Disease choice - A study on how to pick an iGEM project

miRADAR and the application of miRNA technology in diagnostics

During the ideation phase of miRADAR, we were excited to discover that molecular switches could be applied to diagnostics. Detecting combinations of microRNA (miRNA) biomarkers could be an effective strategy for the diagnosis of complex diseases, particularly in cases where physical symptoms alone are inconclusive.^1–3 Additionally, by using logic gates we could detect many miRNAs at the same time, allowing for a wide range of testing. This piqued our scientific interest, and our team immediately started developing the underlying technology to generate a diagnostics test based on miRNA combinations. This is how miRADAR was born.

The problem:
Originally, we aimed to design a modular test platform to diagnose any disease with measurable changes of miRNA biomarker levels. Therefore, our test could potentially detect a panel of diseases at once, which we thought would help the greatest number of patients. Alternatively, we could design a specialised test for a single disease and focus on optimising it for that disease.

Given the two possibilities, we were left with a question. Which approach would be most beneficial for the community?

Who did we talk to and why?
We talked to Dr. Pablo Villoslada, a professor with expertise in neurodegenerative disease, and diagnostics. Besides his academic accomplishments, he is a practicing neurologist at Hospital de Mar in Barcelona, Spain. We wanted to get his opinion on how our diagnostic test could be used in the medical field for multiple diseases, and how it would fit into the current diagnostics field. Dr. Villoslada told us in an interview:

What we learned:
Dr. Pablo Villoslada informed us that our team should focus on a proof-of-concept test that is specific for a single disease, before applying this technology to a panel of disease types. Testing for multiple diseases would require us to pass several regulatory hurdles for each disease per medical regulations on diagnostics.

Integration:
As miRADAR we decided to choose one disease as a proof-of-concept, with the opportunity to expand into the detection of any disease with miRNA biomarkers. This allowed us to focus on making the test as applicable as possible for a single disease.

An evidence-based choice

After determining that the test would be a proof-of-concept, based on one disease, we were left with a choice: What disease should we focus on to have the greatest impact on diagnostics as a whole? To have an impact, it is essential to have data available to base our test on. We found that most miRNA data is available for cancers, followed by autoimmune diseases and neurodegenerative conditions.^1–3

We talked to a variety of medical professionals which included internal medicine, rheumatology, ophthalmology, and neurology. From them we learned that between these different fields, neurodegenerative autoimmune diseases were one of the hardest to diagnose. Compared to cancers the lack of biomarkers, the inability to perform biopsies and the uncertainty of MRI imaging makes the diagnosis of neurodegenerative diseases a more common hurdle in moving towards treatment.

The problem:
There are a few hundred neurodegenerative diseases, but how do we pick one?

Who did we talk to and why?
To find out which neurodegenerative disease has a pressing need for innovation in diagnosis, we went back to Dr. Pablo Villoslada:

What we learned:
We learned about diseases that still lack a proper diagnostic procedure such as rheumatological arthritis, neurodegenerative encephalopathy, multiple sclerosis, Alzheimer’s disease and Parkinson’s disease. We delved into methodologies for diagnosing these diseases and gained a better understanding of the timeline between initial patient symptoms to diagnosis and treatment.
In the end we found that diagnostic developments for many diseases run into one of three problems: they either have no readily detectable biomarkers, the disease is extremely rare, or there is no effective treatment available. These problems make developing a new diagnostic test extremely challenging.
However, there was one disease that stood out from the rest, Multiple sclerosis (MS). MS is a relatively common disease, with almost 3 million patients worldwide. There is no definitive biomarker and there are effective treatment options available that can reduce symptoms and improve the prognosis. This means that a diagnosis could really benefit a patient.

Integration:
Based on our new knowledge, our team felt motivated to focus on MS diagnostics. We searched for miRNA data for MS and discussed with our supervisors whether there was enough data available to make our test effective. Since relapsing remitting MS is the most common form of MS,⁶ it has a relative abundance of miRNA data. Combined with the fact that RRMS also has the most effective treatment, we focussed on this form of MS in our miRNA search models. If you want to know more about how we implemented the available miRNA data in our models, then please visit our dry lab page.

Final selection

Integration:
Based on the input from our stakeholders, we decided to use our diagnostic test to develop the diagnosis of MS.

• MS has a large impact on patients’ lives: MS diagnosis can happen relatively early in life (20’s-30’s) and will affect a patient’s physical, mental and social health for life.⁴

• MS is common, and numbers of MS patients are increasing worldwide: there are currently 2,9 million MS patients in the world.⁴

• MS diagnosis is difficult: even with all the research done in the past few decades, diagnosis remains challenging.

