Note: The Pregistry website includes expert reports on more than 2000 medications, 300 diseases, and 150 common exposures during pregnancy and lactation. For the topic Coronavirus (COVID-19), go here. These expert reports are free of charge and can be saved and shared.
The story of emerging treatments for COVID-19 has been confusing for various reasons. First, it is confusing because infection with SARS-CoV2 (the virus that causes COVID-19) produces a range of disease severity in different people, from a complete lack of symptoms on one end of the spectrum to severe complications in multiple organ systems leading to death at the other end. This means that what might be helpful for some patients might be useless, or even harmful, if administered to others. Second, most people get their news about the pandemic from the general media, as opposed to medical and scientific media, but the general media are set up to report science news in line with how the publication of scientific results works. Consequently, the story has taken various twists and turns, following press releases about one medication or the other, sometimes followed by corrections, due to the relevant data analysis being scrutinized by other researchers, and in a couple of cases, even due to the retraction of papers from prestigious medical journals.
On top of this, when preliminary findings from studies in particular clinical situations suggest that giving one medication could make the difference between life and death, there is pressure to publish data, or summaries of data analysis, early –before the full scientific process has run its course. When this happens, health systems of different countries do not necessarily respond in the same way. Thus, one treatment may be continued, expanded, or discontinued in some countries, but not in others, in turn generating more data, and these new data then are added to the scientific analysis.
This is what happened in the case of a large British study at Oxford University known as the RECOVERY trial in which scientists have been analyzing the effects of several medications on COVID-19 patients using a range of statistical methods. To get scientifically valid information, researchers need to follow certain protocols and wait for the analysis of all data. The statistical analysis is set up in a way to evaluate what scientists call clinical endpoints, which are things that we are interested in knowing whether they are affected by the treatments under study. For example, one end point designated for a research study might be 30-day mortality, the effect that a treatment or situation being studied has on whether patients are still alive after 30 days. Another endpoint might be time in the hospital, or time on mechanical ventilation, while you might also choose the rate of a particular complication as an endpoint. When novel treatments are being studied, we hope that the analysis will show that a given treatment decreases mortality (lowers the percentage of patients who die) compared with standard treatment, shortens hospital stays, shortens the time on mechanical ventilation, reduces rates of complication, reduces the percentage of cases that evolve from mild or moderate disease to severe disease that threatens or takes life, and things like that. Typically, a study will designate one or two primary endpoints (endpoints that are clinically the most important, such as whether a medication keeps people alive), plus various secondary endpoints (endpoints that are still important but not necessarily at the top of the list). In the case of a pandemic, like COVID-19, in which many people are dying, the science can get messy, because the analysis is also designed to let investigators know early if the data are beginning to reveal something that might make a difference between life and death, such that they might need to put people on, or take them off, a medication, or even halt the study. As an extreme example, if there is a study that is supposed to be 6 months long, with researchers blinded (they don’t know which patients are getting the real experimental treatment and which patients are getting a placebo), and after 3 months there are enough data to show that receiving the drug cures people while those getting the placebo die, the study has to be stopped, so that the people who are on the placebo can start getting the real medication. In such an extreme case, there is an ethical imperative that conflicts clearly with the scientific process, but in most cases, there is no such dramatic early revelation, in which case the ethical imperative weighs toward the continuation of the study, as planned.
In the case of the RECOVERY trial, several medications are under study and there is nothing so dramatic as a possible need to stop the study, but study directors were alerted early that a certain medication appears to have beneficial effects on mortality and some other primary endpoints in COVID-19 patients sick enough to require supplemental oxygen. Consequently, on June 16, although RECOVERY data had yet to be published in peer reviewed articles (papers read, analyzed, and criticized by other researchers prior to publication), information about what initial data analysis suggested about the medication was revealed in a press release and presented by researchers in a press conference. Six days later, on June 22, the researchers began publishing the data, still not peer-reviewed, in what are called preprint articles on a health sciences preprint site called medRxiv (pronounced med-archive).
Normally, such pre-prints would have to be taken with a very large grain of salt, but there is a long history with the drug that is under discussion. It’s called dexamethasone, one of a group of drugs known as corticosteroids, which frequently are given to treat conditions that involve inflammation or attack by a person’s own immune system. Given that COVID-19 involves inappropriate activity of parts of the immune system, there was a rationale for giving dexamethasone back in January when COVID-19 was spreading, but studies suggested that the drug could actually be harmful. Consequently, it was not given in western countries, but doctors in China and Japan kept giving it to COVID-19 patients, so more data kept arriving, along with data on another corticosteroid that some were giving called methylprednisolone.
Based on the June 16 announcement and the preprints that have come out since that time, the case appears very strong that dexamethasone, given in a short bout (maximum 10 days), dramatically decreases the death rate when given to people with COVID-19 that is so severe that they are in the intensive care unit (ICU) on mechanical ventilation. The drug also appears beneficial (although less dramatically) in COVID-19 patients who are not severe enough to need mechanical ventilation, but who do need some kind of supplemental oxygen. For example, if you are admitted to the hospital for care that includes low flow oxygen through a tube in your nose, or a mask, the prematurely released study results indicated that you would benefit from dexamethasone. On the other hand, if you have COVID-19, but a more mild case from which you have been ordered to recover at home (the most common COVID-19 scenario), you will not benefit from dexamethasone.
As for pregnancy, dexamethasone is a medication that can be given with caution when the benefits outweigh the risks. It’s actually one of the corticosteroids that can be given to a pregnant women to mature the fetal lungs in cases when delivery is going to be preterm. In the case of COVI-19, the tradeoff means that you can receive dexamethasone, if you are hospitalized with COVID-19 and receiving at least low flow oxygen.
This is a very simplified version of the story, since, as noted earlier, the RECOVERY trial is evaluating several medications and there are hints that some other drugs may also provide some benefit, albeit less dramatic than dexamethasone. One such medication on which researchers are focused is remdesivir, which may interfere with the ability of SARS-CoV2 to make little baby viruses when it infects cells, and which very much needs to be evaluated statistically for its effects when used in combination with dexamethasone, so you will be hearing about it in the weeks to come. The same is true of another drug called tocilizumab, which works against a key agent released by the immune system as part of a massive but counterproductive response against SARS-CoV2 known as cytokine release syndrome.