It's not that Gregorian chant avoided the V-I cadence; it's that the chants were structured around entirely different concepts, and the V-I cadence did not yet exist.
There were eight Gregorian modes, each having an "authentic" and "plagal" version. Each authentic mode was rooted on the "final" (analogous to the "tonic") and extended up to the octave. The plagal versions shared the same final, but extended from a fourth below to a fifth above. The below image illustrates the modes and their finals.1

The Dorian, Hypodorian, and Mixolydian modes sometimes extended one pitch lower, which, given the structure of those modes, would have been a whole-step. Thus the only mode leaving room for a leading tone in the Tonal sense would have been the Hypolydian.
Wikipedia has a good summary of the Gregorian modes and how they functioned. It would be a good starting place for further reading. The article on Gregorian chant is also useful reading. In particular, it discusses the construction of chant melodies. Brittanica also has an illuminating article on Gregorian chant.
An SE question that might be of interest:
1From Wikipedia: By Benjamin D. Esham (bdesham) - Created by bdesham with GNU LilyPond and post-processed with GraphicConverter. The original version of this image was made by Hyacinth. The fs in the image are those designated by Curtis as "final". The sources listed at that image page are Judd, Cristle Collins (ed.) (1998). Tonal Structures of Early Music. New York: Garland Publishing. ISBN 0815323883 Liane Curtis. "Mode"., Public Domain, https://commons.wikimedia.org/w/index.php?curid=63101