COVID: remdesivir RCT, finally

Finally, an actual large RCT on remdesivir in hospitalized Covid-19 patients, with preliminary results from the ACCT-1 study (Adaptive Covid-19 Treatment Trial): see covid remdesivir RCT nejm2020 in dropbox, or DOI: 10.1056/NEJMoa2007764

 

Details:

--1059  patients hospitalized with Covid-19 and evidence of lower respiratory tract involvement (radiologic chest infiltrates, room air oxygen sat <95%, mechanical ventilation/ECMO) were randomized to remdesivir vs placebo for up to 10 days

    --they define the above as “severe disease”, but, as below, they did enroll some who were considered to have mild/moderate disease (oxygen sat >94%, resp rate <24 without supplemental oxygen)

--60 trial sites globally, most in US (45 sites), followed by UK (5 sites), then Greece, Germany, Korea, Mexico, Spain, Japan, and Singapore

--median age 59, 64% male, 53% white/21% black or African-American/13% Asian/23% Latinx,

--median time from symptom onset to randomization: 9 days

--comorbidities: none 21%, one 27%, >1 in 52%

    --hypertension: 50%

    --obesity: 37%

    --diabetes: 30%

    --asthma: 11%

    --CAD: 12%

    --COPD: 8%

    --cancer: 8%

    --CKD: 6%

-- score of Covid-19 severity on ordinal scale (see scale below)

    -- 4: 12%

    -- 5: 40%

    -- 6: 19%

    -- 7: 26%

        --overall 943 patients (89%) had severe disease at enrollment

--exclusion criteria: transaminase >5x upper limit of normal, CKD (not defined further), pregnancy/breast-feeding, anticipated discharge within 3 days

--98% of patients received the treatment as assigned

--primary outcome: time to recovery, defined as the first day after enrollment in which the patient was either discharged from the hospital or hospitalized for infection control purposes only (levels 1-3 in scale below)

 

Results:

--data and safety monitoring board (through NIAID) stopped the study early because of evident benefit

--remdesivir vs placebo:

    -- median recovery time, after randomization: 11 days (9-12) vs 15 days (13-19), rate ratio for recovery time: 1.32 (1.12-1.55), p<0.001

    -- mortality by 14 days: 7.1% vs 11.9%, HR 0.70 (0.47-1.04) [note: this confidence interval is dramatically skewed, a every strong statistical trend of benefit]

-- per baseline ordinal score of disease severity:

    -- baseline score of 4: 127 patients, rate ratio for recovery 1.38 (0.94-2.03)

    -- baseline score of 5: 421 patients, rate ratio for recovery 1.47 (1.17-1.84)

    -- baseline score of 6: 197 patients, rate ratio for recovery 1.20 (0.79-1.81)

    -- baseline score of 7: 272 patients, rate ratio for recovery 0.95 (0.64-1.42)

--serious adverse events 21.1% vs 27.0%; similar numbers of patients in each group discontinued their med because of adverse event

 

Commentary:

--Covid-19 illness scale (ordinal scale):

    1. Not hospitalized, no limitations of activities

    2. Not hospitalized, but limitation of activities, home oxygen requirement, or both

    3. Hospitalized, not requiring supplemental oxygen and no longer requiring ongoing medical care

    4. Hospitalized, not requiring supplemental oxygen but requiring ongoing medical care (can be for other medical conditions besides Covid-19)

    5. Hospitalized, requiring supplemental oxygen

    6. Hospitalized, requiring noninvasive ventilation or use of high flow oxygen devices

    7. Hospitalized, requiring invasive mechanical ventilation or ECMO

    8. Death

-- the above study was shortened because of evidence of early benefit. In addition, the study was a preliminary analysis; there will be an additional analysis looking at the full, proposed, 28-day follow-up

    -- one concern in general with shortened analyses, in this case as dictated by the monitoring board and in the setting of being a preliminary analysis, is that a shortened analysis will tend to exaggerate the benefits (noted before expected) over the longer term adverse events (not have time to emerge/be counted)

--remdesivir is a pro-drug metabolized to an analogue of ATP that inhibits viral RNA polymerases, with broad anti-viral activity: in nonclinical models, remdesivir has both prophylactic and therapeutic efficacy for filoviruses (eg Ebola) and coronaviruses (SARS-CoV and MERS-CoV). And in vitro efficacy has been shown against SARS-CoV-2. see http://gmodestmedblogs.blogspot.com/2020/04/covid-remdesivir-may-help-asymptomatic.html, a review of a drug-company sponsored compassionate use study of remdesivir

