The contact angle is specific for the surface orientation of a solid since every orientation has a different surface energy. If you increase the roughness of a solid, it means that you no longer have a defined crystal plane (assuming you even have a crystal). As a consequence, the resulting apparent contact angle will no longer be a single species, but it will be an envelope of multiple liquid-microstructure interactions.
In normal cases, for a hydrophilic substrate increasing the roughness will expose more reactive and thus even more hydrophilic side facets and this will make the contact angle decrease.
However, if the structure of your material has both hydrophilic and hydrophobic components, an increase would be imaginable. The first hit I had from a quick search is this theoretical work on strychninium crystals:
Article Hydrophilic and hydrophobic pockets at surfaces of strychnin...
I would suppose that if you disturb one of the hydrophilic surfaces by roughening, you could potentially see a contact angle increase, but I'm not sure that this is the scenario you had in mind.
Jürgen Weippert I work with UV-irradiated polymer films and have observed that the contact angle, indicative of surface hydrophilicity, increases with prolonged UV irradiation time and heightened surface roughness.
OK, that opens a whole spectrum of interpretations. An incomplete set of ideas:
changes in the molecular structure at the affected surface areas (just internally in the film)
along with the first point: formation of reactive sites at which water or another species can dock and form hydroxyl groups; although in the particular case of OH I would expect a decrease of the contact angle
formation of pockets with hydrophopic/hydrophilic local transitions as described in the strychnine paper
the added energy did actually not cause damage, but actually healed some defects which were contributing to the hydrophilicity and their absence increases the contact angle
There is definitely a lot more which did not come to my mind spontaneously.
Measurements of A/W surface tension by the Wilhelmy plate method is very sensitive to contact angle. A study of surface roughness and apparent tension has been reported in the literature some time ago. I think back in the '80s. I don't recall if they interpreted the results in terms of contact angle or not. If you can't find it, let me know and I think I retrieve the citation.