Poster Presentation 29th Annual Lorne Proteomics Symposium 2024

Optimisation of laser capture microdissection with tandem mass spectrometry for the diagnosis of amyloid subtype (#134)

Dorothy Loo Oey 1 , James Rowland 2 , Janelle Hancock 1 , Kylie Cuthbertson 2 , Peter Mollee 2
  1. Proteomics Core Facility, Translational Research Institute, Woolloongabba, QLD, Australia
  2. Pathology Department, Princess Alexandra Hospital, Woolloongabba, QLD, Australia

Background: Correct diagnosis of amyloidosis subtype is a critical step to direct patient management, inform prognosis and guide targeted genetic testing. Laser capture microdissection and tandem mass spectrometry (LMD-MS) analysis of formalin-fixed paraffin-embedded (FFPE) biopsy samples is emerging as the new gold-standard diagnostic technique in amyloid subtyping. A novel LMD-MS assay has been developed and implemented at Pathology Queensland, Australia, as a research tool. To allow transition to the clinical diagnostic laboratory, optimisation and validation of this assay was required.

Method: LMD-MS was performed on 78 patient samples and results evaluated assessing key MS measurements and the identification of amyloid signature and amyloid forming proteins. A bioinformatic reporting algorithm was developed and applied to a validation set of 226 samples. 121 samples were assessed by both IHC and LMD-MS for comparative assessment. 11 samples were assayed by this LMD-MS assay with results compared to those obtained at international laboratories.

Results: Critical MS datapoints contributing to confident amyloid forming protein identification were protein mean intensity, number of spectral matches and relative abundance. The addition of the Kabat protein library to the Swiss-Prot/Uniprot database did not interfere with identification of the amyloid proteins and showed superior diagnostic yield. Technical changes including buffer formulation and MS analyser technology variation had no impact on amyloid forming protein identification. A bioinformatic reporting algorithm was created which confidently identified an amyloid subtype in 96.1% of patient samples. LMD-MS was shown to be superior to IHC in subtyping of amyloid proteins with a non-diagnostic rate of only 3.6% The local LMD-MS assay showed 100% concordance with international laboratory analysis.

Conclusions: A novel LMD-MS assay has been developed and validated and shows superiority to the previous IHC based standard of care assessment. Role-out into the clinical diagnostic laboratory in Australia is planned with some opportunities for further development.