Two-thirds of patients with severe COVID-19 treated with remdesivir under a compassionate-use protocol showed clinical improvement, and early reports on phase 3 trials of the drug indicate promise, but researchers stress that definitive randomized, placebo-controlled trial data still are needed to confirm the drug’s safety and efficacy.

Seven such trials are in the works to test remdesivir, a nucleoside analog antiviral agent being developed by Gilead (Figure). In March, Gilead initiated two randomized, open-label, multicenter phase 3 studies in adults with COVID-19 in areas with a high prevalence of the disease in the United States, Asia and Europe (bit.ly/34SIi8I). The first trial is evaluating the safety and efficacy of five- and 10-day dosing of the drug used with standard of care in 2,400 hospitalized patients defined as having severe disease (ClinicalTrials.gov Identifier: NCT04292899). The second study will compare the safety and efficacy of the same dosing regimens of remdesivir plus standard of care with standard of care alone in 1,600 patients defined as having moderate disease (ClinicalTrials.gov Identifier: NCT04292730). 

No Definitive Conclusions Yet

One of the testing sites in the United States, the University of Chicago, has recruited 125 people with COVID-19, 113 with severe disease, treating them with daily infusions of remdesivir. Kathleen Mullane, DO, PharmD, a professor of medicine and the chief of clinical trials in the Section of Infectious Diseases at University of Chicago Medicine, who is leading the remdesivir trials there, recently led a video discussion with colleagues about the trial results, according to a report in Stat (bit.ly/2yBj0ja). 

During the discussion, Dr. Mullane shared some positive findings but refrained from drawing definitive conclusions from the early data, particularly with respect to the information from the study in severely affected patients, which does not have a control group. She did note that they have seen high fevers fall “quite quickly” in remdesivir-treated patients and patients weaning “off ventilators a day after starting therapy.” 

Katherine K. Perez, PharmD, an infectious diseases pharmacist at Houston Methodist Hospital, another trial site, noted that “early results are promising” there, too. In a press release from the hospital, she explained that “courses have been well tolerated, and a number of COVID-19 patients who have undergone treatment with remdesivir are showing signs of recovery and have been released from the hospital to go home.” Acknowledging that it is too early to tell, Dr. Perez noted that there are “indications that treatment with remdesivir can possibly stave off” intubation. 

Preliminary data from the trial in severe COVID-19 patients are expected to be available by the end of April (bit.ly/2KpDL45). Until then, Jeffrey R. Aeschlimann, PharmD, an associate professor at the University of Connecticut School of Pharmacy, in Storrs, and adjunct associate professor at UConn School of Medicine’s Division of Infectious Diseases, in Farmington, cautioned that “these ‘results’ at this point are … nonscientifically analyzed, partial-trial individual clinician impressions” from single sites of an ongoing multicenter clinical trial. He pointed out that patients in the trial are described as having severe disease, but per the protocol on ClinicalTrials.gov, initially patients are “not permitted to be on invasive mechanical ventilation at time of enrollment.” (However, patients may have been permitted to be on mechanical ventilation after enrollment, and a later extension phase of the trial includes patients on mechanical ventilation.)

“Unfortunately, the data leaked gives us no insights as to the numbers of patients who were/were not on vents and weaning ‘off ventilators a day after starting therapy,’” Dr. Aeschlimann said. “From a drug mechanism of action standpoint versus hyperinflammation response pathophysiology standpoint, this timeline is highly [unlikely] to be due to remdesivir” because the drug does not have “anti-inflammatory activity that we are aware of.”

Beyond that, Dr. Aeschlimann stressed that clinical experience with patients not needing invasive mechanical ventilation “indicates that almost all of them will improve and eventually will be discharged even with supportive care only.”

Limitations of Compassionate-Use Study

While all these phase 3 data are awaited, health care teams clamoring for options to treat their patients with SARS-CoV-2 virus are looking at the compassionate-use data published April 10 in The New England Journal of Medicine (doi: 10.1056/NEJMoa2007016). In that study, 36 of 53 patients (68%) treated with remdesivir showed clinical improvement (see sidebar, “Results from Compassionate-Use Study”).

Results From Compassionate-Use Study

Patients in the compassionate-use cohort were hospitalized with confirmed SARS-CoV-2 infection and had an oxygen saturation of 94% or less on ambient air or were receiving oxygen support (New Engl J Med 2020 April 10. doi: 10.1056/NEJMoa2007016). Patients were meant to receive remdesivir for 10 days (200 mg on day 1, followed by 100 mg daily for nine days). 

