Accurate MS-Based Diagnostic Amyloid Typing Using Endogenously Normalized Protein Intensities in Formalin-Fixed...

Area of Science:

• Proteomics
• Biochemistry
• Medical Diagnostics

Background:

• Amyloidoses involve misfolded protein deposition, requiring precise identification of the fibril-forming protein for prognosis and treatment.
• Immunohistochemistry (IHC) is common for amyloid typing but can be subjective and yield inconclusive results.
• Mass spectrometry (MS) is preferred by some centers, often using spectral counts for quantification.

Purpose of the Study:

• To introduce and validate an alternative relative quantification method for proteomic amyloid typing.
• To enhance the accuracy and reliability of amyloidosis diagnosis using mass spectrometry.
• To resolve cases with inconclusive IHC staining and identify rare amyloid subtypes.

Main Methods:

• Analyzed 62 formalin-fixed, paraffin-embedded (FFPE) tissue samples using liquid chromatography-tandem mass spectrometry (LC-MS/MS).
• Employed internal normalization of iBAQ values of amyloid-related proteins relative to serum amyloid P component (APCS).
• Utilized de novo sequencing for samples lacking distinct fibril-forming proteins and developed an XGBoost machine learning model.

Main Results:

• The APCS normalization method showed robust performance across different LC-MS/MS platforms.
• Achieved complete concordance with clear-cut IHC results and resolved cases with inconclusive staining.
• Identified immunoglobulin light chain components in rare AL-amyloidosis subtypes and achieved 94% accuracy with the XGBoost model.

Conclusions:

• The iBAQ APCS normalization method, extended by de novo sequencing, provides robust and accurate diagnostic amyloid typing.
• This proteomic approach offers a reliable alternative to IHC, especially for complex or inconclusive cases.
• AI-based classification complements the proteomic method, though clinical context remains essential for interpretation.