Partnerships Prevail at Roche

Last month, Roche established an agreement with a compatriot Swiss firm, Glycart Biotechnology, for using Glycart's GlycoMab technology to develop next-generation therapeutic antibodies.

It's a small deal, as deals go. But it's an indication of Roche's determination to extend its reach into areas of innovation through partnerships. The company even has a separate team, Roche Pharma Partnering, devoted exclusively to finding, negotiating, and nurturing the Swiss giant's partnerships with biotech minnows.

"In the first quarter of this year, 49% of Roche's sales came from biotechnology products," explains Peter Hug, the firm's global head of pharma partnering. "To maintain and expand our leadership in the antibody market and in disease areas such as oncology, Roche will continue to partner with excellent antibody companies such as Glycart."

To focus on its partnering philosophy, the company held a media briefing last month at its headquarters in Basel. Roche is a partner to more than 50 companies worldwide. Of its current partners, 25 are based in the U.S.; 11 in the U.K.; four each in Japan and Switzerland; and the rest sprinkled through Canada, Denmark, France, Germany, Iceland, Italy, and Sweden, according to Jonathan K. C. Knowles, board director and head of global research at Roche.

The company's total number of partnership deals in 2003 topped 40, compared with Johnson & Johnson, Novartis, and Merck, each with fewer than 35, according to Knowles. "We are working harder to create new partnerships than our competitors are," he contended.

At the briefing, Hug set the stage. "In Europe 10 years ago, there were 90 biotech companies. Today, there are more than 2,000. In this sea of opportunities, how can Roche ensure we don't miss the boat in finding the answers to unmet medical needs?"

One way, he argued, is through the company's collaborative approach to partnering, which emphasizes that any deals must make sense for both sides. And "chemistry"--the chemistry between people and teams--is particularly important in creating long-term relationships.

"Pharma Partnering is a unique group--the single point of contact within Roche for partners," Hug said. The unit begins as a "finder" that searches for the best opportunities. It also includes business analysts and information managers to analyze the value of opportunities and the potential cost of development; negotiators to design the deals; lawyers responsible for contracts; and, perhaps most important, alliance directors who manage collaborations throughout their duration.

"In 2003, we started with 1,683 new opportunities--they come in from everywhere," Hug noted. The company completed 43 deals in 2003, valued at roughly $280 million. "This year," he said, "won't be dramatically less or dramatically more." Moreover, Roche has not abandoned any partnerships since 2001. "We have restructured two, but abandoned none," he emphasized.

FOR HUG, the prime example of Roche's dealings with long-term partnerships is Genentech. Roche made its initial investment in Genentech in 1990 and currently owns a 55.6% stake in the biotech firm. However, Hug said, "if you walk into Genentech, you don't feel you are walking into Roche. In 1990, Roche was looked at as a strange company to invest in biotechnology. Whoever thought of proteins as blockbusters? Now Genentech is the leading biotech success, with 13 protein-based products on the market. And 10 projects in the Roche pipeline are the result of the partnership with Genentech."

Similarly, in Japan, Roche holds a 50.5% stake in Chugai Pharmaceutical. With the merger of Nippon Roche and Chugai in 2002, Chugai moved from about number 10 to number five in Japan, the second largest pharmaceuticals market in the world. Today, Chugai is number four, Hug pointed out. Another 10 projects in Roche's pipeline come from this partnership.

It's all well and good for a big pharma company to thump its chest and boast about how splendid a partner it is. It's another thing, though, when biotech partners agree.

Two executives from partner companies showed up at the briefing to do just that: Gardiner F. H. Smith, vice president of business development at Montvale, N.J.-based Memory Pharmaceuticals, and Glyn O. Edwards, chief executive officer of London's Antisoma.

According to Smith, Memory has two major collaborations with Roche. The original "traditional" deal, directed at Alzheimer's disease and depression, was done very quickly in the summer of 2002. There was due diligence in late June, and the agreement was signed and announced in early August, he said.

A year later, Roche and Memory established what Smith calls "a true partnership" in its NicA7 compounds development program, targeting schizophrenia. But Memory had to overcome some internal problems, he noted.

"Having known each other quite a bit, Roche was interested in this program," Smith recalled. "Our dilemma was in our Memory board of directors, drawn from venture-capital firms. They were worried about our having all our eggs in one basket." Together with Roche, the Memory team worked out a deal to resolve those concerns: Roche made a minority equity investment and gave Memory responsibility to take compounds from proof-of-concept through Phase IIA. Roche maintained an option on developing downstream rights.

"This was a tailored, customized deal structure. We were delighted to have this," Smith said. The deal, he added, preserves Memory's entrepreneurial culture and capabilities and, most important, maintains its operational independence. One important benefit from the partnership, he noted, is that Roche has created an area on its information technology network that is accessible to Memory.

Pharmaceuticals make up nearly three-quarters of Roche's sales. But its remaining operations--diagnostics--are also open to partnering, pointed out Thomas Metcalfe, head of Roche's biomarker program.

"We aren't able to cover all the opportunities. So we work together with partners," he said. The company is looking not so much for nascent technology as for "relatively robust technology with established proof of principle. We are really interested in looking at people who have done work in validated diagnostic markers."

Roche Diagnostics' recent investments into technology and content date back to 2001, when it established deals with deCode Genetics and Partners Healthcare System in markers for complex disease. In 2002, it added partnerships with Epigenomics in the discovery of markers for cancer and Quest Diagnostics in new-marker discovery and validation. And last year, it reached an agreement with Affymetrix in high-density microarray technology.

At the briefing, Hug predicted that Roche's deal-making will continue at its current rate. The emphasis may be shifting, however, into earlier phase work.

The company is flexible on equity investments in potential partners. "If we have full confidence in a company, it makes sense to invest," Knowles pointed out. "If the company is successful, we will benefit. A majority stake--usually to add value to all the partners--is one route. It's not done to keep someone else out, although we could consider that if it made sense to the company."

The risk to an investor company such as Roche "is inherent in the projects, not the partnerships," Hug noted. A partnership project has the same success rate as an internal project. Historically, Phase I projects had a 14% success rate, but now that is down to 8%; Phase III projects once had a 20% success rate, but that is now 15%, he said, citing industrywide data from the Food & Drug Administration.

And none of the Roche executives at the briefing were pleased about the current depressed financial state of the biotech industry. Rather than seeing the situation as offering "fire sale" opportunities for investment, Knowles noted that "our model is risk sharing. If there is nobody there to share the risk with, it doesn't help us."