I would like to add some points about shutting down/pausing studies. It doesn’t take much to make it happen. The IRB has a lot of power in this matter.
My lab was in a “pause” mode for a few weeks and no one could touch the data or has any experiment running because of some paperwork violation when we used a few consent form that was not updated to the correct version.

If the study were to be shut down, I would think that it could be the IRB. I wouldn’t imagine it being the manufacturer unless some serious consequences happened such as many death, serious illness etc. because a study is always time-consuming and expensive, if some data could still be salvaged, the study wouldn’t be shut down.