By Malorye A. Branca
June 12, 2002 | Struck for the third time by poor performance from one of its genomics-based drug candidates, Human Genome Sciences Inc. (HGS) is shifting course. With product revenues farther from reach than expected, the company is adding more traditional drugs to its pipeline and aiming for new, cash-generating deals.
Founded in Rockville, Md., in 1992 as the for-profit counterpart to J. Craig Venter’s The Institute for Genomic Research, HGS was one of the first and fastest miners of genomic data. The company signed a groundbreaking $120-million deal in 1993 with SmithKline Beecham (now GlaxoSmithKline, or GSK) to provide the pharma giant with new gene targets based on Venter’s expressed sequence tag (EST) methodology. HGS moved quickly to gain a lead in the gene patent race. Today, HGS has more than 7,500 gene-based discovery patents filed, about $1.6 billion in the bank, and seven drugs in clinical trials.
HGS asserts that it has set an industry record for getting monoclonal antibodies and protein-based drugs into clinical trials. The company has filed nine investigational new drug (IND) applications in five years, with two more INDs planned this year. But some of those compounds are running into trouble.
In April, HGS announced it would discontinue the development of one of its most advanced genomics-based drugs, Mirostipen, which was in Phase IIa trials for chemotherapy-induced neutropenia (a side effect of cancer therapy). According to David Stump, HGS’s senior vice president of drug development, “the biological activity observed in patients is not sufficiently robust to meet our criteria for continued development of Mirostipen as a single, stand-alone agent.” In other words, Mirostipen works, but not well enough.
Another promising HGS genomic drug candidate, Repifermin, also suffered setbacks in clinical trials last year. The drug was not effective against ulcerative colitis or in treating mouth sores caused by chemotherapy. But the company is not giving up on Repifermin; it’s conducting new trials for the latter indication, as well as for chronic skin ulcers.
“The bet was that they could get drugs to market faster than the industry benchmarks, and I think they lost the bet,” says Jerry Cacciotti of Strategic Decisions Group, a consulting firm doing pharmaceutical and biotechnology portfolio strategy analysis.
At the company’s annual investor meeting on April 30, HGS chairman and CEO William A. Haseltine presented an updated pipeline and said HGS would leverage its compounds for cash. “Before we get a product on the market, I would bet we get a significant number of products licensed to others, or partnered with others to develop,” said Haseltine.
The revamped HGS pipeline now contains an almost equal mix of genomics and “me-too” drugs. Thanks to a technology acquired when it bought Principia in 2000, HGS can now make longer-acting, albumin-tagged versions of common protein drugs. Three such drugs are already in clinical trials, with more on the way. Although these are not the novel products HGS originally planned on creating, investors may not worry about the difference if they reach the market.
“To me, this looks like a very reasonable strategy,” says Jerry Karabelas of venture capital firm Care Capital. Formerly CEO of worldwide pharmaceuticals for Novartis AG, Karabelas was recently appointed to the HGS board. “This is a way of getting cash flow that will help push the science forward.”
The HGS approach is certainly cheaper than buying new products. Fellow genomic pioneer Millennium Pharmaceuticals bought COR Therapeutics for about $2 billion last year, anticipating that sales of COR’s cardiovascular drug Integrilin would reach about $200 million in 2001. The company declined to say whether the drug reached its goal.
Pushing the Pipeline
Selling products, particularly advanced ones, to other companies seems like the best deal at this time. “Big Pharma has a big problem, and needs to fill in [its] own pipeline gaps,” says Karabelas. Aventis just agreed to pay about $480 million for a cancer drug in clinical trials; Schering is acquiring a late-stage cancer treatment from Immunex for $380 million; and Bristol-Myers Squibb committed about $1 billion for ImClone’s colon cancer drug Erbitux. In the preclinical arena, Genentech agreed to pay about $63 million for a compound from Xenova.
“One of the key challenges for [HGS] is to achieve proof of principal that [its] intellectual property is distinct, and to demonstrate that it provides an increased probability of success,” says Cacciotti.
That validation can come from within or without. Through its early database deals, HGS put more than 460 targets into collaborators’ hands. A success with any of those would not only help validate the company’s approach, but could bring milestone or royalty revenues as well. A leader in that pack is GSK’s Lp-PLA2 inhibitor, a possible atherosclerosis treatment now in Phase I trials. DiaDexus also has an Lp-PLA2-based diagnostic in development, as well as other targets that could bring HGS small royalties.
Because its scientific approach is only one among many, HGS’ progress should not reflect on the entire industry, but it is bound to have some effect. The Washington Post has branded the Mirostipen results “the first failure of a genomics-based drug.”
Haseltine sees things differently. “People ask, ‘Where is the beef in genomics?’” he said at the investor meeting. “If you want to see where it is, it’s in Human Genome Sciences today, and if people are clever, it will be in other companies tomorrow.”