Cardiovascular autonomic failure in hereditary transthyretin amyloidosis and TTR carriers is an early and progressive disease marker

Background. The cardiomyopathic and neuropathic phenotype of hereditary transthyretin amyloidosis are well recognized. Cardiovascular autonomic dysfunction is less systematically and objectively assessed. Methods. Autonomic and clinical features, quantitative cardiovascular autonomic function, and potential autonomic prognostic markers of disease progression were recorded in a cohort of individuals with hereditary transthyretin amyloidosis and in asymptomatic carriers of TTR variants at disease onset (T0) and at the time of the �rst quantitative autonomic assessment (T1). The severity of peripheral neuropathy and its progression was strati�ed with the polyneuropathy disability score. Results. A total of 124 individuals were included (111 with a con�rmed diagnosis of hereditary transthyretin amyloidosis, and 13 asymptomatic carriers of TTR variants). Symptoms of autonomic dysfunction were reported by 27% individuals at T0. Disease duration was 4.5 ± 4.0 years (mean ± SD) at autonomic testing (T1). Symptoms of autonomic dysfunction were reported by 78% individuals at T1. Cardiovascular autonomic failure was detected by functional testing in 75% individuals and in 64% of TTRcarriers. Progression rate from polyneuropathy disability stages I/II to III/IV seemed to be shorter for individuals with autonomic symptoms at onset (2.33 ± 0.56 versus 4.00 ± 0.69 years [mean ± SD]). Conclusions. Cardiovascular autonomic dysfunction occurs early and frequently in individuals with hereditary transthyretin amyloidosis within 4.5 years from disease onset. Cardiovascular autonomic failure can be subclinical in individuals and asymptomatic carriers, and only detected with autonomic function testing, which should be considered a potential biomarker for early diagnosis and disease progression.

License:   This work is licensed under a Creative Commons Attribution 4.0 International License.Read Full License Version of Record: A version of this preprint was published at Clinical Autonomic Research on May 20th, 2024.See the published version at https://doi.org/10.1007/s10286-024-01038-z.

Introduction
Hereditary or variant transthyretin amyloidosis (ATTRv amyloidosis) is a multi-systemic disorder characterised by the progressive deposition of extracellular brillary misfolded β-sheets amyloid aggregates in various organs and tissues. 1 Because it can manifest with heterogeneous, nonspeci c symptoms, signi cant diagnostic delay is common.ATTRv amyloidosis is an autosomal dominant inherited condition, with more than 130 point variants in TTR gene reported so far. 1 The Val30Met variant is the most common ATTRv amyloidosis worldwide, with endemic foci in Portugal, Sweden, and Japan.
In the United Kingdom, the majority of individuals presenting with a neuropathy, or with a combination of neuropathy and cardiomyopathy have the Thr60Ala variant, which originated in northwest Ireland, 2,3 while the majority of individuals presenting with cardiomyopathy have the Ile122Val variant.ATTRv amyloidosis should be considered in all individuals with a combination of somatic and autonomic neuropathy, proteinuria, and heart failure with preserved ejection fraction. 4,5 ptoms of autonomic dysfunction are known to be a common and early feature in ATTRv amyloidosis.Despite recent efforts to capture cardiovascular autonomic impairment with different techniques and from different perspectives, a detailed and systematic phenotyping of cardiovascular autonomic failure with standardized quantitative cardiovascular autonomic function testing (AFT) in large cohorts of individuals has yet to be undertaken.Similarly, objective functional testing to explore the presence of subclinical autonomic neuropathy in TTR carriers may in uence individual access to diseasemodifying treatment (DMT), which is currently dependent on the presence of a peripheral neuropathy in many countries, although the scenario is rapidly changing. 6,7 ndeed, autonomic manifestations can cause burdensome symptoms throughout the disease course.Their functional assessment is crucial in monitoring disease onset and progression, in helping decide the best time to initiate DMT, for optimal symptom management, and in monitoring the response to treatment. 8,9 this study, we have characterised autonomic impairment clinically and objectively with cardiovascular AFT in a well-de ned, large cohort with a con rmed diagnosis of ATTRv amyloidosis, and in a subgroup of asymptomatic TTR variant carriers.We also evaluated whether autonomic dysfunction could serve as a marker of disease progression.

