VBL Therapeutics Inc. CEO Dror Harats cautioned investors about jumping to conclusions regarding the late-stage ovarian cancer bid with anti-angiogenic gene therapy VB-111 after the candidate failed in top-line data from the phase III trial against recurrent glioblastoma (GBM).

Harats said results as known so far from the GBM experiment called Globe were shared with investigators in the ongoing phase III ovarian trial. "Instead of them getting cold feet, we actually got a little lecture on why ovarian is so different from GBM," he said, and how drugs that don't work in the one indication may well have efficacy in the other. GBM, he pointed out during a conference call with investors, is "one of the most difficult indications to treat in cancer. We all know that a lot of excellent drugs in [other] cancers didn't work in GBM. We always have a duty to try and treat the very difficult indications, especially when we had such a great result in our phase II," along with phase I evidence that the drug is active in the brain.

Also, he reminded listeners, the phase III ovarian study, unlike the Globe trial, is not a combo-drug test.

Post-phase II hopes in GBM were dashed for VB-111 in combination with Avastin (bevacizumab, Roche Holding AG), as the company disclosed top-line results from Globe. Designed to compare the combo with Avastin alone, the experiment missed its primary endpoint of overall survival (OS). Shares of VBL (NASDAQ:VBLT) closed Thursday at $2.65, down $4.15, or 61 percent.

In March 2015, the Tel Aviv, Israel-based company disclosed interim phase II data that gave cause for optimism regarding VB-111, also known as ofranergene obadenovec. But the compound faltered "at the last hurdle," as H.C. Wainwright analyst Swayampakula Ramakanth put the matter in a research report. "We expect the company to present the detailed analyses for these [other] endpoints at a scientific conference later this year, and if VB-111 is able to show any benefit in patient subpopulations, there may still be a path forward for the drug in this indication," he wrote. But, staying conservative, he excluded "all GBM-associated revenues" from his financial projections. (See BioWorld Today, March 26, 2015.)

Enrolling an estimated 252 patients, Globe is a randomized, controlled, double-arm, open-label study of VB-111 dosed every two months in combination with Avastin given every two weeks, compared to Avastin monotherapy. Key inclusion criteria include first or second progression of GBM following standard-of-care treatment with temozolomide and radiation, a histologically confirmed diagnosis of glioblastoma, and measurable disease by Response Assessment in Neuro-Oncology, or RANO, criteria at progression. The study is conducted under a special protocol assessment (SPA) granted by the FDA, with full endorsement by the Canadian Brain Tumor Consortium. VB-111 has received orphan drug designation in the U.S. and Europe and was granted fast track designation by the FDA.

The candidate emerged from VBL's Vascular Targeting System platform, which deploys an approach that can be tailored to either destroy angiogenic blood vessels or promote their growth, with special efficacy in endothelial cell gene expression.

The stock-boosting interim phase II data two years ago included 46 patients. VB-111 monotherapy was discontinued upon disease progression in 22 who were then treated with Avastin by itself. Another 24 patients, when their tumors advanced with VB-111 monotherapy, could choose to get further treatment with the drug combined with Avastin. Twenty-three patients had received the combo therapy, and one was still stable on VB-111 monotherapy at 424 days. The upshot at the interim look was that VB-111 in combination with Avastin gained a median overall survival of 414 days, compared to 235 days in patients on VB-111 followed by just Avastin (p = 0.05).

'High bar,' confounding variables

But the drug didn't pan out in phase III, and now eyes have turned to the pivotal phase III Oval study, begun in December for platinum-resistant ovarian cancer. The randomized, multicenter, 350-patient experiment is testing the compound in combination with chemotherapy vs. chemo alone, with a primary endpoint of OS.

Wainwright's Ramakanth was skeptical about the so-called lack of read-through. "While we agree that GBM and ovarian cancer are very different diseases and that VB-111 has achieved promising results in a phase II study in ovarian cancer," he said, "we also note that the phase II study only included 21 patients and that the setback in GBM casts doubt on VB-111's therapeutic potential." He lowered the estimated probability of success for Oval to 45 percent from 65 percent. The Oval study design does not include an interim readout, he said, "and in our view, the top-line results from Oval are not expected until the second half of 2021 at the earliest, which may render the stock price range-bound in the near term."

Meanwhile, VBL will be sorting through the GBM results for subgroups that might prove promising to pursue. "It's a little bit too preliminary to discuss it in depth, but we have in the special protocol assessment certain pre-specified subgroup analyses that [were] agreed [upon] with the agency," Harats said. Included are patients with different sizes of tumors.

"Immune response is quite important," too, he said, noting that in the phase II trial, one patient who developed a high temperature, "a special signature of cytokines, actually survived longer and had a better response to the drug. We are going to look at that as well."

The Globe outcome arrived in the form of "big data, and we just got the top-line results a few days ago," he said. "I believe it will take some time to analyze all of this, and we don't want to come with any data or conclusions that are not very solid. A lot of the subgroup analysis depends on the response rate," which is "very tricky" in GBM, especially as measurements made by magnetic resonance imaging. Having more to report will "take some time but not too much time," Harats said.

Piper Jaffray analyst Charles Duncan pressed the point, asking if more information would be coming in "multiple weeks, multiple months or multiple quarters." Harats said "multiple weeks" is the target. "I hope in the next quarter, let's put it this way," he said.

Duncan, in a report, sounded tentative about VB-111's chances. "For now, we have low visibility on any activity signals that may have been seen in Globe, for example on secondary efficacy endpoints (progression-free survival, response rate, quality of life)," he wrote. "In our view, in the 'high bar' indication of an all-comers recurrent GBM study, any improvements on these other endpoints could potentially still represent a clinical benefit." He said he wants to see the subgroup analyses before making a definitive judgment. "While we can imagine any number of reasons a trial could fail in recurrent GBM, we have little insight into the specific confounding variables or execution issues that may have led VB-111 to fail to replicate the phase II open-label data."

More analysis from the company "should help us to understand the broader, future outlook for this candidate in recurrent GBM, ovarian and lung" cancer, he said.