While setting new cancer cell lines from primary tissues what could be the major consideration so that the cell line would be the representative of a population from which it was derived.
We are just finishing to write a paper in which we suggest to use more than on growth conditions - in the case of glioblastoma probably using media for NSCs, OPCs, and even serum would increase coverage, as each type of medium selects for some type of cell.
Most important: that it is actually derived from the population you think it was: perform short tandem repeat (STR) testing to be certain of genotype. Think of all the cell lines contaminated with HeLa.
Next most important: that it has some physiological characteristics of the tissue type. Here the best surrogate for exhaustive phenotypic testing is probably RNASeq to look at mRNA transcript profiles. Obviously, the immortilization method (e.g. p53 status) will affect certain transcripts, but globally you would ideally like your cell lines to cluster with their parental cell type.
I am working on isolating HCC cells and liver cancer stem cells from excessed tumor tissue from patients. Regarding my humble experiments, dozens of factors affecting isolation of specific cells; like isolation method of total cells, culturing media, growth factors, kind of serum, timing...............
Regarding cells characterizations, I think the best is mRNA profiling, DNA sequencing and micro RNA profiling
Generating a primary culture using a tumour tissue specimen as a source is relatively easy as compared to developing a well defined cell line which required several stringent criteria. With regards to starting a primary culture, it could retain most of the diverse cell types from a given organs. However, one needs to remove fibroblast which I assume you do not wish to have in the primary culture. Fibroblasts could easily overtake the epithelial cell population in a rather short period of time. There are several commercially available cell culture media which are designed to selectively remove fibroblast during the early phase of culture. One of the easiest and least expensive means to characterize a primary culture is by immunostaining of a sample at every passage. Such an assay will allow you to monitor the characteristic of the cancer and normal cells as well as the presence or absence of contaminating cell types (e.g., fibroblast) of a primary culture during its early passage.