Of course, the immune system can mount a robust response to SARS-CoV-2, as vaccine tests have made clear. In addition, however, there are many question marks. People exposed to the virus do not always produce many antibodies against it, and there have been several cases of reinfection. We are not sure how long immunity lasts or whether it correlates with antibody levels or something else – there is not even much evidence that antibodies are useful.
To give you an idea of the challenge of solving all of this, let’s examine three recently published articles that address the interaction between the immune system and COVID-19. One finally provides some evidence that antibodies may be protective, another indicates that neutralizing the inflammatory response may help, while the third suggests that immunosuppressants in no way affect the outcome of the disease.
Good antibodies
Antibodies are a relatively easy way to track an immune response and have been used for this throughout the pandemic. But early studies found that the number of antibodies produced in response to an infection varied dramatically between patients. There have also been clinical trials testing whether the use of antibodies obtained from previously infected people could help treat those suffering from COVID-19 symptoms, with the FDA eventually granting it a controversial Emergency Use Authorization. President Trump also received an experimental treatment of mass-produced SARS-CoV-2 specific antibodies.
The strange thing about all this is that we are not sure if the antibodies are really protective. Other trials of antibody treatments for the infected have produced ambiguous results, with no clear benefit of receiving an antibody boost. And although immunity levels appear to correlate with antibody levels in some studies, we cannot be sure that both are not linked to some other aspect of immune function – perhaps antibody levels are simply a reflection of the activity of T cells, to give an example.
A new article from researchers in Argentina is small, but it suggests that antibodies can help people with COVID-19 – but only if a treatment is given early enough. The research design is solid: a randomized, blinded trial in which some people received a transfusion of saline, while others had antibodies to the infected previously mixed with their saline. Critically, all transfusions occurred a few days after the onset of COVID-19 symptoms. The study’s only limitation is that it occurred while the number of cases was decreasing in Argentina, so it was stopped when they had trouble recruiting patients.
Of the 160 patients, all over 65 years of age, who were included, 25 of the 80 in the control group ended up with severe respiratory symptoms. In those who received plasma containing antibodies, only 13 had these symptoms. Eliminating the six individuals who had to drop out of the study further improved the numbers. Finally, those who received plasma with the highest levels of antibodies tend to have an even better prognosis, although the number of patients here is even lower.
Those who received plasma also tend to have less severe results, such as admission to the ICU and the need for ventilation. However, the numbers for each individual question were small, so none of these measures reached statistical significance.
The researchers note that in some other studies, those who received plasma treatments early tended to do better, but the general population treated at different stages of the infection showed no effect. If that’s right – and this study is small enough to really need to be replicated – then it would present the first clear evidence that antibodies are useful. This can be critical not only for treating infected people, but also for tracking immunity and monitoring risk in populations with varying levels of vaccination.
Bad inflammation
The other lesson from studying antibodies is that carefully defining your treatment population – recently symptomatic elderly people in this case – can be instrumental in identifying a clear effect, although it can make it more difficult to find enough patients to do a full study. This lesson can also be applied to a draft manuscript that describes a study on whether we can limit the effects of COVID-19 by neutralizing the inflammatory immune response. Genetic studies of patients with COVID-19 indicated that variations in some immune signaling molecules were associated with the severity of the disease. But studies of drugs that blocked the effects of an inflammatory signaling molecule called interleukin-6 have shown no effect. The researchers suspected that this was because they accepted a wide range of patients.
So, to make matters worse, they started treatment with interleukin-6 blockers when patients were admitted to the ICU. The trial involved about 800 people, about half of whom served as controls. The rest received one of two different inflammatory blockers. Among those who did not receive a drug, the death rate was around 36 percent. For those who were treated, however, mortality was 27 percent.
This may not be a huge difference, but if it persists, it can make a significant difference in population-level survival. And the United Kingdom’s National Health Service has already warned its doctors about the results, by starting a re-evaluation of these drugs.
Is the immune system overestimated?
All of this would apparently put the immune system at the center of the results of COVID-19, which shouldn’t be surprising at all. But another study published this week raises questions about even that. Here, researchers followed the results of more than 2,000 COVID-19 cases that passed through the Johns Hopkins medical system in March. Of these, more than 100 were using drugs that left them immunocompromised. And when patient results were analyzed, there was no noticeable difference between those who were immunocompromised and the rest of the population. The researchers measured mortality, length of stay and the need for ventilation, but none was significantly different.
It is important to emphasize that “immunosuppressed” does not mean “unable to mount any immune response”. But the answer is usually quite limited.
What to do with all this? The good news is that if the antibody results are maintained, they indicate that the antibodies can provide us with not only therapy for those at high risk of serious infection, but an easy way to track who may be protected in the future. These results are not really confused by the results with immunocompromised individuals, since antibodies are not normally produced during an initial infection, unless it drags on for a while (they take a few weeks to start appearing at measurable levels).
In addition, however, things get very complicated. The immune system has several aspects (immunity based on T cells, dendritic cells, innate immunity, etc.), and we do not really know how many of them have been completely suppressed in immunocompromised individuals. In addition, if the inflammation turns out to be harmful in some cases, it is possible that some forms of immunosuppression may actually be useful.
But the picture these articles really show is that both the immune system and its interactions with the virus are extremely complex. If a study does not have enough people to focus on specific patient populations, or provide treatments at specific points during infection, there is a chance that important effects will be calculated. One problem is that, at this point, we have many smaller and less focused studies already published, leading to an incomplete and confusing picture. Finally, there is undoubtedly a lot of patient-to-patient variability that confuses things further.
All of this explains why there are so many confusing and seemingly contradictory publications out there. It reinforces the need to treat any single result as conclusive. Over time, we will build a clearer picture of the course of a SARS-CoV-2 infection and the immune system’s response to it. Given the time it will take, however, the focus will undoubtedly be on the rush to vaccinate as many people as quickly as possible.