Sixty-one patients received at least one dose of remdesivir, but data from eight could not be analyzed (seven patients with no post-treatment data and one with a dosing error), leaving 53 patients with analyzable data. Of these patients, 40 (75%) received remdesivir for 10 days, 10 (19%) received it for five to nine days, and three (6%) received it for less than five days.

At baseline, 30 patients (57%) were receiving mechanical ventilation and four (8%) were receiving extracorporeal membrane oxygenation (ECMO). Patients receiving invasive ventilation tended to be older; were more likely to be male; had higher median serum levels of alanine aminotransferase and creatinine; and had more comorbidities such as hypertension, diabetes, hyperlipidemia and asthma. The median duration of symptoms before initiation of remdesivir treatment was 12 days; it did not differ substantially between patients receiving invasive versus noninvasive ventilation.

At a median follow-up of 18 days (range, 13-23 days), 36 of 53 patients (68%) showed an improvement in the need for oxygen support after receiving the first dose of remdesivir, and eight of 53 patients (15%) showed worsening. Improvement was observed in all 12 patients who were breathing ambient air or receiving low-flow supplemental oxygen, and in five of seven patients (71%) who were receiving noninvasive oxygen therapy. Seventeen of 30 patients (57%) who were receiving invasive mechanical ventilation were extubated, and three of four patients (75%) receiving ECMO stopped it; all those in this group were alive at last follow-up. By the date of the most recent follow-up, 25 of 53 patients (47%) had been discharged: eight of 34 (24%) who had received invasive ventilation and 17 of 19 (89%) who had received noninvasive oxygen therapy. Seven patients (13%) died: six of 34 (18%) who received invasive ventilation and one of 19 (5%) who did not.

By 28 days of follow-up, the cumulative incidence of clinical improvement, as defined by either a decrease of 2 points or more on the 6-point ordinal scale or live discharge, was 84% (95% CI, 70%-99%) by Kaplan–Meier analysis. Clinical improvement was less frequent in patients receiving invasive ventilation than those receiving noninvasive ventilation (hazard ratio [HR] for improvement, 0.33; 95% CI, 0.16-0.68) and patients aged 70 years or older (HR vs. patients aged <50 years, 0.29; 95% CI, 0.11-0.74). Sex, region of enrollment, coexisting conditions and duration of symptoms before initiation of remdesivir treatment were not significantly associated with clinical improvement.

During follow-up, 32 patients (60%) reported adverse events (AEs) and 12% had serious AEs. The most common serious AEs—multiple-organ dysfunction syndrome, septic shock, acute kidney injury and hypotension—were reported in patients who were receiving invasive ventilation at baseline.

The investigators noted that “comparisons with contemporaneous cohorts from the literature, in whom general care is expected to be consistent with that of our cohort, suggest that remdesivir may have clinical benefit in patients with severe COVID-19.” 

However, they acknowledged several limitations of the study: small sample size, relatively short duration of follow-up, lack of data on the eight unanalyzable patients and other potentially missing data, as well as lack of a randomized control group. They also pointed out other factors that may have influenced the results, including types of supportive care and differences in treatment protocols at the study sites. 

More Calls for Caution From ID Pharmacists 

Infectious disease pharmacists also flagged these limitations, underscoring the need for prudent interpretation of the data as well as the early news about the phase 3 trials.

Debra Goff, PharmD, an infectious disease specialist at the Ohio State University Wexner Medical Center, in Columbus, told Pharmacy Practice News that these data offer “promising glimmers of hope for remdesivir, but we should proceed with caution. The NEJM study is a case series without a control group. The University of Chicago data was unauthorized leaked information not meant for public scrutiny; however, the study does reveal encouraging observational data. Many severely ill patients got off a ventilator in a day and most were discharged at six days. We need to wait for the full study results before we can draw any conclusions and recommendations.”

Dr. Aeschlimann also pointed out several shortcomings in the design of the compassionate-use study. “Although there was a lot of positive buzz associated with the publication of the data from the compassionate-use [study], it is exceptionally difficult to try and draw meaningful conclusions from the data about whether the drug is helping patients with COVID-19 infection.” Without a control group, he added, “we can’t even directly compare how patients who received the drug [added to standard supportive care] did versus patients who received standard supportive care alone.” Dr. Aeschlimann shared several more concerns about the methodology used in the study (see sidebar, “Methodological Concerns About Compassionate-Use Study”). 

