(Note: see Questions about primary keys 05 April 2002)
Assuming that a "more unique" code than the ISBN is even possible, it follows that there would be a one-to-many relationship from each such code to the ISBN numbers of the "same" book. Your examples make sense, though, at least to me, but I hasten to point out that I don't know book publishing, or the conditions under which a "new" ISBN is issued to the "same" book.
worldwide code ISBN description 00000101 1230001234 hardback edition 00000101 1230005678 paperback edition 00000101 1230000937 UK version
From a database design perspective, what would you use as the primary key? In my opinion, the ISBN would still be the PK. The "worldwide code" would be a foreign key, although you might not have a separate table for it. In any case, you can see that to implement this many-to-one relationship (as seen from the perspective of the ISBN) all you need to do is add the new worldwide code column to the existing table that uses ISBN as its primary key.
By the way, have a look at The Digital World and the Ongoing Development of ISBN, part of the isbn.org web site, which I discovered when researching this question. It says that the ISBN will be expanding from 10 digits to 13 in 2005. Not as monumental a change as Y2K, but you may want to start making plans...
This was first published in April 2002