mRNA "dosage". Dew, PGS is right, there is often not a linear relationship between mRNA levels and levels of the protein that is encoded by the mRNA. A couple things to consider here. The cell uses feedback mechanisms to regulate protein production at the translational level, in addition to the transcriptional level. Mature, polyadenylated mRNA is quite stable in the cell and can be sequestered in the cytoplasm and used later for the rapid regulation of local protein synthesis through translational activation by biochemical signaling cascades. As PGS points out, a single copy of mRNA can be use to make multiple copies of a protein (a typical 50KD protein takes about 1 minute to translate). Furthermore, much of the mRNA in a cell normally never gets translated; its just there in case it is needed.
Another factor to consider is that mRNA is edited post-transcriptionally and thus there may be multiple forms of mRNA encoding the same or very similar isoform of protein that are resistant to the siRNA.
Finally there is the more obvious: in the tissue targeted by the siRNA, you probably will not achieve 100% "transfection" efficiency. Thus, the "untransfected" cells may work harder at protein production in an attempt to compensate for reduced serum levels. All these factors make siRNA dosing studies pretty complicated when you really think about it.
To my knowledge, there are no approved products for treating humans with siRNA. In fact, are there any approved drugs which target RNA other than antibiotics? Does anyone have a sense of the regulatory hurdles that will have to be overcome with the FDA for say, ocular delivery of siRNA therapeutics?
The obstacle to discovery is the illusion of knowledge.