Plasma Proteome Analysis of Patients with Hereditary Transthyretin-Mediated (hATTR) Amyloidosis Establishes Neurofilament Light Chain (NfL) as a Biomarker of Disease and Treatment Response ============================================================================================================================================================================================ * Simina Ticau * Gautham V Sridharan * Shira Tsour * William L Cantley * Amy Chan * Jason A Gilbert * David Erbe * Mary M Reilly * David Adams * Michael Polydefkis * Kevin Fitzgerald * Akshay Vaishnaw * Paul Nioi ## Abstract **Background** Hereditary transthyretin-mediated (hATTR) amyloidosis is a rare, progressively debilitating, and fatal disease caused by deposition of aggregated TTR amyloid in multiple organs and tissues. Highly variable disease penetrance has made it difficult to predict disease onset and progression. Clinically validated and non-invasive plasma biomarkers may facilitate earlier diagnosis, treatment initiation, and aid monitoring of disease progression and improvement of symptoms. **Methods** Plasma levels of >1000 proteins were measured from consenting patients in the Phase 3 APOLLO study who received either placebo or patisiran and in a cohort of healthy individuals. A linear mixed model determined the impact of patisiran treatment on the time profile of each protein at 0, 9, and 18 months. The most significant protein, neurofilament light chain (NfL), was further assessed using an orthogonal quantitative approach. **Findings** A significant change in the levels of 66 proteins was observed following patisiran treatment relative to placebo, with change in NfL, a marker of neuronal damage, the most significant (p<10−20). Analysis of the changes in the levels of these 66 proteins demonstrated that the proteome of patisiran-treated patients trended toward healthy individuals at 18 months. Plasma NfL levels in healthy controls were 4-fold lower than in patients with hATTR amyloidosis with polyneuropathy (16·3 [SD 12·0] pg/mL vs 69·4 [SD 42·1] pg/mL, p<10−16). Levels of NfL at 18 months increased in placebo-treated patients (99·5 [SD 60·1] pg/mL) and decreased in patisiran-treated patients (48·8 [SD 29·9] pg/mL). Moreover, at 18 months, improvement in mNIS+7 compared to baseline in patisiran-treated patients significantly correlated with a reduction in NfL levels (R=0·43, p<10−7). **Interpretation** NfL may serve as a biomarker of nerve damage and polyneuropathy in hATTR amyloidosis, as illustrated by the observed correlation with patisiran treatment and improvement in mNIS+7. Measurement of NfL levels may potentially enable earlier diagnosis of polyneuropathy in patients with hATTR amyloidosis and facilitate monitoring of disease progression/regression. **Funding** Alnylam Pharmaceuticals. ## Introduction Hereditary transthyretin-mediated amyloidosis (hATTR) is a rare, progressively debilitating, and fatal disease caused by pathogenic mutations in the transthyretin (*TTR*) gene that results in accumulation of amyloid fibrils throughout the body including in the heart, peripheral nerves and gastrointestinal tract.1 This accumulation leads to damage of organs and tissue that can include peripheral sensorimotor neuropathy, autonomic neuropathy and cardiomyopathy, leading to decreased quality of life and eventually death. The majority of patients with hATTR amyloidosis develop a mixed phenotype of both polyneuropathy and cardiomyopathy, as a result of this multisystem involvement. hATTR amyloidosis affects approximately 50,000 people worldwide and has a median survival of 4·7 years following diagnosis, with a reduced survival of 3·4 years for patients presenting with cardiomyopathy. Due to variability of the initially affected tissues, age of onset and penetrance of hATTR amyloidosis, it is difficult to predict disease onset and progression in individual patients. Thus, there is a need for improved understanding of the disease as a whole, including the ability to identify, monitor, and effectively treat patients. Diagnosis of hATTR amyloidosis remains challenging, including identification of patients with manifest disease, and ascertaining if and when asymptomatic carriers with pathogenic *TTR* mutations become symptomatic. Penetrance of the disease in carriers varies widely by region and *TTR* mutation. Additionally, since TTR fibrils aggregate and deposit in tissues over time, subclinical nerve damage is likely to occur prior to presentation of overt symptomatology. Currently, one measure used to assess severity and progression of polyneuropathy in patients with hATTR amyloidosis is the modified Neuropathy Impairment Score +7 (mNIS+7),2 a composite score which assesses sensory, motor and autonomic nervous system functions. A higher mNIS+7 score indicates worsening of neuropathy (range: 0–304) and whilst it is considered a helpful tool, it can be cumbersome to administer. Advancements