The peer reviewed paper showed that the mutant type of SIGMAR 1 has a 25% reduction in the rate of decline. That's enough drug effect for that group to be included in the labeling.
That was not a stat sig result. That is most likely due to the small n in that subgroup.