Almgren's Big Regularity Paper appeared as a monograph in 2000, so not quite ten years ago, but still fairly recent. The results contained in it were 'new' in the sense that they had not been published before, although they had been in circulation as a preprint since 1984.

I'll add a few words about the history of the monograph. Almgren proves that an area-minimizing current, of arbitrary codimension, has a 'small' singular set: the portion of the surface where it is not smoothly embedded is a codimension two subset. (So if the current is $n$-dimensional, then the singular set has dimension at most $n-2$.) The proof is notoriously difficult, and some of the tools that Almgren developed—multivalued harmonic functions and monotonicity of frequency, for example—have found applications elsewhere.

The original preprint was 1728 pages long, which made it too long to publish in virtually any journal. Almgren was 'exploring the possibility of making it available on the web', but passed away in 1997 as a result of myelodysplasia. Two former doctoral students of Almgren, Jean Taylor and Vladimir Scheffer, posthumously published the monograph containing the result. (Jean Taylor was also Almgren's wife.) The book that eventually appeared in 2000, although shorter than the original paper, is still a hefty 970 pages.