Transforming patient care for rare diseases: LynxCare's success in ATTR-CM detection

Real-World Evidence (RWE) studies have the potential to revolutionize the way we approach healthcare for rare diseases such as Amyloid Transthyretin Cardiomyopathy (ATTR-CM).  

Learn how LynxCare’s collaboration with a leading Belgian hospital resulted in an advanced understanding of this condition and an almost 50% relative increase in genetic testing for this disease compared to manual checking of medical records.

Introduction: The challenge of ATTR-CM diagnosis and treatment

Hereditary transthyretin-mediated (hATTR) amyloidosis is a rare and progressive disease caused by the abnormal buildup of amyloid deposits in various tissues and organs throughout the body, leading to organ dysfunction and eventual failure. A subtype of hATTR amyloidosis, ATTR cardiac amyloidosis (ATTR-CM), is characterized by the accumulation of amyloid deposits in the heart, leading to restrictive cardiomyopathy and heart failure (HF) [1].  

The prevalence of ATTR-CM is estimated to be around 9% in HF patients with preserved ejection fraction (HFpEF) [2]. Unfortunately, its general prevalence in the total population is not known due to underdiagnosis and lack of disease awareness.  

ATTR-CM is a particularly severe condition that progresses rapidly and often leads to death within a few years since diagnosis.

However, with timely detection and treatment, the disease progression can be slowed down. The problem with recognizing ATTR-CM in patients lies in its unspecific symptoms and multisystemic manifestations. Due to this, as many as 57% ATTR-CM patients are misdiagnosed and the median delay in diagnosis is 3 years [3]. This means that new methods of early ATTR-CM detection are necessary to transform the lives of patients with this condition.

Goals: Exploring the potential of AI for detecting and managing rare diseases

The aim of this study was to explore the potential of LynxCare’s Natural Language Processing (NLP) model in extracting free text and use of real-world structured data to obtain RWE from electronic health records (EHRs) aiding in the diagnosis and management of ATTR-CM. We collaborated with a leading Belgian hospital that governed our access to anonymized patient records.  

We aimed to develop an algorithm that could accurately identify patients with ATTR-CM and provide insights into the clinical characteristics, treatments, and outcomes of these patients.

By harnessing the power of AI and NLP, we aimed to improve the understanding of this rare disease, accelerate the diagnosis and treatment of affected patients, and ultimately improve their outcomes.

Methods: AI and NLP application in identifying clinical features of aTTR-CM from EHRs

You can find our methodology in the full blog post, which you can download as PDF.

Results: LynxCare AI streamlines the diagnostic process for rare diseases

You can find our results in the full blog post, which you can download as PDF.

Conclusions: Transforming rare diseases patient care through data insights

This research highlights the unique potential of AI and NLP in extracting valuable insights from large-scale, unstructured health data, such as Electronic Health Records, leading to improved patient outcomes.  

The use of NLP and machine-learning algorithms can significantly improve the efficiency of disease screening and diagnosis, especially in the case of rare diseases like ATTR-CM, regularly overlooked with traditional testing.

Further research in this area can lead to the development of similar algorithms for other rare diseases, ultimately leading to improved healthcare outcomes for patients worldwide.