Disclaimer: I am not a medical professional and this is not medical advice.
We’ve come a long way since March of 2020, though frustratingly not long enough to be out of the woods. With the emergence of several highly effective vaccines, monoclonal antibodies, and upcoming oral antivirals, the need to find existing drugs that can be repurposed to treat COVID-19 is certainly nowhere near as urgent as it was last year. As the pandemic drags on, there are unfortunately more and more opportunities to collect data on cases and outcomes – and by now, the sex hormone hypothesis appears likely to be a dead end:
- In a large observational study from this month of 1,106 COVID-19 patients with prostate cancer who were or were not taking an androgen-blocking treatment, there was no significant difference found between these groups in disease severity or likelihood of death after 30 days (Schmidt et al., 2021). Although this isn’t a controlled experiment, it’s also not the outcome we would expect to see if lowering testosterone levels or blocking androgen receptors did have a beneficial effect on the course of disease.
- Another observational study of COVID-19 outcomes in prostate cancer patients found no difference in risk of death between those taking and not taking androgen deprivation therapy (Duarte et al., 2021).
- Additionally, in a study of records from the University of California healthcare system, cis women with conditions associated with elevated testosterone such as PCOS, hirsutism, or acne vulgaris were no more or less likely than other women to be infected, hospitalized, or die from COVID-19 – and although many of these women were taking spironolactone or estradiol to treat these conditions, neither medication was associated with any difference in infection or hospitalization risk (Yale et al., 2021). Sorry, transfems, no help here for us – but this does at least confirm that spironolactone does not increase the risk of infection or severe disease, a concern that was raised early in the pandemic.
- Even a previous finding that the androgen receptor blocker enzalutamide inhibited the entry of SARS-CoV-2 into cells in vitro was later contradicted by another study showing that it had no effect on viral entry in human lung cells and lung organoids (Li et al., 2021).
Meanwhile, clinical data from actual trials of sex hormone treatment for COVID-19 has been even harder to come by:
- The COVIDENZA trial of enzalutamide (Welén et al., 2021; NCT04475601) was subject to a safety pause and was later terminated before completion on the advice of a data and safety monitoring board.
- The RECOVER trial of bicalutamide (NCT04374279) was ended due to “limited resources”.
- The HITCH trial of the GnRH antagonist degarelix (NCT04397718) is currently ongoing.
- And a completed study of 40 men hospitalized with moderate to severe disease (NCT04365127) did find that subcutaneous injection of progesterone 100mg twice daily significantly improved clinical outcomes by day 7, reduced days spent on supplemental oxygen, and reduced days spent hospitalized (Ghandehari et al., 2021).
And now we come to proxalutamide (GT0918), an unreleased drug developed for the treatment of prostate cancer by blocking androgen receptors similarly to other nonsteroidal antiandrogens like bicalutamide and enzalutamide, while also potentially being more effective than existing antiandrogens due to its mechanism of actually degrading androgen receptors (Zhou et al., 2020). This, too, has been trialled as a COVID-19 treatment, and I posted earlier this year on the dramatic results from trials in which proxalutamide entirely prevented hospitalization and death.
It now appears that there are serious data (and ethical) integrity issues with these trials. First, all trials of proxalutamide for COVID-19 have been conducted by the same team led by Flavio Cadegiani of Applied Biology – there does not appear to have been any independent confirmation of these findings, or any other trials at all, by others outside this group, while other researchers have noted an almost unbelievably large reduction in death of 77% in one trial. It would be good to see these results replicated in more than one lab, particularly given that the results from this lab on proxalutamide (as well as dutasteride) appear to have suspicious features.
Nick Brown, a forensic data investigator who previously helped to uncover data fraud by Brian Wansink at the Cornell Food Lab, has found that the baseline characteristics of subjects in two studies by Cadegiani on proxalutamide and dutasteride have statistical features that are highly unusual and should not be present in a trial where subjects have been appropriately randomized. This analysis uses a technique that previously identified possible data fraud in over 5,000 published medical articles (Carlisle, 2017). As others have noted, the fatality rate of 49.4% in the control group of one proxalutamide study is abnormally high and causes the drug to appear very effective in comparison. I would love to see additional forensic analysis of the Cadegiani team’s work from anyone with more statistical knowledge and experience than I currently have – normal, well-done trials don’t throw up red flags like this, and this suggests there are more shoes yet to drop.
Aside from the data, it’s now alleged that the proxalutamide investigators did a lot of things you shouldn’t do to human beings. Taylor (2021) reports that patients in a proxalutamide trial in Brazil were not informed they were being given an experimental drug as part of a trial, meaning that consent was not properly obtained and participants were not made aware of the potential risks. Brazil’s National Health Council described these irregularities as the worst medical ethics violations in the country’s history, while an inquiry in parliament asserted that Cadegiani had committed crimes against humanity.
Additionally, Brazil’s National Research Ethics Commission stated that if proxalutamide was found to be so clearly effective, the trial should have ended as it would be unethical to continue to deprive the control group of access to what’s allegedly a miracle drug. Of course, such a concern would only be in play if the drug really is this effective, which is exactly what these data irregularities call into question. Epidemiologist Jesem Orellana is quoted as stating that “Everything about this trial is suspicious and anything but clinical and randomized”, while medical ethicist Arthur Caplan called these research practices “an ethical cesspool of violations, from consent and design to over-optimistic reporting of results and hiding deaths”. This is, by all appearances, thoroughly sketchy and just a mess – and it’s all we have to go on when it comes to proxalutamide.
Although a treatment this effective did seem too good to be true, we do know now that such a large effect on COVID-19 illness is indeed possible, as shown by Pfizer’s Paxlovid oral antiviral combination pill and its reduction of hospitalization or death by 89%. Its large effect itself, though cause for raised eyebrows, is hardly enough to conclude this is bunk – the statistical irregularities were what personally convinced me of that. At the same time, I saw the possibility of research misconduct in the proxalutamide trials as too bad to be true. Kintor, the maker of proxalutamide, continues to announce new and ongoing trials, as well as deals with Fosun (who are partnering with BioNTech in China to produce the Comirnaty mRNA vaccine) to distribute proxalutamide throughout massive markets in Africa and India. Would they really be continuing to produce and promote this on the basis of unethical studies and suspect data?
I get the distinct sense that we now know less about this drug, and I feel I could be genuinely surprised both if it turns out to work and if it turns out to be ineffective and fraudulent. What I want is for ethically, methodologically, and statistically sound studies to be done by research teams not associated with Kintor, Applied Biology, or Cadegiani and his colleagues. The decisions that are going to be made about the future of a potential COVID-19 treatment require a far more solid basis than is available right now. ■