are therefore necessary to improve our ability to diagnose hATTR amyloidosis earlier and assess its severity with greater objectivity. Discovery of novel plasma biomarkers associated with disease progression could potentially provide minimally invasive measures that facilitate earlier patient diagnosis and improved monitoring of disease progression and response to therapeutic intervention. Due to the low prevalence of hATTR amyloidosis, there has been a limited number of clinical proteomic studies conducted in hATTR amyloidosis patient samples. A study was conducted to investigate plasma protein profiles in patients presenting with hATTR amyloidosis relative to healthy controls, however the sample size was relatively small.3,4 The pivotal phase 3 APOLLO study was a double-blind placebo-controlled trial of patisiran, a lipid nanoparticle-delivered RNA interference (RNAi) therapeutic, that reduces serum TTR levels by inhibiting hepatic synthesis of the disease-causing mutant and wild-type (wt) TTR proteins.5 In APOLLO, patisiran-treated patients experienced improvements in the primary endpoint (mNIS+7) and all secondary endpoints with an acceptable benefit: risk profile. Plasma samples collected at multiple timepoints throughout the course of the study from both patisiran and placebo-treated cohorts enabled the most comprehensive plasma proteomics analysis in patients with hATTR amyloidosis to date. Not only did this evaluation provide additional insights into disease state and biological processes involved with progression, it also supported the first system-wide proteomics interrogation of response to an RNAi therapeutic in humans. The objective of this study was to evaluate the change in circulating protein biomarkers in response to patisiran treatment in patients with hATTR amyloidosis with polyneuropathy, with the aim of identifying potential clinical biomarkers of hATTR amyloidosis that correlate with treatment response. The clinical utility of biomarkers could conceivably include monitoring of disease progression and treatment effect; biomarker development could potentially lead to an early diagnostic or predictor of when disease may become manifest in latent carriers of pathogenic *TTR* mutations. ## Methods ### Study design and participants The present study includes analysis of plasma samples in a subset of patients enrolled in the APOLLO study. Patients who completed the APOLLO study and where plasma samples were available at baseline, 9 months and 18 months for biomarker assessment were included. This patient cohort was demographically representative of the patients enrolled in APOLLO. APOLLO was a Phase 3, randomized, placebo-controlled study of patisiran 0.3 mg/kg intravenously every 3 weeks in patients with hATTR amyloidosis with polyneuropathy.5 Clinical trials identifier: [NCT01960348](http://medrxiv.org/lookup/external-ref?link_type=CLINTRIALGOV&access_num=NCT01960348&atom=%2Fmedrxiv%2Fearly%2F2019%2F11%2F08%2F19011155.atom). Healthy control samples (n=57) were collected under IRB-approved protocols separately and were age-, sex- and race-matched to the baseline demographics of the APOLLO patients included in this analysis (Dx Biosamples, LLC, San Diego, CA). ### Plasma measurement of biomarkers Plasma samples of patients enrolled in APOLLO were collected according to the study protocol prior to dosing.5 Levels of 1196 proteins were measured (1164 unique proteins) using 13 panels containing 92-plex immunoassays (Olink Proteomics, Watertown, MA) with intra-assay coefficients of variation ranging from 2% to 14% and inter-assay coefficients of variation ranging from 5% to 26%. Olink uses a proprietary proximity extension assay technology to combine a detection step involving oligonucleotide-labeled antibodies with a proximity-dependent DNA polymerization step and a real-time quantitative PCR amplification step to measure relative levels of multiple biomarkers simultaneously. Relative protein levels are reported as normalized protein expression (NPX) values which are on a log2 scale. Quantitative measurements of NfL were made using an ultrasensitive single molecule array method,6 in duplicate, in a subset of patients where sufficient volume was available (N=112 patisiran-treated, N=47 placebo-treated), at baseline and 18 months. All NfL values were within the linear ranges of the assay, except for two measurements, which were excluded. ### Statistical analysis A linear mixed model regression analysis was performed for each protein to determine if there was a significant differential time profile based on treatment. The Wilkinson notation representation of the model is given by: ![Formula][1] Where the response Protein NPX is a measure of protein abundance, Treatment denotes whether or not the patient was administered placebo or patisiran, Time is a numeric quantity representing baseline, 9 months or 18 months on treatment, Treatment:Time is the interaction