Because it’s 100% synthetic and has no biologically sourced API, Copaxone is easier than Lovenox to reverse engineer to the lot-to-lot tolerance of the branded product.
Thanks for the cite. It might be true that there is less lot-to-lot variability in the Copaxone product but I am not sure that I would agree with conclusion - everything I've seen implies Copaxone has a much wider distribution of components in the mixture. I would think that would make it more difficult to reverse engineer and certainly more difficult to prove sameness (e.g. because you have to prove sameness for 1000 subcomponents instead of 100).
Nota bene - I am very much a neophyte on this topc, but ...