The Large Molecule Future

In an extraordinary series of deals, Pharma has embraced early-stage large-molecule technology, reflecting both the surging value of biologics and the severity of pipeline anemia. But many observers are skeptical the less experienced Big Pharmas can buy themselves into a brand new business with very different requirements. For start-ups, however, the news is very good--they now have two clear and viable pathways to creating shareholder value: acquisition and alliance.

Big Pharma's extraordinary series of biologics deals reflects both the surging value of large molecules and the severity of pipeline anemia. Meanwhile start ups are—delightedly—revaluing their term sheets.

By Roger Longman

• With more than $2 billion in M&A deals in barely as many months, Big Pharma has embraced early-stage large-molecule technology.

• Many observers are skeptical the less experienced Big Pharmas can merely buy themselves into a brand new business with very different R&D, process development, manufacturing and marketing requirements.

• But all companies, no matter how experienced, see dealmaking as fundamental to

• their ability to build biologics operations.

• New biologics companies now have two clear and viable pathways to creating shareholder value: acquisition and alliance.

"Amazing," says Steve Projan, PhD, Wyeth 's VP and head of biological technologies. "Spectacular," adds his colleague Cavan Redmond, EVP and general manager, biopharmaceuticals, "are we all heading in a brand new direction?" "People have certainly gotten religion," says Tom Bumol, PhD, VP, biotechnology, discovery research for Eli Lilly & Co.

They're talking about Big Pharma's extraordinary string of biologics-company acquisitions—nearly $2.3 billion worth of deals. Pfizer Inc. , Merck & Co. Inc. , AstraZeneca PLC and Roche have all purchased large-molecule companies in the past three months. Add to that total another several hundred million dollars worth of product-based alliances, by those companies and others, and you have all the makings of a feeding frenzy. (See Exhibit 1.)

Two decades after the first antibody made it to the market—Johnson & Johnson 's transplant-rejection drug muromonab-CD3 (Orthoclone OKT3)—biologics, and especially antibodies, are the hottest class of molecules, bar none, in the industry pipeline. According to a report from Datamonitor, biologics will have accounted for 60% of absolute annual sales growth in Big Pharma from 2004 through 2010—double the growth from small molecules (vaccines make up the rest) (See Exhibit 2.)

And that's just from the biologics Datamonitor can see. The business-information company's total sales growth estimate over the six-year period--$45.8 billion, or 3.2% CAGR--is pretty anemic. And in its biggest new sales driver, Big Pharma is way behind the technology curve. Hence the acquisitions, which are designed to transform Big Pharmas into real biologics players. With the purchase of Abmaxis Inc. [See Deal] and GlycoFi Inc. [See Deal], Merck "now has what we need to go from discovery to market," says Mervyn Turner, PhD, SVP, worldwide licensing & external research. Likewise AstraZeneca: With Cambridge Antibody Technology Group PLC (CAT) "at its center, we have the right people and the right technologies to build a world class biologicals capability within AstraZeneca," John Patterson, executive director, development, told investors June 8 [See Deal]. He predicts that by 2010 a quarter of AZ's products in full development will be biologics. To get there, he says, AZ will, each year, put four-to-five new biologics into full development—beginning in 2007. For comparison's sake, that's twice the goal that Wyeth, a much more experienced biologics company, has publicly set for itself.

There's plenty of skepticism about the abilities of the less experienced biologics players among Big Pharma companies to merely buy themselves into what is definitely a brand new business—one in which chemistry, synthetic manufacturing, and primary-care marketing and sales, the traditional strengths of the industry, will be of minimal value. But all companies, no matter how experienced, see dealmaking as fundamental to their ability to build biologics operations. Merck, an antibody and proteins novice, is buying optimization and expression technologies that will allow it to exploit its existing target-discovery and vaccine bioprocessing assets. Wyeth, with plenty of protein infrastructure and process development expertise, is in-licensing products to which it can apply those specific capabilities.

