Review Reports - Wearables for Telemonitoring in ATTR-Amyloidosis: Current Perspectives

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

Kind et al. presented a review of telemonitoring opportunities for ATTR-amyloidosis, describing the most suitable classes of wearable biosensors for monitoring disease progression and potentially enabling the early detection of behavioral alterations.

While their overview is valuable in establishing the role of personalized medicine in the context of ATTR-amyloidosis, an important limitation should be addressed. Most forms of amyloidosis, regardless of the originating protein, share similar clinical manifestations — including heart rate variability, activity patterns, and atrial fibrillation burden, among others. Therefore, the authors should clarify why the described wearable biosensors are specifically suited for ATTR-amyloidosis and not for other forms of the disease. Alternatively, the manuscript should be revised to acknowledge that these biosensors are not exclusive to ATTR-amyloidosis and may have broader applicability across the amyloidosis spectrum.

Author Response

Reviewer 1 Comment 1:
Kind et al. presented a review of telemonitoring opportunities for ATTR-amyloidosis, describing the most suitable classes of wearable biosensors for monitoring disease progression and potentially enabling the early detection of behavioral alterations.

Answer to Reviewer 1 Comment 1:

We agree with the reviewer. The clinical manifestations accessible to wearable biosensors are not unique to ATTR but are present across systemic amyloidoses, including AL and ATTRv. For HRV, AL cardiac amyloidosis produces an autonomic signature at least as pronounced as ATTR (Yamada et al. 2020). Orthostatic hypotension is abnormal in 50 – 96% of AL and 40 – 60% of ATTRv patients (Chompoopong et al. 2024), again overlapping the two entities rather than distinguishing them.

Two considerations nonetheless support retaining ATTR as the central framing. First, the AF signal is quantitatively different: prevalence reaches ~80% in ATTRwt versus 28–33% in AL-CA, with an approximately ten-fold higher adjusted risk in ATTR and a monotonic rise with disease stage (up to 94% in stage III), which is not seen in AL amyloidosis (Papathanasiou et al. 2022; Donnellan et al. 2020). Continuous rhythm surveillance is therefore disproportionately informative in ATTR. Second, ATTR-CM lacks a circulating biomarker analogous to dFLC in AL, so functional, hemodynamic, and rhythm-based endpoints carry more of the monitoring and treatment-response burden. This gap has recently become more consequential with the availability of disease-modifying therapy in ATTR.

In addition, the clinical need for long-term monitoring using wearables is greater in ATTR-CM due to the longer median survival compared to AL amyloidosis.

To reflect this, we have (i) revised the Abstract and Introduction to state explicitly that the sensor modalities discussed apply broadly across systemic amyloidosis, with ATTR serving as an illustrative use case; (ii) added a new paragraph to the 4th section on page 9 lines 411 - 431 summarizing overlapping and divergent features across AL, ATTRv, and ATTRwt; and (iii) added the comparative references (Papathanasiou 2022, Yamada 2020, Chompoopong 2024, Ioannou 2024, Donnellan 2020) to the reference list.

Reviewer 2 Report

Comments and Suggestions for Authors

This review summarizes the current application status of wearable sensors in remote monitoring of ATTR, focusing on the three sensing modes of ECG, PPG, and IMU, and reviews the research progress of digital biomarkers such as arrhythmia, autonomic nervous system function, gait, and heart failure decompensation. There are some problems should be solved before next step, for example:

Only listing device names, without elaborating on sensing principles, sampling rates, signal-to-noise ratios, motion artifacts, placement effects, and other core sensing points.
The waveform distortion and signal loss mechanism of PPG under low perfusion and peripheral neuropathy in ATTR have not been discussed at all.
Lack of key wearable technologies such as Signal Quality Index (SQI), artifact removal, and feature extraction processes.
The English expression is more colloquial, and some abbreviations are not marked for the first time; Some sentences are lengthy and have loose logic.
The references are biased towards clinical practice, and there is a significant lack of core literature in the field of sensing/wearable technology.
The match percent is high.
More pictures should be added.

Author Response

Comment 1:

Only listing device names, without elaborating on sensing principles, sampling rates, signal-to-noise ratios, motion artifacts, placement effects, and other core sensing points.

