re DNDN short interest: There was a concerted effort to attack the stock price on Feb 17 after the ASCO abstract was leaked a day early by ABC-TV in Chicago on 2/16, causing the price to rise. DNDN's 9901 trial presentation was on 2/19, but the UBS analyst issued a reiteration of the $6 price target in premarket on 2/17, stating that there were biases in the stratification of patients in 9901: (1) the survival benefit was due to an imbalance of patients with Gleason score of 7 or less favoring the Provenge arm; and (2) a higher percentage of patients with only soft tissue metastases were in the placebo arm (UBS analyst said that presence of soft tissue mets means more advanced prostate cancer than presence of bone mets).
18 million shares traded on 2/17, and the stock went down despite stories in the WSJ, USA Today, AP, Reuters, and every major city newspaper. I would say that there was a lot of selling by the momo crowd, in addition to the shorting of 2 million shares beginning that day and extending through the next few trading sessions.
On Sat 2/19, the 9901 presentation proved that the UBS analyst was incorrect concerning the Gleason score imbalance. There was actually a greater relative benefit for Gleason 8 and 9 patients. Of course, everyone familiar with advanced prostate cancer already knew that UBS was wrong about the issue of soft tissue mets vs. bony mets. The presence of bone metastases indicates MORE advanced PC, not less, especially when those soft tissue mets are not present in the viscera. This means that the Provenge arm had a tougher set of patients than the placebo arm, and still showed a stat sig survival benefit.
The conspiracy theories on the Yahoo DNDN board are a little wild. It's simply a situation where DNDN has replaced GNTA as the favorite whipping boy of the hedge fund crowd which shorts development-stage biotechs, and the biggest hedgies are trying to keep the smaller hedgies in line. When one hedge fund capitulates, more will follow. Right now they are showing pretty good discipline.
The huge jump in short interest occurred after the limited release of the 9901 results in late October 2004. The initial argument was that DNDN mgmt was lying about the results. Then in early December, the argument became "a few patients in the Gleason 7 or less subgroup lived a long time." Then it became what the UBS analyst said (earlier in this post). After the ASCO presentation, it has become "9902A will not be stat sig, so DNDN won't file the BLA." If DNDN files the BLA after supportive but non-stat sig 9902A survival data is announced, then it will be "FDA won't allow pooled data, and the trials failed the primary endpoint, and 9902A was not stat sig." If DNDN files the BLA after stat sig 9902A survival data is announced, the bear argument will be "trial size too small, and the trials failed primary endpoint." If the FDA approves Provenge, it will be "DNDN has to spend a lot of money for sales force, there will be losses for years, and doctors will use Taxotere instead." If Provenge sales take off at the expense of Taxotere, the shorts will say "competition is on the horizon."