Other new entrants to large molecules, like Pfizer, AstraZeneca, and Roche—outside of its relationship with Genentech Inc. —are all pursuing their own dealmaking strategies, each designed to allow it to play catch-up in a game hitherto left to others. Novartis AG is pursuing the most complicated set of strategies: leading the charge into biogenerics through its Sandoz division, and creating a large-scale discovery effort in its Cambridge research facility, built around a series of platform deals and a handful of licensing arrangements. Indeed, 25% of Novartis' preclinical pipeline is made up of biologics.

And meanwhile, the large-molecule experts aren't ignoring the new technologies, or even the old ones. Amgen Inc. bought Abgenix Inc. largely to get hold of the 50% of the Phase III anti-cancer drug panitumumab that it didn't already own, but it also got with it Abgenix's XenoMouse human antibody technology. [See Deal] (See "Amgen/ Abgenix: All About a Drug," IN VIVO, January 2005 (Also see "Amgen/Abgenix: All About a Drug" - In Vivo, 1 Jan, 2006.).) Genentech, too, has done a number of large-molecule technology deals, including alliances with BioWa Inc. , a unit of Kyowa Hakko Kogyo Co. Ltd. , and Xencor Inc. (See also: "The Next-Generation Antibody Players, February 2006 (Also see "The Next-Generation Antibody Players " - Scrip, 1 Feb, 2006.) and "Glycobiology's Second Wave," May 2005 (Also see "Glycobiology's Second Wave" - Scrip, 1 May, 2005.), both in START-UP.)

All of which is very good news indeed for large-molecule start-ups. Unlike many other residents of biotech country, biologics companies have two clear and viable pathways to creating shareholder value: acquisition and alliance. On the one hand, the biologics platform technologies can interest multiple buyers—unlike most platform plays in the industry. And simply because it acquires one platform, a Big Pharma will not be uninterested in others: it's already clear that no one optimization or manufacturing technology will work for the increasingly varied world of biologics—antibodies, fusion proteins, antibody fragments, synthetic proteins, synthetic protein fragments and a variety of other large molecules decreasingly unlike anything Nature ever conceived. Notes Genentech VP of business development Joe McCracken: "Different technologies are required to design the best antibody for different diseases and targets." Genentech did deals with BioWa and Xencor in part, he says, because they need a toolbox of optimization and expression technologies. "One size will not fit all," he says.

On the other hand, companies will pay significant amounts for large-molecule products which have reached some clinical proof of concept—preparing the way for an IPO, as Wyeth's deal may do for Trubion Pharmaceuticals Inc. [See Deal], which has filed to go public, or even a significant acquisition—like Pfizer's $500 million purchase of Rinat Neuroscience Corp. [See Deal]

Fear and Greed

It's not so much that large molecule technologies have caught up to Big Pharma's expectations as Big Pharma's needs have caught up to large molecules.

No one questions any more the market value of large-molecule drugs. Bevacizumab (Avastin) is on track, most analysts believe, to be a $5 billion drug; two other Genentech drugs have gone beyond the billion-dollar mark, driving that company's market cap to #8 in the pharmaceutical industry, all without any direct sales outside of the US. Three anti-TNF products—two antibodies and a fusion protein—sell a combined $8.9 billion. Indeed, had Wyeth not acquired American Cyanamid Co. [See Deal], getting with it a share of the US profits in the TNF inhibitor etanercept (Enbrel) [See Deal], it likely wouldn't exist as an independent company. J&J's most important product is the protein erythropoietin alpha (Procrit)—which itself, with $3.3 billion in 2005 sales, is merely the number two player in the EPO market behind Amgen Inc. 's sustained-release version of the protein darbepoetin (Aranesp).

And companies have built these drugs with far smaller sales forces than are required by small-molecule primary-care drugs. Indeed, it is in primary care where marketing productivity has slipped fastest, as average number of details per sales rep, and average time per detail, has slipped dramatically—by 50% or more in the 1995-2003 period, according to Bain & Co.

