It has been recently estimated that between 3.5 and 5.9% of the general population suffers from a rare disease even if this percentage is probably underestimated as many people with a rare disease are often not properly diagnosed. Moreover, national definitions of ‘rare’ are not uniform thus leading to group patients under different categories [1].
Approximately 6000 rare diseases are known, 72% of which have an identified genetic origin, most of the time (70%) with pediatric onset. In the last decade rare diseases have become an emerging global public health priority and a lot has been done to raise awareness about the topic, improve the diagnostic process, the access to treatments and last but not least, improve the support to the patients’ families.
So where does the riddle lie?
One of the main challenges for rare diseases is related to diagnosis. Rare diseases are often misdiagnosed because they are characterized by a broad diversity of symptoms, often common, that vary from disease to disease and from patient to patient suffering from the same disease. This leads to the fact that there are still millions of rare disease patients without a proper diagnosis. The so-called diagnostic odyssey for families of children with undiagnosed genetic diseases can be uncertain and unpredictable and often last several years, eight on average.
The advent of cheap and fast genome sequencing, together with the growing availability of bioinformatics tools to process huge amounts of data (genomics will soon generate more data than applications such as social media, earth sciences, and astronomy [2]) has given the medical community an unprecedented opportunity to tackle the challenge of providing the right diagnosis for rare disease patients in a reasonable time.
Despite many rare diseases lacking proper treatments, there are plenty of patients for which a fast diagnosis through sequencing can guide the therapeutic choice thus resulting in immediate life’s quality improvements. Shortening the path to diagnosis has been demonstrated to be both valuable from a healthcare and an economical perspective [3] and the fact that the majority of rare diseases patients are children points to new-born screening strategies. It is therefore extremely important to support initiatives such as the Rare Disease Day to promote new efforts in this direction and increase public understanding.
But what could accelerate and enable the success of a program that aims to shorten the path to rare diseases diagnosis?
Besides innovative new-born screening programs based on genome sequencing, a robust and sustainable strategy in a global landscape would require to properly assess existing rare disease resources as registers, databases platforms and networks. To potentially positively impact the lives of millions of rare disease patients, this would require an active partnership from a range of contributors across the public and private sectors. Such an ambitious effort could then pave the way of federating most of the available resources in metadata repositories that can advance our knowledge about rare disease thanks to the use of machine learning or other advanced digital tools. AI algorithms can for example identify early-onset rare diseases in electronic health records and can help undiagnosed patients for which genome sequencing was not sufficient or not done at all or patients who frequently change the site of medical assistance for which the problem of an undefined picture often arises. Furthermore, the exploration of AI would lead to further options for implementing symptom control into viable solutions for healthcare professionals and patients.
As part of the community involved in the fight for rare diseases we should strongly support the digital transformation in the health sector, change the paradigm in the diagnosis of rare diseases and improve the diagnostic tools. Ideally this would require interconnecting different data sources and making them interoperable and reusable, to develop and exploit algorithms compliant with guidelines for reliable artificial intelligence and validation policies and finally to distribute resulting algorithms to hospitals to run on existing systems.
Despite the recent rise in rare diseases research, most rare diseases remain understudied and therefore under treated and the lack of concrete medical options for patients with rare diseases could weaken the interest in diagnosis / screening initiatives. To stop feeding this conundrum let’s fully embrace the mission of the #RareDiseaseDay and speak out about the importance of properly addressing the challenges of the rare disease community.
