Yeah, "Stop when the results are positive" is one of the classic ways to fudge a study. Twelve patients (the number mentioned in the article) is way, way premature to end this early.

In this case, it's not even the regulatory environment. As the article mentions, the treatment is as fantastically expensive as it is fantastic. A lot of the skepticism isn't that it can't work, but that it might be too expensive to be viable.