Remove/reduce live audio echo

I have an audio file of an interview with a person. Unknown at the time we recorded, the mic was picking up his voice out of a speaker and thus there is an echo throughout.
 
I'm looking for ideas on how to get rid of the echo. Things I have tried, but which had limited results, include:
 
1) EQ on a very narrow freq using a high Q (better but not good enough).
 
2) Noise gating

I tried a demo of Sound Soap once and it does amazing things with noisy background. But I think your issue might be impossible to clean up. Noise removal software looks for,, well noise. 
A repeated word is not really noise. 
Is it only noticeable in the pauses between words? 
If so if a noise gates threshold can't nail it then you might be faced with manually running through the track and silencing the pauses. 
I've done using Wave Lab. 
 
You can almost see the places where the pauses and breath is with a wave editor set at the right zoom level. All you need to do is highlight the offending space and hit backspace and it will silence that part. If the spaces are large I also cut time out of them. 
 
You can do wonders with interviews by tightening up the spaces. A 10 minute speech, sermon can be cut down to 8 min etc. Cut out all the pause to think words like --you know and uhhhh ..you can make someone sound more confident.