Exelixis (Nasdaq: EXEL) gambled a little more than a year ago, deciding to go "all-in" on its lead drug candidate for cancer after many investors lost interest in its strategy of building a broad pipeline with lots of different drugs.

It was risky thing to do in a business where most drugs fail in clinical trials, but it raised about $180 million on early data to support this idea back in March, and now more and more public data is rolling in to buttress the argument.

The South San Francisco-based biotech company has shown off results over the past several days from a trial that enrolled 490 patients with nine different tumor types, to see where cabozantinib (XL184) might work best. The latest batch of data for this drug, from 171 patients with prostate cancer, is being presented to doctors today at the American Society of Clinical Oncology (ASCO) meeting in Chicago.

Exelixis has taken an unusual tack in the prostate cancer drug business, where drugs are primarily judged on whether they can help men live longer. In this early trial, Exelixis has focused on how its drug was able to partially or completely wipe out bone lesions for about three-quarters (82 of 108) patients whose prostate cancer had spread to the bones. Prostate cancer is notorious for migrating into bones and essentially splintering them from the inside out, causing intense pain and helping send many patients into a death spiral. About 30,000 men in the U.S. die of prostate cancer each year, and while promising advances have been made recently by Dendreon's sipuleucel-T (Provenge) and Johnson & Johnson's abiraterone (Zytiga), neither drug has primarily focused on an ability to preserve bones.

"For the first time, we have enough patients on the drug, long enough, to say, 'Wow, cabo resolves bone scans for almost 80 percent of patients, and you can see clinical benefit,'" says Exelixis CEO Mike Morrissey. "We have a lot of support from investigators, especially in prostate cancer. When metastatic prostate cancer goes to the bone, it drives pain, complications, fractures, spinal cord compression, anemia. Everything combined really puts the patient in a bad place. We have investigators who doing this for several decades who have never seen this level of activity in the bone. They are incredibly enthusiastic."

The drug, pronounced "CAH-bo" like the resort town in Baja California, is a daily pill designed to specifically block two biological targets thought to help cancer grow and spread -- MET and VEGFR2. These targets are implicated in a number of forms of cancer, and since Exelixis didn't know the best possible use for sure, it designed the trial to enroll patients with nine different tumor types, see how the drug performed in a few patients from each group. Then it enrolled many more patients with malignancies that appeared most susceptible to the new drug.

Based on this study, prostate and ovarian cancer have emerged as the two settings in which Exelixis has seen the most interesting results, Morrissey says. Small-cell lung cancer, pancreatic cancer, stomach cancer provided less convincing evidence, Exelixis said.

In the group of prostate cancer patients, researchers found that if they took the Exelixis drug and it got rid of their bone lesions, it helped improve their overall clinical outlook compared to those who didn't see an improvement in bone lesions. Researchers, at six months of follow up, found patients who had improved bone scans were much more likely to have their tumors remain in check (61 percent to 35 percent), to have pain relief (83 percent to 43 percent), and to reduce or eliminate dependence on narcotic painkillers (68 percent to 33 percent), Exelixis said.

Once patients' tumors progressed, there was an opportunity for them to go into another phase of the study in which they were randomly assigned to get another round of the Exelixis drug or a placebo. At that point, researchers found that patients' tumors were kept in check for a median time of 21 weeks on the Exelixis drug, compared with about six weeks for the placebo. The overall number of patients in this group (31) is small, but the difference was statistically significant, and the Exelixis drug was shown to offer an 87 percent reduction in risk of a patient's disease spreading.

There were significant side effects reported with this treatment, which doctors will surely be scrutinizing at ASCO. Fatigue (16 percent), high blood pressure (6 percent), a tenderness/peeling of skin known as hand-foot syndrome (6 percent), decreased appetite (5 percent), nausea (4 percent), and vomiting (4 percent) were the most common moderate to severe side effects reported. One patient on the drug died from unexplained causes at Week 33 in the study, according to an Exelixis statement. About half (51 percent) of patients had to reduce the dose at least one time, Exelixis said.

The safety profile of the drug, Morrissey says, was consistent with prior studies, and consistent with other drugs in its class. There was "nothing overall surprising," he says.

Still, the unexplained death is bound to raise at least a few questions at the ASCO conference. Earlier in the weekend, Exelixis presented data from 70 patients on the treatment for their ovarian cancer, and found that 73 percent had at least some tumor shrinkage, and 24 percent had what is considered significant shrinkage of their tumors. Researchers don't yet have data to report on whether the drug keeps tumors from spreading for a meaningful length of time, or whether it helps extend lives.

The side effect profile in these patients appeared similar to what was observed in prostate cancer patients, although two deaths in the ovarian cancer patients were reported by researchers as drug-related. One was characterized as an "enterocutaneous fistula," which the National Institutes of Health defines as an abnormal opening that allows the content of the stomach or the intestines to leak. The other patient who died had an intestinal perforation, Exelixis said.

Cory Kasimov, an analyst with JP Morgan who reviewed the ovarian cancer data, said yesterday in a note to clients that, "overall, cabo clearly looks like an active agent in this setting," but said it's hard to put the findings in context without a full report on survival times or the length of time tumors stay in check. He also noted the two deaths, "which has prompted some incoming questions."

David Miller, president of Biotech Stock Research, said the bone scan data "look great" in prostate cancer patients, especially when it's shown to be linked to better clinical outcomes, like less pain. But Exelixis did report six deaths out of the total database of 490 patients in the trial. The deaths of the patients "are always a concern," that need to be inquired about more, Miller said.

Exelixis will have more to say about all of this on a conference call with investors later today from Chicago, at 7 p.m. Eastern/6 p.m. Central/4 p.m. Pacific.


More from Xconomy.com:

Luke Timmerman is the National Biotech Editor of Xconomy, and the Editor of Xconomy Seattle. Email him at ltimmerman@xconomy.com or follow him on Twitter at twitter.com/ldtimmerman.