Text-to-Speech – eLearning

I am using the text-to-speech component for an eLearning course and have noticed that some words are mispronounced. When I adjust the spelling phonetically to correct the pronunciation, the changes also appear in the closed captions. Is there a workaround for this?

For example, the word ‘Workqueue’ is mispronounced by the English text-to-speech agent. I’ve attached an image for reference.