Why the answer to COVID-19 might lie in your blood

Surrounded by questions about COVID-19, we as the SECTOR set out to find answers. Luckily, we came across professor Hans Zaaijer MD PhD, who can be seen in many newspapers and late-night tv shows like ‘Op1’ to talk about COVID-19 developments or anything blood bank related. Zaaijer is head of the blood-borne infections department at Sanquin, our national blood bank.  His research interests include the detection and epidemiology of blood-borne infections in addition to decision making and ethics concerning preventive measures. 

What does your job entail?

The majority of my time I spend at Sanquin where I focus on three different research departments. The first one researches blood-borne infections. Think about questions like: “How do we keep infectious threats from the national blood supply?” An example of such a threat is West-Nile fever. Eighty percent of the cases do not experience any symptoms, and will not think twice about donating blood. Infectious diseases which only result in symptoms in a fraction of the cases are a big challenge for Sanquin. In the past couple of years, we have seen samples infected with mad cow disease, Q fever and hepatitis E. One year, of every 600 donation samples, one was infected with hepatitis E. The latter is not deadly for most people, but very much so for those who just went through an organ transplantation or for those who are undergoing chemotherapy. The second department I do research at occupies itself with donor studies and the third one with Transfusion Technology Assessment. 

The rest of my time I spend as a clinical virologist at the Academic Medical Centre (AMC). This is to keep my experience as a practitioner on par. 

Aside from this, I am a medical microbiologist at the diagnostic virus lab. You can see me as Sanquin’s main medical microbiologist for whatever disease might occur. The last position mostly has a legal background, since every organisation that works with human tissue is required to employ a BIG-registered medical microbiologist. 

What is the role of Sanquin during this pandemic?

Sanquin is the only organisation in the Netherlands which has quick access to blood samples on a large scale. Every day, Sanquin receives samples from 3000 unique donors. 99% of these donors allow the residue of their sample to be used for non-commercial research. 

In April, Sanquin performed a study to find out what percentage of the Dutch population has COVID-19 antibodies. We do this because we can. If research is possible and useful, we feel like we have to perform it. This was our so-called T-1-study, for which we tested 8000 blood samples. The result: 3% of the tested samples contained COVID-19 antibodies. A month later, this study was repeated and found that the prevalence of antibodies had increased to 6%. 

Upon request of the National Institute for Public Health and Environment (RIVM), this study is currently performed extensively. This means that Sanquin is now officially tasked to deliver this data. For the last five months, every week we test 2000 random donors on antibodies. Random in the sense that it is balanced by gender, zip code and age. It is interesting that we see that donors occasionally lose their antibodies over time.

How are you so sure about the results?

We always test every donation, despite the fact that this particular donor might have already donated many times. Samples are tested for hepatitis B and HIV, among others. After this testing, a few milliliters of sample remain, which is stored in case it has to be retested. Aside from this, we freeze one milliliter and store it for two years as is required by law. In public health emergencies, we are allowed to test a fraction of each sample prematurely. This means Sanquin has an archive of a few million tubes of samples containing every donation from the past two years. 

This database has proven to be very useful as a baseline measurement for our antibodies tests, considering that we did not know the specificity of the test yet. In other words: it was not known how many times the test would result in a false positive. The archive could be used to compare the amount of antibodies in the samples from April 2020 to the amount of antibodies from a year ago, which should be zero, since COVID-19 didn’t exist yet. We used this perfect baseline measurement to support the specificity of our test and to be able to present an accurate result. In the Netherlands, only Sanquin is able to perform this kind of research. 

Do blood donors represent the general population well enough to not lead to a certain bias?

My usual answer to this question is: “Be my guest, perform a study that is more representative!” Of course, the sample might not be completely representative since donors have to be approved first. If you have a wild lifestyle with many different sexual partners, we would rather not receive a donation from you.

“So it's true that the donor population is not entirely representative of the average Dutchman, precisely because they have actually donated.”

However, you could ask yourself whether this matters for diseases which spread through the air or through what you eat. I think that this bias is negligible for diseases like this, including COVID. 

