To the Editor:
We read with great interest the very thoughtful commentary by Joob and Wiwanitkit1. We agree with the authors that, despite conflicting clinical data to date, it is possible that hydroxychloroquine (HCQ) may have a protective effect in the setting of the coronavirus disease 2019 (COVID-19). More importantly, we agree that optimal HCQ dosage and timing is a critical underpinning for clinical trials. However, we highlight several considerations regarding treatment with high-dose HCQ.
To optimize dosing, the relationship between drug concentration in the target tissue and response must be well characterized. Because the precise mechanism of action of HCQ for SARS-CoV-2 is unknown, current dosing strategies are based on extrapolating the in vitro antiviral activity targets (e.g., effective concentration (EC) 50, EC100]. Herein lies one of the first challenges: the reported EC50 target for inhibiting viral replication varies in different reports (0.72–17.31 μM)2,3. Based on in vitro EC50 targets, dosages of HCQ 400 mg orally every 12 h for 2 doses followed by 200 mg every 12 h to achieve target unbound lung concentrations has been proposed2. However, the authors relied on animal-derived lung partition coefficients for HCQ, and it was unclear how they accounted for differences in the fraction of unbound drug in tissue compared to plasma2. Differences in these assumptions significantly alter whether the proposed dosing would achieve target concentrations4. Similarly, we modeled total serum concentrations and observed that only the lowest EC50 target (0.72 μM) is achievable in serum using proposed loading doses, and all concentrations were significantly less than those needed for complete viral inhibition5.
Although our analysis suggests that most in vitro target concentrations cannot be achieved in plasma and pulmonary interstitial fluid even with higher dosing, this does not preclude a potential benefit for HCQ in the setting of COVID-19. In addition to possible antiinflammatory benefits, intracellular lung concentrations may be critical for the drug’s potential antiviral effect. HCQ appears to block intracellular transport of SARS-CoV-2, with only 0.03% of virions in HCQ-treated cells localizing to endolysosomes, compared to 34.3% in untreated cells3. Because animal studies suggest substantially higher HCQ concentrations in lung tissue compared to plasma, it is possible that antiviral target concentrations can be achieved in the lung cells compared to interstitial fluid with conventional dosing3,6. However, because HCQ accumulates in lung tissue over several months6, the duration of therapy (as opposed to the dose) may be the key to understanding the drug’s potential in COVID-19. In addition, we are aware of early reports that suggest patients with systemic lupus erythematosus and SARS-CoV-2 infection develop severe COVID-19 at a similar frequency regardless of prior HCQ use, underscoring the importance of further study7.
Footnotes
Salary and/or research support for this project was provided by the Rheumatology Research Foundation’s Scientist Development Award, the Thrasher Research Fund, the Childhood Arthritis and Rheumatology Research Alliance/Arthritis Foundation, the Derfner Foundation, NIGMS/NICHD (2T32GM086330-06), NICHD (5R01-HD076676-04, HHSN275201000003I), and a Duke Health/Private Diagnostic Clinic ENABLE grant.
Dr. Schanberg serves on the data and safety monitoring board for Sanofi (sarilumab). Sanofi is a maker of hydroxychloroquine. Dr. Swamy chairs the Independent Data Monitoring Committee for GlaxoSmithKline (RSV vaccine trials). Dr. Cohen-Wolkowiez receives support for research from industry for drug development in adults and children.