Do Anti-androgens Have Potential as Therapeutics for COVID-19?

Coronavirus disease 2019 (COVID-19) is characterized by a gender disparity in severity, with men exhibiting higher hospitalization and mortality rates than women. Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), the virus responsible for COVID-19, infects cells following recognition and attachment of the viral spike glycoprotein to the angiotensin-converting enzyme 2 transmembrane protein, followed by spike protein cleavage and activation by cell surface transmembrane protease serine 2 (TMPRSS2). In prostate cancer cells, androgen acting on the androgen receptor increases TMPRSS2 expression, which has led to the hypothesis that androgen-dependent expression of TMPRSS2 in the lung may increase men's susceptibility to severe COVID-19 and that, accordingly, suppressing androgen production or action may mitigate COVID-19 severity by reducing SARS-CoV-2 amplification. Several ongoing clinical trials are testing the ability of androgen deprivation therapies or anti-androgens to mitigate COVID-19. This perspective discusses clinical and molecular advances on the rapidly evolving field of androgen receptor (AR) action on cell surface transmembrane protease serine 2 (TMPRSS2) expression and severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection, and the potential effect of anti-androgens on coronavirus disease 2019 (COVID-19) severity in male patients. It discusses limitations of current studies and offers insight for future directions.

As of June 2021, with over 170 000 000 confirmed cases of COVID-19 worldwide, the percentage of confirmed cases is equal among men and women. However, among more than 3 500 000 deaths globally, for every 10 female confirmed cases that have died from COVID-19, there are 15 males (https://globalhealth5050.org/the-sex-gender-and-covid-19-project/the-data-tracker/). Several mechanisms have been proposed to explain this relative female protection from severe COVID-19 outcomes, with the most likely being related to women’s more robust immune response to viruses (1).

SARS-CoV-2 entry and infection of cells is mediated by recognition and attachment of the viral spike glycoprotein to the angiotensin-converting enzyme 2 (ACE2) transmembrane protein on host cells, followed by spike protein cleavage and activation by TMPRSS2, which facilitates membrane fusion and entry. In the prostate, TMPRSS2 expression is upregulated by androgens via the AR, which has led to the hypothesis that androgen-dependent expression of TMPRSS2 in nonprostatic tissues, especially the lung, may be instrumental to men’s increased susceptibility to COVID-19 severity and mortality. Accordingly, investigators have examined the relationship between androgen depletion and COVID-19 in patients with prostate cancer, as androgen deprivation therapy (ADT) consisting of luteinizing hormone-releasing hormone agonist/antagonists (blocking testosterone production) or AR inhibitors/antagonists (blocking testosterone action) is the standard of prostate cancer treatment.

The first population-based study of the Italian region of Veneto, which was severely affected by the COVID-19 pandemic, looked at patients with prostate cancer and confirmed SARS-CoV-2 infection. They compared 5273 patients receiving ADT, with 37 161 patients not treated with ADT. They reported that patients receiving ADT exhibited a lower risk of SARS-CoV-2 infection than those without ADT (odds ratio [OR] 4.05, 95% CI 1.55e 10.59) (2). The study had limitations: the clinical and biological characteristics of the patient populations with and without ADT were not described and were likely to be different, and the analysis was not corrected for multiple variables. A second study examined 58 patients with confirmed SARS-CoV-2 infection and prostate cancer at Mount Sinai Health System in New York City (22 patients were on ADT and 36 were not) (3). After controlling for multiple variables, they concluded that ADT was associated with reduced odds of hospitalization (OR 0.23, 95% CI 0.06-0.79, P < 0.02), supplemental oxygen requirements (OR 0.26, 95% CI 0.07-0.92, P = 0.036), need for intubation (OR 0.31, 95% CI 0.05-1.81, P = 0.192), and mortality (OR 0.37, 95% CI 0.08-1.80, P = 0.22). However, the study lacked power, and 2 outcomes (intubation and mortality) were not significant. Another Italian study focusing on 36 SARS-CoV-2–positive patients with metastatic prostate cancer and receiving ADT did not find any increased mortality rate compared to the Italian SARS-CoV-2–positive male population of the same age (4). A retrospective analysis compared male subjects with confirmed SARS-CoV-2 infection, and androgenetic alopecia (AGA), treated with 5-alpha-reductase inhibitors (5AIs), which block testosterone conversion to the potent AR agonist dihydrotestosterone, to men with AGA nonusers of 5AIs. They observed a significant reduction in the frequency of clinical symptoms in men with AGA using 5AIs compared to men with AGA not using 5AIs (5). Based on these discordant observational studies and the absence of randomized trial, no definitive clinical evidence indicates that ADT prevents COVID-19 infection or mitigates COVID-19 severity. Thus, several randomized clinical trials (RCTs) have begun to test the efficacy of ADT in the general population of male COVID-19 patients. The most interesting to date, a RCT of proxalutamide (a novel generation AR antagonist; NCT04446429) in Brazilian outpatients with COVID-19, concluded that subjects assigned to the drug (0/134) benefited from a reduced rate of hospitalization compared to those assigned to placebo (35/128). The study has not been peer reviewed yet. A second RCT of 107 men and 128 women with mild to moderate COVID-19 reported that proxalutamide accelerates viral clearance, with 82% of subjects testing negative for SARS-CoV-2 in the treatment group at day 7 compared to only 31% in the placebo group (P < 0.01) (6). In addition, the time to clinical remission was improved with a 5-fold decrease in the average number of days required to be COVID-19 symptoms free in the proxalutamide group vs the placebo. Additional ongoing RCTs, include degarelix (gonadotropin-releasing hormone antagonist; NCT04397718, completion date July 2021), bicalutamide (first-generation anti-androgen; NCT04374279; completion date January 2022), and enzalutamide (second-generation anti-androgen; NCT04475601, completion date July 2021).