Take for example the West-Nile virus which spreads to humans via a mosquito bite. Different lifestyles of donors will probably not lead to a significant bias here as I doubt that donors use a mosquito net more often than non-donors. In contrast, one might argue that donors are more likely to be more observant or careful compared to non-donors. 

At Sanquin, we have a team performing donor studies. They research donors’ motives and characteristics. How do you get the idea to donate? How do you make sure a donor keeps donating? Why do we sometimes lose donors? Why do some donors portray their lifestyle as average, while they may have a more dangerous lifestyle in reality? What’s interesting is that several donors  who turned out HIV-positive decided to keep their dangerous lifestyle a secret. In a so-called exit-interview – so an interview with a donor we “fired” – the donor admits this quite easily. These interviews are highly confidential, but we are interested in studying this behaviour with regard to national blood supply safety. 

So it's true that the donor population is not entirely representative of the average Dutchman, precisely because they have actually donated. Fortunately, our sample size is very large. When somebody tells me this is not a representative sample, my usual response is “At least our sample size is half a million!”. 

There is also a group that abuses the blood bank by living a more careless life and donating blood just to test for diseases like HIV on a regular basis. These so-called “test seekers” are on the opposite side of the spectrum. 

“For a donor, however, it’s free and you get a free coffee and even a “roze koek””

We made a deal with the government that the day we started testing for HIV in blood samples, it should be free to test for HIV at the GGDs (Municipal Health Services) of large cities. Otherwise Sanquin would attract all kinds of unwanted donors, like test seekers. This was of utmost importance, since the first HIV tests did not work particularly well, so repelling the high-risk group was a priority. 

This is why we were not amused when it was recently decided that you should go to your general practitioner for an STD test, so it will cost you money if you’d like to find out if you have a certain STD.

For a donor, however, it’s free and you get a free coffee and even a “roze koek”. In addition to this, you can come in for a blood donation at your own convenience in contrast to an STD test at the GGD. To conclude, test seekers can be a result of the government cutting back on making STD tests easily accessible. 

Is COVID-19 a blood-borne disease?

There has been a lot of research into this. With every pandemic or epidemic, it is interesting to know for me whether the virus is contained in donated blood. 

We start with educated guesses stemming from academic virological analogues. Surely, this is not the first coronavirus. It is actually the seventh, following SARS1 and MERS; and the four ‘common cold’  corona viruses. It is a “snot virus”, and we know that those are not blood transmittable. However, the virus genome can be found in the blood of terminally ill patients. Around 6% of the group of COVID-19 patients in the intensive care unit have COVID-19 RNA in their blood. It is safe to say that those people do not donate blood, so this group is not that interesting to me. 

Meanwhile there has been large-scale research on whether donated blood contains the virus. It is possible to donate blood on Monday, start feeling sick on Tuesday and to get a positive corona test result on Wednesday. This means that you must have already been infected when you donated blood. Many asymptomatic COVID-19 patients must have donated blood, especially at the start of the pandemic. However, the coronavirus (note: the virus itself, not the antibodies) was not found in a large pool of random blood samples that have been tested at the start of the pandemic. 

Anyway, we know from practice that snot viruses are not blood-borne. This makes sense, because the virus has protrusions (spikes) which have to find the right receptor. These receptors are certain molecules on the lining of your lungs. If the virus gets lost in blood, it can never run into a receptor. I should be more careful with "never", but in any case, the first blood-borne flu virus is yet to be found.

Can you treat a patient with the plasma of cured patients?

It is a reflex with every outbreak that some doctors think you can treat very sick people with the plasma of people who have just been healed. So, curing very sick patients with the antibodies of already cured patients. This is called convalescent plasma therapy. I barely believe in that to be honest, which is pretty bizarre because I am one of the people who is very involved in collecting this material. We call on donors: "Have you just had corona and have been better for at least two weeks? Then please come and donate plasma." Thousands of people do that.

As a clinical virologist with 30 years of experience I have come to the conclusion that administering antibodies almost never works for patients with an advanced virus infection, because the patient has already had these antibodies for a long time. But with every epidemic there is an immediate call for convalescent plasma therapy. 

This was also the case for the discontinued study at the Erasmus University. It was discontinued  because these patients already have plenty of antibodies, so it does not make sense to add even more.

