In spite of less-than-ideal results from a Phase III study of its investigational HIV protease inhibitor 433908 (908), Vertex Pharmaceuticals Inc. still called the drug approvable, although the results point to what could become a difficult marketing hurdle.

The Cambridge, Mass.-based company filed for FDA approval late last year based on findings from two prior Phase III trials called SOLO and NEAT. Vertex maintained expectations of FDA approval and U.S. launch in the fourth quarter of this year, with European approval and launch expected next year. (See BioWorld Today, Dec. 23, 2002.)

"These results met our expectations for safety and antiviral activity of 908," Michael Partridge, the director of corporate communications at Vertex, told BioWorld Today. "We've already filed large amounts of clinical data with the FDA, including 48-week results from the first two studies as well as the 24-week results from the study that we've just done. But these data also are being provided to the FDA."

The news, released after the market closed Thursday, nonetheless played a role in a 9 percent drop in Vertex's stock Friday. The shares (NASDAQ:VRTX) fell $1.34 to close at $13.80.

While the latest results do not necessarily call into question the previous data, they could prove troublesome relative to market penetration. Preliminary 48-week data from the open-label, 320-patient CONTEXT study, conducted by development partner GlaxoSmithKline plc, showed the drug failed its primary endpoint: non-inferiority to a comparator drug, lopinavir (Kaletra, from Abbott Laboratories). More specifically, in an evaluation of time-averaged change in viral load, once- and twice-daily doses of 908 boosted with ritonavir produced no better results than twice-daily doses of lopinavir with ritonavir.

"In general, our physician consultants have been rather unenthusiastic about 908," Phil Nadeau, an analyst with SG Cowen Securities Corp. in New York, wrote in a research note. "Although they believe that 908 is approvable, they think that it is unlikely to unseat Abbott's Kaletra as the first-choice protease inhibitor, an opinion that is unlikely to be changed by last night's data release."

He called Kaletra "the HIV protease inhibitor to beat," with worldwide sales projected to reach $1.2 billion by 2006.

At 48 weeks, the mean time-averaged change in viral load (log¹⁰ c/mL) was minus-1.49 for once-daily 908/ritonavir patients, minus-1.53 for twice-daily 908/ritonavir patients and minus-1.76 for twice-daily lopinavir/ritonavir patients. Vertex said non-inferiority could not be established because the upper limits of the 97.5 percent confidence intervals for the differences in mean time-averaged change in viral load between the two 908 arms to lopinavir comparisons were above 0.5.

The findings contrasted those at 24 weeks, in which 908 was demonstrated to be statistically non-inferior to Kaletra.

Patients in the three treatment arms, all of whom had prior protease inhibitor treatment experience, also received two nucleoside reverse transcriptase inhibitors. The patients had been randomized into three groups - 105 received 1,400 mg of 908 combined with 200 mg of ritonavir once a day, 107 received 700 mg of 908 with 100 mg of ritonavir twice a day, and 103 received 400 mg of lopinavir with 100 mg of ritonavir twice a day.

Vertex and London-based GlaxoSmithKline co-discovered the compound, a calcium phosphate ester prodrug of their amprenavir (Agenerase). If approved, GlaxoSmith-Kline would market it and both companies would share co-promotion duties in the U.S. and some European markets.

"These results show that 908 is an active drug in treatment-experienced patients, and that's good news because they are in need of options to construct viable regimens," Partridge said. "For us, it was encouraging to see the level of activity we saw with 908 in this population. And we think that the clinical activity seen in this trial will be meaningful for clinicians as well."

He added that 908's potency, convenience and tolerability profile would bolster its presence in the protease inhibitor-naive market, which he said has moved toward simplified regimens. It has been dosed as both a single tablet once a day or two tablets twice a day.

Nadeau, the investment analyst, also pointed toward the drug's favorable tolerability.

"Our consultants do think that 908 will be used in the 10 percent to 15 percent of patients who cannot tolerate Kaletra," he said. "One consultant in particular notes that 908 will be his first choice in those patients who find Kaletra's gastrointestinal side effects intolerable."

Vertex and GlaxoSmithKline noted 908's antiviral activity through measures of patients with viral load less than 400 copies/mL and less than 50 copies/mL. In an analysis of the proportion of patients with viral load below 400 copies/mL, those in the twice-daily 908/ritonavir arm (58 percent) and those in the twice-daily lopinavir/ritonavir arm (61 percent) demonstrated comparable levels of antiviral activity. Also, 50 percent of once-daily 908/ritonavir patients achieved viral load below 400 copies/mL.

The proportion of patients who achieved viral load below 50 copies/mL at 48 weeks was also similar in the twice-daily 908/ritonavir arm (46 percent) and the twice-daily lopinavir/ritonavir arm (50 percent).

Vertex added that the latest data in treatment-experienced patients complement results from the prior two Phase III trials, which supported the antiviral activity and tolerability of 908 in unboosted and boosted regimens in treatment-naive patients. The NEAT and SOLO studies, which included 249 and 660 patients, respectively, compared 908 to nelfinavir (Viracept, from Pfizer Inc.) in treatment-naive patients.

The NEAT trial showed 109 of 166 HIV-positive patients achieved an undetectable viral load with 908, compared to 42 of 83 patients taking nelfinavir. Both the NEAT and SOLO studies met their primary endpoints at 24 and 48 weeks. (See BioWorld Today, Feb. 18, 2003; Sept. 30, 2002; and May 20, 2002.)

Among the latest findings, the overall incidence of drug-related adverse events was comparable in the two twice-daily treatment groups and there were no statistical differences in any individual adverse events. Vertex and GlaxoSmithKline said they would report a final comprehensive analysis of the 48-week CONTEXT data at a future medical meeting.

Outside of 908, Vertex's pipeline includes several Phase II candidates - merimepodib for hepatitis C virus, pralnacasan for rheumatoid arthritis and osteoarthritis, and VX-148 for psoriasis. Later this year, the company plans to report data from its studies of pralnacasan in osteoarthritis and VX-148.

Its Phase I products include VX-765 for inflammatory diseases, VX-563 for sickle cell disease and VX-702 for inflammatory diseases.