Methods

Study centres and patient population

A total of 15 referral centres for ATTRv, expert in the diagnosis and management of the disease, participated in this retrospective study. Five were located in Northern Italy (Milan, Padova, Pavia, Trieste, Udine), eight in Central Italy (Ancona, Bologna, Florence, Pisa, Rome Sant’Andrea Hospital, Rome Gemelli Hospital, and Rome Umberto I Hospital), and two in Southern Italy (Naples and Messina). Diagnosis of ATTRv and staging were performed according to the evolving clinical practice and specific guidelines1,12–14 (see Supplementary data online, Table S1). From 2004 to 2024, families receiving a diagnosis of ATTRv were offered genetic counselling and presymptomatic testing in at-risk relatives according to each centre’s local practice.

The study was approved by the local Ethics Committees, and all participants gave written informed consent for their clinical data to be used for research purposes in accordance with the Declaration of Helsinki.

Baseline clinical evaluation in index cases

Data on baseline demographic characteristics, clinical evaluation, and instrumental tests were retrospectively collected. Information regarding carriers’ conversion to overt phenotype and date of referral to therapy were also recorded.

Cascade genetic screening, monitoring of mutation carriers, and definition of disease onset

Starting from the diagnosis of ATTRv in index cases, all subjects identified during family screening carrying a TTR pathogenic variant were included. Genetic testing was first offered to first-degree relatives (siblings and offspring) and then extended also to second-degree relatives. TTRv detection was performed on extracted DNA according to centre practice. Cascade genetic screening has always been routinely proposed across all participating centres. According to each local centre practice, ATTRv mutation carriers were then evaluated by a multidisciplinary team to identify or exclude signs or symptoms of overt ATTR disease. Subjects with clinical, instrumental, or biopsy-proven signs of ATTRv disease at the time of genetic screening (baseline) were identified as genotype-positive/phenotype-positive (G+/P+), whereas those with a negative multidisciplinary baseline evaluation were considered asymptomatic carriers (namely genotype-positive/phenotype-negative; G+/P−). Patients were followed per centre practice with a multidisciplinary evaluation every 6, 12, or 24 months according to PADO and centre practice. In particular, index patients and G+/P+ individuals diagnosed at baseline were reevaluated every 6 months, as per standard practice for symptomatic patients. G+/P− patients were monitored according to their proximity to PADO based on the proband family member (index) age at disease onset: those within 10 years of PADO were generally seen annually, while those more than 10 years from PADO were followed every 2 years.1 In addition, PADO was also determined following the time from the average age at diagnosis of the index and the G+/P+ relative according to the family pedigree.15

Phenotype definition

For mutation carriers that developed an overt ATTRv phenotype (defined as the first clinical diagnosis of ATTRv—neuropathy and/or cardiomyopathy—meeting consensus criteria at follow-up), age at conversion, approximate date of phenotype onset and the diagnostic tests indicating conversion were recorded. Incidence was determined based on the time passed from genetic testing to phenotype definition. Standard assessment of patients with positive genetic testing is summarized in Supplementary data online, Table S1 and was equal for index patients and carriers. All individuals underwent comprehensive cardiovascular and neurologic screening. Signs and symptoms suggesting potential disease included: dyspnoea, peripheral swelling, presence of atrial or ventricular brady- or tachy-arrhythmias, presence or onset of ventricular hypertrophy or other common cardiovascular red-flags,1 autonomic dysfunction (e.g. erectile dysfunction, orthostatic hypotension, syncope), gastrointestinal symptoms, unexplained weight loss, bilateral carpal tunnel syndrome, lumbar canal stenosis, renal impairment, ocular involvement, and/or family history of polyneuropathy. The polyneuropathy disability score (PND) consists of five stages: (i) sensory disturbances with preserved walking capacity, (ii) impaired walking capacity with no need for walking aids, (iiia) need for a stick or a crutch for walking, (iiib) two sticks or crutches required for walking, and (iv) wheelchair or bed bound.1,16

Cardiac magnetic resonance was performed according to current guidelines and standard clinical practice.1 Criteria for diagnosis of ATTRv amyloidosis were as follows: (i) bone scintigraphy (99mTc-DPD, 99mTc-PYP or 99mTc-HMDP) showing cardiac uptake grade II–III according to Perugini score in the absence of a monoclonal gammopathy, (ii) Perugini grade I cardiac uptake with a confirmatory cardiac or extracardiac biopsy, (iii) biopsy-proven cardiac TTR amyloidosis performed for patients with cardiac imaging suggestive of cardiomyopathy without scintigraphy, and (iv) symptoms compatible with ATTRv neuropathy with at least one confirmatory instrumental test and/or a positive (skin) biopsy.14

All diagnostic assessments were interpreted locally at each participating centre.

Electrocardiographic abnormalities included rhythm disturbances (atrial fibrillation, conduction blocks (first-degree, second-degree atrioventricular blocks), QRS interval prolongation beyond 120 ms and pseudo-necrosis pattern.

Study objectives

The primary objectives of the present analysis were (i) to evaluate the prevalence of G+/P+ and carrier individuals among those relatives who performed genetic testing, (ii) to determine variant-specific phenotypic conversion rates and identify tests allowing early detection of conversion, and (iii) to assess the prognostic impact of cascade genetic screening exploring long-term outcome across index cases and patients diagnosed through genetic screening.

Statistical analysis

Continuous variables are expressed as median and interquartile (IQR) and were compared with nonparametric tests, while categorical variables are expressed as counts and percentages and were compared with χ2 or Fisher’s exact test, if the predicted count was <5.

Patients were divided into groups according to cascade family screening as follows: index (defined as the first individual within a family diagnosed with ATTRv, whose diagnosis initiates cascade genetic screening to identify other at-risk family members), G+/P+ (defined as individuals who carry a pathogenic TTR mutation and exhibit clinical signs or symptoms of ATTRv), and G+/P− (i.e. carriers; defined as individuals who carry a pathogenic TTR mutation but do not yet show clinical manifestations of the disease at the time of screening).

For each carrier developing overt disease, date of the first diagnostic test was recorded and was used to determine rates of conversion.

Survival analysis was performed with the Kaplan–Meier method. For cumulative incidence of all-cause mortality, rates were expressed as events per 100 patient-years, where ‘patient-year’ represents the sum of individual follow-up times expressed in years. For index, G+/P+, and carriers without conversion, follow-up started on the day that the genetic test was performed. For carriers manifesting phenotype conversion, the date of follow-up started with the date of conversion by study protocol. In addition to determining incidence rate, a Cox regression analysis was performed to determine the potential factors associated with conversion to over phenotype. Sensitivity analyses were performed by censoring patients who were receiving disease-modifying therapies or enrolled in clinical trials at the start date of these interventions. A time-dependent Cox regression analysis was performed to determine factors associated with all-cause mortality among patients with an overt clinical phenotype (variable selection: backward selection with P < .10 as threshold). Disease-modifying therapy was analysed as a time-dependent covariate (stsplit), with exposure switching at the date of treatment initiation; potential time-varying effects were modelled using the tvc() option in Stata. Models did not include variables that would be only collected in the presence of specific phenotypes (i.e. New York Heart Association [NYHA] class or PND scores). Follow-up ended in March 2024. Statistical analysis was performed with IBM SPSS 29.0 (Armonk, NY, USA) and STATA (v19.0).