Answer Comment 1:

We thank the reviewer for this helpful comment. The aim of our review was to present the clinical arguments in favour of digital phenotyping using wearables in ATTR amyloidosis. We therefore intentionally kept the technical description at the device level concise. We agree, however, that a brief technical characterization of the referenced sensor platforms enhance the completeness of the review. We have therefore expanded Section 2 ("Categories and Technology of Wearable Devices") to include, for the introduced device categories, the sensing principle, the typical sampling rate. Table 1 has been updated accordingly. We have deliberately refrained from providing a complete technical description of each platform, as comprehensive technical reviews are already available (e.g. Bayoumy et al. 2021; Charlton et al. 2022), to which we now refer.

Comment 2

The waveform distortion and signal loss mechanism of PPG under low perfusion and peripheral neuropathy in ATTR amyloidosis have not been discussed at all.

Answer to Reviewer 2 Comment 2:

We totally agree with the reviewer and thank for pointing this out. Low peripheral perfusion and autonomic and small-fiber neuropathy seen in ATTRv both have specific consequences for PPG waveform integrity that we did not adequately address. Low perfusion flattens the pulse amplitude and the dicrotic notch, reducing the signal to noise ratio and destabilizing inter-beat interval detection; vasoconstriction and reduced sympathetic modulation further distort pulse morphology; and the high arrhythmic burden in ATTR-CM (frequent ectopy, AF) causes beat-to-beat variability that confounds template-based pulse detection.

We have added a dedicated paragraph to Section 4 discussing ATTR-specific signal distortions and their implications for both pulse rate variability estimation, with references to the PPG-specific literature on low-perfusion signal degradation and AF-related PPG instability.

Comment 3

Lack of key wearable technologies such as Signal Quality Index (SQI), artifact removal, and feature extraction processes.

Answer to Reviewer 2 Comment 3

We thank the reviewer for raising this point. As our review is positioned toward clinical applicability, a comprehensive discussion of signal quality indexing, artifact removal, and feature extraction pipelines is beyond the intended scope. However, the reviewer is correct that these methodological layers directly affect whether any downstream digital biomarker is interpretable in an ATTR cohort. We have therefore added brief descriptions of SQI principles (e.g., beat-detector agreement for ECG, perfusion index and pulse-template correlation for PPG) and artifact-handling strategies (baseline wander, motion artifacts, power-line interference) at the relevant points in Section 2. We have added references to the core methodological literature (Orphanidou et al.2015; Satija et al. 2018; Charlton et al.2022).

Comment 4

The English expression is more colloquial, and some abbreviations are not marked for the first time; Some sentences are lengthy and have loose logic.

Answer to Reviewer 2 Comment 4:

We thank the reviewer for the critical reading. We respectfully disagree with the characterization of the writing as colloquial as our manuscript follows standard academic language throughout. However, we acknowledge that the previous version contained a number of grammatical inconsistencies (missing articles, hyphenation errors, and typographic slips), which we have corrected in the tracked-changes version. We have also carefully re-read the manuscript to verify that all non-standard abbreviations are defined at first mention; we have added expansions for NYHA (New York Heart Association) and AUROC (area under the receiver operating characteristic curve), which were inconsistently introduced. The revised manuscript has additionally been reviewed by a native English-speaking colleague acknowledged in the appropriate section (Haley Hartmann).

Comment 5

The references are biased towards clinical practice, and there is a significant lack of core literature in the field of sensing/wearable technology.

Answer to Reviewer 2 Comment 5:

We fully agree with the reviewer. Our reference list reflected the clinical framing of the manuscript, but it was indeed underweighted toward foundational sensor-engineering literature. This imbalance is inappropriate for a submission to a journal like Sensors with readers from a more technological background. We have added the following core references to Sections 2 and 4:

Bayoumy K, Gaber M, Elshafeey A, et al. Smart wearable devices in cardiovascular care: where we are and how to move forward. Nat Rev Cardiol. 2021;18(8):581-599.
Charlton PH, Kyriacou PA, Mant J, Marozas V, Chowienczyk P, Alastruey J. Wearable photoplethysmography for cardiovascular monitoring. Proc IEEE. 2022;110(3):355-381.
Orphanidou C, Bonnici T, Charlton P, Clifton D, Vallance D, Tarassenko L. Signal-quality indices for the electrocardiogram and photoplethysmogram: derivation and applications to wireless monitoring. IEEE J Biomed Health Inform. 2015;19(3):832-838.
Charlton PH, Pilt K, Kyriacou PA. Establishing best practices in photoplethysmography signal acquisition and processing. Physiol Meas. 2022;43(5):050301.