Methodological Concerns About Compassionate-Use Study

Jeffrey R. Aeschlimann, PharmD, an associate professor at the University of Connecticut School of Pharmacy, in Storrs, and adjunct associate professor at UConn School of Medicine’s Division of Infectious Diseases, in Farmington, noted several concerns about the study’s compassionate use data published in The New England Journal of Medicine (2020 April 10. doi: 10.1056/NEJMoa2007016).

• There was no clear explanation of how the 61 patients were selected to receive compassionate-use remdesivir. “The numbers of applications and demand for compassionate-use remdesivir were so overwhelming that Gilead stopped the program within a week of launching it,” Dr. Aeschlimann said.  “There was no information about how many patients applied, why some would be excluded versus accepted, etc.”

• The posted exclusion criteria included several factors that would select for patients with a much lower likelihood of dying (such as exclusions for vasopressor use, renal failure and alanine aminotransferase level >5 times the upper limit of normal). In addition, important information about the patients—including key laboratory markers of disease severity, such as D-dimer, C-reactive protein, lactate dehydrogenase, ferritin, and disease severity scoring indices such as APACHE score—were missing from the report, Dr. Aeschlimann said.

• Seven patients were excluded from the analyses because they had no clinical data after the first day. “There was no explanation of why this happened,” Dr. Aeschlimann pointed out, adding that this raises several questions. “Did these patients have adverse effects? Did they rapidly deteriorate? Did they get discharged from the hospital unexpectedly early?”

• Remdesivir was started late in the course of the COVID-19 infection (median, 12 days after onset of symptoms). “Most research on the time course of infection indicates that severely ill patients will deteriorate earlier than this … if they are going to deteriorate … ,” Dr. Aeschlimann said, “and that hyper-inflammation is the most important pathophysiologic process going on, as opposed to massive active viral replication.” 

• Over one-third of the patients (36%) were receiving noninvasive oxygen support at the time of enrollment. “These patients are likely to recover from their COVID-19 infection just with supportive care,” Dr. Aeschlimann said. “So, many of the patients enrolled probably would be likely to improve with or without any additional direct antiviral therapy.”

• Without a control group, he said, “it is nearly impossible to reliably assess patient outcomes for possible signs of remdesivir drug toxicities.”

Beyond pointing to flaws in study design, Dr. Aeschlimann took a very tempered approach to the results. Noting that although all 12 patients on ambient air/low-flow oxygen were discharged, he said this outcome was “to be expected even with supportive care.” Of the patients on noninvasive ventilation, one died and another ended up needing invasive therapy, so two of seven patients (29%) in this group “had progressive disease” despite remdesivir treatment, he said. In addition, he noted that of the 34 patients receiving invasive oxygen support, six (18%) died, nine (26%) were still receiving invasive oxygen support at the time of manuscript submission, eight (24%) were weaned to ambient air, and eight (24%) were discharged. “These numbers,” he said, “are well within the range of numbers that have been reported in other case series describing supportive care treatment of patients with COVID-19 infection who were admitted to ICUs and/or on invasive oxygen support.” These results “are perhaps a little bit better,” but should be “balanced with the significant methodological concerns,” Dr. Aeschlimann said.

“The bottom line,” he added, “is that this published case series doesn’t really squash clinicians’ optimism that remdesivir could possibly help patients with COVID-19 infection. But, it shouldn’t objectively do much to cause our optimism to grow significantly.”

Voices of Reason Needed

Dr. Aeschlimann expressed concern that “studies of other possible off-label therapies are being released for public consumption prior to being peer-reviewed, and many news outlets are reporting results from them without applying scientific scrutiny.” In addition, he said that some of the COVID-19–related studies “that have been peer-reviewed and accepted for publication are severely flawed, and some have even been retracted after scientists have formally expressed concerns to the publishing journals about these studies.” He stressed that “it’s a time for … pharmacists to heighten the education of our patients and colleagues about strengths, limitations and concerns about these reports.”

Dr. Goff reiterated the important role of pharmacists in interpreting emerging and evolving data on drug therapy for COVID-19. “Every study is ‘breaking news,’ and there needs to be a voice of reason to provide a critical scientific evaluation of each study or report. In times of desperation, everyone wants that next COVID-19 drug to be the one that works,” she said. “Pharmacists need to work with physicians to help prioritize which drug should be used in which COVID-19 patient. Providing dosing guidelines for each drug is not enough. Pharmacists must recognize that the COVID drug recommended on Monday may no longer be the best one by Friday. Every day is a new day in the world of COVID drug therapy.“