Methods
A retrospective natural history study was conducted in a cohort of individuals with either a genetically con rmed diagnosis of ATTRv amyloidosis or an asymptomatic TTR variant carrier status, referred to the national tertiary referral autonomic unit (at the National Hospital for Neurology and Neurosurgery, University College London Hospitals NHS Foundation Trust and at St Mary's Hospital, Imperial College Healthcare NHS Trust, London, UK) between 1992 and 2022.
Demographics, age at onset, disease duration, genetic variant, cardiac, renal, neuropathic, and autonomic features were collected at disease onset (T0), and at the time of autonomic assessment (T1) by a consultant in autonomic neurology.
Additionally, nerve conduction studies (NCS) and the PND score 10 were collected in a subset of individuals.
Neuropathic onset was de ned by the presence of neuropathic pain, paraesthesia, or numbness in the lower and/or upper extremities.
Isolated carpal tunnel syndrome (CTS) was considered connective tissue onset.
Cardiac onset was de ned by the presence of symptoms of heart failure, sick sinus syndrome, and atrioventricular conduction abnormalities, as it is widely accepted that these should raise the suspicion of cardiac amyloidosis and trigger further investigations. 11As part of the diagnostic algorithm, all our patients underwent at least an echocardiogram and a radioactive technetium-99m ( 99m Tc) and 3,3-diphosphono-1,2-propanodicarboxylic acid (DPD) scan, mostly done within the National Amyloidosis Centre in London. 5nal onset was de ned by the presence of either a reduction in glomerular ltration rate (GFR), increased serum creatinine levels, or proteinuria.
Asymptomatic relatives who tested positive for pathogenic TTR variants, had no de nite ATTRv amyloidosis-related symptoms, normal neurological examination, and normal cardiological investigations were de ned as TTR variant carriers. 7 T1, autonomic function was assessed with a battery of quantitative AFTs evaluating adrenergic and cardiovagal function, as previously described. 12Sympathetic function was evaluated by the standing test, 60° passive head-up tilt table test (HUT) for up to 10 minutes, HR responses to the Valsalva manoeuvre (VM), sustained isometric exercise, mental arithmetic, and cold pressor.Parasympathetic function was assessed by evaluating respiratory sinus arrhythmia (RSA) to deep breathing and heart rate (HR) responses to the VM.Neurogenic orthostatic hypotension (nOH) was de ned by a drop of ≥ 20 mmHg in systolic blood pressure (SBP) or of ≥ 10 mmHg in diastolic blood pressure (DBP) within 3 mins of orthostasis or HUT test. 13lues were de ned abnormal according to age ranges, as previously described. 14e severity of autonomic impairment was classi ed according to the following staging system: Autonomic Stage 0 = no cardiovascular autonomic failure (cvAF): normal RSA, normal BP, and HR responses to VM and no evidence of nOH.
Autonomic stage 1 = isolated parasympathetic or mild mixed sympathetic and parasympathetic cvAF, individuals with impaired RSA or impaired RSA and abnormal BP response to VM, but without evidence of nOH.
Neuropathy phenotyping was done by clinical examination, neurophysiology and PND grading where available.Small bre neuropathy was considered with both abnormal sympathetic skin response (SSR) and thermal threshold (TT) on nerve conduction studies (NCS).The interval delay to progress from a PND score of I/II to a PND score of III/IV was calculated against the presence or absence of autonomic symptoms at disease onset.
Statistical analysis was performed using GraphPad Prism version 8. Data are presented as mean (± standard deviation [SD]).
Where appropriate, explorative analyses were performed using t-test, 1-and 2-way ANOVA, χ 2 test; survival analyses were performed with log rank tests.P values < 0,05 were considered statistically signi cant.
Ethics approval was obtained by national (London Harrow Research Ethics Committee) and institutional (University College London and Imperial College London) research authorities.The study was conducted according to the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) Statement. 15
Nerve conduction studies and AFT were analysed only for those 102 individuals who did not receive liver transplant or any disease-modifying treatment up to the time of the assessment.There was evidence of large bre neuropathy in 48 (47%), small bre neuropathy in 24 (23%), and CTS in 36 (35%) individuals.

