Ivermectin as a potential COVID-19 treatment from the pharmacokinetic point of view: antiviral levels are not likely attainable with known dosing regimens

Abstract

The broad-spectrum antiparasitic agent ivermectin has been very recently found to inhibit SARS-CoV-2 in vitro and proposed as a candidate for drug repurposing in COVID-19. In the present report the in vitro antiviral activity end-points are analyzed from the pharmacokinetic perspective. The available pharmacokinetic data from clinically relevant and excessive dosing studies indicate that the SARS-CoV-2 inhibitory concentrations are not likely to be attainable in humans.

Keywords:

• Ivermectin
• SARS-CoV-2
• COVID-19
• drug repurposing

Introduction

The COVID-19 pandemics has fuelled much research efforts towards repurposing of existing drugs as possible antiviral agents, whereby the therapeutic strategies have been largely based on pre-existing data for the preceding coronaviral outbreaks TORS and MERS [1–3]. The drug regulatory agencies, health authorities, key opinion leaders and policy decision makers have been significantly strained by the dilemma of evidence-based medicine and good clinical practice versus the prompt need for safe and effective treatment [4]. Unfortunately we have been witnessing huge public and political pressure for legitimation of drug-repurposing and off-label use worldwide, which nonetheless could be regarded as an acceptable compromise, pending the emergency of the current situation, but only in case of drugs with well-defined safety profiles and at least some clinical evidence in COVID-19 [4, 5]. Conversely most of the treatment protocols are based on observational studies and anecdotic reports [4, 6–9], albeit with a hope that the promptly emerging data from randomized studies will enable switching COVID-19 treatment back to the avenues of evidence-based medicine [10]. An exceptionally alarming phenomenon however is the public communication of drugs with preliminary in vitro activities against SARS-CoV-2 as potential therapeutics for COVID-19 eventually causing malignant reverberation in social media. Such example is the otherwise very interesting study of Caly et al., recently published in Antiviral Research [11].

This paper describes the in vitro antiviral activity of the antiparasitic agent ivermectin in a model of Vero/hSLAM cells infected with a SARS-CoV-2 isolate (Australia/VIC01/2020) [11]. The authors have performed a pilot experiment using continuous exposure of the cells to ivermectin at 5 μmol/L and found time-dependent decrease of cell associated and supernatant viral RNA. Thereafter the antiviral activity was assessed following continuous exposure to serial dilutions of ivermectin, which caused concentration-dependent antiviral effects with practically total eradication at 5 μmol/L and half-maximal inhibition at approximately 2.5 µmol/L [11].

The academic, virological and pharmacological impact of the newly discovered antiviral effects of ivermectin against SARS-CoV-2 is beyond any doubt. Nevertheless the notion for possible clinical translation and repurposing, which has generated enormous media coverage, needs to be carefully addressed with reference to the pharmacokinetics of ivermectin. In this paper we sought to analyze the dosing regimens of the drug, the available maximal plasma concentration levels to allow detailed juxtaposition with the SARS-CoV-2 inhibitory effects and to question the paradigm for the plausibility of ivermectin repurposing in COVID-19.

Materials and methods

A literature survey was performed in order to analyze the published dose regimens and to collect human exposure data for ivermectin, following clinically relevant (150–800 μg/kg) or excessive dosing (up to 2000 μg/kg). The available pharmacokinetic data for ivermectin in patients with parasitic infection and healthy volunteers were pooled and the maximal plasma concentration levels (C_max) used as surrogates for juxtaposition with the in vitro SARS-CoV-2 inhibitory findings. The published concentrations showing antiviral activity were recalculated in ng/ml to allow direct comparison with the pharmacokinetic data.

