The biggest update about investigational therapies to combat the COVID-19 pandemic is this recent NEJM article, which showed that remdesivir is associated with clinical improvement in 36/53 patients who received this medication for compassionate use. In this post, I’ll take a deeper dive into this specific paper and provide my own personal opinions. For more background about remdesivir, please see a previous post. This was not a randomized clinical trial, so we cannot make an efficacy claim meaning that we cannot tell if remdesivir was the specific cause that these patients improved.
Background
Remdesivir is a nucleotide analog that is metabolized to create a dummy adenosine triphosphate group. Viruses normally need nucleotides like adenosine to replicate their viral genome, so adding these dummy group limits viral expansion.
In January to February, there were no widespread clinical trials in the United States for remdesivir so clinicians had to send these request approvals to get this drug. Gilead only approved requests for patients with RT-PCR confirmed COVID, good kidney function, and no failed liver. Once approved, patients received a 10 day course of remdesivir (200 mg IV day 1 and 100 mg daily afterwards).
Patient Demographics
Below are the baseline patient demographics for the patients included in this study. 2/3 were on a ventilator and 1/3 just had noninvasive oxygen support (i.e. nasal cannula or face mask). The big takeaway for me was that the patients’ kidney function was really good (Cr of 1) and liver function was really good (ALT, AST < 40). This means they did not want to take a chance on having really bad adverse effects if this drug is poorly cleared.
Clinical Improvement
The paper then had three figures showing clinical improvement with remdesivir.
In the first figure, they stratified every patient’s baseline condition based on oxygen requirement along columns and described their resulting medical status in rows. In patients with invasive ventilation, 56% improved with 8 patients getting discharged and another 8 now without any oxygen requirement. The results were even better for those with noninvasive ventilation, in which 5/7 patients were discharged.
The next figure took a deeper dive at the oxygen status of every single patient who received therapy. There are several interesting stories from this figure.
There are 3 patients who were on ECMO (extracorporeal membrane oxygen support) requiring an external device to oxygenate their blood, who made near miraculous recoveries.
Of the patients who were intubated, 1/2 recovered, 1/4 remained the same, and 1/4 died. There was greater clinical recovery with remdesivir in patients with less severe oxygen requirements, in which all but two patients recovered.
The next figure plotted the course of clinical recovery with remdesivir. The authors claimed that it took a few weeks to lead to full clinical improvement with this medication, patients with noninvasive ventilation did better, and younger patients recovered faster. None of these are big surprises.
Adverse Effects
Finally the authors listed all the adverse effects. Honestly, we cannot tell if these adverse effects are from the medication or from worsening clinical status from the virus. This table was not too helpful from my assessment. We know hepatotoxicity is associated with remdesivir, but we do not have an idea of how much hepatic enzymes increased with this medication. Presenting the raw numbers would have been way more helpful.
Conclusions
From this trial, it seems like remdesivir is associated with clinical improvement severe COVID-19.
However, it is difficult to make any large claims about remdesivir efficacy. We do not know the natural history and clinical improvement for patients with COVID-19. Are patients recovering because they have built enough antibodies (which takes weeks to mount) or are they recovering because of the medication?
We can speculate that remdesivir won’t work well for patients who have large viral load causing multiorgan failure. Perhaps this drug works better when given earlier.