term, Age/Sex are additional covariates, and individual patients (Patient_ID) are assigned random intercepts. The model was regressed for each protein using the fitlme function (Matlab 9.3) and the coefficients and p-values associated with the Treatment:Time interaction coefficient were stored and plotted. All box plots represented in the study are based on standard metrics, namely the minimum value, maximum value, lower and upper quartiles, and median value. Outliers are denoted as those values that are more than 1·5 times the IQR (Interquartile Range) above the third quartile or below the first quartile. Plots were made using the Python Matplotlib graphics environment.7 All correlation analysis was performed using the Pearson’s correlation coefficient and all comparisons of means between the two groups were performed using the student t-test in R. Principal component analysis was performed using Python’s sklearn package,8 and the 99·9% confidence intervals were drawn using the variance-covariance matrix of each treatment group’s multivariate distribution using the Ellipse patch as part of Matplotlib. ### Role of the funding source The study sponsor was fully responsible for the study design, data analysis and decision to submit this study for publication. The study sponsor also contributed to interpretation of the data and writing of the report, in collaboration with all listed authors. The corresponding author had full access to all the data and had final responsibility for the decision to submit for publication. ## Results Plasma levels of 1164 unique proteins were analyzed using proximity extension assay in 136 patisiran-treated and 53 placebo-treated patients from the APOLLO study (Fig. 1). The abundance of these proteins was measured from each patient at three separate timepoints: baseline, 9 months, and 18 months. Because the APOLLO study enrolled only patients diagnosed with hATTR amyloidosis with polyneuropathy, also measured were the levels of the same proteins in 57 healthy controls that were age, sex, and race matched to the APOLLO patients (Table 1) to allow comparison of proteome differences between patients at baseline and healthy controls, as well as proteome differences occurring upon treatment with patisiran. View this table: [Table 1.](http://medrxiv.org/content/early/2019/11/08/19011155/T1) Table 1. Baseline demographics of APOLLO study patients participating in the biomarker study and the age/gender/race-matched healthy controls. ![Figure 1.](http://medrxiv.org/https://www.medrxiv.org/content/medrxiv/early/2019/11/08/19011155/F1.medium.gif) [Figure 1.](http://medrxiv.org/content/early/2019/11/08/19011155/F1) Figure 1. Overview of study design and samples analyzed. ### Multiple plasma proteins differ significantly between placebo- and patisiran-treated patients A linear mixed model was used to determine the impact of treatment on the time profile of each protein’s plasma level across the time-course of baseline, 9, and 18 months. A total of 66 proteins were found to show a significant change in levels in placebo vs patisiran-treated patients over time (p < 4·18×10−5; applying Bonferroni correction), of which NfL was the most significant (Figure 2; p-value = 3·95×10−21). Overall, there were proteins that increased upon patisiran treatment (e.g. N-CDase) as well as proteins that decreased upon patisiran treatment (e.g. NfL, RSPO3, CCDC80), many of which have not been described as biomarkers previously (Supp. Table 1). A significant decrease in NT-proBNP levels (p-value = 7·02×10−14), a described biomarker of cardiac health, upon patisiran treatment was also seen, consistent with measurements of NT-proBNP by an orthogonal method in the APOLLO study.5 ![Figure 2.](http://medrxiv.org/https://www.medrxiv.org/content/medrxiv/early/2019/11/08/19011155/F2.medium.gif) [Figure 2.](http://medrxiv.org/content/early/2019/11/08/19011155/F2) Figure 2. Proteins identified that have a change corresponding to patisiran treatment (relative to placebo) over 18 months. Proteins are shown here as a volcano plot, with the strength of the association on the y-axis (−log10(p-value)) and the effect size on the x-axis (shown as the treatment×time coefficient from the model). A full list of the 66 most enriched proteins can be found in Supplemental Table 1. ### Effect of disease progression and treatment on overall plasma proteome signature To elucidate whether a systemic plasma proteomic shift with hATTR amyloidosis and subsequent reversal following treatment with patisiran could be observed, principal component analysis was used to compare samples from healthy controls with baseline hATTR amyloidosis samples using the 66 most significant proteins found in the mixed model analysis (Supp. Table 1). Projecting each individual’s data onto the first two principal components that account for the most variance reveals a distinct separation between healthy controls and patients with hATTR