What I do know as a virologist is that if you give vulnerable people antibodies in advance, or within one or two days after they contract the virus, or in a very early stage of disease, it may work very well. And this has been proven with Hepatitis A and B, chickenpox, RS virus, rabies, smallpox, etc. This is an accepted and proven application of antibodies. In addition, administration of antibodies may work very well in more advanced stages of covid, in immunodeficient patients who are not able to mount an antibody response at all.

Is this how we should tackle the virus?

Yes! So that is why I am working full steam on the collection of the convalescent plasma. We now have a few thousand packets of it, with which we are going to do a new trial. It will be given to patients who have already been admitted to the hospital, but who are not that ill. Perhaps then administering antibodies will work.

When Trump was tested positive, he received a cocktail of medication. But in that cocktail Trump also received two artificial antibodies within one day of his positive test. These artificial ones are grown in labs, while we obtain real antibodies from donations. That's exactly how I think antibodies can work really well. Of course I cannot say anything about that because n = 1. It could just be that the administration of the powerful antibody drug within 24 hours after infection is the reason he was back in the White House so quickly.

We also use that plasma in a pooled way to make intravenous immunoglobulin (IVIG). That is a registered drug and not experimental by any means. This is currently not created from 5000 random donors, but from 5000 donors who all have corona antibodies.

I would like to use this IVIG in 3 different categories. The first group are people who will undergo an organ transplantation, because their immune system is shut down to prevent rejection of the organ. The second group are people who are undergoing chemotherapy. And the third group, which I find the most interesting and also ethically relevant, is nursing home residents. In collaboration with Venessa Harris, an internist infectologist at the AMC and with a professor of elderly care, we have submitted a research proposal to the Ministry of Health, Welfare and Sport. The idea is to use antibodies to passively protect all residents as soon as there is a case in a nursing home. And this is again the situation where I think antibodies work well. This proposal is now under assessment.

So we are working on collecting donations from donors who have gone through corona for the two different products, namely the one-to-one product, a bag of plasma from the healed person to cure a sick person and the pooling of plasma to create ampoules of concentrated antibodies to immunise vulnerable people.

Uptil now, we have only discussed passive immunization. Using a vaccine would be active immunization. Here you train the memory of your immune system. If your immune system is inhibited however, a vaccine won’t work well for you. So if you are undergoing chemotherapy, a vaccine is not the solution. Passive immunization probably is the only option for this group. The downside to passive immunization is that it only works for a few weeks to months, since the half-life of the working molecule is only three weeks.

Through all these epidemiological studies, we at Sanquin are becoming experts in corona antibodies. This is really an incredible amount of work for us, since we are not a very large organisation. In addition, we also work with the handicap of corona in the Amsterdam region where Sanquin is located. One of our most important analysts had a sore throat and had to wait five days for a corona test. This sometimes results in a real personnel shortage. But we keep trying our best.

What kind of modeling does Sanquin do?

Sanquin doesn’t create models themselves, but we do aggregate and report data. We are actually hobbying a bit. We are academics – of course we like to analyse our data, for example by age demographic, or look at whether infections take place more in urban or rural areas. To get to know more about these models, you should ask Jacco Wallinga, the head of the unit for Infectious Disease Modelling at the RIVM.

Modeling is based on the data we receive from test sites, information from GGDs and the serological knowledge of Sanquin. This is how the RIVM knows whether the intensive care departments will be at full capacity if we do not increase measures in time. 

Sanquin’s primary task is to provide safe blood products and to produce medicines that can be produced from blood. Like now, for example, those corona antibodies. Modelling is not what we do, but it is nice that you can see the demand growing for this kind of data. Blood banks are the only organizations where you have a set of samples from all over the country in just one day. We are also the largest PCR test lab in the Netherlands. Every day, all 3000 donations go through a whole series of tests which are 100% automated. 

A great thing is that we have a backup test channel in our lab in case there suddenly is a new blood transmissible infectious disease. As long as we don’t need it, you can do a lot of research there. The question remains whether West-Nile virus will soon occupy this channel. The European law states that if there is at least one human West-Nile patient in the Netherlands, we need to screen the donor blood with the corresponding PCR test.

