Karla Plevová: For better diagnosis of leukemias and more effective treatment

"The LYNX tool is unique in that it combines both laboratory and analytical parts, and it can refine the diagnosis of haemato-oncology patients and ultimately influence their treatment," says Karla Plevová, who works at the Institute of Medical Genetics and Genomics and the Internal Haematology and Oncology Clinic at the Faculty of Medicine of Masaryk University and University Hospital Brno, and at the Centre for Molecular Medicine at CEITEC. Together with Veronika Navrkalová from Šarka Pospíšilová's Research Group, Tomáš Reigl, Jakub Porec and other colleagues, they were awarded for it in the second year of the MUNI Innovation Award, given for projects that have been successfully implemented in practice and have thus contributed to increasing the social relevance of research at Masaryk University.

What led you to work on the project in the first place? And how long did it take you to get from the initial idea to launching a usable tool?
We have been working on leukemias, specifically chronic lymphocytic leukemia, for years. And since we have a very good collaboration with doctors from the Internal Hematology and Oncology Clinic of the University Hospital Brno, we knew it was desirable to design a diagnostic test that would be able to analyze as many molecular markers as possible and look for mutations in genes in a single process. We were therefore led to the project by the interest of physicians. This was based on our knowledge of gene mutations specific to certain leukaemias and lymphomas. It has been at least five years since my colleagues and I first started putting together a list of these genes.

So if a doctor wants to find out the features of a patient's leukaemia, he no longer has to run several tests, but just one, yours?
Yes, previously we would test each of the genes or molecular markers separately, or in smaller sets of, say, three genes. In our package, we test sixty-seven genes, and in addition to that we are able to analyse the regions coding for antigen receptors, i.e. molecules specific for B and T lymphocytes, as well as looking at damage on chromosomes. This is quite unique because chromosomal damage, gene mutations and antigen receptor rearrangements are usually tested separately using procedures based on different principles.

What did you anticipate would be the biggest challenge in the project? And has this assumption been confirmed?
There were many challenges! We mainly wanted our panel - or package - to be applicable in clinical practice, so it was essential to validate it sufficiently. For this reason, we searched for samples in the biobank of the Centre for Molecular Biology and Genetics at the University Hospital Brno. They had to be well characterized by other diagnostic tests that could potentially replace ours. Moreover, although our initial focus was on chronic lymphocytic leukaemia, it soon became clear that our test could be suitable for the analysis of a wide range of leukaemias, lymphomas and other haemato-oncological diseases. So in the end we worked with a set of nearly ninety different blood, bone marrow or tissue samples.

The advantage of the panel is personalization and customization to the needs of specific users. What exactly does it consist of?
We test the same set of markers for each patient, because although leukaemias and lymphomas are different, their DNA damage is often similar. So we have chosen a set of genes whose mutations are present in multiple types of target diagnoses and also occur repeatedly in different patients with the same diagnosis. The personalization is that for each patient we are able to detect a set of prognostically significant lesions in cancer cells, which can be used to further monitor, for example, how they respond to treatment. An important part of it is then also determining the rearrangements of those antigen receptors, because these are completely unique to each patient, to each leukaemia, and are markers that are used to monitor, for example, residual disease.

And this also allows doctors to tailor treatment more effectively?
Yes, based on certain types of defects, doctors can adjust treatment because we can tell that a patient with a particular mutation in a particular gene, for example, will not respond to chemotherapy. The result of the test could also influence the inclusion of patients in different clinical trials.

Your diagnosis relies on data generated by next-generation sequencing, which generates a huge amount of it. How have you approached this problem?
This was also one of the challenges that required close collaboration with our bioinformaticians to overcome. Because the typical lab worker who pipettes and then evaluates genetic data doesn't include the analytical part. You get so much data from next-generation sequencing that you can't simply open an Excel spreadsheet and extract the information you're looking for. The data needs to be well processed using bioinformatics techniques and the analysis needs to be treated preferably in such a way that the average lab worker, who may not have a bioinformatics background, can run it. Therefore, we have created software that any laboratory worker can run through the results.

How do you envision LYNX integrating into existing processes and integrating with other diagnostic tools?
We are fortunate that we have already integrated the panel into diagnostics at the University Hospital Brno, where it is routinely used. They use it to analyse samples from patients with chronic lymphocytic leukaemia, acute lymphoblastic leukaemia, as well as various types of lymphomas and multiple myeloma. Although we didn't really count on it at the beginning, it turned out that it has potential for a number of other lymphoproliferations besides these leukemias.

Can LYNX be used throughout the course of the disease to assess how the patient responds to different therapies?
In theory, yes, but practice depends on insurance reimbursement, which brings us to yet another set of challenges and limitations. Because this is a next-generation sequencing-based diagnosis, which is still quite limited. So we are currently only testing some patients. Either at diagnosis or prior to treatment, or for example in cases where there is not enough sample to perform several different methods, or the diagnosis is not completely clear and we can help distinguish whether the disease will be mild or aggressive. So it's not that we use our panel en masse. We continue to use standard diagnostic tests, of course, which insurance companies are willing to reimburse.

What kind of feedback do you get from doctors?
We see that they are interested and that they like using the panel. Whether it becomes more widely used depends on other system processes. But we also see interest from other hospitals, not only in the Czech Republic. Our partners include, for example, the hospital in Hradec Králové, where they target slightly different diagnoses, so we have adapted the analysis for them. But we are also working with a hospital in Singapore, where they want to use the panel without changes as we originally designed it.

There is an appeal to timeliness for every diagnosis. If your analysis is this comprehensive, it can probably speed that up too, right?
(smiles) I'm probably going to disappoint you a little bit here. The advantage is that you get several questions answered in one test, which can really speed up the diagnostic process. On the other hand, if you only want to target one marker, it's faster to use a test designed for that particular marker, which is then ready in a couple of days. Our method is more time-consuming as it involves sample library preparation, actual sequencing, bioinformatics analysis and the interpretation part, which itself takes a lot of time as it requires double reads. But this is an issue with next-generation sequencing in general, because you do get a lot of data, but it needs to be interpreted well.

When you mentioned that insurance companies only reimburse for next-generation sequencing in small amounts, is one of the next methods to push for those reimbursements to increase so that your system can be used more often? After all, isn't it more cost effective?
Yes, you are right. We may not have calculated it in detail, but knowing the cost effectiveness of other methods, using our method would indeed be cheaper. The push to push for more frequent use of next-generation sequencing is there, but it's a systemic issue and it takes time for the system to respond to such "innovations". After all, next-generation sequencing has been around for more than a decade. Unfortunately, a lot of people still think it is expensive, but when you consider the capacity and what you can evaluate with it, it is cheaper than other methods.