Randomization of patients
A total of 61 patients received at least one dose of remdesivir by March 7, 2020; 8 of these patients were excluded due to the absence of post-initial information (7 patients) and an incorrect start date of remdesivir (1 patient) (Fig. S1 in the additional annex). Of the remaining 53 patients included in this analysis, 40 (75%) received the full 10-day course of remdesivir, 10 (19%) received 5 to 9 days of treatment and 3 (6%) less than 5 days of treatment .
Basic patient characteristics
Table 1 shows the basic demographic and clinical characteristics of the 53 patients in the compassionate use cohort. Patients were recruited in the United States (22 patients), Japan (9), Italy (12), Austria (1), France (4), Germany (2), the Netherlands (1 ), in Spain (1) and in Canada (1). A total of 40 patients (75%) were male, the age range was 23 to 82, and the median age was 64 (interquartile range, 48 to 71). Initially, the majority of patients (34 [64%]) were on invasive ventilation, including 30 (57%) on mechanical ventilation and 4 (8%) on ECMO. The median duration of invasive mechanical ventilation before the start of treatment with remdesivir was 2 days (interquartile range, 1 to 8). Compared to patients who received non-invasive oxygen support at baseline, those who received invasive ventilation tended to be older (median age, 67 versus 53 years), were more likely to be male (79% versus 68%), had a higher median serum ALT (48 U per liter, versus 27) and creatinine (0.90 mg per deciliter, versus 0.79 [79.6 μmol per liter, vs. 69.8]), and a higher prevalence of coexisting diseases, including hypertension (26%, versus 21%), diabetes (24%, versus 5%), hyperlipidemia (18%, versus 0%) and asthma. (15%, versus 5%). The median duration of symptoms before the start of remdesivir treatment was 12 days (interquartile range, 9 to 15) and did not differ significantly between patients receiving invasive ventilation and those receiving non-invasive ventilation (Table 1).
Clinical improvement during treatment with Remdesivir
For each oxygen support category, the percentages were calculated with the number of patients leaving as the denominator. Improvement (blue cells), no change (beige) and worsening (gray) of the oxygen support status are shown. Invasive ventilation includes invasive mechanical ventilation, extracorporeal membrane oxygenation (ECMO), or both. Non-invasive ventilation includes high-flow nasal oxygen therapy, non-invasive positive pressure ventilation (NIPPV), or both.
The baseline (day 0) was the day that treatment with remdesivir (RDV) was started. The stated oxygen support end states are based on the most recent reported data. For each patient, the colors in the line represent the patient’s oxygen support status over time. The colored circles to the left of each line indicate the patient’s overall change in condition from the baseline. A patient’s condition has “improved” if the oxygen supply improved before the last follow-up or if the patient was discharged. The vertical black marks indicate the last day of treatment with RDV. The gray dotted lines represent the missing data between the last state of oxygen reported by the patient and an event (death or discharge) or the last dose of appointment. A solid square at the end of a line indicates that the patient has died; an open diamond indicates that the patient is discharged from the hospital. If there is no square or diamond at the end of a line, no death or discharge has occurred. Patient 2 breathed ambient air until day 36. Patients 19 and 31 were discharged on the 44th day.
During a median 18-day follow-up (interquartile range, 13 to 23) after receiving the first dose of remdesivir, 36 of 53 patients (68%) showed improvement in the category of oxygen support, while 8 of 53 patients (15%) showed worsening (Figure 1). Improvement was seen in the 12 patients who breathed ambient air or received supplemental oxygen at low flow rates and in 5 of 7 patients (71%) who received non-invasive oxygen support (NIPPV or supplemental oxygen at high debit). It should be noted that 17 of 30 patients (57%) who received invasive mechanical ventilation were extubated, and 3 of 4 patients (75%) receiving ECMO stopped receiving it; all were alive at the last follow-up. The changes for each patient in the oxygen supply category are shown in Figure 2. As of the most recent follow-up date, 25 of 53 patients (47%) had been discharged (24% receiving invasive ventilation [8 of 34 patients] and 89% [17 of 19 patients] receiving a non-invasive oxygen support).
Clinical improvement is shown in the full cohort, in the stratified cohort based on baseline ventilation status and in the age stratified cohort.
At 28 days of follow-up, the cumulative incidence of clinical improvement, as defined by a decrease of 2 or more points on the six-point ordinal scale or a living discharge, was 84% (95% confidence interval % [CI], 70 to 99) by Kaplan – Meier analysis (Figure 3A). Clinical improvement was less frequent in patients receiving invasive ventilation than in those receiving non-invasive ventilation (risk of improvement, 0.33; 95% CI, 0.16 to 0.68 ) (Figure 3B) and in patients 70 years of age or older (risk ratio compared to patients under 50 years of age, 0.29; 95% CI, 0.11 to 0.74) (Figure 3C). Gender, region of enrollment, coexisting conditions, and duration of symptoms before initiating treatment with remdesivir were not significantly associated with clinical improvement (Table S1).
Seven of 53 patients (13%) died after the end of treatment with remdesivir, including 6 of 34 patients (18%) who received invasive ventilation and 1 of 19 (5%) who received non-invasive oxygen support (see additional appendix for case stories). The median interval between initiation of remdesivir and death was 15 days (interquartile range, 9 to 17). Overall mortality from date of admission was 0.56 per 100 days of hospitalization (95% CI, 0.14 to 0.97) and did not differ significantly among patients receiving invasive ventilation (0.57 per 100 days of hospitalization; 95% CI, 0 to 1.2)]) compared to those receiving non-invasive ventilation (0.51 per 100 days of hospitalization; 95% CI, 0.07 to 1.1]). The risk of death was higher in patients 70 years of age or older (risk ratio compared to patients under 70 years of age, 11.34; 95% CI, 1.36 to 94.17) and in those whose serum creatinine was higher at the start (risk ratio per milligram per deciliter, 1.91; 95% CI, 1.22 to 2.99). The risk ratio for patients receiving invasive ventilation compared to those receiving non-invasive oxygen support was 2.78 (95% CI, 0.33 to 23.19) (Table S2).
A total of 32 patients (60%) reported adverse events during follow-up (Table 2). The most common adverse events have been increased liver enzymes, diarrhea, rash, kidney failure and low blood pressure. In general, adverse events were more common in patients receiving invasive ventilation. A total of 12 patients (23%) had serious adverse events. The most common serious adverse events – multiple organ dysfunction syndrome, septic shock, acute kidney injury and hypotension – have been reported in patients who received invasive ventilation at baseline.
Four patients (8%) discontinued remdesivir early: one due to worsening of pre-existing renal failure, one due to multiple organ failure and two due to increased aminotransferases, including one patient with a maculopapular rash.
Due to the nature of this compassionate use program, data on a limited number of laboratory measurements have been collected. Median serum ALT, AST and creatinine fluctuated during follow-up (Fig. S2).