A U-shaped dose response would, to me, imply substantially less credibility than the absolute numbers imply.
Agreed with a proviso: The U-shaped dose-response curve doesn’t bother me as much in this instance as it normally would because the MoA of Gammagard in AD is very poorly understood. It’s cases where the MoA is supposedly well understood and the dose-response curve is inconsistent with the MoA that I find the most troubling.
p.s. If IVIG actually pans out as a treatment for AD, it will be one of the biggest stories in the history of the drug/biotech industry, IMO.