Feuerstein: The bear case for EXEL
Exelixis shares were trading around $12 in the run up to last year's ASCO meeting but have since slumped to less than $5. To help explain why Exelixis has fallen and isn't getting up anytime soon (say the bears), I've asked a veteran Wall Street biotech investor -- and a longtime source -- to lay out his negative thesis for cabozantinib in this week's Biotech Stock Mailbag. His job prevents him from going on the record, so let's just call him Bio-investor X. And while he's not short Exelixis, you can assume that he's in the short-selling camp on this stock.
Remember, every stock -- biotechs included -- has a bull and bear thesis. You can't make an informed investing decision without knowing what the other side thinks.
As background, Exelixis is conducting two phase III studies of cabozantinib ("cabo" for short) in advanced prostate cancer. The company's plan is to use data from both studies to seek regulatory approval.
Study 306 is designed to assess durable pain reduction, currently enrolling prostate cancer patients with progressive disease (bone pain) after treatment with Taxotere and either Johnson & Johnson's(JNJ_) Zytiga or Medivation's(MDVN_) MDV3100 . The study's targeted enrollment is 236 patients who are randomized 1:1 to receive either cabo or mitoxantrone/prednisone.
Study 307 is the survival trial and will enroll the same type of prostate cancer patient (post Taxotere and either Zytiga or MDV3100), randomized 2:1 to receive either cabo or prednisone. Target enrollment is 960 patients. An interim analysis of survival will be conducted after 387 deaths, a final analysis at 578 deaths.
Top-line data from both studies is expected in late 2013 or first half 2014, says Exelixis.
Exelixis bear thesis of Bio-investor X:
Exelixis will miss trial enrollment timelines, especially since on top of Zytiga, Amgen's(AMGN_) Xgeva and Dendreon's(DNDN_) Provenge (all approved and competing for prostate cancer patients), Algeta's Alpharadin and MDV3100 will be approved soon and also have compassionate-use programs in place.
The trials are mis-modeled; the assumptions for survival in the control arm of study 307 are too low. This will make timing of events (deaths) slower and hamper cabo's ability to beat the control arm (survival benefit) by 30%.
Study 307 is being conducted at an unproven lower dose of cabo than previously tested, which is likely to reduce the drug's efficacy and make it difficult to win on survival benefit (especially if the control-arm patients live longer than expected.)
The pain trial, study 306, is of no value on its own for a cancer drug. FDA recognizes three endpoints for prostate cancer drugs -- survival, quality of life and pain reduction -- but if you don't get a survival claim, forget about premium pricing.
Bio-investor's main beef with Exelixis is Study 307, which he predicts is more likely to fail to demonstrate a survival benefit favoring cabo. Without a survival benefit, FDA will not approve cabo, he says.
I turn the Mailbag over to Bio-Investor X:
"Exelixis and the analysts supporting the company say study 307 will take a year to enroll, with top-line data ready about a year after full enrollment. This timeline is unrealistic.
"Algeta's phase III study of Alpharadin [another prostate cancer therapy] took 36 months from the start of enrollment to the reporting of interim data. The study enrolled 900 patients, with median survival of Alpharadin reaching 14 months compared to 11.2 months for the control arm.
"Johnson & Johnson's Zytiga study took one year to enroll 1,150 patients (with no competition) and 22 months from the start of the trial until interim results. Survival in the Zytiga arm was 14.8 months versus 11.9 months in the control arm.
"Medivation's MDV3100 trial took 26 months to enroll 1,165 patients and reach the interim look. Survival in the '3100 arm was 18.4 months compared to 13.6 months in the control arm.
"With the timelines of these three prostate cancer trials in mind, it seems aggressive to expect Exelixis' study 307 to enroll 960 patients and gather enough events for a top-line data analysis in two years, especially since the competitive environment will be tougher for them. Of course, Exelixis will try to speed up enrollment by seeking patients outside the U.S., but it will still be difficult to find patients that meet the study's enrollment criteria (more so than the three trials mentioned above) and who aren't too sick to tolerate a toxic drug with a narrow therapeutic window like cabo. Who would enroll in this trial ahead of taking Alpharadin or MDV3100, both of which have survival, pain palliation and quality-of-life data?
"Let's turn to an explanation of how Study 307 is mis-modeled. Exelixis designed the study with the assumption that the control arm (patients treated with prednisone) will live an average of 7 months. But look at the post-progression survival for MDV3100 (10.1 months) and for Zytiga (9.2 months). Is it realistic to expect dosing beyond progression? I say no, it isn't. A blend of 10.1 months and 9.2 months is unlikely to give you 7 months survival. The patients in the control arm of Study 307 will live longer than Exelixis modeled.
"With Zytiga approved and MDV3100 and Alpharadin going to be filed for approval soon (and both available already under compassionate-use programs), Exelixis will have a tough time enrolling patients in Study 307. If we assume the trial starts now, I don't expect the company to have data until the first half of 2015 at the earliest, and that assumes a lightning-fast enrollment pace and an early halt for efficacy. How likely is it that the cabo trial can be halted early for efficacy? Not so much because cabo has yet to show any survival benefit to date. The drug is being studied at a new, lower dose and at every turn, it seems inferior to the competition. On pain, Alpharadin and Zytiga are better; on survival MDV3100, Provenge, Alpharadin and Zytiga win; and on quality of life, Alpharadin and Amgen's Xgeva win.
"I see two possible outcomes: Exelixis is forced to resize the trial, i.e. add more patients, which means more cost and more time; or Exelixis moves forward with the current trial size, setting up a very high-risk event that is more likely to fail."
Thank you, Bio-investor X.
Any comments?
Bladerunner
Market Data 
Markets 
AI