Comment 6

The match percent is high

Answer to Reviewer 2 Comment 6:

We thank the reviewer for pointing this out. We were not able to access the similarity check report independently and would appreciate clarification on which sections or sources were flagged as showing high overlap. In parallel, we have re-examined the manuscript for passages that closely follow previously published wording, particularly in the device-description paragraphs of Section 2 and in the introductions to the Sangha et al. and ATTRibute-CM substudies in Section 3, and have paraphrased parts of these passages which can be seen in the tracked-changes version of the submission. A renewed similarity check is welcome.

Comment 7

More pictures should be added

Answer to Reviewer 2 Comment 7:

We thank the reviewer for this suggestion. The original version included only the graphical abstract (Figure 1) and two tables. We have added an additional Figure 2 that showcases a clinical integration pathway case for the use of wearables in ATTR cardiac amyloidosis. We believe this addition strengthens the visual structure of the review while reinforcing the disease-specific framing as well as clinical integration.

Reviewer 3 Report

Comments and Suggestions for Authors

The manuscript explores a relevant topic and shows a clear conceptual framework for wearable-based monitoring in ATTR amyloidosis. However, before publication, it requires improvement of methodological transparency, a more organized presentation of evidence, and a deeper clinical context.

Suggestions:

To improve methodological transparency, the authors should briefly describe the literature search strategy, including databases, keywords, and inclusion criteria.

A table distinguishing ATTR-specific studies from extrapolated evidence that also indicates the level of evidence would improve clarity.

The manuscript would benefit from an additional discussion of how ATTR-specific pathophysiology may affect signal interpretation and the need for dedicated validation studies.

Adding specific clinical use cases or decision pathways would improve the review's practical relevance.

A more detailed discussion of device variability, interoperability, and regulatory differences would enhance the implementation perspective.

Author Response

Comment 3

To improve methodological transparency, the authors should briefly describe the literature search strategy, including databases, keywords, and inclusion criteria.

Answer to Reviewer 3 Comment 1:

We thank the reviewer for raising this point. The present work was conceived and written as a narrative review rather than a systematic review, and we did not apply PRISMA-compliant search methodology. To improve transparency towards our literature research methodology, we have added a brief paragraph at the end of the Introduction describing how the literature was identified:

"This review is narrative in scope and does not follow a systematic review methodology. Relevant literature was identified through PubMed searches combining the terms ['ATTR amyloidosis', 'transthyretin', 'cardiac amyloidosis'] with ['wearable', 'sensor', 'telemonitoring', 'digital biomarker', 'accelerometer', 'photoplethysmography', 'ECG patch'], supplemented by hand-searching of reference lists of key articles and recent expert consensus documents. No formal inclusion/exclusion criteria or quality appraisal were applied and studies were selected based on their conceptual relevance to ATTR-specific monitoring domains and to illustrate methodological templates from adjacent cardiovascular and neurological cohorts. The clinical recommendations and biomarker framework presented here therefore reflect the authors' synthesis rather than a pre-specified evidence hierarchy."

Comment 2

A table distinguishing ATTR-specific studies from extrapolated evidence that also indicates the level of evidence would improve clarity.

Answer to Reviewer 3 Comment 2:

We agree with the reviewer on this point. The original Table 2 combined ATTR-relevant monitoring domains with evidence from adjacent disease states, which prevents the reader from quickly understanding the important distinction between direct ATTR evidence and extrapolated evidence. We have restructured the evidence presentation into two separate tables:

Table 2 now lists studies conducted specifically in ATTR cohorts indicating the level of evidence
Table 3 now summarizes extrapolated evidence from adjacent cohorts such as heart failure, Parkinson's disease, multiple sclerosis and diabetic peripheral neuropathy. It includes an explicit "ATTR transferability" column and a GRADE-inspired evidence level (high / moderate / low) for each domain.

We believe that this new arrangement might help to understand that direct ATTR evidence remains small and mostly retrospective and observational, while the biological plausibility of monitoring domains rests largely on adjacent-disease data requiring prospective validation.

Comment 3

The manuscript would benefit from an additional discussion of how ATTR-specific pathophysiology may affect signal interpretation and the need for dedicated validation studies.

