Abdominal fat pad biopsies exhibit good diagnostic accuracy in patients with suspected transthyretin amyloidosis

Abstract

Background: The diagnostic accuracy of histopathological detection of transthyretin amyloid (ATTR) by Congo red staining of abdominal fat samples has been questioned since low sensitivity has been reported, especially for patients with ATTR cardiomyopathy. However, the outcome of surgically obtained fat pad biopsies has not yet been evaluated. The aim was to evaluate the diagnostic accuracy of skin punch biopsies from abdominal fat in patients with suspected ATTR amyloidosis.

Material and methods: Data were evaluated from patients who had undergone abdominal fat pad biopsies using a skin punch due to suspected amyloidosis from 2006 to 2015. The biopsies had been analysed using Congo red staining to determine the presence of amyloid, and immunohistochemistry or Western blot to determine the type of amyloidosis. The final diagnosis was based on the clinical picture, biopsy results and DNA sequencing. Minimum follow-up after the initial biopsy was 3 years.

Results: Two hundred seventy-four patients (61% males) were identified, and in 132 (48%), a final diagnosis of amyloidosis had been settled. The majority (93%) had been diagnosed with hereditary transthyretin (ATTRv) amyloidosis, and therefore subsequent analyses were focused on these patients. Overall, our data showed a test specificity of 99% and a sensitivity of 91%. Ninety-eight (94%) of the patients had neuropathic symptoms at diagnosis, whereas 57 (55%) had signs of amyloid cardiomyopathy. Subgroup analyses showed that patients with merely neuropathic symptoms displayed the highest test sensitivity of 91%, whereas patients with pure cardiomyopathy displayed the lowest sensitivity of 83%. However, no significant differences in sensitivity were found between patients with or without cardiomyopathy or between the sexes.

Conclusions: Abdominal fat pad biopsies exhibit good diagnostic accuracy in patients with suspect ATTRv amyloidosis, including patients presenting with cardiomyopathy. In addition, the method enables typing not only of the precursor protein but also of the amyloid fibril type, which is related to the phenotype and to the outcome of the disease.

Keywords: Adipose tissue; Amyloid polyneuropathy; Amyloidosis; Biopsy; Cardiomyopathy; Techniques and procedures; Transthyretin; diagnostic; hereditary; restrictive.

Fig. 1

Abdominal fat pad biopsy procedure. How we perform an abdominal fat pad biopsy using an 8 mm skin punch. a Local anaesthesia with premixed lidocaine (10 mg/ml) with adrenaline (5 μg/ml) after disinfection with chlorhexidine (5 mg/ml). b Skin punch after applying a sterile drape and using sterile gloves. c Biopsy sample. d Cutting out additional fat using a pair of forceps and a pair of scissors. e Suture with thread size 4- 0, two simple interrupted sutures and one central vertical mattress suture is usually adequate. f Final biopsy sample, including skin punch and additional fat, in saline (sodium chloride 9 mg/ml)

Fig. 2

Test sensitivity in ATTRv amyloidosis patients. Sensitivity of abdominal fat pad biopsies in different subgroups of patients with a final diagnosis of hereditary transthyretin (ATTRv) amyloidosis within three years from initial biopsy. a Test sensitivity in subgroups as per symptom/findings and sex. Each patient can present with more than one symptom or complication, and most patients presented with a mixed phenotype. No significant differences were found between the subgroups. b Test sensitivity in patients with a pure neuropathic or pure cardiac phenotype. No significant differences were found between the groups. CM: cardiomyopathy; eye: eye complications; GI: gastrointestinal symptoms; PN: polyneuropathy

Fig. 3

Histopathology of adipose tissue with amyloid deposits. Adipose tissue biopsy from a patient with type A transthyretin amyloid fibrils. The spotty appearance of sharply demarcated deposits is characteristic. Such deposits are sometimes small and quite few, and a systematic review is necessary. Congo red staining, normal light. Bar = 200 µm