PRACTICAL CHALLENGES IN AMYLOID SUBTYPING IN BRAZIL: A COMPARATIVE STUDY BETWEEN MASS SPECTROMETRY AND A CLINICAL-LABORATORY MODEL

Abstract

Background: Systemic amyloidosis is a protein misfolding deposition disease often under-recognized in middle-income countries. The precise identification of the precursor protein is a key factor in the diagnosis challenging clinical practice, especially outside of reference centers. Objectives: To subtype amyloid by mass spectrometry (MS), the gold standard method for diagnosis, and to compare it with the clinical-laboratory model (CLM) for amyloidosis diagnosis when the gold standard is not available. Methods: This is a retrospective, observational, unicentric study. Patients with a histologic diagnosis of systemic amyloidosis from 2009 to 2018 at the Hospital das Clínicas of USP, a Brazilian public university hospital, were included. Patients were identified through database searches and an active search of biopsies positive for amyloid in the pathology laboratory archives. Stored tissue samples were selected by availability to perform MS. All medical records were retrospectively reviewed. Amyloid specimens were analyzed by MS coupled with liquid chromatography following laser microdissection, thermochemical processing, and trypsin digestion. Amyloid subtypes were defined by the CLM using combined data on clinical presentation, monoclonal gammopathy, pyrophosphate scintigraphy, immunohistochemistry, immunofluorescence, and genetic mutations. When exact criteria couldn't be met to define the amyloid subtype, clinical judgment was used based on available data. Results: From the 138 patients with biopsy-proven systemic amyloidosis, 57 biopsies were excluded (17 external diagnoses, 13 unavailable samples, 8 insufficient samples for laser microdissection, 10 insufficient for MS, 9 had no remaining amyloid deposit in the stored block). MS was performed on 81 biopsies from 13 different organs: 30% kidney, 23% heart, 14% subcutaneous fat, 7% gastrointestinal tract, 6% nerve, 5% respiratory tract, 4% bone marrow, 3% muscle, and 1% each lymph node, gingiva, liver, salivary gland and tongue. The following amyloid subtypes were detected: AL (58%, n = 36), ATTR (23%, n = 14), AA (8%, n = 5), AFib (3%, n = 2) and AH (2%, n = 1). In 2 cases amyloid co-depositary proteins were not identified, and another 2 were inconclusive (1 weak detection of fibrinogen with a dispersed pattern and 1 detection of both fibrinogen and transthyretin). Results are still pending for 19 cases. The CLM correctly subtyped amyloid in 100% of AA and AFib subtypes, 94% of AL, and 79% of ATTR cases. In 2 AL and 1 ATTR cases, the CLM was not able to determine the subtype, and patients received only supportive treatment. ATTR was misdiagnosed as AL in 3 patients, who were treated with anti-plasma cell agents. The patient with AH was diagnosed as AL by the CLM. Discussion: Although the CLM showed good performance in subtyping amyloid in our cohort, it failed in 3 patients, resulting in a misdiagnosis that led to inappropriate treatment with chemotherapeutic agents. Additionally, 3 other patients did not receive specific treatment due to the models limitations in identifying the amyloid subtype. Conclusion: Our study emphasizes the role of mass spectrometry as the gold standard method for accurately subtyping amyloid and the importance of establishing referral centers for amyloidosis. This approach can prevent potential harm from misdiagnosis and improve patient care.