Couple of points on adenovirus dosing...
Historically, virus and phage are quantitated based on their ability to infect. In the classical work with bacteriophage lambda, the concentration of the phage was quantified based on its ability to infect and lyse bacteria, thereby causing "plaques" or seemingly empty spots in a lawn of grown bacteria. This is generally why people refer to viral titres as plaque-forming units (PFU). Furthermore, in stark contrast to the mass-oriented measurements for normal drug administration ( x mg dose), PFU measurements are really related to the number of *molecules* of individual virus present in the sample... that's why we get to rather large numbers (to the 8th-12th powers) when discussing amounts. People often work with viral titres that vary sequentially by 10 fold differences in magnitude... so a 3000 fold difference between lowest and highest dose is really ~3 dose increments. Considered in that fashion, it is not unlike the number of dose increments that occur for traditional small molecule drugs in phase I studies.
In reading through some of dew's referenced posts, i noted one correction to an earlier suggestion: the adenovirus used by GenVec is replication defective.
At this point, hard to determine if GenVec has even a fledgling AMD program since they did not provide any data. Sigh...