This is not about the cost of withdrawing or barring drugs as much as it is about the cost of running all these tests and trials. And yes, drugs can potentially have terrible side effects, but not being able to afford the drug can also have terrible results.

If the FDA requirements were much less strict, the drug company would have had to spend much less on R&D. That, of course, would not be enough to lower the price - but the other effect of cheaper R&D is that it’s easier for other companies to compete, and competition does drive prices down. The the point either the mother could either afford it herself, or the insurance wouldn’t be so stingy about paying it.

Now, of course, less strict requirements also mean we know less about the drug’s safety and efficiency. Let’s say that, because of the lack of knowledge, we assign a 50% probability for the drug to kill the patient and even if it doesn’t we only assign 50% probability for it to work (that does not mean it killed half the test subjects and failed for half of the remaining ones - just that we didn’t test enough to get significant results that say otherwise, and these are the worst case estimates under our lack of data). That means, that there is only 25% chance for each one of these twin babies to survive if they take the drug.

Which is better than the 0% they get now, being unable to afford it.