Serological testing in Germany shows resistance to coronavirus SARS-CoV-2 at 14 percent 3

First preliminary results from widespread IgA/IgG SARS-CoV-2 resistance antibody testing in Heinsberg Germany

German Karneval goer ready to party
German Karneval goer ready to party, photo credit Cooper Le on Unsplash

Being an active research scientist or having a life science research background during a once-in-a-lifetime worldwide epidemic is a remarkable thing. Whether reading pre-prints on MedRxiv or looking into the latest news of the day or reading interesting links from a variety of sources, having dozens of tabs open (and bringing your browser and computer to a crawl) is an endless feast of real-time information.

If you were curious how the pandemic has affected media consumption by generation, well here’s a handy visualization. Online press receives a healthy bump; what is it about Gen Z and Millennials about their consumption of online video, up 51% and 44% respectively, topping the increase of all media sources?

Anyway I last wrote about Cellex US having FDA Emergency Use Authorization approval for detection of protective antibodies (in this case IgG/IgM) against the causative virus for COVID-19 disease, which goes by the scientific name of SARS-CoV-2. It is notable the World Health Organization has decided not to use the scientific name, likely to not overly alarm people in the early days of the pandemic. Nonetheless I use the scientific name to be specific, there are some four coronaviruses that cause the ‘common cold’ and three other coronaviruses that cause nasty diseases (SARS ‘classic’ from 2003, MERS in 2012 and now SARS-CoV-2).

In Germany, a small town near the Netherlands border called Gangelt in the larger area called Heinsberg has a festival marking the beginning of Lent called Karneval. This town of Gangelt (about 12,000 residents) has been called ‘Germany’s Wuhan’ as it has had the largest number of infections and deaths in the country; Heinsberg (the larger principality with about 42,000 residents) as of this writing (10 April) has 1,521 cases and 45 deaths.

On 31 March 2020 it was announced they would test 1000 residents that were representative of the population, with the goal of determining at greater detail how the SARS-CoV-2 coronavirus spreads. Dr. Hendrik Streeck a Bonn virologist said at that time

If there are ways of preventing the illness from spreading in our environment, we want to know what they are, with the goal of finding out how we can freely move about in the environment together.

Dr. Hendrick Streeck, in The Guardian on 31 March 2020

It is surmised through active research across countries worldwide that the coronavirus is not spread through casual contact, such as grocery shopping or sitting next to some random people in public transit, but close social interaction such as hugging, singing, laughing and social kissing. (In Rhineland Germany they greet each other with a Bützchen, or kiss on the cheek.)

On 15 February 2020, a group of about 350 gathering for the Karneval listened to live music, mingled with food and drink, heard the town leaders and generally socialized for a total of four hours. 7 individuals, all attending that event, later tested positive, among the first of all cases in Germany.

Of the 1,000 volunteers who submitted for both rRT-PCR testing (via throat swab for active virus infection) and antibody testing for resistance to SARS-CoV-2 (via blood draw to test for anti-SARS-CoV-2 antibodies, specifically IgA and IgG), the results of the first 500 samples were released yesterday.

In the report (link to the PDF in German is here, the page this PDF comes from is located here) are some surprising, and very interesting, results: about 2% have tested positive for the virus via PCR testing, and 14% have tested positive for antibodies against the disease. Originally given the population size and infected rate, the researchers were expecting a smaller number of IgA/IgG-positive samples (perhaps 5%). In other words, one in seven have been infected and have recovered with antibodies against the virus, and can be considered ‘immune’ from being infected again (although that is also being actively studied).

This preliminary data has large (dare I say huge) implications for public health policy, as questions are currently being raised as to how the economies and society of many nations (and literally trillions of dollars of economic activity) will restart again.

From a translation of the German PDF, they claim a >99% specificity; will have to await the publication of the results to find out the details. (Friendly reminder, specificity is the ‘true negative’ rate, thus 1-specificity is the ‘false positive’ rate or <1%.)

The larger implication of this 14% number is the calculation of the death rate, known as a the ‘Case Fatality Ratio’ or CFR. If the denominator is much larger (i.e. many individuals having been infected and recovered from the virus) the CFR is much lower. Here the German virologists estimate a CFR of 0.37%, compared to the ‘official’ CFR of about 2% for Germany.

As a point of reference, the CFR for the United States is about 3.6%.

Many studies of this type are going on right now in the US; Stanford University recently announced serology testing (first among healthcare workers and this week more widely to the general population) and is ongoing in New York City although not to the wider population yet.

Accurate, and widespread serological testing is something to keep on the lookout for, as it will indicate who can be ‘certified’ to be virus-resistant. Antibody testing will be a vital tool in getting past this pandemic, along with a vaccine, and getting back to something or a normal life.