amyloidosis with polyneuropathy at the APOLLO baseline (Fig. 3A), with most of the separation being driven by the first principal component (PC1). The proteomes of placebo-treated patients at 18 months were projected onto the same principal components and the 99·9% confidence ellipse shows a leftward movement along PC1, suggesting that the proteome is further separating from healthy controls (Fig. 3B). When the placebo- and patisiran-treated groups at 18 months were compared, it was found that plasma proteomes of patisiran-treated patients are more similar to those of healthy subjects compared to placebo-treated patients (Fig. 3C). Since there is significant heterogeneity in the proteome of patients at baseline, individual patient trajectories were also captured in 66-dimensional space. For each patient, the course of their disease progression or reversal was defined using two metrics, one for the rate (ϕ) and a second for whether the proteome moved away from or towards the healthy controls (θ; see Fig 3D legend). Plotting these two parameters clearly separates patisiran- and placebo-treated patients (Fig. 3E), emphasizing the distinct differences in the proteomes of the two groups at 18 months, reflective of changes in disease status, as evidenced by improvement in mNIS+7 score. ![Figure 3.](http://medrxiv.org/https://www.medrxiv.org/content/medrxiv/early/2019/11/08/19011155/F3.medium.gif) [Figure 3.](http://medrxiv.org/content/early/2019/11/08/19011155/F3) Figure 3. Global changes in plasma proteomes observed with the proteome of patisiran-treated patients trending toward that of healthy controls at 18 months. (A) A subset of the measured proteins was used to project the differences between patients with hATTR amyloidosis and healthy controls at baseline onto two principal components (PC1 and PC2) that most explained the difference in the data sets; analysis of (B) placebo-treated patients at 18 months and (C) patisiran-treated patients at 18 months is shown in the same PC1 and PC2 space. (D) Illustrative diagram depicting v1 and v2 as well as ϕ and θ. For each patient, a vector (v1) from the mean healthy patient to baseline and a second vector (v2) from baseline to Month 18 were used to compute two metrics, one for the ratio of the magnitude of v2 compared to v1 denoted as ϕ, and θ as the angle between the two vectors. Here ϕ measures the rate of disease progression or regression and θ measures the directionality of the proteome (defined by whether the proteome moves away or towards the healthy controls). (E) Individual patient trajectories are shown, separated by whether patients were on placebo or patisiran treatment ### Plasma NfL is increased in patients with hATTR amyloidosis with polyneuropathy and decreases with patisiran treatment Neurofilament light chain was chosen to be further investigated as it was the most significantly changed protein in the analysis when comparing placebo- and patisiran-treated patients with hATTR amyloidosis with polyneuropathy. Patients diagnosed with hATTR amyloidosis with polyneuropathy had greater than 4-fold higher levels of NfL in their plasma at baseline relative to healthy controls (Fig. 4A; log2 scale). As expected, plasma NfL levels at baseline did not differ between the patisiran and placebo groups. The patisiran-treated group had a significant decline in plasma NfL levels at 9 months that was sustained at 18 months whereas plasma NfL levels in the placebo group increased at both 9 and 18 months relative to baseline. At 18 months, placebo-treated patients had 2-fold higher NfL levels than patients treated with patisiran. Treatment with patisiran significantly lowered NfL levels in patients with hATTR amyloidosis toward levels observed in healthy controls. ![Figure 4.](http://medrxiv.org/https://www.medrxiv.org/content/medrxiv/early/2019/11/08/19011155/F4.medium.gif) [Figure 4.](http://medrxiv.org/content/early/2019/11/08/19011155/F4) Figure 4. Change over time in NfL in patients treated with placebo and patisiran. (A) Levels of NfL (normalized protein expression (NPX) values which are on a log2 scale) in healthy controls and placebo- or patisiran-treated patients at baseline, 9 months, or 18 months. Trajectories of individual patients on (B) patisiran or (C) placebo over time, color-coded by their corresponding change in mNIS+7 (from baseline to 18 months). (D) Correlation between change in NfL levels from baseline to 18 months and the corresponding change in mNIS+7 colored by treatment. (E) Levels of NfL separated by whether patients were in the predefined cardiac subset of Apollo or not, at baseline. * p-value< 0.01 ** p-value <0.001 To assess variability between individuals, plasma NfL levels were plotted for each patient over time, separated by whether they received placebo (Fig. 4B) or patisiran treatment (Fig. 4C) and colored by whether patients had a worsening of disease symptoms (change in mNIS+7>0; red) or improvement