Proxalutamide Reduction of Mortality Rate in Hospitalized COVID-19 Patients Depends on Treatment Duration – an Exploratory Analysis of the Proxa-Rescue AndroCoV Trial

  1. Ricardo Ariel Zimerman
  2. Daniel do Nascimento Fonseca
  3. Michael do Nascimento Correia
  4. Renan Nascimento Barros
  5. Dirce Costa Onety
  6. Karla Cristina Petruccelli Israel
  7. Emilyn Oliveira Guerreiro
  8. José Erique Miranda Medeiros
  9. Raquel Neves Nicolau
  10. Luiza Fernanda Mendonça Nicolau
  11. Rafael Xavier Cunha
  12. Maria Fernanda Rodrigues Barroco
  13. Patrícia Souza da Silva
  14. Raysa Wanzeller de Souza Paulain
  15. Claudia Elizabeth Thompson
  16. Andy Goren
  17. Carlos Gustavo Wambier
  18. Flávio Adsuara Cadegiani

Reviewed by

Read the full article See related articles

Listed in

Log in to save this article

Abstract

Introduction

Proxalutamide, a second-generation non-steroidal antiandrogen (NSAA), primarily developed for castration-resistant prostate cancer, demonstrated reduction in 28-day mortality rate of 77.7% in hospitalized COVID-19 patients in a double-blind, placebo-controlled, two-arm randomized clinical trial (RCT), through intention-to-treat (ITT) analysis. We observed a high 28-day mortality rate of patients that did not complete the 14-day treatment with proxalutamide, compared to the placebo arm. These differences may raise hypotheses to explain the wide differences between ITT and on-treatment (OT) analysis in terms of efficacy. Despite the inherent limitations of OT analysis, we aimed to respond to unanswered questions regarding the drug efficacy when the 14-day treatment with proxalutamide was complete, and secondarily understand the causality relationship between treatment interruption and mortality rate.

Methods

This is a post-hoc exploratory analysis of a double-blinded, randomized, placebo-controlled, prospective, multicentric, two-arm RCT of 300mg-daily 14-day proxalutamide therapy for hospitalized COVID-19 patients not requiring mechanical ventilation. OT population excluded patients that did not complete the full 14-day course of therapy or died from COVID-19 complications within 24 hours of randomization. The primary outcome was the 28-day COVID-19 mortality rate. Secondary outcomes included median hospital length, 14-day and 28-day alive hospital discharge rate and 28-day all-cause mortality rate of those who discontinued intervention.

Results

In total, 580 patients completed the 14-day treatment or died during treatment, including 288 patients in the proxalutamide arm and 292 patients in the placebo arm, with similar baseline characteristics between groups. The 28-day COVID-19 mortality rate was 4.2% in the proxalutamide group and 49.0% in the placebo group. The mortality risk ratio (RR) was 0.08 (95% CI, 0.05-0.15), with a number needed to treat (NNT) of 2.2 to prevent death. The median hospital length stay after randomization was 5 days (interquartile range [IQR] = 3 to 7.2 days) in the proxalutamide group and 9 days (IQR = 6 to 15 days) in the placebo group (p <0.001). The 28-day all-cause mortality rate of patients that received proxalutamide but interrupted treatment before 14 days was 79.3%, while those that received placebo and interrupted before 14 days was 52.8% (p = 0.054 between groups).

Conclusion

The reduction in 28-day all-cause mortality rate with 14-day proxalutamide treatment for hospitalized COVID-19 patients was more significant while on treatment adhesion (92%), compared to the reduction when all patients enrolled in the proxalutamide arm were considered (77.7%). However, the magnitude of statistical significance of the reduction in all-cause mortality and the NNT were similar between the OT and ITT analysis. The apparent high mortality risk rate with early interruption of proxalutamide treatments suggests that strategies for treatment compliance should be reinforced for future RCTs with proxalutamide. ( NCT04728802 )

Article activity feed

  1. Version published to 10.1101/2021.06.28.21259661v2 on medRxiv
  2. ScreenIT

    SciScore for 10.1101/2021.06.28.21259661: (What is this?)