Additionally, as we were narrowing down the choice on which disease to focus on in our project, we encountered friends, acquaintances, and colleagues of the members in our team with MS. MS is truly more common than we initially realised, and we were inspired to use synthetic biology to make the diagnosis procedure of MS more patient-friendly.

We keep them at the heart of the research we do for miRADAR. If you want to learn more about MS, please visit our Awareness page.

Development

Current state of diagnostics

From Dr. Pablo Villoslada we understood that the process of diagnosing MS is very difficult. Before designing a novel diagnostic device, we needed to understand current diagnostic procedure. From literature we found the golden standard for MS diagnosis relies on the McDonald criteria.⁵ These criteria are based on MRI imaging of the central nervous system. The key requirement for a diagnosis of MS is evidence of damage to the central nervous system that is disseminated in time and space. This means that damage has occurred at different moments and to different parts of the central nervous system. This is typical of the most common form of MS, Relapsing-Remitting MS, where attacks occur in relatively short flares that leave lasting damage. These flares, as they are called, distinguish MS from other neurological conditions, as others either occur once or continuously.⁵ For more information on the mechanism of MS disease progression please visit our Description page.

The problem:
After learning about the diagnostic timeline using McDonald criteria, we had to find a specific application of our new test: How can miRADAR improve the diagnostic procedure of MS?

Who did we talk to and why?
We interviewed several neurologists such as Christa Benit M.D. and Joey Kuijpers M.D. By obtaining input from this panel of stakeholders we were able to form a complete picture of the diagnostic procedure. By conducting interviews, we got the following information about the current diagnosis techniques from Dr. Kuijpers and Dr. Benit, respectively:

Both neurologists underlined the difficulties in diagnosing patients with MS. When the McDonalds criteria are not satisfied based on the MRI, neurologist must opt for a very invasive lumbar puncture. This entails the insertion of a needle into the spine and taking a small sample of the cerebrospinal fluid (CSF). CSF contains antibodies that indicate repeated inflammation in the central nervous system, which is correlated with MS. The procedure is highly invasive and painful for patients, which is why some refuse. Unfortunately, this procedure is not always conclusive. From our stakeholders we learned that for patients who cannot be diagnosed with MRI or with a lumbar puncture, diagnosis can take years.

Integration:
We aim to support neurologists in diagnosing challenging cases of MS, by designing a test based on criteria independent of the current procedures: MRI and lumbar puncture. The miRADAR test should be as minimally invasive as possible, which improves patient care and presents an alternative for patients who refuse a lumbar puncture.

Differential diagnosis

As was mentioned in the interview with Dr Joey Kuijpers, neurologists often struggle with differentiating MS from similar diseases, in medical terms referred to as differential diagnosis. Since excluding similar diseases is a core part of the diagnosis of MS, we were wondering whether miRNA biomarkers could contribute in distinguishing between MS and similar diseases. Thus, we wanted to know whether our test would be capable of providing extra information on differential diagnoses.

Who did we talk to and why?
In the interview with Dr. Pablo Villoslada, we asked him an open question if he believed that miRNA alone would be sufficient to create a diagnosis of MS:

Dr. Villoslada really got us thinking about the option of differential diagnosis of MS. He went in depth on how results of a paper between MS groups and patients seems impressive, but almost all of the miRNA analysed are non-specific autoimmune miRNAs. Those miRNAs are therefore not sufficient to perform a differential diagnosis between MS and other similar diseases.

Based on the concept of changing the focus of our test, we were interested in finding out more about how differential diagnosis influences the treatment. That is why we spoke to neurologist Dr. Joey Kuijpers:

Dr. Pablo put us on the path towards differential diagnosis. By informing us about the difficulties with current techniques, and with miRNA, we learned that we have to put each miRNA that we find with our computational models, in biological context, and see if it makes sense that they are actually specific with MS. We have to find a group of miRNAs that is specific for autoimmune inflammation of the central nervous system, and miRNAs that differentiate MS from similar diseases. This could allow specific MS diagnosis based on miRNAs.

Dr. Joey Kuijpers' explanation supported the approach of distinguishing between MS and similar diseases. He mentioned several diseases that are similar to MS and explained how their treatment differs. Some diseases such as neuromyelitis require immediate and heavy intervention to prevent further damage. This requires heavy immune cell inhibitors, which are excessive for MS. Other diseases such as encephalitis also require immediate medicine, but they require antibiotics. Giving the wrong diagnosis, and medication can lead to devastating results. That is why neurologists must be certain about a diagnosis, otherwise the results can be severe.

From Dr. Christa Benit we learned that differential diagnosis can be very difficult for 10-30% of patients. When the lesions on the spine are not conclusively linked to MS, no specific diagnosis can be reached.

