Abstract
Objectives Sensitive biomarkers of disease and progression are needed in hereditary transthyretin amyloidosis (ATTRv). Neurofilament light chain (NfL) has emerged as a potential biomarker, however, appropriate cut-off values, alteration longitudinally and applicability in real-world settings requires evaluation.
Methods NfL levels were measured longitudinally (2015–2022), in presymptomatic and symptomatic ATTR variant carriers. Correlations between NfL and demographics, renal function, examination, and staging scores were performed. ROC analyses were performed to determine cut-off values. NfL changes longitudinally were explored, pre-treatment, in asymptomatic and symptomatic cohorts, and converters to sensory or motor neuropathy via mixed-effect models.
Results 59 individuals with mixed ATTR variants (26=T60A, 12=V30M, 21=other [including G47V/V112I/S77Y/A97S]) and PND scores (0=18, 1=19, 2=9, >3=13) were studied over maximum follow-up of 4.75 years.
No correlations were found between NfL and age, creatinine, eGFR, gender or mutation. NfL correlated with CMTNS (r=0.56, p=0.002), NIS (r=0.5, p=0.0014), and MRC scores (r=-0.57, p<0.001). NfL significantly differed between PND0 and PND2, PND3A and PND3B (all p<0.003), and PND1 and PND3B (p= 0.047). NfL differed between FAP2 and FAP0 (p= 0.001) and FAP1 (p=0.03). Patients with PND>2, were discriminated from PND0–1, by NfL>52.2pg/ml (AUC=0.83; 95%CI:0.71–0.95; sensitivity=100%, specificity=55.5%).
NfL was higher in symptomatic and motor converters, than in asymptomatic or sensory converters, irrespective of time (all p<0.001). Symptomatic or motor converters were discriminated from asymptomatic by NfL>64.5pg/ml (AUC=0.95; 95%CI:0.90–0.99; sensitivity= 91.9%, specificity=88.5%).
Conclusions This study proposes NfL cut-off values for conversion to symptomatic disease and validates the use of NfL to monitor disease activity and progression in ATTRv.
