On Friday, Amgen Inc. masked disappointing findings on cancer immunotherapy talimogene laherparepvec, commonly known as T-vec, by disclosing that secondary endpoint data on overall survival (OS) in melanoma showed a “strong trend” despite a statistical miss.

In March 2013, Amgen reported the global, randomized, open-label, phase III OPTiM study met its primary endpoint of durable response rate (DRR), and the oncolytic immunotherapy generated buzz at last year’s meeting of the American Society of Clinical Oncology (ASCO). (See BioWorld Today, March 31, 2013, and June 4, 2013.)

But analysts had eagerly awaited OS data from the trial, which evaluated the efficacy and safety of T-vec in unresected stage IIIB, IIIC or IV melanoma compared to treatment with subcutaneous granulocyte-macrophage colony-stimulating factor (GM-CSF). Following release of the data, most gave Amgen the benefit of the doubt.

The Thousand Oaks, Calif.-based company reported that the trend in favor of T-vec achieved a “p” value of 0.051, giving the drug an estimated OS hazard ratio and improvement in median OS similar to findings reported at the interim analysis. That result was good enough for Piper Jaffray analyst Joshua Schimmer to declare that, with respect to T-vec phase III OS data, “close counts in horseshoes and melanoma.”

Schimmer pointed out that, although OS was a technical miss, the “p” value was “as close as you can come to positive without actually being positive.” Since the DRR was hit by a margin of 16 percent to 2 percent for GM-CSF and interim OS trend data were strong in treatment-naïve and earlier-stage populations, “it is hard to imagine the FDA not accepting the application for approval,” he wrote.

RBC Capital Markets analyst Michael Yee was similarly unphased by the miss. “FDA might be open to approving it given a very durable response and trend in OS, and a benign safety profile, as an alternative to Yervoy which does not have benign tolerability,” he wrote in a first glance. “If approved, T-vec could be attractive for patients with earlier, slightly less aggressive disease.”

T-vec is designed to selectively replicate in tumors and to initiate an immune response to target cancer that has metastasized. In the phase III trial, patients were randomized 2:1 to T-vec every two weeks through direct tumor injection or GM-CSF subcutaneously for the first 14 days of each 28-day cycle, for up to 18 months.

Amgen gained the technology in its $1 billion acquisition of Biovex Group, of London, in 2011. At the time, the immunotherapy candidate had already moved into pivotal phase III trials in melanoma and head and neck cancer, both under FDA special protocol assessments. (See BioWorld Today, Jan. 26, 2011.)

Amgen was short on details about the next steps. Spokeswoman Kristen Davis said the company was encouraged by the response, which merits “further research of talimogene laherparepvec to better understand its role in melanoma, both as a single-agent and in combination with other therapies.”

Amgen has a phase II combination study with Yervoy (ipilimumab, Bristol-Myers Squibb Co.) under way in treatment-naïve stage IIIb/IV melanoma, with final OS results expected in 2016. The trial is comparing T-vec in combination with Yervoy to Yervoy as a single agent. Secondary endpoints include safety, objective response rate (ORR), time to response, duration of response, progression-free survival, resection rate and one- and two-year survival rate, according to Thomson Reuters Cortellis Clinical Trials Intelligence (CTI). Outcome measures will be assessed 24 months following randomization of the last enrolled patient.

Amgen also has an open-label phase Ib/II study under way with Merck & Co. Inc. examining T-vec in combination with the PD-1 antibody MK-3475 in mid-to-late stage melanoma, with ORR as the endpoint, according to Cortellis CTI.

Although the focus, for now, is squarely on melanoma, “we believe that it is possible that the effects seen in melanoma may be evident in other tumor types,” Davis told BioWorld Today.

In the meantime, Amgen remains “in close communication with regulatory authorities” around the T-vec phase III data, she said, declining to comment on a potential filing timetable. The company plans to present full data at a future medical congress, which ISI Group analyst Mark Schoenebaum speculated could be as early as ASCO.

As Schoenebaum observed, interim OS data in November showed a median benefit of 4.3 months for T-vec. At the time, Amgen also reported subgroup analyses suggesting potentially greater OS benefit in patients receiving first-line T-vec and in those with stage IIB, IIC or IV M1a disease.

However, “based on a lack of OS benefit, the likelihood of regulatory approval and (more so) the commercial opportunity (given that T-vec administration entails injections of the tumors) for T-vec monotherapy in melanoma appear less,” Schoenebaum wrote in an email.

Investors were equally cautious. Although the stock (NASDAQ:AMGN) was barely dented at Friday’s opening bell, shares slid throughout the day, falling $5.02 to close at $119.11. The loss of 4 percent, or about $3.6 billion dollars for the company, was more than three times the price tag of Biovex.

DENOSUMAB DEAL WITH GSK ALSO RE-TOOLED

The T-vec miss came the same week Amgen and Glaxosmithkline (GSK) plc, of London, quietly re-tooled their collaboration to commercialize and market denosumab in postmenopausal osteoporosis (PMO). That deal, inked in July 2009, called for Amgen to receive up to $120 million, with an undisclosed amount up front, in return for GSK’s right to sell denosumab in Europe, Australia, New Zealand, Mexico and other territories following approval. Amgen retained rights in PMO and cancer in the U.S. and Canada and in oncology indications in Europe and certain other markets. (See BioWorld Today, July 28, 2009.)

After a complete response letter (CRL) in June 2010, the FDA green-lighted the drug in PMO, where it’s branded as Prolia. But the bigger win came in November of that year, when the drug was approved as Xgeva to prevent skeletal-related events in patients with bone metastases from solid tumors. Even before European approval of Xgeva in July 2011, sales of the drug already outpaced those of Prolia. (See BioWorld Today, June 10, 2010, Nov. 22, 2010, and April 22, 2011.)

Blue-sky revenue forecasts hit turbulence, however, when the drug received another CRL in its bid to expand the label as a treatment to prolong the time to bone metastasis in men with castration-resistant prostate cancer. The FDA’s action came after an Oncologic Drugs Advisory Committee meeting in which members voted 12-1 against recommending approval in the indication. (See BioWorld Today, Feb. 9, 2012.)

In modifying their agreement, Amgen and GSK terminated their collaboration in all countries except Australia. Amgen is set to resume commercial activities previously assigned to GSK by year-end. Amgen agreed to pay GSK $275 million during the transition period, including an undisclosed initial payment and milestones.

The biotech also will reimburse GSK $15 million to cover certain costs during the transition period.

The original agreement gave GSK the right to register and commercialize denosumab for all indications in countries where Amgen did not have a commercial presence. At the time, those territories included China, Brazil, India and South Korea. Amgen had the option for an expanded role in selling denosumab in Europe and certain emerging markets. Terms of the expansion agreement remain in force, according to Amgen’s 8-K filing.