Evaluation of Potential Therapeutic Options for COVID-19.

With regard to the current coronavirus disease 2019 (COVID‐19) outbreak, the recent publications of inhibitory concentrations for a range of antiviral treatments , , , permits the evaluation of possible translation of therapeutic utility to humans. Using estimated half‐maximal viral inhibitory concentration (IC50) values and published pharmacokinetic exposure measures (area under the curve, clearance, and population variability) sourced from prescribing information, , , published studies , , , , , and published regulatory agency report, it is possible to calculate doses necessary to inhibit virus in vivo based on well‐established translation of anti‐infective agents (assuming free drug hypothesis and translational serum albumin protein binding to humans). Table 1 shows the inhibitory potencies, exposures, and predicted average concentration population range for 10 drugs tested in vitro. , , ,

Table 1

The Inhibitory Potencies, Exposures, and Predicted Average Concentration and Doses for Evaluated Drugs

	In Vitro			Regimen			Target Achievement
Drug	MWT	IC50 (μM)	IC50 (μg/mL)	Defined Daily Dose (mg/d)	Average Conc (95% Population Range) (μg/mL)		Prob (>IC50), %	Prob (>IC80), %	Prob (>IC90), %	Dose (IC50) (mg/d)	Dose (IC80) (mg/d)	Dose (IC90) (mg/d)	Dose (90% > IC90) (mg/d)	Primary Safety Concern
Ribavirin	244.2	109.5	26.7	1000	0.9	(0.5‐1.7)	0	0	0	28 736	114 944	258 625	379 082	Hemolytic anemia, teratogenic
Penciclovir	253.3	95.96	24.3	1400	1.6	(1.3‐2.1)	0	0	0	20 778	83 111	187 000	217 974	Acute renal failure
Favipiravir	157.1	61.88	9.7	1600	22.4	(19.8‐25.4)	100	0	0	692	2767	6226	6756	Teratogenic
Nafamostat	347.4	22.5	7.8	480	0.8	(0.5‐1.1)	0	0	0	4877	19 508	43 894	56 700	Cardiac arrest
Nitazoxanide	307.3	2.12	0.7	1000	3.5	(2.9‐4.2)	100	100	0	187	746	1679	1894	Gastrointestinal
Remdesivir	602.6	0.77	0.5	150	0.14	(0.1‐0.2)	0	0	0	499	1997	4494	5759	Liver damage
Chloroquine	319.9	1.13	0.4	600	1.7	(1.2‐2.6)	100	81	0	125	501	1128	1469	Cardiomyopathy and QT prolongation
Hydroxychloroquine	335.9	0.72	0.2	620	2.9	(1.4‐5.8)	100	100	77	52	210	472	751	Cardiomyopathy and QT prolongation
Ivermectin	875.1	2.48	2.2	14	0.03	(0.02‐0.1)	0	0	0	908	3634	8176	12 027	The Mazzotti reaction, teratogenic
Niclosamide	327.1	0.28	0.1	2000	0.08	(0.03‐0.2)	36	0	0	2416	9663	21 741	42 257	Gastrointestinal

IC50, half‐maximal viral inhibitory concentration; IC80, 4× IC50; IC90, 9× IC50; MWT, molecular weight.

Of note, we find that only 4 drugs (favipiravir, nitazoxanide, chloroquine, and hydroxychloroquine) are predicted to achieve in vivo average concentrations in excess of in vitro IC50 values. Three drugs (nitazoxanide, chloroquine, and hydroxychloroquine) are predicted to match IC80s, (defined as 4× IC50), and only hydroxychloroquine to match IC90 (9× IC50). An increased daily dose of 500‐mg remdesivir, or 2500‐mg niclosamide is also predicted to match in vitro IC50s in half of patients. At a population level, clinical doses necessary to provide 90% inhibition in at least 90% of patients (as opposed to 50%) are not accessible for any agent except for hydroxychloroquine. Further clinical evaluation of nitazoxanide (1000 mg/day), hydroxychloroquine (620 mg/day), chloroquine (600 mg/day), remdesivir (500 mg/day), and niclosamide (2500 mg/day) appear merited based on in vitro to in vivo translational evidence. These clinical studies must be conducted using proper study design (randomized, double blinded, etc) and must assess efficacy and toxicity using appropriate measures and evaluation of safety and tolerability. Caution is additionally advised for agents that are renally cleared, where exposure may be higher than indicated in our population analysis.

Conflicts of Interest

The authors are employees and stockholders of GlaxoSmithKline.