Exciting news this week as the FDA has approved the use of Convalescent Plasma (CP) for treating COVID-19 patients in emergency hospital settings.
It is not my purpose here to discuss the politics surrounding this approval, nor do I want to address any potential hurdles created by this decision for researchers who are trying to do better controlled therapies. I will leave those discussions to others.
What I would like to do is to take some time to put into perspective the value of this therapy, the associated problems and how this advance can lead to further, more specific treatments.
Bottom line: CP provides a new and potentially useful tool for treating patients, but it is not a cure nor a major broad-spectrum therapeutic for COVID-19.
Plasma, as you may know, is the yellow liquid remaining after all of the red and white blood cells have been removed from your blood. It consists of proteins, nutrients, platelets and waste products that have not yet been removed by your kidneys. The use of CP in this case is with the hope that specific antibodies to COVID-19 left in this liquid from recovered patients will be able to have a therapeutic effect on other infected people.
ARE ANTIBODIES A GOOD PROTECTION?
Antibodies are expected to be protective against virus in the blood as long as they bind to the right parts of the virus, in this case the binding section of the spike protein.
Antibodies DO NOT enter infected cells; they DO NOT kill virus that is replicating in the cells; their effectiveness is in preventing released virus from infecting OTHER cells.
DOES CP ACT LIKE A VACCINE, TO PREVENT INFECTIONS IN THE FUTURE?
- The antibodies in the plasma should be able to neutralize existing virus, but they themselves convey no lasting immunity. They will be cleared from the blood over a period of time.
When you are vaccinated and develop your own immune response and manufacture your own antibodies, the cells that make those antibodies remain circulating in your body, ready when confronted by a virus in the future to start making antibodies. This is why vaccinations work. CP is not a vaccination; it is a short-term therapeutic.
ARE ANTIBODIES THE BEST IMMUNE RESPONSE?
Antibodies are an important element of protection, but not the only one. As you may know, antibodies are produced by B-Cells and are the blood-borne protein protections. There are also T-Cells which provide cellular immunity and are able to attack infected cells directly. CP provides no cells of any type; therefore, it does not provide any cellular immunity.
IS CP SAFE?
It is precautionarily safe when administered under strict controls and monitoring. It should be used only in hospital settings with inpatients and not by physicians in their offices.
There are several potential serious side effects to the use of any CP. The most common adverse reactions associated with CP therapy are transfusion-related events, involving chills, fever, anaphylactic reactions, transfusion-related acute lung injury, circulatory overload and hemolysis. These are serious side effects that can result in death if not treated quickly.
There is also a risk of transfusion-transmitted infections, such as HIV, hepatitis B virus, hepatitis C virus and syphilis. When collected plasma is well-controlled, tested, stored and monitored, the risk of these infections is very low; problems arise if there is pressure to collect and distribute plasma too quickly without these controls in place or if unregulated individuals provide plasma outside of normal distribution systems.
THE NEW STUDY THAT HAS LED TO FDA APPROVAL
Earlier this month, the Mayo Clinic announced a study of the use of CP in COVID-19 patients. This article has not been published as yet, but preprints of the study were made available.
It is critical to understand that this preprint is NOT a publication. As clearly stated in the preprint itself, this “preprint has not been peer-reviewed. It reports new medical research that has yet to be evaluated and so should not be used to guide clinical practice.” Further, there is no indication of which journal it will be submitted.
I have read that preprint and my comments are based on my reading, not on the talking heads who have been opining on multiple news channels.
Patients included in the study presented between April 4 and July 4
The Mayo Clinic studied more than 35,000 patients. The preprinted study claims that patients who received transfusions within three days of their Covid-19 diagnosis had a seven-day death rate of 8.7%, while patients who received plasma treatment after four or more days had a mortality rate of 11.9%. The difference met the standard for statistical significance and may be interpreted as a 26% decrease in mortality.
This type of study, by the nature of its construction is difficult to analyze. The study did not include a placebo group for comparison. A better study would look at groups of individuals treated at day 2, 3, 4 and 5 for example. If there is, in fact a valuable window of opportunity for treatment, such a study would be much better able to identify it.
I have read the preprint of the study and it has a variety of issues that would cause any reviewer to be cautious when drawing conclusions. Here are a few of those issues:
- Patients were divided into two groups, identified in the 3 months between April 4 and July 4. We do not know what changes in protocols were made during that period, a period in which much progress was made in treating patients in hospital settings. We also do not know the patient distribution regarding this time period, how many were identified in each month, and if there were differences in the distribution of each of the groups.
