A SciFinder search revealed 3 patents with published syntheses of the drug; surprisingly, no papers so far with SAR studies or anything of the sort, and just enough clinical trials to allow the drug to be marketed. The drug wasted no time getting to the market - at least, according to the literature I looked through - which for a mental disorder like schizophrenia for which there is no truly good treatment, only adequate treatment, is wonderful.
Two patents had two different approaches (and the others were either duplicates or elaborations), and the most recent was an elaboration on one of the others. The first, from 2006, sought to make the trans-pyrrolidine as efficiently as possible to improve upon the former synthesis. The synthesis on a whole, though, is extremely steppy. The second, in 2008 (Int. Pat. #: WO 2008/003460) instead chopped the molecule in two via a cyclization between a stilbene and an in-situ generated azomethine, followed by an Ullman condensation to connect the phenol with the other ring; use of the trans-stilbene naturally leads to the trans-ring juncture.
To make the stilbene, HWE: