Re: Chen Dong, Sung-Lang Chen, and Wen-Wei Sung's Letter to the Editor re: Karin Welén, Ebba Rosendal, Magnus Gisslén, et al. A Phase 2 Trial of the Effect of Antiandrogen Therapy on COVID-19 Outcome: No Evidence of Benefit, Supported by Epidemiology and In Vitro Data. Eur Urol. 2022;81:285-93.

We read with interest the letter from Dong et al regarding our three-pronged study [1]. We use a three-dimensional model of primary human lung cells to evaluate the antiviral effect of enzalutamide against SARS-CoV-2. We agree that our experimental data do not give the final answer as to whether an antiandrogen-mediated effect on SARS-CoV-2-infected lung cells is possible. However, we have verified the expression of the androgen receptor (AR) at the mRNA level in these cells, and positive controls in the lung setting are difficult to define. Dong et al suggest that antiandrogens may have an indirect effect on lung cells via inhibition of the production of AR-regulated cytokines in the prostate and recommend studying these in blood to clarify the underlying mechanism. However, the critically ill patients in the present study are not suitable for that purpose. Furthermore, in terms of indirect effects, we believe that it is more likely that inhibition of AR on immune cells would have an impact on the lung environment in COVID-19 patients.

The SARS-CoV-2 we used was isolated in Sweden early in the pandemic (GenBank accession number is provided in the Supplementary material [1]). When the clinical trial was conducted, the first variant dominated in Sweden, while the epidemiologic study also includes a time period dominated by the alpha variant. To the best of our knowledge, no data on different mechanisms for viral entry or changes regarding the male-dominated disease profile have been presented, and thus we believe that the role of different virus variants on the data presented is minimal.

Regarding the clinical trial, two issues were discussed by Dong et al. The first relates to the different nature of the grade 3 adverse events (AEs) reported for the two arms of the trial. Gastritis and urine retention were reported as AEs for both groups (Supplementary data [1]) although the severity differed. Furthermore, the majority of the grade 3 AEs in the treatment group occurred after more than 20 d in hospital, a time point by which all patients in the control group had left the hospital, so comparisons of the nature of these AEs are impossible.

The second issue discussed was a possible impact of differences in baseline parameters between the groups on their respective outcomes. There were no statistically significant differences between the groups regarding any of the parameters listed in Table 1 [1]. In a sensitivity analysis, we separately excluded diabetes cases (8 in the treatment group, 0 in the control group) and cases with an ordinal scale score of 5 (4 in the treatment group, 1 in the control group). This did not change the median times for hospital stay. It also did not affect the hazard ratio regarding hospital stay disfavoring the enzalutamide group, although removal of cases with an ordinal scale score of 5 increased the p value to 0.059. Even though some less obvious residual confounders may be present, we believe that it is highly unlikely that they would influence the results enough to even out the worse outcome for the treatment group or alter our conclusion that no benefit of enzalutamide was observed for hospitalized COVID-19 patients.

: Andreas Josefsson has received an unconditional research grant for COVIDENZA from Astellas Pharma and honoraria from Astellas, Ipsen, Sandoz, and Janssen-Cilag. The remaining authors have nothing to disclose.