of disease symptoms (change in mNIS+7<0; blue) at 18 months. Consistent with the previous analysis, patients receiving placebo showed increasing plasma NfL levels over time, whereas patients receiving patisiran demonstrated decreasing plasma NfL levels. These changes also largely correlated with whether patients had a corresponding decrease or increase in mNIS+7 score. Plotting change in mNIS+7 for each patient against change in plasma NfL levels at 18 months shows a correlation of 0·4 (Fig. 4D), indicating that decreasing plasma NfL levels are associated with an improvement in polyneuropathy. Interestingly, no correlation was observed between plasma NfL levels and mNIS+7 score at baseline (Fig. S1). Additionally, patients belonging to the pre-specified cardiac subpopulation of the APOLLO study (defined as baseline left ventricular wall thickness ≥13 mm in the absence of a history of aortic valve disease or hypertension) saw similar elevation of plasma NfL levels (Fig. 4E) suggesting that NfL can serve as a biomarker of polyneuropathy in patients regardless of cardiac involvement. The initial NfL measurements were further validated with a widely used quantitative assay to provide absolute plasma concentrations of NfL in healthy controls and hATTR amyloidosis patient samples from APOLLO at baseline and 18 months (Fig. 5). There was a very strong correlation between the two assays used to measure plasma NfL levels (Fig. 5A; *R*=0·96). NfL levels in healthy individuals were on average 16·3 pg/mL, consistent with published reports (std. dev. = 12·0 pg/mL; Fig. 5B).9 Patients with hATTR amyloidosis with polyneuropathy had plasma NfL levels that were approximately 4-fold higher at baseline, 69·4 pg/mL (std. dev. = 42·1 pg/mL) which decreased following patisiran treatment (48·8 pg/mL ± 29·9 pg/mL) and increased in the placebo group at 18 months (99·5 pg/mL ± 60·1 pg/mL). One outlier in the patisiran-treated group, with 747 pg/mL of NfL that was excluded from the calculations (Fig. 5B, 5E). Consistent with the initial plasma NfL measurements (Fig. 4), there is a strong correlation was observed for each patient between change in mNIS+7 and change in plasma NfL levels at 18 months (correlation of 0·43; Fig. 5C), suggesting that improvements in polyneuropathy are associated with a decrease in plasma NfL levels. To determine whether plasma NfL levels could discriminate between healthy individuals and patients with hATTR amyloidosis with polyneuropathy, a receiver operator curve was used yielding an area under the curve of 0·956 (Fig. 5D) and the distribution of each was plotted (Fig. 5E). Based on the current data, a cutoff of 37 pg/mL can be considered a conservative threshold to distinguish between healthy and hATTR amyloidosis patients, at which the false positive rate is 3.6% and the true positive rate is 84.9%. ![Figure 5.](http://medrxiv.org/https://www.medrxiv.org/content/medrxiv/early/2019/11/08/19011155/F5.medium.gif) [Figure 5.](http://medrxiv.org/content/early/2019/11/08/19011155/F5) Figure 5. Quantitative measurement of NfL confirm previous findings and show potential of plasma NfL levels to distinguish between healthy patients and patients with hATTR amyloidosis with polyneuropathy. (A) Correlation between Olink and Quanterix Simoa platform is 0.96. (B) Levels of NfL in healthy controls and placebo- or patisiran-treated patients at baseline, or 18 months. (C) Correlation between change in NfL levels from baseline to 18 months and the corresponding change in mNIS+7 colored by treatment. (D) Receiver operator curve for the ability of NfL plasma levels to distinguish between healthy patients and patients with hATTR amyloidosis with polyneuropathy. The area under the curve is 0.956. (E) Histograms showing the distributions of NfL concentrations in healthy controls (green) and patients with hATTR amyloidosis at baseline (dark grey). ## Discussion Plasma proteomics and the identification of minimally invasive biomarkers are emerging as integral parts of modern drug discovery and clinical development. In an effort to leverage this approach, the plasma proteomes of patients with hATTR amyloidosis with polyneuropathy were investigated over time in the largest clinical proteomic study of this disease to date. Proteomics analyses of patient plasma samples demonstrated that patisiran treatment resulted in a general shift of patients’ proteomes toward those of healthy individuals relative to placebo; this finding suggests that the plasma protein milieu is reflective of hATTR amyloidosis disease progression and response to treatment. Across > 1000 unique proteins that were assessed from plasma samples collected in the APOLLO study, 66 proteins were found to exhibit a significantly different plasma level time profile in patisiran-compared to placebo-treated patients over the course of 18 months, of which NfL was identified as the protein most significantly different between the two