--there are certainly issues with having so many sites in so many countries as in the above study:

    --their inclusion criteria were for people hospitalized for severe disease, and though this was reflected in their patient skew to more advanced disease (as noted above), there were people with less severe disease (127 patient, 12% of the total, had ordinal score of 4), which raises teh possibility that entrance into the study may not have been consistent between sites (and/or countries), which also suggests that there might have been other significant differences in terms of their standard management of patients, or general adherence to protocol. there was no explanation of the reasons these less sick Covid-19 patients were included. and, there were no data on site- or country-specific outcomes. 

    --was there a difference in thresholds for admission from site to site? Did some sites systematically admit patients at an earlier time than others (and the global median may have been distorted by the overwhelming presence of US sites?)

    --they did not breakdown outcomes by the number of days the remdesivir was actually given ("up to 10 days"): was there a systematic difference in different localities? were there differences in adverse events?

-- By baseline score of disease: it is clear that those who did the best had a baseline ordinal score of 5, which was the only one by itself that was statistically significant. Those with baseline score of 4 were very close to being statistically significant, and actually had a rate ratio of recovery very close to those with baseline score of 5. The sicker patients, with baseline scores in 6 or 7, seem to have had essentially no benefit. This was not so different from the findings in the compassionate use article, per the blog cited above

    -- it would be useful to know the mortality rates stratified by the baseline ordinal illness scale: did those on remdesivir who had initial ordinal scores of 4 or 5 have a significantly lower mortality rate?? if so, that would support giving remdesivir to patients earlier in the course of the disease

 

--the adverse event rates are a bit confusing: placebo was worse, but (i think) that is likely because many of these events may have been from the untreated virus but considered a potential drug-related adverse event. This is also likely true for the remdesivir group, but the relative equivalence of these adverse events may actually speak for the benefit of remdesivir (ie: the increase in remdesivir-associated adverse events might have been compensated for by the decrease in the illness-related ones)

 

Limitations of the study:

-- this study was done in a time of restricted travel, hospitals were restricted at the entrance of nonessential personnel, much of the training and monitoring visits were performed remotely, many sites did not have adequate PPE and trial-related supplies (e.g. swabs). So, these results might have been different under different circumstances, as seems to be happening now (for better or worse)

 

so, remdesivir does seem to decrease the recovery time in those with Covid-19, though a few comments:

-- the effects seem to be highest in those hospitalized who were not so sick (either requiring regular supplemental oxygen or were not needing oxygen at all). There was no clear benefit in those who were sicker. This all suggests that earlier treatment is more beneficial (also found in the compassionate use article mentioned above), as found in their rate ratios for Covid-19 recovery

-- though there was no statistically significant mortality benefit, it was really pretty close to significant (and the 30% found is pretty impressive). we shall see what the final report finds....

    -- it would be great to have the breakdown by initial illness severity: specifically, would giving remdesivir early to those admitted to the hospital with ordinal scores of 4 or 5 significantly decrease mortality?? if so, that would further support having a more concerted effort to identify people with early signs of lower respiratory tract Covid19 and treating them with remdesivir.

-- But, despite its benefits in the above preliminary publication, remdesivir does not appear to be a silver bullet: people still they found long hospital stays and significant mortality. But, it certainly increases the potential of remdesivir as a component of combined, perhaps triple therapy, being significantly more effective, as suggested in http://gmodestmedblogs.blogspot.com/2020/05/covid-early-use-of-lopinavirritonavir.html . (triple therapy being used successfully in some other difficult-to-treat diseases, such as HIV or TB)

    --and, it might be a good idea to make sure that people with Covid-19 are vitamin D replete: http://gmodestmedblogs.blogspot.com/2020/05/covid-vitamin-d-deficiency-may-lead-to.html

geoff

 

If you would like to be on the regular email list for upcoming blogs, please contact me at gmodest@uphams.org

 

to get access to all of the blogs:

1. go to http://gmodestmedblogs.blogspot.com/ to see them in reverse chronological order

2. click on 3 parallel lines top left, if you want to see blogs by category, then click on "labels" and choose a category​

3. or you can just click on the magnifying glass on top right, then  type in a name in the search box and get all the blogs with that name in them

 

please feel free to circulate this to others. also, if you send me their emails, i can add them to the list