Not really. Safe Harbor protects companies at this stage from a host of activity related statements, savvy investors know this. In this case, however, several possibilities may apply.
1. At the end of a study all data records are reviewed by the sponsor at each study site to correct errors or omissions after the last datapoint has been collected. In the industry once that process is completed we say that the dataset is 'locked', no further changes are allowed. For example; in the PVCT HCC PI study, 6 patients were enrolled with 'hepatocellular carcinoma', but in the end it turned out that 5 patients in the study had HCC, one had metastatic cancer growing in the liver. So in this case, one interpretation of the statement would be that the information presented in Australia 20 months ago was from an 'unlocked' data set which the company has taken 20 months to scrub (with the benefit of insight from ongoing FDA discussion) to the new data set to be presented this year. 20 months is a long time to scrub 80 records by anyones definition.
2. We have talked about this before, use of the arbitrary division of PV-10 patients into 'responder' and 'nonresponder' groups to evaluate PFS and other outcomes is a manipulation of the data. In this manipulation, the reader has to assume that PV-10 in the non-responder was benign, acted as a control. But the company has presented no independent evidence for this in their clinical trials (leaving equally valid the possibility that PV-10 killed patients faster in the 'non-responder' population). Further, by simply changing the definition of a "responding patient" the company can change the 'look of the data', the p value associated with that look, and the conclusions they want you to take away.
3. So to, they can re-group data by cancer stage to create a similar effect, although in this case we assume they are not manipulating the 'staging definition', just the tabulation of numbers (think about Congressional re-districting as an example).
In a PIII trial all these points are mute as end points are clearly defined (there are no do-overs), ask the Regeneron folks what it means to have a well designed PIII study...they should have looked at the Savient study designs a little more closely (although Savient is an classic example of what happens to a company when investors pay for PIII trial and no one in big pharma wants the drug).*
* Purely a disclaimer so The Master does not jump to the wrong conclusion that I am endorsing Savient, as he is inclined to do.
The company wants investors to believe that there is a survival advantage for PV-10 (and with a properly constructed study that may turn out to be the case), but manipulation of datasets with arbitrary classifications or grouping is not the approach to take. I think they are trying to re-launch the data [under the guise of new long term outcome data as indicated in the first post] because they can only project $ through 2013 and the risk of not getting a SPA on first application has started to sink in. They are in a two step process here...have to get a protocol approved and have to get the money to run a trial (although technically I suppose if they get the PIII approved but don't have the $ they could BK the company at the end of 2013 and hold the approval as an asset for future business interests? Recent comments about not wanting to dilute the company further are interesting in this light). PVCT is not known for speed or agile handling of development challenges, 2013 is not that far off now.
Hope that is helpful for you |
