The DCVax-L treatment arm used a crossover dosing when patients were diagnosed as recurring GBM (progression). 90% of the treatment arm had crossover (2nd dose) during the trial. That is the double dosing. It is in the SAP document.
The external comparator trials were not screened for dosing and most were single dosed. Thus DCVax-L had two doses, most comparator trials used single dose. That is confounding the trial results. NWBO statisticians statistically manipulated the OS endpoint data to try to compensate for the confounding of the data. That was their attempt to make it acceptable for regulators. FDA not likely to buy it.