It depends on how well studied the treatment is. If it's been on the market for 70 years, I'd take the chance. Otherwise, why risk waking up one morning with a tail, or having my arm fall off ;)
Let's revisit which medications have harmful side effects a century from now. It's impractical to run studies with thousands of patients, and we haven't had more than half a century of data on most drugs. Not that there's much interest in replicating studies anyways.
I'm not keen to pop a bunch of pills unless there is significant (absolute, not relative) benefit. Drugs are overprescribed in my part of the world, partly because people assume "cuts risk by 50%" must mean "a lot." It might... it might not.