>> Seems to me the EMA does this sort of thing better (or at least in a much more co-operative spirit).
No, EMA doesn't do any better. It seemed that way because there is no formal rejection for this type of issues in EMA review process and companies don't feel compelled to disclose 120-day, 180-day, xxx-day questions - some companies do disclose on inconsistent basis - thus no one knows for sure when the review clock is actually stopped when the clock is actually started. I would prefer EMA review process if at least there is mechanism to let public know when review clock is started or stopped and at what point. If this type of CMC issue were to happen in EMA review process, companies would get xxx-day letters, and no one would know what those questions would be because few would disclose them.
BTW: The canister filling process improvement was in previous CRL too.
Market Data 
Markets 
AI