In parallel to these clinical studies, investigators have begun to explore the molecular underpinnings of AR antagonism regarding TMPRSS2 expression and SARS-CoV-2 infection using prostate and lung cell models. Two studies identified coexpression of ACE2, TMPRSS2, and AR in human alveolar and bronchial epithelial cells, suggesting that AR drives ACE2 and TMPRSS2 expression in the lung (7,8). Indeed, in male C57 mice, TMPRSS2 is upregulated by testosterone in bronchial cells. In human prostate LNCaP cancer cells, treatment with AR antagonists approved for the treatment of prostate cancer (apalutamide, darolutamide, and enzalutamide), all inhibited SARS-CoV-2 infection (7). Unfortunately, the effect of AR antagonists was not tested in lung cells, as the authors were unable to identify a lung epithelial cell line susceptible to SARS-CoV-2 infection that also had adequate AR signaling (7). Still, in male human AR-expressing lung carcinoma cells (H460), treatment with androgens increased, and the anti-androgen enzalutamide or the AR degrader ARD-69 decreased TMPRSS2 and ACE2 messenger RNA and protein levels (9). Using a replication-defective SARS-CoV-2 pseudovirus, they found that androgen-deprivation, enzalutamide, and ARD-69, all reduced SARS-CoV-2 pseudovirus entry into LNCaP prostate and H460 lung cells (9). However, although enzalutamide reduced the entry of authentic SARS-CoV-2 into LNCaP prostate cells, surprisingly, the authors did not test the effects of anti-androgens on the SARS-CoV-2 entry in lung cells (9). Another more translational study used human lung organoids derived from normal human lung tissues and expressing AR and TMPRSS2. They used a similar SARS-CoV-2 pseudovirus as previously described. However, in contrast to the previous study, in these human lung organoids, enzalutamide showed no effect on TMPRSS2 expression or on infection by pseudovirus or authentic SARS-CoV-2 (8). Similarly, in human lung cancer cell lines (H1437 and H2126 cells), dihydrotestosterone and enzalutamide produced no change in TMPRSS2 expression and did not prevent SARS-CoV-2 pseudovirus infection. The authors next used mouse models to examine the role of AR in lung TMPRSS2 expression and SARS-CoV-2 infection in vivo, delivering human ACE2 with a replication-deficient adenovirus (adenovirus-expressing human ACE2) (8). They observed no effect of either castration or enzalutamide on SARS-CoV-2 infection or inflammation of the lungs. To identify differences between lung and prostate with regard to AR action on SARS-CoV-2, the authors performed AR chromatin immunoprecipitation with sequencing (to determine AR binding sites) and assay for transposase-accessible chromatin using sequencing (to assess chromatin accessibility at AR binding sites) in AR-positive lung compared to prostate cells. Prostate cells exhibited the reported specific AR binding sites, with open chromatin at the TMPRSS2 locus, and robust AR binding. In lung cells, however, they observed no AR binding at the TMPRSS2 locus, suggesting that failure of castration and enzalutamide to prevent SARS-CoV-2 infection of the lungs is due to the lack of AR binding to the TMPRSS2 promoter in lung cells. Notably, when transduced with Ad-ACE2, TMPRSS2-negative H23 lung cells were still permissive for robust SARS-CoV-2 infection, also suggesting that other factors besides TMPRSS2 can promote SARS-CoV-2 infection.

