There are two possibilities: a) UCI covered up a suspicious passport; b) USADA is trying to make a case out of a suspicious passport that was already vetted by the expert panel. I think 2) is much more likely.
Roughly, there are three levels of passports: 1) a normal one that falls within the acceptable parameters; 2) a suspicious one that trips the algorithms; on a graph, this is observed as a plot that extends above or below the baseline plots; 3) a suspicious plot that is examined by experts in more detail, and concluded to indicate doping.
Most likely, LA’s blood values fell in the 2) range. Either UCI ignored it (a), or they passed them on to the expert panel, which concluded that they were not sanctionable (b). My understanding is that most suspicious passports are in fact rejected as evidence of doping by the panel, in an efforty to minimize false positives. Of course, experts may disagree, and USADA may have a panel of its own experts that has concluded LA’s values are sanctionable. This would be really interesting, but in that case the values would still most likely be supporting evidence, not standalone evidence of doping.
It wouldn’t be a protease, as the amino acid sequence of synthetic and natural are the same. It might be a sialidase, which selectively degrades the carbohydrate residues on synthetic, but I doubt very much that there is an enzyme that is selective for one and not the other. It could be an enzyme directed to one of the newer synthetic EPOs, though, that removes the residues specific to it.
Nothing suspicious about that. Many urine samples show no detectable levels of natural EPO.