Development of a Novel and Rapid Diagnostic Technique for Cardiac Amyloidosis Using Raman Spectroscopy

Introduction: Amyloidosis is a group of disease in which misfolded proteins aggregate as fibrous amyloid proteins with a β-sheet structure and deposit in various organs, causing progressive organ failure. Systemic amyloidosis is a rare and its diagnosis is often delayed because of non-specific clinical symptoms and its complicated diagnosis protocol. Despite the recent advances of therapeutic agents for ATTR and AL amyloidosis, cardiac amyloidosis remains life-threatening. Therefore, establishing an early and accurate diagnosis is a challenge for systemic amyloidosis patients, especially in cardiac amyloidosis. Raman spectroscopy is a label-free and non-invasive technique for estimating the molecular structure of a target, being adapted for biomedical analysis. Raman spectroscopy is an effective and rapid method for studying target proteins that are locally aggregated in a cell. Recent studies have also demonstrated that Raman spectroscopy is useful for detecting tumor, myocardial infarction, and intracellular lipids, suggesting that Raman spectroscopy could be powerful and rapid technique for analyzing biological samples and disease diagnosis. This study was undertaken to clarify whether the rapid diagnosis of amyloidosis and amyloid precursor protein type can be achieved by detection of Raman spectra specific for amyloid deposits in Raman spectroscopy.

Methods: Tissue samples were obtained from patients who were diagnosed with amyloidosis in Tokushima University Hospital. Biopsy samples were formalin fixed and paraffin embedded, and two or more serial sections with a thickness of 3 μm were prepared using a microtome. One of the serial sections was stained with Congo red for polarized light microscopy, and the adjacent sections were left unstained for Raman spectroscopy. Raman spectra of normal and amyloid-deposited areas were measured using a laser confocal Raman microscope. This study was approved by the Institutional Review Board of the Tokushima University (approval number 3507-3).

Results: Thirty-one biopsy specimens from 15 amyloidosis patients (9 AL, 5 ATTR, and 1 AA) with a median age of 70 years old (range: 49-82) were examined. The specimens were taken from heart (10; 5 AL and 5 ATTR), stomach (3; 2 AL and 1 AA), duodenum (4), rectum (4), kidney (4) and skin (6). We found a specific peak of Raman shift at 1680 cm^-1 for the amyloid-deposited area in all specimens, which represents the β-sheet structure. We next explored the patterns of Raman spectra specific for AL and ATTR amyloid deposits and underwent principal component (PC) analysis on the spectra in 1680 cm^-1. In the analysis of receiver operating characteristics (ROC) curves for the prediction of amyloid-deposited area on Raman spectra in patients with AL cardiac amyloidosis, ROC curves based on PC3 that appeared to include some spectral features, revealed the high accuracy rate with an AUC value of 0.922, sensitivity of 91.1%, and specificity of 96.2%. ROC curves for the prediction of amyloid-deposited area on Raman spectra in patients with ATTR cardiac amyloidosis, which based on PC1 showed the high accuracy rate with an AUC value of 0.735, sensitivity of 60.0%, and specificity of 100%. Interestingly, evaluating the PC analysis of the Raman spectra of amyloid-deposited area from AL and ATTR cardiac amyloidosis samples, ROC curves based on PC8, which showed significantly higher accuracy rate with an AUC value of 0.867, sensitivity of 75.0%, and specificity of 88.5% to distinguish AL from ATTR amyloidosis in cardiac biopsy specimens.

Discussion and conclusion: These results collectively suggest that Raman spectroscopy may contribute to a novel and rapid differential diagnosis in patients with amyloidosis, especially in AL and ATTR cardiac amyloidosis. Further study is warranted to analyze the detailed molecular structure of amyloid by Raman spectroscopy.