Results and discussion

Ivermectin has a valuable clinical role for the management of different parasitic diseases whereby the described therapeutic regimens, could be summarized as follows: 150 µg/kg once yearly for treatment of onchocerciasis, 200 µg/kg as a single dose for strongyloidiasis, 150 to 200 µg/kg twice yearly or alternatively 300 to 400 µg/kg once yearly in endemic areas for lymphatic filariasis, and 200 µg/kg in conjunction with topical drugs for hyperkeratotic scabies, also known as crusted or ‘Norwegian’ scabies [12–14].

Ivermectin is a semisynthetic analogue of the natural product avermectin B_1a, a lipophilic macrolide isolated from Streptomyces avermitilis developed as a crop management insecticide. Ivermectin affects a plethora of ivertebrate species, incl. parasitic nematodes, arachnids and insects. Its mode of action on target species is by potentiating GABA-mediated neurotransmission and by binding to glutamate-gated Cl⁻ channels, found only in invertebrates [13]. The drug induces tonic paralysis of the musculature of susceptible parasites, and eventually death [13]. At the recommended doses, ivermectin does not readily penetrate the central nervous system (CNS) of mammals, where GABA functions as a neurotransmitter [13, 15]. Conversely, in healthy volunteers and infected patients, the drug is usually well tolerated at the therapeutic dose ranges [12–14]. A recent meta-analysis has shown that even larger doses (up to 800 µg/kg) with a several years period of follow-up could be well tolerated in patients with parasitic infections [16]. The largest dose intensity with registered pharmacokinetic parameters in healthy subjects is 120 mg, corresponding to up to 2000 µg/kg [12].

As evident from the analyzed pharmacokinetic data, both the clinically applied dosage schedules and the aforementioned excessive 120 mg dose yield blood levels at the ng/mL i.e. nanomolar range (Table 1). These concentrations are orders of magnitude lower, as compared to the in vitro antiviral end-points described in the study of Caly et al. [11]. Table 2 summarizes the in vitro inhibitory concentrations recalculated in ng/mL (based on a molecular weight of 875.1) to allow direct juxtaposition with the pharmacokinetic parameters in Table 1. Moreover the in vitro data were compared to the C_max values obtained after 36 mg and 120 mg doses corresponding to dose intensities of up to 700 µg/kg [25] or 2000 µg/kg [12] respectively, with calculation of the corresponding exposure ratios.

Table 1. Published pharmacokinetic parameters for ivermectin following clinically relevant and excessive dosing.

Pharmacokinetic study	Dose	Absorption parameters				Elimination
		Cmax (ng/mL)	Cmax (nmol/L)	tmax (h)	t1/2(abs) (h)	t½ (h)	Cl (1 kg^−1·day^−1)
Krishna et al., 1993 [17]	6 mg (tablets)	23.1	26.4	4.3	0.5	12.6	4.28
Krishna et al., 1993 [17]	12 mg (tablets)	30.4	34.7	10.3	2.5	13.4	4.03
Long et al., 2001 [18]	6 mg (tablets)	20.2	23.1	4.7	1.4	11.1	7.57
Long et al., 2001 [18]	12 mg (tablets)	23.5	26.9	5.3	1.4	21.1	6.53
Long et al., 2001 [18]	18 mg (tablets)	31.2	35.7	5.1	1.7	16.7	10.6
Edwards et al., 1988 [19]	12 mg (solution)	81	92.6	3.6	–	–	–
Edwards et al., 1988 [19]	12 mg (tablets)	50	57.1	3.4	–	–	–
Edwards et al., 1988 [19]	12 mg (capsules)	46	52.6	3.6	–	–	–
Ogbuokiri et al., 1993 [20]	150 μg/kg	37.9	43.3	–	–	–	–
Baraka et al.,1996 [21]	150 μg/kg	52.2	59.7	5.2	–	35.0	–
Elkassaby, 1991 [22]	150 μg/kg	39	44.6	5.6	–	16	–
Okonkwo, et al., 1993 [23]	6 mg (tablets)	38.2	43.7	4.7	–	54.5	3.1
Njoo, et al. 1995 [24]	150 μg/kg	–	–	–	–	19.9	–
Muñoz, et al., 2018 [25]	36 mg (550–700 μg/kg)	96.2	109.9	3.64	–	91.77
Guzzo et al., 2002 [12]	90 mg	158.1	180.7	4.9	–	18.8	–
Guzzo et al., 2002 [12]	120 mg (1404–2000 µg/kg)	247.8	283.2	4.2	–	19.1	–

Table 2. Comparison between the in vitro inhibitory concentration of ivermectin [11] with exemplar published C_max values of the drug [1, 127].

