3 Results
The popPK model by Carmichael et al was successfully replicated as shown
in Figure 1. Results of external validation were satisfactory: the model
by Carmichael was able to acceptably predict previously published data
from external sources in CLE and in COVID-19 patients (see Figure 1).
This model was therefore used to simulated different dosing scenarios
including in national guidelines and ongoing/planned clinical study
protocols in Belgium. Results of the different simulations are shown in
Figure 2.
Selected clinical response markers including time to discharge from the
hospital, survival, C-reactive protein, blood oxygen levels and absolute
lymphocyte counts were collected from a cohort of 172 patients treated
with HCQ monotherapy at Saint Pierre Hospital in Brussels. Summary
descriptive statistics are included in Table2 together with patient age
and relevant comorbidities. Most of these patients received 5 day
treatment scheme with 400mg BID on day 1 followed by 200mg BID from days
2 to 5. Three patients received 6 or 7 days of treatment. As regards
impact of HCQ treatment on CRP, blood oxygen levels and absolute
lymphocyte counts, while all the patients received the same dosing
regimen for HCQ, variable responses were observed for each of the
biomarkers: a subset had their levels decreasing (i.e. negative slope)
after start of treatment with HCQ, while others had an increase of the
levels (positive slope). Moreover, time delay from apparition of
symptoms to implementation of treatment was a significant predictor of
admission to the intensive care unit (p<0.01, AUC ROC curve=
0.63) and death (p<0.01, AUC ROC curve= 0.61) based on
logistic regression analysis, while patient age, hypertension,
cardiomyopathies, cancer and obesity were significantly correlated with
patient death (p<0.05).
As regards clinical safety, case reports were found from literature data
and public vigilance database (EV). The dosing regimens information were
extracted when provided in the report and in most of the cases 800 mg
daily or lower doses were reported for day 1 and 400-600 mg daily were
reported for the following days. The time delay from start of treatment
to onset of ADR was very variable (from 1 to several days). Cardiac
disorders and especially QT prolongation were the most frequently
reported adverse drug reaction. At least one of the following additional
risk factors were reported in patients experiencing adverse drug
reactions: concomitant medication with at least one drug known to carry
QT prolonging drugs or cardiac toxicity, relevant comorbidities (e.g.
renal impairment), cardiovascular disease, cardiomyopathies and
hypokalaemia (see Table 2).