Again this is a trial of which the outcomes can vary depending on the patient profiles (especially with a low n like here). Do you really expect a linear improvement from interim to final results. If you really think that, then you're a tool!