In vitro end-points [11]	Inhibitory concentrations (µmol/L)	Inhibitory concentrations recalculated in µg/mL	Inhibitory concentrations recalculated in ng/mL	Ratio between the inhibitory concentrations and clinically relevant Cmax (96.2 ng/mL)	Ratio between the inhibitory concentrations and Cmax after excessive dosing 247.8 (ng/mL)
IC50 – cell associated (E gene)*	2.8	2.45	2450	25.5	9.9
IC50 – supernatant (E gene)	2.4	2.1	2100	25.9	8.5
IC50 – cell associated (RdRp gene)**	2.5	2.19	2190	27.0	8.8
IC50 – supernatant (RdRp gene)	2.2	1.95	1950	24.1	7.9
Average IC50	2.475	2.166	2166	22.51	8.74
IC90–100 (highest and most effective concentration tested)	5.0	4.3755	4375	55.0	17.7

* E – envelope protein gene.

** RdRp – RNA-dependent-RNA-polymerase gene [11].

The data values have been bolded as they represent the ratios between the average IC₅₀ and te IC₉₀ endpoints and the C_max values.

The analyzed data show that, at least at the clinically relevant dose ranges of ivermectin, the published in vitro inhibitory concentrations and especially the 5 µmol/L level causing almost total disappearance of viral RNA are virtually not achievable with the heretofore known dosing regimens in humans. The 5 µmol/L concentration is over 50 times higher than the levels attainable after 700 μg/kg [25] and 17 times higher vs. the largest C_max found in the literature survey (247.8 ng/ml) [12]. Moreover the authors` claim for achieving viral inhibition with a single dose is inappropriate because practically the infected cells have been continuously exposed at concentrations that are virtually unattainable even with excessive dosing of the drug. In other words, the experimental design is based on clinically irrelevant drug levels with inhibitory concentrations whose targeting in a clinical trial seems doubtful at best.

The mechanistic rationale could be questioned as well – ivermectin has been proposed to mediate its antiviral effects via host cells mechanisms, namely inhibition of the importin α/β-mediated nuclear transfer of viral cargo proteins [26–28]. The clinical translation of this mechanism however is greatly debateable, as for instance, the IC₅₀ for importin α/β binding to a DENV non-structural protein is 17 μmol/L [26].

The repurposed antimalarial drugs hydroxychloroquine and chloroquine, which have been included in numerous COVID-19 treatment protocols, also have micromolar inhibitory concentrations against SARS-CoV-2 [29] and nanomolar C_max values [30]. Nevertheless these agents have enormous apparent volumes of distribution and presumably disproportionally larger tissue levels relative to the plasma concentrations, which makes the translation of the in vitro data plausible [30–32]. The other very promising agent tested in clinical trials and applied as compassionate use for COVID-19 is remdesivir [1, 33]. This broad-spectrum antiviral drug originally developed for Ebola shows potent inhibitory effects against SARS-CoV-2 with an IC₅₀ of 0.77 µmol/L (464 ng/mL) [34]. This concentration is readily attainable as the drug is given as venous infusion. The typical dosing schedule for remdesivir – initial infusion of 200 mg, followed by 100 mg/daily for a total of 5 days – yields maximal plasma concentrations of 5440 ng/mL in the first day and 2610 on day 5 [33].