    Please note, not all rigor criteria are appropriate for all manuscripts.

    Table 1: Rigor

    Ethicsnot detected.
    Sex as a biological variable, immunosuppression, baseline alanine transferase (ALT) above five times ULN (> 250 U/L), creatinine above 2.5 mg/ml, and not currently using antiandrogen drugs, and, in case of women, not pregnant or planning to become pregnant within 90 days and nor breastfeeding.
    RandomizationThis is a post-hoc analysis of a double-blinded, randomized, placebo-controlled, prospective, two-arm RCT22 describing the results of the RCT restricted to OT analysis, i.e., excluding patients that discontinued treatment before 14 days of drug treatment.
    BlindingA detailed description of the trial design, sample size calculation, settings, recruitment, patient selection criteria, randomization, blinding, procedures and statistical analysis are published in another manuscript and detailed elsewhere.
    Power AnalysisA detailed description of the trial design, sample size calculation, settings, recruitment, patient selection criteria, randomization, blinding, procedures and statistical analysis are published in another manuscript and detailed elsewhere.

    Table 2: Resources

    Software and Algorithms
    SentencesResources
    Statistical analysis was performed in Stata/SE version 16.1 for Mac (StataCorp LLC, College Station, TX, USA).
    StataCorp
    suggested: (Stata, RRID:SCR_012763)

    Results from OddPub: We did not detect open data. We also did not detect open code. Researchers are encouraged to share open data when possible (see Nature blog).

    Results from LimitationRecognizer: We detected the following sentences addressing limitations in the study:
    This study has limitations inherent to post-hoc and retrospective analyses, although all the major barriers that could lead to confounding biases were addressed.25 Unbalanced actives and placebo distribution across sites and minor protocol modifications were the major weaknesses of the present RCT, while the lack of any modification of the ITT analysis prevented additional biases.21 The major confounding factor in the present analysis is the causality relationship between interrupting of treatment with proxalutamide and apparent higher mortality rate. Initially, differences in the mortality rate observed between the OT and ITT analysis could be due to the fact that patients tend to discontinue treatment in case of lack of response. Since the mortality rate was substantially lower in the proxalutamide arm compared to the placebo arm, and the number of non-completers was similar between these groups, similar increases in mortality rate would represent a higher increase in terms of percentage in the group with lower mortality rate, i.e., in the active arm. However, treatment non-compliance subjects in the placebo arm was not associated to poorer response compared to completers of the placebo arm. Conversely, non-completers of the proxalutamide arm had higher mortality rate compared to completers of the same arm and to the placebo arm. This reinforces the hypothesis of potential harm caused by abrupt and early interruption of proxalutamide. Indeed, in the male outpatient trial wi...

    Results from TrialIdentifier: We found the following clinical trial numbers in your paper:

    IdentifierStatusTitle
    NCT04728802CompletedProxalutamide Treatment for Hospitalized COVID-19 Patients

    Results from Barzooka: We did not find any issues relating to the usage of bar graphs.

    Results from JetFighter: We did not find any issues relating to colormaps.

    Results from rtransparent:
    • Thank you for including a conflict of interest statement. Authors are encouraged to include this statement when submitting to a journal.
    • Thank you for including a funding statement. Authors are encouraged to include this statement when submitting to a journal.
    • No protocol registration statement was detected.

    Results from scite Reference Check: We found no unreliable references.

    About SciScore

    SciScore is an automated tool that is designed to assist expert reviewers by finding and presenting formulaic information scattered throughout a paper in a standard, easy to digest format. SciScore checks for the presence and correctness of RRIDs (research resource identifiers), and for rigor criteria such as sex and investigator blinding. For details on the theoretical underpinning of rigor criteria and the tools shown here, including references cited, please follow this link.

  3. Version published to 10.1101/2021.06.28.21259661v1 on medRxiv