groups. Collectively these findings demonstrate that patient proteome changes are detectable following treatment, as well as support the identification of proteins that have both the potential to serve as biomarkers and also offer insights into hATTR amyloidosis disease biology. The results highlighted herein indicate that patisiran-mediated reduction of TTR protein may impact multiple biological pathways. Of highest clinical interest are the significantly high plasma NfL levels detected in patients with hATTR amyloidosis with polyneuropathy, and the subsequent decrease following patisiran treatment. This is important given the emerging role of NfL as a general biomarker of central and peripheral nervous system dysfunction. NfL is an integral component of the axonal structure of neurons and has been described extensively as a biomarker of neuroaxonal injury across many central nervous system diseases including multiple sclerosis,10 Alzheimer’s,11 and Huntington’s,6 as well as peripheral nervous system diseases, such as vasculitis,12 chronic inflammatory demyelinating neuropathy (CIDP),13 Guillain-Barre syndrome,14 and Charcot-Marie-Tooth disease (CMT).15 Mutations in NfL itself can also lead to Charcot-Marie-Tooth disease, a group of inherited peripheral neuropathies,16 indicating that NfL plays a crucial role in peripheral nerve function. Average NfL levels described here in patients with hATTR amyloidosis with polyneuropathy (69·4 pg/mL) are higher than those observed for other peripheral nerve disorders (CIDP 42 pg/mL, CMT 26 pg/mL) and consistent with a previous report in patients with hATTR amyloidosis (58·1 pg/mL).17 Elevated NfL levels, thought to be caused by the release into circulation from damaged neurons,18 have been identified in both cerebrospinal fluid,19 and blood making it a proximal biomarker for nerve damage, but not specific to hATTR amyloidosis. Since treatment with patisiran lowers *TTR* mRNA levels, and therefore circulating TTR protein, the reduction in amyloidogenic aggregates likely results in a reduction of neuronal damage, thereby lowering circulating NfL. Reductions in NfL levels observed following only 9 months of patisiran treatment indicate that treatment reduces hATTR amyloidosis-associated neuronal damage in a fairly short timeframe. No further reduction in NfL levels toward those detected in healthy individuals were observed following 18 months of patisiran treatment. This may indicate that not all damage to nervous tissue can be halted or reversed, and thus, a plateau in NfL decline has been reached. Alternatively, future studies may reveal that levels of NfL continue to decline in patients treated with patisiran for >18 months. A correlation was observed between change in polyneuropathy (as measured by mNIS+7) and change in plasma NfL levels over 18 months (Fig. 4D, 5C), indicating that decreasing levels of NfL detected in patients treated with patisiran are associated with improvement in polyneuropathy in the pivotal APOLLO study. There was, however, no correlation between NfL levels and mNIS+7 at baseline. One possibility for this observation is that significant heterogeneity exists in inter-patient plasma NfL levels, while intra-patient NfL changes over time may be more informative regarding disease status. Another possibility for the lack of correlation with mNIS+7 at baseline is that patients randomized in the APOLLO study may have already entered with significant disease burden, and thus may have had saturated NfL levels, as reported in other diseases.6 NfL is a promising biomarker for response to treatment, as NfL levels are elevated in patients with hATTR amyloidosis with polyneuropathy and decreased in response to patisiran treatment. In addition to monitoring treatment response, NfL may serve as a potential diagnostic biomarker or as an indicator for the transition from an asymptomatic carrier to a symptomatic patient. Studies reported in the literature suggest NfL can serve as a prognostic biomarker in pre-symptomatic amyotrophic lateral sclerosis,20 and Alzheimer’s disease.21 The data described here suggest that a plasma NfL concentration of 37 pg/mL may represent a clinically meaningful threshold to aid the diagnosis of neuronal damage in hATTR amyloidosis with polyneuropathy, however, supplementary assessments will also be required for diagnosis. Due to the heterogeneity observed in patients with hATTR amyloidosis with polyneuropathy, the rate of change in NfL levels over time may prove to be a more sensitive indicator of neuronal damage than absolute levels. In plasma samples collected from the phase 3 APOLLO study, an average increase of 30 pg/mL was observed in the placebo group over the course of 18 months, indicating that NfL may also serve as a biomarker of disease progression. Moreover, NfL may potentially play an important role in indicating response to treatment. Notably, levels of NfL were elevated at baseline in