In the Brazil trial, 645 hospitalised Covid-19 patients who were breathing on their own received either proxalutamide for 14 days or a placebo, plus standard care. After two weeks, recovery rates were 81.4 per cent for the proxalutamide group versus 35.7 per cent for those who got a placebo.
After four weeks, 49.4 per cent in the placebo group had died, versus 11 per cent with proxalutamide.
Benefits were similar across genders, even though women might not be expected to respond to drugs that block male hormones, McCoy said.
The study was done during a Covid-19 surge in a Brazilian state where the Gamma variant, formerly known as P1 and first discovered in Brazil, was dominant.
The mortality reduction could be even higher in other settings where infection rates are lower and that variant is not predominant, said coauthor Carlos Gustavo Wambier of the Alpert Medical School of Brown University.
"We still don't know," he said. "Someone has to initiate an investigation there.