Meanwhile, most Big Pharmas have begun to accept an increasingly conventional bit of wisdom: small molecules fail in clinical trails at greater rates than large molecules—in getting from Phase II to the market, at roughly twice the rate of large molecules, according to data from Booz Allen Hamilton and CS First Boston. (See "Know Thy R&D Enemy: The Key to Fighting Attrition," IN VIVO, January 2005 (Also see "Know Thy R&D Enemy: The Key to Fighting Attrition" - In Vivo, 1 Jan, 2005.).) "People are looking for faster, cheaper and safer ways to innovate," notes Lilly's Tom Bumol. "What's important is getting drugs into humans—and you can do that faster with biologicals since you don't get all those tox problems you get with small molecules." Antibodies are especially important. "It's the one proven technology," says Bumol. "Evolution gave us a great therapeutic design."

And if a disease is amenable to both kinds of molecules, it makes sense to attack it with both. That's why Novartis AG is going after hepatitis C both with small molecules, like Idenix Pharmaceutical Inc.'s oral small molecule polymerase inhibitor valopicitabine (NM283) [See Deal] [See Deal] and Anadys Pharmaceuticals Inc. 's TLR-9 oral prodrug [See Deal], and large ones, like Human Genome Sciences Inc. 's long-acting alpha interferon, Albuferon. [See Deal]

But the immediate impetus for at least some of these transactions is "fear and greed," argues Jan Turek, CEO of Biolex Inc. , which is developing an antibody optimization and manufacturing technology. "Big Pharma is fearful that someone else may limit access to technologies and become greedy."

Amgen's acquisition of Abgenix, for example, threatened AstraZeneca's most important large-molecule discovery and development program—indeed, virtually its entire large-molecule strategy in cancer. AZ had signed up with Abgenix in October 2003, paying $100 million for its stock, in order to get access to the large-molecule discovery, clinical and process development, and manufacturing capabilities it needed [See Deal]. When Amgen bought Abgenix, notes one AZ executive, it was almost certain to lose the discovery portion of the deal—Amgen wouldn't renew it, he says. Nor, it might be supposed, would an Amgen-owned Abgenix have been energetically diligent in developing the collaboration's products through Phase II—one of its responsibilities in the deal—when it would have to share the profits with AZ.

And so AZ turned to its other large-molecule partner, CAT—paying a 72% premium to take it over. Notes Don Drakeman, the CEO of Abgenix-competitor Medarex Inc. : AZ needed to "make sure the same thing didn't happen twice."

The Manufacturing Opportunity

The idea of getting shut out of the large-molecule opportunity has driven many of the deals. Even Roche, which profits hugely in large molecules from its investment and licensing deal with Genentech, has apparently decided it needs its own large-molecule pipeline—one reason for its $182 million acquisition of GlycArt Biotechnology AG . [See Deal]

Getting shut out means, for many of these companies, getting shut out of the lower-cost manufacturing technologies—one of Roche's rationales for the GlycArt transaction. Producing biologics is expensive. Bristol-Myers Squibb Co. just agreed to spend $650 million to build a factory in Devens, MA for its rheumatoid arthritis fusion protein abatacept (Orencia). And while Massachusetts won the competition for the plant by agreeing to supplement Bristol's spending with $34 million in state funding for infrastructure improvements, it's hardly reducing Bristol's capital outlay by very much.

Moreover, many of the new drugs—like Orencia--will apparently require substantially larger doses than earlier protein therapeutics, and thus substantially larger manufacturing lots. Says Sherrill Neff, a partner at Quaker Bioventures and an investor in several protein companies: "EPO is a huge drug, but you still don't need to manufacture more than a couple dozen kilos. Now we're making the transition to far less potent molecules that will be used in greater volumes, where the supply requirements will be measured in metric tons." For many companies, he continues, developing a biologic is "like having a Lamborghini raring to go, but when you get to the clinical development stage you hit a Model T production line. Something's got to give to put the production on a more even footing."

