How large should the sample be before it's substantive?
The British polling institute advises a sample of 1000 for a total UK population of around 60,000,000. Trans people make up 1% so roughly 600,000, with that in mind 57 doesn't seem that low?
The required sample size doesn't depend on population size that much.
Bigger than 57. Bad power is bad power whether or not it's the best you can get.