My understanding is that it isn't necessary since it appears that the idea behind a feeder layer is that the stem cells produce their own bFGF & EGF along with other growth factors. Yet I've still seen some papers (e.g. this one which I am trying to reproduce), use it in mediums used for NSC maintenance (their neural progenitor culture contains both). This paper seems to suggest that a feeder layer for NSCs is unnecessary which would make sense but I wanted to know what is common practice and the most robust method and how it is related to these findings. Because of that I would like to hear from personal experience but references are also obviously useful.