"Given the amount of effort [many Big Pharmas are] spending on the discovery side," says Wyeth's Steven Projan, "there's going to be a bottleneck even in their ability to produce enough material for clinical trials." And figuring that problem out involves more than simply building pilot plants. Contrary to most of Wyeth's competitors, "our reading is that large and small molecules have the same failure rates in each phase of development," says Redmond. But whatever the failure rate, a clinical failure of a large molecule is more expensive than a small-molecule failure because the cost to make the clinical material is far higher. Tom Bumol of Lilly figures the cost of enough GMP material for toxicology and early clinical trials at $4-5 million. Says Redmond: "That's why we put as much money into process development as we do into discovery and why we spend it at about the same time—to make sure we've got a production process that makes sense. By spending that money early, we'll bring down the risk of getting a drug we can't make economically."

Wyeth learned its lesson on Enbrel, the first of the TNF blockers to win approval. Unable to meet demand, Wyeth and partner Immunex scrambled to add capacity, not merely losing millions in sales but also giving a wide opening for second-to-market competitor infliximab (Remicade) to help fill the unsatisfied patient demand and thus establish itself in the market. Likewise, Schering AG opened up the market for Biogen Idec Inc. 's interferon beta 1-a (Avonex) when, unable to make enough of its own interferon product, Betaseron, it instituted a product lottery—infuriating many patients who didn't want to wait for the drug.

And every biologics player is carefully watching the linked debates around pricing and biogenerics. The high prices of biologics are causing big problems for payors. They constitute the fastest-growing segment of insurers' pharmacy expenses and there seems to be relatively little they can do to mitigate the cost, other than to invest in more efficient specialty pharmacy distribution and management systems.

But the high prices are certainly hastening the arrival of biogenerics. Sandoz got the most publicity from the US approval in May of its biogeneric human growth hormone, Omnitrope. But in fact the FDA has already snuck through the system a number of biogeneric approvals in new-molecule disguise. For example, over the last several years, the Agency approved four different brands of the long-generic hyaluronidase, each on the basis of relatively little clinical-trial data. But now that Omnitrope has made it through the political minefield, the floodgates for other follow-ons are opening: Sandoz has already announced it will seek approval on half-a-dozen biogenerics.

Reducing Cost of Goods

If prices do begin to come down, reducing the cost-of-goods will be critical to creating viable, competitive products. That's a theme Wyeth is playing, loudly, in its strategy for in-licensing biologics in competition with far richer companies. To attract them, Wyeth is holding out the carrot of its production-optimization expertise and manufacturing facilities in Ireland and Andover, MA. "Our model," says Redmond, "is to optimize the in-licensed products enough so that we can use our facilities to make them," rather than build new facilities. "We want to trade stainless steel for science," he says.

Wyeth is betting it can use its process development and manufacturing assets to out-compete richer Big Pharmas for sought-after, clinical-stage products. And Wyeth contends the strategy worked in its first deal, with Trubion, which makes engineered biologics called SMIPs (Small Modular ImmunoPharmaceutical), roughly a third to half the size of normal antibodies. The deal is certainly rich, but not as rich as an almost directly comparable deal in which Pfizer acquired rights to Coley Pharmaceutical Group Inc. 's anti-cancer toll-like receptor 9 (TLR9) agonist aptamer ProMune. [See Deal]

In the Trubion deal, Wyeth garnered a Phase II anti-CD20 SMIP, in development for rheumatoid arthritis, lupus and an undisclosed niche indication; any further CD20 products Trubion comes up with; one further SMIP focused on a different target; and a series of time-limited options to work on other pre-specified targets. In return, Wyeth paid $40 million up front (Pfizer paid $50 million up front, getting access to just one product) and agreed to support an IPO by purchasing 20% of Trubion's offered shares (Pfizer agreed to purchase a flat $10 million worth of Coley's IPO). Trubion will get $250 million in regulatory and sales-based milestones for the Phase II program (Coley got $455 million, more than 90% of which were based on regulatory milestones).