patients with and without evidence of cardiac amyloid involvement, suggesting that NfL may be useful as a biomarker of polyneuropathy in patients previously diagnosed with hATTR amyloidosis with cardiomyopathy. An investigation of plasma NfL levels in patients with predominantly cardiomyopathic phenotype may further support hATTR amyloidosis as a single multisystem disease, regardless of *TTR* mutation or initial presenting symptoms. In conclusion, the widest range plasma proteomics analysis reported to date on samples collected from the pivotal Phase 3 APOLLO study has led to an improved understanding of polyneuropathy progression in patients with hATTR amyloidosis. Identification of plasma NfL as a protein that is not only differentially expressed in patients with hATTR amyloidosis vs. healthy individuals, but also a protein that exhibits changes in response to treatment over time, highlights the potential of this protein to serve as a biomarker in several capacities, including: as an early diagnostic of hATTR amyloidosis to enable earlier treatment, as a marker of disease prognosis, and as an objective method to monitor disease progression and/or regression over time. The results from this comprehensive, proteomic analysis using samples from a placebo-controlled study have (1) identified NfL as a potential multi-purpose biomarker of hATTR amyloidosis, (2) provided new insights into hATTR amyloidosis disease biology, and (3) enabled the first system-wide proteomics interrogation of response to an RNAi therapeutic in humans. ## Data Availability Not applicable ## Declaration of interests Simina Ticau, Gautham V Sridharan, Shira Tsour, William L Cantley, Amy Chan, Jason A Gilbert, David Erbe, Kevin Fitzgerald, Akshay Vaishnaw, and Paul Nioi are all employed by Alnylam. Michael Polydefkis has participated in clinical trials sponsored by Alnylam, Ionis and Pfizer and has received consulting fees from Alnylam, Ionis and Pfizer. David Adams has participated in clinical trials sponsored by Alnylam, Ionis and received consulting fees advisory from Alnylam and Pfizer. Mary M Reilly has participated in a clinical trial sponsored by Ionis and has received consulting fees from Alnylam, Ionis and AKCEA. ## Supplemental Data View this table: [Table S1.](http://medrxiv.org/content/early/2019/11/08/19011155/T2) Table S1. Comprehensive list of all proteins that significantly changed (Bonferroni threshold p-value < 4·18×10−5) found using a linear mixed model to compare placebo- and patisiran-treated patients over time. Beta coefficient of time:treatment term and −log10(p-value) are listed. ![Figure S1.](http://medrxiv.org/https://www.medrxiv.org/content/medrxiv/early/2019/11/08/19011155/F6.medium.gif) [Figure S1.](http://medrxiv.org/content/early/2019/11/08/19011155/F6) Figure S1. Association between NfL levels and mNIS+7 at baseline. * Received November 6, 2019. * Revision received November 6, 2019. * Accepted November 8, 2019. * © 2019, Posted by Cold Spring Harbor Laboratory The copyright holder for this pre-print is the author. All rights reserved. The material may not be redistributed, re-used or adapted without the author's permission. ## References 1. 1.Adams D, Koike H, Slama M, Coelho T. Hereditary transthyretin amyloidosis: a model of medical progress for a fatal disease. Nat Rev Neurol 2019; 15: 387–404. 2. 2.Adams D, Suhr OB, Dyck PJ, Litchy WJ, Leahy RG, Chen J, et al. Trial design and rationale for APOLLO, a Phase 3, placebo-controlled study of patisiran in patients with hereditary ATTR amyloidosis with polyneuropathy. BMC Neurol 2017; 17(1): 181. [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1186/s12883-017-0948-5&link_type=DOI) [PubMed](http://medrxiv.org/lookup/external-ref?access_num=28893208&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2019%2F11%2F08%2F19011155.atom) 3. 3.Chan GG, Koch CM, Connors LH. Blood Proteomic Profiling in Inherited (ATTRm) and Acquired (ATTRwt) Forms of Transthyretin-Associated Cardiac Amyloidosis. J Proteome Res 2017; 16(4): 1659–1668. 4. 4.Chan GG, Koch CM, Connors LH. Serum Proteomic Variability Associated with Clinical Phenotype in Familial Transthyretin Amyloidosis (ATTRm). J Proteome Res 2017; 16(11): 4104–4112. 5. 5.Adams D, Gonzalez-Duarte A, O’Riordan WD, Yang CC, Ueda M, Kristen AV, et al. Patisiran, an RNAi Therapeutic, for Hereditary Transthyretin Amyloidosis, NEJM 2018; 379: 11–21. [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1056/NEJMoa1716153&link_type=DOI) [PubMed](http://medrxiv.org/lookup/external-ref?access_num=29972753&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2019%2F11%2F08%2F19011155.atom) 6. 6.Byrne LM, Rodrigues FB, Blennow K, Durr A, Leavitt BR, Roos RAC, et al. Neurofilament light protein in blood as a potential biomarker of neurodegeneration in Huntington’s disease: a retrospective cohort analysis. Lancet Neurol 2017; 16(8): 601–609. 7. 