Disease progression according autonomic symptoms
Strati cation into PND I/II and timing to further progression to stages III/IV was available for 23 individuals who did not receive any DMT from disease onset.The small size of this subgroup was motivated by the fact that the PND scoring was not systematically assessed in the earlier years of the study.These were divided into two subgroups, according to the presence or absence of autonomic symptoms at disease onset.Progression rates to reach PND stages III/IV were shorter for individuals with autonomic symptoms at onset, compared to those without autonomic onset (mean progression 2.3 ± 0.5 versus 4.0 ± 0.7 years respectively).There was a suggestion of more rapid progression of neuropathy in those with autonomic dysfunction, but this did not reach statistical signi cance in this analysis, likely due to small numbers in each group (Log Rank Mantel-Cox 2.75, p = 0.09) (Fig. 2).

Autonomic phenotype in less common TTR variants
15 individuals with TTR other than Thr60Ala and Val30Met presented with symptoms of autonomic dysfunction and had quantitative evidence of cardiovascular autonomic failure on AFT (Table 5).In two variants, Ile84Thr and Ser23Asn, we described subjective autonomic dysfunction and objective cardiovascular autonomic failure for the rst time.The carrier of the Ile84Thr variant, with 7 years of disease duration, reported orthostatic intolerance and syncope, alternating diarrhoea and constipation, and weight loss.The carrier of the Ser23Asn variant, who had an 8-year disease history, reported orthostatic intolerance and syncope, diarrhoea, anhidrosis, erectile dysfunction and nocturia.On AFT, they both showed evidence of stage 1 cardiovascular autonomic failure.The key ndings of this study can be summarized as follows: 1. Symptoms of autonomic dysfunction were reported by up to 78% of individuals with ATTRv amyloidosis in our cohort, where 40% carried a Thr60Ala variant, within a mean disease course of 4.5 years.Gastrointestinal autonomic dysfunction was the most prominent, both at disease onset and during diseases progression.2. Functional autonomic tests detected cardiovascular autonomic failure in 75% of individuals after an average disease course of 4.5 years.To this end, a thorough, standardized autonomic functional assessment is far more sensitive in detecting cardiovascular autonomic failure compared to a simple standing test (nOH was present on HUT in 33% individuals versus 20% on the standing test).Indeed, it identi ed autonomic failure in 51% of individuals who did not report autonomic symptoms on clinical questioning.
3. Stage 1 cardiovascular autonomic failure was present in 64% of TTR variant carriers as well, suggesting that cardiovascular autonomic quantitative tests are potential pre-clinical biomarkers in ATTRv amyloidosis.
4. Cardiovascular autonomic failure may have an impact on prognosis and thus survival of individuals with ATTRv amyloidosis.The provision of objective evidence of end-organ involvement is increasingly important now that disease modifying therapies are available for this progressive condition.
Several previous studies described autonomic dysfunction in the early stages of ATTRv amyloidosis, mainly for the Val30Met variant.However, these were performed in smaller cohorts, 16-20 de ned orthostatic hypotension not according to the current criteria, 21 were based on clinical symptoms and questionnaire outcomes only, 22 or did not use a standard battery of orthostatic challenges and cardiovascular re exes. 23few studies used cardiac MIBG scintigraphy as a measure of autonomic function.To our knowledge, there is no literature conclusively comparing the sensitivity of cardiac MIBG scintigraphy with functional cardiovascular autonomic testing, and the studies available so far have not been able to correlate imaging outcomes with clinical ndings or the results of autonomic function tests. 21,24 One study suggested that abnormal cardiac MIBG scintigraphy might re ect cardiomyopathy rather than cardiac autonomic neuropathy in ATTRv patients and carriers in case of heart failure.25 The use of cardiac MIBG scintigraphy should be encouraged, but it should not be used as a substitute for standard autonomic testing, which re ects not only the cardiac innervation, but is an accurate way to assess clinical symptoms such as orthostatic hypotension.
In our opinion as a long-running tertiary referral autonomic unit, the assessment of cardiovascular autonomic function cannot disregard a minimum battery of head-up tilting, cardiovascular re exes (Valsalva manoeuvre, pressor responses, respiratory responses) and catecholamine spillover.The adoption of standardised testing procedures and a shift towards data sharing between autonomic centres is desirable, given that such tests are highly specialised, and their results can be challenging to interpret, considering that there may be concurrent cardiac involvement due to amyloid cardiomyopathy, volume depletion due to diarrhoea, and the effects of vasoactive medications. 26In a recent review of 128 individuals with ATTRv cardiomyopathy, cardiac syncope was diagnosed in 47% individuals, while out of the 53% individuals without a clear cause for syncope, 12% of cases were attributed to nOH. 27In another review, nOH has been estimated to affect 40-60% of individuals with ATTRv amyloidosis. 28r study demonstrated that the autonomic nervous system is involved at an early stage in ATTRv individuals and at a pre-  23 It is interesting that hypohidrosis was reported by 17% of individuals at autonomic assessment, while it was not described at disease onset.Sudomotor dysfunction is often subclinical and can represent a possible early involvement in individuals with ATTRv amyloidosis, although its presence is di cult to ascertain retrospectively and without speci c objective testing (i.e. thermoregulatory sweat test, quantitative sudomotor axon re ex test, or dynamic sweat test). 