issues because you don't even know what the questions are yet."

We Are Family

This open-endedness may be why, as Stuart Kliman, founder of Vantage Partners, an offshoot of the Harvard Negotiation Project, says, "Folks use the marriage or family metaphor for alliances all the time." An alliance, like a marriage, he says, "is a complex, interdependent relationship." According to Professor Mike Peng, "Formal regulations and contracts can only govern a small (although important) portion of alliance/network behavior," much as in "the institutions governing human marriages." In the end, "the success and failure of such relationships depend, to a large degree," Peng says, "on the day-in-day-out interaction between partners influenced by informal norms and cognitions."

The flip side of trust is risk. A good nontechnical definition: whatever can go wrong. And, as Murphy observed, whatever can go wrong will (O'Toole's Commentary on Murphy's Law is: "Murphy was an optimist"). Alliances thrive on trust while limiting the cost of risk.

Why did Saltzman call GSK Theravance smart? "Because it attempts to re-create in microcosm what is unquestionably the most successful strategic alliance, the one between Roche and Genentech." What do they have in common? Both used (private, nontradeable) financial options that gave both parties the right, but not the obligation, to buy (or sell) equity at a set price within a fixed time. (See "Back to the Future.")

A real option is the application of financial options to actual investment decisions. Real option reasoning encourages the notion that decisions are not all-at-once or all-or-nothing. For a relatively small initial outlay, a financial option gives you the right to buy or sell something later. It lets you sequentially increase your stake as you learn more about the risks and benefits. It's like poker not everything is wagered on one hand. You get more cards (more information) and make your decision to call, raise, or fold accordingly. That's also how alliances work. "Every deal," says Michael McCully, director of consulting at biotech alliance experts Recombinant Capital, "is basically structured as an option. All have some sort of termination component, provisions that state how each party can get out of them." Companies enter alliances because they can get out of them. Impermanence entails flexibility.

The open-endednesss and complexity of alliance life make it more difficult to grade the two alliances discused here than one might think.

Is Abbott CAT a failure? Scientifically and commercially, no. CAT might yet win its suit and Abbott, should it triumph, may gain only a Pyrrhic victory. It could win the royalty battle but lose the reputational war. Indeed, a survey of 17 biotech CEOs conducted by Cambridge Healthcare and Biotech, unrelated to CAT, Abbott was rated the least desirable pharma licensing partner of the 20 named. (Bristol-Myers Squibb was tops.)

Is GSK Theravance a sure thing? Decidedly not.

The uncertainties and ambiguities of alliances is precisely why such structures exist.

The only certainty is that alliances play a vital role in pharma today.