This has been covered in a few places like the BBC and ScienceDaily. It seems that some researchers at Queen Mary College in London have recently come with a non animal 3D human breast cancer model.
This is interesting at more than one level. Their research was funded by a research charity (Dr Hadwen Trust) that supports the development of methods that avoid animal experiments. Working with rats is not that satisfactory for ethical reasons and also because there are many cases in which the results of rat experiments cannot be extrapolated to humans (seems we are not so similar in some respects after all). It is also much more realistic than just taking some human cancer cells and studying them on a petri dish.
Being capable of performing experiments using realistic 3D models quickly and efficiently is one of the holy grails of theoreticians since it would make experimental validation of our models much easier (confusingly enough what theoreticians call model, eg, equations or computer rules, is not what experimentalists understand as a model, eg. rat, arabidopsis or drosophila). This validation is quite complicated as I have already mentioned in another post. Making this validation easier and more convenient will go a long way in terms of making our work more reliable and quantitative.