In fact, my job is to predict how many tests we will need next year. Every two weeks we have a meeting where we look at all signs of what is happening in the area or what we could expect in terms of outbreaks. Of course, we are keeping a close eye on the West-Nile virus, which is slowly creeping up on us. Furthermore, we follow other outbreaks of tropical viruses in the Mediterranean, such as dengue fever or chikungunya, which you normally only catch on tropical vacations far away, but can now  also be found in Marseille for example.

The real modeling is a bit more advanced. The RIVM asks for example: “How much higher should the average temperature be before the tiger mosquito becomes native to the Netherlands?”. As Sanquin, we concluded from this that we should buy West-Nile virus test robots either way.

How are immunity and the reproduction number related?

The real reproduction number is the current transmission rate, that is the expected number of people one sick person infects. This is the number we hear about in the news almost every day. The real reproduction number is in turn related to the basic reproduction number R0. If you were to put an infected person in a group of people who are all 100% not immune and who do not exhibit any behaviour different from their normal lives, the disease would spread on average to R0 people. For COVID-19 we have an R0 of 2,5. So, without any measures like social distancing or a reduction of the maximum allowed group size, the average case would infect 2,5 others. 

Articles about COVID vaccination sometimes argue that at least 60% of the population must be immune to the virus. This metric has everything to do with the R0 value. The virus will die out if the reproduction number is 1 or lower. This happens if at least 1,5 of the 2,5 people who would normally become infected are immune. So 1,5 / 2,5 = 60% of the population has to be immune for the spread of the virus to decline while having relaxed the measures. Note that R0 is inherent to the virus, while the real reproduction number can be influenced through measures and vaccination. Passive immunization, with donor plasma, has hardly any effect on the real reproduction number because it is only temporary and can only be done for thousands of people and not for millions of people. If safety measures, like social distancing and washing your hands more often, become a natural part of our life, I believe that a much lower percentage than ‘60 %’ could be enough to get rid of the outbreak. 

The question is also whether it is possible to increase real group immunity in a controlled manner. They tried this in Sweden, the country that is well-known for not taking very strict COVID measures. Simulating this is pretty easy. You just have to think about some balls bouncing around in a box. Two balls hitting each other represents an infection. The fact is however that no society is homogeneous because it consists of different demographic or socio-economic groups for example. So maybe not exactly 60% of the Dutch population has to be immune to stop the spread of the virus. Purely based on that, different age groups have different amounts of contact. So if a subgroup that is responsible for the majority of contacts in society is more immune than the vulnerable group that just sits on the couch, this can help contain the virus. These considerations are also relevant considering whom to vaccinate first, once vaccines become available.

Does this mean that we could let all the young people party?

Yes, that could be an option. Maybe we even need to report a reproduction rate per age group in order to make this more clear. The RIVM does not only get one number from us like “11,2% is immune”, but they also get this information per age category and per region. We also see that adolescents are on average more immune than other age categories. 

Do you think the Netherlands is handling COVID-19 the right way compared to Germany for example?

There is something that they do in Germany that I do not understand. Maybe that is because they admit everyone to the intensive care unit much too quickly? Maybe they have more money than we do? I'm not quite sure. We are not doing that bad in the Netherlands. However, we are not the best either. Germany has a higher intensive care capacity per inhabitant than we do and they also have larger test sites. This is very useful in a pandemic. On the other hand, my profession, as a medical microbiologist, is extremely rare in Germany. It turned out that they regretted this during an E-coli outbreak a few years ago, when there was no medical microbiologist who initially could identify the E-coli strain involved. In response, Germany adapted their education to train more medical microbiologists such that they could make more sense of the results in future outbreaks.

We are also very small in comparison to Germany. From an epidemiological point of view, we actually only have one large city, namely the Randstad. Measures per region, city and province make less sense in the Netherlands than in countries like Germany or Italy.

Finally, I think the Dutch government handles the pandemic very well. I have also noticed a shift in the role of the RIVM from decision making to giving advice. After all, decision making is something inherently political.

Next week, an article from Luc Zeinstra will be published where he talks about his experience of organizing the Forum.