Answer to Reviewer 3 Comment 3:

We thank the reviewer and agree that the disease-specific impact on signal interpretation deserves more detailed treatment. We have expanded Section 4 ("Why Extrapolation from Adjacent Diseases Is Plausible — But Requires Validation") with a new subsection, "ATTR-Specific Signal Distortions and Implications for Validation," covering the following mechanisms:

Low voltage and altered QRS morphology: amyloid infiltration reduces ECG voltage and distorts depolarization morphology, compromising automated R-peak detection and template-matching algorithms trained on non-infiltrative cohorts.
High arrhythmia burden: frequent AF, atrial ectopy, and non-sustained VT confound HRV metrics that assume sinus rhythm, and destabilize PPG-based pulse detection.
Conduction disease: prolonged AV-nodal delays and bundle branch blocks alter the beat-level features that feed into AF-detection and morphology-based classifiers.
Low perfusion and autonomic dysfunction: reduced stroke volume, peripheral pulse deficit due to arrythmia and impaired vasomotor control may influence the PPG waveform and attenuate the autonomic signatures that HRV frequency-domain metrics depend on.

We have added a resource demonstrating early work applying ECG-morphology analysis (Symmetric Projection Attractor Reconstruction, SPAR) in ATTR cohorts as a proof-of-principle that ATTR-specific signal features can carry prognostic information for worsening heart failure (Aston et al., Circulation 2025, Abstract 4370866), and we explicitly argue that prospective, rather than further extrapolation, prospective ATTR-specific feasibility and validation cohorts are the critical next step.

Comment 4

Adding specific clinical use cases or decision pathways would improve the review's practical relevance.

Answer to Reviewer 3 Comment 4:

We agree with the reviewer and have added a new subsection, "ATTR-specific Clinical Use Cases", at the end of Section 4, outlining a possible scenario in which wearable-derived data could be integrated into ATTR care pathways:

Early detection of worsening heart failure in ATTR-CM, identification and early treatment intervention to prevent heart failure hospitalization.

To visualize this pathway, we have added Figure 2 to depict a possible clinical integration pathway for wearables in ATTR cardiac amyloidosis.

Comment 5

A more detailed discussion of device variability, interoperability, and regulatory differences would enhance the implementation perspective.

Answer to Reviewer 3 Comment 5:

We thank the reviewer for this constructive suggestion. We have substantially expanded the "Implementation Barriers and Future Directions" section with the following paragraphs addressing device variability, interoperability, and regulatory differences:

Device variability. Wearable platforms differ in sampling frequency (e.g., patch-based ECG typically at 200–512 Hz versus consumer wrist PPG at 25–100 Hz), preprocessing pipelines, and proprietary feature-extraction algorithms. For a single underlying physiological metric (i.e. resting heart rate) the value reported by a Fitbit, an Apple Watch or an medical-grade ECG patch may differ in ambulatory conditions, even more in the presence of AF (Bent et al. 2020; Bayoumy et al. 2021). This limits reproducibility across studies and complicates meta-analytic synthesis. In ATTR, where low perfusion and high arrhythmia burden further degrade signal quality, device-specific validation is essential.

Interoperability. Routine integration of wearable-derived data into clinical workflows requires that sensor output be translatable into structured electronic health record entries. HL7 FHIR "Observation" resources, combined with LOINC codes (e.g., 8867-4 for heart rate) and UCUM units, have emerged as the prevailing standard for this translation. However, most consumer-grade devices do not provide API interfaces for clinical integration. Therefore, specialized disease centers might require either device-agnostic middleware or partnerships with manufacturers willing to provide clinical-grade data streams.

Regulatory differences. Regulatory classification of wearables varies markedly by jurisdiction. In the European Union, devices intended for diagnosis, monitoring or treatment fall under the Medical Device Regulation (EU 2017/745) and are classified I, IIa, IIb, or III depending on risk. Most diagnostic wearables for arrhythmia detection are IIa. In the United States, the FDA uses a Class I–III framework with 510(k), De Novo, or PMA pathways, and in 2024–2026 has issued revised guidance on Clinical Decision Support software and Software as a Medical Device.

Germany's digital health applications (Digitale Gesundheitsanwendungen,DiGA) pathway, administered by the BfArM, provides a structured reimbursement route for digital health applications classified as MDR Class I/IIa (and, since 2024, IIb), requiring CE marking, demonstrated positive care effect, and adherence to data-protection and interoperability standards. Comparable pathways have since been adopted or announced in France (PECAN) and Belgium (mHealthBelgium). No such pathway currently exists specifically for wearable-based rare-cardiomyopathy monitoring, which represents a significant barrier to translation of the monitoring concepts discussed in this review.

We have added references to Bayoumy et al. (Nat Rev Cardiol 2021) and Gehrmann et al. (Telemedicine and e-Health 2025)

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

The authors have adequately improved the manuscript. The revised version is acceptable in its present form.