7.Hunter JD. Matplotlib: A 2D Graphics Environment. Computing in Science & Engineering 2007; 9(3): 90–95. [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10. 1109/MCSE.2007.55.&link_type=DOI) 8. 8.Pedregosa F, Varoquaux G, Gramfort A, Michel V, Thirion B, Grisel O, et al. Machine Learning in Python. J Mach Learn Res 2011; 12: 2825–2830. [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1016/j.cpc.2010.04.018&link_type=DOI) [PubMed](http://medrxiv.org/lookup/external-ref?access_num=23755062&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2019%2F11%2F08%2F19011155.atom) 9. 9.Disanto G, Barro C, Benkert P, Naegelin Y, Schädelin S, Giardiello A, et al. Serum Neurofilament light: A biomarker of neuronal damage in multiple sclerosis. Ann Neurol 2017; 81(6): 857–870. [PubMed](http://medrxiv.org/lookup/external-ref?access_num=http://www.n&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2019%2F11%2F08%2F19011155.atom) 10. 10.Gunnarsson M, Malmeström C, Axelsson M, Sundström P, Dahle C, Vrethem M, et al. Axonal damage in relapsing multiple sclerosis is markedly reduced by natalizumab. Ann Neurol 2011; 69(1): 83–89. [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1002/ana.22247&link_type=DOI) [PubMed](http://medrxiv.org/lookup/external-ref?access_num=21280078&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2019%2F11%2F08%2F19011155.atom) 11. 11.Lewczuk P, Ermann N, Andreasson U, Schultheis C, Podhorna J, Spitzer P, et al. Plasma neurofilament light as a potential biomarker of neurodegeneration in Alzheimer’s disease. Alzheimers Res Ther 2017; 10(1): 71. 12. 12.Bischof A, Manigold T, Barro C, Heijnen I, Berger CT, Derfuss T, et al. Serum neurofilament light chain: a biomarker of neuronal injury in vasculitic neuropathy. Ann Rheum Dis 2018; 77(7): 1093–1094. [FREE Full Text](http://medrxiv.org/lookup/ijlink/YTozOntzOjQ6InBhdGgiO3M6MTQ6Ii9sb29rdXAvaWpsaW5rIjtzOjU6InF1ZXJ5IjthOjQ6e3M6ODoibGlua1R5cGUiO3M6NDoiRlVMTCI7czoxMToiam91cm5hbENvZGUiO3M6MTE6ImFubnJoZXVtZGlzIjtzOjU6InJlc2lkIjtzOjk6Ijc3LzcvMTA5MyI7czo0OiJhdG9tIjtzOjM5OiIvbWVkcnhpdi9lYXJseS8yMDE5LzExLzA4LzE5MDExMTU1LmF0b20iO31zOjg6ImZyYWdtZW50IjtzOjA6IiI7fQ==) 13. 13.Van Lieverloo GGA, Wieske L, Verhamme C, Vrancken AFJ, van Doom PA, Michalak Z, et al. Serum neurofilament light chain in chronic inflammatory demyelinating polyneuropathy. J Peripher Nerv Syst 2019; 24(2): 187–194. 14. 14.Mariotto S, Farinazzo A, Magliozzi R, Alberti D, Monaco S, Ferrari S. Neurofilament Light Chain as a Biomarker in Multiple Sclerosis. J Peripher Nerv Syst 2018; 23(3): 174–177. 15. 15.Sandelius A, Zetterberg H, Blennow K, Adiutori R, Malaspina A, Laura M, et al. Plasma neurofilament light chain concentration in the inherited peripheral neuropathies. Neurology 2018; 90(6): e518–e524. 16. 16.Jordanova A, De Jonghe P, Boerkoel CF, Takashima H, De Vriendt E, Ceuterick C, et al. Mutations in the neurofilament light chain gene (NEFL) cause early onset severe Charcot-Marie-Tooth Disease. Brain 2003; 125(Pt 3): 590–597. 17. 17.Kapoor M, Foiani M, Heslegrave A, Zetterberg H, Lunn MP, Malaspina A, et al. Plasma neurofilament light chain concentration is increased and correlates with the severity of neuropathy in hereditary transthyretin amyloidosis. J Peripher Nerv Syst 2019; doi: 10.1111/jns.12350. [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1111/jns.12350&link_type=DOI) 18. 18.Bacioglu M, Maia LF, Preische O, Schelle J, Apel A, Kaeser SA, et al. Neurofilament Light Chain in Blood and CSF as Marker of Disease Progression in Mouse Models and in Neurodegenerative Diseases. Neuron 2016; 91(1): 56–66. [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1016/j.neuron.2016.05.018&link_type=DOI) [PubMed](http://medrxiv.org/lookup/external-ref?access_num=27292537&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2019%2F11%2F08%2F19011155.atom) 19. 19.Lycke JN, Karlsson JE, Andersen O, Rosengren LE. Neurofilament protein in cerebrospinal fluid: a potential marker of activity in multiple sclerosis. J Neurol Neurosurg Psychiatry 1998; 64(3): 402–404. [Abstract/FREE Full Text](http://medrxiv.org/lookup/ijlink/YTozOntzOjQ6InBhdGgiO3M6MTQ6Ii9sb29rdXAvaWpsaW5rIjtzOjU6InF1ZXJ5IjthOjQ6e3M6ODoibGlua1R5cGUiO3M6NDoiQUJTVCI7czoxMToiam91cm5hbENvZGUiO3M6NDoiam5ucCI7czo1OiJyZXNpZCI7czo4OiI2NC8zLzQwMiI7czo0OiJhdG9tIjtzOjM5OiIvbWVkcnhpdi9lYXJseS8yMDE5LzExLzA4LzE5MDExMTU1LmF0b20iO31zOjg6ImZyYWdtZW50IjtzOjA6IiI7fQ==) 20. 20.Benatar M, Wuu J, Andersen PM, Lombardi V, Malaspina A. Neurofilament light: A candidate biomarker of presymptomatic amyotrophic lateral sclerosis and phenoconversion. Ann Neurol 2018; 84: 130–139. [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1002/ana.25276&link_type=DOI) [PubMed](http://medrxiv.org/lookup/external-ref?access_num=30014505&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2019%2F11%2F08%2F19011155.atom) 21. 21.Weston PSJ, Poole T, O’Connor A, Heslegrave A, Ryan NS, Liang Y, et al., Alzheimers Res Ther 2019; 11(1):19. [1]: /embed/graphic-1.gif