Ivermectin Effect on SARS-CoV-2 Replication in Patients With COVID-19

Ivermectin (IVM) is a semisynthetic antiparasitic agent belonging to the avermectin group,
drugs isolated from Streptomyces avermitilis. Ivermectin is widely used for humans and
animals, with millions of doses annually administered through mass drug distribution programs
held by the World Health Organization (WHO - 2016). Since 1980, Ivermectin has been included
in the Essential Medicines List of the World Health Organization (WHO - 2019).

This medicine is orally administered and is usually used for the treatment against nematodes
and ectoparasites, making this drug the first-choice medication for onchocerciasis, lymph
filariasis, itch and strongylosis.

Until now, SARS-CoV-2 viral load dynamics has not been clearly determined. However, works
tending to a preliminary characterization of the viral load (VL) behavior have emerged. One
of them, the most substantial one, includes the work done by Kai-Wang showing the VL behavior
during the 30 days before the onset of COVID-19 symptoms (Kai-Wang To et al; 2020). In this
work, an average of 7.5 oropharyngeal samples of individuals with severe (n=10) and moderate
(n=13) COVID-19 have been assessed. Time between the onset of symptoms and hospitalization
ranged between 0 and 13 days, with a mean of 4 days. VL median at day 0 in all patients was
5.2 log10 copies/mL, and no significant differences between severe and moderate COVID-19
groups occurred. The viral load peak observed during the first week from the onset of
symptoms had a median equal to 6.91 (Q1-Q3: 4.27-7.40) and 5.29 (Q1-Q3: 3.91-7.56) log10
copies/mL in severe and moderate COVID-19 patients, respectively. There was no significant
difference between both groups (p=0.52). Likewise, no difference between patients with and
without comorbidities has been observed (n of patients equal to 12 and 11, respectively) as
per initial VL (p=0.49) and peak VL (p=0.29). However, it has been observed that the VL peak
was directly associated with the age of the patient (R2=0.48 and p=0.02). In a combined
analysis of all patients (n=23), it has been observed that the VL grows in the first days
following the onset of symptoms and that, in the 5-6 days, VL falls sharply, reaching a lower
mean value, yet similar to that of the day 0. General VL behavior from day 9 to 30 showed a
negative slope (VL fall) equal to -0.15 (95% confidence interval: -0.19 to -0.11) log10 per
daya. It must be emphasized that by day 20 from the onset of symptoms, VL mean continues to
be quantifiable (4 log10 copies/mL), and that 7/23 (30%) patients show viral RNA detection
after such day.

In other paper, the VL of 17 patients with an age median of 59 (range 26-76 years) has been
studied, who tested positive for SARS-CoV-2 (Zou et al; 2020). Nonetheless, VL measurement by
nasal swab was performed in 16 patients and quantification was conducted relatively,
expressed in Ct values (cycle threshold), which is related with the VL copies detected in the
molecular reaction: the lower the Ct value, the greater the VL, and vice versa. The feature
worth noting of such work is that VL dynamics varies widely from one patient to another.
Contrary to the work of Kai-Wang To et al. [1], the VL peak seems to occur earlier (first
three days), whereas the sharp VL fall is at day 6 approximately, from which day the VL is
undetectable in most patients. Only very few patients show detectable viral RNA after 10 days
from the onset of symptoms.

In other work, pharyngeal swabs performed in 67 patients have been studied (Pan et al; 2020).
VL dynamics similar to that reported by Kai-Wang To et al. is observed, with a VL of
approximately 4 to 5 log10 copies/mL at day 0 of the onset of symptoms, a peak at day 6/7 (8
log10 copies/mL) and a sharp fall from day 8 of the onset of symptoms. A temporary onset of
viremia is observed with a VL of 4 log10 copies/mL up to day 15 from the onset of symptoms.

Finally, in the work published by Wölfel et al. [4] much more accelerated dynamics than that
in the work of Kai-Wang To et al. and Pan et al. (Wölfel et al; 2020) is observed, and the
peak appearing much earlier than the day 4 from the onset of symptoms in the nine studied
patients. Additionally, in 8 patients, the VL falls sharply reaching values below the
quantification limit (2 log10 copies/mL) at day 10-11 following the onset of symptoms.
Although there are transient VL relapses after day 10, this value remains very close to the
study quantification limit.

In conclusion, if the works of Kai-Wang and Pan are considered, a greater viral RNA half-life
is expected at day 10 following the onset of symptoms. Even though in the works of Zou and
Wölfel the peak occurs much earlier than the first five days and viral negativity occurs
after day 10-11 following the onset of symptoms. Method variability among the many works is a
limiting factor when comparing the outcomes. That is why, due to the sample size in Kai Wang
et al (n=23) and Zou et al (n=67), it is suggested to use these works as reference and to
consider the works of Pan (n=17) and Wölfel (n=9) as part of behavior diversity. It is
important to define the period in which therapy is to be initiated from the onset of symptoms
so that the sharp fall observed post-peak is not a factor that biases the potential antiviral
effect of the drug.

Considering these backgrounds along with the preliminary study with ivermectin, it is not
possible to define a specific, progressive outcome for the reduction of the VL, but it may be
expressed in a variation percentage with respect to the control population (without therapy)
at the end of treatment. This percentage difference between the treated population and the
untreated population must be greater than the variation observed for the study day or period
in the patients included in the above mentioned works, since such percentage difference would
replicate the behavior in our control population.