Cancer cells often show heterogeneity. One aspect of heterogeneity is the presence of multi-nucleated and/or large cells in the general cell population. How can one separate this population while keeping it alive?
Depending on the quantity of material availible it is also quite feasible to sieve an aliquot of cells through a suitable sized mesh. Sieves are commercially available for this use and it produces satisfactory results.
If you want to isolate multinucleated cells, then it depends on how these cells are formed. If they result from cell fusion, then you can label your cells with either one of two vital dyes (say green or red), co-culture the labeled populations and isolate double-labeled cells by sorting. Alternatively, you can introduce one of two drug resistance markers (say to puromycin and blasticidin) and select from cells that have both of them.
Another possibility, which would also apply to multinucleated cells that result from failed cytokinesis, is to flow sort the cells using a vital DNA dye.