Figuring the CD-20 SMIP and Coley's Promune are roughly comparable in the size of the markets they might serve—admittedly a huge assumption—and figuring a roughly comparable royalty (about 15% in the Coley/Pfizer case), Wyeth's return could be significantly higher than Pfizer's. Not that Trubion will do badly from the deal: given Wyeth's expertise, it is far more likely to see a product commercialized, faster, than with a company which doesn't have Wyeth's process development and manufacturing capabilities.

No Retrofitting

Merck also believes that getting an early handle on manufacturing costs will be fundamental to the success of its nascent biologics business. As with Roche and GlycArt, efficient production was one key reason Merck bought GlycoFi: its technology for inexpensively making large quantities of the right product. For antibodies to help kill the cells they target, they need a particular pattern of decorating sugars, or glycosylation. The problem is that the traditional manufacturing medium in which most antibodies are made--Chinese Hamster Ovary (CHO) cells—produces a mix of glycosylation patterns, or glycoforms. That mixture must be further purified to tease out the properly glycosylated, higher-potency variants, a process which can dramatically increase the cost of manufacturing.

There are now a number of technologies which deal with this problem to greater or lesser extent. Some companies—like GlycArt and BioWa--manipulate CHO cells to produce the higher-potency antibodies. A few others dispense entirely with CHO cells, like Biolex, which harnesses an aquatic plant called lemna to manufacture proteins.

GlycoFi, too, dispenses with CHO cells. Its optimization and manufacturing technology uses specially engineered yeast strains to help find and then produce a single high-potency glycoform of the desired antibody.

All this was quite attractive to its two partners, Lilly and Merck. Both, apparently, offered to acquire it. But Merck was willing to pay more, at least in part because it hadn't yet made the CHO cell manufacturing commitment. It could therefore use GlycoFi as a basis for a much larger operation without having to do as much re-fitting of existing facilities. In fact, it already understood yeast production methods pretty well: that's how the company makes two of its vaccines, Recombivax for heptatitis B and the newly approved human papillomavirus vaccine, Gardasil.

Still, Merck wants more from GlycoFi than cheaper drug. It also wants to more broadly exploit its own discovery infrastructure, and in particular its Rosetta target discovery and validation system, which it believes can deliver the right biological mechanisms for its large- and small-molecule efforts. GlycoFi's technology can help optimize any antibodies it raises to such targets. And Abmaxis will also help Merck on the discovery side: its in silico immunization (AISIM) technology optimizes existing antibodies.

But that still leaves a question in Ronald Pepin's mind. The Medarex SVP of business development wonders where Merck—and other acquirers—are going to get the actual molecules to optimize. Perhaps, he says, Merck has the ability to create the basic antibodies. But navigating the patent field is treacherous. The PDL BioPharma Inc. humanization patent, for example, remains in force till at least 2014. One advantage of Abmaxis—and Trubion, too--is that their technologies could get around the PDL patent. But then again—maybe not. "It's something we've looked at very carefully," notes Merck's SVP, external research, Merv Turner, non-committally.

Growing Corporate Understanding, Overnight

Meanwhile, many patent-savvy observers, notes Edwards Angell Palmer & Dodge attorney Kathleen Madden Williams, PhD, were surprised that Genentech was able to get its Cabilly patent extended from covering merely chimeric antibodies to covering antibodies in general, including fragments. And while most people, she says, figure the new patent won't hold up, Genentech is insisting that it will continue to prosecute it aggressively, hoping to convince companies to take a license as insurance. Granted, of course, that it agrees to provide a license, which on competitive targets it may not.

It's possible that some of the new technologies will avoid Cabilly, too. But at the end of the day, says Pepin, "the cost of these workarounds is probably higher than the cost of a license to Cabilly." Or indeed that of a deal with a discovery engine like Medarex (one of the few antibody discovery companies left independent) since the general perception is that Medarex's licensees—through a non-exclusive licensing deal with Genentech--won't also need a separate license to Cabilly. At least that's what appears to be true from the publicly available documents, according to patent attorneys Vern Norviel and Shirley Chen, PhD, of Wilson Sonsini Goodrich & Rosati.

