I am a software engineering researcher (read: academic) and my expertise includes regression testing techniques. I'm mostly interested in academic publications, so the context of patent is definitely to me. As far as papers go, however, the earliest prioritisation paper that I'm aware of goes back to 1999:
Rothermel G, Untch RH, Chu C, Harrold MJ. Test case prioritization: An empirical study. Proceedings of International Conference on Software Maintenance (ICSM 1999), IEEE Computer Society Press, 1999; 179–188.
Now, I don't know what qualifies for the "Continuous Integration" part. If you mean prioritisation as part of an established CI tools, such as Jenkins, than very little academic work will count for that. However, if you broadly mean the context of regression testing (where you are testing the new version using, partly, the existing test suite from the previous versions), then the above is definitely one of the earliest.
To see a bit more detailed timeline, I'd recommend the following paper, which is a survey of regression testing techniques, and also happens to be written by yours truly: http://dl.acm.org/citation.cfm?id=2284813