Nested case-control studies: should one break the matching?

Nested case-control studies: should one break the matching?