You can describe alphabetic characters using the inclusive regex matcher:
Note that the Computed Column function REGEXP only replaces matching strings, simply duplicating the input if it does not match. Therefore, to make your step bulletproof you need to produce a result that is guaranteed not to be identical to the input. For example, create a Computed Column that creates a column named "Valid First Name":
This produces a value where the input is prefixed with an underscore _, marking that value so you know it's different from the input. If it marks, then you know there was an alphabetic character; otherwise, there were no alphabetic characters.