Sixteen human autopsy brains of patients with ATTRv, aged 27–69 years, were studied. The cases were identified from the archive of Unidade de Neuropatologia of Centro Hospitalar Universitário do Porto (CHUP) (case #1 to #12) and from the Portuguese Brain Bank (PBB) (cases #13 to #16). Seven cases (#1 to #3 and #6 to #9) were previously described [21]. All cases were followed in the Neurology outpatient clinic of CHUP, and clinical data were retrospectively assessed. Genetic testing in cases #1 to #3 was not performed, because at that time, no molecular diagnosis was available. Appropriate consent procedures for the collection and use of human brain tissues were obtained and the study was approved by the Ethics Committee of CHUP.

Tissue was fixed in a 4% aqueous solution of formaldehyde. Paraffin sections (4 µm) were obtained from the following anatomical regions: frontal cortex, entorhinal region, cerebellum, thalamus, basal ganglia, midbrain, pons, medulla, and upper cervical cord. Cases #1, #5, #8, and #11 did not have cervical cord section. When available, particularly from the PBB cases, additional regions were included: temporal, parietal and occipital cortices, cingulate cortex, amygdala region, and olfactory bulb.

All cases were studied for the presence of neurodegenerative disorders and vascular pathology. Cortical cerebral microinfarcts were defined as microscopic regions of cellular death or tissue necrosis up to 5 mm in dimension and lacunar infarcts as cavitating infarcts measuring up to 15 mm [14]. Territorial infarcts and hemorrhages were also recorded. Microhemorrhages were distinguished from perivascular haemosiderin leakage by the accumulation of haemosiderin in the brain parenchyma [28].

Six μm-thick paraffin-embedded tissue sections from the representative blocks were cut on a microtome. Haematoxylin & Eosin, Perls’ Prussian blue, and Congo red staining were performed using standard protocols. For the Perls’ protocol sections were incubated with a 1:1 mixture of 5% hydrochloric acid and 5% potassium ferrocyanide (30 min), and counterstained with eosin (1 min). For the Congo red, protocol sections were place in Congo Red solution (8 min), followed by differentiation into 0.2% alcoholic Potassium Hydroxide (quickly), 0.2% aqueous Potassium Hydroxide (quickly), washed in distilled water, and counterstained with hematoxylin (less than 1 min).

Immunohistochemistry (IHC) staining was performed using the Ventana OptiView DAB IHC detection kit and the Ventana BenchMark Ultra processor (Ventana, Tucson, AZ, USA). Paraffin tissue sections after dewax were pre-treated with heat using Ultra Cell Conditioning Solution (CC1 or CC2; Ventana) and the endogenous peroxidase was inactivated before the incubation with the primary antibodies: α-synuclein (Novocastra, KM51, 1/80), β-amyloid (Dako, 6F/3D, 1/120, Tau (Thermoscientific, AT8, 1:1200), TDP-43 (Proteintech, ref: 10,782–2-AP, 1/2000), or TTR (Dako, ref: A 0002, 1/1500). The slides immunostained with α-synuclein and β-amyloid antibodies suffered an extra pre-treatment with formic acid for 2 and 1 min, respectively, before the heat pre-treatment.

Morphological analysis of TTR CNS involvement (leptomeninges, vessels, and parenchyma) was based on anti-TTR immunostained sections. The presence and degree of amyloid deposition was analyzed semi-quantitatively by a neuropathologist (RT) blinded to patient´s clinical features. Amyloid vessel deposition was classified as negative if no TTR positive vessels were found; mild if scattered positivity was found in few vessels; moderate if scattered positivity in many vessels or strong positivity in few vessels was found; and severe if strong positivity in many vessels was found [4] (Fig. 1). In the leptomeninges and brain parenchyma, the amyloid deposition was classified as negative, if no amyloid deposition was found, and as mild, moderate, and severe according to the amount of immunoreactivity present [21] (Fig. 1).

Representative examples of TTR vascular (upper row) and leptomeningeal/subpial deposition (lower row) according to the scale description. TTR immunohistochemistry. Scale bar: 200 µm

Age at death, sex, mutation, age at disease onset, disease duration, cause of death, and clinical data (TTR-FAP disease stage, intracerebral hemorrhage, Ischemic stroke, cognitive complaints, cranial nerve dysfunction, and liver transplant) were collected for the study. Data are presented in percentages for qualitative variables, mean, and range for quantitative variables.

Normality of the variables age at death, disease duration, and amyloid burden (total number of affected areas and sum of the amyloid scores) was tested using Shapiro–Wilk normality test. Pairwise comparison between groups, based on parametric or non-parametric tests, was performed using one-way ANOVA and Kruskal–Wallis, depending upon the assumptions required. Post hoc testing was performed to identify the groups between which significant differences existed (Dunn–Bonferroni). Pearson´s correlation coefficient test was used to assess correlation between disease duration and number of brain regions with TTR deposition or sum of TTR scores.

The publicly available tool Morpheus (https://software.broadinstitute.org/morpheus) was used to perform unbiased hierarchical clustering. Due to missing data, cervical cord was excluded for this analysis.

Rating data were entered into an Excel spreadsheet and analyzed using Statistical Package for Social Sciences (SPSS) software (version 26.0), Armonk, NY. A P value of < 0.05 was considered statistically significant.