29,30 At our autonomic unit, we have facilities for the thermoregulatory sweat test, the dynamic sweat test, and nerve conduction studies with small bre assessment.The sympathetic skin response (SSR) is a fairly simple, quick and innocuous test that has been proven useful in the diagnosis of small bre damage in ATTRv amyloidosis, 31 but its limited reproducibility hinders its use as a reliable technique.32 Electrochemical skin conductance with the Sudoscan has been gaining momentum as a quick way to assess sudomotor dysfunction.There are several reports of its use in ATTRv amyloidosis, however, these do not provide correlations with quantitative cardiovascular autonomic dysfunction.[33][34][35][36] In our view, the major limitation of electrochemical skin conductance is its low speci city for detecting neuropathy.Hence, it may perform poorly on its own, but it may be useful in combination with other exploratory techniques.37 In this retrospective cohort, we did not speci cally collect quantitative data on sudomotor function, but our ongoing prospective study includes formal sudomotor assessment.We suggest that sudomotor function testing could be implemented in the autonomic workup of individuals with ATTRv amyloidosis as a marker of small bre involvement, with potential to further improve sensitivity of assessment for TTR end-organ effect.38 Similary, current novel promising additional biomarkers include skin biopsy, 39,40 plasma neuro lament light chain, 5 and magnetic resonance neurography.41 It would be of interest to compare these early markers of disease onset to autonomic measures.These issues will be addressed in future publications.
Currently, transthyretin suppressor therapies (RNA interference [RNAi] agents and antisense oligonucleotides [ASO] are formally approved in the UK, US, and many European countries for the treatment of stage 1 and stage 2 neuropathy caused by ATTRv amyloidosis.These stages are based on the PND and/or Coutinho's systems, 10,42 which focus on the individual's ambulatory capacity with or without aids, and only partially depict motor and sensory qualities, ignoring small breassociated sensory and autonomic symptoms.However, over the last few years, several research groups have proposed consensus recommendations to recognise both onset and progression of neuropathy as early as possible, thus granting access to DMT. 23,43,44 If these recommendations were widely implemented in clinical practice, carriers of TTR variants may bene t by having reliable techniques to monitor their preclinical status with autonomic function testing, among other techniques, to capture objective markers of disease onset, enabling them to access the newer therapies sooner, and shifting the standard of care for ATTRv amyloidosis towards preventative medicine.
We have described the autonomic phenotype in 15 individuals with rarer TTR variants, who presented with symptoms of autonomic dysfunction and had quantitative evidence of autonomic failure on autonomic assessment.The clinical presentation of rarer variants in our cohort is in line with previous literature reports, except for two variants, where cardiovascular autonomic failure had not been described yet.Individuals with the Ile84Thr variant had only reports of gastrointestinal dysfunction so far, 45 while individuals with Ser23Asn variant have been associated with cardiomyopathy and only in one case with peripheral neuropathy. 46Both of our individuals presented with cardiovascular autonomic failure, which broadens the spectrum of these phenotypes and highlights the need for formal quantitative autonomic assessment in rarer variants.
Survival analysis of 23 individuals suggested a possibler trend towards faster disease progression in individuals with autonomic symptoms at disease onset, although this did not reach statistical signi cance, probably due to the small sample size.Median survival in ATTRv amyloidosis depends on various factors, such as the type of variant and the number of organs and systems involved.Survival in untreated individuals with predominantly neuropathic disease courses is estimated between 5-15 years from onset, and this time span is decreased to 2.5-4 years, if cardiac involvement is present. 47Similar to our cohort, the large transthyretin amyloidosis outcomes survey (THAOS) study showed faster disease progression in individuals with nOH. 48Our survival analysis de ned the time to reach PND stages III/IV as its endpoint, rather than death, and therefore it might not be accurately comparable to the numbers available in the literature.We focused on the prevalence of autonomic symptoms at onset, to determine whether this clinical variable could be a possible predictor of a shorter and more aggressive disease course.Indeed, in keeping with larger natural history studies, the presence of early autonomic involvement appeared to affect the progression of neuropathy in our cohort as well.
This work has some limitations, most importantly the retrospective nature of the study.Secondly, formal autonomic cardiovascular assessment will need to be performed on larger number of carriers to estimate the prevalence of subclinical autonomic involvement.Neuropathy staging and detailed assessment was not routinely available at the time of autonomic testing, and this may might have affected our survival analysis, likely due to small numbers in each group.

