Although I don’t really support doing a brute-force approach doing manual variants review, if you only have some of your top signal that you would like to confirm for further wet-lab experiment validation, IGV might still proves helpful.
This review by Robinson, et al from a group at UCSD shed some lights and detail into how you can do the manual review in IGV: http://cancerres.aacrjournals.org/content/77/21/e31
I also found the IGV manual describing all the options in the preference menu to be quite useful: https://software.broadinstitute.org/software/igv/Preferences
Lost in Translation 