How do we interpret and integrate the information provided herein, and what additional studies are needed to address the question of anti-androgens potential as therapeutics for COVID-19? A few answers are discussed next.

RCTs of proxalutamide in outpatients with COVID-19 suggest that AR antagonism accelerates viral clearance and reduces rate of hospitalization (6). However, if androgens enhance TMPRSS2 expression in lung cells to promote higher infection of male compared to female cells, one would expect men to be more susceptible to SARS-CoV-2 infection than women. Infection rates are similar between sexes globally. In the initial Veneto study reporting that men receiving ADT exhibited a low risk of SARS-CoV-2 infection, surprisingly women exhibited higher prevalence of infection than men (44% men vs 56% women) (2). In the near future, results from the RCTs using gonadotropin-releasing hormone antagonist and other AR antagonists will provide the necessary information regarding the putative therapeutic effects of AR antagonism in COVID-19.

Molecular studies are consistent with the efficacy of AR antagonists to decrease TMPRSS2 expression and prevent SARS-CoV-2 infection in rodent and human prostate. In human lung cells, however, the effects of androgens and AR antagonists on TMPRSS2 expression and SARS-CoV-2 infection are not consistent (7), and AR antagonists show no effect in preventing SARS-CoV-2 infection of the lung, suggesting that the lung is not the target of AR. Lung endothelial cells express AR, and evidence suggests that endothelial cells are central in COVID-19 (10). The possibility that androgens target endothelial cells and that AR antagonists displays SARS-CoV-2 antiviral activity by altering TMPRSS2 expression in endothelial cells should be investigated. In addition, androgens affect male immune responses via AR in myeloid and lymphoid cells (11). Androgens increase circulating neutrophils and decrease antibody response to viral infections. Therefore, androgens may predispose to severe COVID-19 by altering immune responses, and ADT may protect through effect on immune cells. Further studies are also needed to define the role of AR in immune cells in COVID-19 severity.

Finally, testosterone seems to play a bidirectional role in COVID-19 pathogenesis. Most men with COVID-19 exhibit testosterone deficiency, and low testosterone at admission was associated with intensive care admission and death in multivariate analysis. Evidence suggests that low testosterone levels actually worsen COVID-19 severe outcomes, as lower testosterone concentrations during hospitalization are associated with increased inflammation and mortality in men (12,13). The X-linked AR gene contains a highly polymorphic N-terminal polyglutamine (polyQ) tract, ranging from 9 to 36 polyglutamine coding cytosine-adenine-guanine repeats, with the polyQ tract-length being inversely correlated to AR functionality. A case-control study examined the genotypes of 638 male and female Italian COVID-19 patients, comparing those with severe COVID-19 with those who were oligo-asymptomatic. Among males, they found an association between length of the AR polyQ tract and disease severity, with shorter alleles (predicted to increase AR action) appearing to protect against worse clinical outcomes independent of age (14). In contrast, men with long polyQ repeats (predicted to decrease AR action) exhibited increased biomarkers of inflammation. They suggest using testosterone as adjuvant therapy for patients with severe COVID-19 exhibiting low levels of circulating androgens and defective AR action, defined in this study as >23 PolyQ repeats. Therefore, the potential bidirectional modulation of COVID-19 severity by testosterone and AR action require further investigation.

In conclusion, to what extent antiandrogens mitigate COVID-19 severity in men requires further demonstration and mechanistic study. While we wait for the results of clinical trials testing the efficacy of ADT and anti-androgens in COVID-19 outcomes, studies are also needed to assess the effect of AR action in endothelial and immune cells in preventing or mitigating SARS-CoV-2 infection.