Effect of Dalcetrapib in Patients With Confirmed Mild to Moderate COVID-19

Sustained clinical resolution is defined as occurring when no key COVID-19 related symptom has a score higher than 1 over a 72-hour period (as documented using an electronic patient-reported outcome [ePRO] instrument), except for sense of smell and taste where the score should be 0 over a 72-hour period. The time to resolution was taken as the time from randomization until the first day of the last 72-hour period where the patient met the definition of resolution within 28 days. Patients who did not meet the definition of resolution 28 days after randomization were considered not resolved.

The scale is "Assessment of 14 Common COVID-19-Related Symptoms: Items and Response" from the Food and Drug Administration (FDA). The symptoms are scored as on a scale of 0 to 3 for 12 of the symptoms where 0 is none and on a scale of 0 to 2 for the two other symptoms where 0 is "as usual". A higher score is a worse outcome.

Log10 viral load, as assessed using the saliva, was summarized by treatment group using descriptive statistics (N, mean, median, standard deviation, minimum, and maximum) for each visit as well as for changes from baseline where an 80% CI were also presented. A repeated ANCOVA model was used for the data shown below, showing the mean changes from baseline to study visits (Day 3, Day 5, Day 10, and Day 28/EOS) in log10 viral load including treatment groups by study visit interaction, baseline value of log10 viral load and baseline value of log10 viral load by study visit interaction.

Log10 viral load, as assessed using the nasal swab, was summarized by treatment group using descriptive statistics (N, mean, median, standard deviation, minimum, and maximum) for each visit as well as for changes from baseline where an 80% CI were also presented. A repeated ANCOVA model was used for the data shown below, showing the mean changes from baseline to study visits (Day 3, Day 5, Day 10, and Day 28/EOS) in log10 viral load including treatment groups by study visit interaction, baseline value of log10 viral load and baseline value of log10 viral load by study visit interaction.

Sustained clinical resolution is defined as occurring when no key COVID-19 related symptom has a score higher than 1 over a 72-hour period (as documented using an electronic patient-reported outcome [ePRO] instrument). The time to resolution was taken as the time from randomization until the first day of the last 72-hour period where the patient met the definition of resolution within 28 days. Patients who did not meet the definition of resolution 28 days after randomization were considered not resolved.

The scale is "Assessment of 14 Common COVID-19-Related Symptoms: Items and Response" from the Food and Drug Administration (FDA). The symptoms are scored as on a scale of 0 to 3 for 12 of the symptoms where 0 is none and on a scale of 0 to 2 for the two other symptoms where 0 is "as usual". A higher score is a worse outcome.

Viral clearance based on polymerase chain reaction (PCR) test for Severe Acute Respiratory Syndrome CoronaVirus 2 (SARS-CoV-2) using nasal swab and saliva samples was performed on the intention-to-treat (ITT) population. Viral clearance was summarized by treatment group using Kaplan-Meier methods. Median and associated 80% confidence interval (CI) was presented. The number and percentage of patients who did not show viral clearance, did show viral clearance, and patients censored were presented.

Complete clinical resolution is defined as occurring when no key COVID-19 related symptom (excluding cough, sense of smell and taste) has a score higher than 0 over a 72-hour period. The time to resolution was taken as the time from randomization until the first day of the last 72 hour period where the patient met the definition of resolution within 28 days. Patients who did not meet the definition of resolution 28 days after randomization were considered not resolved.

The scale is "Assessment of 14 Common COVID-19-Related Symptoms: Items and Response" was used. The symptoms are scored as on a scale of 0 to 3 for 12 of the symptoms where 0 is none and on a scale of 0 to 2 for the two other symptoms where 0 is "as usual". A higher score is a worse outcome.

Sustained clinical resolution is defined as occurring when no key COVID-19 related symptom has a score higher than 0 over a 72-hour period. The time to resolution was taken as the time from randomization until the first day of the last 72 hour period where the patient met the definition of resolution within 28 days. Patients who did not meet the definition of resolution 28 days after randomization were considered not resolved.

The scale is "Assessment of 14 Common COVID-19-Related Symptoms: Items and Response" was used. The symptoms are scored as on a scale of 0 to 3 for 12 of the symptoms where 0 is none and on a scale of 0 to 2 for the two other symptoms where 0 is "as usual". A higher score is a worse outcome.

Resolution must have occurred within 28 days. Time of resolution of 29 days was imputed in censored subjects.

COVID-19 total symptom severity score was summarized by treatment group using descriptive statistics (N, mean, median, standard deviation, minimum, and maximum) for each visit as well as for changes from baseline where an 80% confidence interval (CI) was also presented. Mean changes from baseline were analyzed using a repeated measures ANCOVA model.

The scale is "Assessment of 14 Common COVID-19-Related Symptoms: Items and Response" from the Food and Drug Administration (FDA) document "Assessing COVID-19-Related Symptoms in Outpatient Adult and Adolescent Subjects in Clinical Trials of Drugs and Biological Products for COVID-19 Prevention or Treatment Guidance for Industry". Symptoms are scored as on a scale of 0 to 3 for 12 the symptoms and on a scale of 0 to 2 for two symptoms. The sum of all 14 symptom scores is reported, where 0 is the minimum and 40 is the maximum. A higher score is a worse outcome.

The number and percentage of patients for each WHO clinical outcome score was summarized. Scores were compared using the Mann-Whitney-Wilcoxon test.

This scale is called the "WHO Clinical Outcome Scale". It is scored from 0 to 9 where 9 is the most severe disease presentation. A higher score is a worse outcome.

The analysis of this endpoint was performed on the intention-to-treat (ITT) population. The percentage of patients who were hospitalized was compared using a binary logistic regression analysis. The model included only the treatment group. The results were presented as odds ratios, with associated 80% CIs and p-value.

The analysis of this endpoint was performed on the intention-to-treat (ITT) population. The number and percentage of patients who progressed to oxygen therapy was presented. The percentage of patients who had progressed to oxygen therapy was compared using a binary logistic regression analysis. The model included only the treatment group. The results were presented as odds ratios, with associated 80% confidence intervals (CIs) and p-value.

The analysis of this endpoint was performed on the intention-to-treat (ITT) population and only on those who received oxygen therapy. The number and percentage of patients who received different types of oxygen therapy was presented using descriptive statistics.

The duration of hospitalization was performed on the intention-to-treat (ITT) population in subjects who were hospitalized. Duration of hospitalization was summarized by treatment group using descriptive statistics (N, mean, median, standard deviation, Q1, Q3, minimum, and maximum). An ANOVA model was performed to analyze the difference between treatment groups. The model included only the treatment group. Contrasts under this model allowed for the comparisons across treatment groups. The results were presented as mean treatment difference with associated 80% confidence interval (CI) and p-value.

The analysis of this endpoint was performed on the intention-to-treat (ITT) population. The number and percentage of patients who died was presented. The percentage of patients who died was compared using a binary logistic regression. The model included only the treatment group. Contrasts under this model allowed for the comparisons across treatment groups. The results were presented as odds ratios, with associated 80% CIs and p-values.