In the case of ivermectin, however, the potential repurposing plausibility if any is at present not very likely, because the antiviral concentrations would be attainable only after massive overdose. The therapeutic application of ivermectin is usually not associated with significant toxicity, whereby the majority of documented adverse effects, such as: nausea, rash, dizziness, itching, eosinophilia, abdominal pain, fever, tachycardia, could be generally attributable to the gross lethality of invading microfilarias giving rise to Mazzotti-like reactions [13, 14, 35]. Nevertheless, at large doses, the drug could penetrate the blood-brain barrier and could affect GABA-ergic transmission causing CNS depression and potentiation of the effects of benzodiazepines [15]. Human exposure at doses multifold higher than the therapeutic one is expected to give rise to side effects similar to those documented in preclinical mammalian testing [13, 15]; noteworthy, the anti-SARS-CoV-2 levels overlap with the inhibitory concentrations of ivermectin on P-glycoprotein, and other ATP-binding cassette transporters [36], which limit the penetration of the drug in the CNS [15, 37]. Human overdoses have been associated with vomiting, tachycardia and ECG abnormalities, significant blood pressure fluctuations, CNS effects (drowsiness, ataxia) and visual disturbances (mydriasis). Accidental self-injection of a veterinary medicinal product has produced signs of clinical toxicity, albeit the drug was applied at therapeutically relevant dose (approximately 200 μg/kg) [15].

It has to be emphasized that general public communication of drugs as potential COVID-19 therapeutics, based solely on in vitro data, is neither scientifically nor ethically appropriate. Ivermectin has been previously shown to exert antiviral activity in vitro against Dengue fever virus (DENV) [28], influenza virus [38], West Nile Virus [26], Venezuelan equine encephalitis virus [27] and heralded as a possible antiviral drug, but so far there has not been any clinical translation of these data. Noteworthy, a clinical trial for the treatment of Dengue fever in Thailand failed to show clinical benefits [11]. In light of the aforementioned pharmacokinetic considerations, this is not surprising given that the published inhibitory concentrations against DENV1-4 ranged within 1.66–2.32 µmol/L [28].

The data-processing and analysis is an oversimplification of the pharmacokinetic processes because it is based solely on total C_max as a drug exposure surrogate. To address this issue, a Supplemental data file has been compiled to analyze the influence of the tissue accumulation and protein binding of ivermectin.

The world has already seen epidemics of self-medication, drug shortages and even overdoses with chloroquine and hydroxychloroquine [39, 40]. Unfortunately the Caly et al. study, which prompted enormous public interest, has the potential to evoke similar tragic sequels, especially having in mind that in many countries the drug is only available as solutions for injection for veterinary use, whose potential for serious toxicological outcomes in humans is undisputed. In Bulgaria, in particular this study has prompted the National Veterinarians` Union to share their concerns about the hysteria this study has evoked and to firmly discourage self-medication with ivermectin, which in this country is only available for use in animals [41]. In a 10^th April letter to stakeholders FDA has similarly shared its concerns on this issue and explicitly advised against any attempts for self-medication with ivermectin for COVID-19 [42].

Similar general pharmacological considerations and toxicological concerns have been shared in other recent contributions, questioning the rationale for ivermectin repurposing for SARS-CoV-2 infection [43, 44]. Nevertheless, at the time of the writing of this paper several clinical trials of ivermectin for COVID-19 have been registered [45] and an international registry-based case-control study has been described in a non-peer reviewed preprint [46].

Conclusions

The available pharmacokinetic data for ivermectin indicate that at the doses routinely used for the management of parasitic diseases the SARS-CoV-2 inhibitory concentrations are practically not attainable. At present any empiric treatment with ivermectin or its inclusion in therapeutic protocols are not scientifically justifiable. The very consideration of the drug as a broad spectrum antiviral agent is incorrect because it has failed to demonstrate antiviral effects beyond the in vitro level. Pending the paucity of reliable data from controlled studies and the aforementioned pharmacokinetic considerations, the application of ivermectin in COVID-19 patients is to be decisively discouraged.

Disclosure statement

No potential conflict of interest was reported by the authors.