REGN/SNY—FDA’s statistics reviewer is critical of dataset’s inability to quantify efficacy difference between low and high doses.