I believe the Manual Tracking plugin has a directionality feature -- you can define the reference point to be the cancer cell that is nearest to the macrophage at the last timepoint.
Then you can get the % of macrophages that are moving towards a cancer cell (though with the high density you might have some false positives there).
Other simple measures that come to mind:
* average distance between the cancer cell and the 5 (or some suitable number of) closest macrophages
* number of macrophages within a certain radius of the cell -- calculate the difference between that number at the end vs. at the beginning of the time-lapse.
You could get all these values by recording the cancer cell coordinates using the Tracking plugin or the point tool, and calculating distances in a spreadsheet or math application.
I'd recommend trying a few different methods on positive and negative controls, and get an idea of what measure best captures the actual behavior.
Hope this helps.