Create .wav or .mp3 from mix of pre-recorder and text-to-speech

We need a Windows application that will take pre-recorded voice files (questions) and add text-to-speech (answers) and combine them into one .wav or .mp3 file. The text-to-speech needs to be in Spanish and consists of reading a name, yes/no answers to the questions, and either a maiden name or birthdate. The completed file needs to be filtered to make it sound authentic (best text-to-speech quality possible). The text-to-speech needs to include different "voices" as well. The text-to-speech "answers" can be provided in a method of your choosing (database, xml file, text file, etc.)