Integration:
We acknowledged the importance of selecting an MS-specific combination of miRNAs to distinguish between MS patients, healthy controls, and similar diseases. Please visit our dry-lab page to find out how we found our unique set of biomarkers. Since a miRNA-based test has not yet been proven to be reliable for a differential diagnosis, and misdiagnosis can have severe effects, we felt the best application for our proof-of-concept test would be to use it alongside the current McDonald criteria. We now knew that for about 10-30% of patients, the McDonald criteria are insufficient to diagnose patients. These are complex cases of MS, for which diagnosis can take months to years. The miRADAR miRNA test would give additional information about a patient’s condition, which would allow neurologists to be confident in giving patients the medication they need, reducing the risk of incorrect medication and preventing progressive damage due to lack of treatment. By complementing the current diagnostic procedure, the miRADAR test can simplify diagnosis of complex cases of MS.

MS screening

Since the onset of MS symptoms is particularly prevalent among young adults (20-30), we thought it might be a good idea to use the test to screen young people for the presence of MS before they develop any symptoms.⁵ That way even more damage can be prevented, and perhaps treatment might be more effective. We wanted to understand if doctors would appreciate such a wide-scale screening step, or if our test would be preferred as a specific tool at the doctor’s office.

The problem:
Should miRADAR design a test to screen the general public for MS, or should it be a specific test used at the hospitals?

Who did we talk to and why?
We spoke with Dr. Wim de Kieviet, an expert in clinical chemistry and laboratory medicine. Dr. de Kieviet has experience with the requirements needed for general screening. In an interview he informed us about his doubts towards this goal:

Since miRNA levels can vary slightly from patient to patient, developing a test with few false positives would be very difficult. Even if the test would be 99,9% accurate, this would result in almost as many false positives as true positives.⁶ This meant we couldn’t do the wide screening as we were originally planning. We had to find another target where the test could be useful.

We considered to opt for risk groups, fractions of the population that due to genetic variation or other measurable factors could have a higher risk of developing MS. To learn more about risk group screening, we consulted the neurologists Dr. Pablo Villoslada and Dr. Christa Benit again:

We asked Dr. Pablo Villoslada if it was possible to target a risk group:

Dr. Pablo Villoslada informed us that the risk group is too big of a target and would cause unnecessary harm due to the lack of treatment options. Dr. Christa Benit went on to explain that even though it might be possible to diagnose patients earlier it might not be very useful for patients today. The current treatments for MS mostly relies on preventing flares. Preventing these flares with heavy immune system supressing medication, years before they are likely to appear, might not be the best application.⁵

Integration:
Based on the stakeholders’ advice, we decided to make our test more specific. This would be applied when a patient comes into the doctor’s office with symptoms correlated with MS. This allowed us to also focus our test platform to be tailored towards laboratory analysis, rather than a point-of-care test (a simpler ‘covid style’ test) that can be performed at home. In the end, this also proved to be a much more realistic target.

Implementation

Where and when to test?

As the project progressed, we realised that where this test would be taken is an important factor. From our previous interviews we knew that this test was not appropriate for screening, however there are many steps in the process of diagnosis. Should this test be used immediately when patients come to the general practitioner (GP) with complaints, at neurologists when they are unable to confirm the diagnosis with the McDonald criteria or can it only be done in extremely specialised hospitals for a second opinion? This is necessary because we need to know what components should be included, and what equipment could be expected to be available at the testing location.

Who did we talk to and why?
We spoke with both doctors and patients, so we could understand the question from both perspectives. Starting with Marianne Vogel, a spokesperson of the Dutch MS association (MS vereniging). And asked her, where a test would be best added to the current diagnostic process:

Standard laboratory compliance

We wanted to ensure that the test would be accessible. Many patients worldwide have a delayed diagnosis due to the unavailability of MRI scans and neurologists. We wanted to design a test that did not rely as much on personnel or equipment so the cost wouldn’t be a major problem. By having a simple test, both the neurologist and radiologist are spared a lot of time analysing the ambiguous MRI images. Instead, our test has to be used by an pathologist. We designed it so that only RNA purification and amplification is necessary for the pathologist to use the test, minimising the required labour. We wanted to confirm whether these isolation and amplification procedures would be possible using standard hospital laboratory equipment.

Who did we talk to and why?
We explained our test to Dr. Wim de Kieviet, clinical laboratory specialist. He gave us information about which tools are generally available in a hospital laboratory, and what our test should be capable of:

He was enthusiastic about our project and stated that it had potential. He mentioned some specific criteria about what hospital laboratories are capable of in terms of equipment, personnel and how commercial test kits are used. Also, we see that there is room for improvement in the RNA analysis techniques, as currently this is delegated to larger hospitals.

