An open-label randomized, controlled trial of the effect of lopinavir/ritonavir, lopinavir/ritonavir plus IFN-β-1a and hydroxychloroquine in hospitalized patients with COVID-19 – Final results from the DisCoVeRy trial

Trial design and oversight

DisCoVeRy is a phase 3 open-label, adaptive, multicenter, randomized, superiority-controlled trial which evaluates the efficacy and safety of repurposed drugs in adults hospitalized for COVID-19. Sponsored by the Institut national de la santé et de la recherche médicale (Inserm, France), the trial was approved by the Ethics Committee (CPP Ile-de-France-III, approval #20.03.06.51744). Written informed consent was obtained from all included participants or their legal representative, when unable to consent. The trial was conducted in accordance with the Declaration of Helsinki and national laws and regulations and declared on the clinicatrials.gov registry (NCT 04315948) and on the European Clinical Trials Database (2020-000936-23).

Study population

Eligible participants were adults (≥ 18-year-old) hospitalized with a PCR-positive (< 72 hours) SARS-CoV-2 infection and pulmonary rales or crackles with a peripheral oxygen saturation ≤ 94% or requiring supplemental oxygen. Inclusion and exclusion criteria are presented in the Supplementary Appendix.

Interventions and randomization

Participants were randomly assigned to treatment arms in a 1:1:1:1 ratio, through computer-generated blocks of various sizes and stratification by administrative region and severity of disease at enrolment (moderate: hospitalized participants not requiring oxygen or receiving low-flow supplemental oxygen; severe: hospitalized participants requiring non-invasive ventilation or high-flow oxygen devices, invasive mechanical ventilation or extracorporeal membrane oxygenation (ECMO)). Randomization was implemented in the electronic Case Report Form to ensure appropriate allocation concealment. Investigational arms were standard of care (SoC, control), SoC plus lopinavir/ritonavir (400 mg lopinavir and 100 mg ritonavir orally twice on day for 14 days (3,16)), SoC plus lopinavir/ritonavir plus IFN-ß-1a (44 μg of subcutaneous IFN-ß-1a on days 1, 3, and 6), SoC plus hydroxychloroquine (400 mg orally, twice on day 1 as a loading dose followed by 400 mg once daily for 9 days) (17). Supportive treatments corticosteroids, anticoagulants or immunomodulatory agents were allowed except antivirals. Enrolment in another investigative trial was not allowed.

Clinical and laboratory monitoring

Participants were assessed at days 3, 5, 8, 11, 15±2 and 29±3 while hospitalized. If discharge occurred before day 15, face-to-face visits were set up for days 15±2 and 29±3, for efficacy and safety evaluations. Clinical data, concomitant medications, adverse events (AEs) and measurements for safety biological data (blood cell counts, serum creatinine and liver aminotransferases) were collected. Nasopharyngeal swab (NPS) and lower respiratory tract (LRT) specimens were collected for SARS-CoV-2 RNA quantification. For lopinavir and ritonavir, trough plasma concentrations were obtained at days 1 and 3, 12h (±2h) after the last administration and for hydroxychloroquine, at day 1 12h (±2h) and at day 3 24h (±4h) after the last administration.

Outcomes measures

The primary outcome measure was the clinical status at day 15 as measured on the 7-point ordinal scale of the WHO Master Protocol (v3.0, March 3, 2020): 1. Not hospitalized, no limitation on activities; 2. Not hospitalized, limitation on activities; 3. Hospitalized, not requiring supplemental oxygen; 4. Hospitalized, requiring supplemental oxygen; 5. Hospitalized, on non-invasive ventilation or high flow oxygen devices; 6. Hospitalized, on invasive mechanical ventilation or ECMO; 7. Death.

Secondary efficacy outcome measures were the clinical status at day 29 and the time to an improvement of 2 categories as measured on the 7-point ordinal scale or hospital discharge until day 29, the time to National Early Warning Score 2 (NEWS2) ≤2 or hospital discharge until day 29, the time to hospital discharge until day 29, oxygenation- and ventilator-free days until day 29, in-hospital, 29-day and 3-month mortality, and the SARS-CoV-2 detection and quantitative normalized viral loads. Trough plasma concentrations of lopinavir, ritonavir and hydroxychloroquine were measured at days 1 and 3.

Secondary safety outcomes included the cumulative incidence of any grade 3 or 4 AE, or of any serious adverse event (SAE, according to the DAIDS Table for Grading the Severity of Adult and Paediatric Adverse Events, v2.1, July 2017) and the proportion of patients with a premature suspension or discontinuation for any reason of the investigational treatments.

