From my experience with dialogue editorial, you're usually SOL when it comes to this - the problem being that handling noise tend to fill all frequencies. Depending on how the frequency content is from what you're hearing, a hi pass may help you, automated of course, but even then usually such noise sits too high into the mid-lows and mids (like a plosive) that the rolloff will then have a huge affect upon your actual sound content. The other thing is to cut it out piece by piece, find replacement words, syllables, etc with alts (or even stolen from other lines) as needed. This is one of the hallmarks of successful dialogue editorial, done carefully to be as transparent as possible.
There is no such thing as a free lunch for getting good results with a dialogue track, especially so with a case like this. The more well-edited and strategic those edits are, the more success there will be with processing options on the stage. So my guess is that your best bet will be (assuming there's no ADR options) rolling up your sleeves and doing some creative edits.