Conclusions
Autonomic dysfunction in ATTRv amyloidosis affects several domains of the autonomic nervous system in the early stages of the disease, and it progresses over time.Given the multi-systemic nature of the disease and the advancement in DMT, biomarkers are needed to aid early diagnosis, monitor disease progression, determine response to therapy, and to prognosticate outcomes.In the setting of ATTRv amyloidosis, the identi cation of autonomic dysfunction is essential to optimal symptom management and early identi cation of end-organ involvement, with potential to in uence earlier access to disease modifying therapies.

Con icts of Interest
Nothing to report

Figure 1 A
Figure 1

Table 2
Demographics, clinical and autonomic features at disease onset (T0) and at autonomic function testing (T1).Results are shown as number of individuals/total individuals (%).CTS = carpal tunnel syndrome.

Table 3
Cardiovascular autonomic function testing in the whole cohort, and in the 3 main subgroups by variant.Data are expressed as the mean ± SD.SBP = supine blood pressure; DBP = diastolic blood pressure; HR = heart rate; OIR = orthostatic intolerance ratio (ΔSBP/time tolerated); IE = inspiration-expiration; Δ change compared to basal values.Within the entire cohort, cardiovascular autonomic failure was detected in 76 (75%) individuals.By subgroup, 31 (74%) Thr60Ala, 23 (92%) Val30Met group, and 22 (76%) other TTR individuals had evidence of cardiovascular autonomic failure respectively.Out of the 11 carriers, 7 (64%) had subclinical evidence of stage 1 cardiovascular autonomic failure (Table4).52%) individuals did not report any symptoms of cardiovascular autonomic dysfunction.Of these, 51% had evidence of stage 1 or 2 cardiovascular autonomic failure.Conversely, 4 individuals with subjective of autonomic dysfunction had no objective evidence of cardiovascular autonomic failure on quantitative AFT.

Table legends Discussion
We provided a detailed subjective and objective assessment of the autonomic phenotype in a large cohort of DMT-naïve individuals with ATTRv amyloidosis and asymptomatic carriers of TTR variants.Despite multiple advances in this eld, especially in recent years, a knowledge gap still exists in recognising, qualifying, and quantifying cardiovascular autonomic dysfunction in ATTRv amyloidosis and in monitoring the progression of cardiovascular autonomic failure in these individuals.
stage in asymptomatic carriers.Recently, Guaraldi et al. reported an impairment of cardiovagal responses at during the Valsalva manoeuvre in 5/18 (28%) individuals with pre-symptomatic ATTRv amyloidosis.This data provides a strong argument for the performance of a thorough, standardised functional assessment of cardiovascular autonomic function as part of the initial workup and ongoing follow up of individuals with ATTRv amyloidosis and pre-symptomatic TTR variant carriers.A recent study on a far larger cohort of TTR carriers demonstrated that, in the absence of overt cardiac and neuropathic involvement, multimodal investigations including, among others, measures of autonomic function (in this case a combination of electrochemical skin conductance with the Sudoscan, heart rate variability analysis, cardiac MIBG scintigraphy and atropine test) can detect early, subclinical denervation, and might grant access to DMT.