The targeting here is definitely the point, its the core hypothesis in the abstract. In general this increased targeting ability has been enormously powerful in

The targeting here is definitely the point, its the core hypothesis in the abstract. In general this increased targeting ability has been enormously powerful in solving the response rate problem you mentioned. Fundamentally the problem isnt that the response rate is low accross the board, it's that it's incredibly high in a small subset and incredibly low in the rest. Figuring out that subset is basically wins all around - you can identify the actual subpopulation You're treating so your study is better powered (less noise from people who wouldn't be responsive), you can codevelop a diagnostic to identify chance of response, you don't treat and give side effects to patients who won't respond, also saving costs. This is a bit of a weird post and comment thread because this is an older paper about a technology that is genuinely very successful, but it's being responded to as if it's one of those random mouse trials of a brand new therapeutic strategy. This drug already had accelerated approval and was on the market when this trial was run! In case the rest of the world missed the memo - some of those mouse trials y'all read about in the early 2000s or so? Yeah they worked in humans too. This is one of them.