I am a layman too who will definitely fail biology 101 exam. But fortunately for this investment, we can treat many things as blackbox and still be able to draw definitive conclusion.
From my understanding about the poly-iclc paper and the one on neoadjuvant therapy, Dr. Liau and her team were trying to find among tens of thousands of gene sets the specific gene set with its GSVA score that can be used to predict OS or t-cell infiltration. As shown in the paper on the combination of DCVax-L with poly-iclc, Dr. Liau showed that the specific gene set called HALLMARK INTERFERON GAMMA RESPONSE with its GSVA score greater than median specifically related to the significant OS. I also list the genes in this specific gene set here.
My conjecture about the new expansion cohort of neoadjuvant therapy is that they used the GSVA score to evaluate the trial results and found out that the above specific GSVA score showed there was no survival benefit for neoadjuvant therapy.
However, we did not confirm a survival benefit to neoadjuvant pembrolizumab in recurrent glioblastoma. Our new data suggests some patients may exhibit innate resistance to pre-surgical ICI and require other concomitant therapies to sensitize effectively.
You remind me one point that I forgot to mention before. When I first took look at the data from the poly-iclc trial, I thought Dr. Liau measured the data at 0th and 29th day after the initiation of the treatment during the trial. I think I was wrong about this. How could I know I was wrong? As shown below, Dr. Liau used Illumina NovaSeq 6000 for the sequencing. But Illumina NovaSeq 6000 was available on market after 2017 and the poly-iclc trial was finished with enrollment before 2014. So Dr. Liau did the data analysis at least after 2017 using the tissue samples collected during the trial.