Development of a Non-Invasive Biomarker for Bleeding Risk Stratification of Gastroesophageal Varices

The aim of this application is the development of a non-invasive method to predict the presence of treatable but potentially lethal gastroesophageal varices in patients with end-stage liver disease (cirrhosis).In 2017, cirrhosis was the 7th most common cause of premature deaths in Germany. One of the most critical complications of cirrhosis is the development of gastroesophageal varices that can cause fatal internal bleeding. To identify and treat at-risk patients, guidelines recommend regular esophagogastroduodenoscopy and variceal band ligation which can reduce mortality by 50-70%. However, endoscopy is invasive and often unnecessary when no treatment is required. For these reasons, the American Association for the Study of Liver Diseases has identified the development of “non-invasive markers that predict the presence of high-risk varices” as a major unmet need. Such markers would facilitate the identification of high-risk patients, while reducing unnecessary invasive procedures.The primary biological mechanism for the development of gastroesophageal varices is elevated portal pressure and reversal of flow in the left gastric vein. Detection and quantification of this reversed flow is the primary innovation and fundamental biological premise by which this application proposes to detect gastroesophageal varices at risk for bleeding.This application proposes the necessary technical development and preclinical validation of new quantitative 4D Flow MRI-based biomarkers. 4D Flow MRI is a non-invasive MRI technique that measures flow velocities in all 3 dimensions and over time. It can quantify blood flow in complex anatomy, providing comprehensive anatomic and hemodynamic characterization. Compelling preliminary data shows that 4D Flow MRI can detect high-risk gastroesophageal varices by quantifying reversed flow in the left gastric vein.This application’s technical innovations and preclinical validation aims to pave the way for clinical implementation of 4D Flow MRI-based detection of high-risk varices. Essential aspects are a fast sequence with high sensitivity for detection of varices at risk for bleeding. Therefore, today’s lengthy acquisitions of 12 min will be reduced to 3 min by exploiting the low pulsatility of portal venous flow and using novel time averaging acceleration strategies. Difficulty in visualizing slow flow in the left gastric vein will be addressed with velocity-optimized strategies including deep learning algorithms. Rigorous pre-clinical validation of the proposed MRI method will include patient specific 3D printed phantom evaluation and in vivo measurements in healthy volunteers and patients with low and high-risk varices.If successful, the proposed velocity-optimized, accelerated 4D Flow MRI strategy will provide accurate, precise and reproducible non-invasive risk stratification for variceal bleeding. It has the potential to fundamentally improve clinical management and long-term survival of cirrhotic patients.