In light of your comments, I think one important aspect of Icagen as an analogy is that, even with multiple failures, Pfizer saw value in targeting ion channels, and therefore saw value in Icagen's platform and compound library. Trial design, indication or specific target may lead to failure, but the broader class of targets can still have great potential. Similarly, despite Lpath's failures (to what degree still TBD), Pfizer probably sees value in targeting bioactive lipids, and therefore values Lpath's platform for targeting them and the resulting portfolio of molecules they're starting to develop. The platform has value, even if iSONEP or ASONEP never reach the market.
As far as what Pfizer did, I think they did extend/restructure the deal once. Ultimately, after Icagen was down in the dumps following several failures it was sub-$1. That was 2010. In 2011 Pfizer bought them for $6.
Icagen had an ongoing collaboration with Pfizer starting in 2007. Their platform involved ion-channel blockade. They had a string of failures during a 3 year period but Pfizer stayed with them. They are now a wholly owned subsidiary of Pfizer.
The point is that even with the platform failures, Pfizer still saw potential. Who knows what will happen, but it just lends some credence to the notion that Lpath is not necessarily doomed.
The painful thing is that these results are materially similar to what was already reported. When you have a bimodal distribution split roughly 50/50, with an endpoint based on median OS, you're already on the raggedy edge. Forty percent of patients could have great results and yet this benefit could be invisible in the median OS.
If a biomarker/responder profile emerges then this drug could be significant. I know, "if" is such a big word for so few letters.