IN VITRO EXPERIMENTS WITH PRIMARY MAMMALIAN CELLS: TO POOL OR NOT TO POOL? EDITORIAL

There are various strategies that can be adopted when performing in vitro experiments with primary cells such as mesenchymal stem cells. It is generally accepted that multiple donors need to be investigated to take into account donor to donor variability; this is especially critical when investigating primary human cells. However, increasingly it is being seen that studies are pooling the cells from multiple donors prior to performing the experiment. This has obvious advantages but also many disadvantages, the greatest being loss of statistical power. Pooling donors reduces variability within, and between, experiments. In some cases, such as the allogenic transplantation of primary cells, the pooling of cells may be a necessity due to the low numbers of harvested cells. To overcome the loss of statistical power, multiple experiments using various donor pools can be performed. When the experiment is of long duration, as is often the case in tissue engineering studies, pooling the cells offers the opportunity to perform a large experiment with multiple replicates. Statistically speaking, each independent experiment is an n = 1 and the replicates are an indication of measurement variation. While it is sometimes argued that three independent experiments using the same pool of cells is therefore an n = 3, this is incorrect as any variation measured would be due to subtle changes in methodology introduced during each experiment. Ultimately, one pool of cells is considered to be a single sample and any repeat experiments using the same sample will only determine slight variation between experimental procedures. It is also undeniable that valuable information is lost when taking this approach. The most obvious being, how reproducible is a phenomenon within a population? Do all donors respond to the same extent or is there a variation in the magnitude of the response? More critically as we move towards autologous cell based therapies, how many donors respond to a particular treatment and how many do not? For novel cell based therapies to be translated into a clinical setting, it is critical to know what per cent of the population are likely to respond. If it is determined that a IN VITRO EXPERIMENTS WITH PRIMARY MAMMALIAN CELLS: TO POOL OR NOT TO POOL?