Chloroquine for treatment of COVID-19 - a pig in a poke?

Introduction

At the end of 2019, the SARS-CoV-2 virus emerged as a cause of severe respiratory illness beginning in Wuhan, China. Soon thereafter, the World Health Organization declared COVID-19 a pandemic [1]. One study found 90% inhibition of viral replication at a concentration of 6.9 μM (IC₉₀), which, according to the authors, was a clinically achievable concentration[2]. Another study found an even higher IC₉₀ of chloroquine (>10 μM) against SARS-CoV-2[3]. These preclinical studies have caused the wide scale repurposing of chloroquine as a treatment for COVID-19. At present, it remains unknown whether the preclinical determined best-case therapeutic threshold of 6.9 μM is achieved in COVID-19 patients and whether this threshold can be directly translated to the clinic, while there are major concerns about the cardiac safety of chloroquine[4, 5]. We, therefore, investigated the pharmacokinetics and cardiac safety of chloroquine in hospitalized COVID-19 patients.

Materials and Methods

This was a prospective observational study of hospitalized COVID-19 that were treated with chloroquine from March 20, 2020 until May 1, 2020 in 2 academic hospitals and 3 teaching hospitals in the Netherlands. The institutional review boards of the Radboud University Medical Center and Leiden University Medical Center waived the need for formal ethics approval as all data were collected anonymously using minimally invasive measures. Chloroquine was dosed at a loading dose of 600 mg followed by 300 mg 12 hours later on day 1, followed by 300 mg bidaily for a total treatment duration of five days[6]. Plasma concentrations of chloroquine and its metabolite desethylchloroquine were measured in samples collected as a 4-point pharmacokinetic curve over an 8 hour interval or in randomly collected surplus samples from routine laboratory monitoring by means of a validated bioanalytical assay. Furthermore, chloroquine and metabolite protein unbound concentrations were determined in a subset of 20 random samples using ultrafiltration. Lastly, the unbound fraction of chloroquine and desethylchloroquine were determined at a concentration of 1 and 10 μM in six-fold in the same medium as used for the experiments by Wang et al (Dulbecco’s Modified Eagle Medium with 10% fetal bovine serum). Electrocardiography was performed at baseline and during treatment at the discretion of the treating physician and the QTc-interval was calculated by computerized interpretation. An increase in QTc of >60 ms or a QTc of >500 ms was considered a major risk factor for Torsades de Pointes[7].

Results

A total of 83 patients were included in this prospective observational study. The patient characteristics are reported in table 1. The patients in our population had a median age of 65 years with an interquartile range (IQR) of 57-73. All but 2 patients were admitted to the intensive care unit and 60% of the patients was male. The majority of the patients concomitantly used other QTc-prolongating medication (see table 1), with propofol as the most frequently used drug in this critically ill a population. The measured concentrations of chloroquine, its active metabolite desethylchloroquine and the individual sum concentration of both chloroquine and its metabolite per day are depicted in figure 1. None of the measured concentrations of chloroquine, nor the sum of chloroquine and its metabolite reached the best-case IC₉₀ found in vitro for chloroquine. Furthermore, it was found that the median (interquartile range) protein unbound fractions of chloroquine and desethylchloroquine in clinical samples were 18.3% (16.2-22.4%) and 20.2% (18.2-25.2%), respectively. The median unbound fractions of chloroquine and desethylchloroquine in cell medium with 10% fetal bovine serum at the 1 μM concentration level were 62.0% (IQR 58.4-633%) and 71% (IQR 69.1-73.0%), respectively. At the 10 μM concentration level these were 84.1% (IQR 83.8-84.4%) and 88.3% (IQR 87.6-88.7%)

• Download figure
• Open in new tab

Figure 1.

Left panel: Total (free and protein bound) concentration of chloroquine in μmol/L for every treatment day of chloroquine. Middle panel: metabolite desethylchloroquine in μmol/L. Right panel: sum of chloroquine and desethylchloroquine in μmol/L. Dashed line in left panel represents the IC₉₀ from the study by Wang et al[2].

QTc measurements at baseline and during chloroquine treatment were available in 41 patients (table 2). In 14 (34%) of these patients a ΔQTc >60 milliseconds was observed after initiation of chloroquine. QTc measurements during chloroquine treatment were available in 69 patients and a QTc ≥500 or ≥600 milliseconds during treatment was observed in 32 (46%) and 7 (10%) of these patients. Torsade de pointes was not reported in any of the patients. No statistical different plasma concentrations of chloroquine were observed in patients with or without QTc >500 ms (p=0.90 with an unpaired t-test on log-transformed pharmacokinetic data).

Discussion

The previously in vitro determined chloroquine concentrations needed to inhibit viral replication in cell medium were suggested to be clinically achievable[2, 3]. We here show that treatment with chloroquine does not result in plasma concentrations reaching the best-case IC₉₀ of 6.9 μM. In vivo, chloroquine is metabolized to desethylchloroquine[8]. It is likely that desethylchloroquine has antiviral properties against SARS-CoV-2. Desethylchloroquine is, like chloroquine, a cationic amphiphilic molecule able to accumulate in lysosomes and thereby exerting broad antiviral activity[9]. We found that even the sum of the individual chloroquine and desethylchloroquine concentrations did not exceed the proposed 6.9 μM target (figure 1). Furthermore, chloroquine is known to bind to plasma proteins. In our population, the protein unbound and pharmacologically active fraction of chloroquine and desethylchloroquine in plasma was approximately 20%, while the unbound fraction in cell medium was much higher at approximately 60-80%. Consequently, the gap between the reached pharmacologically active concentrations in vivo and in vitro IC₉₀ is even larger than presented in figure 1 due to this discrepancy. It is known that chloroquine accumulates in lung tissue. The lungs are considered lysosome-rich tissue and the sequestration of cationic amphiphilic drugs like chloroquine in lysosomes, therefore, results in a relatively high overall abundance of chloroquine in lung tissue[10]. However, as the accumulation of chloroquine into the lysosomes is likely to be similar as in other tissues due to the physicochemical properties of chloroquine and the intralysosomal concentration of chloroquine is the putative cause for antiviral effect, it is unlikely that effective exposure is reached in lung tissue.

Patients in our cohort received a total of 3300 mg chloroquine (as chloroquine base) during a 5-day treatment period. Although a simple solution to inadequate exposure would be administration of a higher dose, there are serious concerns for QTc prolongation. Although other drugs than chloroquine and conditions in critically ill patients may cause QTc prolongation, it has been recently shown that high dose chloroquine (a total of 12000 mg as chloroquine base during 10 days) is associated with increased mortality and life-threatening QTc prolongations compared to low dose chloroquine (2700 mg during 5 days) in COVID-19 patients[4]. Further dose increases should, therefore, not be considered. Our findings cast serious doubts on the role of chloroquine for treatment of COVID-19 and underline the necessity of dose finding studies before rushing a drug from the bench to the bedside of a vulnerable population.