West Coast Editor

Mixed news from a Phase II study testing its lupus drug sent shares of Human Genome Sciences Inc. into a tailspin, but the company plans to push ahead with development based on a clearly identified target patient group and a dose that works for it.

HGS' stock (NASDAQ:HGSI) closed Wednesday at $9.87, down $4.10, or about 29 percent.

"We're planning for Phase III," said Jerry Parrott, vice president of corporate communications for Rockville, Md.-based HGS, adding that the firm has "a lot of numbers crunching to do before we put them in front of the FDA," and partner GlaxoSmithKline plc, of London, has "only begun to look at the data."

LymphoStat-B (belimumab) missed its overall primary efficacy endpoint of reducing the signs and symptoms at week 24 as measured by the Safety of Estrogens in Lupus Erythematosus National Assessment Systemic Lupus Erythematosus Disease Activity Index, known as SELENA SLEDAI.

An antibody designed to inhibit the biological activity of B-lymphocyte stimulator, LymphoStat-B also missed the primary endpoint of increasing the time to first lupus flare over a 52-week period.

The largest-ever such trial at multiple centers in the U.S. and Canada, HGS' study randomized 449 patients to get one of three doses (1 mg/kg, 4 mg/kg or 10 mg/kg) of LymphoStat-B or placebo intravenously over the 52-week treatment period, along with standard-of-care therapy.

All patients were dosed on days 0, 14 and 28, and then every 28 days for the remainder of the 52 weeks. Eighty-two percent of the patients were receiving background prednisone therapy, either alone or in combination.

The primary efficacy endpoints were the SELENA SLEDAI score at week 24 and time to first disease flare over the 52-week period. Secondary efficacy and biological endpoints, as well as safety, were evaluated, too.

Good news amid the bad: In seropositive patients - those with lupus in their blood, a subgroup that made up 75 percent of the trial population - the drug reduced the signs and symptoms at week 52 at a level of statistical significance as measured by SELENA SLEDAI (p=0.021) and by the Physician's Global Disease Assessment (p=0.016).

Investors may have had some difficulty grasping how 75 percent of patients ended up as a subgroup, but tricky-to-diagnose lupus has no fewer than 11 symptoms including blood tests, and no single test exists to determine whether a patient has the disease. Experts do not yet agree entirely on an index to measure lupus activity, either.

"There are six or seven, each of them with its proponents," Parrott said, though opinions are "beginning to coalesce around a couple of them," including SELENA SLEDAI. The company has "gained a treasure trove of information about these disease activity indices, and we're going to publish it," eventually, he said.

"If you don't deal with the realities of trial design, it's difficult to understand" the Phase II study's outcome, Parrott acknowledged. "Add to that the tremendous complexity of this disease."

Still, in the seropositive group, top-line data from the trial show a "100 percent better reduction" over placebo in disease based on SELENA SLEDAI scores, he said. Using the physician's assessment score, the benefit tripled. Parrott conceded that the latter is "more subjective, but it's still double-blinded."

Trends toward greater reduction in prednisone therapy also showed up in the trial. As expected based on earlier research, LymphoStat-B produced targeted and statistically significant reductions vs. placebo in all active-treatment arms in both circulating B cells (CD 20(+) and other subsets) and anti-double-stranded DNA autoantibodies.

Specifically, results in active treatment groups showed a 29 percent median reduction in anti-dsDNA autoantibodies among patients who were positive for anti-dsDNA at baseline (p=0.0018), and a 33 percent median increase in C4 complement at week 52 among patients with low C4 at baseline (p=0.014).

There was no dose response observed, with signs of drug activity noted across the range of doses studied, including the lowest, and the drug was well tolerated with no clinically significant differences from placebo in adverse events, serious adverse events or laboratory abnormalities. Clinically significant infusion reactions proved rare, and infection rates did not differ from placebo.

LymphoStat-B met its primary endpoint in a Phase II trial in rheumatoid arthritis patients and gained fast-track status for lupus, which has proven to be one of medicine's higher hurdles. Nothing has been approved by the FDA in 40 years for the disease, typically treated with steroids, immunosuppressants and others - "everything on God's green earth," Parrott said, that might help patients without too many side effects.

Others trying their hands in the lupus space include Aspreva Pharmaceuticals Corp., of Victoria, British Columbia, which last month with partner F. Hoffmann-La Roche Ltd., of Basel, Switzerland, finalized a special protocol assessment with the FDA for a Phase III study with CellCept (mycophenolate mofetil) against lupus nephritis. The global study will enroll 358 patients with biopsy-proven lupus. The first patient was treated in July, and results are due at the end of next year.

HGS is "doing our planning [for a Phase III study with LymphoStat-B], but you're probably talking a couple of months, easily" before meetings with the FDA and partner GSK are done, Parrott told BioWorld Today.

In July, GSK rewarded positive LymphoStat-B data from the Phase II trial against rheumatoid arthritis by exercising an option to develop and commercialize the drug for that indication and lupus with HGS. (See BioWorld Today, July 8, 2005.)

"It doesn't do you much good to worry about how the market responds to these things," Parrott said, referring to the hit taken by HGS' shares Wednesday. Instead, the firm looks ahead to using the findings from the Phase II study in further work.

"When we did the rheumatoid arthritis trial, we knew what we were doing," he said. "With lupus, there was not much to learn from."