Lyme disease: why tests are unreliable

ELISA. The release of the National Protocol for Diagnosis and Care (PNDS) of Lyme disease, renamed "Recommendations of good practices", far from appeasing the spirits, only increases discontent. At the center of the discord are the famous serological screening tests known as Elisa and Western Blot. For some, they are considered perfectly reliable, for others, it would simply remove them as they are outdated. As a reminder, in the PNDS, these tests are always prescribed in first intention to detect the disease. " It is also recalled in Annex 3 of the text, that they are almost 100% reliable in certain situations where they must be used. However, we consider that these data are largely erroneous " explains Hugues Gascan *, immunologist, director of research at the CNRS, having participated in the drafting of the PNDS.

These serological tests are done in two stages: first the Elisa which, if it is positive, must be confirmed by the Western Blot. If the latter is also positive, then the patient is considered to have Lyme disease. The controversy is due to the fact that many patients have negative tests while they complain of symptoms that could enter the clinical picture of Lyme disease. Hence the suspicions about their reliability

But the doubts are not confined to the testimony of the patients. A meta-badysis conducted at Imperial College London, published in 2016 in the journal International Journal of General Medicine, reveals that the sensitivity of these tests does not exceed on average 60%. Thus, 40% of sick people would not be detected.

Beyond this study, it is the very methodology of screening that could be problematic in the case of Lyme disease. And in particular, the one about the Elisa test. The faults are threefold:

1 / Unclear calibrations

To understand the purpose of the debate, we must return to the principle of this examination.

The idea of ​​the Elisa test is to check whether the patient is or has been infected with a Borrelia bacterium responsible for Lyme disease. This verification is indirect. That is to say that we will not seek to detect the bacterium in the patient, but rather the antibodies produced by its immune system and directed specifically against the antigens of the bacterium. Antigens are those pathogen proteins specifically recognized by the immune system. This antibody-antigen reaction is one of the foundations of immunity. Antibodies, made by B cells (cells of the immune system), bind naturally with antigens. It is this reaction that the Elisa test exploits

These tests are marketed by many manufacturers. They are in the form of a plate with a number of micro-wells (often 96), at the bottom of which are fixed Borrelia antigens. The medical badysis laboratory that performs the test will then fill these wells with the patient's serum. The serum is in fact the blood freed from its cellular elements (red blood cells, white blood cells, platelets) and enzymes promoting its coagulation. It has a lot of chemical compounds like sugars, lipids and of course proteins. Among these are the antibodies produced by the immune system against Borrelia, but also against all other kinds of pathogens to which the person has been exposed during his lifetime.

Once the serum is placed in the wells only antibodies against Borrelia will bind to the antigens at the bottom of the well. Thus, all the other molecules and in particular the antibodies directed against other pathogens can be eliminated by a simple washing. Only the antibodies specific for Borrelia that we seek to detect are left in the wells.

The quantity of these antibodies must now be measured to determine whether the patient is well infected with the bacterium . This step is based on the measurement of the optical density, that is to say on the intensity of the color of the solution which is in each well. For this, the laboratory operator adds a product that changes color if the antibody is attached to the antigen. Thus, in each well, the more antibodies attached to the antigens of the bacteria, the more intense the color and the higher the optical density will be.

The optical density is therefore a reflection of the immune response of the person against the bacteria. But to affirm whether or not the person is "HIV-positive", the Elisa test must be calibrated in order to determine from what optical density the patient is considered to be HIV positive.

It is in the definition of this threshold that the problems begin. To achieve this calibration, we must start from a serum of a sick person, so that we already know positive. This is provided by each test manufacturer so that it can be compared to the measured optical densities for each patient. The plates thus have a series of wells reserved for the pure standard serum (the one that is known to be positive) which is diluted several times in order to obtain a range of less and less intense colors, with the corresponding optical densities. The optical density of the sample of each patient is then compared to the optical densities of this standard range.

Finally, it is necessary to establish the threshold optical density from which the patient is considered to be ill. " This threshold was defined in the 2000s statistically by the European Concerted Action on Lyme Borreliosis (EUCALB), a European consortium for Lyme disease research, created in the 1990s. Another surprising element this threshold must be defined in the region in which the test is performed. However, we know that there is a strong geographical variation. For example, there are many more cases of Lyme in Alsace and Limousin than in the PACA region. This calibration is not standard. It varies from one region to another and from one manufacturer to another "explains Hugues Gascan. In other words, the thresholds are not the same throughout France. The same person can then be positive in Montpellier and negative in Paris.

" A recent meta-badysis by Michael Cook and Basant Puri shows that serological tests for Lyme disease generate 500 times more false negatives that tests for AIDS Imagine an HIV test as weak as those used for Lyme. This would be a scandal ", comments Hugues Gascan.

To do it right, it would be necessary to make a calibration with weight values ​​that is to say by accurately dosing the amount of antibody present in the patient's serum. . The result would then be indicated by a quantitative value (for example in microgram per milliliter) as is the case for most measurements made during blood tests. " This standard calibration would make it possible to obtain comparable results whatever the region in which the test is practiced, and whatever the manufacturer ", badures the researcher.

2 / The diversity of the bacteria detected in tests

Every year or so a new Borrelia appears among the agents involved in Lyme disease. For example, in 2016 in the United States, the Centers for Disease Control and Prevention and the Mayo Clinic in Rochester discovered a new bacterium called Borrelia mayonii and involved in the disease. However, these new bacteria are not detected by the tests currently available on the market.

3 / A serology that varies over time

As we have seen, the detection of Lyme disease is based on the detection of antibodies against the bacteria. However, this antibody concentration can vary greatly over time in the same person. Thus, depending on when the test is performed, the test can be positive or negative. A mouse study published in 2015 in PLOS and directed by Nicole Baumgarth of the University of California, Davis, shows that the bacterium attacks B cells. These cells of the immune system are those that just produce the antibodies that we tries to detect in Elisa tests. This action of the bacterium on lymphocytes could partly explain the low level of antibodies in the blood of sick people and therefore a negative test. These results are further supported by a genomic study conducted by Jerome Bouquet of the California University of San Francisco that shows in humans a bad response of lymphocytes to Borrelia bacteria. Thus, given the modulatory effects of Borrelia on the immune response, is it appropriate to use an Elisa test, based precisely on the badysis of this immune response, to detect Lyme disease?

* Hugues Gascan CNRS research director, former director of Inserm Unit and the PADAM platform (Automated Production of Monoclonal Antibodies) – Rennes, Angers. As a member of the French Federation against Tick-borne Diseases, he participated in the development of the PNDS.