Rarely, however, do companies make major acquisitions as mere patent workarounds. Along with the capabilities they bring, acquisitions are seen as a short-cut to experience. Notes Richard DiMarchi, PhD, the former group VP, biotechnology and product development at Lilly, "Acquisitions allow companies to bring competency into the organization. They want the experience and higher intellect in bioproducts. That's certainly true for how AZ looked at CAT. Or take the J&J acquisition of Centocor. Its view has shaped in large part how J&J views biomolecules."

But such a large-molecule understanding doesn't grow corporate-wide overnight, counters DiMarchi's former Lilly colleague, Bumol. "Yes, you can acquire capabilities, but it takes time to accumulate the expertise." Lilly, he says, has been serious about making a business in large molecules since 1997. "We've done 13 deals in large molecules since then. We've worked with lots of companies to find out what works."

Some of what works is process—expecting to parallel-track manufacturing development with discovery along with the recognition of the additional financial requirements for such programs. And part is an understanding of large-molecule patients. What companies often don't recognize, Bumol says—indeed, Lilly didn't for too long—was the necessity for "ergonomic devices." Large molecules will be a highly competitive space, he notes. "If you're competing with an easy-to-use, stable formulated product a patient can keep in a purse, while you're trying to sell a drug with a needle and syringe, you'll lose."

Experience may indeed be a higher-than-expected barrier to entry into the biologics business, says Genentech's Joe McCracken. "But we worry nonetheless." Up until just a few years ago, he notes, only a few companies were working seriously on VEGF—the target for Avastin: "Now there's not a company out there without an anti-angiogenesis program." Hoping that experience in large molecules will keep the early-movers ahead of the competition is foolhardy. "There are lots of ex-Genentech employees out there who can show you how it's done; there are lots of contract manufacturing groups. People will figure out how to do it. We don't feel comfortable for a moment that our experience in large molecules is going to continue to be a competitive advantage."

Anticipating Acquisitions

Meanwhile, investors grow increasingly excited about the possibility of acquisitions. That's at least part of the reason the stock of the antibody company Genmab AS nearly doubled once Amgen announced the purchase of Abgenix—and it's also a big reason why shares of Genmab's corporate founder, Medarex, have stayed flat: that company will be a much tougher acquisition to make, given the number of licensing deals it has outstanding. "If you buy us," says Medarex's Drakeman, "you'll have to supply [antibodies] to a lot of other companies."

Acquisition, of course, isn't an automatic option for every experienced and validated biologics company. Serono SA hung out a very public for-sale sign, but got no takers, at least at the price it required. (Novo Nordisk SA is a rumored merger partner for a 50-50 deal that would split control of the combined organization between Serono's Bertarelli family and Novo's major shareholder, the Novo Foundation.)

But no one thinks the dealmaking is anywhere close to finished. The speed with which GlycoFi was snapped up, and the $400 million price tag, should indicate the industry's appetite for large-molecule platforms. "We were all set to expose the technology to all the companies who were interested" by doing a variety of small deals, says Michael Ross, PhD, a partner at SV Life Sciences and one of the GlycoFi investors. But "we got taken off the table before we even got wound up."

Indeed, companies are far more willing to take a flyer. Contrast the GlycoFi deal with Lilly's acquisition in November 2003 of Applied Molecular Evolution Inc. [See Deal]. The two companies aren't exact matches: AME lacked a manufacturing technology. But it was also a far better established company. It had raised $130 million in equity by the time Lilly bought it (better than four times GlycoFi's total) and had signed nine partnerships to GlycoFi's two.

