The assumption that Sr-90 and childhood cancer are correlated is best supported for a supralinear dose-response, meaning the greatest per-dose risks are at the lowest doses.
"The absence of a dose-response curve was disappointing, " says Jeffrey Cummings of UCLA. Moreover, he noted that the drug only very modestly reduced levels of a nasty protein called tau in the spinal fluid.
Nevertheless, while the primary endpoint in the PRE-SURGE 2 trial does not appear to have any dose response, as you noted, the endpoints in PRE-SURGE 1 and the secondary endpoints in PRE-SURGE 2 appear to trend slightly better with the 160 mg dose.