- The first group consisted of patients who received treatment WITHIN 3 days (on day 1, 2 or 3) while the second consisted of patients receiving treatment AFTER 4 days (between days 4 and 11). It is important to understand that these time points actually refer to the days AFTER a test was done and infection was confirmed. It DOES NOT relate to the date of ACTUAL infection. We do not know if the individual presenting at the hospital had been infected for 2, 3, 4 ,5 or more days before they were admitted. The assignment to a group was based on their testing at admission, so the results need to be questioned as to whether that selection process had any built-in prejudice, including their clinical status at their time of presentation to the hospital.
- The plasma used to treat the patients was NOT controlled. In a well-controlled trial, the plasma would have been pooled so that each patient received the same plasma. This is extremely important. The immune response from one individual IS NOT identical to the immune response from another individual. As I have explained in the past, the immune system does not “construct” an antibody to react with a virus, it rather randomly creates billions of antibodies and expands those cells that coincidently bind to the virus. Therefore, the use of plasma from patient A is not identical to the use of plasma from patient B. The lack of control here adds variability to the results.
- Plasma needs to be “typed” before transfusion. Much like your blood type, there are several factors that need to be checked before a particular plasma can be determined to be safe to transfuse into a specific individual. Lack of complementary individuals can result in extremely severe reactions including death. We do not know at this point whether there are any links between blood type, histocompatibility antigens and other identifiers in the effects of the disease or the ability to recover. These types of genetic links and predispositions are relatively common with most diseases and may very well be important with this virus also. Any such link, which will be understood several years from now, might skew the results.
- The designation of the two groups is potentially problematic. The study is clear; one group was treated WITHIN 3 days of determination. Some may have been treated on the first day. The second group was treated AFTER 4 days; their treatment may potentially have actually occurred a week or more after identification. If CP were to be effective, we would expect that for an antibody transfusion, the sooner you get that treatment, the more likely you would be able to reduce viral replication and subsequent infection of additional cells. Late transfusion would be expected to have less effect. The scale of the effect would be expected to be proportional to the quick application of the therapy. The net effect of the groupings in this study then would be expected to amplify those actual differences. Comparing someone treated on day 2 to someone treated on day 11 would be quite different than comparing individuals actually treated on day 3 to those treated on day 4.
Also, there is no “control” for mortality. We do not know what the mortality percentages were for patients who were not treated at all. Was there a significant reduction in deaths for those patients that were treated after day 4? After all, if the mortality rate for untreated patients was comparable to the 11.9% seen in patients treated after 4 days then it would suggest that there is little if any benefit for using such a treatment for patients with severe disease in the hospital.
- The other major issue with the groupings used in this study has to do with the treatment and condition of the patients. As this study was retrospective (they looked at the mortality rates at a point well after the treatments were done, not in association with a choice to do the treatment) there are multiple factors that will influence the results.
For example, we do not know if the patients who received the transfusions in the first 3 days of their admission were part of a subgroup of physicians who were treating their patients more aggressively than those who waited to begin treatment. We do not know what other therapies, tests and protocols were used with these patients that might have predisposed those patients to do better. A simple question would be, “were those patients who were treated early also given steroids like dexamethasone earlier than the other patients?”
As another example, waiting for over 4 days before beginning treatment may have resulted in patients developing more severe disease, additional symptoms or having required or been given treatments, tests or protocols different from the early onset patients.
WHAT ARE THE IMPLICATIONS?
Positive results are always encouraging.
We need to understand that if these results are correct, that this therapy would be useful primarily at the very beginning of infections; and yet those early infected people seldom present to the hospitals where this therapy would need to be delivered.
Also, this would require maintaining a supply of plasma from recovered patients. Some of those patients will have plasma that has high levels of neutralizing antibodies, others may have lesser levels, and some may have scant levels. Which one do you want? How will you know which you are getting? Who will standardize the plasmas so that physicians will know what the antibody titer (the amount of antibodies in each standard volume of plasma) is for each plasma? Who will get the “best” plasma? Who will pay for the hospital stay, the transfusions, the monitoring, etc.?
NEXT STEPS
If these studies are confirmed by subsequent research, it gives even more strength to the development of monoclonal antibodies to these viruses. Monoclonal antibodies promise a very consistent, directed and scalable antibody therapeutic. The use of cocktails of multiple monoclonal antibodies with independent specificities would provide a strong source for immune therapy in advance of a successful vaccine.