The End of the Validating Deal

Alliance or acquisition, the new large-molecule deals are having a big effect on start-up strategies as well on Big Pharma's. "More than anything," says Robert Connelly, the CEO of Domantis Ltd. , an antibody fragment company, the recent acquisitions and alliances require the company "to accelerate what we are already doing. Giving partners broad access isn't what we want to do anymore." The acquisitions, he says, have "forced us to raise overnight the price of our deals." In the midst of partnering discussions on the company's preclinical asthma program, Connelly has suddenly stepped back to re-evaluate. "It's true that you can only develop a certain number of things at one time, but you also don't want to give away access to a huge number of targets" just because you can't develop them yourself, now. That only drains value from any potential acquisition or future alliance.

One reason for doing its smaller-value, broader-access deals, like those Domantis signed with Bristol [See Deal] and Abbott Laboratories Inc. [See Deal], was validation for investors. But such validation deals, at a time when the financial markets apparently value biotech at a significant discount to what Big Pharma might pay for them or even their programs, are increasingly less important. Especially so in biologics.

"The investment community looks at these deals with a much more jaded eye than they did historically," says DiMarchi, a co-founder and chairman of Ambrx Inc. , which optimizes existing proteins using a kind of medicinal chemistry for large molecules. That's because the companies doing the purchasing "don't have the credibility they did a decade ago. Small companies are trying to mature their assets, making them as robust as they can, to do deals from positions of strength. There's also experience in the small companies—people at Ambrx, for instance, have already successfully conducted the development drill and are respected for what they've done." Big Pharma, by and large, hasn't—and thus its imprimatur isn't worth much.

But while validation deals might not be important to Wall Street, few Big Pharma deals get done without some sort of validation—usually a previous collaboration. The process development team at Wyeth, for example, knew the top Trubion scientists well. The biotech's SVP, R&D and chief medical officer had both worked on Enbrel with the Wyeth team that helped figure out how to make it—much the same team that did the due diligence on Trubion. Doing the deal "was like intermarriage," says Trubion's CEO, Peter Thompson, MD.

The acquisitions likewise depended on previous collaboration. GlycArt had already worked with Roche as had GlycoFi with both its bidders, Lilly and Merck. AstraZeneca and CAT had been working together in an inflammation collaboration since November 2004 [See Deal].

The challenge, says SV Life Sciences' Ross, is to give "partners a taste without letting them eat too much." Thus GlycoFi signed two relatively cheap but limited deals—enough, apparently, to assure both Merck and Lilly that the company's technology was real. The company had raised about $28.6 million since its founding in June 2000; not a lot, notes Ross. "We wanted to use non-equity cash to keep growing" so the deals they did—given that they didn't have much to sell—were quite small.

The contrasting strategy—the big deal and the IPO—requires a lot more investor cash. Trubion has raised twice as much as GlycoFi, $46 million, using the money to create two licensable programs—the Phase II TRU 015 that it sold to Wyeth and its preclinical anti-CD37 project, which it's retained. But the deal set up its IPO filing, which it hopes will raise an additional $86 million [See Deal], though few biotechs have realized their optimistic IPO projections. (See "The Price is Wrong: Finding Alternatives to Traditional IPOs," START-UP, June 2006 (Also see "The Price is Wrong: Finding Alternatives to Traditional IPOs" - Scrip, 1 Jun, 2006.).)

Given the interest in large molecules, it looks like both strategies can work. And for a biotech industry hungry to prove value in early-stage technology, the passion for large molecules is invigorating. Big Pharma is still skeptical, and cheap, when it comes to the platforms companies have been creating in target and small-molecule discovery—dealmaking there is still moribund. For most small-molecule discovery companies to net a major alliance or acquisition, they have to have plenty of clinical data.

But Big Pharma does want platforms in biologics. It's telling that Pfizer, now in the midst of a wave of layoffs, has kept Esperion Therapeutics Inc. 's large molecule program and people [See Deal], and plans to keep Rinat's. For the foreseeable future, large-molecule expertise may be the best full-employment insurance in the pharmaceutical industry.