Exact shares are up 37% this year -- respectable, especially given the long timelines. The pivotal validation study of Cologuard won't be completed until late 2012, with FDA submission to follow quickly thereafter.
"@adamfeuerstein, Any thoughts on PPHM news this morning?"
(PPHM - Get Report)
released preliminary results from a phase II study of bavituximab ("bavi" for short) in front-line non-small cell lung cancer patients. Eighty-six lung cancer patients treated with bavi plus carboplatin and paclitaxel (chemo) demonstrated an overall response rate of 39% compared to an overall response rate of 26% for patients treated with carboplatin and paclitaxel alone.
Kudos to Peregrine for running a randomized phase II study.
The response rate in the bavi arm is numerically higher than the control arm but obviously not statistically significant. How do I know? Because Peregrine would have been shouting from the rooftop if the result was statistically significant.
In the headline of its press release, Peregrine characterizes the difference between the 39% and 26% response rates as a "50% improvement." That's suspect, especially since the result is not statistically significant. The relative difference between the two response rates is irrelevant.
The 26% response rate in the control arm -- those patients who received chemo only -- might be unusually high. In its benchmark phase III "E4599" study of
Avastin in non-small cell lung cancer, researchers reported a 15% response rate in patients treated with carboplatin and paclitaxel. [The response rate for Avastin plus chemo was 35%.] Both Peregrine and Genentech enrolled the same type of lung cancer patients in their respective studies so making cross-study comparisons isn't totally out of bounds, at least for our purposes.
Other lung cancer studies employing carboplatin and paclitaxel found in the medical literature report response rates in the 20% range, or more in line with what Peregrine reported from its bavi study on Tuesday.
In response to my questions, Peregrine says the 26% control-arm response rate is "not unexpected" for a phase II study which enrolled 86 patients. Each patient, Peregrine notes, accounts for 2.5% of the total.
OK, but one implication is that I (and investors) shouldn't necessarily fully trust the data because the study is small. If the control arm response rate is susceptible to relatively large swings due to the outcome of just one or two patients, the response rate observed in the bavi arm is similarly volatile.