Integration:
We designed all of our cell-free toehold switches to work at room temperature and have a long shelf life. To see how we did this, please visit our Design page.

Regulation

Once we had a clear image in our mind of what our test should look like, and how it is applied, we were wondering: how does a test go from a lab test to a real clinical test? What criteria and laws does it have to abide by to be approved in a real clinical setting?

Who did we talk to and why?
This we asked to laboratory expert Dr. Wim de Kieviet:

Dr. de Kieviet stated that a diagnostic test such as ours must undergo clinical testing, which takes a lot of time and money. This reduces the flexibility of a test, and the rate of innovation. This was confirmed when we had a talk with a diagnostic start-up from Wageningen. They confirmed that the IVDR is the main European regulation for in vitro diagnostic tools. After covid, and due to budget cuts, the waiting lists for approval of tests have been growing, which increases the cost and time to bring a test to market. Since miRNA data has only just been coming up in the past few years, the risks must be taken into account: how long should you wait for the body of miRNA data to expand, to ensure you have a good test, but not so long that someone else beats you to a patent? And how ethical is it to wait, considering patients are waiting for diagnosis?

Integration:

To ensure that the test is simple to use, all test components should be supplied in a ready-to-use kit. Therefore, in our test, the necessary materials and reagents for the isothermal amplification of miRNA will be provided, along with the correct positive controls. The logic circuit and threshold module are already embedded in the paper during assembly. Please visit our implementation page for more information on the details of our test kit design.

Ethics

The problem:
After we had summarised all the information from our stakeholders, we felt we were well underway towards generating a test that could be used in the real world. That is when the discussion arose: even if we manage to design and produce a ‘perfect’ test for MS with high accuracy, how do we implement it in an ethical way? While discussing it with the team, we found several disturbing ways a test might cause unwanted adverse results.

• What if a test is biased, because it was based on biased data (small patient samples that do not reflect the general population)?

• What if the test becomes very expensive due to a market monopoly?

• What if the test is used to replace lumbar punctures but proves to be much less accurate?

• And would you want to know that you would get MS in the future, if there is very little you can do to stop it?

Who did we talked to and why?
To discuss these questions and more, we shared them with a panel of students and our ethics coordinator Dr. Zoey Robaey, from which we received the following interesting perspectives:

We were talking about the lack of a risk group, which eliminated the option of screening, when one of the students said something unexpected:

The core response we got that even with a near perfect test, the way it is used can change everything. And if the test should ever come to the market, all these criteria must be accounted for. We do not know yet which ethical dilemma will be most critical, therefore all ethical concerns must be analysed in depth to create an honest and patient-oriented test.

miRADAR

Through an integrated human practices approach, we have shaped our project to better meet the needs of the MS community, medical professionals, and researchers. By engaging with neurologists, patients, and MS researchers, we learned that early diagnosis is crucial for improving patient outcomes, while also ensuring that any new diagnostic tool must be highly specific and reliable to complement existing methods. Patient interviews highlighted the emotional and physical toll of delayed diagnosis, reinforcing our goal to develop a minimally invasive, quicker, and more accessible blood-based test for MS. The feedback from researchers helped us prioritise biomarker accuracy and the importance of differentiating MS from other autoimmune diseases, a critical factor in ensuring the practicality of our test.

Future

Research
We recommend further data collection on miRNA patterns associated with MS and MS mimics, particularly in diverse patient populations and across different stages of the disease. Expanding the miRNA dataset would improve the sensitivity and specificity of our diagnostic test, helping to better differentiate between MS and other autoimmune conditions. This would validate the clinical utility of these biomarkers in early detection and disease monitoring, ensuring the test provides neurologists with a reliable tool to support timely diagnosis and treatment of MS.

miRADAR in the real world
miRADAR's test is designed from the very beginning for implementation in the real world– from doctors to researchers to patients. In our project we have been focussing on diagnosing complex cases of MS. For this we want the diagnosis to be accessible. However, for many MS patients in the world, the hurdle to diagnosis is not due to inconclusive McDonald criteria, but a lack of available resources such as neurologists and MRIs. Unfortunately, to do this our test would have to be reliable when used without complementation from MRI. It is still unknown whether our test is capable of this. Secondly, even when diagnosis is made more accessible, the medication and treatment for MS should also follow for it to have a meaningful impact. As miRADAR, we want to contribute to health for all, but we should also be transparent and realistic in what we can achieve. We hope that by progressing the technology, it can be developed in the future into a standalone test that can be accessible for regions with limited funds. Want to see exactly how we envision putting miRADAR's test into action? Visit implementation.