Liver Prognostic Model Combining Elastometry and Biomarkers
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current prognostic methods for liver-related events and mortality in patients with hepatic disorders are inadequate, particularly for those with lesser fibrosis stages, as existing scores like MELD and Child-Pugh are limited to cirrhosis prognosis, and diagnostic tests are not designed for accurate prediction of survival or liver-related deaths.
Innovation Solution
A non-invasive prognostic method combining multivariate mathematical combinations of blood biomarkers, clinical markers, and elastometry data, using variables such as age, sex, alpha-2 macroglobulin, hyaluronic acid, platelets, and liver stiffness measurement, to generate a prognostic score through binary logistic regression and Cox models for assessing the risk of death and liver-related events.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If diagnostic tests (FibroMeter, FIB-4) are used to assess liver fibrosis severity, then diagnostic accuracy is improved, but prognostic accuracy for predicting death and liver-related events remains insufficient
Solution Approach 1:
The patent combines multiple diagnostic tests (FibroMeter, FIB-4, APRI) with elastometry data (liver stiffness measurement) and clinical variables (age, bilirubin, creatinine, INR) into a unified prognostic model. This merging of previously separate diagnostic tools creates a comprehensive prognostic system that accurately predicts both fibrosis severity and patient outcomes including death and liver-related events
Solution Approach 2:
The prognostic model serves multiple functions: it diagnoses liver fibrosis stage, predicts short-term mortality risk, predicts liver-related event risk, and guides clinical decision-making for patients across all fibrosis stages (F0-F4). This multi-functional approach replaces the need for separate diagnostic and prognostic assessments
2Reliability
If MELD and Child-Pugh scores are used for prognosis assessment, then cirrhosis prognosis is improved, but applicability to patients with lesser fibrosis stages (F0-F3) is lost
Solution Approach 1:
The patent develops a universal prognostic model based on the multivariate combination formula that applies to patients across the entire spectrum of liver fibrosis from F0 to F4 stages. Unlike MELD and Child-Pugh which are cirrhosis-specific, this model provides accurate prognosis for all fibrosis stages by incorporating elastometry data and clinical variables that reflect disease severity at any stage
Solution Approach 2:
The model uses parameter changes in elastometry measurements (liver stiffness) and clinical biomarkers to dynamically assess prognosis across different fibrosis stages. As fibrosis progresses from F0 to F4, the model captures parameter changes in liver stiffness, bilirubin, creatinine, and INR to provide stage-appropriate prognostic estimation
3Ease of manufacture
If diagnostic tests are applied to prognosis prediction, then existing resources are utilized, but prognostic accuracy is compromised
Solution Approach 1:
The patent merges existing diagnostic test results (FibroMeter, FIB-4, APRI) with elastometry data and routine clinical laboratory values (bilirubin, creatinine, INR, age) to create a prognostic model. By combining these readily available resources in a specific multivariate formula, the model achieves high prognostic accuracy without requiring new or expensive diagnostic tools
Data Source
Figure 1A~1B
Figure 1C~1D
Figure 2A~2B
AI summary
The present invention relates to in vitro prognostic method for assessing the risk of death or of liver-related event in a subject, comprising: a. obtaining at least one, preferably at least 2, of the following variables from the subject: i. biomarkers measured in a sample from the subject; ii. clinical data; iii. binary markers; iv. blood test results; b. optionally obtaining at least one blood test result by univariate combination, preferably with a binary logistic regression, of the at least one variable obtained in step (a), said blood test not being a Fibrotest, c. obtaining at least one physical data from medical imaging or clinical measurement, preferably from elastometry, more preferably from Vibration Controlled Transient Elastography (VCTE), and d. mathematically combining in a multivariate time-dependent model said at least one variable obtained in step (a) and/or said at least one blood test result obtained in step b); and said at least one physical data, obtained in step (c) thereby obtaining a prognostic score.