Update April 12, 2020:

Thanks to a good friend, I discovered a publication (Okba et al on the pre-print server now up on the CDC website located here, with the details on the antigen used for their Eliza test with useful clinical-sample data. Inexplicably the CDC journal Emerging Infectious Diseases does not have this article as a tidy PDF.

About Dale Yuzuki

A sales and marketing professional in the life sciences research-tools area, Dale currently is employed by Olink as their Americas Field Marketing Director. For additional biographical information, please see my LinkedIn profile here: and also find me on Twitter @DaleYuzuki.

Leave a comment

Your email address will not be published. Required fields are marked *

3 thoughts on “Serological testing in Germany shows resistance to coronavirus SARS-CoV-2 at 14 percent

  • AMac78

    Prof. Hendrick Streeck actively researches in virology (96 citations in PubMed). His main focus is HIV, and he has also published on host response to infection and to vaccination. Maybe one of the other authors (Gunther Hartmann, Martin Exner, Matthias Schmid) is expert on coronaviruses, but in any case they seem to know their stuff.

    I wonder about the claim that their serology approach has “99% specificity”, which the abstract doesn’t elaborate on. This must refer to the test detecting antibodies to one or more SARS-CoV-2 viral proteins (Spike, Nucleocapsid, etc.), with <1% false positives due to cross-reactivity to antibodies against proteins of the endemic coronaviruses (which are common). Any other definition of specificity isn’t meaningful in the context of this study.

    John Bell at Oxford warned that many of the serological tests for SARS-CoV-2 that are entering the market have unacceptable performance. Presumably, they don’t stack up to the claims on their own labels. Which raises the question: how did Streeck validate the performance of the test he used?

    If Streeck’s results hold up, it will require a major re-think of SARS-CoV-2 virus epidemiology. Models of the pandemic will have to be stripped down and built back up, if they are to be useful for guiding public policy. Important stuff.

    • Dale Yuzuki Post author

      Indeed we’ll have to wait the publication (which also depends on how quickly the remaining 500 samples can be collected and processed).

      It would be interesting to see how long the antibody resistance lasts – it appears expectations are set by the virologists in the 6 to 18 months range.

      Interesting Oxford read, the tests coming out of China have been reported to be terrible and publicly so (i.e. hundreds of thousands of tests rejected at the country-level). No wonder Mt. Sinai and Stanford had to develop their own (highly likely these are standard ELISAs) with better mAbs.

      I wonder how convalescent plasma therapy measures their amount of neutralizing antibodies (obviously polyclonal). Here’s a review paper that I found yesterday (but haven’t read completely) with their recomendations across institutions (Hopkins, Mayo, Stanford, Columbia etc.)

  • AMac78

    An article posted yesterday adds some information on the specificity of the antibody test Streeck used.

    “Implausible figures” – criticism of the Heinsberg study, in the Bandera County Courier. (I’m not sure what this publication is.)

    If it turns out that cross-reactivity has actually occurred, the conclusions of Streeck’s team would be in question. Because then maybe not 15 percent of people fought in Heinsberg with the new virus, but part of it could simply have antibodies to the mostly harmless seasonal coronaviruses in the blood. As a result, significantly fewer people would be immune.

    Hendrik Streeck reached “Zeit Online” by phone in the evening. He said the test used could, of course, with a certain margin of error, distinguish between infection with Sars-CoV-2 and other corona viruses. The Lübeck-based company Euroimmun, which produced the antibody test that was used in the Heinsberg study, had previously done this on 1. 600 sera checked by blood donors, says Streeck.

    Assuming that maybe five to ten percent of the population had already been infected with other coronaviruses than Sars-CoV-2, the procedure would also be wrong have to find positive test results. But that is apparently only a very small part of the case, the manufacturer give the specificity with more than 99 percent.

    That would mean, in less than one percent of the cases the test used shows a false positive result. However, this statement cannot be verified without specific data from Euroimmun.

    As part of the development of their own antibody test, an international research group led by Christian Drosten also has prototypes of the antibody test from the company Euroimmun checked (MedRxiv: Okba et al., 2020). To this end, the scientists examined sera from people who could not have been infected with the current Sars-CoV-2, since their samples come from a time before the outbreak.

    And indeed, in some cases the tests reacted to antibodies from seasonal cold viruses from the Corona family. The researchers used the test again on twelve serum samples from the blood of two patients who apparently had such a seasonal infection. Again he reacted to the antibodies. And these are just not antibodies against Sars-CoV-2.

    This very small check contradicts Streeck’s statement, the cross-reactivity – that is, the test for antibodies seasonal corona viruses responded – hardly play a role in the Heinsberg study. No data is currently available on the extent to which the Euroimmun test in the study by the Bonn working group had been developed to such an extent that this cross-reactivity could be reduced or even ruled out.