Clinical tests of drugs known to cause harm

[JellySaurus]

Has there ever been a clinical test of drugs known to cause greater harm than good? Sodium valproate and thalidomide spring to mind - once it was recognised that these drugs were causing irreversible harm to children, were clinical studies set up to assess their value, or were they stopped in populations that could be harmed by them?

Thalidomide had such severe harms it was very rapidly withdrawn as an antisickness agent. It's still used as treatment of myeloma with very strict pregnancy prevention measures.

Sodium valproate has been around for a long time - the harms to a foetus weren't recognised early on, then we had a period were they were known but not robustly communicated to women patients. The very strict pregnancy prevention requirements have happened due to public campaigns- there are problems with the data though, so it's still being argued about.

Doing a clinical trial of any drug in pregnancy is very, very tough and rarely done unless your treating something life threatening (eg new surgery on a usually fatal foetal problem) Most safety data is retrospective and a lot of drugs aren't used due to lack of safety data.

The acceptability of a trial of a drug with known side effects really depends on the condition you're treating - if it's a cancer with high mortality and no current treatment it would be acceptable to do very experimental therapy on 'compassionate grounds' - with current best treatment you'll be dead soon anyway so it's ethically acceptable to offer you something with high risks and/or little data in humans. For example, drilling a catheter into your brain to give chemo directly to a tumour which usually kills you in a few months.

If it's a low risk condition (say heartburn) then the threshold for acceptable harm from a new treatment would be virtually zero.

Gender dysphoria in children I would put at the low risk end of the spectrum- most kids grow out of it post puberty so the justification for any medical intervention is very weak.

Sodium valproate effects babies in the womb. That didn't become clear until much later. It is still an important drug used for epilepsy and bipolar disorder. If you are a woman of childbearing age you have to be using contraception and sign all sorts of forms each year to show you understand. It has to be signed off by 2 consultants. So they are very strict.

Those drugs are valuable for life-threatening conditions. Their harms in those cases are outweighed by their benefits. But in situations where the benefits are outweighed by their harms - ie pregnant women - they were withdrawn once their harms were recognised. Not tested further in pregnant women.

Doing a clinical trial of any drug in pregnancy is very, very tough and rarely done unless your treating something life threatening (eg new surgery on a usually fatal foetal problem) Most safety data is retrospective and a lot of drugs aren't used due to lack of safety data.

Why does this not apply to children, too?

https://www.bbc.co.uk/news/world-africa-14493277