By Randall Osborne


Talk about mixed messages.

Nobody's disputing that last year's boom in biotechnology is over, and the mad scramble to go public is but a memory. It was a stretch of time when some of the least fit companies sought to file for initial public offerings (IPOs) only to have underwriters turn them away.

That's over. By the end of 2000, the IPO house was as quiet as a transgenic mouse, and you could count the number of firms that have priced IPOs on one hand, if you had a couple of extra fingers. Many pulled out of the IPO game after registering.

The industry is all about genomics now, right? Especially, it's about those companies that have established themselves, and achieved a position from which to cash in on the much-ballyhooed map of the human genome. Right?

So it seems. Such names as CuraGen Inc., Incyte Genomics Inc. and Millennium Pharmaceuticals Inc. continue strong on Wall Street. In fact, last Monday, the main gainer in the sector was Human Genome Sciences Inc. (HGS), which closed 14 percent higher after disclosing the previous Friday that a new compound the first genomics-based small molecule has been shown to reduce levels of lipoprotein-associated phospholipase A2, an enzyme associated with formation of atherosclerotic plaques.

Called SB-435495, the compound emerged from Rockville, Md.-based HGS's collaboration with GlaxoSmithKline plc, of Brentford, UK, which said gene-based drugs such as those developed with HGS will be important to the firm's growth.

Then, along comes Lehman Brothers.

Actually, New York-based Lehman Brothers came ahead of the HGS news, issuing early in February "The Fruits of Genomics: Drug Pipelines Face Indigestion until the New Biology Ripens." Done with McKinsey & Co., the report is the kind of work that might make a short-term genomics investor, or even the middle-term investor, think twice.

Among other things, it says the high cost of developing drugs from much-celebrated genomics findings won't start to drop off until 2006, and it will be another decade before expenditures sink below where they are today.

Gene-based drugs take a long time, the report says. Longer than many folks are taking into account. And they cost, in a big way.

What's going on, then? Are more investors the longer-haul kind, or are they just not paying attention? Nathan Myhrvold, a newcomer to the industry, spoke at the Biotechnology Industry Organization (BIO) CEO & Investor Conference in New York last month. He predicted "exponential" growth as a result of genomics, noted BIO president Carl Feldbaum.

"His argument was surprising . . . not irrefutable, and in some ways persuasive," Feldbaum said, adding that, although optimistic, he would stop somewhat short of the forecast.

Myhrvold, president and co-founder of Intellectual Ventures LLC in Bellevue, Wash., formerly was chief technology officer at Microsoft Corp.

"He's a very forward thinker," Feldbaum told BioWorld Financial Watch. "He's had people say to him, 'You're late to the party,' and he said, 'Yes, that's what people said to me when I arrived at Microsoft, in 1986.'"

The Lehman-McKinsey report, however, suggests Microsoft-style leaps and bounds are hardly to be expected, and "exponential" is not the right word at all at least for the moment, and possibly not quite a few moments to come. One cause: poor target validation, with technology that isn't yet up to snuff.

"Perhaps the most surprising and compelling discovery of the work," the report says, "is that genomics threatens to increase not only the overall associated [research and development] costs, but also the average cost per new chemical entity (NCE) or drug."

Not contacted for the report, however, was William Haseltine, chairman and CEO of HGS. Haseltine told BioWorld Financial Watch he was less than impressed with the Lehman-McKinsey effort, although it's having an impact and has begun to appear in analysts' remarks.

Said Haseltine: "They got it absolutely wrong."

He called the report, which questions of the quality of genomics drug targets, "something a group of management consultants would do." The report, "based on interviews with [R&D] directors, none of whom had put a single drug into trials at that time," had a weak equation, in which "N" equaled zero, Haseltine said.

Compilers of the report said they interviewed more than 40 subjects, "including R&D executives of top pharmacos, executives of leading biotechs, academic leaders and patent lawyers." In any case, Haseltine said that, with one example which would be HGS' compound, partnered with GlaxoSmithKline "the central thesis of their argument fails."

One conclusion of the report was correct, in Haseltine's view: "It takes longer than the industry believed to bring a small-molecule drug to market. [But] that phenomenon has been going on independent of genomic targets. It's a reality of small-molecule drug discovery."

He said a chemical treatment will take 13 years to 15 years to develop, and a therapeutic protein or antibody treatment will take six years to seven years. The success rate is different, too, he said. One in 10 small-molecule drug targets results in an investigational new drug (IND) application, with one in 10 a success after that, Haseltine said. For therapeutic proteins and antibodies, out of 100 targets, 20 reach the IND stage, and there's a 50 percent success rate thereafter, he said.

He chided the report for not addressing the efficiency of turning genomic targets into therapeutic proteins and antibodies, an idea "included in a broad brush, by anecdotal example. We have had a 100 percent success rate in seven Phase I trials. It's 20 to 30 times more productive. That entire argument, and our record of achievement, was ignored. It is bizarre."

Lehman and McKinsey said "one of the recurring themes in these early post-genomics years is the poor quality of target validation. Ten years ago, the average number of literature references per target under consideration at a pharmaceutical company was more than 100, now it is eight. The flood of new targets has not been matched with a flood of biological information about these targets."

The cost of more compounds going through development with "decreased efficiency" also was analyzed in the report. In 1995, the R&D budget per NCE was $700 million, and now is only $800 million.

"By 2005 [when, the report said, genomics will dominate drug pipelines], we expect that with no improvement in the genomics technologies, this will actually increase to $1.6 billion or, assuming modest technological improvements, to $1.3 billion. By 2010, the numbers improve to $1 billion, or $600 million with improvements."

Haseltine bristled at the notion of blaming the quality of genomic drug target validation. "[Poor] is one thing it isn't," he said.

Haseltine instead blamed difficulties in development on a number of other possible factors, including drug-drug interactions, the fact that targets are "extracted from their natural environment, and are being examined as pure enzymes or receptors outside of a biological context," and "process- rather than results-oriented" procedures.

Haseltine said the longer development time only favors companies such as HGS, which started earlier than most. HGS's deal structure gives more than most, too, he said.

"If [GlaxoSmithKline] takes it all the way to approval, we get no exposure at all," he said. "We get 10 percent royalties, if we just sit on our hands. For 20 percent, we have to co-promote, either with people or cash. It's a great deal, and it's over dozens of products."

Rachel Leheny, head of U.S. biotechnology research for Lehman, could not be reached for comment. But, at an investors' meeting last month, the chairman of R&D for GlaxoSmithKline, Tadataka Yamada, referred to the report and somehow managed to agree with the report, and at least partly with Haseltine.

"Normally I don't quote from a holy collaboration between an investment bank and a consulting firm," he said. "But . . . I think they came up with the right conclusion, which is, in effect, that this is where the industry has to go. Once the industry goes there and they don't realize something, they're shocked that their calculated attrition rates are maybe wrong. They're going to have to invest a lot more money to be successful in that game. I think all those are true. The fact is, we've been there."