Virological methods

Determination of normalized viral load blinded to treatment arm was performed on NPS and LRT specimens by RNA extraction on the EMAG® platform (bioMerieux, Marcy-l’Étoile, France). The SARS-CoV-2 load was measured by quantitative RT-PCR, according to a scale of calibrated in-house plasmid, using the RT-PCR RdRp-IP4 developed by the Institut Pasteur (Paris, France) (18). The amplification protocol was developed using QuantStudio 5 rtPCR Systems (Thermo Fisher Scientific, Waltham, Massachusetts, USA). The number of cells in sample (quality criteria for NPS and normalization tool for viral load determination) was checked using the CELL Control r-gene® kit (Argene-BioMérieux, Marcy-l’Étoile, France). If cell quantification was below 500 cells/reaction, the quality of the sample was considered too low to be measured. We computed a normalized SARS-CoV-2 load by dividing the viral load by the number of cells. All viral loads strictly below 1 log₁₀ RNA copies/10 000 cells were considered under the limit of detection and were reported as negative.

Pharmacological methods

Plasma concentrations of lopinavir, ritonavir and hydroxychloroquine were determined using liquid chromatography coupled with tandem mass spectrometry (19,20). The limits of quantification were 30 ng/mL for lopinavir and ritonavir, and 10 ng/mL for hydroxychloroquine.

Sample size calculation

The sample size was determined assuming the following scenario under SoC for each item of the ordinal scale at day 15: 1, 42%; 2, 38%; 3, 8%; 4, 7%; 5, 2%; 6, 1%; 7, 2%. At the time of the trial design, there was a significant uncertainty with these assumptions. We powered the study for an odds ratio of 1.5 (an odds ratio higher than 1 indicates superiority of the experimental treatment over the control for each ordinal scale category), with 90% power and using an overall two-sided type I error rate of 0.05. Adjusting for multiplicity of 4 pairwise comparisons with the control arm in a 5-arm setting, the two-sided false positive error rate would be 0.0125. We determined that the inclusion of 620 patients in each treatment arm was required.

Statistical and interim analyses

An independent data safety and monitoring board (DSMB) externally reviewed the trial data periodically. Based on interim analyses (see Supplementary Appendix), enrolment in the hydroxychloroquine arm was prematurely stopped on June 17^th, and enrolment in lopinavir-containing arms was stopped on June 29^th 2020.

For the 7-point ordinal scale, data were analyzed using a proportional odds model, which assumes a common odds ratio between the 7 points of the ordinal scale. All analyses were stratified by severity at randomization, and adjusted effect measures are reported. Full statistical methods are presented in Supplementary Appendix.

Patient’s characteristics at baseline

Between March 22^nd and June 29^th 2020, 603 participants were randomized across 30 sites in France and 2 in Luxembourg; 593 were evaluable for analysis (Supplementary Figure S1): control arm, n=149; lopinavir/ritonavir arm, n=147; lopinavir/ritonavir plus IFN-ß-1a arm, n=147; hydroxychloroquine arm, n=150. Participants’ baseline characteristics are presented in Table 1. Participants were mostly male (n=421, 71.0 %), median age was 64 years (IQR, 54-71). The median time from symptoms onset to randomization was 9 days (IQR, 7-12). The most frequent underlying conditions were obesity (n=169, 28.7%), chronic cardiac disease (n=155, 26.2 %) and diabetes mellitus (n=134, 22. 7%). At baseline, severe disease accounted for 214 (36.1%) participants. Concomitant treatments are listed in Supplementary Table 1.

Primary endpoint

The distribution of the 7-point ordinal scale at day 15 is presented in Figure 1 and Table 2. Adjusted OR for clinical improvement (aOR) were not in favor of investigational treatments (i.e., below 1): lopinavir/ritonavir versus control, aOR 0.82 (95% confidence interval [CI] 0.54-1.25, P=0.36); lopinavir/ritonavir plus IFN-ß-1a vs. control, aOR 0.69 (95%CI 0.45-1.05, P=0.08); hydroxychloroquine vs. control, aOR 0.94 (95%CI 0.62-1.41, P=0.76).

• Download figure
• Open in new tab

Figure 1. Clinical status, as measured by the 7-point ordinal scale, at day 15 and day 29 of patients from the intention-to-treat population of the DisCoVeRy trial, according to treatment arm and disease severity at baseline.

Reported numbers refer to the proportion of patients with the corresponding level in each group. L/r, Lopinavir/ritonavir; L/r + IFN, Lopinavir/ritonavir + interferon ß-1a; HCQ, Hydroxychloroquine.

Secondary endpoints

There was no significant difference between any of the treatment and control arms on the 7-point ordinal scale at day 29 (Figure 1 and Table 2). The time to improvement of 2 categories of the same scale or hospital discharge within day 29 was significantly longer in lopinavir/ritonavir-containing arms than in the control arm: lopinavir/ritonavir versus control, HR=0.70 (95%CI 0.54-0.92, P=0.01 and lopinavir/ritonavir plus IFN-ß-1a versus control, HR=0.68 (95%CI 0.52-0.89, P=0.008). The time to NEWS ≤2 or hospital discharge within 29 days was significantly longer in the lopinavir/ritonavir plus IFN-ß-1a arm than in the control arm (HR=0.72, 95%CI 0.54-0.96, P=0.02), as was the time to hospital discharge within day 29 (HR=0.72, 95%CI 0.54-0.97, P=0.03). Participants assigned to the lopinavir/ritonavir plus IFN-ß-1a arm exhibited a higher risk of 3-month mortality than participants assigned to the control arm: aOR 2.11 (95%CI 1.01;4.39, P=0.04), while no significant effect was observed in the lopinavir/ritonavir arm (aOR 1.41, 95%CI 0.65;3.06, P=0.39) nor in the hydroxychloroquine arm (aOR 1.17, 95%CI 0.53;2.58, P=0.70).

No other significant difference was observed for other secondary outcomes (Table 2 and Supplementary Figures S2-S4).

Virological endpoints

The slope of the decrease of the viral loads in NPS over time was not significantly affected by any of the investigational treatments (Figure 2). No significant difference in the proportion of participants with detectable viral loads at each sampling time was observed in the NPS nor in the LRT specimens (Supplementary Table S2 and S3).

• Download figure
• Open in new tab

Figure 2. Evolution of the normalized SARS-CoV-2 viral load in nasopharyngeal swabs between baseline and day 15 in the intention-to-treat population of the DisCoVeRy trial: means (95%CI) of the log viral loads (panel A), mean changes from baseline (95%CI) of the log viral loads (panel B).

L/r, Lopinavir/ritonavir (blue line); L/r + IFN, Lopinavir/ritonavir + interferon ß-1a (yellow line);

HCQ, Hydroxychloroquine (red line); control (black line).

LSMD, least-square mean difference; 95%CI, 95% confidence interval.

Trough concentrations of experimental treatments

At day 3, median trough plasma concentrations of lopinavir were 20 328 ng/mL (IQR, 13 251; 26 980) and 20 926 ng/mL (16 510; 25 930) and of ritonavir were 536 ng/mL (312; 1 028) and 609 ng/mL (388; 1 164) in the lopinavir/ritonavir and in the lopinavir/ritonavir plus IFN-β-1a, respectively (Supplementary Table S4). Median trough plasma concentrations of hydroxychloroquine were 126 ng/mL (67; 276).

Safety

The safety analysis included 589 participants (control, n=149; lopinavir/ritonavir, n=147; lopinavir/ritonavir plus IFN-β-1a, n=145; hydroxychloroquine, n=148). Safety outcomes are presented in Table 3. Among 2524 reported AEs, 570 were graded 3 or 4 in 243 patients and mostly reported in lopinavir/ritonavir-containing arms (Table 3).

A total of 856 SAEs were reported in 282 participants; 189 (22.1%) were related to the investigational drug according to investigator’s judgment (lopinavir/ritonavir arm, n=54 lopinavir/ritonavir plus IFN-β-1a arm, n=89; hydroxychloroquine arm, n=46). A significantly greater number of patients experienced at least one SAE in the lopinavir/ritonavir-containing arms than in the control arm (Table 3). The most frequently reported SAEs were acute respiratory failure (n=78, 13.2%), acute kidney injury (n=64, 10.8%), acute respiratory distress syndrome (n=74, 12.5%), arrhythmia (n=58, 9.4%), pulmonary embolism (n=34, 55.7%), and sepsis including those related to super-infections (n=32, 5.4%). Twelve percent (n=103) of participants developed at least one kidney-related SAE. Among them, 23 had acute renal failure upon admission, and 78 were critically-ill ventilated patients with acute kidney injury. Among 64 fatal SAEs, 32 had a pulmonary origin, and 32 had a non-pulmonary origin. Three non-pulmonary-related deaths were linked to investigational treatments by investigators, all in the - lopinavir/ritonavir plus IFN-ß-1a arm.